Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
VOLUME 1
A–M
EDITED BY
2004
v
vi
did an outstanding job of checking all the many cross references and the consistency of the notation. The
Academic Press Dictionary of Science and Technology has been an invaluable source of information for
clarifying various issues during the editing.
I wish to thank my wife for her patience with me during the long hours expended in the editing process.
The page proofs of this work arrived at my home on the octave day of our 50th wedding anniversary, which
we celebrated on October 17, 2003, with our five children and fifteen grandchildren.
vii
This Page Intentionally Left Blank
FOREWORD OF THE EDITORIAL BOARD
OF THE ORIGINAL RUSSIAN EDITION
This publication is an attempt at a systematized presentation of modern knowledge within all areas of solid
state physics from the most fundamental ideas to applied questions. There has been no similar encyclopedic
compilation of solid state physics available in the world until now. The Dictionary contains more than
4000 articles. Among them there are relatively long articles devoted to the basic concepts and phenomena
of solid state physics, moderate-length articles on particular problems and physical concepts, and, finally,
brief definitions of terms used in the literature. Considerable attention is paid to technical applications of
solid state physics. Brief historical information is given in many articles such as the names of authors and
the dates of discoveries. In a number of cases the history of the question is covered in the article. Together
with the description of classical topics, special attention is paid to those new ideas that arose only recently in
solid state physics and, therefore, often still do not have commonly accepted definitions. Many articles are
provided with illustrations which supplement the text material to help the reader to acquire more complete
and more visual information.
More than 500 scientists participated in the preparation and the writing of the articles of the Dictionary;
these articles were written by specialists working in the particular fields that they cover.
The goal of the publication is to provide needed reference material to a wide audience of readers, which
will help them to understand the basic principles of solid state physics. Therefore, the authors endeavored
to probe deeply into some salient details of the subject, while preserving a broad perspective throughout the
text of the article. The content of articles is interconnected by a system of cross-references. The appended
Subject Index will help the reader to orient himself easily in the scope of the various articles.
To our deep sorrow, during the period of preparation of the manuscript of this Dictionary, several out-
standing scientific members of the Editorial Board Valentin L’vovich Vinetskii, Yakov Evseevich Geguzin,
Emanuil Aizikovich Kaner, Boris Ivanovich Nikolin, Kirill Borisovich Tolpygo, Leonid Nikandrovich
Larikov and Alexander Bronislavovich Roitsyn have passed from among us. The recently expired well-
known scientist and prominent publishing specialist Aleksandr Aleksandrovich Gusev, who shared with us
his vast knowledge and encyclopedic expertise, provided invaluable aid in the publication of the dictionary.
ix
x FOREWORD OF THE EDITORIAL BOARD OF THE ORIGINAL RUSSIAN EDITION
xi
xii LIST OF CONTRIBUTORS
xv
xvi INTRODUCTION
physics are less systematically presented, the following areas are covered rather deeply and broadly: phys-
ical properties of metals, including superconductivity, electrical and optical properties of semiconductors
and insulators, magnetism and magnetic materials, mechanical properties of crystals, and physics of dis-
ordered systems. One may also notice an emphasis on the applied physics of plasticity and strength. This
is because of the fact that, although physics is an objective international science, some of its presenta-
tions or overviews found in textbooks, monographs, or encyclopedic publications unintentionally introduce
subjective ideas and personal preferences of authors.
The Dictionary includes articles reflecting the latest achievements in solid state physics, sometimes with
results that have not yet appeared in monographs. This relates both to experimental discoveries and theo-
retical ideas. There are articles on high-temperature superconductivity, the quantum Hall effect, Kosterlitz–
Thouless phase transitions in two-dimensional systems, autowave processes, fractals, etc. Novel nontra-
ditional materials and techniques for investigation are described: amorphous substances, liquid crystals,
quasi-crystals, polymers, low-dimensional systems, etc.
In addition let us point out to the reader the “quantum-mechanical emphasis”, which may be noted
in many articles devoted to diverse macroscopic phenomena and properties of solid bodies. Traditionally,
use is made of the point of view that the quantum statistics of quasi-particles undoubtedly defines such
macroscopic low-temperature properties of crystals as specific heat and thermal conductivity, electrical
conductivity and magnetism. However, some authors only assign to quantum mechanics the role of de-
scribing microscopic laws explaining the dynamics of atomic particles. The discovery of high-temperature
superconductivity transferred the quantum manifestations of particle dynamics into the region of moder-
ately high temperatures. State-of-the-art experimental techniques make it possible to observe a number of
phenomena involving macroscopic quantum tunneling. Thus, macroscopic solids fairly often exhibit some
particular quantum features.
Finally let us stress the role of models in solid state physics. As is well known, a model is a much
simplified speculative scheme which allows a researcher, on the one hand, to propose a simple qualitative
explanation of a property or phenomenon under study, and, on the other hand, to give a quantitative descrip-
tion of properties and to perform a complete theoretical calculation of characteristics of the phenomenon
being studied. Models are utilized very often in solid state physics. Apparently, the most ancient model
is that of an “absolutely solid body”. The model of perfect gas of quasi-particles is frequently employed;
Ising’s model is well known in the study of magnetism. The model of a two-dimensional lattice of clas-
sical spins in the theory of phase transitions proposed by Onsager allowed him to give the first statistical
description of second-order phase transitions. The Frenkel–Kontorova and Peierls models are widely used
in dislocation theory. References to them can be found in many of the articles included in this work.
The need for this publication arises from the recent tremendous progress in this field of physics, espe-
cially by the extension of its areas of practical applications. There have appeared new fields of electron-
ics (optical electronics, acoustoelectronics); semiconductor electronics has changed and now employs an
increasingly wider range of complex structures such as metal–insulator–semiconductor devices and com-
puters with superconducting elements (cryotrons); and memory elements based on new physical principles
have been developed.
There are many promising new solid state materials based on garnets, rare-earth metals, magnetic semi-
conductors, amorphous metals, semiconductors, and, finally, the discovery of high-temperature supercon-
ductors in 1986.
The compilers of this Dictionary hope that this publication will be useful for researchers working in
solid state and other fields of physics (nuclear physics, plasma physics, etc.) and also for a wide audience of
other readers, including, but not limited to, engineers investigating and utilizing solids, as well as doctoral
and advanced students at universities.
V.G. Bar’yakhtar
A.M. Kosevich
HOW TO USE
THE ENCYCLOPEDIC DICTIONARY
In this Dictionary, we follow standard fundamental rules generally adhered to in encyclopedic publications.
1. Articles are arranged alphabetically; if a term (typeset in boldface) has a synonym, then the latter is
given in the entry after the main meaning of the term (after comma, in usual print).
2. Some terms in the titles of articles adopted from other languages are followed by a brief etymolog-
ical note (with clear abbreviations: abbr., abbreviated, fr., from; Gr., Greek; Lat., Latin; Fr., French;
Germ., German, etc.).
3. The names of scientists appearing in the title or text of an article are commonly followed by the date of
obtaining the result (effect, phenomenon) under consideration.
4. We use a system of cross-references (in Helvetica italics) to other articles. Additional terms explained
within the same article are typeset in Times italic and, together with the main entries, are listed in the
Subject Index at the end of Volume 2. Many terms synonymous to those in the article titles may be also
found in this Subject Index.
5. Units of physical quantities and their symbols are given in the systems of units most commonly used
in solid state physics. For the relations which convert these units into SI, see the tables included in the
article Units of physical quantities.
6. The figures are explained either in their captions or in the text of the article.
7. Usually, the notation used in formulae is explained in the text of that article. However, some sym-
bols have the same meaning throughout the Dictionary (unless otherwise specified): c, speed of light;
h and = h/2π , Planck constant; T , absolute temperature; kB , Boltzmann constant; λ, wavelength;
ν and ω = 2πν, frequency. The notation of some elementary particles is also standardized: ν, photon
and gamma quantum; e, e− , electron; p, proton; n, neutron; d, deuteron; π and π ◦ , pions; μ± , muons;
K ± and K ◦ , K-mesons.
xvii
This Page Intentionally Left Blank
Aa
ABLATION (fr. Lat. ablatio, removal, taking away) ABRIKOSOV–SUHL RESONANCE
The removal of mass from the surface of a solid (A.A. Abrikosov, H. Suhl, 1965)
brought about by a heated gas flow sweeping past The abnormal increase of the electron spin- ip
it. scattering amplitude of conduction electrons on a
paramagnetic impurity when the energy ε of the
This is observed during the motion of mater-
scattered electron approaches the Fermi energy εF
ial bodies and space crafts through the atmosphere of the metal.
at high velocities. Ablation occurs as a result of The cause of the Abrikosov–Suhl resonance is
erosion, melting, and sublimation. It is used to the creation of a long-lived bound state of a con-
cool key parts of super- and hypersonic aircrafts, duction electron with a paramagnetic impurity due
spacecrafts, and rocket engines by coating them to the indirect interaction of conduction electrons
with layers of an ablatable material (e.g., ceram- with each other via the impurity spin when the
ics, or pyrolytic graphite). The ablation material spin–paramagnetic impurity exchange interaction
must have high melting and sublimation tempera- constant is negative, i.e. J < 0. The Abrikosov–
tures and a large speci c heat. Suhl resonance manifests itself below the Kondo
temperature T K by an anomalous temperature de-
pendence of the behaviour of the resistivity and the
ABRASIVE MATERIALS, abrasives (fr. Lat. magnetic susceptibility of a metal (see Kondo ef-
abrasio, scraping) fect). The expression for the resistivity of a metal
Fine-grained solid materials intended for work- is derived by the selective summation of the “dan-
ing (polishing, finishing) a material surface. gerous” terms of the perturbation expansion in
the parameter |J |/εF 1 (diagrams calculated in
The working consists in cutting and scrap-
the so-called parquet approximation), and has the
ing the surface by a great number of randomly
form
oriented sharp edges of abrasive grains. Typical −2
abrasive materials are diamond (natural or syn- (0) J ν(εF ) εF
ρ = ρs 1 + ln ,
thetic), carborundum, the cubic modification of n kB T
boron nitride, and boron carbide. Abrasive mate- (0)
where ρs is the resistivity calculated in the first
rials are characterized by hardness, strength, gran-
Born approximation, ν(εF ) is the conduction den-
ularity (describing the magnitude and dispersion
sity of electron states at the Fermi energy, and n is
of the grain size); knowing these parameters per-
the metal density. The denominator vanishes at the
mits the selection of abrasive materials for par- Kondo temperature
ticular purposes, and provides the expected pre-
cision of the working. Depending on their relative εF n
TK = exp − .
abrasive abilities, abrasive materials can be clas- kB |J |ν(εF )
sified in terms of the following scale: diamond In the temperature range T TK , perturbation
(1.00), cubic boron nitride (0.78), boron carbide theory is no longer valid, and a more detailed
(0.60), silicon carbide (0.50), and titanium carbide analysis shows that a spin complex of a conduc-
(0.42). tion electron with an impurity is created, where at
1
2 ABRIKOSOV VORTICES
tor (called a hard superconductor) the Abrikosov (4) the extraction of hydrocarbon gases (natural
vortices should be held in position by pinning sites and synthetic) from their mixtures (e.g., propanes,
or defects of the crystal lattice (see Kosterlitz– butanes and pentanes).
Thouless transition, Vortex pinning).
ABSORPTION COEFFICIENT, absorption factor,
ABSOLUTE DIAMAGNETISM, ideal absorptance, absorptivity
diamagnetism The ratio of energy (particle) flux, absorbed by
The state of a superconductor in which a mag- a given sample, to the incident flux. The term “ab-
netic field is completely excluded from its interior sorption factor” is sometimes used in reference to
at a temperature below its critical temperature Tc . an absorption index that is defined as the natural
This occurs when the applied magnetic field B logarithm of the reciprocal of the absorption fac-
is lower than the thermodynamic critical magnetic tor.
e ld Bc (T ) for a type I superconductor, and lower
than the lower critical eld Bc1 (T ) of a type II su- ABSORPTION EDGE
perconductor (see Meissner effect). This implies See Intrinsic light absorption edge.
that in the superconducting state the magnetic sus-
ceptibility χ attains its highest negative value χ =
ABSORPTION INDEX
−1 in SI units (χ = −1/4π in Gaussian units). A quantity α, which is the reciprocal of a dis-
Absolute diamagnetism is caused by the persistent tance, by which radiation flux in a parallel beam
supercurrents which circulate in the outer surface
is attenuated by the factor of 1/e as a result of
layer of the material and shield the interior from
its absorption by the material (cf. Absorption co-
the penetration of magnetic flux (see Penetration
ef cient). It is introduced to characterize radiation
depth of magnetic eld). This diamagnetism re-
intensity I = I0 e−αx which decays exponentially
sults in repulsion between a superconductor and
with the distance (see Beer–Lambert law). The
a magnet, and the levitation of a magnet above a
value of the absorption index in crystals may de-
superconductor.
pend on the direction of propagation, and the type
Absolute diamagnetism is utilized in the man-
of polarization of the radiation (see Dichroism of
ufacture of frictionless bearings, and the operation
crystals, Pleochroism).
of trains levitated above their tracks.
of majority carriers near a surface is called accu- plane of polarization ϕ, as in the Faraday effect for
mulation, while that of minority carriers is called linearly polarized light, depends on the distance l
inversion. traversed by the acoustic wave in the elastically
anisotropic medium: ϕ = (1/2)(k+ − k )l, and
−
ACOUSTIC ANALOG OF COTTON–MOUTON the ellipticity ε is linearly connected with the mag-
EFFECT netic circular dichroism (or differential adsorption
Consists in the acoustic wave ellipticity and the of the circularly-polarized waves), having the as-
rotation of acoustic wave polarization plane (orig- ymptotic behaviour with increasing l
inally linearly polarized) as it propagates normal l) − exp(−k l)
exp(−k+ −
to a static applied magnetic field (magnetic induc- ε= l) + exp(−k l) ,
tion) B. The polarization plane of the initial wave exp(−k+ −
should not be parallel to or normal to B. This ef- where k± and k are the real and imaginary com-
±
fect is caused by linear magnetic double refrac- ponents, respectively, of the complex wave vectors
tion (due to different phase velocities of transverse k = k + ik of the circularly polarized acoustic
waves polarized parallel and normal to B) and waves. In case of strong coupling (e.g., under con-
linear magnetic dichroism (differential absorption ditions of helicon–phonon resonance) the elliptic-
of transverse waves polarized parallel and nor- ity and rotation of the polarization plane results
mal to B, denoted as α and α⊥ , respectively). from interference not only of two, but of three cir-
For α = α⊥ the dependence of the ellipticity ε cularly polarized waves, with ϕ and ε having os-
on the distance l covered by an acoustic wave in cillation dependences on l and B. The acoustic
an elastically-anisotropic medium (see Anisotropic analog of the Faraday effect was observed in met-
medium) is represented by a saw-toothed curve, als under conditions of Doppler-shifted cyclotron
and the dependence of the rotation angle of the resonance (Al, Cu, W), of doppleron–phonon res-
polarization plane Φ on l jumps by π/2 at points onance (In, W), of helicon–phonon resonance
where ε = ±1. A difference between α and α⊥ (K, In), and in magnetic substances: yttrium iron
smooths out the jogs on the curves ε(l) and Φ(l). garnet and some other iron garnets.
This effect provides information about magnetoe-
lastic constants. It was first observed by B. Luthi ACOUSTIC BREAKDOWN
(1963) in iron-yttrium ferrite, and later it was Change of electron path topology in metals in
seen in rare-earth garnets, magnetite, nickel, and a magnetic field when acted on by an intense ul-
RbNiF3 crystals. trasonic wave.
During an intraband acoustic breakdown the
ACOUSTIC ANALOG OF FARADAY EFFECT periodic strain acting on the sonic wave splits the
Rotation of acoustic wave polarization plane energy band of a metal into several subbands, each
and acoustic wave ellipticity during its propagation of which is related to a specific electron path in
parallel to an applied static magnetic eld B along the external magnetic field. An interband acoustic
a symmetry axis of rotation not lower than third breakdown appears when the quasi-momentum of
order (see Crystal symmetry). This acoustic ana- the sonic wave is close to the minimal distance be-
log of the Faraday effect arises from the interac- tween the electron paths in momentum space with
tion of the elastic subsystem with either the spin no sound present. The interband acoustic break-
subsystem, or with the of conduction electron sub- down always manifests itself in combination with
system. This interaction may have either a nonres- magnetic breakdown: in the presence of sound, the
onant, or a resonant character. In the case of a non- transitions linked with the magnetic breakdown
resonance interaction, and also of a weakly cou- take place in much lower magnetic fields, and can
pled resonant interaction (characterized by small lead to changes in the electron path topology. The
variations of dispersion curves of normal modes acoustic breakdown causes the appearance of new
near resonance) the rotation of the polarization periods of Shubnikov–de Haas oscillations (see
plane is caused by magnetic circular birefringence Shubnikov–de Haas effect), as well as to changes
(different phase velocities of oppositely circularly- of the more gradually varying part of the electrical
polarized waves). By this, the rotation angle of the conductivity tensor in strong magnetic fields.
6 ACOUSTIC CONTACT
Experimentally, acoustic emission signals that segments fixed at ends. Also described are the
accompany the individual acts of inelastic defor- acoustic signals generated by the motion of sim-
mation are also observed: during the formation of ple dislocation accumulations, sides of cracks, and
twins, martensite inclusions (see Martensitic poly- nucleation in a rst-order phase transition. An im-
types), reemission of the sound by an individual portant trend in the theory of acoustic emission is
slip band, appearance of microcracks. The the- the analysis of the transient procedures in record-
oretical description of acoustic emission in the ing and processing of acoustic emission signals,
framework of continuum mechanics is based on as well as the statistical and correlation analy-
the assumption that the structural changes tak- sis of pulse fluxes. The detection and analysis of
ing place in the sources (spontaneous transforma- acoustic emission yield important information on
tion strain) are matched to the ambient material the structure and its change in time for solids ex-
without upsetting the macroscopic continuity, but posed to various factors. That is why the phe-
cause the emergence of time-varying forces local- nomenon of acoustic emission is widely used as
ized at the source (internal stresses), which excite the basis for various nondestructive testing tech-
the acoustic signals by their action. In some cases, niques, in particular for control of corrosion crack-
use is made of formal modeling of the source by ing, and technological processes such as diffusion
a system of point sources (dipoles, quadrupoles, saturation, welding, hardening, cutting. The use of
etc.) varying in time according to a given law. acoustic emission control requires comparison of
Then, the problem is reduced to finding wave so-
acoustic emission parameters with those of emerg-
lutions of the dynamic equations of elasticity the-
ing and developing defects, especially at the stage
ory, which are satisfied by these forces. The dy-
when their further development becomes critical
namics of reconstructing the structure at the semi-
for the object under investigation. The acoustic
microscopic level is described in terms of elemen-
emission method is also used in physical research
tary “carriers”. In the cases of slipping, twinning,
for studying plastic deformation processes, failure,
and martensitic transformations, the part played
and phase transformation kinetics.
by such carriers is related to corresponding per-
fect dislocations, twin dislocations, and transfor-
mation dislocations (see Dislocations). The phase ACOUSTIC GENERATOR
interfaces and the sides of the cracks can also be See Acoustic wave transducer.
treated as the elementary “carriers”. The physi-
cal description of an acoustic emission source is
ACOUSTIC GYROMAGNETIC PHENOMENA
reduced to establishing the relationships between
in magnetic materials
spontaneous structure transformation, the strain
Rotation of acoustic wave polarization plane
and dynamic characteristics of the carriers, and the
and the appearance of ellipticity of the initially
formulation of the dynamic equation of elasticity
theory involving the distribution of the forces de- plane-polarized transverse elastic wave propagat-
termined by the “carrier” fluxes. The end purpose ing along the magnetization direction.
of the physical theory of acoustic emission is to as- The phenomena take place in a medium with
certain the connection between the wave solutions non-zero magnetization M (spontaneous or in-
of the dynamic equation and the spatial position, duced by an external magnetic field). They are
speed and accelerations of “carriers”. In the frame- conditioned by the emergence of an antisymmet-
work of the approach, a great number of elemen- ric component of the dynamic elastic modulus (if
tary mechanisms of acoustic emission connected M is parallel to the z axis, then the components
with the dislocation motion is described: annihi- cyzxz = −cxzyz ). The first effect is related to the
lation of dislocations of opposite signs, escape of real part of the antisymmetric components, result-
dislocations to the crystal free surface (transition ing in a difference between the phase velocities
radiation by dislocations), the work of the dislo- of right and left circularly polarized waves. The
cation Frank–Read source, the oscillatory motion second effect arises from the imaginary part of
of straight-line dislocations, and the dislocation these components, which causes the damping of
ACOUSTIC METHODS 9
the right- and left-polarized waves to be differ- The latter equals the product of the area S where
ent (magnetic circular dichroism). Both effects in- acoustic impedance is determined times the nor-
crease sharply (in a resonant manner) with the fre- mal component of the velocity of the oscillating
quency approaching magnetoacoustic resonance, medium particles, averaged over S. The real part
as has been proven experimentally. of the acoustic impedance is associated with the
losses in the system itself plus the radiation loss.
ACOUSTIC HOLOGRAPHY The imaginary part is due to the reaction of the
A direct analog of optical holography, i.e. an forces of elasticity and inertia. One can also de-
interference mode of recording three-dimensional fine the specific acoustic impedance za and the me-
images using reference and object waves. The chanical impedance Zm , and these are related to
recording conditions are selected so that the image the acoustic impedance by Zm = Sza = S 2 Za . At
can be reproduced using the diffraction of light in great distances from the source, Za coincides with
the visible range. A phase interface between a liq- the wave resistance of the medium, equal to ρv,
uid and a gas can serve as a recording medium. where ρ is the medium density and v is the sound
Under the action of the acoustic wave, the inter- velocity. The ratio of the amplitudes of transmit-
face surface undergoes a deformation, and the re- ted and reflected plane waves from the interface of
lief which appears represents a phase hologram (1) (2)
two media, za /za = ρ (1) v (1) /(ρ (2) v (2) ), also
for the optical radiation. The use of cholesteric liq- determines the matching conditions for acoustic
uid crystals for recording the acoustic holograms waveguides. The efficiency of sound transducers,
is also available. At the locations of the antinodes the power of emitters, etc., depend on the acoustic
of the acoustic standing wave the arrangement of impedance match.
liquid crystal molecules becomes disordered, con-
serving its initial ordered condition at the nodes.
ACOUSTIC METHODS of studying solids
As a result, the extinction coefficient becomes
space-modulated, and an amplitude hologram is Determining properties of solids by measur-
detected. ing their acoustic characteristics. The acoustic
The possibility of directly measuring the methods are especially important for the study of
acoustic wave phase with the help of linear de- opaque objects.
tectors (in contrast to optical detectors capable These methods cover a wide range of frequen-
of recording intensity waves only) provides some cies and are used, in particular, as nondestructive
unusual recording methods in acoustic hologra- testing techniques for characteristics of solids. De-
phy. For example, during the recording of an fectoscopy can be complemented by investigating
acoustic hologram, the reference wave can be in- objects using an acoustical microscope, and this
troduced in the form of an electrical signal that is is especially effective for non-transparent media,
summed with the acoustic signal transformed into or searching for inhomogeneities in transparent
an electric one. The instantaneous distribution of ones (see Acoustic microscopy). In addition to de-
the acoustic waves in space is registered with the termining the macroscopic parameters of a me-
help of a rapidly moving range of detectors which dium, acoustic methods are successfully applied
record the complete information about the struc- in investigations of its microscopic properties.
ture of the wave front scattered by an object. The Acoustic paramagnetic resonance and acoustic
reference wave is not needed for this recording nuclear magnetic resonance provide information
method. on the spin–phonon interaction (see Spin–phonon
Hamiltonian), and the character of crystal e lds
ACOUSTIC IMPEDANCE (their symmetry, the presence of defects). Met-
A complex-valued quantity Za characteriz- als can also be studied. The investigation of the
ing the resistance of a medium to the excitation acousto-electronic interaction provides character-
or propagation of acoustic waves. Acoustic im- istics of the electron–phonon interaction, and the
pedance is equal to the ratio of the complex am- electrical conductivity of a medium. The observa-
plitudes of sound pressure and the volume speed. tion of nonlinear acoustic effects enables the study
10 ACOUSTIC MICROSCOPY
the phonon–phonon interaction and related phe- moduli and their anisotropy (see Elastic modulus
nomena: thermal expansion and thermal conduc- tensor, Surface acoustic waves). The resolution
tivity of materials. depends on the ultrasound frequency, and is typ-
Ultrasonic spectroscopy is an important meth- ically comparable to optical microscopy resolu-
od of acoustic investigation. It can provide the tion. The highest resolution achieved in cryogenic-
spectral dependences of acoustic parameters, such scanning-raster microscopy is 25 nm. Acoustic
as velocity, absorption, scattering of ultrasound. microscopy is useful for studying the structure and
This method yields information on such physical defects of crystals and polycrystals, mono- and
characteristics of solids as relaxation time, electron polycrystal lm s, many-layer systems, coatings
effective mass, Fermi surface shape, as well as on and their adhesion, the structure of composite ma-
the presence of structural inhomogeneities, in par- terials, alloys and minerals, biological tissues and
ticular, magnetic domain structure. cells; nondestructive testing of microelectronic
elements, etc.
ACOUSTIC MICROSCOPY
A technique which provides magnified acoustic ACOUSTIC NUCLEAR MAGNETIC
images of small objects and measures the distrib- RESONANCE (ANMR)
ution of local physico-mechanical properties. Selective absorption of ultrasonic oscillation
In acoustical microscopy, ultrasound and hy- energy by nuclear spins in solids, occurring when
persound are used as probing radiation. As dis- the ultrasound frequency coincides with the en-
tinct from optical microscopy where the image ergy level spacing of the nuclear spins in the ex-
is formed by refractive index variations, the con- ternal magnetic or intracrystalline electric field.
trast of an acoustic image results from changes The crystal lattice vibrations at the ultrasonic
in the acoustic impedance and in the sound ab- frequency lead to the appearance of intracrys-
sorption coefficient. Acoustic images have high talline oscillating electric or magnetic fields which
contrast for both optically transparent and opaque induce electric quadrupole or magnetic dipole
objects. There are several schemes for acoustical transitions between nuclear spin levels. Therefore,
microscopy implementation, including scanning- ANMR can be thought of as an acoustic ana-
raster, scanning-laser, photo- and electronic logue of nuclear magnetic resonance (NMR). The
acoustical microscopy. The latter two are based ANMR is detected by the absorption of ultrasound
on the detection of an acoustic signal excited in at resonance, or by changes in the NMR signal via
the object by a focused optical or electron beam acoustical saturation of the spin levels. Further de-
scanning its surface. In scanning-laser acoustic velopments include acoustic nuclear–nuclear and
microscopy the object under study is placed in electron–nuclear double resonances (see Double
a liquid and exposed to an acoustic plane wave. resonances), nuclear polarization (see Nuclear ori-
After the interaction of the wave front with the entation), and spin echo.
object at the liquid boundary, the surface relief ANMR has been studied in insulators, semi-
that is formed is read by a laser beam and forms conductors, magnetic materials, and metals. These
an image on the display after signal processing. measurements provide values of the spin–phonon
The scanning-raster approach is based on expos- interaction tensors (see Spin–phonon Hamiltonian)
ing the object under study to a focused electron and make it possible to clarify the crystal e ld
beam created by an acoustical lens; with the ob- theory and spin system dynamics descriptions.
ject located near the focus of the acoustic mi- A method of crystal defectoscopy based on ANMR
croscope. It is possible to obtain the acoustic has been developed. For metals and high-conduc-
image by detecting the transmitted, reflected or tivity semiconductors ANMR is the only method
scattered radiation, and scanning the focal re- for investigating a spin system in regions not lim-
gion near the sample. In addition, this scanning- ited by the skin depth (see Skin-effect). These
raster method permits the measurement of the lo- magnetoacoustic methods are useful for the study
cal speed of the Rayleigh waves on the object’s of ferro- and antiferromagnets. See also Acoustic
surface, and, hence, the distribution of the elastic paramagnetic resonance.
ACOUSTIC PLASMON 11
The branches of acoustic plasmons exist in practically the same, and only the acoustic plas-
multicomponent or low-dimensional Fermi sys- mon damping in the energy range ω < 2Δi van-
tems with Coulomb interactions between parti- ishes, where Δi is the gap in the quasi-particle
cles (e.g., in a degenerate electron–hole plasma, spectrum of the ith superconducting component.
or in a space-quantized electron gas in an in- The acoustic plasmons can be viewed as a
version layer of a semiconductor) having a large Goldstone mode (see Goldstone theorem) in sys-
difference between the effective masses of the tems with a long-range Coulomb interaction,
“light” and “heavy” current carriers, or a strong which arises due to the spontaneous symmetry
anisotropy of the electron spectrum. For a two- breaking of the initial Hamiltonian as a conse-
component isotropic system subject to the condi- quence of the difference in the effective masses
tions m∗1 m∗2 and vF1
vF2 , where m∗i and vFi of the particles. That is, when the “light” fermi-
are the effective mass and Fermi velocity of the ith ons have the time to shield the “adiabatically
component, respectively, regardless of the charge slow” (low-frequency) density oscillations of the
carrier sign, the acoustic plasmon spectrum in the “heavy” fermions, and they effectively block the
momentum range q kF1 has the following form long-term interaction which prevents the acoustic
at sufficiently high densities: branch (hydrodynamic sound) from appearing.
1/2 The “heavy” component can also be acted upon
2 + 3 q 2 v2
ωp2 by the charged bosons, e.g., bipolarons of large
ωq ∼
5 F2
=q 2 , (1)
enough mass.
κ1 + q 2 [1 − κ12 /(8kF2 )]
In a two-dimensional electron gas, the plas-
where kF1 is Fermi wave number of the light mon spectrum is quasi-acoustic with a square-root
√
fermions, ωp2 is the plasma frequency of the behaviour (ωq ∼ q) and weak damping in the
heavy fermions, and κ1−1 is the Thomas–Fermi case of one component, and it is linear (ωq ∼ q
screening radius of the “light” component (set- at q → 0) in the case of two components. In
ting = 1). At q ωp2 /vF2 , the acoustic plas- layered quasi-two-dimensional crystals and chain
mon branch falls into the region of strong quantum quasi-one-dimensional crystals with the Coulomb
Landau damping due to the decay into “heavy” interaction between conducting layers and chains,
electron–hole pairs, while at q < ωp2 /vF2 it lies in the plasmon spectrum is three-dimensional even
the region of relatively weak damping due to the when disregarding the electron tunneling (see Tun-
decay into “light” Fermi excitations. The phase neling phenomena in solids) between neighbor-
velocity and group velocity of acoustic plasmons at ing layers and chains, but it is anisotropic, ωq ∼
q → 0 are (q /q)ωp , where q = (q2 +q⊥2 )1/2 , q and q be-
⊥
∗ 1/2 ing the longitudinal and transverse components of
ωp2 m1 n2 the momentum, respectively, so that ωq → 0, but
u= = vF1
κ1 3m∗2 n1 q⊥ = 0. The virtual low-frequency acoustic plas-
∗ 1/2 1/6 mons can play an important role in weakening the
m2 n2 Coulomb repulsion between electrons and thereby
= vF2 , (2)
3m1 n1 favor superconductivity in anisotropic (multiband)
and low-dimensional metals (see Plasmon mecha-
where ni is the concentration of the ith com-
nism of superconductivity).
ponent, so the weak damping condition of the
long-wave acoustic plasmons has the form
ACOUSTIC RADIATION BY DISLOCATIONS
vF2 u vF1 , i.e.
Generation of acoustic waves in a crystal,
1/3
m∗2 n2 n1 which accompanies dislocation motion and vari-
max , 3 . (3)
m∗1 3n1 n2 ous dislocation reactions. This is one of the most
important mechanisms of acoustic emission by a
After the transition of one or both components into under strain crystal.
the superconducting state (see Superconductivity), Acoustic radiation by dislocations in a homo-
the acoustic plasmons dispersion relation remains geneous medium arises in nonstationary processes,
ACOUSTIC SPECTROMETER 13
as well as at the initiation and annihilation of dis- hypersound pulse splits into 2π -pulses. Acoustic
locations. In inhomogeneous media, elastic waves self-induced transparency has been observed ex-
are also excited under stationary motion of dislo- perimentally in dielectric and ferroelectric crys-
cations at points where changes occur in elastic tals, alloyed by paramagnetic ions of the iron
moduli (transition radiation by dislocations). A par- group, satisfying the conditions of acoustic para-
ticular case of transition radiation is observed at magnetic resonance at liquid helium tempera-
the escape of a dislocation to the free surface of tures.
a body. A dislocation segment (see Dislocation
string) fixed at its ends can vibrate under the action ACOUSTIC SOLID-STATE RESONATOR
of external oscillating stresses in the manner of a A device involving acoustic vibrations in a
tight string. While moving, a dislocation also ex- mode close to that of a standing wave, viz. when
periences the action of the periodic Peierls poten- the distance between the end plates of an acoustic
tial relief (see Peierls–Nabarro model) and alter- resonator is an integral multiple of half-wave
nating fields of internal stresses; it can be pinned lengths.
by impurity atoms, and break away from them. All Typically, an acoustic resonator is made from
of these nonstationary processes make contribu- a piezoelectric material (see Piezoelectricity), and
tions to the radiation of elastic waves. the oscillations are excited by an electric field.
Piezoelectric acoustic resonators are electrical two-
ACOUSTIC RESONANCE, X-RAY pole and multipole elements with the impedance
See X-ray acoustic resonance. exhibiting resonant behavior near resonant fre-
quencies. These resonators exhibit a high stabil-
ACOUSTICS ity of the natural oscillation frequency, a long-
See Geometrical acoustics, Nonlinear acoustic term one of ∼10−6 –10−8 Hz, and a short-term
effects, Quantum acoustics. one of 10−8 –10−10 Hz and higher. These are used
to stabilize oscillator frequencies (of resonators),
ACOUSTIC SELF-INDUCED TRANSPARENCY and form amplitude responses of filters (filter res-
Hypersound pulse propagation without reso- onators).
nant absorption of energy by impurity paramag-
nets and other resonant species, brought about by ACOUSTIC SPECTROMETER
the coherent self-action of the hypersound via the An experimental installation for measuring
resonant electron system of impurity centers. sound absorption and sound velocity dispersion.
The impact of a hypersound pulse on these An acoustic spectrometer comprises the fol-
resonant centers causes the inversion of the pop- lowing basic components: a source of electro-
ulations of their energy levels E1 and E2 , with magnetic ultra-high frequency (UHF) oscillations,
the subsequent stimulated emission of coherent a component to pass on these oscillations to an
phonons into the region behind the pulse. Because acoustic wave transducer with a connection to the
of this, the pulse energy remains constant, but its sample under study, a UHF section from the sam-
velocity and envelope shape change noticeably. ple to the receiving end of the spectrometer, and
Acoustic self-induced transparency may be real- a recording device. Acoustic spectrometers typi-
ized only for pulses which are shorter than the re- cally employ pulse and pulse-phase measurement
laxation time of the impurity centers. The region techniques, those with continuous oscillations be-
of the input pulse ing less common. As a rule, the spectroscopic
+∞
measurements determine the absorption and sound
Θ = −1 F12 (t) dt velocity dispersion as functions of external para-
meters such as magnetic field, temperature, pres-
−∞ sure, etc., at a fixed UHF frequency. In pulse
(F12 is an electron–phonon interaction Hamil- acoustic spectrometers, these quantities are the
tonian matrix element) must exceed π . When Θ = pulse amplitude and its traveling time through the
2π the pulse is stationary. If Θ > 4π , the input sample under study, while in pulse-phase acoustic
14 ACOUSTIC SPECTROSCOPY
spectrometers the pulse that passes through the and velocity dispersion, geometrical resonance,
sample is compared to a signal coherent with it. acoustic cyclotron resonance, Doppler-shifted cy-
The changes in amplitude and traveling time of clotron resonance, coupled acoustic and electro-
the pulse through the sample are determined by magnetic waves. These magnetoacoustic effects
beating the measured pulse with a reference sig- can provide extremum cross-sections, Fermi sur-
nal. Pulse and pulse-phase acoustic spectrometers face geometry, cyclotron masses, Fermi veloci-
operate in both “transmission” and “reflection” ties and other energetic and kinetic parameters of
modes. In the first case, two acoustic wave trans- conduction electrons (see also Electron–phonon
ducers are placed on opposite sides of the sample, interaction, Electromagnetic generation of sound,
whereas in the second one, the same transducer is Acoustic wave ellipticity, Rotation of acoustic wave
used for exciting and recording the acoustic os- polarization plane). Acoustic spectroscopy is also
cillations. The pulse duration must be much less an effective technique for studying an energy gap
than its transit time through the sample. Con- and its dependence on temperature and crystallo-
tinuous wave acoustic spectrometers operate in graphic direction.
transmission only, and are used in measurements Acoustic spectroscopy of paramagnetic centers
requiring low sensitivity to absorption factor vari- in insulator and semiconductor crystals is based
ations. To increase the acoustic spectrometer re- on the application of hypersound waves for in-
sponse, use is made of low-frequency modulation vestigating the electron energy spectrum structure
of external parameters, with subsequent synchro- and the electron–phonon interaction of paramag-
nous demodulation or digital accumulation of the netic centers based on hypersound acoustic para-
signal. magnetic resonance (APR) spectroscopy. The res-
onant absorption of hypersound corresponding to
ACOUSTIC SPECTROSCOPY of solids APR in dielectric or semiconductor crystals con-
A collection of methods for investigating the taining paramagnetic centers is associated with in-
resonant and oscillatory effects induced by the duced transitions between the levels of the elec-
interaction of acoustic oscillations with different tron paramagnetic centers’ energy spectrum, and
subsystems (electron, nuclear, spin, etc.) in a solid. the APR line intensity (resonant absorption value)
Acoustic spectroscopy of normal metals. The is proportional to the electron–phonon coupling
acoustic spectroscopy of pure metals in external of the center. That is why the APR characteris-
magnetic fields at low temperatures is the most di- tics such as the resonant magnetic eld and line
verse case since a large number of non-damping width, as well as their dependence on the tempera-
low-frequency collective waves can exist in met- ture, relative orientation of the magnetic field and
als. Owing to the cyclotron character of con- crystallographic directions, yield direct and accu-
duction electron motion and its interaction with rate data on the positions of the ground and excited
acoustic waves in pure metals at low temperatures, energy levels of the paramagnetic center, and its
the oscillating and resonant features of the sound interaction with selected crystal lattice vibrations
absorption and the sound velocity dispersion de- (electron–phonon interaction). Hypersound APR
pend on the external magnetic field. If the condi- spectroscopy is an effective technique for study-
tions kl, Ωτ
1 are met (l, τ are the length and ing the energy spectrum of non-Kramers centers
time of the mean free path of conduction electrons, with forbidden magnetic-dipole transitions and,
k is the wave vector of the acoustic wave, and Ω especially, paramagnetic centers with orbital de-
is the electron cyclotron frequency), one can ob- generacy in cubic crystal e lds, involving a strong
serve a magnetoacoustic effect whereby particular electron–phonon–ion interaction.
electrons most effectively interact with the sound, Acoustic spectroscopy of magnetically ordered
as indicated by the magnitude and direction of the crystals. These crystals exhibit elementary ex-
constant magnetic field relative to k, and by the citations of magnetization, such as magnons or
value of parameter ωτ (ω is the frequency of the spin waves. The magnetic subsystem energy is
sound wave). Particular phenomena appear, such determined by the electric (exchange) and mag-
as giant quantum oscillations of sound absorption netic (dipole–dipole and spin–orbit) interactions
ACOUSTIC WAVE AMPLIFICATION 15
which depend on the distances between atoms. ACOUSTIC VIBRATIONS of crystal lattice
Magnons interact with phonons and undergo mu- A type (branch, mode) of small oscillations of
tual scattering. This interaction is especially pro- crystal atoms, when a displacement of the center
nounced at crossings of magnon and phonon spec- of mass of a unit cell occurs; sometimes they are
tral branches in the region of magnetoacoustic called translational vibrations.
resonance. In this case, coupled magnetoelastic The dispersion law (the dependence of the fre-
waves having the amplitude equally dependent quency ω on the wave vector k) of acoustic vi-
on the magnetization and elastic strain propagate brations for small values of k is linear: ω = vph k,
in the crystal. Hypersound investigations of mag- where vph is the phase velocity of the atomic dis-
netoacoustic resonance, which underlie acoustic placement waves. It is an ordinary sound wave
spectroscopy of magnetically ordered crystals, (hence the term “acoustic branch”), i.e. a mat-
provide direct data on the structure of the magnon ter density wave with frequency 0–1013 s−1 . At
spectrum. One measures the resonant absorption high frequencies, the dispersion law becomes non-
value and the hypersound velocity change in a linear. In a three-dimensional crystal, there are al-
magnetic field as a function of the temperature for ways three branches of vibrations: with k directed
several relative orientations of the external mag- along the symmetric direction in the crystal, one
netic field and the crystal. Acoustic spectroscopy longitudinal and two transverse modes exist (the
is highly effective for the study of phase transi- same as for an elastic-isotropic medium), the dis-
tions in magnetically ordered crystals. placement vectors in transverse waves being mu-
tually perpendicular. The quanta of vibrations are
ACOUSTIC SPIN ECHO referred to as acoustic phonons. See also Crystal
Coherent acoustic response of an electric or lattice dynamics and Phonons.
nuclear spin system in a paramagnetic crystal
to the application of an acoustic or acoustico- ACOUSTIC WAVE AMPLIFICATION by current
electromagnetic pulse sequence during its reso- carrier drift
nant interaction with a spin system. An effect which is the reverse of the elec-
Through the spin–phonon interaction (see Spin tronic attenuation of acoustic waves. Amplifica-
Hamiltonian), the acoustic pulse creates an in- tion of acoustic waves by charge carrier drift oc-
phase precession of dipole moments or quadrupole curs in semiconductors and layered structures of
moments which, after the end of the pulse, become the piezodielectric-semiconductor type when the
dephased in the time T2 under the influence of re- drift velocity of free current carriers exceeds the
laxation processes. If the excitation pulse length phase velocity of the acoustic wave. For this am-
is less than T2 , the spin system dynamics is in- plification to occur the acoustic waves must in-
dependent of the excitation field amplitude at that teract with free charge carriers. The strongest
given time, but it is determined by its value in the and most extensively studied interaction mech-
preceding period. Therefore, the application of the anisms are those involving the piezoelectric ef-
second pulse after the time τ < T2 causes the re- fect (see Piezoelectricity) and the strain poten-
versal of the dephasing process so that the initial tial. The piezoelectric interaction is dominant
phased state is recovered after the time interval 2τ in piezosemiconductors at low frequencies
(so-called phase memory effect). In materials with (<10 GHz). When travelling through a piezoelec-
a strong spin–phonon interaction acoustic oscilla- tric crystal, an acoustic wave is accompanied by
tions are generated (acoustic echo-response) at the a wave of electric potential ϕ. The free charge
moment of full phasing, in addition to the genera- carriers (electrons) screen the wave field and be-
tion of electromagnetic field pulses. Acoustic spin come clustered in such a way that a space charge
echo is the acoustic analogue of the magnetic res- travelling wave arises. As this occurs energy ex-
onance spin echo and photon echo. Acoustic spin change between the wave and the electrons takes
echo enables one to measure the relaxation time place. The physical features of this exchange de-
and the spin–phonon interaction. Spin memory ac- pend essentially on the ratio between the acoustic
counts for such nonlinear effects as acoustic self- wave length λ, and the length l of the charge car-
induced transparency and induction. rier mean free path. If l/λ 1, then electrons
16 ACOUSTIC WAVE AMPLIFICATION
K 2q ωτM (1 − vd /vs )
αe = ,
2 ω2 τ 2 (1 − vd /vs )2 + (1 + q 2 r 2 )
M D
(1)
where K is the electromechanical coupling con-
stant for the given direction of wave propagation,
τ M is the Maxwell relaxation time, r D is the De-
bye screening radius, ω is the wave frequency, and
q = 2π/λ. Eq. (1) provides, for the most part, an
adequate description of experiments on acoustic
Fig. 1. Phase relationships between potential wave os- wave amplification in piezosemiconductors which
cillations (solid lines) and electron concentration oscil- exhibit moderate electron mobility: CdS, CdSe,
ZnO, GaAs, etc. (μ = 50–5500 cm2 ·V−1 ·s−1 ).
lations (dashed lines) for three successive times (a), (b)
and (c).
The general nature of the relation between the
electron absorption coefficient and the drift veloc-
at wavelength λ undergo a large number of colli- ity is shown in Fig. 2. Experimentally obtained
sions. Due to momentum relaxation, the motion of electron amplification coefficients are of the order
electrons with the wave is not inertia-free: as a re- of tens and even hundreds of dB/cm.
sult, the peak values of the electron concentration The nature of the interaction changes with in-
lag behind the potential minima, and the majority creased frequency and with increased mean free
of grouped electrons become located at those po- path of the carriers. As the wavelength approaches
sitions of the charge pattern where the wave field the carrier mean free path the energy exchange
accelerates electrons in the wave propagation di- between free carriers and the waves ceases to be
rection (see Fig. 1(a)). As this takes place, the en- dependent on collisions. In the case of a purely
ergy of the wave is transferred to electrons, and the
wave decays.
The generation of a current carrier drift in the
wave propagation direction in a semiconducting
crystal brings about a change of the value of the
absorption. Once the drift velocity v d equals the
acoustic wave velocity v s , the concentration max-
ima coincide in space with the negative poten-
tial peaks, and the electron absorption discontin-
ues because the wave no longer expends energy
on inducing electron transport (Fig. 1(b)). When
v d > v s , the peak values of electron concentration
are ahead of the negative peaks of the potential Fig. 2. Dependence of the absorption coefficient α on
(Fig. 1(c)). The electrons, which had been accel- the ratio v d /v s of the electron drift velocity to the wave
erated by the drift field are now decelerated by the velocity for the two limits ql 1 (curve) and ql
1
alternating wave field. The energy, which is ac- (straight line). Note that for both limits at v d = v s the
quired by electrons from the electron drift genera- coefficient α changes sign.
ACOUSTIC WAVE ELLIPTICITY 17
ACOUSTIC WAVE FOCUSING acoustic channels in space, and increasing the ef-
Concentration of the sound energy in a small ficiency of the non-linear wave interaction.
region by the formation of a convergent wave front.
In an isotropic medium the convergent wave fronts ACOUSTIC WAVE REFRACTION
are portions of spherical and cylindrical surfaces. Variation of the direction of wave propagation
The minimum sizes of focused regions are of the in a nonuniform medium caused by the depen-
order of the sonic wave length. Focusing is used dence of the phase velocity on the coordinates. The
for its local effect in small regions of space in the term “refraction of acoustic waves” refers to cases
medium under investigation. Acoustic lenses and where geometrical acoustics is applicable, and it
reflectors of corresponding shape may be used as means the refraction of sonic beams at a planar or
focusing elements. smoothly bent (on the scale of wavelengths) in-
terface between two media. In the medium with
ACOUSTIC WAVEGUIDE a continuous dependence of the phase velocity on
A region of a medium relatively small in cross- the coordinates the beam direction varies continu-
section compared to its length, which localizes ously, which involves refraction of waves. At the
acoustic wave energy for the transmission and interface between two anisotropic media (see Bulk
processing acoustic signals.
acoustic waves, Elastic modulus tensor) one inci-
The energy localization inside the wave-guid-
dent plane harmonic elastic wave in the general
ing structure is influenced by the acoustic nonuni-
case excites three waves – one quasi-longitudinal
formity of the medium. In the case of a smooth
and two quasi-transverse refracted, and three re-
nonuniformity, the velocity profile of an acoustic
flected waves (see Re e ction of acoustic waves).
plane wave must have a minimum inside the struc-
The wave vectors of all these waves lie in the
ture in order to focus acoustic waves into the
plane of incidence, determined by the normal to
waveguide, taking into account wave front cur-
the interface and by the incident wave vector. In
vature. In the case of a sharp nonuniformity of
anisotropic cases refraction of the wave vector in
the acoustical properties at the boundary, almost
the reflection of energy (sonic beam) from the in-
all the energy is localized inside the waveguide,
and is retained there owing to the repeated reflec- terface can also occur (see Group velocity).
tions from the boundaries. The interference of the
waves produces a complicated field pattern in the ACOUSTIC WAVE ROTATION
waveguide arising from the superposition of nor- See Rotation of acoustic wave polarization
mal modes of acoustic oscillations. plane.
Most widespread are planar acoustic wave-
guides which are inhomogeneous structures on ACOUSTIC WAVE SCATTERING in solids
the surface of substrates: strip waveguides formed The process of formation of weak additional
from applied strips of another material, topo- sound waves, differing in their characteristics from
graphic waveguides involving grooves and ridges the incident wave, as a result of acoustic diffrac-
etched on the substrate, and diffusion waveguides tion in a medium with inhomogeneities or obsta-
wherein the inhomogeneities arise from the im- cles. The process of scattering may be either co-
plantation of impurities in the substrate. There are herent or incoherent, depending on whether the
also fiber acoustic waveguides. frequency of the vibrations is conserved or not
Distinctive features of the waveguide propaga- during the scattering. In either case the scatter-
tion are the presence of many acoustic vibration ing results in the attenuation of the sound wave.
modes with strong dispersion. The latter is widely An example of incoherent acoustic wave scatter-
used for the development of dispersion devices for ing is the observation of hypersound in a dielec-
acoustic signal processing: ultrasonic delay lines, tric crystal at low temperatures when the condi-
filters, waveguide lenses. The pronounced energy tion ωτ > 1 is satisfied (ω is the hypersound angu-
localization inside a waveguide enables the di- lar frequency, and τ is a thermal phonon lifetime).
recting and transmitting of acoustic signals: with- Absorption in that case is due to a three phonon
out loss from the wave divergence, multiplexing scattering process whereby the incident phonon
ACOUSTO-ELECTRIC EFFECT 19
flux forms a hypersonic wave of thermal phonons resonators or transmission lines. A thin magne-
with long lifetimes, a result which arises from tostriction transducer of acoustic waves is a mag-
anharmonic vibrations of the crystal lattice. Irre- netic lm applied at the end of the sample. The
versible absorption of the hypersound takes place, film magnetized by an applied magnetic field is
with the incident energy transferred to the ther- deformed by the oscillating magnetic component
mal phonons. For coherent acoustic wave scat- of the UHF field. Improved matching of the im-
tering off static inhomogeneities in a solid, scat- pedance of the acoustic wave transducer with the
tered waves of the same frequency appear, with UHF guides and resonators may be achieved with
characteristics that depend on the type of incident the aid of piezoelectric transducers based on the
wave, the inhomogeneities, and the boundary con- depletion layers of piezosemiconductors, formed
ditions at the surface. An example of such a co- with many-layered and many-element construc-
herent process may be found in the scattering of tions. For the excitation and recording of sur-
sound from crystalline inhomogeneities, such as face acoustic waves one may employ counter-pin
the temperature-independent residual attenuation transducers, in which the distribution of the elec-
of hypersound in dielectric crystals at low temper- tromagnetic field and the corresponding deforma-
atures. tion at the sound guide surface is pregiven or fixed.
These acoustic wave transducers are also used for
ACOUSTIC WAVE, SUBSURFACE excitation and recording of volume modes, spread
See Subsurface volume acoustic waves. at an angle to the surface at which the transducer
is placed. This mode of exciting volume oscilla-
ACOUSTIC WAVE, SURFACE tions is used in dispersion delay lines when the
See Surface acoustic waves. effect of the sound velocity dispersion is reflected
by the passage through different distances of vi-
brations with different wave lengths. For measure-
ACOUSTIC WAVE TRANSDUCER
ments of the absolute values of crystal lattice vi-
A device which transforms an electromagnetic
brations capacitance transducers have found their
vibration into acoustic ones, and vice versa. The
wide application, with their design in the shape of
majority of acoustic wave transducers are recip-
flat plate capacitors, one plate of which is rigidly
rocal, i.e. they may serve both as emitters and
connected with the sample. See also Phonon gen-
as receivers of acoustic vibrations. In ultrasonic
eration, Electromagnetic generation of sound.
and hypersonic investigations of bulk acoustic
waves the most common transducers make use
of the piezoelectric effect (see Piezoelectricity) or ACOUSTO-ELECTRIC DOMAIN
the phenomenon of magnetostriction. In resonance A stationary or moving region of strong elec-
piezoelectric transducers, which have a thickness tric field and an elevated intensity of acoustic noise
equal to an odd number of half wavelengths, which forms in semiconductors as a result of the
monocrystalline quartz or lithium niobate (see amplification of thermal crystal lattice vibrations
Niobates) plates are used as operating elements by the electric current at a charge carrier drift ve-
at frequencies up to 1 GHz. Alternative materials locity faster than that of the sound (see Acousto-
are barium titanate, lithium tantalate, bismuth ger- electric effect).
manate, and piezoelectric ceramics. At frequen-
cies in the UHF range the resonance acoustic wave ACOUSTO-ELECTRIC EFFECT
transducers are applied at the samples under in- Establishment of a constant current or electro-
vestigation, or at the sound wave guides in the motive force caused by the transfer of momen-
form of thin lms of ZnO, CdS and other piezo- tum and energy from a traveling ultrasonic wave
electric materials. The alternating electric field is to conduction electrons (see Acousto-electronic in-
applied and recorded at the piezoelectric element teraction).
either by the direct insertion of the latter into the The capture of electrons by the local electric
electromagnetic circuit, or with the help of UHF fields arising in the conductor under the action
20 ACOUSTO-ELECTROMAGNETIC EFFECT
of an ultrasonic wave leads to their being car- the acting force field is always nonuniform be-
ried along in the direction of the sound propa- cause the oscillatory displacement of the particles
gation, and an acousto-electromotive force arises decays into the bulk of the sample.
along the sample. When the sample is introduced
into a closed electric circuit an electric current ACOUSTO-ELECTRONIC AMPLIFIER
begins to flow through it. The current density is See Acoustic wave ampli cation.
jA.E. = −2μΓ I /v (G. Weinreich, 1957), where
Γ is the coefficient of sound absorption by the ACOUSTO-ELECTRONIC INTERACTION (AEI)
conduction electrons, I is the ultrasonic wave in- Interaction of acoustic waves with charge car-
tensity, v is the sound velocity, μ = eτ/m is the riers (conduction electrons and holes in metals,
mobility, τ is the mean free path time. As seen semimetals and semiconductors).
from the expression for jA.E. , the sign of the In metals, due to the Pauli’s principle, the inter-
acousto-electric current or acousto-electromotive action with acoustic waves (sound) involves only
force EA.E. = jA.E. σ (σ is the conductivity) electrons with energies close to Fermi energy EF .
depends on the sign of Γ , and can change de- There are two main mechanisms of AEI: (1) the
pending on the ultrasonic wave propagation direc- deformation mechanism involves a change in pe-
tion, its damping, or its amplification. During the riodic potential of the crystal lattice and a local
propagation of a surface acoustic wave a trans- change of the dispersion law of quasi-particles
verse acousto-electric effect is observable. When E(p) in the electron Fermi liquid, (2) the elec-
a transverse acousto-electric current directed into tromagnetic mechanism is due to forces exerted
the sample bulk arises along with the longitudinal on the electrons by electric and magnetic fields
one; this is linked with the wave localization on arising from the displacement of ions during the
passage of the ultrasonic wave. For longitudinal
the surface. See also Acoustomagnetoelectric phe-
sound, the fields are electrostatic and are screened
nomena, Surface acoustic wave.
by free electrons (see Electric charge screening)
to a great extent. This leads to renormalization
ACOUSTO-ELECTROMAGNETIC EFFECT (p)
of the strain potential Λik = λik (p) − λik (p),
The appearance of a magnetic moment in a where the angle brackets indicate averaging the
semiconductor crystal under application of a suf- bare deformation potential λik over the Fermi sur-
ficiently strong electric field causing the amplifi- face. The quantity Λik can be interpreted as a
cation of acoustic noise (phonons). The generated distinctive phonon electric-charge in the metal.
flux of acoustic energy results in the movement of During propagation of the transverse sound wave
charge carriers (see Acousto-electric effect), and in the electrons are affected by the vortex electric
some cases the charges may not move along the fields (see Acousto-electric effect). In an external
electric field direction (e.g., in an anisotropic crys- magnetic field B, an additional, so-called induc-
tal where the direction of maximal noise amplifi- tion interaction of the electrons with the acoustic
cation may not coincide with that of the applied waves arises, its strength being e(v · [U × B]),
electric field). As a result, a circular electric cur- where v is the electron velocity, and U is the
rent begins to flow around the sample, and hence ion displacement velocity. This interaction in-
the appearance of a magnetic moment. Acoustic volves the additional parameter (qRL )−1 , where
energy flux injected into a sample from outside q is the wave vector of the sound wave, and
can induce a magnetic moment in the absence of RL = pF /eB is the electron Larmor radius. The
the external electric field (acoustomagnetic effect). induction interaction prevails over the deforma-
The nonlinearity of charge carrier motion is re- tion in sufficiently strong magnetic fields when
lated both to the anisotropy of crystals, and the qRL < 1. Besides, in the presence of a magnetic
acoustic energy flux nonuniformity. This phenom- field AEI leads to various magnetoacoustic ef-
enon is also observable in metals. The acousto- fects, in which electromagnetic resonances, bal-
magnetic field arises, in particular, during the listic and collective effects add their characteristic
propagation of surface acoustic waves. In this case features.
ACOUSTO-ELECTRONICS 21
The great difference between the electron and light and acoustic waves). Acousto-electronics
ion masses leads to the fact that the sound velocity was formed as an independent part of electronics
is many times less than the Fermi velocity of the in the decade of the 1960s with the start of in-
electrons. This circumstance (adiabaticity) plays tensive studies stimulated by the discovery of the
an important role in the interaction of electrons acoustic wave ampli cation by conduction elec-
with sound. Adiabaticity supports introducing the trons in cadmium sulfide crystals. The rapid de-
concept of electrons and phonons as weakly in- velopment of acousto-electronics was caused by
teracting quasi-particles, and makes it possible to the necessity to create simple, reliable and small-
neglect inertial terms (Stewart–Tolman effect) in scale radio signal processing devices for radio-
the AEI energy. In pure metals with a long mean electronic equipment. At the present time acousto-
free path l (ql
1) the adiabaticity underlies the electronic devices perform the following signal
fact that only electrons with a velocity v almost transformations: (1) in time (delaying and chang-
perpendicular to the sound wave vector can ef- ing the duration of signals), (2) in phase and
fectively interact with the sound (autophasing ef- frequency (frequency and spectrum conversion,
fect). phase shift), (3) in amplitude (amplification, mod-
In semiconductors, the AEI is mainly deter- ulation), and (4) those involving more complicated
mined by the strain interaction related to the lo- functional transformations (coding and decoding,
cal change in the band gap under the action of the signal convolution and correlation). In some cases,
strain arising when a sound wave travels through the acousto-electronic methods of signal transfor-
the crystal. In ionic crystals and semiconductors mation are simpler (as compared, e.g., to elec-
with a large number of charged impurity atoms in tronic methods) and, sometimes, they constitute
the lattice, the electromagnetic mechanism of AEI the only possible choice. The use of these de-
is also possible. vices is predicated on the low propagation veloc-
In superconductors, the AEI at temperatures ity of acoustic waves (compared to that of elec-
lower than critical temperature of superconductors
tromagnetic waves), weak absorption of acoustic
differs in some features from the AEI in normal
waves in crystals with a high acoustic Q-factor
metals. Due to the Cooper pairing of electrons (see
(quality factor), as well as favorable interactions
Cooper pairs), the phonon absorption probability
of these waves with electromagnetic fields and
decreases and, as a consequence, a sharp (expo-
electrons. In acousto-electronic devices, both bulk
nential) decrease of the ultrasound absorption fac-
acoustic waves and surface acoustic waves (SAW)
tor occurs. Besides, the Meissner effect precludes
are used. The manufacture of these devices em-
the electromagnetic mechanism of AEI.
ploys piezoelectric materials and laminated struc-
ACOUSTO-ELECTRONICS tures composed of piezoelectric and semiconduc-
A branch of electronics dealing with the effects tor layers.
of excitation and propagation of acoustic waves Many acousto-electronic devices convert high
in condensed media, the interaction of the waves frequency electric signals into acoustic waves
with electromagnetic fields and electrons, and the (acoustic wave transducers) to propagate in a
development of acousto-electronic devices based sound guide and then be converted back again
on these effects. into a high frequency signal. Components used for
The diversity of effects utilized in the design excitation and reception of bulk acoustic waves
of acousto-electronic devices suggests the divi- are mainly piezoelectric transducers: piezoelec-
sion of acousto-electronics into the high frequency tric plates (at frequencies up to 100 MHz) and
(microwave) acoustics of solids (effects of ex- film transducers (at frequencies above 300 MHz),
citation, propagation and reception of high fre- while those employed for the excitation and recep-
quency acoustic waves and hypersonic waves), tion of a SAW are opposite-pin transducers.
acousto-electronics proper (interaction of acoustic The ultrasonic delay lines operating in the fre-
waves with conduction electrons in the bulk and on quency range from several MHz to several tens of
the surface of solids, see Acousto-electronic inter- GHz, as well as quartz resonators for the radio fre-
action), and acousto-optics (interactions between quency generator stabilization (see Acoustic solid-
22 ACOUSTO-ELECTRONICS MATERIALS
Laboratory product
Laboratory product
vices based on the SAW due to small loss for
Industrial product
Industrial product
Industrial product
Industrial product
wave excitation, the possibility of wave prop-
agation control at any point of a sound guide
(along the wave propagation path), and the fea-
sibility for designing devices with controllable
frequency, phase and other responses. Such de-
vices include resonators on the SAW, which are
loss in air
100 MHz
Acoustic
(dB) at
0.0145
0.03
0.05
0.05
0.95
0.10
0.10
are also introduced into a generator circuit for
its frequency stabilization; delay lines (for long
signal delay in memory elements); various filters
coefficient of delay
Existence of cut
yes
yes
yes
yes
yes
no
no
no
no
no
3488
3808
1681
2270
2300
3138
3295
2640
2505
3100
1010
coupling
0.014
0.015
0.015
0.073
0.062
0.05
0.09
0.01
0.01
[110]
[100]
ST, x
ACOUSTO-ELECTRONICS MATERIALS
Pb0.91 In0.03 [Zr0.44 Ti0.53 (Li3/5 W2/5 )0.08 ]O3
and possibilities offered by applying a piezoelec- composition and pressure of the ambient gaseous
tric material layer to a semiconductor substrate (or medium. See also Acoustoluminescence.
vice versa) are realized in these systems. Parame-
ters for the materials of such systems are given in ACOUSTOMAGNETOELECTRIC PHENOMENA
Table 2. Kinetic effects in semiconductors and metals
arising from the entrainment of conduction elec-
ACOUSTOLUMINESCENCE
trons by an ultrasonic wave in the presence of
Luminescence in a solid arising under its ex-
an external magnetic field. As is the case with
citation by an acoustic wave with an energy flux
galvanomagnetic effects, acoustomagnetoelectric
density in excess of a certain threshold value.
phenomena are due to charge carrier trajectories
The threshold energy flux density (typically
bending in the magnetic field (see Lorentz force).
1–10 W/cm2 ) decreases as the density of struc-
Identified as acoustomagnetoelectric phenomena
tural defects (e.g., dislocations) rises. The lumi-
are the acoustic analogs of the Hall effect, Etting-
nescence is nonuniform throughout the bulk, and
is concentrated in regions of defect accumulation. shausen effect, and so on. Their distinctive feature
The spectrum of acoustoluminescence at the en- is that the ultrasonic flux rather than the external
ergy flux threshold is close to that of photolumi- electric field creates the initial current. There also
nescence of the corresponding material, it con- exists the acoustomagnetoelectric effect consist-
tains bands associated with intrinsic structural de- ing in the appearance of an electromotive force
fects and impurity centers, and it is stable in time in the direction perpendicular to the ultrasonic
(some hours). The increase of energy flux den- flux under the action of the magnetic field, when
sity by a factor of 1.5 to 2 relative to the thresh- there is no acousto-electric current present (see
old leads to a sharp change in the acoustolumi- Acousto-electric effect). In bipolar semiconductors
nescence spectrum. A large number of narrow and semimetals, the acoustomagnetoelectric ef-
lines having half-widths 1 meV appear instead fect is caused by the separation of the electron–
of wide bands with half-widths of some tens of hole flux drawn along by the ultrasonic wave in
meV. The restructuring of the spectrum is accom- the magnetic field. In monopolar semiconductors
panied by an instability in the amplitude and sign and metals, the acoustomagnetoelectric effect is
of the acousto-electric current. Under these condi- associated with the fact that carriers with differ-
tions the luminescence is unstable in time. Along- ent energies are entrained by the ultrasonic wave
side of the bulk luminescence there is a surface lu- and deflected differently by the magnetic field.
minescence with a spectrum which depends on the Therefore, the magnitude and even the sign of the
24 ACOUSTO-OPTIC DEFLECTOR
monopolar effect are sensitive to the carrier scat- of the intensity of the radiation diffracted in two
tering mechanism. Acoustomagnetoelectric phe- acousto-optic cells replicates the correlation func-
nomena are also possible in a planar configuration tion of the acoustic wave amplitude in the cells.
when the directions of the ultrasonic flux, mag- These acousto-optic devices are components of
netic field, and resultant electromotive force lie various systems for radio signal processing.
in the same plane. Acoustomagnetoelectric phe- Of particular importance are acousto-optic ma-
nomena have been observed in Bi, Te, InSb, and trix processors intended for algebraic matrix op-
graphite. erations over numeric data arrays because the lat-
ter can be entered into the processor through the
ACOUSTO-OPTIC DEFLECTOR acoustic and optical input.
See Light de ectors.
ACOUSTO-OPTIC DIFFRACTION, diffraction of
ACOUSTO-OPTIC DEVICES light by sound
A class of active functional devices based on A class of phenomena involving the deviation
light diffraction by acoustic waves (see Acousto- of light from a straight path in a medium under the
optics, Acousto-optic diffraction).
action of an ultrasonic wave.
These devices are used for measuring opti-
By changing the refractive index of the medium
cal radiation characteristics, controlling its para-
in a periodic manner the ultrasonic wave acts on
meters, and processing optical, radio-frequency
the light as a phase diffraction grating moving at the
and acoustical signals. The basic element of an
sound velocity. When a monochromatic light beam
acousto-optic device is an acousto-optic cell, a
passes through the region of a medium where an
crystal where the ultrasound is excited by the
ultrasonic wave propagates, then beams of dif-
acoustic wave transducer, and where the acousto-
fracted light appear. The frequencies and wave
optic interaction takes place. Since the cell has
vectors of the incident (subscript I) and diffracted
two inputs (optical and acoustical), two classes of
(subscript D) light waves are interrelated through
acousto-optic devices can be distinguished.
the acoustic wave parameters ωac and kac :
One type for controlling light beam parame-
ters includes adjustable optical lters, light de- ωD = ωI ± mωac , kD = kI ± mkac ,
e ctors and monochromatic radiation modula-
tors (see Modulation spectroscopy). They have where the order m of the diffraction is an integer.
the advantages: fast response (adjustment time The following basic types of acousto-optic diffrac-
1–30 μs), large light-gathering capability (angular tion are known:
aperture 1–15◦ , spatial aperture 1–10 cm2 ), con-
siderable resolution (number of resolvable points 1. Raman–Nath diffraction (sometimes called
is 103 –104 ), and operational flexibility. The ad- Debye–Sears diffraction or Lucas–Biquard dif-
justments of the transparency window filters and fraction), which is equivalent to the diffrac-
the beam scanning deflectors are implemented tion by thin holograms, arises for normal light
by changing the ultrasound frequency, whereas incident on a low-frequency ultrasonic beam
2 L/k 1 holds, where
when the relation kac
the light intensity modulation is carried out by I
controlling the sound power. These elementary L is the length of the domain of interac-
acousto-optic devices can be components of more tion of the light and ultrasound in the direc-
complicated instruments: spectrometers, group tion of the first diffraction maximum. After
switches, scanners, etc. passing through the ultrasonic wave, the light
Another class of acousto-optic devices is in- beam exhibits a sinusoidal phase variation in
tended for processing signals carried by the ultra- the plane perpendicular to the direction of its
sound, e.g., radio-frequency signals. In spectrum propagation. As a result, a series of diffracted
analyzers the angular dependence of the diffracted beams appears with frequencies ωD charac-
monochromatic beam repeats the ultrasound spec- terized by m = 1, 2, . . . , propagating at an-
trum, while in correlators the time dependence gles θ = arcsin(mλ/Λ), where λ and Λ are
ACOUSTOTHERMAL EFFECT 25
the wavelengths of the light and sound, respec- the action of the ultrasonic wave. See also Brillouin
tively. The Raman–Nath diffraction is negligi- scattering, Raman scattering of light.
ble in the hypersound frequency domain where
the above inequality is not satisfied. ACOUSTO-OPTICS
2. At high frequencies when the inequality Study of the diffraction of optical radiation
2 L/k
1 holds, then only the zero and first
kac I by acoustic waves in solids, and the development
order maxima are observed for the diffracted
on this basis of devices for controlling light flux
light. This phenomenon is usually known as
parameters and processing optical and radio sig-
Bragg diffraction, because it arises only for
nals (see Acousto-optic diffraction, Acousto-optic
the light incident on the ultrasonic beam at
devices).
the Bragg angle θB = arcsin[λ/(2Λ)]. This is
The basis of acousto-optic diffraction in crys-
equivalent to diffraction by a thick hologram.
The diffraction efficiency is specified here by tals is the photoelastic effect (see Photoelastic ten-
sor). The elastic wave creates a periodic distur-
the formula
bance of the crystal dielectric constant which acts
ID kI L
= sin2 as a phase diffraction grating for the light. The fre-
II 4(M2 Iac )1/2 quencies and wave vectors of the incident and dif-
and can be as high as 100%, where M2 = fracted light waves are coupled through the para-
p2 n6 /ρv 3 is the acousto-optic quality factor meters of the acoustic wave. The spectral and an-
of the crystal determined by its density ρ, re- gular selectivity of acousto-optic diffraction can
fraction index n, photoelastic constant p, and be controlled by changing the sound frequency,
sound speed v. i.e. the period of the acoustic diffraction grating.
The sound intensity determines the energy distrib-
Under acousto-optic diffraction in anisotropic
ution between the initial and diffracted rays. These
media, the polarization of the incident and dif-
effects underlie acousto-optic methods and visual-
fracted light are generally not the same. Therefore,
ization of acoustic elds, measurement of different
under this diffraction with a rotated plane of po-
larization in an anisotropic medium (anisotropic acoustic characteristics of materials, and a wide
diffraction), as distinct from diffraction in an range of acousto-optic devices.
anisotropic medium without any polarization plane
rotation or in an isotropic medium, the angles ACOUSTOTHERMAL EFFECT
of incidence and diffraction are generally not the Appearance of a temperature difference in a
same. For oblique light incidence on a solid sur- thermally isolated body as a consequence of a heat
face, along which a surface acoustic wave prop- transfer by a traveling acoustic wave.
agates, the diffraction of the light is observed in The acoustothermal effect is a result of the de-
both the transmitted and reflected beams. This is parture of the sound wave regime from adiabatic
explained by the fact that, instead of photoelas- behaviour. It is more pronounced in materials with
tic changes of dielectric constant, there is a pe- a high thermal conductivity (metals and metal al-
riodic bending of the surface, which corresponds loys). If the thermal conductivity κ rises with in-
to the normal component of the displacements in creasing pressure, P , the sample-averaged gradi-
the surface acoustic wave, and creates a varying ent of temperature ∇T is oriented opposite to the
difference in path-length for the light waves. In a wave propagation direction
conducting medium the acoustic wave creates two
diffraction gratings with the same spatial period, 1 ∂κ K 2 b2 ω 4
typically phase-shifted, for the electromagnetic ∇T = − αT 2 2 ,
2 ∂P vω κ + v 5 c2
wave. One of the gratings corresponds to a change
in dielectric constant owing to the deformation of where α is the volume thermal expansion coeffi-
the lattice by the acoustic wave. The other one cient, K is the bulk modulus, b, ω and v are the
constitutes an electron diffraction grating due to amplitude, angular frequency and phase velocity
the local change in electron concentration under of the wave, respectively; and c is the speci c heat
26 ACTINIDES
concentration of adsorbed molecules in the tran- Examples of thermally activated processes are the
sient activated complex. This complex forms at the following: diffusion of impurity atoms or vacan-
potential energy maximum along the most ener- cies; electrical conductivity of semiconductors; de-
getically favorable path of chemisorption. cay of metastable states at rst-order phase transi-
tions by the formation of nuclei of the new phase;
ACTIVATION ANALYSIS plastic deformation of a crystal due to the motion
A method for determining the chemical com- of dislocations overcoming the Peierls barriers
position of a substance, together with the impu- (see Peierls relief) or local defects. Such processes,
rities that are present, with the aid of parameters referred to as elementary events (elementary struc-
induced by radioactivity. tural rearrangements), are the passage of a dif-
The sample is exposed to a particular type of fusing atom between neighboring equivalent po-
nuclear radiation that induces nuclear reactions in sitions in a crystal lattice, the birth of a current
atoms of the material. Then, the type of induced carrier in the conduction band of a semiconductor,
radioactivity is determined experimentally, along the breaking loose of a dislocation from its pin-
with the secondary particle energy and the half- ning center, the formation of a critical embryo of a
life. Using these data with the aid of nuclear re- new phase, and so on. Therefore, one can charac-
action tables one can deduce the isotopes of the terize the rate of a thermally activated process by
chemical elements that underwent the radioactive the number of elementary events per unit time ṅ:
transformations. The intensity of the secondary ra-
diation and the cross-sections of the correspond- U
ṅ(T ) = ṅ0 exp − , (1)
ing nuclear reactions provide the initial concen- kB T
trations of these isotopes. The sensitivity of the
method varies over a wide range for isotopes of where ṅ0 is a normalization constant. This expres-
different elements. A sufficiently large nuclear re- sion, called the Arrhenius law, is used in chemical
action cross-section is required, and the half-life kinetics to describe thermally activated reactions.
should be not be too long and not too short. Activa- This specific equation for the activation energy,
tion analysis is efficient for determining the gold, and its relationship with the parameters describing
platinum, oxygen, boron content in silicon; the nio- the structure and the state of a material, depend
bium in tantalum, and others. Activation analysis on the type of process under consideration. The
establishes the total impurity content in the sam- detailed expression for Eq. (1), and the calcula-
ple, but is insufficient for deducing physical states tion of the activation energy U for the majority of
in a crystal (phase, doping atoms versus substitu- cases require solving a complicated statistical me-
tional atoms, etc.). Depending on the type of acti- chanical problem. A simple qualitative interpreta-
vating radiation, one can distinguish neutron ac- tion of Eq. (1) based on the thermodynamic the-
tivation analysis, proton activation analysis and ory of uctuations consists of the following: an in-
gamma activation analysis. dividual event of the thermally activated process
includes overcoming a certain energy barrier of
ACTIVATION ENERGY height U ; the exponential probability law for its
Basic characteristic of a thermally activated occurrence as a function of the temperature is a
process which determines the rate of the process at consequence of exponential Gibbs distribution for
a given temperature T , and its sensitivity to a vari- the energy states of the subsystem where the given
ation of the temperature. The activation energy U elementary process takes place.
is related to the probability W of the occurrence of The value of the activation energy can depend
an elementary event whose successive occurrences on external parameters a which determine the ma-
implement the corresponding process through the terial condition: the mechanical stress, the electric
expression
field strength, and so on (U = U (a)). This circum-
U stance allows one to control the rate of a thermally
W = W0 exp − . activated process, not only through a temperature
kBT
28 ACTIVATION VOLUME
change, but also by varying the external parame- (or the number n of elementary acts per unit time)
ters a: ṅ = ṅ(T , a), where is described by the formula
U (a) V (σ )
ṅ(T , a) = ṅ0 exp − . (2) ṅ(T , σ ) = ṅ0 exp − . (1)
kBT kB T
One of the experimental methods for determin- The activation volume v is introduced formally
ing the activation energy is the measurement of as the negative derivative of the activation energy
the rate of a process at various temperatures while volume with respect to the stress (pressure), v =
keeping the parameter a constant, then drawing −dV (σ )/dσ . The activation energy change under
a plot of ln ṅ versus 1/T , and finding the slope a small stress change δσ can be expressed via the
of the resulting straight line. Another method is activation volume using the relation V (σ + δσ ) =
based on the measurement of the sensitivity of the V (σ ) − v(σ )δσ . It should be noted that the acti-
rate of a thermally activated process to small varia- vation volume does not always have the sense of a
tions of the temperature δT (again keeping a con- physical volume. The experimental determination
stant), with the subsequent application of the fol- of an activation volume is based on the measure-
lowing formula deduced from Eq. (2): ment of the activation energy for different values
of the stress, following numerical differentiation
ln ṅ(T + δT , a) − ln ṅ(T , a) of the dependence so obtained, or on the change
U = kBT 2 in the process rate sensitivity to small stress vari-
δT
(3) ations at a fixed temperature using the following
∂ ln ṅ
≈ kBT 2 . formula derived from Eq. (1):
∂T a
ln ṅ(T , σ + δσ ) − ln ṅ(T , σ )
The measurement of the activation energy and the v = kB T
δσ
analysis of its dependence on external parame-
ters is called thermal activation analysis. This pro- ∂ ln ṅ
kB T . (2)
vides information on the physical nature of the el- ∂σ T
ementary reactions involved in the thermally acti- Measurement of the activation volume and analy-
vated process (see also Activation volume). sis of its dependence on the value of σ (or P ) is
one of the important components of the procedure
ACTIVATION VOLUME of thermo-activation analysis, and it provides in-
A quantity with the dimension of volume, formation on the energy and geometric parameters
which characterizes the sensitivity of the rate of of activation processes.
activation analysis in a solid relative to the change
of the applied mechanical stress (i.e. a change of ACTIVE CARBONS
stress tensor σ or pressure P ). Highly porous carbon materials containing
Initially, the term activation volume arose in 85–97% carbon.
the description of the pressure effect on point de- These materials are obtained by carbonization
fect diffusion in a solid, and it had the sense of and activation of carbon-containing substances
the volume change involved in the transition of a such as naturally occurring wood, mineral coals,
point defect from an initial equilibrium state to a peat, oil and coal cokes, as well as synthetic
so-called activated state. At present it is used in polymers such as phenol-formaldehyde, styrene-
a more general sense, and reflects the dependence divinylbenzene, and other resins. Carbonization
of the activation energy of any thermodynamically products, the so called carbonizates, comprise an
activated process on σ or P . For example, in the ordered part (graphite-like carbon crystallites) and
case of the diffusion of vacancies the parameter is an amorphous part (high carbon content radicals),
uniform pressure, while in the case of the motion associated mainly with the prismatic faces of crys-
of dislocations it is the shearing stress acting in its tallites. Upon activation (850 to 950 ◦ C), the oxi-
slip plane. In such cases the rate of the process ṅ dizing gases (CO2 , H2 O et al.) interact first of all
ACTIVE CENTERS 29
with the amorphous carbon, and then with the car- portance are surface compounds with oxygen. Es-
bon of the crystallites. The burning-out of impu- pecially prepared oxidized carbons form a sepa-
rities and partially of carbon from the contacting rate class of active carbons with specific properties
and coalescent carbonizate crystallites results in such as an extremely high cation-exchange selec-
the formation of microporous regions containing tivity. Surface compounds with nitrogen and sul-
pores of two types: intracrystallite ones with di- fur exert a strong effect on the properties of active
mensions up to 1.5 nm, which are flat slots with carbons.
walls formed by the basic faces of the crystallites; Active carbons, because of their well devel-
and intercrystallite ones with the walls in the form oped porosity, have important applications as ad-
of prismatic faces. A portion of the external sur- sorbents and catalysts: for air and gas purifica-
face of the microporous areas forms the surface tion, for extraction of solvents, for preparation
of larger mesopores (1.5 to 100 nm) and macrop- of drinking water, for purification of sewage, in
ores (200 nm). The dimensions of single zones in the production of food, and for the extraction of
active carbons are between 10 and 60 nm, while valuable components (e.g., rare and noble met-
their number per unit of mass for the majority of als) from their complex mixtures. Oxidized car-
active carbons is the order of 1016 . The number of bons are used for the acquisition and analysis of
micropores in a single microporous area is as high particularly pure materials for electronic engineer-
as several thousand. ing, and they are used in medicine for detoxifica-
Typical of active carbons is the turbostrate tion of blood and other biological liquids (hemo-,
structure characterized by the absence of strict lympho-, plasmo-, enterosorbtion), during treat-
periodicity of the elementary layers. The dis- ment of many diseases, and also as carriers of
tances between the latter are unequal, and fluctu- medicinal preparations.
ate within the range 0.334 to 0.365 nm, in contrast
to the equidistant graphite structure which has an
ACTIVE CENTERS
inter-layer separation of 0.33554 nm. The carbon
atoms in the crystallites of the turbostrate structure These centers are impurity atoms injected into
are displaced relative to the lattice plane within solids to produce changes in their electrical, op-
the range 0.014 to 0.017 nm. Active carbons pos- tical, magnetic and other properties. An injected
sess high electrical conductivity due mainly to impurity which brings about no change is called a
π -electrons in the planes of hexagonal rings along passive impurity. The impurity activity in a given
the system of conjugated bonds. The electric re- material is determined not only by its chemical
sistance of the active carbons depends strongly on nature, but also by its position in the crystal lat-
the pretreatment temperature and its duration, and tice, and this depends on the manner and condi-
can be 108 to 1010 ·cm for the carbonizates, and tions of its injection, and the subsequent process-
0.1 to 10 ·cm for well-activated carbons. This ing of the sample. Activation or passivation of
is associated with structural changes: of the con- impurities can be caused by heat treatment, the
ductive systems in cokes (carbonizates) separated effect of nuclear radiation, pressure, and other fac-
from one another by compounds with an aliphatic tors. For instance, electrically inactive interstitial
character which burn out during an oxidizing ther- atoms of oxygen in silicon transform under high-
mal treatment. The ordering of the structure is as- temperature annealing into donor centers (called
sociated with the reduction of the electric resis- thermo-donors). The injection of Au, Cu and other
tance. atoms into photoelectrically inactive silicon crys-
Active carbons, in comparison with other tran- tals results, after subsequent exposure to radiation,
sitional forms of carbon, are characterized by a in the appearance of photoconductivity. Impurity
broad spectrum of atomic carbon hybrid electronic activation can arise from the change in position of
states – sp2 , sp, sp2 π , etc. Dangling bonds at the impurity atom as a result of irradiation. Ac-
the surface of active carbons may bind oxygen, tive centers are not necessarily impurities, but can
nitrogen, hydrogen, sulfur, etc., with the forma- be intrinsic crystal lattice defects, as well as com-
tion of strong surface compounds. Of great im- plexes of intrinsic defects and impurity atoms.
30 ACTIVE LASER ELEMENT
Table 1. Important properties of active solid state laser materials doped with neodymium
primary radiation with the irradiated system; it is designate the trajectories of the motion of three
the active spectrum that causes radiation effects in particles, whose energies during the motion in the
solids. This spectrum is characterized by the set of time interval
t lie within the range (ε, ε +
ε).
functions Ni (ε, r), where ε is the particle energy, During the interval
t particles 1 and 2 intersect
r is the space coordinate, and the dimensionality the area
A at a right angle and contribute to
of N is MeV −1 ·cm−2 ·s−1 . The expression Eq. (1). Particle 3, though it succeeded in pene-
trating the sphere, did not proceed far enough to
Ni (ε, r)
ε
A
t (1) pass through the large cross-section area at right
angles, and hence it does not make any contribu-
gives the number of fast particles with energies in
tion to the value (1).
the range from ε to ε +
ε, which intersect at right
angles any region of the large cross-section of the The active spectrum, as well as the absorbed
sphere
A centered at the point r during the time dose, has a macroscopic value, unambiguously de-
interval from t to t +
t (see Fig.). Vectors 1–3 fined only for such
A and
t at which it is possi-
ble to neglect fluctuations of the value N(ε, r). For
a sufficiently large volume the Bragg–Grey ap-
proximation is valid, namely that the energy given
by the particle to the ionization and excitation of
atoms remains entirely within the volume. In this
case the Bragg–Grey correlation holds between
the absorbed dose of radiation D(r) and the active
spectrum:
D(r) = T N(ε, r)S(ε) dε, (2)
ε
ranges contribute to different radiation induced ADDITIONAL LIGHT WAVES, Pekar waves,
processes (ionization of different atomic shells, re- supplementary light waves
moval of an atom from its site, etc.). Thus, besides Waves arising from effects of spatial dielec-
the spectrum, it is necessary to know the cross- tric dispersion in the range of exciton resonances
sections of the processes σ (ε) as a function of the in crystals. In contrast to classical crystal optics,
incoming particle energy. It is possible to calcu- which implies that in a given direction s and at a
late the yield G of the radiation induced forma- frequency ω two mutually orthogonally polarized
tion of Frenkel defects within a crystal irradiated waves can propagate (birefringence), the appear-
by accelerated electrons, with the aid of the for- ance of additional light waves means that there
mula can be more than two such waves with the same s
and ω. In the case of isolated exciton transitions,
G(r) = N(ε, r)σF (ε) dε, (3) there are two waves with the same polarization
ε but with different complex refractive indices, i.e.
with different velocities and attenuation coeffi-
where σF (ε) is the cross-section of Frenkel defect cients. The number of additional light waves is
formation upon collision of fast incoming elec- equal to the number of excitons that are at reso-
trons of energy ε with crystal atoms. nance. The dispersion of additional light waves is
shown in Fig. for the cases of positive (a) and neg-
ACTIVITY of particles (quasi-particles) ative (b) exciton effective masses, where a dashed
A quantity that represents the effective concen- line indicates curves that do not take into account
tration of a solute in an ideal solution. It has the the spatial dispersion.
dimension of particle concentration, and is used In the presence of additional light waves the
in the equation of the law of mass action in cases Maxwell boundary conditions imposed on the
when the Boltzmann distribution does not apply. fields for reflection and transmission at the inter-
The classical expression for the relationship be- face between two media are insufficient for the
tween the particle concentration n and the chemi- determination of the amplitudes of all the waves.
cal potential μ is Therefore, it is necessary to introduce additional
boundary conditions that are obtained from the
μ = μ0 + k B T ln n, (1) law of exciton wave reflection from a crystal sur-
face [P ex (r, t)]z=0 = 0, where z = 0 is the sur-
where μ0 = const. This equation does not hold
face, and P ex (r, t) is the partial exciton contri-
in the presence of degeneracy (e.g., when Fermi–
bution to the polarization P (r, t). For excitons of
Dirac statistics or Bose–Einstein statistics apply),
large radius, due to the image forces in the vicinity
or when the interaction (electric, elastic, etc.) be- of the surface, there appears a “dead” layer with a
tween particles and quasi-particles leads to a redis- thickness of l ≈ 2–3 exciton radii into which the
tribution and spatial nonuniformity of their con- excitons cannot penetrate; and one should impose
centration, with a component of shielding. Under the additional boundary conditions on the surface
these conditions a quantity a called the activity z = l. There also exist other boundary conditions.
is introduced instead of n to retain the form of The additional light waves exist in a narrow
Eq. (1). The ratio A = a/n, called the activity co- spectral range in the vicinity of an exciton reso-
efficient, is a measure of the effectiveness of par- nance at low temperatures. When moving away
ticular particles (atoms, ions, etc.) in undergoing from the resonance toward higher or lower fre-
reactions or other actions. The concept of activity quencies a transition takes place to the classical
is widely used in physical and chemical research. theory of birefringence because the amplitudes of
In descriptions of non-equilibrium processes the the additional light waves go to zero. When the
form of Eq. (1) is sometimes retained, but with temperature is raised and the exciton–phonon in-
the chemical potential μ associated with the Fermi teraction constant Γ is increased, then a rearrange-
level replaced by a Fermi quasi-level counterpart ment of the dispersion curves takes place; and
that corresponds to a non-equilibrium concentra- when Γ reaches a certain critical value Γ cr they
tion n. acquire the classical form.
ADIABATIC APPROXIMATION 33
ADIABATIC APPROXIMATION
A specific method of approximate solution of
quantum mechanical problems, based on the dis-
tinction between “fast” and “slow” variables.
In solid state physics, this approximation is
used to separate the electronic (r) and nuclear
(ionic, R) variables in the Schrödinger equation.
A solid body is an electron–ion system bound to-
gether by strong interactions, in which the char-
acteristic frequencies of the electron subsystem
are much higher than those of the vibrating ions.
This makes it a good candidate for applying adi-
abatic perturbation theory. The basic parame-
ter that permits this application is the mass ratio
κ = (m/M)1/4 , where m is the electron mass and
Additional light waves.
M is the mass of an ion. The kinetic energy of the
The existence of additional light waves is sup- ions is κ 2 times less than the characteristic elec-
ported by many experiments. For example, two tron energy so it is neglected in the zeroth-order
equivalently polarized waves with different refrac- approximation, i.e. the Schrödinger equation is
tive indices were observed after the transmission solved for the electrons with the ions fixed in posi-
of light through a wedge-shaped crystal of CdS. tion at their equilibrium configuration. The result-
ing wave function n (r, R) and energy En (R)
of electrons in the state n depend on the coor-
ADHESION (fr. Lat. adhaesio , sticking)
dinates R of all the ions. At the next stage of
Coupling or attachment of two dissimilar bod-
the problem the ion movement is solved for; the
ies brought into contact.
motion, as a rule, being oscillations of the ions
In the case of bodies of the same physico-
around their equilibrium positions. In the adia-
chemical nature, the phenomenon is referred to batic approximation, the motions of the ions do
as cohesion. Adhesion is caused by interatomic not induce transitions in the electron subsystem,
(intermolecular) surface attraction, the nature of but there can be a non-dissipative transfer of en-
the latter being related to the deformation of the ergy between the electronic and ionic subsystems.
atomic electron shells at the interface of the bod- The adiabatic approximation makes it possibile to
ies. Various mechanisms can be operative, such as treat the electronic states of the system indepen-
polarization attraction via the van der Waals inter- dently, to a considerable extent, of the oscillations
action (see Van der Waals forces) with character- of the lattice, and hence, every subsystem makes
istic energies ε ≈ 10−3 –10−1 eV/atom, and the an additive contribution in the total energy. In this
interchange or sharing of valence electrons (chem- approximation, the oscillations of ions occur in
ical interaction, ε ≈ 1–10 eV/atom). High adhe- the field of an effective potential (adiabatic poten-
sion durability (several units of eV/atom) is ex- tial) which equals to the sum of the electron en-
hibited by contacts of metal with most (but not ergy En (R) and ion–ion interaction potential. The
all) other metals and semiconductors, as well as by adiabatic description assumes that terms higher
those of polymeric films with different materials. than κ 5 are neglected. Higher-order terms describe
The durability of contacts of metals with insula- interactions between the electron and ion systems
tors and graphite is much lower (10−1 eV/atom). (electron–phonon interaction) and higher-order an-
Adhesion plays an important role in some inter- harmonicity (phonon–phonon interaction).
face phenomena such as friction, wetting, and the The main adiabaticity criterion is that the elec-
heterogeneous growth of crystals. Adhesion finds tron spectra have no excitations with energies of
practical use in soldering, welding, and gluing. the order of nuclear oscillation frequencies: i.e.
that ωD |En − Em |, where En and Em are the
34 ADIABATIC DEMAGNETIZATION COOLING
energies of the free and occupied electron states, ADIABATIC FAST PASSAGE
and ωD is the characteristic nuclear vibrational Manner of scanning the applied magnetic field
frequency. This criterion is well satisfied in in- Bapp through the Larmor magnetic resonance con-
sulators and semiconductors due to the presence dition Bapp = B0 = ω0 /γ in a time subject to
of a band gap Eg in the electron spectrum. For two criteria: (1) the effective magnetic field vec-
this case the adiabaticity parameter is ωD /Eg . tor direction, B e = B 0 − B app + B 1 , changes so
This criterion can be violated in narrow-gap semi- slowly that the magnetization vector, M, remains
conductors. In metals, the excitation energy is of approximately parallel to B e ; and (2) the passage
the order of the Fermi energy EF (see Fermi en- through the resonance is so rapid that the relax-
ergy), so for electrons far below the Fermi sur- ation processes have no time to deviate M from
face the criterion is satisfied due to the smallness the direction of B e . The mathematical statement
of the ratio ωD /EF . The delocalized electrons in of these two criteria in solids is as follows:
the energy range ∼ωD near the Fermi surface are
1 1 dB app
non-adiabatic. The relative smallness of the vol- |γ B 1 |,
ume occupied by such states permits the use of the T1 B dt
1
adiabatic approximation for bulk characteristics of where B 1 is the alternating magnetic field vector
a metal such as the binding energy and the phonon which oscillates at the fixed frequency ω0 ; B 0 is
spectrum. the resonant field, γ is the gyromagnetic ratio, and
T1 is the spin–lattice relaxation time.
ADIABATIC DEMAGNETIZATION COOLING,
magnetic cooling
ADIABATIC PARAELECTRIC COOLING
Decrease of crystal lattice temperature, T , aris-
Decrease of the temperature of a dielectric (in-
ing from demagnetization without heat exchange
sulator) via a decreasing electric field due to adi-
with the surrounding medium of paramagnetic
abatic depolarization; the electric analog of adia-
particles (spins) contained in the crystal lattice.
batic demagnetization of a spin system (see Adia-
This was predicted independently by P. De-
batic demagnetization cooling).
bye and W. Giauque (1926). The first experiments
Cooling by adiabatic depolarization occurs in
were carried out in 1933–1934. It is used to ob-
systems with impurity electric dipoles, such as
tain temperatures considerably lower than 1 K
(e.g., ∼10−3 K). Transition metal ions of the iron KCl:OH− , KCl:Li+ , RbCl:CN− . In these sys-
group or rare-earth elements can be used as para- tems, the temperature was observed to decrease
magnetic particles, as well as other nuclei with to 4 mK. Systems of reorientable electric dipoles
non-zero magnetic moments. The cooling is per- having large dipole moments and a small inversion
formed in several stages: (1) the sample is pre- splitting are favorable for obtaining low tempera-
cooled to an initial moderately low temperature tures. The measurement of the temperature change
(∼1 K) by other methods; (2) at these tempera- (ΔT ) dependence on the electric field strength and
tures, the sample is magnetized isothermally in a on the initial temperature elucidates some basic
field B ∼ 1–10 Tesla; (3) then the sample is adi- characteristics of these dipoles, such as the impu-
abatically demagnetized and its temperature falls rity concentration and the magnitude of the dipole
rapidly far below 1 K. The physical reason for this moment, as well as the inversion splitting of the
decrease is associated with the transfer of entropy impurity energy levels.
from the magnetic subsystem to the crystal lattice
(and vice versa). At stage 2 the spin alignment ADIABATIC POTENTIAL
of the magnetic subsystem occurs, diminishing its The energy of a many-atom system as a func-
entropy and increasing the crystal lattice subsys- tion of nuclear coordinates, obtained in the so-
tem entropy. Since step 2 is isothermal there is no lution of the electronic part of the Schrödinger
change in the sample temperature. Since stage 3 equation with the nuclei fixed in position. If the
is isentropic the demagnetization is accompanied criterion for the validity of the adiabatic approxi-
by withdrawal of the heat from the lattice, and the mation is met, the adiabatic potential has the sense
sample temperature decreases. of the potential energy of the nuclei in the field
ADSORPTION 35
The compounds that are formed are not necessar- and r is the distance of the particle from the sur-
ily stoichiometric. Their components are atoms or face. A more general description of the absorp-
individual radicals of molecules dissociated dur- tion potential is obtained by summing over over
ing the process of adsorption. Sorption is a more the contributions of each pairwise interaction of the
general concept which includes adsorption and ab- particle being adsorbed with each atom of adsor-
sorption (see Activated adsorption). bent. This potential depends not only on the dis-
tance between the particle and the surface, but also
ADSORPTION, ACTIVATED on its particular position relative to the substrate
See Activated adsorption. atoms.
AIRY STRESS FUNCTION (G.B. Airy, 1863) temperature measuring by the radiometer method,
A function used to express the solution of the and in neutron diffusion theory.
problem of planar strain of a homogeneous, linear
elastic solid. Planar strain is the strain at which ALFVEN WAVES
one of the components of the displacement vec- See Solid-state plasma.
tor u, called uz , is equal to zero, and the other two
components ux and uy do not depend on z. In this ALGEBRA, COMPUTER
case the equations for the equilibrium of the elastic See Computer algebra.
body can be written in the form
∂σxx ∂σxy ALKALI-HALIDE CRYSTALS
+ + ρX = 0, Ionic crystals of compounds of alkali metals
∂x ∂y
and halogens. They crystallize mainly in the NaCl-
∂σyy ∂σxy type FCC crystal lattice (except CsCl, CsBr, CsI
+ + ρY = 0,
∂y ∂x which adopt the CsCl BCC structure). NaCl is a
typical representative of alkali halides: the ionic-
where σij are the components of the stress ten-
ity is 0.94; interionic spacing is 0.282 nm; bind-
sor, ρ is the density, X and Y are components
ing energy per pair of ions is 7.93 eV; melting
of the bulk force density vector. If this vector can
temperature 1074 K; Debye temperature 281 K;
be derived from a potential V , i.e. X = −∂V /∂x,
speci c heat 48.86 J·mol−1 ·K−1 (250 K); ther-
Y = −∂V /∂y, then there exists a function χ called
mal conductivity 0.07 W·cm−1 ·K−1 (273 K); sta-
the Airy stress function, in terms of which one can
tic dielectric constant is 5.9, optical dielectric con-
express the components of the stress tensor that
stant 2.25. Alkali halides are one of the most
are the solutions of the problem at hand:
important model systems in solid state physics.
∂2χ During investigations of these crystals many fun-
σxx = + ρV , damental phenomena were discovered, such as
∂x 2
self-localization of electron excitations, the exis-
∂2χ tence of free and self-localized states of quasi-
σyy = + ρV ,
∂y 2 particles, the decay of electron excitations per pair
of defects, light absorption and luminescence, re-
∂2χ
σxy = − . lations of intraband transitions in insulators.
∂x∂y
The valence band (width 2–5 eV) develops
Under certain conditions the Airy stress function from p-states of the halogen, the conduction band
is biharmonic. (somewhat wider) arises from the empty s-state of
the alkali metal. The broad band gap (6–14 eV)
ALBEDO causes the transparency in the visible and UV
The reflecting ability of a matte surface. The spectral regions. In the IR range the transparency
plane (Lambert) albedo is the ratio of the light region is limited by the lattice absorption (excita-
flux scattered by a flat element of surface to the tion of optical phonons ν ∼ 1013 Hz).
incident flux. This is measured by an albedometer. In the UV region the transparency limit is de-
The apparent albedo is the ratio of the brightness termined by the exciton absorption bands situated
of a flat surface element illuminated by a beam of on the long-wave edge of the fundamental absorp-
parallel rays to that of an absolutely white sur- tion. These bands involve the excitation of the
face oriented normal to the incident beam. The halogen s-state to the p-state (Frenkel exciton). In
albedo for monochromatic radiation is called spec- the transparent region absorption bands caused by
tral albedo. A similar physical quantity in neutron structural or impurity defects (color centers) may
optics, neutron albedo, is the probability of the be observed.
neutron reflection that results from multiple scat- A strong electron–phonon interaction results
tering in the material. The concept of albedo is in a sharp difference between free and relaxed
used, e.g., in optical engineering calculations, in quasi-particle state properties. Effective masses
ALKALI METALS 39
are: for the electron band m∗e ≈ 0.5me , and for tron and hole migration (recombination lumines-
the polaron m∗p ≈ me . The drift mobility cence), or by exciton migration. At T 200 K the
is about 104 cm2 ·V−1 ·s−1 at 5 K, and about ionic conductivity arises mainly from the motion of
10 cm2 ·V−1 ·s−1 at 300 K. For holes and excitons cation vacancies, while for T 300 K it is due to
relaxation leads to the self-localization of quasi- the motion of anion vacancies.
particles in the regular lattice (holes as a molec- The main technical applications of alkali
ular halogen ion X− 2 , exciton as an excited quasi-
halides are for optical materials, for the detectors
∗ of nuclear radiations (scintillators NaI doped with
molecule (X2− 2 ) ). Tl, CsI with Tl, CsI with Na, and thermolumines-
Holes travel a distance of about 102 –103 nm cence dosimeters based on LiF), and lasers based
in a time of about 10−11 s before their self- on color centers.
localization. For the self-localization of an exciton
the activation energy of about 10 meV is required, ALKALI METALS, alkali elements
therefore, the lifetime of a free exciton and its shift Chemical elements of Group I of the periodic
depend on the temperature. With the increase in table (Li, Na, K, Rb, Cs, radioactive Fr). Owing
temperature the thermally activated motion of self- to the high radioactivity of Fr its properties have
localized holes and excitons is observed, with ac- been rarely examined. The distribution of Na and
tivation energies 0.18–0.54 eV for holes and ap- K in the Earth’s crust is rather high (2.64 and 2.5
proximately half that value for excitons. It is likely mass %); the remaining alkali metals are rare and
that for excitons this motion represents a series of widely dispersed. All alkali metals have low elec-
reorientation jumps of (X2− ∗ tronegativities, with electronegativity values rang-
2 ) , terminating in a
radiative or nonradiative annihilation, or capture ing from 0.96 (Li) to 0.69 (Fr). The outer elec-
by a defect. For holes there is a combination of re- tronic shell of an alkali atom contains a single
orientation jumps of X− s-state electron (configuration s 1 ), with a stable
2 and “great leaps” for a inert gas shell (s 2 or s 2 p6 ) lying below it. The
distance about 100 nm. This hopping, which pro-
atomic radius expands from Li (0.155 nm) to Fr
vides the major part of the contribution to the hole
(0.280 nm). At the same time, the number of elec-
motion, is due to a thermally activated transition
tron shells increases, resulting in a decrease of the
of the Vk -center in a hole band, or a hole polaron
electrostatic attraction force of the outer valence
of large radius.
electron to the nucleus, and a decrease of the ion-
The recombination of an electron with a self-
ization energy from 5.392 eV for Li to 3.894 eV
localized hole, or a nonradiative exciton annihila-
for Cs. This is why alkali metals easily give up
tion, takes place with a high probability of the for- their valence electron and transform into ions M+ .
mation of a Frenkel defect pair in the anion sub- Alkali metals are light metals: density in-
lattice. This is responsible for the low radiation re- creases from 0.539 g/cm3 (Li) to 1.9039 g/cm3
sistance of alkali halides. Luminescence of alkali (Cs). They melt easily: Tmelting changes from
halides is related to both radiative exciton annihi- 453.69 K (Li) to 301.75 K (Cs). These metals
lation (free and self-localized exciton) and radia- have the ability to color a colorless flame: Li into
tive transitions in local centers (luminescent cen- red, Na into yellow, K and Cs into violet, Rb into
ter). An impurity ion or impurity atom (e.g., Ag+ , a purple-red color. Alkali metals are similar in
In+ , Tl+ , Pb2+ ), or a structural defect, is typically their physical and chemical properties, all being
a nucleus of such centers. very active chemically. They interact with oxygen
The luminescence of local centers in alkali and water at room temperature. Hydroxides of al-
halides is observed both at the photoexcitation of kali metals with the general formula MOH are all
absorption bands of these centers (the intra-center strong bases, and are highly water soluble. They
luminescence), at the excitation in a region of fun- are often referred to as caustic alkalis. Most alkali
damental absorption, and by interactions involv- metals salts are very water soluble, exceptions be-
ing incident high-energy particles and quanta. In ing the Li compounds LiF, Li2 CO3 , Li3 CO4 , and
the latter case the energy transfer from the lattice also perchlorate salts MClO4 , where M = Li, Na,
to the luminescence centers is provided by elec- K, Rb, Cs, Fr, NH4 .
40 ALLOTROPY
Uses of alkali metals in the free state are value unity as well, that is they also become al-
severely limited: Na and K are used as thermo- lowed, e.g., ΔM = ±2, ±3, . . . or ΔM = 0 (inte-
carriers; Cs and other alkali metals as getters in ger spin and B 1 B 0 ). When spins interact with
vacuum equipment, etc. each other, transitions involving a change in the
projection of just one of the two are considered al-
ALLOTROPY (fr. Gr. αλλoς , other, and τρoπη , lowed transitions as well: ΔM1 = ±1, ΔM2 = 0,
turn) Δn = 1 or ΔM1 = 0, ΔM2 = ±1, Δn = 1.
The ability of a chemical element to exist as
two or several distinct forms (allotropic forms) ALLOY AGING
having different physical and chemical properties. Change of crystal structure and physical prop-
The allotropy can be caused by various processes erties of alloys as a result of the decomposi-
such as the formation of molecules having differ- tion of a supersaturated solid solution (see Alloy
ent numbers of atoms (e.g., oxygen O2 and ozone decomposition), which is generated when an al-
O3 ), crystals of a different structure (e.g., graphite, loy undergoes rapid cooling (quenching) from the
diamond, carbyne, fullerene forms of carbon), or single-phase region of the phase diagram to the
high pressure (diamond). In the case of carbon the two-phase (low-temperature) one. The process of
allotropy is a kind of polymorphism. alloy aging takes place when the mutual solubility
of alloy components in each another decreases be-
ALLOWED SPECTRAL LINES cause of a lowered temperature. The resulting su-
Spectrum of lines of allowed transitions, i.e. persaturated solid solution is in a metastable state
those which satisfy quantum-mechanical selection and attains equilibrium, i.e. reaches a state char-
rules, and have intensities that far exceed the in- acterized by a minimum of free energy (see Ther-
tensities of forbidden quantum transitions. modynamic potentials), by precipitating the excess
amount of alloying element in the form of second-
ALLOWED TRANSITIONS phase particles (see Alloying). This process is re-
Transitions which are permitted by the selec- ferred to as natural aging if it occurs at room
tion rules of quantum mechanics. The dimen- temperature, as distinguished from the artificial
sionless squared magnitude of the matrix element aging process, which requires heating of the hard-
of the periodic perturbation that induces transi- ened alloy. Natural alloy aging is common to alu-
tions between particular initial and final states pre- minum alloys (e.g., Al–Cu, Al–Zn, Al–Ag, etc.),
scribes the probability of such a transition to oc- whereas the artificial variety is characteristic of al-
cur, and for an allowed transition this matrix el- loys featuring higher heat resistance (e.g., Cu–Be,
ement is close to unity. The relative intensity of Cu–Ti, Ni–Be, Co–Be, Co–Ti, Co–Cu, Ni–Al,
transitions in magnetic resonance is proportional Ni–Ti, Fe–Be, Fe–Cu, etc.).
to |ini |gμB J |f nf |2 B 21 , where B 1 is an alter- Alloy aging is a rst-order phase transition in-
nating microwave or radio-frequency field, g is volving nucleation of a new phase followed by
the dimensionless g-factor, μB is the Bohr mag- growth of the nucleus. The late stages of alloy ag-
neton (the nuclear magneton is used for NMR), ing are characterized by the coalescence process,
J is the spin, ni,f are the numbers of photons, i with larger particles growing at the expense of
and f are the quantum numbers of the spin sys- smaller ones. The formation of the new phase nu-
tem initial and final states. If one denotes the states cleus calls for overcoming a potential barrier due
of an isolated spin by the value of its projection to the generation of a new volume which exhibits a
M along the applied magnetic field B 0 , then the new crystal structure (different from original ma-
transitions with ΔM = ±1, n = 1 are allowed in trix), composition and phase boundary. When the
strong magnetic fields for B 1 ⊥B 0 . For very weak matrix and precipitated phase have different spe-
applied magnetic fields (e.g., weaker than a crys- cific volumes, an additional contribution to the
tal field term that causes magnetic quantum num- new volume is made by the elastic energy, which
ber splittings) the intensities of other transitions is due either to elastic coherent stresses arising
(forbidden in high fields) sometimes approach the in the case of a coherent interface, or to fields
ALLOY BRITTLENESS 41
of stresses due to interphase dislocations in the other intermediate state is determined by the ki-
case of an incoherent interface (see Coherent pre- netics of alloy aging (i.e. depends on the activation
cipitates). Two decomposition types are distin- energy of the process). Aging is accompanied by
guished. If the structures of the matrix and of the considerable changes in alloy properties: increase
precipitated phase have much in common, then the of strength and high-temperature strength, coer-
process is called isomorphic decomposition (see cive force, magnetic energy, electrical conductivity,
Isomorphism), otherwise it is referred to as non- etc. It is for this reason that alloy aging is used in
isomorphic decomposition. The potential barrier industry for the purposes of upgrading alloy char-
may be substantially lowered if the nucleus of the acteristics.
new phase arises at grain boundaries, dislocations,
stacking faults, i.e. a discontinuous-type heteroge- ALLOY BRITTLENESS
neous nucleation takes place. In the event that the High susceptibility to brittle failure, appearing
nuclei are formed at point defects, the process is at the alloying of metals, and at the formation of
called “continuous homogeneous nucleation”. An- alloys. The brittleness of alloys appears due to:
other mechanism of alloy aging is spinodal de- (1) Hardening, introduced by atoms of the doping
composition. element. The increase of flow stress σS within
The form and nature of the spatial arrangement the solid solution limits results in a decrease
of the precipitated particles in an aged alloy are de- of plastic characteristics (elongation δ, trans-
termined by the following factors: type of decom- verse narrowing ψ ), and in an increase of the
position (continuous or discontinuous), magnitude cold brittleness temperature Tb . Alloys where
of coherent interphase strains, value of the dif- a rhenium effect occurs are an exception.
ference between the elastic moduli of the phases, (2) Dispersion hardening (see Precipitation-hard-
type of anisotropy of elasticity exhibited by the ened materials). When phases of refractory
matrix, volume fraction of precipitates in the al- compounds are introduced into the alloy as
loy. Provided a certain balance is reached between dispersed particles, then the mean free path
the above parameters (elastically isotropic matrix, of dislocations is restricted. Also this favors
high coherent strains, and volume fraction of pre- their conglomeration at barriers, leads to an
cipitates), the particles of the coherent phases in an increase of σS , and to a decrease of δ and ψ ,
alloy can achieve a regular distribution in the form thereby inducing an increase of Tb .
of a macrolattice, forming a so-called modulated (3) Formation of intermetallic brittle phases (see
structure during the process of coalescence. The Intermetallic compounds) (σ -, χ -phases, etc.).
shape of precipitate particles depends on the co- The reasons are the same as in case of dis-
herent strains and on the ratio between the elastic perse hardening, but they are compounded by
moduli of the matrix and of the precipitated phase. the primary release of these phases along grain
There exist spherical, laminar-shaped, and needle- boundaries.
shaped precipitate particles. (4) Decay of solid solutions oversaturated by
The process of aging does not ordinarily in- interstitial impurities (dispersion hardening).
volve the immediate formation of the precipi- The reasons are the same as in two previous
tated phase: the alloy passes through a sequence cases (see Alloy decomposition).
of intermediate metastable states which corre- (5) Solid solution decay of the stratification type
spond to certain free energy minima. At the ini- (see Strati cation-type failure). At the decay
tial stages (pre-precipitation), either zones en- extremely small (of the order of nanometers)
riched with the alloying element (Guinier–Preston solid solution regions based on each alloy
zones), or dispersed particles of the coherent in- component are formed (e.g., in the Cr–Fe sys-
termediate phase are formed. The dimensions of tem: α-phase based on Fe, and α -phase based
the latter range from 0.5 to 1.0 nm. At subsequent on Cr). The presence of fine-grained phases
stages their structure and composition approxi- presents difficulties for the formation and mo-
mate the equilibrium values of the equilibrium- tion of dislocations, and thereby decreases δ
precipitated phase. The realization of one or an- and ψ and increases σS and Tb .
42 ALLOY DECOMPOSITION
element, rather than the element in its pure form. the relative concentrations of atoms A and B may
Alloying elements are obtained by melting to- be introduced: cA = NA /N , cB = NB /N , where
gether the necessary components, or by reducing N = NA + NB = N 1 + N 2 . The long-range or-
ores, concentrates and oxides. Alloying elements der is characterized by the long-range order para-
find their widest application in ferrous metallurgy, meter η (see Order parameter), which is given by
(1) (1)
mainly for the modi cation and doping of steel and the formula η = (pA − cA )/(1 − ν), where pA
cast iron. is the a priori probability of site 1 being occupied
by atom A, and ν = N 1 /N . The name “disordered
ALLOY ORDERING alloy” is commonly given to alloys that exhibit
The process of redistribution of atoms in al- no long-range order (short-range order may exist,
(1)
loys, whereby atoms of each type begin to oc- however), i.e. pA = cA and η = 0. For a com-
cupy their individual crystal lattice sites so long- pletely ordered alloy the stoichiometric composi-
range order becomes established (see Long-range tion is cA = ν, so the long-range order parameter
and short-range order). As this takes place, the lo- η = 1. Also other types of long-range order para-
cal increase (or reduction) of the concentration of meters (as well as short-range order ones) may be
atoms in the coordination spheres around a cen- introduced.
tral (arbitrarily chosen) also atom takes place; this The most direct and widely used methods
local change of concentration corresponds to the for experimentally investigating ordered struc-
presence of short-range order along with the long- tures (superlattices) in alloys are radiographic (see
range order (see Correlation in alloys). The phe- X-radiography) and neutron diffraction (see Neu-
nomenon of alloy ordering is of common occur- tron diffractometry) methods, which allow the de-
rence; it is observed in both substitutional alloys termination of the arrangement of atoms in var-
and interstitial alloys (vacancies at the sites and ious superlattice structures (Fig. 1). It has been
interstices of the lattice may also be involved), found that η decreases with increased tempera-
in oxides, in two-dimensional structures consist- ture. For many alloys, η becomes equal to zero
ing of atoms adsorbed on the crystal surface, in at a certain absolute temperature T = T0 (order-
ferromagnetic (see Magnetic materials) and anti- ing temperature) below the melting temperature.
ferromagnetic metals (where differently oriented Here T0 is called the order–disorder phase tran-
atomic spin moments become ordered), etc. Con- sition temperature, which may occur at either a
sider the simplest case of a binary alloy consist- second-order phase transition (e.g., in alloys like
ing of NA atoms of type A and NB atoms of β-brass) or a rst-order phase transition (in alloys
type B that are distributed over respectively N 1 of AuCu3 and AuCu type). The temperature de-
positions of type 1 and N 2 positions of type 2 (po- pendence of the long-range order parameter for
sitions of types 1 and 2 being, respectively, the these cases is given in Fig. 2. The value of the
sites and interstices of the lattice). In this case, ordering temperature T0 depends upon the alloy
Fig. 1. Distribution of atoms A (open circles) and B (closed circles) in perfectly ordered alloys with three common
structures of the types (a) β-brass, (b) AuCu3 and (c) AuCu.
44 ALLOY ORDERING
composition. The short-range order does not die obtained for one- and two-dimensional crystal
out at T = T 0 , but reduces gradually with increas- lattices (Onsager method). The difficulties asso-
ing T . Certain alloys (e.g., Fe–Al) exhibit succes- ciated with the three-dimensional problem neces-
sive formations of various superlattice structures sitate the application of approximate methods, of
at various temperatures; i.e. these alloys feature which the simplest is the Gorsky–Bragg–Williams
not only order-disorder transitions between super- approximation (V.S. Gorsky, 1928; W.L. Bragg,
lattice structures, but also transitions of the order- E.J. Williams, 1934–1935), which completely ig-
order type. Various phase transitions involving a nores correlation in alloys. The application of this
change of atomic order have been found in the method gave rise to the theory of ordering in
subsystem of interstitial atoms in interstitial alloys binary A–B alloys with two equivalent atom po-
(e.g., in metal–hydrogen systems). sitions. In the context of this theory the equi-
The theory of atomic ordering is ordinarily librium equations for determining the T - and
based on one of two approaches: the thermody- cA -dependences of the equilibrium value of η
namic approach or the statistical one. The ther- have been derived. The general forms of these
modynamic theory treats alloy ordering in the equations, which are applicable to all types of
neighborhood of the T0 temperature on the ba- structures, are given by
sis of the general Landau theory of second-order
1 ∂E
phase transitions. Another thermodynamic ap- kB T Z = ,
proach (E.M. Lifshits, 1941, 1944) consists in ana- ν(1 − ν)N ∂η
lyzing the symmetry of the atomic probability den- [cA + (1 − ν)η](1 − cA + νη)
sity function using group theory methods, which Z = ln ,
[1 − cA − (1 − ν)η](cA − νη)
allows the determination of structures prone to
second-order transitions. where E is the pairwise interaction energy of an
The statistical theory of alloy ordering is based atomic system. The η(T ) dependences are pre-
on employing simplified models of alloy atomic sented in Fig. 2, and the corresponding phase tran-
structure. It provides a means of determining the sitions follow from these expressions. One can
equilibrium values of order parameters in alloys also obtain the T 0 (cA ) dependence, which takes
over the whole range of variation of these pa- the form given in Fig. 3 in the case of alloys of
rameters. The commonly used model is based the β-brass type. Generalization of the alloy or-
on the pairwise interactions of atoms in the con- dering theory to alloys with inequivalent sites, and
text of the configuration approximation. Even to interstitial alloys with inequivalent interstices,
when the only interatomic interactions consid- brings the equilibrium equation to the following
ered are those between nearest neighbor atoms form: k B T Z = R 1 η + R 2 , where R 1 and R 2 are
(Ising model), the determination of thermody- constants, and R 2 = 0 for the case of equivalent
namic functions remains a complicated task. The positions of atoms. The latter equation suggests
exact solution of this statistical problem has been the occurrence of order–order phase transitions,
ALLOYS 45
which do not involve the generation of new su- Alloy ordering strongly affects the energy spec-
perlattice structures. The alloy ordering theory is trum of electrons (theoretically, there is a cer-
extended also to cases when the type of the result- tain probability of generating a gap in the elec-
ing superlattice structures is not known in advance tron energy spectrum (see Band gap) which may
(method of branch points, and method of static in principle induce metal–insulator transitions),
concentration of waves); it allows the prediction therefore, alloy ordering exerts a strong influence
of the types of superlattice structures arising from on galvanomagnetic effects. The Hall constant of
the disordered state of an alloy of a given struc- the alloy AuCu3 even exhibits a change of sign.
ture. The theory of alloy ordering under pressure Alloy ordering has a pronounced effect on mag-
is developed by taking into account the relation- netic properties (e.g., the alloy Ni3 Mn is weakly
ship existing between the energies associated with magnetic in the disordered state, and becomes a
the interatomic interaction and the interatomic dis- ferromagnet in the ordered state).
tance.
Allowance for correlation in alloy ordering
ALLOYS
theory has been made using various approaches,
Macroscopically homogeneous many-com-
of which the Kirkwood approximation (J.G. Kirk-
ponent systems. The main method of alloy produc-
wood, 1938) and the quasi-chemical approxima-
tion is crystallization from the melt. There are also
tion should be mentioned. The former is based
other techniques of alloying: sintering, diffusion of
on expanding the free energy in a power series
of w/(k B T ), where w is the ordering energy. The the alloying elements into a solid; accumulative
latter approximation treats groups of neighboring crystallization of vaporized components; electro-
atoms as quasi-molecules. Allowance made for plating; ion implantation, etc. In certain cases the
correlation results in the T0 (cA ) curves being con- concentrations of the components are in a strictly
tracted to the medium region of the phase dia- defined ratio, i.e. the alloy is a chemical com-
gram, in the increase of the η0 jump at first-order pound (with only slight deviations from stoichio-
transitions, in a more correct determination of the metric composition); e.g., brass structures of the β
phase transition type in alloys of the AuCu type, (CuZn), γ (Cu5 Zn8 ) and ε (CuZn3 ) types, or com-
and in other refinements of the theory. pounds AuCu3 , AuCu, etc. The alloy name is of-
In a theory of alloy ordering, consideration ten determined by the name of its major com-
must be given to the electronic subsystem of the ponent (copper alloys, titanium alloys, aluminum
crystal. Alloy ordering has a pronounced effect alloys, etc.). Certain groups of alloys go under
on the mechanical properties of alloys, strongly conventional names: cast irons and steels (Fe–C),
affecting the elastic moduli, critical shear stress, brasses (Cu–Zn), bronzes (remaining Cu-based al-
creep, high-temperature strength (e.g., the hard- loys). Alloys are divided into single-phase (homo-
ness of the alloy CuPt more than doubles on or- geneous) and multiphase (heterogeneous) types.
dering). Alloy ordering is responsible for abrupt The following individual alloy phases are distin-
changes in atomic diffusion coef cients and the guished: solid solutions formed when atoms or
specific electrical resistance (see Resistivity of or- ions of components share a common lattice (char-
dered alloys), as well as for profound changes acteristic of one of the components) when mixed
in thermal properties (characteristic anomalies of in arbitrary proportions; intermetallic compounds
speci c heat in the neighborhood of T 0 , abrupt characterized by certain relations among the con-
changes of alloy energy during the course of first- stituent elements; crystal lattices of intermetal-
order transitions). The assumption of the existence lic compounds that differ from those of the con-
of ordered structures is based on the observation stituent elements. Alloys are grouped according
of minima on the plots of ρ(cA ) for the Au–Cu to common inherent features: nonferrous metal
alloys. Quantum perturbation theory (with no re- alloys, refractory metal alloys, ferrous metal al-
gard for correlation) provides the dependence of loys, semiconducting alloys, etc. The most widely
residual resistivity on cA and η under conditions used types are metal alloys which are divided into
of alloy ordering:
casting alloys and wrought alloys depending on
ρ0 ∼ cA (1 − cA ) − ν(1 − ν)η2 . the technique of production of semifinished items
46 ALLOYS OF METALS
or complete articles. In terms of applications, al- Magnesium alloys, Molybdenum alloys, Nickel al-
loys are classified into constructional alloys, tool loys, Noble metal alloys, Rhenium-bearing alloys,
alloys, alloys with special electrical properties, Tantalum alloys, Titanium alloys, Tungsten alloys,
solders, etc. In terms of the principal operating Heusler alloys, Mishima alloys.
characteristic, high-temperature alloys, thermal-
environment resistant alloys, wear-resistant (see ALLOY STRATIFICATION
Wear) ones, corrosion-resistant (see Corrosion re- A phase transition consisting in isostructural
sistance) alloys, extra-hard alloys, etc., are distin- decay of a supersaturated solid solution (see Al-
guished. The large group of alloys with special loy decomposition) into two or perhaps three so-
characteristics due to their definite chemical com- lutions with differing concentrations of doping
position, lack of harmful (exerting strong adverse elements. Experiments have demonstrated alloy
effect on properties) impurities, are called pre- stratification in systems exhibiting immiscibil-
cision alloys. This group includes soft magnetic ity regions in their solid states (such as Au–Pt,
alloys and hard magnetic alloys (see Soft mag- Au–Ni, Cu–Ni–Fe). A spinodal mechanism of the
netic materials, Hard magnetic materials), alloys phenomenon is suggested (see Spinodal decompo-
with specified electric, thermal, elastic properties sition).
(including alloys exhibiting particular temperature
dependences of these properties), as well as super- ALLOY STRUCTURE AND PROPERTIES
conducting and bimetallic alloys. High purity of For structure and properties of different kinds
alloys is achieved by using components of extra- of alloys, see Amorphous metals and metallic
purity, and performing the smelting in vacuo, inert, alloys, Binary alloys, Eutectic alloys, Eutectoid
or refining (hydrogen-containing) media, by appli- alloys, High-coercivity alloys, Interstitial alloys,
cation of zone re nement, electroslag remelting, Monotectic alloys, Peritectic alloys, Superconduct-
vacuum refining, methods of floating zone melt- ing alloys, High-temperature alloys, Substitutional
ing. In order to increase the anisotropy of prop- alloys, Thermal-environment resistant alloys, Fer-
erties, directional crystallization and thermomag- romagnetism of metals and alloys, Radiation phys-
netic treatment are used. The finishing stage of ics of metals and alloys, Failure of amorphous
the development of alloy properties is heat treat- metallic alloys, Order parameters in alloys, Resis-
ment, which facilitates eliminating the nonuni- tivity of ordered alloys.
form distribution of alloying components within
the boundaries of individual grains, establishing ALNI (fr. aluminum and nickel)
the needed dislocation structure (see Dislocations), Hard magnetic materials of the system
phase composition, distribution of atoms within Fe–Ni–Al with Cu commonly added. The opti-
separate phases (see also Alloy ordering, Alloy ag- mal composition is close to Fe2 NiAl. The prod-
ing, Order parameters in alloys). The thermody- ucts from these alloys are manufactured by casting
namic description of alloys usually involves the and grinding. A high-coercivity state results from
temperature T and composition (molar ratios of the decomposition of an oversaturated solid solu-
components) as independent state variables. The tion. Properties of some alni alloys are given under
characteristic functions are entropy S, enthalpy H , the entry Hard magnetic materials.
and Gibbs free energy G = U − T S + pV (U is
internal energy, V is volume) (see Thermodynamic ALNICO (fr. aluminum, nickel and cobalt)
potentials). The components of alloys are de- Hard magnetic materials based on the system
scribed using partial molar derivatives, e.g., chem- Fe–Co–Ni–Al. They are commonly used with the
ical potential μ = ∂G/∂ni . addition of up to 4% Cu. They are similar in
many aspects to alni alloys. Alloys with contents
ALLOYS OF METALS of Co 18 to 25% are sensitive to a thermomagnetic
For different kinds see Aluminum alloys, Beryl- treatment resulting in the formation of anisotropic
lium alloys, Chromium alloys, Copper alloys, magnetic materials. Their properties are given in
Germanium–silicon alloys, Iron alloys, Lead alloys, the table under the entry Hard magnetic materials.
ALUMINUM ALLOYS 47
ALPHA PHASE (α -phase) (introduced by most abundant metal in the earth’s crust. It can-
F. Osmond, 1885) not occur free in nature due to its high chem-
A phase on the basis of the low-temperature ical activity. The crystal lattice of aluminum is
modification of a polymorphous chemical element face centered cubic, with lattice parameter a =
(see Polymorphism). The term alpha phase was in- 0.4049 nm. Its density is 2.70 g/cm3 (at 293 K),
troduced to designate the low-temperature modi- Tmelting = 933 K, Tboiling = 2793 K. Heat of
fication of iron and related solid solutions; later it melting is 10.55 kJ/mole; heat of vaporization is
was extended to other polymorphous elements. In 291 kJ/mole, specific heat is 25.1 J·kg−1 ·K−1
binary systems, the alpha phase indicates the solid (at 273 K). Debye temperature is 663 K; thermal
solution of the second component in the substrate linear expansion coefficient is 24.56·10−6 K−1
metal. The temperature and concentration regions from 293 to 473 K. Specific thermal conduc-
of the alpha phase are determined by the ther- tivity is 225 W·m−1 ·K−1 , electric resistivity is
modynamic potentials of its initial and concurrent 2.66 μ·cm (at 273 K). The superconducting tran-
phases. If the alpha phase exists over wide con- sition temperature is 1.2 K. Aluminum is weakly
centration limits, then alloy ordering, alloy strati - paramagnetic. Its elastic modulus is 68.6 GPa,
cation, and magnetic transformations may occur at Brinell hardness for calcined aluminum is
its cooling. 167 MPa. Aluminum is easy to process by press-
ing, rolling, forging, stamping and drawing. Its
ALUMINOSILICATES ultimate tensile strength is 78.5 to 98.1 MPa,
A group of minerals: various salts of silicic acid yield limit upon stretching is 29.4 MPa, its rigid-
where aluminum occupies the same position in the ity is 25.5 GPa. After cold rolling, its ultimate
crystal structure as silicon. Also present are an- strength grows to 177 to 245 MPa. Adiabatic elas-
ion complexes, namely SiO4 tetrahedra, and also tic moduli of the aluminum monocrystal: c11 =
AlO4 . The number of Si4+ ions replaced by Al3+ 106.43; c12 = 60.35; c44 = 28.21 GPa (at 300 K).
ions does not exceed 50%. The ratios Si : Al equal Pure aluminum has a high reflection factor which
3 : 1 and 1 : 1 are found. The ratio (Si + Al) : O in makes it suitable for the manufacture of reflec-
anion radicals is also 1 : 2. Aluminosilicates can tors. Aluminum and aluminum alloys are used as
have chain, laminar or framework structures. In construction materials in machinery, buildings, air-
minerals with the chain structure, Al can replace craft, civil engineering, etc.
up to 1/4 of the Si, and in the laminar and three
dimensional framework ones – up to 1/2 of Si. In ALUMINUM ALLOYS
this case, the charge balance is restored by the in- Alloys based on aluminum.
troduction of ions of monovalent (Na, K) or di- They are characterized by low density, as well
valent (Ca, Ba) elements into the structure. The as high corrosion resistance, thermal conductivity,
most common aluminosilicates are feldspar, zeo- electrical conductivity and specific strength. There
lites, scapolite, micas, chlorites, and others. are deformable and cast aluminum alloys.The de-
formable alloys possess high plasticity, they can be
ALUMINUM, Al welded (see Welding), and machined, do not be-
A chemical element of Group III of the peri- come brittle (see Embrittlement) at low tempera-
odic table with atomic number 13 and atomic mass tures. They are separated into hardenable and non-
26.98154. hardenable types by a thermal treatment. The me-
The electron configuration of the outer shell chanical properties of nonhardenable alloys can be
is 3s 2 3p1 . Successive ionization energies are improved by alloying and work hardening, while
5.99; 18.83; and 28.44 eV. Atomic radius is those of the hardenable alloys are improved by
0.143 nm; radius of Al3+ ion is 0.051 nm. The quenching and aging (see Alloy aging). To increase
main oxidation state is +3, with +2 and +1 pos- their plasticity the alloys are annealed (see Anneal-
sible at high temperatures. Electronegativity is ing). Basic alloying elements of deformable alloys
1.5. Free aluminum is a silvery-gray metal, un- are Cu, Mg, Mn, Si, Zn, and Cr. Cast alloys pos-
der normal conditions covered with a thin ox- sess fluidity, high density, and low linear shrink-
ide film. It has one stable isotope 27 Al. It is the age. They do not tend to form voids or cracks. The
48 ALUMINUM NITRIDE
main alloying elements of cast alloys are: Cr, Cu, where ne,h , μe,h , De,h are the concentrations, mo-
Mg, Mn, Ni, Ti, and Zr. Aluminum alloys are used bilities and diffusion coef cients of electrons and
as antifriction materials. In volume of production holes. Non-degenerate carriers obey the Einstein
and range of application aluminum alloys are sec- relation De,h = (T /e)μe,h , where e is the elemen-
ond after ferrous metals. tary charge, T is the temperature expressed in en-
ergy units, and as a result, the previous equation
ALUMINUM NITRIDE, AlN simplifies to
The only chemical compound of aluminum ne + nh
D= .
with nitrogen. It has a hexagonal wurtzite-like ne /Dh + nh /De
4 ); crystal
(ZnS) lattice of space group P 63 mc (C6v
lattice parameters are a = 0.3112 nm,
c = 0.4981 nm. Aluminum nitride decomposes AMBIPOLAR DRIFT
at 2400 ◦ C; density is 3.26 g/cm3 ; heat con- The drift of the quasi-neutral perturbation of
ductivity of polycrystalline samples varies from charge carrier density under the action of an exter-
50 to 150 W·m−1 ·K−1 depending on purity; nal electric field: E = −q∂n/∂x. Since the elec-
heat conductivity of monocrystals is 270 to tric force acts in opposite directions on electrons
360 W·m−1 ·K−1 ; microhardness is 1.2 GPa. Alu- and holes, an additional electric field arises due to
minum nitride is an insulating material with elec- charge separation in the density perturbation. In
trical resistivity ρ > 1015 ·m and band gap width addition to the drift fluxes from each-type carrier
exceeding 6.2 eV. It reacts with water, diluted by density irregularity, fluxes in the disturbed field
mineral acids; it does not interact with melt Al, arise which to some extent compensate the former.
Ga, Cu, cryolite, gallium arsenide. Aluminum ni- When the undisturbed electron and hole concen-
tride is obtained by nitration of Al, by reduction trations are equal (ne = nh ), complete compensa-
of Al with simultaneous nitration, by thermal de- tion takes place, and the bunches become immo-
composition of aluminum-containing compounds bile. The ambipolar drift velocity is given by v =
(AlCl3 ·6NH3 ). It is used as a fireproof and elec- μE, where μ = μe μh (ne − nh )/(ne μe + nh μh )
trical insulating material. is the ambipolar mobility, where μe,h are the
electron and hole mobilities. It follows from the
ALUMINUM–YTTRIUM GARNET formula that in the limiting cases ne
nh and
See Yttrium–aluminum garnet. ne nh the ambipolar mobility μ is equal to μh
and −μe , respectively, i.e. it coincides with the
AMBIPOLAR DIFFUSION
mobility of the minority carriers. The expression
for the ambipolar drift mobility is valid if the car-
The diffusion of quasi-neutral disturbances of
rier mobilities are isotropic and independent of ne
an electron–hole plasma density n driven by the
and nh and there is no heating of the electrons and
gradient ∂n/∂x (see Solid-state plasma). If the
holes by the field. Otherwise v can differ from zero
electron and hole diffusion rates differ then some
even when ne = nh .
spatial charge separation occurs. An electric field
arises inside the disturbed density region to re-
AMERICIUM, Am
tard the fast carriers and accelerate the slow ones.
A chemical element of Group III of the peri-
Therefore, the carriers of either type, while dif-
odic table with atomic number 95 and atomic mass
fusing in the direction of their concentration gra-
243.0614.
dient, simultaneously drift in the self-consistent
Fifteen radioactive isotopes are known with
(so-called imbedded) electric field of the separated mass numbers 232, 234 to 247. There are no sta-
charges. With this field accounted for, it is possi- ble isotopes. Electronic configuration of filling
ble to describe the density cluster dynamics by just and outer shells is 5f 7 6d 0 7s 2 . The ionization en-
one ambipolar diffusion coefficient ergy is 5.99 eV. Atomic radius 0.174 nm; radius
ne μe Dh + nh μh De of Am3+ ion is about 0.10 nm, of Am4+ ion
D= , 0.09 nm, of Am5+ ion 0.086 nm, of Am6+ ion
ne μe + nh μh
AMORPHOUS EPITAXY 49
and stacking faults. Immediately after their syn- disordered ferromagnets with ferromagnetic ex-
thesis, amorphous magnetic materials have high change interactions which fluctuate noticeably in
internal stresses bringing about the appearance magnitude. The class is exemplified by metal-
of magnetoelastic anisotropy. To eliminate these lic glasses Fe80 B20 , Fe40 Ni40 B20 . Most physi-
stresses, thermal and thermomagnetic annealing is cal properties of these systems differ only slightly
performed at temperatures below the crystalliza- from those of ordered ferromagnets. The main dis-
tion temperature, and alloys with a zero magne- tinction is that amorphous ferromagnets are char-
tostriction coefficient are used. One disadvantage acterized by local frozen magnetization mi and
of amorphous magnetic materials is the presence its first moments M = mi c and q = m2i c (av-
in them of structural relaxation. eraging over the configurations of exchange inte-
Amorphous magnetic materials have prospec- grals). As a result, some interesting new phenom-
tive applications in electrical and electronic engi- ena arise which are also present in other classes
neering. The replacement of conventional electri- of amorphous magnetic substances, e.g., the crit-
cal sheet steels, precision alloys of the permalloy ical magnetic scattering of neutrons by the con-
type (see Soft magnetic materials) and ferrites in figuration fluctuations of magnetization, and fluc-
different electromagnetic devices and transform- tuations decaying with distance following the ex-
ers with amorphous magnetic materials manufac- ponential dependence exp(−κr) rather than the
tured by a simpler and cheaper process allows Ornstein–Zernike law.
reducing the dimensions of the devices, the en- The second class of amorphous magnetic sub-
ergy loss, and the cost. In electronic engineer- stances comprises systems with concurrent ferro-
ing, the low-coercive field amorphous magnetic and anti-ferromagnetic exchange interactions, ex-
materials find wide application in the produc- emplified by the alloys Eux Sr1−x S and metal-
tion of magnetic heads, delay lines, and magnetic lic glasses Fex Mn1−x P16 B6 Al3 . The exchange
shields. Ferromagnetic films (see Magnetic lms) interaction can have different signs at different
of amorphous magnetic materials of a transition
bonds, and depending on these signs these mate-
metal or rare-earth metal type, with high single-
rials may be paramagnets, ergodic ferro- or anti-
axis magnetic anisotropy (up to 5·105 erg/cm3 )
ferromagnets, nonergodic ferro- or antiferromag-
induced during the process of synthesis, can be
nets (asperomagnets) as well as spin glasses.
used as inexpensive media for the recording de-
These states are characterized by different sets of
vices in magnetic bubble domains. With the films
first and second moments, possible local frozen-
Co–Gd, the diameter of magnetic bubble do-
in magnetization, and the presence or absence of
mains achieves submicron dimensions with mobil-
nonergodicity. The most interesting property of
ity 20 m·s−1 ·Oe−1 . The Fe–Tb films are prospec-
this class of substances is the emergence of non-
tive materials for contact printing and optical in-
formation recording with magneto-optic reading.
ergodic phases; experimentally the phases are as-
Their advantages are low Curie temperature (310 sociated with various irreversible phenomena.
to 370 K) and, hence, low optical energy re- The third class of amorphous magnetic sub-
quired for the thermomagnetic information record- stances comprises alloys and the metallic glasses
ing, strong magnetooptical effects, absence of the which include rare-earth elements and noble met-
intercrystallite boundaries which bring about ad- als of the DyCu and TbAg type. These exhibit fer-
ditional scattering and absorption of light, and, as romagnetic interactions and high local magnetic
a result, high signal-to-noise ratio in the reading of anisotropy with random directions of anisotropy
information. axes. The last property is the salient one char-
acterizing the physics of this class of magnetics.
AMORPHOUS MAGNETIC SUBSTANCES Amorphous magnetic substances are asperomag-
Involves solid substances having long-range nets (typically, nonergodic) or spin glasses, and
magnetic order, but lacking long-range crystalline they are used as hard magnetic materials; magnetic
order (see Long-range and short-range order). bubble domains can be observed inside them. De-
There are three classes of amorphous magnetic spite the dissimilarities of their interactions, this
substances. The first such class includes highly class of materials is quite close to the previous one
52 AMORPHOUS MAGNETISM
levels of broken bonds D 0 and of charged D + transfer to delocalized states of the conduction
and D − states (Fig. 3(b)). The presence of two band, arriving first at the localized states (the
unfilled electronic levels lying below the single- “tail”) of this band. In this case
electron states in chalcogenide ASCs due to the
E − EF + W
effective attraction Ueff between electrons and of σ = σ1 exp − A ,
free electronic levels at the broken bonds in tetra- kB T
hedral ASCs brings about the so-called Fermi level
pinning in the middle of the gap. Doping with
where typically σ0 10 −1 cm−1 , W is the ac-
donor (acceptor) impurities fails to shift the Fermi tivation energy of the transitions between local-
level EF from the middle of the pseudo-gap to ized states within the tail of the conduction band.
the mobility edge of electrons (holes) at doping At low temperatures and large energy gap widths,
with donor (acceptor) impurities because the elec- jump conductivity prevails. This latter conduction
trons (holes) fill the empty levels of defects in the mechanism is characteristic of tetrahedral ASCs,
middle of the pseudo-gap. Experimental measure- while the former one applies to chalcogenide
ments reveal rapidly decaying “tails” of the va- ASCs. The mobility of electrons and holes is usu-
lence and conduction bands, and the presence of ally low, depending on the electric field strength
several peaks in the of density of states in the and on the sample thickness. The conductivity of
pseudo-gap, apparently associated with different such amorphous semiconductors can be changed
types of structural defects. by 7 to 8 orders of magnitude with the help of dop-
A possible way to change electronic proper- ing.
ties of tetrahedral ASCs with doping is accom- Many chalcogenide amorphous semiconduc-
plished by preliminary hydrogenization, or the tors have the property of conductivity switching,
introduction of hydrogen, which saturates the bro- i.e. the ability to transform reversibly from a state
ken bonds and thus counteracts the Fermi level with high electrical resistance to one with low
pinning in the middle of the gap. Changes in the resistance under the influence of a strong elec-
electronic properties of ASCs without the prelim- tric field with intensity 105 V/cm (and higher).
inary saturation of broken bonds can be brought The switching commonly occurs during short
about by the introduction of relatively higher con- (10−10 s) time intervals, and the underlying cause
centrations of doping elements (see also Doped of the phenomenon is twofold. First, it can oc-
amorphous semiconductors). cur as a result of injection of additional electrons
Electrical conductivity. In amorphous semi- (or holes) into ASCs with an associated increase
conductors the Fermi level lies in the energy gap. of the concentration of carriers in the delocal-
The transfer of electrons to the external electri- ized state. It can also be due to ohmic heating of
cal field may occur either by crossing over to the the sample in the area of a current channel. The
conduction band above the mobility edge Ec1 , or formation of current channels with small cross-
by hopping (jumping) from the level of one de- sections occurs due to the local heating of the sam-
fect to an empty level of another, adjacent defect ple upon the reduction of resistance, so that the
(see Hopping conductivity). The prevalence of ei- switching in this case involves the rearrangement
ther mechanism of conductivity depends on the of the material, uniform in cross section, by the
electronic structure of ASCs and on the temper- flow of electrons into narrow current channels. In
ature. Prevailing at high temperatures and small some chalcogenide ASCs, the state of low elec-
gaps between EF and Ec is the transfer connected trical resistance persists for a long time, which is
with activation transitions of electrons to the con- explained by the partial crystallization of the sam-
duction band; here ple in the region of a current channel. The return to
the state with high electrical resistance occurrs by
Ec − EF
σ = σ0 exp − , the transmission of a short pulse of current. Amor-
kB T
phous semiconductors exhibit the phenomenon of
where σ0 ∼ 103 to 104 −1 cm−1 . At lower tem- photoconductivity (see Photoconductivity of amor-
peratures, the electrons from the pseudo-gap may phous materials).
AMORPHOUS SILICON 55
Optical properties. The absorption spectra of light, known as the phenomenon of photostruc-
light in amorphous semiconductors reflect the tural transformations. These amorphous semicon-
structure of the electron density of states. The co- ductors form the basis of elements of electronics
efficient of absorption of the photon α(ω) with and fiber optics in the IR range, including the wave
energy ω is proportional to the density of elec- length 10.6 μm. Fast switches and memory arrays
trons which are capable, having absorbed the en- based on switching phenomena are available. Hy-
ergy, of transferring it to unfilled levels. Accord- drogenated silicon and other tetrahedral ASCs are
ingly, it is high when ω is approximately equal used in the manufacture of solar cells, effective
to or greater than the width E0 of the pseudo-gap. electroluminophors, videocons, etc.
In this case α(ω) ∝ c0 (ω − E0 )/(ω), where
c0 ∼ 105 to 106 cm−1 eV−1 . For a photon energy AMORPHOUS SILICON (α -Si)
slightly below E0 , the absorption coefficient expo- Silicon is one of the most comprehensively
nentially decreases along with ω (Urbach rule): investigated amorphous semiconductors. It has
α(ω) ∝ exp[−β(E0 − ω)], where E0 differs lit- tetrahedral coordination, and is characterized by
tle from E0 , and the constant β depends weakly the same short-range order (see Long-range and
short-range order) as crystalline silicon: the co-
on the temperature. An explanation of the Urbach
rule given by J. Dow and D. Redfield (see Dow– ordination number is close to four, interatomic
distances in α-Si differ from corresponding crys-
Red eld model) is based on local fluctuations of
talline values by several percent, and uctuations
the potential, which occur due to the topologi-
of tetrahedral angles are about 10◦ . Usually α-Si
cal distortions of the network of atoms. The ab-
is obtained in the form of lm s by spraying or
sorption of the photons with energies consider-
decompositing silane SiH4 in a high-frequency
ably lower than E0 is associated with transitions
glow discharge. The latter method is used to obtain
of electrons to localized states within the pseudo-
amorphous lms of hydrated amorphous silicon
gap.
(α-Si:H) with hydrogen content of several atomic
The phenomenon of photoluminescence, which
percent.
was discovered in chalcogenide and tetrahedral In the pseudogap of the non-hydrated mater-
ASCs, is inherent in ASCs, the number of radi- ial, the density of the localized states is high (1016
ated light quanta being comparable to that of ab- to 1019 eV−1 cm−3 ). As a result, the Fermi level
sorbed phonons (or an order of magnitude less). undergoes almost no shift upon doping. Dangling
The luminescence spectra, usually wide, lie in the bonds are considered to be the main type of de-
interval of energies below E0 and lack n e struc- fect that creates deep levels in the mobility gap.
ture. Chalcogenide ASCs exhibit a reduction of lu- For these deep defects, the intracenter energy of
minescence in time during the excitation process, electron correlation is positive. The concentration
accompanied by the detection of electron para- of defects changes under annealing, and also un-
magnetic resonance. This phenomenon of photo- der the Stebler–Vronsky effect, which consists in
induced electron paramagnetic resonance is ex- the reversible (after annealing) reduction of the
plained by the redistribution of D 0 , D + and D − conductivity of films after long-term preliminary
defects under the action of light. irradiation with white light. The direct current
Applications. Amorphous semiconductors find conductivity σ at low temperatures (T 300 K)
a variety of applications along with crystalline is described by the Mott law ln σ ∼ T −1/4 , and
semiconductors. In electrical photography, they the alternating current electrical conductivity de-
are used as photosemiconductor layers with high pends on the frequency according to the power law
electrical resistance and low charge mobility. They Re σ (ω) ∼ ωs , where typically s = 0.8 to 0.95; the
serve as materials for nonsilver photography which thermal electromotive force at low temperatures is
includes a stage of chemical development; in holo- independent of temperature. All this indicates that
graphy, for manufacturing of masks, microfiches, the hopping conductivity mechanism with a vari-
etc., where use is made of the unique property able jump length is realized in amorphous silicon.
of chalcogenide ASCs to change optical parame- The hydrated material differs by a low den-
ters under the irradiation with strongly absorbed sity of localized states in the pseudo-gap. By
56 AMORPHOUS SOLID
bringing the hydrogen concentration in films to material in the amorphous state is a thermody-
5–15 atomic %, the density of localized states can namically unstable or metastable system that usu-
be reduced to 1015 to 1016 eV−1 cm−3 . The mate- ally becomes crystalline upon annealing. How-
rial can be doped by adding phosphine PH3 or di- ever, polymeric glasses consisting of high molec-
borane B2 H6 to the source gas mixture during the ular weight compounds typically do not yield to
fabrication of films. In both doped and undoped crystallization. The physical properties of amor-
films the resistance usually depends exponentially phous solids differ substantially from those of
on the inverse temperature. crystalline solids with the same composition. See
The optical width of the band gap of also Vitreous state of matter, Amorphous metals
α-Si:H depends on the hydrogen content, and is and metallic alloys, Amorphous semiconductors,
usually 1.6 to 1.8 eV. Films of α-Si:H exhibit a Metallic glasses.
high coefficient of light absorption (in the maxi-
AMORPHOUS SUPERCONDUCTORS
mum region of the solar spectrum, it exceeds the
Superconducting materials with an amorphous
absorption coefficient of crystalline silicon by ap-
structure.
proximately an order of magnitude) and have high
These materials are, typically, amorphous met-
photosensitivity (ratio of photoconductivity to dark allic lm s obtained by deposition on cold sub-
conductivity is as high as 106 –107 ). The above- strates and metallic glasses produced by rapid
mentioned properties of α-Si:H, and the possibil- cooling (quenching) from the melt. Many com-
ity of obtaining large-area multilayer structures on pounds which are nonsuperconducting in the crys-
its basis within the scope of a continuous auto- talline phase become superconducting in the amor-
mated process, make α-Si:H favorable for the pro- phous state, e.g., Be, Bi, transition metal alloys
duction of inexpensive solar cells. Maximum effi- such as Zr1−x Nix , Zr1−x Cox , and others. Con-
ciencies achieved for such structures of small area versely, many crystalline compounds lose super-
exceed 13%. Amorphous silicon is used in electric conductivity upon amophization, e.g., compounds
photography and microelectronics, in particular for with the A-15 structure (e.g. Nb3 Sn), and high-
creating thin-film transistors acting as drivers in temperature superconductors.
flat liquid-crystal displays. See also Amorphous Amorphous superconductors based on non-
semiconductors, Doped amorphous semiconduc- transition metals have a strong electron–phonon
tors. interaction, while those based on transition met-
als and their alloys have a weak or intermediate
AMORPHOUS SOLID strength electron–phonon interaction, even though
See Polycluster amorphous solids. in the crystalline phase these metals have a trend
toward strong coupling. At temperatures below
the critical temperature of superconductors Tc ,
AMORPHOUS STATE
amorphous superconductors are the most favor-
A state of a solid without any long-range or- able materials for studying two-level systems in
der (see Long-range and short-range order) of the metals. In these materials, due to their short co-
atomic arrangement. As distinct from liquids and herence length ξ , the domain of strong thermo-
gases, the mean atomic positions in amorphous dynamic uctuations covers a considerably wider
solids do not change in time, i.e. the changes in temperature range than in ordered systems. In the
atomic configurations due to the thermal u ctua- case of extensive disorder (at the threshold of An-
tions occur randomly about fixed lattice sites and derson localization), thermodynamic fluctuations
can be ignored in the structure description. Solids bring about a reduction of Tc . Here, statistical spa-
can be transformed to the amorphous state by sev- tial fluctuations of the superconductor order pa-
eral different methods: deposition of molecular, rameter become essential; they may involve the
atomic, ion and plasma fluxes on substrates, rapid percolation nature of the superconducting phase
cooling (quenching) of a liquid, amorphization of transition. A reduction of Tc in amorphous super-
crystalline solids. Amorphous solids obtained by conductors can also occur as a result of: (a) the en-
cooling of liquids are called glasses. The solid hancement of the Coulomb repulsion at the slow
AMPLITUDON 57
the symmetric incommensurate phase type transi- and satisfies the Hamilton–Jacobi partial differen-
tion may correspond to a non-unidimensional rep- tial equation
resentation of the symmetry group of the symmet-
∂S ∂S
ric phase, so that several amplitudon branches may + H ..., , . . . , qk , . . . , t = 0,
exist in this phase. During vibrations correspond- ∂t ∂qk
ing to amplitudons, quantum effects are usually with the momenta in the Hamiltonian being re-
not too noticeable. placed by the partial derivatives pk → ∂S/∂qk .
After integration, the Hamilton–Jacobi equation
ANALYSIS, CHEMICAL provides n + 1 arbitrary constants, S = (t, q1 , . . . ,
See Chemical methods of analysis. q0 , α1 , . . . , α0 ) + αn+1 , and it is possible to obtain
the solutions of the equations of motion by means
ANALYTICAL MECHANICS (introduced by of the set of partial differentiations ∂S/∂αk = βk ,
J.L. Lagrange) where βk are additional constants. All qk are ex-
Theoretical mechanics based on mathematical pressed in terms of them as functions of time and
equations without the aid of geometric construc- the constants αk . Instead of solving the differen-
tions. tial equations, one can use the least action prin-
ciple and seek the extremum trajectory from the
Modern analytical mechanics applies a set of
condition δ L dt = 0 (Hamilton’s principle), or
effective and quite general methods, based in part
from other principles related to it. The application
on minimization principles, for the solution of me-
of variational calculus methods often allows solv-
chanical problems; in particular utilizing symme-
ing the dynamic problem approximately, or inves-
try for finding the various integrals of motion.
tigating the character of the solution. If some vari-
One expresses the equations of motion in the form
ables from the set q1 , . . . , qk are explicitly absent
of Lagrange’s equations in the presence of holo-
from α or H (they are called cyclic coordinates),
nomic constraints, and adopting generalized coor-
then the integrals of the motion can be found di-
dinates can make it possible to eliminate some un-
rectly. For example, if ∂H /∂qk = 0, then the cor-
necessary coordinates.
responding canonical momentum pk is conserved.
The Lagrange function L (the difference be- When the Hamiltonian has no explicit dependence
tween the kinetic and potential energies of the on the time t , and the forces are derivable from a
system), is expressed via the generalized coordi- conservative potential (i.e. work is independent of
nates qk and generalized velocities q̇k . The n La- the path), then the Hamiltonian is the total energy:
grange’s equations have the form H (p, q) = E.
d ∂L ∂L
− = 0, where k = 1, 2, . . . , n. ANDERSON–BOGOLYUBOV MODE
dt ∂ q̇k ∂qk
in superconductors
If one defines the generalized momenta as Oscillations of the order parameter phase when
pk = ∂L/∂ q̇k , and composes
the Hamiltonian a superconductor is treated as a neutral superflow-
H (. . . , pk , . . . , qk , . . . , t) = pk q̇k − L, the sys- ing Fermi uid (see Super uidity). The Anderson–
tem of 2n first-order equations q̇k = ∂H /∂pk , Bogolyubov mode corresponds to a Goldstone
ṗk = −∂H /∂qk (Hamilton’s canonical equations) excitation arising from a spontaneous symmetry
is obtained. Another approach is to define Hamil- breaking at the transition to the superconducting
ton’s principal function S in terms of the action in- state. At zero temperature and small wave vec-
tegral S = tt0 L(. . . , qk , . . . , q̇k , . . . , t) dt + const. tors k, the Anderson–Bogolyubov mode has a lin-
While evaluating the integral, the time depen- ear dispersion
√ law ω = vk, where in a pure case
dences of qk and q̇k are assumed to correspond v = vF / 3 (vF is the electron Fermi velocity). At
to the actual motion between the initial and fi- the excitation of this mode, an electron density
nal qk (t) values of the system coordinates. Then disturbance takes place, and the spectrum of such
the function S depends on these coordinate values, oscillations in real superconductors arises due to
ANDERSON MECHANISM OF HIGH-TEMPERATURE SUPERCONDUCTIVITY 59
Coulomb interactions involving the plasma fre- close to the top and bottom of the band still are.
quency (see Collective excitations in superconduc- In this case, the electron system is in the insulat-
tors). ing state if the Fermi energy EF is lower than the
threshold energy (see Mobility edge) Ec ≈ W , and
ANDERSON HAMILTONIAN it is in the conducting state, if EF > Ec . The elec-
See Anderson model. trical conductivity behaves in the close vicinity of
the metal–insulator transition point as follows:
ANDERSON LOCALIZATION
σ ∼ (t − γ W )ν or
The disappearance of electrical conductivity in (2)
a system of noninteracting electrons when the im- σ ∼ (EF − Ec )ν , ν = 0.5.
purity concentration becomes sufficiently high. If the random potential in which the electrons
This can explain metal–insulator transitions at move changes only slightly on a scale of the or-
a low temperature in amorphous semiconductors, der of the electron wavelength, then the quantum
doped amorphous semiconductors, and other dis- value of the critical index ν = 0.5 “works” in a
ordered solids. Anderson localization is also stud- very narrow region, whereas in a wider region near
ied in disordered spin systems (disappearance of the critical point the index ν takes on the quasi-
spin diffusion), and for normal modes of dis- classical value 1.7 predicted by percolation theory.
ordered lattices (fractons). Under consideration All the preceeding remarks concern ordinary
when investigating Anderson localization is the three-dimensional systems. It has been proven that
correlation function of the density–density type in the one-dimensional case an electron has only
n(r, t)n(r, 0), or the Green’s function Grr (t). If localized states for any value of the disorder para-
these functions do not vanish in the limit t → 0, meter W . Apparently this is also valid in the two-
the electron “remembers” its initial state and, dimensional case, except for the only infinitely ex-
hence, is in a localized state. Serving as a local- tended state in the middle of the band.
ization criterion in the energy representation is Closely related to Anderson localization are
the presence of a pole of the Fourier-transformed such phenomena as weak localization and meso-
Green’s function G(ω, t) at zero frequency. In the scopics, in which the interference of electron
initial formulation, use was made of a model in scattering processes by several impurities, al-
which an electron jumps from site to site, the en- though weak, still plays a significant role. It
ergy of a site having random values in the range should be noted that Anderson localization is de-
(−W, W ), with the tunneling amplitude depend- scribed in the single particle approximation. At
ing on the distance according to a power law r −α finite temperatures, the inelastic particle interac-
(see Anderson model). If the tunneling amplitude tion becomes significant and cancels Anderson lo-
decays rapidly enough (α > 3), all the states are calization. However, it can be shown that self-
localized provided the following inequality holds: localization takes place even in an ideal crystal if
t < γ W, γ ≈ 1, (1) inequality (1) is satisfied, where the quantity W
has the sense of a characteristic level difference at
where t is the mean tunneling amplitude, or a neighboring sites. In an ideal crystal, this differ-
quantity proportional to it, the width of the con- ence of levels is not due to impurities, but to the
duction band that would be created for W t . intrinsic interaction of particles with one another.
The value of γ is determined numerically. In the Autolocalization has been observed in studies of
Lifshits model, the disorder is not specified by the the diffusion of 3 He atoms in solid 4 He. See also
spread in the site energies, but by the randomly Elementary excitation spectra of disordered solids.
positioned short-range impurities. In this model,
Anderson localization takes place when the mean ANDERSON MECHANISM OF HIGH-
free path decreases to the order of the de Broglie TEMPERATURE SUPERCONDUCTIVITY
wave length. If condition (1) for complete local- (P.W. Anderson, 1987)
ization is not satisfied then the states in the central A mechanism of superconductivity in systems
region of the band are not localized, but the ones with “resonating” valence bonds (RVB).
60 ANDERSON MECHANISM OF HIGH-TEMPERATURE SUPERCONDUCTIVITY
This magnetic mechanism of Cooper pairing ter are the topological defects (kinks), their ther-
involves a fundamentally new type of ordering modynamically equilibrium number being deter-
for two-dimensional systems with an antiferro- mined by the temperature. The kink displacement
magnetic Heisenberg interaction (see Antiferro- corresponds to the “valence bond transfer” from
magnetism). Proceeding from the experimental one site to another, with no change in energy and
data on the insulating (La2 CuO4 ) and conducting no transfer of charge taking place. It is in this sense
(La2−x (Ba, Sr)x CuO4 ) phases of lanthanum com- that the state (1) corresponds to “resonating va-
pounds, Anderson proposed several hypotheses re- lence bonds”, while spinons (S = 1/2) turn out to
garding their properties, and a possible supercon- be neutral, zero mass fermions.
ductivity mechanism of high-temperature super- The transition from the insulating RVB-state
conductors. The presence of bivalent copper Cu2+ to the metallic one is caused by introducing new
with one hole in the unfilled 3d-shell classifies holes into the system (by doping or a change in
La2 CuO4 as a magnetic insulator adequately de- stoichiometry). The “extra” (free) holes are cou-
scribed by Hubbard’s Hamiltonian (see Hubbard pled to a spinon forming a new charged Bose-
model) with a strong intra-atomic Coulomb re- particle called a holon. The motion of holons
pulsion hindering the free motion of the carri- both inside and between planes, their scattering
ers (holes), and causing their localization (Mott by spinons, and the phenomenon of holon con-
metal–insulator transition). The Mott–Hubbard in- finement determine the finite resistance in the nor-
sulator ground state with one electron (hole) per mal metallic phase, as well as the attraction with
atom (half-filled band) is equivalent to Heisen- the formation of holon Cooper pairs, thereby fa-
berg’s antiferromagnet (see Heisenberg magnet) cilitating a transition to the superconducting state
with one spin S = 1/2 per site. A second An- at temperatures on the order of Heisenberg ex-
derson suggestion based on H. Bethe’s works on change interaction constant. The holon gas does
one-dimensional magnetic materials, as well as not undergo a Bose–Einstein condensation. There
on L. Pauling’s chemical bond theory, matches is a relationship between the wave function (1)
with the essentially quasi-two dimensional char- and the condensate wave function of the Bardeen–
acter of the structure of superconducting cuprates. Cooper–Schrieffer theory
+ +
The suggestion means that under certain condi-
tions quantum fluctuations, even at T = 0, are able ΨBCS = uk + vk ak↑ a−k↓ |0,
to completely destroy antiferromagnetic ordering k
in the CuO2 planes, so that the complete wave |ΨBCS |2 = 1, (2)
function of the ground state of an individual plane
+
can be represented by a linear combination of the where uk , vk are coefficients, and akσ (σ =↑, ↓)
wave functions of “valence-coupled” pairs of the are Fermi creation operators of carriers from the
localized carriers: vacuum state |0. The wave function (2), projected
on the pairs and then transformed into real co-
ΨRVB = C{n, m} (αn βm − βn αm ), ordinate space, acquires the functional form (1).
{n,m} n,m
Another feature of the Anderson mechanism lies
|ΨRVB |2 = 1, (1) in the possibility of a qualitative description of
some unusual properties of high-temperature su-
where n, m are the bidimensional site vectors, perconductors (such as the anomalous tempera-
{n, m} represent all possible configurations with ture dependence of the heat capacity, the resis-
contributions determined by the coefficients tance in the normal phase, anisotropy of conduc-
C{n,m} , and αn , βn are functions of “up” and tance, etc.). Experiment shows that cuprate metal
“down” spin projections. In this state, the mean oxides in the normal state are ordinary Néel anti-
value of a spin at each site is zero, but the phase co- ferromagnets. However, some arguments favor in-
herence is retained. The lowest excitations above ducing the RVB-state in metal phases under con-
the ground state (1), known as spinons, are dan- ditions of strong spin interactions of free and lo-
gling bonds or individual unpaired spins. The lat- calized carriers.
ANDREEV REFLECTION 61
ANDERSON MODEL (P.W. Anderson) the presence of a dielectric gap on a part of the
(1) In disordered systems theory – a strong cou- Fermi surface, etc.
pling model with diagonal disorder described by
Anderson’s Hamiltonian ANDREEV REFLECTION of quasi-particles
(A.F. Andreev, 1964)
H =J |nm| + εn |nm|, A specific reflection of quasi-particles (elec-
n,m n trons and holes) from the interface between the
superconducting (S) and normal (N) phases of a
where n, m are the sites of a d-dimensional lat-
metal (NS-boundary).
tice, n, m are pairs of the nearest neighbors,
During this process the quasi-momentum p
J is the transfer integral, and the energies εn at dif-
of a reflected particle is conserved, but the signs
ferent sites n are independent, uniformly distrib-
of the charge and all three velocity components
uted random quantities. The pioneer work of An-
change. When an electron of energy ε (measured
derson gives arguments in favor of the fact that for
from the Fermi level) and velocity v = ∂ε/∂p
d = 3, if εn are uniformly distributed in the inter-
is incident from the normal metal side on the
val [−W/2, W/2] and the quantity W/J exceeds
NS interface, the reflected quasi-particle is a hole
some critical value (W/J )c , then all the states of
with the same energy ε and velocity v. The re-
this Hamiltonian are localized (Anderson localiza-
flection of holes proceeds in a similar way. The
tion). The Anderson model provides an effective
value of the Andreev reflection coefficient R de-
analysis of the transition from quasi-Bloch (prop-
pends on the quasi-particle’s energy: R(ε) = 1 for
agating) states to localized ones, and clarifies the
0 < ε < Δ (Δ is the value of the energy gap in
presence and position of the mobility edge, char-
the superconducting phase); R(ε) → 0 for ε
Δ.
acteristics of the fluctuation elementary excitation
The essence of Andreev reflection is that at the
spectra of disordered solids, and other properties
NS-boundary the incident electron enters the su-
of disordered solids.
perconducting condensate, forming a Cooper pair
(2) In the theory of magnetism this approach with an electron having the opposite momentum.
provides a model Hamiltonian to describe the for- A hole arising in the process moves back into the
mation of the local magnetic moment of an im- normal region. When the normal metal layer is
purity atom in a nonmagnetic metal as the impu- situated between superconducting slabs (e.g., in
rity level changes relative to the Fermi level. The the intermediate state of a superconductor or in
model admits an exact solution by means of the SNS-junctions), Andreev reflection gives rise to
Bethe substitution (Bethe ansatz). distinctive quantization of the low-energy (ε < Δ)
(3) In the theory of metal-oxide supercon- excitations of the N-layer (Andreev quantization).
ductors – the model provides an explanation of This quantization corresponds to a periodic mo-
resonating valence bonds (see High-temperature tion of quasi-particles inside the N-layer, with the
superconductivity, Anderson mechanism of high- quasi-particle being an electron for half of the pe-
temperature superconductivity). riod and a hole for the other half.
Andreev reflection manifests itself in a number
ANDERSON THEOREM (P.W. Anderson, 1959) of features of the thermodynamic and kinetic prop-
A theorem stating that the electron scattering erties of the intermediate state of a superconduc-
by nonmagnetic impurities in superconductors has tor. In particular, it leads to geometrical oscillation
a slight effect on the critical temperature only. This effects in ultrasonic attenuation, and to an addi-
theorem is based upon the invariance of the Hamil- tional thermal resistance of the intermediate state
tonian with respect to time inversion. The Ander- during thermal flux propagation normal to N- and
son theorem is not necessarily valid outside the to S-layer, etc. As distinct from thermal flux prop-
framework of the Bardeen–Cooper–Schrieffer the- agation, the passage of an electric current through
ory, namely, if one takes into account the weak lo- the NS-boundary at low temperatures (kB T Δ)
calization of carriers, anisotropy of the Fermi sur- is not associated with overcoming the excess resis-
face, fine structure of the density of electron states, tance of the S-boundary due to the quasi-particle
62 ANGULAR DEPENDENCE OF SPECTRA
reflection, since the conversion of the normal exci- meters, Invariants), and the local symmetry groups
tation current into that of Cooper pairs takes place of its interactions. An example of the angular de-
under Andreev reflection. Andreev reflection has pendences of the magnetic resonance spectra of
been directly observed experimentally in studies the Gd3+ ion at a site of symmetry C2 is given
of the radio-frequency size effect, and of transverse in Fig. which plots the values of the resonant mag-
magnetic focusing of electrons in NS-structures. netic e ld Br versus the azimuthal and polar angles
θ and ϕ, respectively, for the indicated transitions
ANGULAR DEPENDENCE OF SPECTRA M → M − 1, where M is the eigenvalue of the
of magnetic resonance projection of the electronic spin S = 7/2 on B.
The dependence of the resonant magnetic eld When the symmetry and arrangement of crystallo-
Br (in case of electron paramagnetic resonance, graphic axes (see Crystal symmetry) are unknown,
EPR) or resonance frequencies (in case of electron– they can be determined from EPR (NMR) spec-
nuclear double resonance, ENDOR; nuclear mag- tra recorded for rotations around three mutually
netic resonance, NMR) on the relative orientation perpendicular axes fixed in the crystal. In certain
of the constant external magnetic field B and the cases, the values of θ and ϕ also affect the line
crystal axes, i.e. on the values of the polar angle θ width and line intensity.
and azimuthal angle ϕ that specify the B direc- An EPR spectrum is customarily called an
tion in a coordinate system relative to the crystal. anisotropic EPR spectrum, if the positions and in-
The parameters of a generalized spin Hamiltonian tensities of its lines depend on the orientation of
are obtainable by comparing experimentally ob- the sample relative to the external magnetic field.
served angular spectral dependences to those cal- For spin S 1 and in the absence of anisotropic
culated from the Hamiltonian. Examining angular hyper ne interactions, the anisotropic spectrum is
spectra in three mutually perpendicular planes per- observed only for centers belonging to non-cubic
mits determination of all parameters of the effec- spin Hamiltonian symmetry groups, whereas in
tive spin Hamiltonian (see Spin Hamiltonian para- the presence of anisotropic hyperfine interactions
with surrounding nuclei an anisotropic spectrum levels. Accounting for the anharmonicity is a typ-
can be observed even for cubic symmetry groups. ical many-body problem, with its solution provid-
If the positions and intensities of EPR spectral ing the answers to some fundamental questions
lines are independent of the orientation of the sam- of solid state physics, such as the possibility of
ple under investigation relative to the direction of forming two-phonon coupled states (biphonons),
the external constant magnetic field, then the spec- the role of soft modes in structural phase transi-
trum is called an isotropic EPR spectrum. The con- tions, etc. Slightly and highly anharmonic crystals
cepts “anisotropic spectrum” and “isotropic spec- can be distinguished. Quantum crystals, systems
trum” are valid also for other resonance phenom- with structural phase transitions, in particular fer-
ena such as NMR. roelectrics, fall into the latter type. The presence of
solitons, which can contribute to the energy trans-
ANGULAR MOMENTUM, moment of momentum
fer and to the process of the restoring the equilib-
Vector product r × mv, where r is the posi- rium state after an external perturbation, is possi-
tion vector, v is the velocity of a material point of
ble in an anharmonic crystal.
mass m. The angular momentum of an arbitrary
system of points with masses mi , which move
with velocities v i , is the sum of vector products ANISOTROMETER
See Magnetic anisotrometer.
i r i × mv i , where r i is the position vector of
the ith mass point relative to an origin of coordi-
nates 0. The angular momentum depends on the ANISOTROPIC MAGNETIC MATERIALS
choice of origin. Magnetic materials having different magnetic
properties along different directions.
ANHARMONICITY, VIBRONIC The most common anisotropic magnetic mate-
See Vibronic anharmonicity. rials have one easy magnetization axis. Along this
axis, high magnetization and low values of the co-
ANHARMONIC VIBRATIONS of crystal lattices
ercive force are observed in soft magnetic materi-
Deviation from the simple harmonic nature of
als and the maximal remanence is realized in hard
atomic vibrations in a crystal, i.e. a nonlinear re-
magnetic materials. The magnetic anisotropy is
lationship between the displacements of atoms
natural for monocrystalline magnetics. The widely
(ions) from their equilibrium positions and the
used polycrystalline anisotropic magnetic materi-
restoring forces.
als are obtained by a variety of methods. One of
While the potential energy of a crystal in the
them is the production of crystallographic texture,
harmonic approximation includes only terms of
second order in the relative atom displacements, i.e. preferable orientation of easy magnetization
taking anharmonic vibrations into account leads to axes of individual crystallites along one common
the appearance of cubic and higher order terms in direction. Thus, in sheets of anisotropic electri-
the potential energy expansion in powers of the cal steel, the crystalline structure is produced by
displacements. Anharmonicity of vibrations in- cold rolling and annealing, and in magnetically
volves interaction between the normal modes (see hard materials it is obtained by powder metallurgy
Normal vibrations) of a harmonic crystal, which which involves pressing in a magnetic field. An-
corresponds to taking into account both elastic other method of obtaining anisotropic magnetic
and inelastic phonon collisions from the quan- materials involves producing magnetic texture by
tum point of view. Anharmonic vibrations are re- means of thermomagnetic treatment or thermome-
sponsible for the thermal conductivity of insula- chanical treatment. In some anisotropic magnetic
tors, the thermal expansion of bodies, the change materials, both the crystallographic and magnetic
in the dielectric function of the crystal with tem- texture are used simultaneously. Thus, e.g., the
perature, IR absorption and Raman scattering, etc. best magnetic properties are realized after ther-
The anharmonic interaction of phonons leads to momagnetic treatment of alloys of the alnico type,
their scattering and decay which, in turn, deter- in which the columnar crystallographic texture is
mines the width (lifetime) and shift of the energy produced by directed crystallization.
64 ANISOTROPIC MEDIUM
where λ and μ are Lamé coef cients. Any devi- of higher orders. The effective elastic moduli in
ation of a tensor Ciklm from the form (1) corre- a polycrystalline material are found by averaging
sponds to the presence of anisotropy of elasticity. over orientations of grains.
In Voigt notation, the tensor Ciklm is replaced by
the matrix cI J (I, J = 1, 2, . . . , 6) which, for the ANISOTROPY OF g -FACTOR
lowest symmetry, has 21 independent elements. A property which reflects the dependence of
The symmetry of interatomic interactions in a the splitting of electronic (nuclear) energy levels
crystal imposes restrictions on the number of in- in a magnetic field (see Landé g -factor) on the mu-
dependent elements of cI J . If central forces act tual orientation of the crystal and the field. It man-
between atoms and every site of a crystal lattice is ifests itself in the angular dependences of spec-
a center of symmetry, six Cauchy relations are sat- tra of magnetic resonance, and is described by the
isfied: c12 = c66 ; c13 = c55 ; c23 = c44 ; c14 = c56 ; second-rank g-factor tensor gpq (p, q = x, y, z)
c25 = c46 ; c36 = c45 , which, for instance, in crys- with unequal principal values. The magnitude of
tals with a heteropolar bond reduce to one rela- the tensor in the principal axis system for the po-
lar angles θ, ϕ is given by g = [(gxx 2 sin2 ϕ +
tion: c12 = c44 . In hexagonal close-packed crys- 2 2 2 2 2 1/2
tals no site is a center of symmetry, and the Cauchy gyy cos ϕ) sin θ + gzz cos θ] . For axial sym-
metry g = (g⊥ 2 sin2 θ +g 2 cos2 θ)1/2 , where g =
relations do not hold even if the interatomic forces ⊥
are central. The deviations from the Cauchy rela- gxx = gyy and g = gzz . Anisotropy is present
tions give an insight into the contribution of an- in paramagnetic centers with a local symmetry
harmonicity to cI J , and about nonpairwise (many- lower than cubic. Ordinarily the tensor is symmet-
particle) forces of interaction. Noncentrality of the ric, with gpq = gqp .
interaction has a strong effect on the temperature
ANISOTROPY OF LIQUID CRYSTALS
dependence cI J = cI J (T ), as well as on the value
Dependence of physical properties of liquid
cI I . The elements of cI J must satisfy a number
crystals on the direction of a measurement.
of relations which are corollaries of the mechani-
In oriented samples the direction of the ori-
cal stability conditions for a crystal. For example,
entational order is specified by the director. If
in a hexagonal close-packed crystal
n , χ , and ε are respectively the refractive in-
cI I > 0 (I = 1, . . . , 6); dex, diamagnetic susceptibility and dielectric con-
2 − c2 > 0; stant along the director, and n⊥ , χ⊥ and ε⊥ are
c11 12 the same quantities in a transverse direction, then
2 > 0;
c33 (c11 + c12 ) − 2c13
n = n − n⊥ , χa = χ − χ⊥ and εa = ε − ε⊥
are referred to as optical anisotropy, diamag-
2 > 0.
c11 c33 − c13 netic anisotropy and dielectric anisotropy, respec-
tively. For example, for thermotropic liquid crys-
Cubic crystals exhibit anisotropy of shift, which tals
n ∼ 0.1–0.4; χa ∼ 10−7 emu, εa ∼ 5–30.
can be characterized by the factor Ashift = 2c44 / The sign of the anisotropy depends on the
(c11 − c12 ). In addition to the shift anisotropy with chemical structure of liquid crystals. For nemat-
the factor Ashift = c44 /c66 , crystals of hexago- ics (see Nematic liquid crystal), χa > 0,
n > 0,
nal, trigonal and tetragonal symmetry have also εa > 0 and εa < 0. For cholesterics (see Choles-
an anisotropy of linear compressibility (with the teric liquid crystal), χa > 0 and χa < 0,
n < 0,
factor Acompr = (c11 + c12 − 2c13 )/(c33 − c13 ) εa > 0 and εa < 0. For smectics (see Smectic
for hexagonal symmetry). Near the polymorphic liquid crystal), χa > 0,
n > 0, εa > 0 and
transformation temperature, some combinations εa < 0. Typical values for lyotropic liquid crystals
of elasticity constants can diminish abnormally are: χa ∼ 10−8 emu,
n ∼ 10−3 . For lyotropic
(softening of normal modes). The defects in a crys- discotics χa > 0 and χa < 0,
n > 0, for lyotropic
tal cause the components of the tensor of effec- calamitics χa > 0 and χa < 0,
n < 0. Because
tive elastic moduli to decrease by several percent. of the anisotropy, the description of elasticity and
When studying crystal distortions localized near viscosity of liquid crystals requires using several
defects, one needs to know elastic moduli tensors coefficients.
66 ANISOTROPY OF MECHANICAL PROPERTIES
ANISOTROPY OF MECHANICAL PROPERTIES fects of a particular type can lead to either their
Differences in mechanical characteristics: elas- annihilation or their transformation into defects of
ticity, plasticity, and durability, along various direc- more complicated types. The kinetics of isother-
tions of a solid. mal annealing yield information on the nature of
There are two types of anisotropy of mechan- the reactions that result in the annealing.
ical properties: the natural or fundamental (pri-
mary) anisotropy, and the induced (secondary) ANNEALING, IRRADIATION
type. The former is determined by the crystal See Irradiation annealing.
symmetry (see Anisotropy of crystals) and can
be easily observed. The latter is acquired during
the process of mechanical and thermal treatment ANNEALING OF SEMICONDUCTOR
of a metal. There are two origins of secondary See Pulsed annealing of semiconductors.
anisotropy: textural (see Texture) and structural.
The textural type arises as a result of reorienta- ANNEALING, RECRYSTALLIZATION
tion of crystallites along preferred crystallographic See Recrystallization annealing.
directions. The structural secondary anisotropy is
caused by the oriented arrangement of grains and
ANNIHILATION OF DISLOCATIONS
subgrains, inclusions of the second phase, and
pores (see also Structural texture of metals). The Merger and mutual destruction of two disloca-
elastic anisotropy for a number of textures can be tions or elements of one dislocation loop with op-
derived analytically (see Anisotropy of elasticity). posite signs (see Dislocation sign).
Anisotropy of plasticity and anisotropy of dura- The annihilation of dislocations leads to the
bility have more involved dependences than the disappearance (mutual compensation) of elastic
anisotropy of elasticity, and are determined exper- e lds of dislocations. If parts of a dislocation with
imentally in mechanical tests. opposite signs come closer together as a result of
dislocation slip in one common plane, then anni-
ANISOTROPY, OPTICAL hilation restores the perfect crystalline structure
See Optical anisotropy. at their meeting point (see Fig. 1). The annihila-
tion of screw dislocations occurs with the restora-
tion of the structure even when they initially were
ANNEALING
not in the same slip plane. Nonscrew dislocations
Heat treatment of materials at elevated tem-
of opposite signs lying in different, but not too
peratures aimed at investigating or improving
widely separated, slip planes form a dipole due
their properties. Material annealing can lead to
to their mutual attraction. Its subsequent annihila-
phase transitions, recrystallization, polygonization,
tion is possible by means of a head-on encounter
homogenization, relaxation of internal stresses,
with the creation of interstitial atoms or vacancies
removal of aftereffects of cold plastic deformation
(strain hardening), annihilation and rearrangement
of defects, and so on. The results of annealing
depend significantly on its kinetics: the rates of
heating and cooling and the time of exposure at
a given temperature. To investigate annealing of
radiation-induced and other defects, use is made
of isothermal annealing (long time exposure at
constant temperature), and isochronal annealing
(exposure during a given time interval at succes- Fig. 1. Annihilation of edge dislocations of opposite
sively increasing temperatures). Various stages of signs, which move in the same slip plane: (a) approach
isochronal annealing correspond to the presence of of the dislocations; (b) restoration of the perfect crystal
defects of different kinds. The annealing of the de- structure upon annihilation.
ANOMALOUS DIFFUSION 67
ANOMALOUS AVERAGES
The values of products of two (or more) cre-
+
ation akσ (or annihilation akσ ) operators of half-
integer spin particles (fermions), averaged over the
lowest level (ground) quantum-mechanical states
(at absolute zero temperature T = 0), or the ther-
modynamic states (at T = 0) of a system. The
anomalous averages were introduced by N.N. Bo-
golyubov (1958) to describe superconducting (su-
perfluid) Fermi systems in the self-consistent eld
approximation. They provide the complex-valued
order parameter (wave function) of a macroscopic
coherent quantum state (see Macroscopic quantum
coherence).
Fig. 2. Annihilation of edge dislocations of opposite In a superconductor at a temperature below T c ,
+ +
signs, which move in neighboring slip planes: (a, c) ap- the anomalous averages of the akσ a−k−σ and
proach of the dislocations; (b) formation of vacancy at a−k−σ akσ types are nonzero. They describe the
the location of annihilation; (d) formation of interstitial condensate of Cooper pairs of electrons with an-
atom at the location of annihilation. tiparallel momenta (k = −k) and spins (σ =
−σ ), which is analogous to a Bose condensate.
In the super uid phases of 3 He at temperatures
(Fig. 2). When dislocations of opposite signs in below 1 K (T < 0.003 K) and under a pressure
different slip planes approach each other and reach (p > 30 atm), the triplet pairing of 3 He atoms
a characteristic distance ra referred to as the anni- with parallel spins (σ = σ ) is possible along with
hilation radius, then the region of the crystal be- the singlet Cooper pairing (σ = −σ ), when the
tween them loses its stability, and annihilation oc- + +
anomalous averages akσ a−kσ and akσ a−kσ
curs dynamically, rather then by diffusion. Stud- do not vanish. In anisotropic crystals with a de-
ies of the dislocation annihilation process on an generate electron spectrum ε(k) = −ε(k + Q)
atomic scale by mathematical modeling provide and flattened, nested regions of the Fermi sur-
the value ra ≈ (5–10)b, where b is the Burgers face separated by the wave vector Q, a spatially
vector. The annihilation of dislocations plays an nonuniform state can arise. This state results from
important role as an effective channel for reduc- electron–hole pairing or a structural Peierls insta-
ing the dislocation density (see Dislocation density bility, and has nonzero anomalous averages of the
tensor) during the annealing process. This annihi- + +
type ak+Q,σ ak,±σ and ak,±σ ak−Q,σ which
lation is accompanied by the emission of sound, describe charge density waves and spin density
and it is one of the most important mechanisms of waves.
acoustic emission.
ANOMALOUS DIFFUSION
ANODIZATION, anode oxidation A diffusion process in which the time evolution
The creation of a layer of oxide on the surface of the concentrations of diffusing particles fails
of a metal at a positive potential during electroly- to follow the Fick law, occurring asymptotically
sis. Anodization is used to change the color, dura- faster or slower than in the normal case.
bility, hardness, electrical conductivity, catalytic As for any diffusion process, anomalous diffu-
activity, and other properties of surfaces of vari- sion is associated with a random walk of particles.
ous metals and alloys. There is a great variety of However, in contrast to an ordinary random walk,
electrolytes, and either direct or alternating current the mean square displacement of a particle in the
can be used. Depending on this, one obtains either time t exhibits the behavior r 2 (t) ∼ t ν , with
a thin (less than 1 μm), or a thick (tens to hundreds ν = 1. The causes of anomalous behavior are as-
of μm) enamel-like coating. sociated with the disorder of the system where the
68 ANOMALOUS DISPERSION OF LIGHT
random walk occurs. For example, anomalous dif- semiconductors, such departures from classic de-
fusion behavior is observed in systems with a dis- pendences (both in sign and in characteristic field
tribution of activation energies, when the distrib- values) are related to the influence of the magnetic
ution of the wait time of the next jump becomes field on the spin subsystem. In the case of non-
very broad and the mean value of this time di- magnetic conductors, the anomalous magnetore-
verges; in a random walk in fractal systems the sistance is caused by quantum-interference cor-
diffusion is slower (ν < 1); in a random walk with rections to the conductivity, their contribution be-
flights, when a particle can travel a great distance ing determined by the coherence length (phase
during a short time (ν > 1), etc. To character- memory time) of the carriers. Such corrections
ize anomalous diffusion, one often does not use arise both at weak localization (causing negative
the power index ν, but rather a related quantity magnetoresistance), and due to interference of
called an anomalous diffusion index that character- Coulomb and impurity scattering when the mag-
izes properties of diffusion and random walk in a netoresistance is positive.
fractal system. In the normal case, for instance, on
any Euclidean lattice, the square of the recession ANOMALOUS MASS TRANSFER
of a particle in random walk for sufficiently long Transfer of inherent and impurity atoms in crys-
time t is proportional to this time: L2 (t) ∼ t . In tals at a rate several orders of magnitude greater
contrast to such universal behavior, the diffusion than that of diffusion. It is observed during pulse
in a fractal system occurs more slowly, L2 (t) ∝ exposure causing plastic deformation with a rate
t 2/(2+θ) , where θ > 0 is the anomalous diffusion greater than 1 s−1 . It results from dynamic recov-
index. The quantity θ for fractal lines is related to ery of nonequilibrium point defects formed dur-
the fractal dimension of the line D: θ = 2D − 2, ing the process of plastic deformation under forg-
although in other cases the quantity θ is some in- ing, explosion, pulse heating (thermoplastic defor-
dependent characteristic of a fractal system. The mation), in a pulsed magnetic field (due to the
index θ for an infinite cluster at the percolation action of ponderomotive forces). The mechanism
threshold is related to well-known critical indices of anomalous mass transfer consists in the cre-
of percolation theory θ = (τ − β)/ν, where τ is ation of interstitial atoms under the nonconserva-
the critical index of the percolation system con- tive motion of dislocation thresholds at screw dis-
ductivity, β is the critical index of the extent of the locations with a velocity higher than the critical
infinite cluster, and ν is the critical index of the one. The interaction of the moving dislocations
correlation length (see Correlation length). with interstitial atoms causes the drift of the lat-
ter. Hence, the anomalous mass transfer can be
ANOMALOUS DISPERSION OF LIGHT treated as diffusion of nonequilibrium interstitial
Dispersion of light characterized by the anom- atoms under the conditions of drift (in the presence
alous dependence of the refractive index n of a ma- of driving forces of transfer processes). The distri-
terial on the frequency ν (wavelength λ) of the bution of the concentration C of diffusing atoms is
light: n decreases as ν grows (λ decreases). The described by the equation
anomalous dispersion of light occurs in the ab- (x − V t)2
C(x, t) = C0 (4πDt)−1/2 exp −
sorbtion bands of materials. 4Dt
V Vx x +Vt
ANOMALOUS ELASTICITY − C0 exp erfc √ ,
See Superelasticity. 4D D 2 Dt
where C0 is the concentration at the instant of time
ANOMALOUS MAGNETORESISTANCE t = 0; D is the diffusion coef cient, and V is the
The dependence of a conductor resistivity on drift velocity in the x direction. The temperature
the external magnetic field, which is not describ- dependence of the anomalous mass transfer rate is
able within the classical picture of electron or- weaker than that of the diffusion coefficient, and
bit bending (see Galvanomagnetic effects, Mag- cannot in general be described by a simple expo-
netoresistance). In magnetic materials: ferromag- nential over a wide temperature range. The criti-
netic and antiferromagnetic metals, semimagnetic cal rate of deformation is defined by the nonlinear
ANOMALOUS PENETRATION 69
dependence of the yield limit on the deformation of the surface layer of machine parts is achieved
rate. Below the critical rate, the energy of exter- by means of electric spark alloying. Taking anom-
nal stresses causing the dislocation slip is insuffi- alous mass transfer regularities into account leads
cient for the formation of interstitial atoms. There- to improving the alloyed surface layer obtained
fore, only dislocation thresholds that create vacan- by laser processing (see Laser technology). Anom-
cies move conservatively. At fast rates, both va- alous mass transfer occurs also at shot-blasting or
cancies and interstitial atoms are created. The lat- ultrasonic treatment of a metallic coating on the
ter prevail at small dislocation densities, so anom- surface of products. The mechanical mixing of
alous mass transfer diminishes with the extent of surface layers of contacting metals under shock
the plastic deformation. It diminishes also when loading known as mechanical diffusion does not
an elevated concentration of excess vacancies is fall into the category of anomalous mass transfer.
created in a crystal before the pulse arrival. The
rise of the speed of dislocation displacement due ANOMALOUS PASSAGE OF X-RAYS, Borrmann
to the swift rise of the external stresses leads to effect (G. Borrmann, 1941)
the diminishing of the vacancy contribution to the Propagation of X-rays in monocrystals along
anomalous mass transfer. The dynamic multiplica- directions satisfying the Bragg law, when the ef-
tion of dislocations is superimposed, and the drag- fective (so-called interference) absorption coeffi-
ging along of interstitial atoms by shock waves cient is considerably smaller than the photoelectric
begins to play a significant role in the anomalous absorption coefficient μ.
This is observed in the case of Laue diffraction
mass transfer. This transfer is faster in metals with
for sufficiently thick crystals at absorption levels
face-centered cubic and hexagonal close-packed
μt 3–5 (t is the sample thickness). The anom-
lattices than it is in those with a body-centered cu-
alous passage of X-rays is the result of multiple
bic lattice. The effect of anomalous mass transfer
scattering by atomic planes of a crystal, and the
acceleration is more pronounced for substitutional
interference of the scattered waves, which leads to
impurities than for interstitial ones. As a result
the creation of a total wave field with a spatially
of the anomalous mass transfer, nonequilibrium
modulated amplitude. If the antinodes of this field
solid solutions can also be created from the compo-
lie to the side of atomic planes, it is absorbed very
nent with low mutual solubility in the solid phase.
weakly (much more weakly than in ordinary inter-
Under the low-temperature deformation of crystals
ference). As a result, the X-rays can travel in crys-
with ionic bonds, other mechanisms of anomalous
tals for relatively great distances (up to 1–10 mm).
mass transfer appear. Thus, rather small deforma- The anomalous passage was first observed in per-
tion of LiF crystals at liquid helium temperatures fect crystals; and has subsequently been seen in
leads to helium penetration into the surface layer mildly distorted crystals. It occurs for the diffuse
of a crystal due to the dragging along of helium background as well, which leads to considerable,
atoms by the dislocation thresholds, and their mo- often asymmetric deformation of the diffuse scat-
tion along the dislocations. Indentation (see Mi- tering of X-rays profile compared to the kinematic
crohardness) of MgO crystals leads to directed case (amplification of central part of the profile
plastic deformation (without the creation and mo- relative to its wings). See also Dynamic radiation
tion of dislocations due to those of crowdions) scattering, Two-wave approximation.
(see also Anomalous diffusion). Engineering ap-
plications of anomalous mass transfer are diverse. ANOMALOUS PENETRATION
Thus, hot forging or rolling of ingots of alloys at of the electromagnetic eld into a metal
a deformation rate greater than the critical one al- This occurs in an external magnetic field, and is
lows considerable acceleration of the homogeniza- caused by the specific dynamics of the conduction
tion of dendrite liquation. Percussion welding in the electrons. The effects of anomalous penetration
solid phase forms welded joints of heterogeneous are subdivided into two groups. The first group,
metals during a time too short for the creation of that of plasma or collective phenomena, comprises
intermetallic compounds responsible for the em- the aggregate of weakly decaying waves in the de-
brittlement of the weld. The effective hardening generate electron plasma of a metal (see Helicon,
70 ANOMALOUS PENETRATION OF ULTRASOUND
Magnetic plasma waves, Cyclotron waves, Quan- provide bursts of anomalous penetration at depths
tum waves, Doppleron, etc.). The second group in- that are multiples of the displacement along the
cludes ballistic effects of anomalous penetration normal to the sample surface, extremal over the
which are caused by a trajectory transfer of an cyclotron period. This mechanism of anomalous
electromagnetic wave from the skin depth layer δ penetration is related to a sharp rise of the Fourier
into the bulk metal by numerous groups of elec- component of the conductivity due to Doppler-
trons, involving individual motions of the charged shifted cyclotron resonance. The following effects
particles and quasi-particles. Because of the skin- arise from the drift type of anomalous penetration:
effect, the most effective interaction of the elec- focusing of electrons from the vicinity of the ref-
trons with the electromagnetic field occurs for erence point of the Fermi surface, bursts on the
those electrons which move along the wavefront, helical paths of the Fermi surface boundary cross
i.e. parallel to the surface of the metal (so-called sections, trajectory transfer by effective electrons
effective electrons). Electrons moving at large an- on an open Fermi surface, etc. Finally, in a mag-
gles with the surface spend less time in the skin netic field perpendicular to the metal surface, the
layer, and thus are less effective in gaining en- ballistic effects of anomalous penetration are as-
ergy from the wave. In a magnetic field the direc- sociated with the magnetic focusing of drifting or
tion of the electon velocity changes periodically ineffective, electrons, the so called Gantmakher–
due to the cyclotron motion. Therefore, the most Kaner effect.
intense interaction with the electromagnetic wave The various phenomena of anomalous pene-
occurs in the vicinity of points of electron trajec- tration of trajectories of weakly damped waves
tories where the velocity is parallel to the surface are closely interrelated, and can transform into
(so-called effective points). Due to the cyclotron each other when the frequency ω or the magnetic
motion the number of these points is infinite, and field B is changed. The physical pattern of anom-
under the anomalous skin-effect δ D (D is the alous penetration depends on the nature of the re-
characteristic length of a trajectory), some of these flection of electrons from the metal surface. The
points are located outside the skin layer. Their anomalous penetration of radio waves manifests
existence is related to resonant excitation of the itself experimentally as a radio-frequency size ef-
electron collective motion. When moving inside fect. It is used to investigate Fermi surface parame-
this layer the electron acquires a velocity increase ters, the temperature dependence and anisotropy
v and creates the current
j = −e
v, where e of the mean free path, and mechanisms of surface
is the magnitude of the electronic charge. At the electron scattering.
next effective point lying in the depth of a metal,
the electron again moves parallel to the sample ANOMALOUS PENETRATION OF
surface and reproduces the current increment
j ULTRASOUND into a metal
there. This is what constitutes anomalous pene- A significant nonlocal phenomenon appearing
tration of the ballistic type: bursts of current and at high frequencies ω, when the damping depth
magnetic field spaced by the distance D arise in- of a sound wave (calculated within the framework
side the metal. The various types of electron tra- of a local theory), L ∼= vF /ω, is smaller than the
jectories lead to different mechanisms of anom- mean free path of the carriers, l = vF τ (the lat-
alous penetration. ter is the penetration depth of electron density os-
In a magnetic field B parallel to a metal sur- cillations caused by sound, and interacting with it
face, due to the complicated structure of the Fermi in a conductor; see Anomalous skin-effect). Thus,
surface, there exists anomalous penetration along the anomaly criterion is ωτ
1 for the asymptote
the chain of (closed) electron trajectories with ex- of the acoustic field u(x, t) in the bulk of a metal
tremal diameters Dext (see de Haas–van Alphen (x
L), when it is a relatively weak but slowly
effect). The chain of trajectories from the central damped quasi-wave (non-exponentially damped in
section of the Fermi surface is created in a slightly the collision-free limit) propagating with the elec-
tilted B. At a greater inclination of B there oc- tron velocity vF .
curs a kind of focusing of effective electrons drift- The appearance of electron-acoustic quasi-
ing into the interior of the metal. These electrons waves is associated with the degenerate statistics
ANOMALOUS PHOTOVOLTAIC EFFECT 71
of charge carriers in metals, i.e. with the pres- unambiguously, the generating ultrasound pulse
ence in the electron spectrum of a Fermi velocity in width must be much less than the Larmor diam-
every direction. Anomalous penetration of ultra- eter. Otherwise, the effect appears as oscillations
sound occurs for an arbitrary electron dispersion of the sound speed as the magnetic field changes.
law, and for any polarization of the sound. In the The shape of the forerunners repeats that of the
typical case of a convex Fermi surface and mono- generating pulse.
chromatic longitudinal vibrations, we have The analogue of the anomalous penetration of
ultrasound is the anomalous penetration of an al-
s2 ωx
u(x, t) ∼
= Cu(0, t) ln−2 ternating longitudinal (i.e. normal to the interface)
vF0 ωx vF0 electric field. Though its main part decays expo-
iωx x nentially into the metal for a distance that ap-
× exp − ,
vF0 l0 proximates a Debye radius, the nonuniformity of
vF0 the electron density remains down to macroscopic
x .
ω depths ∼= l. Owing to the deformation interaction
Here C is a numerical factor of the order of unity of a nonequilibrium electron gas with the ionic lat-
determined by the form of the electron spectrum, tice vibrations, this leads to a new (linear in the
and the nature of their reflection from the metal applied field amplitude) mechanism of electron-
boundary (x = 0); s is the sound velocity, and acoustic transformation, where efficiency grows
vF0 = (vx )max is the electron velocity at the refer- as ωτ increases. Such an effect related to the
ence point on the Fermi surface. The quasi-waves anomalous penetration of ultrasound was first ob-
are generated also by other particular values of vx served in a tin single crystal at liquid helium tem-
on the Fermi surface; the power law of asymptotic peratures.
decay u(x) ∼ x −3/2 as l → ∞ is characteristic for
an internal extremum of vx on a nonconvex Fermi ANOMALOUS PHOTOCONDUCTIVITY
surface. Photoconductivity which is characterized by
With respect to short (compared with τ ) acous- unusual behavior, such as: independence of the
tic pulses, the anomalous penetration of ultrasound photocurrent from the photoexcitation intensity
is revealed in the formation of an acoustic sig- over a wide range, marked dependence on the
nal “forerunner” by electrons. In the absence of spectral content of the light, and long-term relax-
a magnetic eld the quasi-wave precursor trav- ation of the photoconductivity after extinction of
els with the velocity vF0 . Hence, the observation the light. Anomalous photoconductivity was dis-
of the anomalous penetration of ultrasound in the covered and investigated in detail on layers of
pulse mode could make it possible, by varying the amorphous selenium treated in mercury vapors.
geometry and width of samples, to measure di- It is explained by the existence of deep traps for
rectly both electron velocities and local values of mobile particles with small probabilities of both
their mean free paths over the entire Fermi surface thermal excitation of electrons to the conduction
of the metal under study. Observed in the magnetic band, and recombination with a free hole. The
field B in the pulse mode is the ultrasound field traps are assumed to be clusters containing about
transfer by conduction electrons along the electron 103 atoms, and localized in a region near the sur-
orbit chain (the analogue of bursts of the electro- face of the sample. Anomalous photoconductiv-
magnetic field). Having effectively interacted with ity is sometimes referred to as the phenomenon
the ultrasonic field in the region where vx = s, the of current reduction under illumination, but in this
fast electron carries away the information on the case the term negative photoconductivity is more
sound pulse to create its “image” at its path loca- often used.
tions where vx = s again. As a result, the forerun-
ners and afterrunners of the sound signal arise, the ANOMALOUS PHOTOVOLTAIC EFFECT
distance between them (in the absence of electron Appearance of a voltage Vp during illumina-
drift along x) being a multiple of the extremal di- tion of a sample by light of energy in excess of the
ameter of an electron orbit. To identify the effect band gap Eg .
72 ANOMALOUS SKIN-EFFECT
Under ordinary mechanisms of the photoelec- value of the impedance in the anomalous skin-
tromotive force: eVp Eg (see Photoelectric phe- effect depends on the character of electron scat-
nomena). Anomalous photovoltages are observed tering at the surface: under diffuse reflection, and
in semiconductor films (e.g., CdTe), in lithium nio- the impedance is 12.5% greater than under specu-
bate crystals, selenites and other materials exhibit- lar reflection. The anomalous skin-effect has been
ing unipolarity along some direction. The values observed in pure monocrystals of metals at low
of Vp in the anomalous photovoltaic effect can be (helium) temperatures in the frequency range from
as high as many kilovolts. This effect can occur several to hundreds of GHz. It was discovered by
either with the absence of a center of symmetry H. London (1940) in tin.
of the crystal lattice, or with a certain nonunifor-
mity of the distribution of defects in the sample.
ANOMALOUS THERMAL EXPANSION
In the first case, the appearance of a photovoltage
Excessive thermal expansion of the lattice sur-
provides for the transfer of free current carriers in
face (in the first 1–2 monolayers), which takes
one of the directions along the unipolar axis during
the process of photoexcitation (such a mechanism place in semiconductors, metals and simple chem-
was discovered experimentally in lithium niobate ical compounds. For example, the ratio of the lin-
crystals), or is responsible for the unipolar charac- ear thermal expansion coefficient (normal and tan-
ter of current carrier scattering by oriented dipoles gential) of the surface to its bulk value for a Si
created by lattice defects. In the second case, the crystal is approximately 8, for Ge: 4, for GaAs: 3,
cause of the anomalous photovoltaic effect can for NaF and LiF: 2, for W: 2–3. Anomalous ther-
be the formation of interconnected microscopic mal expansion was established by various meth-
sources of electromotive forces, e.g., semiconduc- ods: low-energy electron diffraction, scattering of
tor junctions. This mechanism has been used to ex- H and D atoms, spin polarization of reflected low-
plain the anomalous photovoltaic effect in semi- energy electrons. This expansion does not harm
conductor films. the crystal surface (for moderately high temper-
atures) due to the relaxation (compression) of the
ANOMALOUS SKIN-EFFECT surface layers of the lattice, and the presence of
A variety of phenomena of electromagnetic monatomic steps and other defects on the crystal
field damping in metals (and plasma) under the surface. For instance, under heating to 500 K the
condition that the mean free path l of conduc- lattice parameter of bulk LiF will rise by 1.5%,
tion electrons considerably exceeds the skin layer while the increase on the surface is 3%. The extra
depth, δ. In contrast to the normal (classical) skin- expansion (1.5%) of the surface is compensated by
effect when l δ and electromagnetic absorption surface relaxation or structural defects.
is determined by collisions of conduction elec-
trons with scatterers, the mechanism of absorp- ANOMALY, KOHN
tion in the anomalous skin-effect has the collision- See Kohn anomaly.
free character of Landau damping. The screening
of the external wave by the conductor in this
effect is caused by the sliding (so-called effec- ANTHRACENE, C14 H10 (fr. Gr. ανϑρας , coal)
tive) electrons that move almost in parallel to An aromatic conjugated, condensed three ring
the surface at angles on the order of δ/ l. The hydrocarbon with molecular weight 178.22.
salient features of the anomalous skin-effect are A crystal of anthracene (specific weight
the non-exponential decay of the amplitude of an 1.25 g/cm3 ; T melting = 217.8 ◦ C; T boiling =
electromagnetic wave deep inside the metal, inde- 355 ◦ C) consists of anthracene molecules bonded
pendence of the skin layer depth δ on l (in the nor- by weak van der Waals forces; it is a typical rep-
mal skin effect δ ∝ l −1/2 ), a different frequency resentative of molecular crystals. Crystals of an-
dependence of the surface impedance Za ∝ ω2/3 thracene belong to the prismatic monoclinic sys-
(whereas Zn ∝ ω1/2 ) and the change of the re- 5 ) and two mole-
tem with space group P 21 /c (C2h
lation between its real and imaginary parts. The cules per unit cell. The monoclinic axis b, the
ANTIFERROELECTRICITY 73
crystal symmetry axis, coincides with the mid- Concerning their mode of application metal-
dle axis of the refractive index ellipsoid. Crys- lic coatings are subdivided into electroplated, hot-
tals of anthracene can be produced by sublima- dip (of the melted metal) and metallized ones.
tion, as well as by crystallization from the melt The nomenclature of metallized coatings is quite
(Bridgeman’s method), or from solution. Using diverse, and is differentiated by the deposition
monocrystal growth by sublimation one obtains method (electric arc, gas-flame, plasma, and gas-
thin (from a fraction to hundreds of microns) col- thermal metallization). To create coatings or cor-
orless foils, their blue uorescence being readily rosion layers on the metal surface, one can also
observable under ordinary illumination (quantum use electrolysis from melted salts, deposition from
yield of fluorescence close to unity). The plane gas-vapor phase, and implantation of metal ions or
of these foils contains the a and b axes of the noble gases. To improve the operating protective
crystal. Anthracene is easily soluble in hot ben- properties of the surface layer, use is often made
zene, poorly soluble in alcohol, ether, and is in- of diffusive heating that provides for diffusion of
soluble in water. It is crystallized from solution coating components into the underlying metal.
to form colorless prisms. Using the Bridgeman The greatest scope of application (about 70%)
method, one can obtain large crystals with any de- involves protective polymeric (and film) lacquer
veloped plane. Crystals of anthracene begin to ab- paint coatings, protective fettle coatings, protec-
sorb light in the near ultraviolet region. The Davy- tive enamel and other inorganic nonmetallic coat-
dov splitting of the 0–0 band of the first electron ings differing in their chemical composition, tech-
transition is equal to 230 cm−1 . Under irradia- nology of application and range of use. Compos-
tion with ultraviolet light anthracene can be dimer- ite metal-polymer coatings, inhibited metallic and
ized or oxidized to anthraquinone. A crystal of an- nonmetallic coatings that provide better reliability
thracene is an organic semiconductor, and it is the and protection duration have gained wide use as
basis of several large groups of dyes. Many valu- well.
able natural substances are anthracene derivatives
(alizarin, cochineal, chrysophanic acid, etc.). An- ANTIEMISSION MATERIALS
thracene lm s produced by spraying can be used Metals, alloys, chemical compounds exhibiting
as nonsilver carriers for information recording. low intensity of electron emission or its absence.
The antiemission materials include gold, platinum,
ANTICORROSIVE COATINGS titanium, zirconium, carbon, carbides of tungsten
Layers of material applied to metal parts for and molybdenum, titanium silicides, and others.
corrosion protection. They are used for coating the working surfaces of
The protection is assured either by insulation instruments, as well as their nonoperational sur-
of the construction metal from the corrosive envi- faces to suppress electron emission during their
ronment, or due to the protective or inhibitory ac- useful lifetime.
tion of the coating upon the metal being protected.
In terms of composition, anticorrosive coatings are ANTIFERROELECTRICITY (fr. Gr. αντ ι, against,
classified into nonmetallic, metallic and compos- and ferroelectricity)
ite; and in terms of their mode of action they can A set of phenomena related to the appearance
be insulating, protecting and inhibiting. Concern- of ordered antiparallel electric dipole moments in
ing their durability, coatings are protective (for a the crystal lattice of antiferroelectric crystals.
given service life) and preserving (temporary pro- The spontaneous polarizations (see Ferroelec-
tection). Some important properties of the coatings tricity) of neighboring ionic chains or dipole groups
are their chemical resistance, porosity (permeabil- of the same type are oriented in opposite direc-
ity), adhesion to the basic metal, and deformabil- tions, so the total polarization is zero. Phases hav-
ity. The tin, lead, copper, nickel, chrome, and ti- ing the free energy (see Thermodynamic poten-
tanium coatings are classified as corrosion resis- tials) close to that of pyroelectric and ferroelectric
tive (insulating) metallic coatings, while zinc, cad- states are usually considered favorable for forming
mium, aluminum coatings on steel are protective antiferroelectric phases, because the phase tran-
types. sitions to paraelectric and ferroelectric states are
74 ANTIFERROELECTRICS
magnons. The optical absorption spectra were also The values of T N of most antiferromagnets lie be-
investigated with binary fluorides like KMnF3 , low room temperature.
CsMnF3 . Brillouin scattering of light by magnons
was observed in antiferromagnetic FeBO3 and ANTIFERROMAGNETIC DOMAIN
CoCO3 . In uniaxial CoF2 the piezomagnetic ef- A region of an antiferromagnet with a con-
fect was discovered (see Piezomagnetism), and stant value of the antiferromagnetism vector L =
in Cr2 O3 the magnetoelectric effect. In uniaxial M 1 − M 2 . The directions of L in different antifer-
crystals with anisotropy of the easy-plane type romagnetic domains are different. Antiferromag-
(α-Fe2 O3 , MnCO3 , CoCO3 , NiF2 ) the weak fer- netic domains are separated by antiferromagnetic
romagnetism due to the Dzialoshinskii interaction domain walls. It is not always true that the division
was discovered. Of special interest among sub- of antiferromagnets into domains results in a de-
stances with weak ferromagnetism are the or- crease of the energy of the demagnetization elds.
thoferrites (YFeO3 and others) where orienta- The appearance of these domains either in the in-
tional phase transitions (the change of the an- termediate state of magnets, or for antiferromag-
tiferromagnetic ordering axis) are observed, and nets with a weak ferromagnetism, is thermody-
transparent FeBO3 with T N above room tem- namically favorable. In other cases, these are ther-
perature which displays a magnetoelastic inter- modynamically unfavorable. In addition in some
cases, e.g., for domains with zero magnetization
action. The strongest magnetoelastic interaction
and differing in the sign of L (antiferromagnetic
among antiferromagnets is observed in α-Fe2 O3 .
domains with 180◦ neighbors), states in different
This is the first compound to exhibit a wide gap
domains, within the macroscopic framework, are
in the spin wave spectrum that equals
E =
considered identical because they differ from one
γ (2B ME B E )1/2 and is caused by the effective
another only in the redesignation of the magnetic
field of magnetoelastic anisotropy B ME (B E is
sublattices. This is a substantial distinction of the
the exchange field) (see Magnetoacoustic reso-
properties of antiferromagnetic domains in com-
nance). Some antiferromagnets contain more than
parison to, e.g., those of ferromagnets with differ-
two magnetic sublattices. For example, uranium
ent directions of magnetization.
oxide (UO2 ) is a four-sublattice antiferromag- The presence of antiferromagnetic domains in
net characterized by three antiferromagnetism vec- real antiferromagnets can be linked with the fact
tors that form an orthogonal vector triplet. There that a magnetic phase transition to an antiferro-
is a group of antiferromagnetic semiconductors magnetic state due to nonuniformity can occur in-
(chalcogenides of Mn, Eu, Gd and Cr) with very dependently in various finite regions of a crystal.
strong magnetooptical effects. If these regions differ in the sign of L, then the
The low-dimensional antiferromagnets are of spreading of the transition over the whole sam-
particular theoretical interest. The already men- ple causes antiferromagnetic domains to appear.
tioned Fe and Co fluorides as well as some binary In addition, the division of antiferromagnets into
fluorides BaCoF4 , Rb2 CoF4 are two-dimensional domains can reduce the energy of elastic strain,
antiferromagnets. The well-known metal-oxide or (at moderate temperatures, near the Néel point)
compounds displaying properties of high-tempera- become thermodynamically favorable due to the
ture superconductivity (La2 CuO4 , YBa2 Cu3 O7−δ ) contribution of entropy to the free energy of the
in their insulating phase are many-sublatticed antiferromagnet. These domains significantly af-
quasi-two-dimensional antiferromagnets. One- fect such properties as the electrical conductivity
dimensional antiferromagnets include KCuF3 , of a metal antiferromagnet, and sound absorption
CuCl2 , RbNiCl3 and others. Especially strong in- in some antiferromagnets. The presence of antifer-
teractions of electron and nuclear spin system os- romagnetism hinders the observation of a series of
cillations were observed in KMnF3 , MnCO3 , and effects sensitive to the sign of L, e.g., piezomag-
CsMnF3 . Anhydrous sulfates of copper and cobalt netism. Antiferromagnetic domains of the non-
as well as CoF2 exhibit antiferromagnetic order- 180◦ kind (so-called T-domains or twin domains)
ing induced by a magnetic field at temperatures are accompanied by a distortion of the crystal lat-
above T N due to the Dzialoshinskii interaction. tice, thus providing a way to detect, for example, a
76 ANTIFERROMAGNETIC METALS
than what could be expected from the magnetic antifriction materials are manufactured by melting
dipole–dipole transitions alone. Exciton–magnon source components with subsequent casting. As to
excitations are observed in the absorption spectra. the structure type, they are subdivided into those
with harder inclusions of intermetallic compounds
ANTIFERROMAGNETISM VECTOR placed in a relatively light matrix (e.g., aluminum
A macroscopic parameter of antiferromagnets bronze, tin bronze, tin babbitt), and those with
and ferrimagnets; the order parameter at the mag- soft inclusions with metallic structural compo-
netic phase transition to the antiferromagnetic nents in a stronger solid solution matrix (e.g., lead
state (see Antiferromagnetism). The vector of an- bronze, aluminum alloys containing tin, lead and
tiferromagnetism is defined as such a linear com- cadmium). A typical feature of baked (metallo-
bination of the magnetization vectors M α of in- ceramic) antifriction materials is the presence of
dividual magnetic pores (8–27% of the volume) which are impreg-
sublattices, that it differs from
a sum M = kα=1 M α which defines the total nated with a fluid lubricant. The metallic base of
magnetization, where α = 1, 2, 3, . . . , n, and n is such materials is iron, steel, copper, bronze, and
the number of sublattices. To describe the atomic nickel. In addition to these, baked materials con-
magnetic structure, n − 1 antiferromagnetism vec- tain components (graphite, sulfides, selenides and
tors Li are introduced. The expression for Li others) to serve as a dry lubricant. Baked antifric-
depends on the magnetic structure, e.g., in the tion materials in most cases have higher tribotech-
simplest two-sublattice antiferromagnet (or ferri- nical properties than the ones obtained by casting,
magnet), only one such antiferromagnetism vec- and can provide unique properties unattainable in
tor exists, L = M 1 − M 2 . Sometimes, normalized cast alloys.
antiferromagnetism vectors are used, l = L/|L|,
l 2 = 1. It is possible to write dynamic equations ANTIMONY (Lat. stibium), Sb
for the antiferromagnetism vector to define linear A chemical element of Group V of the periodic
and nonlinear spin waves in antiferromagnets and system with atomic number 51 and atomic mass
ferrimagnets (see Landau–Lifshits equation). 121.75. It is related to As and Bi in its proper-
ties. Natural antimony has 2 stable isotopes Sb121
ANTIFRICTION MATERIALS (57.25%) and Sb123 (42.74%). Important radioac-
Materials having low values of the friction co- tive isotopes are Sb122 (T1/2 = 2.8 days), Sb124
efficient. (T1/2 = 60.2 days), Sb125 (T1/2 = 2.0 years).
Besides low values of the friction coefficient Electronic configuration of the outer shell is
(0.004–0.10 with lubricant, and 0.12–0.20 with- 5s 2 5p3 . Successive ionization energies are 8.64,
out lubricant), antifriction materials are charac- 16.5, 25.3, 44.0, 55.4 eV. Atomic radius is
terized by considerable wear resistance, adequate 0.161 nm; radius of Sb3+ ion is 0.090 nm, of Sb5+
mechanical strength and plasticity, corrosion re- 0.062 nm, of Sb3− 0.208 nm. Work function is
sistance, and absence of sticking. These materi- 4.08 eV. Oxidation states are −3, +3, +5. Elec-
als slightly wear the adjoining surfaces. Depend- tronegativity is ≈2.0.
ing on the operating conditions antifriction mate- Antimony is a silvery-white metal with strong
rials with special properties are created to be used brilliance. In a free form antimony exists in sev-
with lubrication by oil or water, without a lubri- eral crystalline and amorphous modifications. Un-
cant, in the atmosphere or in vacuo, at a high tem- der normal conditions, the rhombohedral crys-
perature, in chemical environments, etc. Classified talline modification is stable with parameters a =
among antifriction materials are the following: 0.45064 nm and α = 57.1◦ . Under increased
cast materials (babbitts, bronzes, and brasses); pressure the cubic and hexagonal modifications
baked ones manufactured by powder metallurgy, appear. Besides, three amorphous modifications
including composite materials on the base of iron are known: yellow, black and explosive ones.
or bronzes; combined (mostly metal-fluoroplastic) The rhombohedral modification is characterized
materials; antifriction cast irons and antifriction by gray star-like crystals. For this modification
steels (copper steel, graphitized steel). The cast the density is 6.690 g/cm3 ; Tmelting = 903.5 K,
ANTISCREENING FACTOR 81
Tboiling = 1907 K. Heat of melting is 20 kJ/mole, two-layer coating films almost totally eliminate re-
heat of evaporation is 124 kJ/mole; specific heat e ction of light from the surfaces of the optical el-
cp = 25.2 kJ·mole−1 ·K−1 . Thermal conductiv- ements. Note that this is true in a rather narrow
ity is 18.84 W·m−1 ·K−1 , temperature coefficient spectral range only. To attain antireflection protec-
of linear expansion of polycrystalline antimony tion over a broad spectral range triple and multiple
is 9.2·10−6 K−1 (at 273 K), 10.3·10−6 K−1 layer coating systems are commonly used. Antire-
(at 873 K). Debye temperature is 204 K. Brinell flection coating plays a very important role in low-
hardness of technical grade antimony is 320 to ering optical losses in multilens optical systems, in
580 MPa; of zone refined antimony it is 260 MPa. laser technology where coating makes it possible
Shear modulus is 20 GPa, normal elasticity mod- to lower feedback losses from the optical cavities,
ulus is 56 GPa. in opto-acoustic instruments, etc.
Antimony is a diamagnet; magnetic suscepti-
bility of polycrystalline antimony is 0.81·10−9 (at ANTIRESONANCE, MAGNETIC
293 K); electrical resistivity 0.39 μ·m (273 K), See Magnetic antiresonance.
temperature coefficient of electrical resistivity is
5.1·10−3 K−1 . Compounds of antimony with in- ANTIRESONANCE, OPTICAL
dium, gallium, aluminum, cadmium and other See Optical antiresonance.
metals (so-called antimonides) possess semicon-
ductor properties, with indium antimonide InSb ANTI-SCHOTTKY DEFECT
possessing the highest carrier mobility among all A two-component defect in crystals of binary
known semiconductor materials. Antimony is also (MX) or more complex composition. Its compo-
used as a component of various alloys, it is con- nents are IM and IX , the amounts of interstitial
tained in lead storage battery plates (4 to 12% by atoms of chemical elements M and X which them-
mass). Antimony oxide Sb2 O3 is introduced into selves constitute the regular lattice of the crystal.
the composition of glasses characterized by a low Anti-Schottky defects can be equilibrium stoichio-
refractive index. metric defects arising at elevated temperatures,
e.g., because of “vaporization” of M and X atoms
ANTIREFLECTION COATING OF OPTICAL from the surface to the inside of the MX crystal.
ELEMENTS The creation of anti-Schottky defects is hampered
A means of increasing light transmission of op- by the necessity of inserting the larger M and X
tical elements by coating their surfaces with non- atoms into the smaller-sized interatomic “cavities”
absorbtive lm s. The optical paths through such (i.e. interstitial sites). These defects have the name
films should be comparable to the radiation wave- anti-Schottky because a Schottky defect is a va-
length. Without an antireflection coating the loss cancy formed by transferring an atom from the
of light during reflection from their optical sur- bulk of the crystal to its surface.
faces may reach 10% or more. The antireflec-
tion effect involves interference of light beams re- ANTISCREENING FACTOR
flected from the front and back sides of the di- A parameter characterizing antiscreening (the
electric layers that form the coating. Provided the phenomenon which is inverse to electric charge
angles of incidence remain close to normal the screening).
antireflection coating has a maximum effect. The If P0 is a response arising as a result of some
thickness of the thin coating film should be an odd influence in the absence of a medium, and P is the
number of quarter wavelengths, and the refractive same response in the presence of the medium, then
index of the film, n2 should satisfy the relation one can write P = P0 (1 − γ ). If the dimension-
n22 = n1 n3 , where n1 and n3 are the indices of less parameter γ is less than zero it is commonly
refraction of the two media adjacent to the film referred to as an antiscreening factor. The Stern-
(the first is often air, n1 = 1). For optical elements heimer antiscreening factor (R.M. Sternheimer,
made of glass with a low value of n2 a single layer 1951) γ S describes the enhancement of an in-
antireflection coating is often insufficient, whereas tracrystalline electric field gradient at a nucleus
82 ANTISITE DEFECT
due to the polarization of closed atomic shells by anti-Stokes lines depends on the initial level popu-
external electric charges. The value of |γ S | can be lation (excited) and decreases with a lowering of
rather large (50–100). the temperature. The presence in a spectrum of
Stokes and anti-Stokes pairs of lines, symmetric
ANTISITE DEFECT relative to exciting one, is often used to discrim-
An atom of one sublattice of a two- or mul- inate between the processes of Raman scattering
ticomponent crystal, occupying a site of another
and luminescence, while their intensity ratio char-
sublattice; actually, a substitutional atom, denoted
acterizes the scattering medium temperature.
as XM (atom X at a site of sublattice M). Adjacent
defects XM and MX represent an antistructural de-
fect pair. For unequal concentrations of MX and ANTISTRUCTURAL DEFECT
XM the antisite defects determine an excess of A defect in a binary or more complex crystal
component M or X, i.e. a deviation from stoichio- containing two chemical elements M and X.
metric composition. The defect consists of two components – sub-
stitutional atoms MX (atom M at site X) and XM
ANTI-STOKES LUMINESCENCE (atom X at site M). The formation of antistructural
Photoluminescence, or luminescence at wave- defects in a nonmetallic crystal can be accompa-
lengths less than that of the exciting light, which nied by changing the charge state of its compo-
is in contradiction with Stokes’ rule. nents. For example, atoms M and X in the normal
Anti-Stokes luminescence is common in the re- positions at their “own” sites MM and XX can be
gion of overlap between the excitation and the lu- neutral, while as components of an antistructural
minescence wavelengths. The emission of a quan- defect, if M is a metal, MX tends to become an ac-
tum with an energy exceeding that acquired in the
ceptor and X (metalloid) a donor XM . Antistruc-
absorption occurs due to the additional involve-
tural defects either exist in the equilibrium state
ment of the thermal motion of the luminescent sys-
at elevated temperatures (e.g., in crystals Bi2 Te3 ,
tem. This results in the Vavilov law, which reflects
Mg2 Sn, and others) or are created as nonequilib-
the fact that the probability of such a process de-
creases rapidly as the involved energy rises, and rium metastable defects, e.g., as a result of nuclear
it depends on the thermal motion intensity (tem- radiation (see Antisite defect).
perature). Of greater importance are the funda-
mental violations of Stokes’ law when the emit- ANTISYMMETRIC VIBRATIONS of crystal lattice
ted luminescence lies in the region of wavelengths Normal vibrations which reverse their sign un-
less than that of the exciting light. Such effects der the application of an appropriate symmetry op-
are possible by making use of several exciting eration (see Crystal symmetry). In ionic crystals
light quanta (most often, of two) in the emission antisymmetric vibrations involve a change of the
process, by the simultaneous absorption of two electric dipole moment.
photons by one electron at a high excitation inten-
sity, by sequential absorption of two quanta by the ANTISYMMETRY of crystals, Shubnikov
electron, and by pooling the energy of two excited symmetry
electrons to produce the emission (cooperative lu- One of the particular cases of colored symme-
minescence), etc.
try (hypersymmetry). In ordinary crystal symme-
ANTI-STOKES RAMAN LINES try, only spatial (position or orientation) symmetry
Lines in the spectrum of Raman scattering of is considered, while in hypersymmetry, another at-
light that are located on the high-frequency side of tribute (conventionally, color or spin direction) is
the exciting line. assigned a value at every point of the crystal. If
Anti-Stokes light scattering lines are caused by the additional attribute can take on only two al-
annihilation of the excitation quantum in the sys- ternative values (black and white), such a partic-
tem by the inelastic scattering of an incident pho- ular case of hypersymmetry is referred to as anti-
ton on it (a phonon, magnon, plasmon, etc., is typ- symmetry, in other words as black and white sym-
ically annihilated in a crystal). The intensity of metry. If there are many values then one speaks
ARMCO IRON 83
about color symmetry. Another term for antisym- many types). The permutation conditions for the
metry is Shubnikov symmetry after A.V. Shub- latter have the form Pj2k ψ = exp(iπα)ψ , and at
nikov who introduced the concept. In analogy to α = 0, 1 one can speak about an in-between statis-
the ordinary crystal symmetry groups, there exist tics.
122 point, 36 translational and 1651 space Shub- The ambiguous nature of anyon wave functions
nikov groups. All of them have been derived and results in the fact that the quantum-mechanical
classified, a system of notation has been devel- problem of many anyons cannot be solved even
oped, and they are used in concrete physical stud- for the simplest case when there is no interaction.
ies. Examples of the physical realization of crys- The usual procedure of seeking solutions in terms
tal antisymmetry can vary. Treated as the color of single-particle wave functions is not applicable
change operation may be, for example, a change here.
in the chemical type of atoms in an ordered binary The statistics of all elementary particles is de-
alloy, or in the sign of the displacement of atoms termined by their spin in accordance with the Pauli
and of the electric charge, and others. Solid state principle. Therefore (since real space is three-
physics makes the widest use of this approach for dimensional), elementary particles with an inter-
the Shubnikov classification of the magnetic sym- mediate statistics cannot exist in free space; this
metry of crystals, where the color change oper- statistics can be valid only for quasi-particles in
ation has the sense of the time inversion opera- quasi-two-dimensional and quasi-one-dimensional
tion R, and, as a consequence, the inversion of the systems. In particular, the elementary excitations
magnetic moments and magnetic elds (see Mag- in the fractional quantum Hall effect (see Quantum
netic symmetry group). Operations of antisymme- Hall effect) obey intermediate statistics, i.e. they
try, which occur in magnetic crystal symmetry and are anyons. At a filling factor ν = 1/q where q is
are combinations of rotation or translation with R, an odd integer, the statistical parameter of these
are referred to as antirotations and antitransla- excitations is α = ν. The appearance of the ef-
tions. fect at filling factors ν = p/q kept constant is ex-
plained quite well in the framework of so-called
ANYONS (fr. any) hierarchic theory. This implies that there appear
Particles that obey a statistics which is interme- excitations over a “primary” electron gas that rep-
diate between the Bose (see Bose–Einstein statis- resent the anyon gas. Other excitations can arise
tics) and Fermi (see Fermi–Dirac statistics) types. which also represent the anyon gas but with an-
Such particles can exist in two-dimensional space other statistical parameter, and so on infinitely.
since the topology of this space admits the pres- One can demonstrate that under particular condi-
ence of indeterminate wave functions for (identi- tions an anyon gas is a superconductor. As far as it
cal) particles a with strong short range repulsion is known high-temperature superconductivity is of
(hard core). These functions obey the permutation a quasi-two-dimensional nature, and an hypothesis
rules Pj k ψ = exp(iπα)ψ , where Pj k means the about an anyon mechanism of superconductivity
permutation of j th and kth particles, and α is any has been suggested, but there is no experimental
real number which is sometimes called a statistical confirmation of it. See also Anderson mechanism
parameter. The value α = 0 corresponds to bosons of high-temperature superconductivity.
and α = 1 denotes fermions (see Second quantiza-
tion), while intermediate values are indicative of APW METHOD
an intermediate statistics. The possibility of the See Augmented plane waves method.
existence of an intermediate type statistics was
first pointed out by J.M. Leinaas and J. Myrheim ARMCO IRON (ARMCO is an acronym of
(1977), and the term “anyon” was introduced by American Rolling Mill Corporation)
F. Wilczek (1982). Intermediate statistics can also Highly purified iron with high plasticity, en-
be obeyed by particles in one-dimensional space hanced electrical conductivity, high magnetic sat-
(to which a term “anyon” is inapplicable, as a uration and resistance against corrosion. ARMCO
rule), and by distinguishable particles (anyons of iron falls into soft magnetic materials, its magnetic
84 AROMATIC COMPOUNDS
AROMATIC COMPOUNDS
A group of organic compounds possessing a
system of conjugated π -bonds.
The term originated from the specific odor
of natural substances containing aromatic com-
pounds. These compounds are exemplified by Activation energies U1 and U2 for a double minimum
the polycyclics – benzene (C6 H6 ), naphthalene potential.
(C10 H8 ), anthracene (C14 H10 ), tetracene
(C18 H12 ), pentacene (C22 H14 ), and others, with
a common formula C4n+2 H2n+4 , where n is the K0 has a temperature dependence weaker than ex-
number of benzene-type rings. The molecules of ponential. The value of K is determined from the
polycyclics are planar and form molecular crys- expression for the reaction rate nA+mB+lC+· · ·
tals in the solid phase. Most of them belong to which is proportional to the product of the appro-
the monoclinic system with two molecules in a priate powers of concentrations (denoted by brack-
unit cell. Electronic properties of aromatic com- ets): v = K[A]n [B]m [C]l · · · . The Arrhenius law
pounds are determined mainly by pz electrons was deduced empirically and can be justified from
of C atoms forming a delocalized π -orbital. The the Boltzmann distribution. The figure schemati-
π -electrons excited by light undergo a π –π ∗ cally presents the dependence of the energy of the
transition and account for the absorption spec- reacting particles on their generalized coordinate.
tral shift from the ultraviolet to the visible part The activation energies U1 of the reaction 1 → 2
of the spectrum in the above-indicated sequence and U2 of reaction 2 → 1 are determined by the
of polyacenes. The mobility of π -electrons de- energies of the initial states 1, 2 and the height
termines important physical and chemical prop- of the potential barrier separating them. The bar-
erties of aromatic compounds: photoconductivity rier energy U varies from several electron-volts
of their crystals, migration of electron excitation to almost zero; it can equal zero, e.g., in ionic-
energy, catalytic activity. Aromatic compounds molecular reactions or in the reactions with mech-
are model systems for organic semiconductors. anisms based on the tunnel effect. Such reactions
The dark electrical conductivity of pure crystals can occur at temperatures close to absolute zero.
is low, σ ∼ 10−2 –10−12 −1 cm−1 . The typi- Reactions in solids with participation of the in-
cal mobility of excess electrons and holes moving herent and impurity atoms, conduction electrons,
in narrow bands takes on values of the order of and various quasi-particles also follow the Arrhe-
1 cm2 ·V−1 ·s−1 . nius law. In them, a part of the activation en-
ergy U D can arise from the diffusion of reagents,
ARRHENIUS LAW (S. Arrhenius, 1889) determined by the temperature dependence of the
Relation stating the temperature dependence of diffusion coef cient D = D0 exp(−U D /RT ). See
the reaction rate constant K
also Interdiffusion, Activation energy, Thermal ac-
U tivation.
K ∼ K0 exp − ,
RT
where U is the activation energy of the reaction, ARTIFICIAL EPITAXY
and R is the universal gas constant. The parameter See Amorphous epitaxy.
ASTERISM 85
ASBESTOS (fr. Gr. ασβεσ τ oς , unquenchable) made of metal alloys the concentration of atoms
Fine-fiber minerals capable of splitting into with the greater atomic radius grows at the exte-
flexible and thin fibers (up to 0.5 μm thick). rior side of the flexure (a partial stress relaxation
The most significant is serpentine asbestos. accompanies this process). By providing for the
There are such varieties as chrysotile-asbestos and segregation of atoms of different types, ascend-
picrolite. Chrysotile-asbestos Mg6 Si4 O10 (OH)8 ing diffusion, evidently, plays an important role in
has a lamellar-tubular structure with a monoclinic the phase separation in alloys, and at the forma-
lattice. Felted and parallel-fiber aggregates are tion of atmospheres of the impurity atoms in the
typical of it. The cleavage is imperfect along (110) strained regions of a crystal, e.g., interstitial atoms
at the angle 130◦ . Density is 2.5 to 2.65 g/cm3 . preferentially concentrate in the stretched region
Mohs hardness is 2.0 to 3.0. Impurities: Fe (up to of the field of stresses at dislocations (see Cottrell
3%), Ca, Mn, Ni, Al, Na. Water content is 13.0 to atmosphere).
14.8%. Refractive indices: ng = 1.522 to 1.555;
np = 1.508 to 1.543. The fibers of asbestos are ASPEROMAGNET
rupture-strong (285 to 365 kg/mm2 ). Asbestos is An amorphous magnetic substance in which
fireproof (T melting is 1500 ◦ C); it has high sorp- the random frozen magnetization mi has a first
tion ability (see Sorption), swelling ability; it con- moment M = mi c and two second moments,
ducts heat poorly; it is stable against alkalis; upon q = (mi )2 c and q⊥ = (m⊥ 2
i ) c .
heating, it loses water and transforms to forsterite. The averaging is performed over configura-
Asbestos is used in textile, cement, paper, radio tions, the indices and ⊥ denoting projections of
industry, road surfacings, etc. mi onto the vector M or onto the plane perpendic-
ular to M, respectively. The asperomagnetic state
ASCENDING DIFFUSION is observed in two types of amorphous magnetic
A process of transport, whereby the flow of materials: in alloys and metallic glasses contain-
particles is directed along the concentration gra- ing rare-earth elements and noble metals, such as
dient. In contrast to conventional diffusion, in this DyCu, TbAg; in systems of the spin glass type.
case the difference in concentration of the diffus- In the first case, the origin is the large value of
ing component does not diminish but rather grows. the magnetic anisotropy constant (typical of rare-
The possibility of ascending diffusion arises from earth elements) and the random orientations of the
the fact that the driving force of the diffusion is anisotropy axis (typical of alloys and glasses). In
the gradient of the chemical potential which, in the second case, the cause is the presence of con-
general, can differ from the concentration gradi- current random ferro- and antiferromagnetic inter-
ent. Different fields which contribute to the chem- actions. Asperomagnets often exhibit nonergodic
ical potential in addition to the entropy of mixing, behavior (see Nonergodicity).
can cause the ascending diffusion; examples are
thermal fields (thermodiffusion), electrical fields, ASTERISM
vacancy fields, phase non-uniformity of the ma- The characteristic oblong shape of X-ray dif-
terial, and the field of elastic stresses. These fields fraction spots from deformed monocrystals ex-
make the potential barrier that the diffusing parti- posed to X-radiation of a continuous type.
cle encounters an unsymmetrical one, and the re- In the case of transmission images (Laue pat-
sult is a preferred drift in the direction of the field. terns), the diffraction spots are elongated toward
In a general phenomenological theory ascending the center of the picture. Asterism arises due to
diffusion can be described in a systematic uniform the reflection of the X-ray beam from the bent part
manner, but the detailed mechanisms operating in of a crystal. The horizontal diameter of the spot
the above mentioned cases are different. in angular units is ϕ1 = 2α and the vertical one is
The Gorsky effect (V.S. Gorsky, 1935), which ϕ2 = −2α sin θ , where θ is the Bragg angle (see
involves ascending diffusion caused by local gra- Bragg law) and α is the flexure angle of the re-
dients of elastic stresses in crystals, is specific for flecting plane. At small θ , ϕ2 ϕ1 which explains
solids. For example, during the bending of a plate elongation of spots at asterism, and proves that the
86 ATHERMIC CREEP
length of asterism bands can serve as a measure of very low temperatures. It emphasizes that the ther-
e xure. In reflection imaging (epigrams), the di- moactivating character of elementary dislocation
rectional stretching of diffraction spots is observed processes in determining the deformation kinet-
according to the flexure axis orientation relative to ics, such as overcoming Peierls barriers by dis-
the initial beam. locations (see Peierls–Nabarro model) or local
barriers determined by the defects of crystalline
ATHERMIC CREEP structure, is of limited applicability for the de-
A type of low-temperature logarithmic creep, scription of low-temperature plasticity. Athermic
independent of temperature. Its rate after the in- plasticity is not fully understood; it has been pro-
stant jump of the initial strain ε0 at the mo- posed as (i) playing a crucial role in quantum
ment of load application quickly decreases in time, fluctuations that arise during the motion of dislo-
ε = α/t . Typical for this creep is the time depen- cations through barriers, and (ii) playing a role in
dence of strain caused by the motion of disloca- the replacement of a thermoactivated barrier pene-
tions, ε = ε0 + α ln(νt + 1), α and ν being con- tration by an activation-free process of the over-
stants. A feature of athermic creep is the small- barrier motion, which occurs owing to the high
ness (0.1 to 1%), in comparison with ε0 , of the level of effective forces acting upon dislocations,
strain under creep. The deceleration of athermic and the low level of dynamic friction restraining
creep is associated either with the increasing den- the dislocations (see Internal friction), typical for
sity of support of intersecting dislocations or with low-temperature deformation.
the “depletion” of mobile dislocations. At temper-
atures close to absolute zero (∼1 K), the barriers ATMOSPHERE
of the support type are overcome with the help See Cottrell atmosphere, Snoek atmosphere,
of a quantum-mechanical tunnel effect (see Low- Suzuki atmosphere.
temperature yield limit anomalies). According to
Mott, in this case α is independent of tempera- ATOM
ture, α ∝ 2πh/(2Mu)1/2 , where h is the Planck An atom consists of a positively charged nu-
constant, M is the effective mass of dislocations, cleus with a size on the order of 10−15 m and a
u is a quantity on the order of the energy of the surrounding electronic cloud with a size of about
step transformation at the dislocation (3·10−19 J). 10−10 m (see Atomic radius).
Within the range of temperatures between 1 K and An atom as a whole is electrically neutral; the
T ∼ 0.1Θ, where Θ is the Debye temperature, the positive charge of the nucleus Ze (Z is the atomic
crossing of the support may be activated by zero- number and e is the elementary electron charge) is
point vibrations of dislocation segments. The fac- compensated by the negative charge of its Z elec-
tor trons. An atom can accept or donate one or more
electrons to become, respectively, a negatively or
2
1 kBΘ T positively charged ion. The complicated structure
α= 1+
2 hν Θ of the electron cloud is determined mainly by
Coulomb interactions, the attraction of electrons to
insignificantly changes with temperature at the nucleus and their mutual repulsion, as well as
T < Θ. The value of ε at the athermic creep is by the Pauli exclusion principle. Relatively small
not large, but it needs to be accounted for in the but fundamentally important contributions to the
operation of precision parts and products. energy arise from the spin–orbit interaction and the
spin–spin interaction.
ATHERMIC PLASTICITY A single-electron approximation is widely used
Lack of sensitivity of the characteristics of a to describe an atom: the wave function of the
plastic deformation process of a solid to changes electrons is a product of one-electron wave func-
in temperature. tions (Hartree method) or a linear combination
Athermic plasticity characterizes the regulari- of such products which satisfies the requirement
ties of plastic deformation of crystals at low and of interchange invariance (Hartree–Fock method).
ATOMICALLY CLEAN SURFACE 87
In this approximation one can speak about one- is substantiated by data of X-ray emission spec-
electron energy levels and corresponding electron troscopy, by the observed differences of charac-
shells of an atom. To describe them approximately, teristic X-ray frequencies for the inner shells of
the model of a hydrogen-like atom is employed, atoms of a given chemical element incorporated
with the shells characterized by the principal quan- into different chemical compounds. It is not fea-
tum number n = 1, 2, 3, . . . , and their substruc- sible to solve exactly the problem of the states
ture by the orbital quantum numbers l = 0, 1, 2, of atoms forming (comprising) a solid. The ap-
. . . , n − 1 and m = 0, ±1, . . . , ±l. The electrons proximation often used in condensed matter the-
of the K-shell with n = 1 are closest to the nu- ory is based on reducing the number of indepen-
cleus with the deepest (most negative) energy dent variables per atom down to 3. Here, the atom
levels. Then follow the L, M, N, . . . shells with is represented as a structureless entity, while inter-
n = 2, 3, 4, . . . . In the ground states of the atoms actions involving atoms are described by various,
H and He only the K-shell is occupied. As Z mostly empirically chosen, potentials with lim-
grows, the orbits with higher quantum numbers ited applicability (see Interatomic interaction po-
become filled. The electron configuration of an tentials). In a more precise approach the valence
atom is specified by symbols that indicate the val- electrons are singled out, and the nucleus with the
ues of quantum numbers n and l of filled shells deep electron shells is treated as a point particle
in ascending order, and the number of electrons with an effective charge.
in them; the values l = 0, 1, 2, 3, 4, . . . being de- The potential of an atom is determined by the
noted with corresponding letters s, p, d, f, g, . . . . distribution of its electron density. An approach-
For example, the configuration 1s 2 2s 2 2p6 of the ing test particle perturbs this distribution in a man-
neon atom has two electrons with n = 1 and l = 0, ner that depends on its nature and its speed rela-
plus two electrons with l = 0, and six electrons tive to the atom. Atom–atom potentials differ for
with l = 1 at n = 2. Each l shell contains 2(2l + 1) different pairs of atoms. In the quantum descrip-
electrons when it is full. tion one cannot speak about each particular elec-
Another approximate method for describing an tron being affiliated with one particular atom of a
atom (Thomas–Fermi method) is based upon a solid. The delocalization of electrons among many
structureless cloud of atomic electrons with their atoms causes the splitting of atomic energy lev-
density distribution determined by the Poisson els and the creation of bands of allowed energies
equation. in crystals. Such broadening is rather small for
The outer valence electrons determine the electrons in deep levels. For those in outer atomic
chemical properties of atoms, their ability to form shells, broad (several electron-volts) allowed en-
this or that compound, the structure, binding en- ergy bands arise, and are particularly prominent in
ergy, and other characteristics of compounds. Al- metals, with their structure determining the elec-
though atoms are the structural elements of solids, tric, optical, and other properties of the crystal.
the distribution of the atomic electrons in a solid
differs from that of a free atom. The configura- ATOM DISPLACEMENTS
tions of valence electrons undergo considerable See Static displacements of atoms, Thermal
changes, depending on the type of bonding force displacements of atoms.
between atoms, the number and positions of near-
est neighbors, etc. In a crystalline solid, the states ATOMICALLY CLEAN SURFACE
of valence electrons depend on whether the atoms A surface free of impurities and having a pris-
occupy sites of the regular lattice, or they are dop- tine crystal structure.
ing atoms or substitutional atoms. The presence of To obtain and preserve an atomically clean sur-
the two latter types introduces additional levels in face, various methods of cleaning in a super-high
the energy spectrum of electrons in crystals and in- vacuum (10−8 Pa) are used. The main methods are
fluences the optical, electrical and other properties high temperature heating, bombardment with ions
of a crystal. The inner atomic shells also undergo of noble gases with annealing and chemical pre-
some change, but to a much lesser extent. This treatment to obtain products that are removable at
88 ATOMIC CHARGE
moderate temperatures. An absolutely clean sur- because of data obtained from scanning tunneling
face is impossible to obtain: the sources of conta- microscopy. According to estimates based on dif-
mination are the impurities in a crystal bulk and ferent data, the reconstruction encompasses sev-
the adsorption of particles from residual gases. eral atomic planes of a crystal. There are only a
Therefore, the surface is considered to be clean, limited number of orientations of planes which co-
if the concentration of impurities is so small that incide with atomic planes having the same crystal-
it fails to affect phenomena under study. Auger lographic indices and characteristic surface struc-
electron spectroscopy is the most suitable method tures. Such surfaces are referred to as atomically
of monitoring surface cleanliness (the sensitivity smooth. In other cases, energy can be gained by
is 10−2 –10−3 monolayers). The investigation of the decay of a flat surface to micro facets, or by
an atomically clean surface structure is performed the formation of a stepped surface including ter-
by low-energy electron diffraction and high-energy races of low-index facets connected by equispaced
electron diffraction methods (see Electron diffrac- atomic steps as high as one or more interplane dis-
tion). From an analysis of the data from struc- tances. Besides the structure, there are also other
tural investigations, the atomically clean surfaces important properties of a surface such as the elec-
of crystals, especially semiconductor ones, can be tronic energy band structure and dynamical char-
reconstructed. Surface structures can form with acteristics (frequencies of atomic vibrations, their
periods other than those in atomic planes in a crys- vibration amplitudes) that are different from the
tal bulk parallel to the surface. A special notation analogous properties in the bulk. Therefore, the
is adopted for these surface structures. They in- near-surface region is considered as a special sur-
clude the chemical symbol of the crystal, the ori- face phase with its own conditions of equilibrium.
entation of a face, and the ratio of the unit cell vec- The atomically clean surfaces of some crystals of
tors of surface structure to the corresponding vec- semiconductors and metals with particular orien-
tors in the bulk, e.g., Si (111) – 7 × 7, Au (100) tations are the place where order–disorder phase
– 5 × 1 and so on. According to modern con- transitions as well as order–order ones occur, while
cepts, the cause of the reconstruction of the up- there are no phase transitions in the bulk. The
per atomic planes of a crystal is the tendency to order–order reconstruction can occur as transitions
reduce the excessive free energy of a surface lat- between surface structures, both with unit cells ly-
tice relative to that in the bulk. The supposition ing in one plane, and with changes of facet struc-
that the reconstruction is determined by the un- ture – morphological transitions.
controllable contamination of a surface failed to
be confirmed experimentally. The difficulties con- ATOMIC CHARGE
nected with clarification of the nature of surface
See Effective atomic charge.
reconstruction are complicated by the fact that
the problem of the determination of the coordi-
ATOMIC FORM FACTOR
nates of the unit cell atoms for the surface struc-
tures with large unit cells is yet to be solved be- A parameter of the elastic scattering of a plane
cause of the complexity of the calculations in the wave (X-ray, electron, neutron) by an isolated
structural analysis. Therefore, various hypotheti- atom, which determines the amplitude of the scat-
cal models are used in discussions of this recon- tered wave as a function of the scattering angle.
struction, their ability to approximate reality be- In the scattering of waves in crystals, the am-
ing tested using data obtained by various meth- plitudes of scattering by individual atoms are am-
ods. All the varieties of models can be divided plified in some directions, while they are de-
into two classes: those based on small static dis- creased in others; the total result being determined
placements of atoms from their normal positions by the values of the atomic form factors of the
in the surface lattice, and those involving the con- atoms forming the crystal. A general expression
cept of orderly-spaced defects at the surface (va- for an atomic form factor is
cancies, adatoms, etc.). There has been consider-
able progress in understanding surface structures f (s) = K ρ(r) exp(is · r) dr ,
ATOMIC MAGNETIC STRUCTURE 89
quences i.e. it may produce a long-period mod- magnetic moment orientations (e.g., in thulium,
ulated magnetic structure in a crystal. Here, the n1 = 4, n2 = 3).
exchange forces order the magnetic moments ac-
cording to either a ferro- or antiferromagnetic pat- ATOMIC ORBITALS
tern, while the relativistic forces lead to the vec-
See Orbitals.
tors M = M(ξ ) or L = L(ξ ) gradually changing
along a certain direction ξ that is called the axis
of the modulated structure. In the simplest mod- ATOMIC ORDER on a crystal surface
ulated structures, M (or L) just rotates around The ordered arrangement of surface atoms
the axis (long period helices); there are also more (ions, molecules) that corresponds to a minimal
complicated modulated magnetic structures. The free energy (see Thermodynamic potentials) of a
formation of a magnetic helix in some crystals is finite crystal. The order is determined by the in-
due to the exchange interaction. The helix period teraction of surface atoms with one another and
is of the same order of magnitude in this case as with bulk atoms. It can differ from the arrange-
the distance between neighboring magnetic atoms. ment in atomic planes inside a crystal, but parallel
A magnetic helix differs fundamentally from a to the surface (see Surface superlattice, Surface
noncollinear antiferromagnetic structure (e.g., of reconstruction). The possible symmetry types of
YMnO3 type) in that the angles between the mag- periodic surface structures are described with five
netic moments may not correspond to 360◦ /n val- parallelogram systems of equivalent points (two-
ues (n = 1, 2, 3, . . .). The helix period is not com- dimensional analog of Bravais lattices, see Two-
mensurate with the interatomic distances ξlm in dimensional crystallography) with primitive unit
this case. cells in the shape of a general parallelogram, rec-
Another common type of magnetic ordering tangle, square, rhombus with angle 120◦ , rhombus
(ferrimagnetic structure) occurs in ferrimagnets, with angle unequal to 120◦ (in this case, one usu-
where the exchange forces position the atomic ally chooses a centered rectangle as a nonprimitive
magnetic moments in a manner resulting in non- cell). Sometimes, the equilibrium state of a crys-
zero values of both M and L. In particular, a tal surface is in correspondence with the ordered
ferrimagnetic structure appears when the mag- arrangement of vacancies, adsorbed atoms, atomic
netic atoms occupy crystallographic positions of steps and terraces on it, e.g., a system of equidis-
two different types, with each type constituting a tant atomic steps of the same height on the facets
magnetic sublattice. Magnetic structures of a spe- with high Miller indices (see Crystallographic in-
cial type, impossible in magnetic crystals, have dices) in silicon and germanium. The results of
been discovered in chemically disordered solid surface structure studies by scanning tunneling mi-
solutions (alloys) (see Amorphous magnetic sub- croscopy, and by low-energy electron diffraction,
stances, Spin glass, Asperomagnet). A particu-
indicate the presence of atomic order on a crystal
lar form of modulated magnetic structure, the spin surface.
density wave whose period may be incommensu-
rable with the crystal lattice period, may appear in
a system of delocalized electrons. ATOMIC RADII
All the above magnetic structures have been The approximate sizes or radii r a of particu-
observed in a number of materials, such as those lar atoms in crystals, assuming them to be spheres
involving 4f -elements and their alloys (see Mag- touching each other. A typical interatomic spac-
netism of rare-earth metals), as well as actinide- ing between two atoms in a crystal is the sum of
containing compounds. There is still another form their atomic radii. In neutral atom crystals such
of magnetic structure called the antiphase domain as solid argon the electron density distribution of
structure which possesses an alternating sequence outer shells of atoms displays a relatively weak de-
in space of the groups of n1 atoms with the mag- pendence on their nearest neighbors. The value of
netic moments oriented along a certain axis, and the outer shell radius averaged over the sphere,
of the groups of n2 = n1 atoms with opposite which characterizes the atomic size, varies with
ATOM–SURFACE INTERACTION 91
temperature, and so on. Reconstruction of the up- type of attenuated total internal reflection to ex-
per atomic planes caused by the tendency to re- cite surface polaritons. To that end, one introduces
duce the free surface energy, and the phenomenon an air (or vacuum) gap between the prism and the
of a surface phase transition in the absence of one material under study [Otto modification (A. Otto,
in the bulk (see Atomically clean surface) are typi- 1968)], or places the material on the base of the
cal for a number of crystal surfaces. The screening prism as a thin film [Kretschmann modification
of the Coulomb interaction of adsorbate atoms by (E. Kretschmann, 1968)].
charge carriers in a nearby-surface space charge
region can appreciably modify the atom–surface ATTO. . .
interaction. A decimal prefix denoting a 10−18 part of the
initial unit of a given physical quantity. For exam-
ATTENUATED TOTAL INTERNAL ple, 1 as (attoseconds) = 10−18 s (seconds).
REFLECTION, ATR method
AUGER EFFECT in semiconductors (P.V. Auger,
Technique for studying optical properties of
1925)
solids (see Optics of solids), based on the phenom-
A semiconductor analogue of the Auger effect
enon of total internal re ection of light, which oc-
in an atom; a double electron transition with one
curs when a light wave penetrates from an opti-
electron crossing to a lower level, and another one
cally denser medium 1 into medium 2 of lower
to a higher level which is often in the continuum,
optical density (n2 < n1 ). Provided that the angle
outside the atom. The presence of energy bands
of incidence ϕ exceeds the critical value ϕcrit =
of two kinds of current carriers, and the variety
arcsin(n2 /n1 ) there forms a nonuniform (decay-
of structural defects in semiconductors, leads to
ing) field E(z) = E0 exp(−z/dpenetr ) (z is normal the wide diversity of the Auger effect. The role
to the interface between the media), with an effec- of this effect in nonequilibrium processes in semi-
tive penetration depth dpenetr that is close to the ra- conductors may be quite different: recombination,
diation wavelength. Now, assuming that medium 2 generation, impact ionization of local centers, par-
is absorbing, the re e ctance is R = 1, which indi- ticipation in scattering processes, relaxation of the
cates the presence of attenuated total internal re- energy and momentum of charge carriers. This ef-
flection. The characteristics of spectra recorded by fect can also be a cause of the strength and of the
this reflection (1 − R) method, and by the standard temperature extinction of luminescence. The set
absorption method, are similar to each other. That of Auger effect variants is shown in Figs. (a–h).
is why attenuated total internal reflection is used to There are processes with participation of a mul-
study strongly absorbing media, and in particular tiply charged local center (a), singly charged local
optical properties of their surfaces (e.g., perturbed center (f, g), two singly charged local centers (b, c,
unstructured layers, adsorption films, transition d, h), and free charge carriers (e). Included among
layers of various origins). To enhance the contrast Auger processes are also all processes that are in-
accumulated during repeated internal reflections, verse to those represented in this figure. A salient
specially designed prisms are used that provide feature of an Auger processes is the pronounced
multiple reflections which effectively extend the dependence of its probability on the distance R be-
optical path of the light wave through medium 2. tween the interacting carriers (∝ R −6 , correspond-
The propagation of light through medium 2 may ing to the energy of one dipole in the field of an-
result in its polarization, so one may speak about other). In the case of an Auger interaction between
the propagation of a volume polariton wave. From free carriers, this means that the rate varies as the
the law of the conservation of the tangential com- square of their concentration. Thus, in process (e),
ponent of the wave vector kx it is possible to the hole lifetime with respect to the Auger process
determine the dispersion of volume polaritons in is proportional to the inverse square of the electron
medium 2: kx = (ω/c)[ε2 (ω)]1/2 where ε2 (ω) is concentration, which means the dependence of the
the relevant dielectric constant. If medium 2 ap- rate on R −6 . Both the theory of the Auger effect
pears surface active in the respective frequency and experimental data demonstrate that the proba-
range, i.e. ε2 (ω) < 0, one may apply a modified bility of this process in semiconductors is weakly
AUGER ELECTRON SPECTROSCOPY 93
dependent on the temperature. The dependence of at level z, and the emission of the Auger elec-
the probability on the energy exchanged between trons. The identification of chemical elements is
the carriers is a power function with an exponent based on comparison of calculated and experi-
between 2 and 5. When the distance between the mentally found values of Auger electron energy
electrons involved in the Auger interaction mech- Exy = Ex − Ey − Ez + U eff . Here, Ex , Ey , and
anism is about 10 nm in order of magnitude (cor- Ez are, respectively, the binding energies of elec-
responding to a free charge carrier concentration trons at x, y, and z levels (generally designated by
about 1018 cm−3 ), the cross-sections of the Auger X-ray terms) of the neutral atom participating in
processes can be quite large (close in a crystal the Auger process; U eff is a factor to allow for the
unit cell size, i.e. ∼10−15 cm2 ). In this case, the presence of two vacancies in atomic shells. Auger
Auger processes can dominate the recombination lines of different elements are well resolved. The
and scattering, thus determining the lifetime of the
change in the electron density at an atom results in
carriers, the probability of radiative and nonradia-
a shift of the Auger line, the magnitude and sign of
tive quantum transitions, the relaxation times of en-
the shift being characteristic of the atom’s chem-
ergy and momentum, and other parameters of the
ical state. Analysis of the width and shape of the
material.
Auger lines, whose formation is participated in by
valence band electrons, yields information on the
AUGER ELECTRON SPECTROSCOPY
band structure of a solid. The total Auger electron
A method of analysis of the elemental and
current is proportional to the number of vacan-
phase composition of surface layers of solids,
based on the eld ionization of atoms excited cies at atomic levels formed per unit time within
by a beam of primary electrons with energies the thickness of the region of the exit from the
1 to 10 keV. The sensitivity of the method is surface. This circumstance underlies the applica-
1010 –1012 atom/cm2 , and the depth of the an- tion of Auger electron spectroscopy to the quanti-
alyzed layer is 0.5 to 2.5 nm. The mechanism tative analysis of solids. In the scanning mode of
for the emergence of Auger electrons is a com- the electron beam deflection, Auger electron spec-
plicated multistage process that includes the ion- troscopy is used for studying the planar distribu-
ization of an inner atomic level x, the filling tion of chemical elements with a spatial resolu-
of the produced vacancy by an electron from tion of a few tens of nanometers. In combination
a higher atomic level y, the nonradiative trans- with ion etching, it is applicable for determining
fer of the released excess energy to an electron the depth profiles of element distributions.
94 AUGER ION SPECTROSCOPY
AUGER ION SPECTROSCOPY means of the Auger process (see Auger effect).
Determination of the elemental and chemi- A condition for Auger neutralization is that the
cal composition, the structure, and the state of a ionization potential of the ion Vi must exceed the
solid and its surface from Auger electron spectra work function of the solid ϕ (Vi > ϕ), and for
(see Auger electron spectroscopy) excited by ion electron potential emission we require Vi > 2ϕ.
bombardment. An important feature which dis- The Auger neutralization rate W depends on the
tinguishes Auger ion spectroscopy from Auger scattering of ions to the surface S in accordance
electron spectroscopy is its selectivity, i.e. Auger with W = A exp(−S/a) where A ∼ 1016 s−1 ,
peaks of a particular kind specific for a given a ≈ 0.2 nm. A consequence of intensive Auger
solid are typically observed in Auger spectra ex- neutralization is a low (≈1%) degree of ionization
cited by ion bombardment. In addition, these peaks of atoms with high ionization potentials, e.g., inert
appear only under bombardment by certain ions. gases reflected from a metal surface.
In contrast to the electron bombardment case, Auger neutralization underlies the ion-neutral-
the character of ion-stimulated spectra is deter- ization spectroscopy of a solid surface that allows
mined by characteristics of the ionization process one to determine the electron density in the con-
in the inner shells, and the emission of Auger elec- duction band, the position of the Fermi level, and
trons at atomic collisions. In particular, the ion other characteristics of the electronic structure of
Auger spectra have a great number of satellites, a solid surface. The Auger neutralization of multi-
and represent quasi-molecular spectra with a typ- ply charged ions proceeds stepwise, as a series of
ical continuous background. Yet many of Auger successive acts of Auger neutralization and Auger
peaks are narrower, since they are emitted by relaxation in the system of levels of ions with de-
free atoms dislodged from the solid. Furthermore, creasing charge, with a mean step of about 25–
the recoil atoms that emerge under ion bombard- 30 eV. As a result of the neutralization, the number
ment and emit the Auger electrons are very fast of emitted Auger electrons is proportional to the
moving, and induce an extra “Doppler” broaden- resultant energy of the ion neutralization, and can
ing (see Doppler effect). Under ion bombardment be as high as a few tens of electron volts. On the
of monocrystals, one can observe both azimuthal surface of nonmetals this effect leads to the accu-
and polar orientation effects in the Auger spectra, mulation of a positive charge (considerably larger
which are related to the crystal structure of the sur- than the initial charge) in a microscopic region,
face. Also classified as Auger ion spectroscopy is and could possibly lead to a Coulomb explosion.
ion-neutralization spectroscopy that deals with the This causes sputtering of the solid, and represents
analysis of the spectrum of secondary electrons a mechanism for metal surface erosion under ac-
emitted during the Auger neutralization of slow tion of slow multiply charged ions. Auger neutral-
positive ions (including multiply charged ones) at ization of multiply charged ions on the metal sur-
the solid surface. Auger ion spectroscopy is a con- face can serve for their identification on the back-
venient method for diagnostics of the solid sur- ground of singly charged ions due to stimulated
face, and is effective when one has to detect atoms electron emission.
of a certain element or a group of elements on
the surface, or during a layer-by-layer analysis of AUGER RELAXATION
a thin layer. This is particularly convenient in de- Nonradiative relaxation of an excited atom near
vices where ion beams are used for implantation, a solid surface, an analog of the Penning effect in
etching, spraying, epitaxy, etc. gases. During Auger relaxation, the excitation en-
ergy is transferred by the Auger process to an elec-
AUGER NEUTRALIZATION tron of the solid, which can be emitted outward.
The process whereby an ion near a solid sur- Along with Auger neutralization, Auger relaxation
face traps an electron (or several electrons in the is a mechanism of potential emission of electrons.
case of multiply charged ions), with subsequent This relaxation is possible for both metastable ex-
transfer of the resulting energy of neutralization cited atoms coming to the surface, and ions that
to another (or other) electron(s) in the solid by trap electrons from the solid at excited levels. The
AUTOIONIZATION 95
Auger relaxation time is 10−14 to 10−15 s, which pearlite class (pearlite, sorbite, troostite) that dif-
is six orders of magnitude less than the time of fer in the degree of dispersion of cementite plates.
radiative relaxation. This circumstance suppresses At temperatures below 450 ◦ C, when the diffusion
strongly the radiation of the excited atoms near rate of iron and alloying elements is insufficient,
the metal surface. On the surface of nonmetals or austenite transforms into bainite; the transforma-
metals covered by adsorbed layers, the rate of the tion needs only the diffusion of carbon. At lower
Auger relaxation can be slightly smaller. temperatures and at a cooling rate greater than
100 deg/s, austenite transforms into martensite in
AUGMENTED PLANE WAVES METHOD, APW a diffusionless manner (see Martensitic transfor-
method (J.C. Slater, 1937) mation). When overheated above the critical point,
One of the main band structure computation perlite transforms into austenite at a varying rate;
methods. It considers crystal potentials in the form however, the newly formed austenite is heteroge-
of a muf n-tin potential (MT potential) and is based neous even within a single grain volume. To ob-
on the variational principle. Its basis functions are tain a homogeneous austenite, it is necessary to
the APWs. Outside the MT-sphere an APW is a operate at a temperature above the completion of
the perlite–austenite transformation, which pro-
plane wave, ei(k+G)r (G is a reciprocal lattice
duces homogenization. The rate of the homoge-
vector). Inside the MT-sphere the APW method
nization is largely determined by the initial struc-
employs a linear combination of solutions of the
ture of the steel. The trends of the processes of
Schrödinger equation in a spherically symmet-
decomposition (see Alloy decomposition) and for-
ric potential. The coefficients for that combina-
mation of austenite underlie the processes taking
tion are selected to satisfy the continuity condi-
place during the heat treatment of steel, with the
tion for APWs at the MT-sphere boundary. The treatment providing the required combination of
amount of computations needed follows from the strength and durability properties of steel.
necessity of solving a nonlinear secular equation
of high order. The order of the equation, i.e. the AUTOEPITAXY, homoepitaxy
number of APWs, is usually quite large, from sev- The oriented and ordered build-up of a material
eral dozen for the simplest lattices to several hun- on a monocrystal surface of the same material.
dred for polyatomic structures. For the linearized Autoepitaxy is observed in the deposition of
APW method, see Linear methods of band struc- material from liquid solutions or melts, by chem-
ture computation. ical reactions at crystallization from the gas (va-
por) phase, and at condensation from vapors in
AUSTENITE (named after W. Roberts of Austen) vacuo. In autoepitaxy the crystal lattice of the
A high-temperature gamma phase that is a building-up phase is oriented relative to the lattice
solid solution of carbon (or nickel, manganese, ni- of the substrate monocrystal. Important influences
trogen) in γ -iron with a face-centered cubic lattice on autoepitaxy are the chemical purity and the dis-
(see Face-centered lattice). tribution of defects on the monocrystal surface,
Alloying with these elements stabilizes γ -iron. the chemical composition of the environment, the
Alloys in which the γ -state is present at room temperature, the goodness of the vacuum cham-
temperature are called austenite alloys. They are ber, and the chemical composition of the residual
easily deformable and possess high corrosion re- atmosphere.
sistance. Carbonaceous austenite is stable under The phenomenon of autoepitaxy is used to pre-
cooling to 725 ◦ C; at lower temperatures, a phase pare thin lms of germanium, silicon, gallium ar-
senide and other semiconductors on monocrystal
transition occurs which can proceed either in a dif-
substrates of the same materials (see also Epitaxy).
fusion or a nondiffusion manner, depending on the
cooling rate or the isothermal exposure. At cool- AUTOIONIC MICROSCOPY
ing rates of a few tens of a deg/s or at isothermal The same as Field ion microscopy.
exposure in the temperature range 450–723 ◦ C,
the austenite decomposes into a ferrite–cementite AUTOIONIZATION
mixture (see Cementite) and forms structures of a The same as Field ionization.
96 AUTOPHOTOELECTRONIC EMISSION
drops, etc.); in phase transitions, e.g., a metal– AVOGADRO NUMBER (A. Avogadro, 1811)
semiconductor transition, in ferroelectric phase The number of molecules (atoms, ions) in one
transitions, in active non-linear optical solid me- mole, identical for any substance in any state of
dia (lasers, real-time holography), etc. The au- matter; NA = 6.02214·1023 mole−1 . One mole
towave processes in solids are characterized by a (gram-molecule) is the mass of a substance in
variety of spatial–temporal scales (diffusion length grams, numerically equal to its atomic or molecu-
105 –10−7 cm, frequency 10−2 –1010 Hz), and lar weight. The unit of molecular weight is 1/12 of
mechanisms of the excitation transfer (fluxes of the mass of the carbon isotope 12 C. In the SI sys-
electrons and holes, excitons, magnons, phonons, tem, the mole is accepted as the unit of the quantity
photons). of a substance.
Applications of autowaves are extensive, in-
cluding, for example, optical, radio-, and UHF am- AXIAL SYMMETRY
plifiers and generators; generation of ultra- and The symmetry of a body, which comes into co-
hypersound; recording and processing of informa- incidence with itself when rotated through an arbi-
tion; image recognition; processes (propagation of trary angle about some axis. Such a rotation sym-
metry axis (∞) pertains to cones and cylinders
nerve impulses and regimes of cardiac muscle op-
with the limiting symmetry groups ∞, ∞2, ∞m,
eration), etc. ∞ , ∞ m corresponding to them.
m m
99
100 BALLISTIC PHONONS
dependence of the magnetization of a band ferro- important part of band structure computation.
magnet on B and T : Pseudopotential methods generally incorporate a
weak pseudopotential; and full-electron methods
m(B, T ) 3 m(B, T ) T2 2χ B
− 1− 2 = 0 , place no restrictions on the potential strength.
m0 m0 TC m0 Early band structure computation methods had
(2)
been developed even before high-speed computers
where m0 ≡ m(0, 0), and χ0 is the longitudi- were available. These were the augmented plane
nal (in B) static magnetic susceptibility at T = 0. waves method (APW), Korringa–Kohn–Rostoker
Eq. (2) is quantitatively valid for weak band fer- method (KKR), orthogonalized plane waves meth-
romagnets, e.g., ZrZn2 , ScIn, Ni3 Ga, and a num- od (pseudopotential method, version I), and tight
ber of Fe–Ni–Mn alloys. The quadratic depen- binding method. The APW and KKR methods are
dence of the spontaneous magnetization on the designed for a potential of a special form. The
temperature, which follows from Eq. (2), viz. further development of band structure computa-
m20 − m2 (T ) ∝ T 2 , was also experimentally ob- tion methods was chiefly determined by the emer-
served in strong band ferromagnets (3d-metals gence of linear methods of band structure compu-
and alloys). tation, e.g., linear combination of atomic orbitals
The theory of magnons in band ferromagnets (LCAO), linear APW, etc., and generalizations of
is based on the work of A.A. Abrikosov and the methods of the APW and KKR families to
I.E. Dzyaloshinskii. The temperature dependence potentials of arbitrary form. Standing apart are
of the magnon spectrum in band ferromagnets was methods related to the tight binding approach, viz.
studied by V.P. Silin and others. the linear combination of atomic orbitals method
If the Stoner criterion (1) is not met then (Gaussian orbitals and Slater orbitals). One can
the metal is paramagnetic. If, however, |κ| 1, notice their connection to the methods of the KKR
then strong band paramagnetism occurs (for band theory. The time required for computing the band
antiferromagnetism see Antiferromagnetic metals, structure is determined, mainly, by the number of
Spin density wave). atoms in a unit cell. In most of the methods the
computational time grows approximately propor-
BAND STRUCTURE tional to the cube of this number.
This is the energy spectrum (plots of aver-
age energy E versus wave number k) of single- BAND STRUCTURE, SURFACE
electron states represented in the form of a disper- See Surface band structure.
sion law of electrons Ekλ , where k is the quasi-
wave vector, λ is a quantum number called the BAND THEORY of solids
band index or band number. Occasionally, elec- A theory describing a crystalline metal or solid
tronic wave functions are also included in this def- as a system of noninteracting electrons, moving
inition. The band structure is a key notion of the in a periodic potential. The potential is assumed
band theory of solids. See also Band structrure to arise from nuclei or positive ions screened by
computation methods. electrons. The effects of mutual interactions are
considered to be included in the effective poten-
BAND STRUCTURE COMPUTATION tial. The Bragg scattering of electrons in the crys-
See Linear methods of band structure compu- tal potential results in a complicated dependence
tation. of electron energies on quasi-momenta (so-called
band structure), with the appearance of band gaps,
BAND STRUCTURE COMPUTATION METHODS i.e. energy gaps where the electrons have no al-
Procedures and algorithms for solving the sin- lowed states.
gle-particle Schrödinger equation in a periodic In the ground state, at zero temperature, all
potential by using modern computers. The con- electron states with energies lower than the chem-
struction of a crystalline potential, self-consistent ical potential (referred to in band theory as the
in the sense of density functional theory, is an Fermi energy E F , or Fermi level) are occupied,
BAND THEORY 103
Energy bands of (a) the insulator crystalline xenon, (b) the semiconductor silicon, (c) the transition metal molybdenum,
and (d) the nearly free electron metal sodium.
and those with higher energies than the chemical Schrödinger equation. However, in the vicinity of
potential are empty. In semiconductors and insula- nuclei the influence of crystal symmetry is not pro-
tors the chemical potential lies within the gap, so nounced, and the potential is close to an atomic
there are no allowed states with the energy E F it- one. Hence, it is possible to speak about electronic
self. There are such states in a metal, and they form states (or bands) of predominantly s-, p- or d-type
a surface in k -space, called the Fermi surface. An character, similar to the classification of states ac-
important parameter of the electronic spectrum is cording to the atomic angular momentum.
the density of electron states, which indicates the Band theory permits the classification of solids
number of allowed states per unit energy at a given according to the qualitative features of their band
energy. The density of states at the Fermi level is structure. In insulators the allowed bands do not
of special importance. overlap, and some are completely full (valence
The main task in band theory is to solve the bands), while the remainder are empty (conduc-
one-electron Schrödinger equation in a crystal po- tion bands). Semiconductors differ from insulators
tential with translational symmetry. The latter im- only by their small (fraction of an electronvolt)
poses certain constraints on a solution ψk with width of the forbidden gap. In metals the Fermi
quasi-momentum k: ψk (r + R) = ψk (r)eikR , level lies within the allowed band, or more com-
where R is a lattice translation period (Bloch monly, the allowed bands overlap one another. In
theorem). Such solutions are called Bloch wave semimetals, the overlap of these bands also oc-
functions. The requirements of crystal symmetry curs, but to a lesser extent. There is also a separate
also impose other constraints on the wave func- class of gapless semiconductors, distinguished by
tions and the electron energy spectrum, and these the width of the forbidden gap becoming zero at
are often used to simplify the solution of the one or several points of the Brillouin zone.
104 BAND THEORY RELATIVISTIC EFFECTS
Quantitative band theory is based on the nu- the interface. In the limit of high concentrations,
merical solution of the Schrödinger equation in the Bardeen barrier height does not depend on the
a crystal. Initially used for this purpose were the contact potential difference due to screening of the
nearly-free electron approximation and the tight Coulomb interaction by the charges in the sur-
binding method, which are valid within the lim- face electron state. In practice, the Fermi level of a
its, respectively, of a weak crystal potential, and Bardeen barrier is fixed at the surface (at the inter-
of a potential consisting of isolated wells sepa- face). In most cases, this term is used with respect
rated by high barriers. Later the pseudopotential to contacts between a semiconductor and a metal
method was developed, which expanded the ap- (called a Schottky barrier), an insulator, or another
plicability of the method of nearly-free electrons. semiconductor.
More widely used are non-pseudopotential meth-
ods, capable of dealing with more generalized BARDEEN–COOPER–SCHRIEFFER THEORY,
types of crystal potentials. With the appearance of BCS theory (J. Bardeen, L. Cooper, J. Schrieffer,
density functional theory, a formal basis emerged 1957)
for the transformation in the effective potential The first and the currently accepted micro-
from a system of strongly interacting electrons scopic theory of superconductivity based on the
to that of noninteracting particles, i.e. to band presence of so-called Cooper pairs in a supercon-
theory proper. Linear methods of band structure ductor at low temperatures T < T c (where T c is
computation have been elaborated, which permit the critical temperature). The Cooper pairs consist
solving the Schrödinger equation in a crystal 2 of electrons with antiparallel momenta and spins
to 3 orders of magnitude faster than by using (see Cooper effect) that form as a result of their
classical methods of computation. Typical exam- attraction via the electron–phonon interaction. The
ples of band structures are shown in Fig. (see starting point of the theory is the model Hamil-
previous page): (a) an insulator (crystalline Xe) tonian
with a narrow valence band, a wide gap, and
+
HBCS = 2εk ckσ ckσ + Vkk bk+ bk , (1)
bands well described by the tight binding method;
k k,k
(b) a semiconductor (Si) with its indirect gap
width small compared to that of the valence band; where εk is the initial energy of the conduction
+
(c) a transition metal (Mo) with valence electrons electrons with momentum k, ckσ and ckσ are
comprising the group of bands formed predom- the creation and annihilation operators of elec-
inantly by s- and d-electrons; where the Fermi trons, respectively, bk+ and bk are the same for the
level lies inside the d-band with a high den- Cooper pairs (see Second quantization), and Vkk
sity of states; and (d) a simple metal (Na) with are the matrix elements of the effective electron–
s-electrons forming an almost parabolic energy electron interaction. The minimization of the en-
band with a low density of states, in accordance ergy of the ground state of the electron subsys-
with the model of nearly free electrons, where X, tem of the superconductor yields the following ex-
Γ , L, N , and P are symmetry points of the Bril- pression for the quasi-particle spectrum through
louin zone. the use of the variational method within the BCS
framework:
BAND THEORY RELATIVISTIC EFFECTS 1/2
See Relativistic effects in band theory of met- Ek = (εk − μ)2 + Δ2k , (2)
als. where μ is the chemical potential (Fermi level) of
the degenerate electrons in the metal, and Δk is the
BAND-TO-BAND TUNNELING
energy gap determined by the following nonlinear
See Interband tunneling.
equation:
BARDEEN BARRIER (J. Bardeen) Δ Δ
A potential barrier which emerges at the bound- Δk = − Vkk k tanh k . (3)
2E k 2k BT
ary between two bodies in contact (e.g., two solids, k
a solid and a liquid or gas) in the presence of a sig- A nontrivial solution of Eq. (3) Δk = 0 exists only
nificant concentration of surface electron states at for the condition Vkk ≡ V < 0, i.e. so in a cer-
BARIUM TITANATE 105
tain range of energy (momentum) of the Fermi sur- ble isotopes, with 138 Ba predominating (71.66%).
face of a metal, the effective attraction due to the The electronic configuration of the outer shell is
electron–photon interaction exceeds the screened 6s 2 . Successive ionization energies are 5.21 and
Coulomb repulsion (see Electric charge screen- 10.00 eV. Atomic radius is 0.221; ionic radius of
ing). In the weak coupling limit Eq. (3) yields the Ba2+ is 0.134 nm. Oxidation state is +2. Elec-
exponential BCS formula for the gap Δ at T = 0: tronegativity is ≈0.94. In a free form barium is
a silvery-white soft metal. It has a body-centered
1
Δ = 2ωD exp − , (4) cubic lattice with parameter a = 0.5019 nm at
N(0)V
room temperature; Space group is I m3̄m (Oh9 ).
(where ωD is the Debye phonon frequency, and Density is 3.76 g/cm3 ; T melting = 998 K; heat
N(0) is the electron density of states on the Fermi of melting is 7.66 kJ/mole; heat of sublima-
surface in a normal metal), and for the critical tem- tion is 205 kJ/mole; T boiling = 1912 K; heat
perature T c of the transition to the superconduct- of evaporation is 175 kJ/mole; specific heat is
ing state we have: 284.7 kJ·kg−1 ·K−1 ; Debye temperature is 110 K;
2eγ 1 coefficient of linear thermal expansion is
Tc = ΘD exp − , (5)
π N(0)V ≈19.5·10−6 K−1 (at 273 to 373 K); adiabatic
bulk modulus is 9.575 GPa; Young’s modulus is
where ΘD = ωD /k B is the Debye tempera-
15.57 GPa; rigidity is 6.33 GPa; Poisson ratio is
ture. Eqs. (4) and (5) give the BCS relation-
0.229 (at 293 K); Brinell hardness 0.042 GPa;
ship k B T c = eγ Δ0 /π = 0.567Δ0 , where k B is
Mohs hardness is 2; ion-plasma frequency is
the Boltzmann constant and γ = 0.577. . . is the
4.52 THz, Sommerfeld linear electronic specific
Euler–Mascheroni constant. The BCS relationship
heat coefficient γ is 2.72 mJ·mole−1 ·K−2 ; elec-
is valid solely for superconductors with weak cou-
tric resistivity is 357 n·m (at 293 K); tem-
pling, whereas for strongly-coupled superconduc-
perature coefficient of the electrical resistance
tors k B T c < γ Δ(0)/π . The latter is due to the ef-
fects of the retardation of the electron–phonon in- is 0.00649 K−1 ; work function of a polycrystal
teraction, which cause the gap Δ to depend on the is 2.49 eV; molar magnetic susceptibility is
energy of the quasi-particles and, hence, the value 20.4·10−6 cgs units; nuclear magnetic moment of
137 Ba is 0.931 nuclear magnetons. Barium is used
of Δ at T = 0 increases compared to the value
πk B T c /eγ . in alloys with lead, aluminum, and magnesium.
The BCS theory has explained the principal Barium and its compounds are added to the ma-
physical properties of superconductors in terms of terials intended for protection against radioactive
the presence of the energy gap in the quasi-particle and X-ray radiation.
spectrum. This includes magnetic vortices, flux
quantization, the exponential dependence of the BARIUM TITANATE, BaTiO3
electron speci c heat on the temperature and its A colorless solid substance; T melting ∼
jump in value at the phase transition point T = T c ; 1618 K; it has the structure of perovskite (tetrag-
the temperature dependence of the thermodynamic onal crystal lattice with lattice constants a =
critical magnetic eld; the features of ultrasound 0.3992 nm, c = 0.4036 nm); density is 6.08 g/cm3 .
attenuation; relaxation of nuclear spins, and the Each titanium is surrounded by a somewhat dis-
absorption of electromagnetic waves in supercon- torted octahedron of oxygens. Chemically it is
ductors; the current–voltage characteristics of the very resistant: it does not change in air or react
superconducting tunnel junctions, and so on. The with water, diluted acids or alkali; it dissolves in
originators of the BCS theory received the Nobel hot strong sulfuric acid. It is an insulator with
prize in 1972. a very high dielectric constant (1000 at 20 ◦ C),
and one of the most important ferroelectrics with
BARIUM, Ba a high piezoelectric modulus (see Piezoelectric-
A chemical element of Group II of the peri- ity) and specific optical properties. It is obtained
odic system with atomic number 56 and atomic by caking and melting together of titanium diox-
mass 137.33. It consists of a mixture of seven sta- ide with barium oxide, hydroxide or carbonate, or
106 BARKHAUSEN EFFECT
by heating up to 800 to 1200 ◦ C of mixture ob- is highly sensitive to various external factors and
tained by common deposition of carbonates or ox- structural changes in the sample. The connection
alates from solutions. It is used in the form of of the Barkhausen effect with the magnetic domain
monocrystals and solid solutions for manufactur- structure, the main elements and the shape of the
ing of capacitors, piezoelectric cells, pyroelectric hysteresis loop, allows it to be used as a method
collectors of radioactive energy, etc. of investigation of the domain structure of ferro-
magnets. By analogy with ferromagnets, the ap-
BARKHAUSEN EFFECT (H.G. Barkhausen, pearance of repolarization jumps in ferroelectric
1919) crystals under variation of an applied electric field
A jump-like change of magnetization M in is also referred to as a Barkhausen effect.
magnetically ordered materials under continuous
variation of external conditions (e.g., applied mag- BARKHAUSEN NOISE
netic field B). Irregular sequences of jumps in the the mag-
Barkhausen discovered that clicks could be netic moment of a ferromagnetic sample placed in
heard in a telephone supplied by an electron am- a periodically varying magnetic field (see Barkhau-
plifier and connected with a coil containing a fer- sen effect). The component is due to fluctuations
romagnetic sample during very slow (quasi-static) of specific parameters of Barkhausen jumps (the
magnetizing of the latter. The main cause of a shape and duration, the critical field, the magnetic
Barkhausen effect is the various inhomogeneities moment varying at the jump, the number of jumps
of the crystalline structure of a ferromagnet (inclu- in a given range). Studies of Barkhausen noise
sions, residual stresses, dislocations, grain bound- yield substantial information on complicated phe-
aries, the heterogeneity of the composition, and nomena involving the rearrangement of the mag-
other defects in the crystals). When moving in netic domain structure in ferromagnets.
the magnetic field, the magnetic domain walls at-
tach to these inhomogeneities until the increas- BARNETT EFFECT (S. Barnett, 1909)
ing field reaches a certain (critical) value. Then Magnetizing of ferromagnets under a change in
the domain wall is released (Barkhausen jump), their mechanical moment.
so that its motion turns out to be jump-like, in Barnett confirmed Ampère’s hypothesis that
contrast to a smooth movement in a perfect crys- magnetism can arise from rotational motion of
tal. These jump-like changes are irreversible. An- charges in a material (molecular current hypoth-
other mechanism of the Barkhausen effect is pos- esis). In a body with no atomic magnetic structure,
sible: the nucleation of new phase centers with the intrinsic magnetic moment is proportional to
their subsequent growth. The Barkhausen effect the angular velocity and oriented along the rota-
is best observable in soft magnetical materials at tion axis. In some cases, the Barnett effect allows
the steepest part of the magnetization curve (of one to determine the gyromagnetic ratio or Landé
stepped shape in that region, see Fig.). This effect g -factor of atoms of some materials. In particu-
lar, this ratio for Fe, Co, Ni and their alloys cor-
responds to the electron magnetic moment and in-
dicates that ferromagnetism is mainly due to the
electron spin. The Barnett effect, along with the
converse Einstein–de Haas effect, are called mag-
netomechanical phenomena. Magnetic resonance
methods are much more precise for determining
g-factors.
BARODIFFUSIONAL EFFECT
An effect of hydrostatic pressure on the diffu-
sion healing of defects, or the retardation of pore
Barkhausen effect. growth (at low pressures), and on the decrease of
BARRIER INJECTION TRANSIT TIME DIODE 107
the diffusion coef cient in crystals (at high pres- Barrett formula transforms into the classical ex-
sures). pression since ε(T ) satisfies the Curie–Weiss law.
At low pressures, barodiffusional effects are The system behavior at T Ω0 /k B depends
caused by the influence of pressure on the equi- on the ratio of T c and Ω0 /(2k B ). If T c >
librium between the pore and the vacancy subsys- Ω0 /(2k B ), the system has a low-temperature
tem (see Vacancy) taking into account oversatu- phase transition at T = T cL < T c . There is no
ration Δ = cv − cv0 in a crystal (where cv and ferroelectric phase transition, if T c < Ω0 /(2k B ),
cv0 are the actual and equilibrium concentrations in this case the crystal is a virtual ferroelectric.
of vacancies, respectively). Under the character- The Barrett formula is used to explain the ob-
istic pressure Pk = 2α/R0k (where α is the sur- served low-temperature values ε(T ) in SrTiO3
face tension, R0k is the critical radius of a viable and KTaO3 , but is only valid when the soft mode
pore embryo in the absence of pressure), the nu- frequency is high relative to acoustical frequen-
cleation and the growth of pores are forbidden. cies. In the general case, taking into account the
At R0k = 10−5 cm, Pk ≈ 102 atm. Such pressure interaction of optical vibrations and acoustic vibra-
has a significant effect on the kinetics of the diffu- tions results in a characteristic modification of the
sional coalescence of the pore ensemble due to the Barrett formula in a low-temperature range.
transformation of some of the growing pores into
dissolving ones. Under high pressures, the baro- BARRIER
diffusional effect is caused by the decrease of cv See Bardeen barrier, Lomer–Cottrell barrier,
and by the increase of the energy outlay for an el- Potential barrier, Schottky barrier, Local barriers,
ementary migration event with growing pressure. Surface barrier, Surface barrier structures.
There is an exponential diffusion coefficient de-
pendence BARRIER CAPACITANCE, junction capacitance
Electric capacitance caused by the existence of
ω P
DP = DP 0 exp − A , a potential barrier in a semiconductor junction, in a
kBT heterojunction, at a metal–semiconductor junction,
where ωA = ωv + ωm is the activation volume, and so on. Such a system possesses a high-ohmic
ωv < ω is the vacancy volume dependent on the layer that is depleted of current carriers and situ-
extent of relaxation of the atoms surrounding the ated between high-conduction layers, i.e. is simi-
vacancy, ω is the atomic volume, ωm is the volume lar to a capacitor. The electric charge in the deple-
of the activated complex at a crossing point be- tion layer is concentrated at local electronic levels.
tween two neighboring positions. It follows from The barrier capacitance value is determined by the
the dependence DP (P ) that at a characteristic thickness of the depletion layer and the dielectric
pressure P ∗ = k B T /ωA the value of DP reduces constant of the crystal. A layer with an increased
e times. For metals, P ∗ ≈ 104 atm. current carrier concentration also has a barrier ca-
pacitance, e.g., an accumulation layer in the metal–
BARRETT FORMULA (J.H. Barrett) insulator–semiconductor structure.
An expression describing the temperature de-
pendence of the dielectric constant of ferroelectrics BARRIER INJECTION TRANSIT TIME DIODE
in the paraelectric phase in a temperature (T ) (BARRITT diode)
range that includes the quantum region: A superhigh-frequency (SHF) semiconductor
diode consisting of a semiconductor plate with
ε(T ) − ε∞ contacts made of semiconductor junctions
−1 (p+ –n–p+ , p+ –n1 –n2 –p+ , where n1 > n2 , and
Ω0 Ω0
=C coth −Tc , other structures), Schottky barriers (m–n–m struc-
2k B 2k B T
tures, m is metal), or their combinations. With the
where Ω0 is the frequency of zero-point vibra- bias voltage applied to the diode, one contact op-
tions, T c is the “high-temperature” value of the erates in the forward direction and the other in the
Curie–Weiss temperature. At T > Ω0 /k B , the reverse direction. A BARRITT diode operates in a
108 BARRITT DIODE
punch-through mode where the depletion regions The almond-shaped basalts, or mandelstones,
of the forward- and reverse-biased contacts over- are porous varieties. The pores (voids) are of
lap, so that a small change of the voltage across a rounded, ellipsoidal, more rarely elongated,
the diode results in the current growing by sev- tubular shape and are filled with minerals, mostly,
eral orders of magnitude. Dynamic negative re- agate, chalcedony, opal, quartz, or sometimes,
sistance in the BARRITT diode arises in a fre- amethyst, zeolites, calcite, and so on. The den-
quency range ω where the phase delay of carrier sity is 2.8–3.2 g/cm3 ; the resistance to compres-
transit through the depletion layer of the reverse- sion in a dry state is 300–500 kg/cm2 (3–5 MPa);
bias junction is θ = ωτ ≈ 3π/2 (τ is carrier transit electric resistivity ρ = 103 –105 ·m; porosity is
time). The barrier drift time through the depletion 7.9%, water absorption is 1.07%, and magnetic
layer of the reverse biased junction depends on the susceptibility χ = (125–15500)·106 CGS units.
width of the depletion layer and the saturation ve- The basalts are used as building, electric insula-
locity. The effect is due to the phase shift between tion, and acid-resistant materials.
the SHF electric field and the conduction current.
The required phase shift is caused: (i) by the phase BAUSCHINGER EFFECT (J. Bauschinger, 1886)
delay of the injection of the minority charge carri- A partial dehardening of single and polycrys-
ers through the forward biased p–n junction, and talline solids observable at the change in the sign
(ii) by the finite time of the current carrier drift of the strain. The effect manifests itself as a reduc-
through the depletion layer. BARRITT diodes are tion of some initial mechanical properties (elas-
used for signal generation and amplification in the tic limit, yield limit, and so on) when, prior to a
frequency range from 3 to 15 GHz. The main ad- test, a sample was loaded in the opposite direction
vantage of these diodes is a low noise level. until an insignificant residual strain appears. The
Bauschinger effect is observable both in static tests
BARRITT DIODE and in the course of fatigue loading (see Fatigue).
The same as Barrier injection transit time diode. At the same time, its value depends essentially on
the kind of material, of the material pretreatment,
BASAL PLANE of crystals and the strain conditions. Numerous attempts have
A plane containing equivalent crystallographic been made to explain the Bauschinger effect on
(coordinate) axes x and y. Ordinarily it is the plane the basis of dislocation theory (see Dislocations);
that forms the bottom of the conventional unit however, a satisfying agreement with the experi-
cell, with the index (001) in tetragonal and (0001) mental data has been achieved only in individual
in trigonal and hexagonal crystals (see Crystallo- cases.
graphic system). The spatial arrangement of the
BEER–LAMBERT LAW (also called
faces and the sides of crystalline polyhedra is de-
Bouguer–Lambert–Beer law, P. Bouguer,
scribed with the help of symbols relative to these
J.H. Lambert, A. Beer)
axes. The faces that intersect only one axis and
A law determining the dependence of the in-
possess numbers (100), (010) and (001) (see Crys-
tensity I of a parallel monochromatic light beam
tallographic indices) are called basal (coordinate)
on the thickness d of the absorbing matter:
faces.
I = I0 e−kd , (1)
BASALT
where I0 is the incident light intensity, k is the ab-
An extrusive and congealed dark gray or black
sorption coef cient (cm−1 or m−1 ). The latter is
volcanic rock consisting mainly of plagioclase,
related to the dimensionless absorption index κ as
pyroxenes and often olivine, typically with an ad-
k = 4πκ/λ, where λ is the wavelength. Eq. (1) is
mixture of magnetite or ilmenite.
used to evaluate the absorptivity of a solid or liq-
Color is dark sulfur-like to black. Basalt forms
uid layer. For a dilute solid or liquid solution of
frequently have a rod-like or a “pencil”-like sep-
concentration n, Eq. (1) takes the form:
aration. Holocrystalline basalt is called dolerite;
basalt of vitreous texture is called hyalobasalt. I = I0 e−χnd , (2)
BERYLLIUM 109
where χ is the molecular absorption index, cm2 . There are 11 known radioactive isotopes with
In the particle beam radiation of solids the term mass numbers 240, 242 to 251 (no stable iso-
absorption cross-section is often employed for χ , topes have been discovered). Electronic config-
and k is referred to as the absorption factor. Some- uration of outer shells is 5f 7 6d 2 7s 2 . Ionization
times the Beer–Lambert law is expressed in terms energy is 6.30 eV; atomic radius is 0.174 nm;
of common logarithms (i.e. a power of 10) as radius of Bk3+ ion is 0.0955 nm, of Bk4+ ion
is 0.0870 nm. Oxidation state is +3 (most sta-
I = I0 · 10−εcd , (3) ble in solution) and +4 (strong oxidizer). Elec-
where c is the molar concentration, and ε is the tronegativity is 1.2. In a free form, berkelium is
decimal molar absorption index, also called the a silvery-white metal. Stable at a temperature be-
absorption or extinction coefficient. low ∼1250 K is α-Bk with double hexagonal
close-packed structure (hexagonal layers alternate
ABAC. . . ; space group P 63 /mmc (D6h 4 ), basic
BEILBY LAYER (G.T. Beilby, 1901)
parameters a = 0.3416 nm, c = 1.1069 nm at
A glass-like surface layer that appears during
room temperature). Stable above ∼1250 K up to
mechanical processing (grinding, polishing, forg-
T melting ≈ 1290 K is β-Bk with hexagonal close-
ing, punching, cutting, and so on) at the sur-
face and over the grain boundaries of crystalline packed structure (space group F m3m, Oh5 ; pa-
solids (metals, quartz, various minerals and chem- rameter a = 0.4997 nm at room temperature af-
ical compounds). A Beilby layer has its normal ter hardening). Density is 14.80 g/cm3 ; T boiling
crystal structure obliterated. Long term grinding is about 2880 K. At low temperatures, α-Bk be-
or polishing is unable to remove the layer, but comes antiferromagnetic with Néel point T N ≈ 22
rather causes its thickness to grow. A Beilby layer to 34 K). In β-Bk, no magnetic ordering has been
is removable by chemical methods. Its thickness detected. Berkelium isotopes are used for research
ranges from a several to 100 or more microns. in nuclear physics and radiochemistry. Berkelium
A Beilby layer along grain boundaries imparts is highly toxic.
great strength to a material, it is characterized by
pronounced hardness, and specific optical, elec- BERNAL’S FLUID MODEL
tric, and magnetic properties. A Beilby layer crys- See Random close-packing model.
tallizes under heating at a temperature between 0.4
and 0.6 that of the melting temperature. BERYLLIUM, Be
A chemical element of Group II of the periodic
BENDING VIBRATIONS of a crystal lattice system with atomic number 4 and atomic mass
Vibrations of chains (layers) of crystal atoms 9.01218.
in a direction normal to the chains (layers). Bend- Natural beryllium consists of the stable iso-
ing vibrations exist in those crystals where the tope 9 Be with electronic configuration 1s 2 2s 2 .
bonding between the chains (layers) is weak com- The ionization energies are 9.323 or 18.21 eV.
pared to that of the atoms within the chain (layer). Atomic radius is 0.113 nm, Be2+ ion radius is
The linear dispersion law for bending vibrations is 0.035 nm. It is chemically active; oxidation state
valid only for the lowest values of the wave vec- is +2. Electronegativity is 1.47. In a free form,
tor along the chain (layer). At large values of the beryllium is a light bright gray metal, which is
wave vector, a quadratic dispersion law similar to brittle under normal conditions. The properties of
that for the vibrations of strings and membranes beryllium depend on its purity, grain size, tex-
applies. ture and degree of anisotropy that depend on
its pretreatment. Some properties change with an
BERKELIUM, Bk increase of pressure and temperature, and upon
An actinide element of Group III of the peri- strong neutron irradiation (>1023 neutrons/m2 ).
odic system with atomic number 97 and atomic Beryllium crystallizes in two allotropic modifi-
mass 247.07. cations. Below 1527 K, α-Be is stable (hexa-
110 BERYLLIUM ALLOYS
gonal close-packed lattice with parameters a = α-Be is 35.4 n·m along the principal 6 axis
0.2283 nm, c = 0.3584 nm at 300 K and impurity and 31.3 n·m perpendicular to it at 298 K;
content no more than 1%; space group P 63 /mmc, temperature coefficient of electrical resistance
D6h4 ). Above 1527 K up to T
melting = 1560 K, is 0.00628 K−1 (at 293 K). The Hall constant
β-Be is stable (body-centered cubic with para- is +24.3·10−11 m3 /C at room temperature
meter a = 0.2546 nm at 1527 K; space group (+77·10−11 m3 /C for the magnetic field applied
I m3m, Oh9 ). Bond energy of pure beryllium is along the 6 principal axis in α-Be at room temper-
3.33 eV/atom at 0 K; density is 1.848 g/cm3 ature). Work function from polycrystal is 3.92 eV;
(at 293 K); melting heat is 11.7 kJ/mole; subli- coefficient of secondary electron emission is 0.4
mation heat is 314 kJ/mole; T boiling = 2743 K; to 0.5 for the primary electron energy 0.07 to
evaporation heat is 309 kJ/mole; specific heat is 0.2 keV; superconducting transition temperature is
1.80 kJ·kg−1 ·K−1 ; Debye temperature is 1160 K. 0.064 K. Beryllium is diamagnetic with the mo-
Coefficient of linear thermal expansion is lar magnetic susceptibility 9.02·10−6 cgs units at
9.2·10−6 K−1 along the principal axis 6 and 300 K (the value along the 6 axis is higher than
12.4·10−6 K−1 perpendicular to it (at 300 K). that in the perpendicular direction); nuclear mag-
Heat conductivity reduces from 230 W·m−1 ·K−1 netic moment of the 9 Be isotope is 1.18 nuclear
to 178 W·m−1 ·K−1 in the temperature range 273 magnetons. Thus, Be advantageously combines
small atomic mass and low density, high specific
to 323 K; adiabatic coefficients of elastic rigid-
heat, vaporizing heat, thermal conductivity, heat
ity of α-Be monocrystal: c11 = 288.8, c12 = 20.1,
resistance, elastic modulus, hardness, strength and
c13 = 4.7, c33 = 354.2, c44 = 154.9 GPa at 298 K
capability of retaining the stability of dimensions,
(within the range 423 to 1530 K c13 0 GPa).
high electrical conductivity, corrosion resistance,
Bulk modulus is ≈109 GPa, Young’s modulus is
small thermal neutron trapping cross-section and
≈304 GPa, rigidity is ≈143 GPa; Poisson ratio
satisfactory stability under irradiation, high per-
is ≈0.056 (at 298 K); ultimate tensile strength
meability for X-rays and intense emission of neu-
is 0.196 to 0.539 GPa; relative elongation is
trons under the bombardment by α-particles. Due
0.2 to 2%. Pressure treatment considerably im-
to all this, beryllium is used in aviation, rocket and
proves mechanical properties; ultimate strength in
space engineering, in underwater instrument mak-
stretching direction reaches 0.392 to 0.784 GPa;
ing, in reactor material science, radioelectronics,
yield limit is 0.245 to 0.588 GPa; and relative X-ray engineering, metallurgy. Its low plasticity
elongation up to 4 to 12%; mechanical proper- (due in part to impurities) is the main obstacle to
ties perpendicular to stretching undergo minimal its use as a structural material. The processing of
change. Brinell microhardness is 1.18 GPa; impact beryllium is complicated because of the high toxi-
strength is 0.001 to 0.005 GPa; transition temper- city of its volatile compounds in the form of pow-
ature of commercially pure beryllium from brit- der and dust.
tle to plastic state is 473 to 673 K. Highly pure
beryllium (above 99.9%) with grain size <10 μm BERYLLIUM ALLOYS
looses its red-shortness, becomes easily forgeable Alloys based on beryllium. The main advan-
and rollable and, at temperatures within the range tages of beryllium alloys are high specific strength
873–973 K, becomes superplastic. Coefficient of and specific rigidity up to temperatures of 600
self-diffusion of α-Be is 32.9·10−12 m2 /s along to 800 ◦ C, high speci c heat and low neutron
the 6 principal axis and 53.4·10−12 m2 /s perpen- trapping cross-section. The main disadvantages
dicular to it at 1370 K. Effective cross-section of are low plasticity at room and cryogenic temper-
the thermal neutron trapping is 0.009 barn/atom, atures, and toxicity. Hardness varies with low-
their scattering cross-section is 6.1 barn/atom. alloyed (see Alloying), and disperse-hardened al-
The properties of irradiated beryllium are re- loys (including those with intermetallide harden-
stored upon annealing. Ion-plasma frequency is ing). There are structural low-alloyed ones with
49.22 THz, Sommerfeld coefficient γ of the lin- the solid solution structure, two-phase compos-
ear low temperature electronic specific heat is ite ones with pliable base, and those with a dis-
≈0.19 mJ·mole−1 ·K−2 . Electric resistivity of perse structure. The most common are beryllium
BETHE–PEIERLS APPROXIMATION 111
for the uctuations in a spin system, and can de- Here F and F are surface force densities, u and
scribe a number of physical effects not accounted u are the corresponding body displacement vec-
for by the molecular eld approximation. These tors, and S is the elastic body surface area. In other
are the absence of the long-range order (see Long- words, the work that would be done by F acting
range and short-range order) at T = 0 in one- through u equals the work that would be done
dimensional systems, the existence of short-range by F acting through u. The theorem is a gener-
order above the transition temperature to an or- alization of the Maxwell reciprocity theorem for a
dered phase, and the appearance of the percolation linear elastic response.
concentration threshold (see Percolation theory)
in disordered dilute magnetic materials at T = 0. BIAXIAL CRYSTALS
However, like the molecular field approximation, Crystals which have two optical axes with the
the Bethe–Peierls approximation fails to describe property that light travelling along either of these
collective excitations (spin waves, magnons etc.) directions is not subjected to birefringence or po-
in magnetic materials with vector spins. larization. The angle between the two optical axes
(2θ) is called the optical axis angle. Biaxial crys-
BETHE SPLITTING tals are optically positive if the bisector angle
The splitting of degenerate energy levels of 2θ < 90◦ and negative if 2θ > 90◦ (see also Uni-
atoms, ions, or molecules under the action of crys- axial crystals).
talline electric fields of various symmetries (see
Crystal e ld). BICRITICAL POINT
The latter is lower than the spherical symme- A point on the phase diagram of a crystal
try typical of a free atom so that there is a total where two curves of second-order phase transi-
or partial removal of the degeneracy in a crys- tions and one curve of a rst-order phase transi-
tal, and several closely spaced bands (lines) ap- tion meet. The latter curve usually separates the re-
pear in the absorption or emission spectra instead gions of existence of two different ordered phases
of a single line. Their maximum number corre- (see Multicritical point).
sponds to the degree of degeneracy. In a weak
crystalline field, the spin–orbit interaction is re- BIEXCITON, exciton molecule
tained and the Bethe splitting is less than the mul- A bound state of two excitons in semiconduc-
tiplet one. In an intermediate field, the Bethe split- tors and insulators.
ting exceeds the multiplet one and remains less A biexciton in a semiconductor is a stable four-
than the distance between neighboring multiplets. particle complex containing two electrons and two
In this case, the spin–orbit coupling is broken. holes. Since the effective masses of an electron
The Bethe splitting in a strong field exceeds the me and a hole mh in semiconductors are close
distance between neighboring multiplets. This un- to each other, the biexciton stability problem has
couples (quenches) the total orbital angular mo- no small parameter for carrying out a power se-
mentum from the spin angular momentum, and it ries expansion. Therefore, the adiabatic approx-
can also uncouple the individual spin orbital an- imation widely used in molecular spectroscopy
gular momenta from each other. Bethe splitting is is inapplicable in this case; and the calculations
observed in phonon optical spectra as well. are usually carried out with the help of the vari-
ational method. In the limit me /mh = 1, the
BETTI RECIPROCITY THEOREM (E. Betti) biexciton stability problem is analogous to the
A statement of elasticity theory according to positronium molecule problem. The binding en-
which the following equality holds for an elastic ergy of the positronium molecule is extremely
body in the equilibrium state when no bulk forces small:
M ≈ 0.02Ex where Ex is the binding en-
are present: ergy of the positronium atom (electron–positron
bound state analogous to hydrogen atom). For ex-
(F u ) dS = (F u) dS. ample, the biexciton binding energy in Ge and Si is
S S
ΔM 1 meV. Variational calculations of this
BIOLOGICAL CRYSTALS 113
inside the polarization potential well. The possi- concentration and m∗B is the effective mass of the
bility of forming bipolarons was first established bipolarons) of an almost ideal charged Bose gas
analytically in ionic crystals and then extended to 1/3 ∗
of high density under the condition n0 aB >1
amorphous insulators, metals and other materials. (where aB ∗ = ε 2 /(4e2 m∗ ) is the effective Bohr
0 B
In bipolarons the electrons with opposite spins are radius of a bipolaron with charge of 2e, and ε0
bound, which leads to the absence of paramag- is the static dielectric constant of the crystal). An
netism of the free charge carriers. The experimen- estimate T0 2–50 K was obtained for bipo-
tal evidence of the existence of bipolarons was ob- larons of intermediate radius in ionic crystals with
tained in a series of oxide crystals with variable ε0
1 of SrTiO3 type at m∗B 10m0 (m0 is
valence (e.g., Ti4 O7 ), and in some compounds of
free electron mass) and n0 ≈ 1017 –1020 cm−3 ,
linear organic molecules. The space–time proper-
while T0 8–80 K for bipolarons of small radius
ties and energy scale of a bipolaron are essentially
with m∗B ≈ (100–1000)m0 at n0 1022 cm−3 .
different from those in a Cooper pair. See also
The bipolaron mechanism of superconductivity
Bipolaron mechanism of superconductivity.
has been proposed theoretically for heavy fermi-
ons (CeCu2 Si2 , UBe13 , UPt3 ), semiconducting
BIPOLARON MECHANISM OF
perovskites BaPb1−x Bix O3 and possibly the new
SUPERCONDUCTIVITY high-temperature superconductor cuprates.
A possible mechanism to explain supercurrent
flow through a crystal without resistance at low BIPOLAR TRANSISTOR
temperatures due to the super uidity of a charged A semiconductor device with its active zone
Bose gas or a Bose u id of bipolarons. containing two semiconductor junctions separated
The elementary excitation spectrum in such a by a thin region called a base. The outer zones hav-
system of charged bosons possesses, due to the ing the same conduction type are called the emitter
long-range Coulomb repulsion, a finite energy gap and collector (see Fig.). In an n–p–n type transis-
E0 = ωp at zero momentum (where ωp is the fre- tor the minority current carriers of the base (elec-
quency of the collective plasma oscillations in the trons) diffuse through the p-type base region and
bipolaron gas). This also satisfies the Landau su- penetrate the n-type collector region; in a p–n–p
per uidity criterion, which in this case corresponds type bipolar transistor the minority current carri-
to superconductivity. The bipolaron mechanism of ers (holes) diffuse through the n-type base region.
superconductivity differs from the Cooper pairing A bipolar transistor is designed in such a way that
of the Bardeen–Cooper–Schrieffer theory (BCS) small variations of the electric current or voltage
in that bipolarons of small and intermediate radii could cause significant current variations between
are separated in space, and are not expected to de- the emitter and the collector. As to their design,
cay above the transition from the superconducting bipolar transistors can be either homostructural,
to the normal state. To estimate the critical tem- heterostructural, or have graded bands. Bipolar
perature of such a transition, one can use the ex- transistors serve for the generation and amplifi-
pression for the Bose–Einstein condensation tem- cation of electromagnetic oscillations, and for the
perature T0 = 3.312 n0 /m∗B (where n0 is the
2/3
control of voltage and current.
Structure (left) and energy diagram (right) of an n–p–n bipolar transistor showing the emitter (e), base (b) and collec-
tor (c) regions. The donor and acceptor energy levels are indicated by dashed horizontal lines.
116 BIRADICAL
decreases 12.5 times). Cross-section for thermal The effect is observed in technological proces-
neutron capture is not high in Bi (0.034 barn). Bis- ses involving the shape-changing of metals (such
muth is used for the production of low-melting al- as drawing, punching, rolling, and so on). The
loys (e.g., Wood’s metal alloy with melting point effect is especially noticeable when using ultra-
of 343 K). Bismuth alloys are used in the helices sound (see Ultrasonic treatment); in this case, the
of devices for measuring strong magnetic fields, reduction of static stress depends on the ultra-
and it also serves as a heat-transfer agent in nu- sound intensity, the nature and the condition of the
clear reactors. material, the temperature, and other conditions.
The effect of the ultrasound is explained by a local
BISTABILITY absorption of the energy of the high-frequency os-
A property whereby a system has two stable cillation by various defects of the crystalline struc-
states in a certain region of variation of its parame- ture (see Defects in crystals), which facilitates the
ters. The existence of bistability is connected with multiplication of dislocations and vacancies, as
nonlinear properties of the system, and produces well as their motion in the crystal.
hysteresis during a fast enough variation of the
parameters. It is exemplified by optical bistability, BLISTERING
polarization optical bistability, an N- or S-shaped An appearance of blisters on a crystal surface
current–voltage characteristic, the magnetic polar- under bombardment with atoms (ions) of inert
ization of a ferromagnet, in a magnetic field, etc. gases, protons, and other particles.
Nonlinear systems may exhibit more than two sta- The blisters “ripen” with an increase in dosage
ble states (multistability). (see Dosimetry) up to a stage when the blister
dome breaks due to the increased pressure inside
it. A new generation of blisters can develop on the
BITTER PATTERN
uncovered area. The subsequent stages of the blis-
See Magnetic powder pattern.
tering are shown in the figure. The phenomenon
of blistering is caused by migration of irradiation-
BLACK-AND-WHITE GROUPS induced vacancies and vacancy pores towards the
The same as Shubnikov groups. surface. The mechanical stresses arising at embed-
ding of the bombarding ions into the subsurface
BLAHA–LANGENECKER EFFECT (F. Blaha, layer can also induce blistering. If the defect den-
B.-Z. Langenecker, 1935) sity is sufficiently uniform along the surface, the
Reduction of the force needed for plastic defor- scabbing of thin plates off the surface of the ir-
mation of a specimen under simultaneous action of radiated material takes place instead of blister for-
static and alternating loads. mation. The blistering and flaking phenomena lead
Blistering process.
118 BLOCH EQUATIONS
to strong surface erosion of irradiated material at (iii) in an external magnetic field (comprised of a
high irradiation doses. strong constant and weak alternating components)
the relaxation motion of the magnetization vector
BLOCH EQUATIONS (F. Bloch, 1946) is superimposed on the motion of free spins.
Phenomenological equations describing the Under the conditions of resonance, the sta-
magnetic properties of ensembles of electrons tionary solutions of Eqs. (1) lead to the following
(γ < 0) and nonzero-spin nuclei (γ > 0) in exter- expressions for the real (χ ) and imaginary (χ )
nal magnetic fields: parts of the magnetic susceptibility:
dM Mx i + My j χ = (ω0 − ω)T2 χ ,
= γ (M × B) −
dt T2 1 −1
χ = γ M0 T2 1 + (ω − ω0 )2 T22 + S ,
Mz − M0 2
− k, (1)
T1
where ω0 = γ B0 , χ determines the Lorentzian
where Mx , My , Mz are the components of magne- shape of the line with a half-width at a half-height
tization vector M; M0 is the equilibrium value of of Δ = (1 + S)1/2 /T2 ; and the saturation factor
Mz ; i, j and k are the unit vectors of the labora- (see Saturation effects) S = γ 2 B12 T1 T2 indicates
tory coordinate system; B = kB0 + i2B1 cos ωt , the dependence of the line shape on the alternat-
B0 is the constant magnetic field component di- ing field intensity. In the absence of saturation
rected along z axis; ω is the frequency; and (S 1), the half-width is determined by the spin–
B1 is the amplitude of alternating magnetic field spin relaxation time T2 . The Bloch equations are
(B0
B1 ); γ is the gyromagnetic ratio. The con- also used for the analysis of pulsed experiments.
stant T1 is called the longitudinal or spin–lattice One can obtain Eqs. (1) from the general theory
relaxation time and T2 is the dephasing time, also of relaxation (see Bloch–Red eld theory). For two-
referred to as the transverse or spin–spin relax- level systems they are valid if alternating fields are
ation time. In the general case, T1 is unequal to small and there are no spin–spin interactions. In
T2 (T1 T2 ) since T1 characterizes the rate of es- solids the value of T2 is often estimated from the
tablishing thermal equilibrium between the spins width. In this case Eqs. (1) lead to a discrepancy
and the lattice, while T2 is related to rate of es- with the experimental data in the limit of strong
tablishing thermal equilibrium within the spin sys- saturation. Despite the limits to their applicabil-
tem, and this latter is equivalent to the decay rate ity, the Bloch equations play an important role in
of the phase coherences between precessing spins elucidating many aspects of nuclear magnetic res-
(as a result of the interaction between each other onance and electron paramagnetic resonance in
and with their environment, the spins precess in solids.
different local fields).
The Bloch equations are based on the follow- BLOCH FUNCTION
ing reasoning. In an arbitrary uniform magnetic See Bloch theorem.
field B the equation of motion of the magneti-
zation vector M of the ensemble of free spins BLOCH–GRÜNEISEN FORMULA (F. Bloch,
has the form dM/dt = γ (M × B). It is postu- E. Grüneisen, 1930)
lated that: (i) in a constant magnetic field Bz , the
An interpolating expression describing the tem-
longitudinal magnetization Mz relaxes to M0 in
perature dependence of that part of electric resis-
an exponential manner, i.e. satisfies the equation
tance of metals ρT that is caused by electron scat-
dMz /dt = −(Mz − M0 )/T1 ; (ii) transverse com-
tering by lattice vibrations (phonons):
ponents Mx and My decay exponentially to zero
as 5 Θ/T
T z5 dz
dMx Mx dMy My ρT = 4 ρ0 , (1)
=− , =− ; (2) Θ (e − 1)(1 − e−z )
z
dt T2 dt T2 0
BLOCH LINE 119
play an important role in the formation of the phe- from Eq. (1) by a uniform rotation. (The Bloch
nomenon of ballistic aftereffect in the motion of point in Fig. takes the form m = r/r after the ro-
magnetic bubble domains. tation of m by π/2 about the y axis.) In materi-
als with strong uniaxial magnetic anisotropy, the
BLOCH POINT loss of domain wall energy due to the presence
A point inhomogeneity, a singular point of the of a Bloch point is not significant. Bloch points
magnetization M(r) at the merging of two Bloch must always arise under certain distributions of
lines of different signs in a magnetic domain wall. the magnetization. Bloch points and their displace-
The magnetization on a closed surface surround- ments under the action of an appropriately directed
ing a Bloch point takes on all possible values. magnetic field (along x axis) are detectable in do-
Thus, m = M/M s is equal to +ez and −ez far main walls of materials with magnetic bubble do-
from the domain wall, to +ey and −ey in the mid- mains by the deviation of the direction of domain
dle of the domain wall far from the Bloch line, and motion from the direction of the magnetic field
to +ex and −ex in the middle of Bloch line seg- gradient (see Rigid magnetic bubble domain).
ments of opposite sign (see Fig.). Hence, when ap-
proaching a Bloch point, the direction of magne- BLOCH–REDFIELD THEORY
tization m depends on the direction of approach, A magnetic resonance theory describing the
thus producing a “hedgehog”-type singularity of time evolution of a spin system, located in an
the magnetization field (see Topological inhomo- external magnetic field, and interacting with its
geneity). The following distribution corresponds molecular environment (with the lattice).
to this singularity at a < r < Δ (a is the inter- The main result of the theory is the Markov-
atomic distance, Δ is the domain wall thickness): ian kinetic equation for the spin density matrix.
r A consistent quantum-statistical derivation of the
m=± , (1)
r kinetic equation was first presented by F. Bloch
where the different signs determine the sign (i.e. and R.K. Wangness, then Bloch generalized the
topological charge) of a Bloch point. There are theory for alternating magnetic fields of arbitrary
other possible distributions that can be obtained magnitude. Later A.G. Redfield put forth a semi-
classical theory of relaxation where an approach to
the calculation of the kinetic equation coefficients
was treated for the first time. The basic physical
concepts of the Bloch and Redfield theories are the
same. Later, both theories were described within
the framework of a general formalism by other au-
thors.
Let a closed region comprise a spin system and
a lattice with the Hamiltonian H = Hs + Hs (t) +
Hsf +F where Hs is the time-independent Hamil-
tonian of the spin system, Hs (t) is the interaction
operator of the spin system
with alternating mag-
netic fields, Hsf = α uα vα is the spin–lattice
interaction operator, uα are the lattice operators,
vα are the spin operators, F is the Hamiltonian of
the lattice. The exact equation of motion for the
density matrix ρ of the closed system, idρ/dt =
[H, ρ], is iterated in the interaction representation,
and the infinite series obtained is truncated at the
Distribution of magnetization indicated by arrows in a second term (spin–lattice interaction considered as
domain wall with a Bloch point BP crossed by Bloch small perturbation). Then the spin density matrix
lines BL-1 and BL-2. evolves with the expression: σ = Trq ρ where the
BLOCH WALL 121
trace (Tr) of the matrix is taken over the lattice the density matrix σ changes insignificantly
states. The kinetic equation for elements σmn has within the time
tl , thus allowing one to ap-
the following form: ply perturbation theory);
dσmn (iv) alternating magnetic fields are small; γ B1
τc−1 , γ B0 , where B1 is the amplitude of the
dt
external alternating magnetic field, B0 is the
= i[σ, Hs + Hs (t)]mn amplitude of the externally applied magnetic
+ (2Γmkln σkl − Γklmk σln − Γlnkl σmk ). field, and γ is the gyromagnetic ratio;
kl
(v) the interaction between the lattice and the
(1)
spin system results in a small shift of the lev-
Here the indices m, n, k, l enumerate the energy els of the latter that is ignored in Eq. (1); and
levels of Hamiltonian Hs ; (vi) the relaxation coefficients Γ are time-inde-
pendent and comparable to the reciprocals of
Γmkln = π jαβ (ωmk )vβmk vαln Δ(ωmk +ωln ), the relaxation times.
αβ
It follows from the relation
where jαβ (ωmk ) is the Fourier transform of the
lattice correlator: Γmkln = Γlnmk exp(βωln ) ≈ Γlnmk exp(βωkm )
jαβ (ωmk ) that in the absence of alternating magnetic fields,
−βFr δ(ω
the solutions of Eq. (1) are elements of the equi-
r,r uαr r uβrr e mk + ωrr ) librium spin density matrix, i.e. due to the relax-
= −βF
,
2πTrq e ation the spin system equilibrates with the lattice.
Eq. (1) describes the magnetic resonance and the
where ωr r are the lattice eigenfrequencies;
relaxation of magnetically dilute spin systems.
ωmk are the eigenfrequencies of the spin sys-
tem; vβmk and vαln are the matrix elements
BLOCH THEOREM
with respect to eigenfunctions of Hamiltonian Hs ;
uαr r and uβr r are the matrix elements of eigen- One of the most important theorems in solid
functions of operator F ; β = /(k B T ), where state physics, which states that the intrinsic states
T is the lattice temperature, Fr is the energy of of a particle in a periodic potential V (r) sat-
rth lattice level. One must retain on the right- isfying condition V (r + t) = V (r) (where t is
hand side of Eq. (1) those Γmkln for which the constant distance vector called the direct-
lattice vector) can be characterized by two quan-
|ωmk + ωln | τc−1 . This is formally achieved by
tum numbers, viz., the band number λ and a real-
the introduction of the Heaviside step function
valued vector k called the quasi-wave vector (lat-
Δ(x) = 1 at |x| < τc−1 and Δ(x) = 0 otherwise,
tice wave vector or crystal momentum). The wave
where τ c is the lattice correlation time (perturba-
function Ψkλ (r) called the Bloch function satis-
tions in the lattice disappear within time τ c ).
fies the condition Ψkλ (r + t) = Ψkλ (r) exp(ikt);
The derivation of the equation requires a num-
and can be written as Ψkλ (r) = ukλ (r) exp(ikt)
ber of assumptions:
where ukλ has the symmetry of the potential, so
(i) the lattice stays close to thermal equilibrium; that ukλ (r + t) = ukλ (r). The function Ψkλ (r)
(ii) the influence of the spin system on the lat- can also be written as a linear combination of
tice is negligible, since the lattice has many Wannier functions Φλ (r), i.e. Ψkλ (r) =
ikr
more degrees of freedom than the spin sys- R e Φλ (r − R), where the set of points R
tem (ρ = σρf , where ρf is the equilibrium transforms to itself upon displacement by the vec-
density matrix of the lattice); tor t , i.e. it possesses the symmetry of the poten-
(iii) there exists a time interval
t such that tial.
t
τ c and simultaneously
t τ r , where
τ r is the shortest relaxation time of the spin BLOCH WALL
system (if the latter condition is met then See Magnetic domain wall.
122 BLOCH WAVE FUNCTION
weak coupling limit, for the first time, an equa- More generally, the distance between the cen-
tion for the energy gap parameter Δ in the spec- ter of the hydrogen atom and the maximum of the
trum of a superconductor taking into account ef- quantum-mechanical probability for detecting an
fects of retardation. In addition, with the help electron; or the distance through which the elec-
of the Bogolyubov method, a further weakening tron wave function reduces by the numerical fac-
of the screened electron–electron repulsion was tor of e = 2.71828. The Bohr radius has the value
shown to exist (see Bogolyubov–Tolmachyov log- a B = 4πε0 2 /(me2 ) = 5.2918·10−11 m, where
arithm), which facilitates Cooper pairing due to m and e are the electron mass and charge, re-
the electron–phonon interaction. When the Bo- spectively, = h/(2π) is the reduced Planck con-
golyubov method was applied to collective exci- stant. In the physics of insulators and semiconduc-
tors the effective Bohr radius aB ∗ is used which
tations involving undamped branches in the en-
ergy range √ ω < 2Δ with a linear dispersion law differs from the Bohr radius a B with m replaced
ωq = qv F / 3 (where v F is the Fermi velocity), with the effective mass of an electron in a crystal
√
they were found to exist in neutral superfluid and e replaced by the effective charge e∗ = e/ ε,
Fermi systems. That branch represents a Gold- where ε is the relative dielectric constant. The ef-
stone mode. In charged Fermi systems with long- fective Bohr radius determines the approximate
range Coulomb repulsion (electrons in a metal), dimensions and energy, En = −e∗2 /(8πε0 aB ∗ n2 )
the branch transforms into plasma oscillations with (where n is the principal quantum number), of
the frequency ωq = 0 at q → 0. shallow hydrogen-like centers and Wannier–Mott
excitons in semiconductors. It is a convenient scale
BOGOLYUBOV–TOLMACHYOV LOGARITHM for expressing the dimensions of localized elec-
(N.N. Bogolyubov, V.V. Tolmachyov, 1958) tron states in nonmetallic crystals.
The logarithm of the ratio of the Fermi energy
in metals E F ∼ 1–10 eV to the mean energy of BOHR–VAN LEEUWEN THEOREM (N. Bohr,
phonons ωD ∼ 0.1–0.01 eV (ωD is the Debye 1911; generalized by J. van Leeuwen, 1919)
frequency) that characterizes the extent of the ef- A theorem of classical statistical physics,
fective weakening of the Coulomb repulsion be- which states that the magnetization of a system of
tween the electrons in a superconductor: charged particles in a constant external magnetic
field under the conditions of statistical equilibrium
c = Vc is equal to zero. The theorem demonstrates that it
V ,
1 + N(0)Vc ln[EF /(ωD )] is impossible, within the framework of the clas-
sical statistical mechanics of charged particles, to
where V c is the matrix element of the screened
explain ferromagnetism, paramagnetism and dia-
Coulomb interaction that is averaged over the
magnetism. As was shown later, the magnetism of
Fermi surface, and N(0) is the density of states a substance is caused by quantum properties of the
at the Fermi level. The extra (relative to screen- particles comprising it.
ing) weakening of the repulsion is caused by the
decreased probability of the appearance of two BOLOMETER
electrons within the limits of the screening region A device for measuring the radiated energy
due to Coulomb scattering. It facilitates electron (power) in the infrared and optical wavelength
Cooper pair formation via an effective attraction,
ranges. It is in common use for various detection
which results from virtual phonon exchange. schemes with narrow ranges of extreme sensitivity
to particular excitations.
BOHR MAGNETON It was first used by S. Langley (1880) for mea-
See Magneton. surements in the IR range of the solar spectrum.
The principle of bolometer operation is based on
BOHR RADIUS the change of the resistance of the thermosensitive
The radius a B of the orbit of an electron in the element under the action of radiation. Bolome-
ground state of a hydrogen atom. ters are nonselective optical radiation detectors. To
BOLTZMANN EQUATION 125
obtain equivalent spectral sensitivity in a given integrating Eq. (1) over all momenta p is referred
range, absorbing coatings are applied to the sensi- to as the Boltzmann distribution function:
tive elements of the bolometer. There are metallic,
U (r)
semiconductor (thermistor) and insulator bolome- f (r) = A exp − ,
ters. The use of cryogenic effects allowed pro- kB T
ducing cooled bolometers based on the conduc- while Eq. (1) is called the Maxwell–Boltzmann
tion of shallow levels in semiconductors, as well as
distribution. Integration of Eq. (1) over coordi-
superconducting (isothermal and nonisothermal)
nates yields the Maxwell distribution of velocities.
bolometers.
In a quantum description of particles characterized
by a set of single-particle energy levels Ei , the
BOLTZMANN CONSTANT (L. Boltzmann) Boltzmann distribution takes the form
One of the fundamental physical constants, a
parameter of the statistical Boltzmann distribution μ − Ei
ni = exp , (2)
and Gibbs distribution. Through them, the Boltz- kB T
mann constant k B enters various thermodynamic
relations, for example, it relates the thermody- where ni is the average number of particles in the
namic probability W of a system state to the en- state with energy Ei , μ is the chemical potential
tropy S through the expression S = k B ln W , and of the system that acts as the normalization
factor
determines the mean thermal energy k B T /2 of a determined by setting the sum i ni equal to the
particle per degree of freedom. The gas constant R total number of particles, N . The Boltzmann dis-
involved in the equation of state for one mole of tribution is the high temperature limiting case of
the ideal gas, pV = RT (p is pressure, V is gas the more accurate Bose–Einstein and Fermi–Dirac
volume), is related to the Boltzmann constant by distribution functions (see Bose–Einstein statis-
k B = R/N A , where N A is the Avogadro number. tics, Fermi–Dirac statistics) that take into account
The standard value based on the 1986 CODATA the quantum character of particle motion, and this
recommendations is k B = 1.38066·10−23 J/K. limiting case is valid for an ideal gas with a small
density of particles at high temperatures. In the
BOLTZMANN DISTRIBUTION physics of semiconductors, the Boltzmann distri-
A distribution function of coordinates and mo- bution is widely used in the statistics of a gas of
menta for particles or quasi-particles of an ideal quasi-particles (electrons and holes) to determine
nondegenerate gas in the state of thermodynamic their equilibrium concentrations and other specific
equilibrium. The probability dW of a particle to parameters.
have a momentum in a range p ⇒ p + dp and a
coordinate in a range r ⇒ r + dr is related to the BOLTZMANN EQUATION (L. Boltzmann, 1872)
Boltzmann distribution fn (r, p) by An integro-differential kinetic equation that
provides the time (t) evolution of the one-particle
dW (r, p) = fn (r, p) dr dp, distribution function (f ) of a low-density mon-
where in the classical case we have the approxi- atomic gas. The Boltzmann equation has the form
mation ∂f 1
+ (p, ∇x f ) + (F, ∇p f ) = ρLp (f, f ),
ε(p) + U (r) ∂t m
fn (r, p) = A exp − . (1)
kB T
where f (x, p, t) is the probability density func-
Here ε(p) is the kinetic energy, U (r) is the poten- tion in the phase space of coordinates and mo-
tial energy, k B is the Boltzmann constant, and A is menta of a particle (x, p), m is the particle mass,
the normalization factor found from the condition F is the external force field applied at the point
that the probability of a particle to be present in (x, p) of the phase space, and ρ is a dimension-
the phase space equals unity, i.e. W = dW = 1. less parameter proportional to the mean density of
Sometimes the following expression obtained by gas particles. The collision operator (or collision
126 BOND
orbit (see Bohr radius), v = Ze2 /(4πε0 ) (Z is where Li = Ni ai , i = x, y, z, are the Cartesian
the atomic number, e is the electron charge). The coordinates; ai and Ni are the crystal lattice con-
Born approximation finds wide use in the theory stant and the number of lattice periods in the di-
of current-carrier scattering and scattering of other rection i, respectively. Born–von Kármán bound-
microscopic particles in solids, in the treatment of ary conditions allow one to determine the number
ionization processes, e.g., light absorption by lo- of allowed states of electrons with different quasi-
cal electron centers (see Local electronic levels) in wave vectors k in a unit volume of the recipro-
nonmetallic crystals, etc. cal space. Then, an extra (in addition to the Bloch
theorem) condition ψkλ (r + L) = ψkλ (r) (where
BORN–HABER CYCLE (M. Born, F. Haber, 1919)
L is a linear combination of vectors L1 , L2 , and
A method of approximate determination of
L3 with integer coefficients) is imposed on the
the atomic bond energy in crystals (mainly alkali-
wave function. Hence, the number of allowed k
halide crystals) from experimental data on reaction
values equals the number of unit cells in the vol-
energies whose sequence (cycle) results in the for-
ume V = L1 × L2 × L3 and the density of permit-
mation of a crystal from its gaseous atoms.
ted states in k -space equals V /(2π)3 . None of the
The bonding energy at the temperature of ab-
observed values, including the density of electron
solute zero is determined as a sum of energies
states referred to one unit cell, or per unit volume
needed for (i) heating the crystal to room tem-
perature (1.5 to 3 kcal/mole); (ii) separating the of direct space, are dependent on the selection of
crystal into the pure metal and the diatomic halide L1 , L2 , L3 or V .
gas (60 to 150 kcal/mole); (iii) sublimation of
the metal (20 to 40 kcal/mole); (iv) dissocia- BORON, B
tion of the halogen into a monatomic gas (25 to A chemical element of Group III of the pe-
30 kcal/mole); (v) cooling the gases to absolute riodic system with atomic number 5 and atomic
zero (to 2.9 kcal/mole); (vi) removing the valence mass 10.81.
electrons from alkali atoms (90 to 120 kcal/mole); The isotope composition of natural boron:
10 B (19.57%) and 11 B (80.43%); the radioac-
(vii) formation of negative halide ions from atoms
and electrons (70 to 80 kcal/mole). As a result of tive isotopes 8 B, 12 B and 13 B are known (half-
summing, the bonding energy turns out to equal lives are fractions of a second). Outer electronic
∼180 kcal/mole for LiI, ∼185 kcal/mole for shell configuration is 2s 2 2p1 . Successive ion-
NaCl, ∼170 kcal/mole for KCl, ∼160 kcal/mole ization energies are 8.298; 25.16 and 37.93 eV.
for KBr, ∼145 kcal/mole for CsI, and so on. Ap- Atomic radius is 0.091 nm; ionic radius of B2+ is
plication of methods similar to the Born–Haber 0.023 nm. The most typical oxidation state is +3,
cycle allows one to estimate the energies of vari- and in compounds with metals −3 (borides). In
ous reactions in solids, which can be included into the free state boron exists in the form of brown
one or another cycle, using the known energies of fine-crystal powder (so-called amorphous boron)
other components of the cycle, e.g., from the en- and of dark-gray crystals (crystalline boron). The
ergy of formation of Frenkel pairs (see Unstable most important boron modifications are simple
Frenkel pair) or Schottky pairs in different charge α-rhombohedral B (I), tetragonal B (II), and com-
states (see Schottky defect). plicated β-rhombohedral B (III). The α-rhombo-
hedral boron lattice parameters are: a = 0.506 nm,
BORN–VON KÁRMÁN BOUNDARY
α = 58.1◦ , density is 2.365 g/cm3 ; the tetrago-
CONDITIONS (M. Born, Th. von Kármán, 1912)
nal boron lattice parameters are a = 1.012, c =
Periodic boundary conditions imposed on the
1.414 nm, and the density is 2.365 g/cm3 ; for
Bloch function of a particle in a periodic potential,
β-rhombohedral boron we have: a = 1.014 nm,
or on a plane wave
α = 65.3◦ , density is 2330 kg/m3 . The modi-
ϕ(x + Lx , y, z) = ϕ(x, y + Ly , z) fications B (I) and B (II) under heating above
1773 K irreversibly transform to B (III). T melting
= ϕ(x, y, z + Lz )
of B (III) is 2346 K; T boiling is around 3973 K;
= ϕ(x, y, z), thermal coefficient of the linear expansion is
128 BORON CARBIDE
(1.1–8.3)·10−6 K−1 (from 293 to 1023 K). Ther- with coal in arc furnaces, by synthesis from the
mal conductivity coefficient of crystalline elements, by crystallization from melts, from their
boron is 25.96 W·m−1 ·K−1 (at room temper- gaseous compounds. Boron carbide is used as an
ature); specific heat of amorphous boron is abrasive material, for the borating of metals and
12.97 J·mole−1 ·K−1 , specific heat of rhombo- alloys, for adsorption of neutrons.
hedral boron is 11.1 J·mole−1 ·K−1 . Standard
enthalpy is 5.895 J·mole−1 ·K−1 for crystalline BORON NITRIDE, BN
boron and 6.548 J·mole−1 ·K−1 for amorphous Chemical compound of boron with nitrogen,
boron; Debye temperature is 1220 K. with three modifications: α-BN, β-BN, γ -BN.
Crystalline boron is a semiconductor; the for- The α-BN compound is hexagonal with space
bidden gap width determined from electrical and group P 63 /mmc (D6h 4 ), crystal lattice parame-
optical measurements is 1.42 and 1.53 eV, respec- ters: a = 0.2504 nm, c = 0.6661 nm; β-BN is
tively. Boron is a diamagnet, with specific mag- cubic with sphalerite-type space group F 4̄3m
netic susceptibility −0.62·10−9 (at room temper- (Td2 ), a = 0.3615 nm; γ -BN has the hexago-
ature). It is characterized by high hardness. At nal wurtzite-like structure, space group P 63 mc
low temperatures, boron is a brittle material; above 4 ), a = 0.255 nm, c = 0.423 nm. The α-BN
(C3v
1800 ◦ C, it begins to deform plastically. compound transforms into β-BN under high pres-
Chemically, boron is a nonmetal. Elementary sure and temperature (P [kbar] = 11.5 + 0.025T ).
boron has low chemically activity: in the air it In the course of heating under normal condi-
burns at 973 K, forming B2 O3 ; with nitrogen tions, β-BN transforms to α-BN at temperatures
above 1473 K, it forms boron nitrides; with car- above 1400 ◦ C; α-BN transforms to γ -BN at
bon above 1573 K, it forms boron carbides. It does shock compression 120 to 130 kbar. The density is
not interact directly with hydrogen, but a whole 2.20 g/cm3 for α-BN, 3.45 g/cm3 for β-BN and
number of hydride compounds Bn Hm is known 1.8 g/cm3 for γ -BN; the thermal conductivity is
(see Hydrides), obtained indirectly using silicon. 21 W·m−1 ·K−1 for α-BN and 1800 W·m−1 ·K−1
Above 1273 K, there are silicon borides; formed for β-BN. Microhardness of β-BN is 60 GPa;
with phosphorus and arsenic above 1173 K are elastic modulus is 953 GPa. Boron nitride is an
phosphides and arsenides. It finds wide application insulating material with electrical resistivity ρ >
in the production of materials (boroplastics), high- 1015 ·m and band gap width of β-BN ∼10 eV.
temperature strong and superhard materials (e.g., Boron nitride is obtained by nitration of boron.
borazon), neutron-adsorbing materials for nuclear The α-BN form is used as fireproof and electrical
power engineering, semiconductor materials, cor- insulating materials; β-BN is a tool and abrasive
rosion resistant and antifriction coatings. material.
Cubic boron nitride is a superhard material
BORON CARBIDE, B4 C having no natural analogs. It was obtained by
The only chemical compound of boron with Robert H. Wentorf, Gr. (1956) under high pressure
carbon with a broad homogeneity range (molar (above 4.0 GPa) and temperature (above 1470 K)
content of C is 8.6 to 21%). It forms a rhombo- from hexagonal boron nitride in the presence of
hedral lattice of space group R3m (C3v 5 ), and the
alkali and alkali-earth metals, lead, antimony, tin
parameters of the crystal lattice in the homogene- and nitrides of the said metals. The rules of forma-
ity range vary from a = 0.5602 to 0.5616 nm, tion of binary compounds with a tetrahedral con-
c = 1.2066 to 1.2120 nm. T melting = 2400 ◦ C, figuration are as follows. The components should
density is 2.51 g/cm3 , thermal conductivity is belong to groups equidistant from Group IV of
121.4 W·m−1 ·K−1 at 300 K, microhardness is the periodic system; the average number of va-
45.0 GPa, elastic modulus is 450 GPa. Boron car- lence electrons in this compound should equal
bide is a semiconductor of p-type, band gap width four. The majority of binary alloys with sphalerite
is 0.84 eV, electrical resistivity is 0.01 ·m. Boron and wurtzite structures satisfy these rules. Cu-
carbide does not decompose in mineral acids and bic boron nitride with sphalerite-type (β-BN) and
alkalis. It is obtained through interaction of B2 O3 wurtzite-type (γ -BN) lattices are characterized by
BOSE–EINSTEIN CONDENSATION 129
the tetrahedral arrangement of atoms with each level population, as well as the possibility of such
atom of one type surrounded by four atoms of the an accumulation of photons in a resonator. An ana-
other type. In β-BN and γ -BN lattices the four log of a Bose condensate is the ground state in sys-
B–N bonds have the same length 0.157 nm, with tems with spontaneous symmetry breaking, which
109.5◦ angles between them. The crystal lattice is described by a complex order parameter, in par-
of β-BN is similar to that of diamond with spac- ticular, the so-called Higgs field (see Higgs mech-
ing 0.3615 nm. Cubic boron nitride is an insulator anism).
with a density ≈3.465 g/cm3 . Nanotubes can be
made from boron nitride. BOSE CONDENSATION
The same as Bose–Einstein condensation.
BORRMAN EFFECT
The same as Anomalous passage of X-rays. BOSE–EINSTEIN CONDENSATION, Bose
condensation
BORRMAN TRIANGLE (G. Borrman) The property of Bose particles (bosons) with
A triangle in the scattering plane with one apex integer spins to accumulate in macroscopically
lying in the entry surface of a crystal, the two side large amounts in a single quantum state described
lines directed along the wave vectors of the inci- by a unique (coherent) wave function (see Bose
dent wave K 0 and the Bragg-diffracted wave K d condensate). Bose–Einstein condensation is a
(see Bragg diffraction), respectively, and the base property of three-dimensional Bose systems only.
formed by the segment of the exit surface of the It is missing in two- and one-dimensional systems.
crystal intercepted by these lines. If an X-ray beam Bose–Einstein condensation of excitons and
is incident along K 0 at a certain point on the en- biexcitons is a variety this condensation for the
try surface, and the crystal is oriented for Bragg case of quasi-particles (excitons and biexcitons,
reflection, then the wave field inside a thin crys- between which repulsion forces dominate). Bose–
tal propagates only within the Borrman triangle. Einstein condensation of excitons interacting with
Inside a Borrman triangle the response functions phonons leads to renormalization of the energy
also have finite values (see Dynamic radiation scat- spectrum of both subsystems and alters the shape
tering). of light absorption and luminescence bands. The
bands contain additional sharp peaks generated
BOSE CONDENSATE by a condensed phase and sloping wings aris-
A macroscopically large number of particles ing from condensate-free excitons. The Fermi na-
with integer spin (bosons) in an ideal or slightly ture of electrons and holes forming excitons leads
nonideal Bose gas, which are in their lowest state to the non-Boseity of exciton operators. Excitons
with zero point energy and momentum at a tem- metallize when their concentration nex becomes
−3
perature below the Bose–Einstein condensation comparable to the inverse of the volume aex of
temperature. Since the Pauli exclusion principle the exciton, where aex is its Bohr radius. The
does not apply, bosons can accumulate in the low- Bose–Einstein condensation of excitons is realiz-
est energy quantum state, all with the same en- able under the condition naex3 < 1 only.
ergy forming a coherent condensate. The latter is Bose–Einstein condensation of dipole-active
described by a common complex wave function excitons and photons (or polaritons with wave vec-
Ψ = |Ψ |eiΘ with a magnitude |Ψ | and phase Θ tor k0 = 0) is of special interest. In this case,
that are constant or slowly varying in time. This the Bose-condensed state can be considered as
fundamental property of bosons underlies the phe- an intense coherent wave described by a coher-
nomena of super uidity in liquid 4 He (see Bose ent Glauber state with a macroscopic amplitude.
u id) and superconductivity in metals (see Cooper Equations describing inhomogeneous in space and
pairs). Related to this is the possibility of laser time weakly of excitons and of photons have been
generation (see Lasers) of a macroscopically large derived; the diagram techniques for nonequilib-
number of photons with the same wavelength un- rium states has been developed to describe coher-
der the conditions of a nonequilibrium inverted ent quasi-particles existing simultaneously with
130 BOSE–EINSTEIN STATISTICS
a condensate. On this basis, the features of co- the lowest states with ε = 0; their number N0 =
herent nonlinear propagation of light in densely N[1 − (T /T0 )3/2 ] where
condensed media are assigned to the exciton
N 2/3 2
spectral range. The values of wave vectors and fre- T0 ∼
= 3.31
quencies have been found, where, due to the ab- gV mk
solute and convective instability of the spectrum, is called the condensation temperature, g is the
generation and amplification of waves is possi- degeneracy factor, m is the particle mass, V is
ble. Rearrangement of the energy spectrum of the the system volume. Other particles are distributed
phonoriton type (see Phonoriton) is due to pecu- over states according to Eq. (1) at μ = 0. Such
liarities of the distribution functions of incoherent a transition is called a Bose–Einstein condensa-
quasi-particles and conditions of stimulated Bril- tion (Bose condensation). The Bose–Einstein dis-
louin scattering of light. The role of quantum fluc- tribution determines the efficiency of scattering of
tuations and the influence of incoherent polaritons particles moving in a solid (electrons and holes,
upon the statistical properties of an intense polari- neutrons, photons etc.) by photons, nonradiative
ton wave have been clarified. A stimulated Bose– quantum transitions in solids, phonon–phonon in-
Einstein condensation of excitons and polaritons teractions and also values of electrical conductiv-
under the action of coherent resonant laser radi- ity and thermal conductivity, speci c heat, acousti-
ation, and a stimulated Bose–Einstein condensa- cal properties, parameters of optical bands, spec-
tion of biexcitons obtained by two-photon excita- tral composition of black body radiation, and so
tion with counter-directed or parallel laser beams, on.
were experimentally carried out.
BOSE FLUID
BOSE–EINSTEIN STATISTICS (Sh. Bose, A quantum liquid that condenses at low tem-
A. Einstein, 1924) peratures and consists of interacting bosons, i.e.
The statistics describing the distribution in particles with integer (in units) intrinsic angu-
quantum states of a system of very many identi- lar momentum (spin). An example of a Bose fluid
cal particles with zero or integer (in units) spin. is liquid helium (4 He) at T < 4.2 K. Below some
Particles as well as quasi-particles that obey critical temperature (see Lambda point) Bose flu-
Bose–Einstein statistics are called bosons; includ- ids can transform into the superfluid state (see Su-
ing photons, phonons, magnons, excitons, bipo- per u idity).
larons, π -mesons, and so on. The wave function
of bosons is symmetric with respect to the ex- BOSE GAS (Sh. Bose)
change of any pair of particles. A particular fea- A quantum system of weakly interacting par-
ture of bosons is the possibility for any number of ticles (quasi-particles) with integer spins (bosons)
them to be in the same quantum state. In view of subject to Bose–Einstein statistics with the distri-
this property, it follows from the Gibbs distribution bution function (at spin σ = 0)
that the average number of particles ni in quantum −1
ε −μ
state i with energy εi for a boson ideal gas (Bose nk = exp k −1 ,
gas) in thermodynamic equilibrium is kB T
−1 where εk is the energy of the kth quantum state
εi − μ
ni = exp −1 , (1) or of a free particle with momentum k, μ is the
kB T chemical potential (μ 0), T is the absolute tem-
where k B is the Boltzmann constant, μ is the perature. A Bose gas of neutral particles is ideal
chemical potential determined by the total
number for a small density n, when n1/3 r0 1, where
N of particles in the system, and N = i ni (μ). r0 denotes the action radius of exchange or van
In a system with a variable number of particles, der Waals forces, and a Bose gas of electrically
μ = 0; under constant N , μ < 0. Eq. (1) is called charged bosons (e.g., bipolarons) becomes weakly
the Bose–Einstein distribution. At low temper- nonideal at a high density, when n1/3 a0
1,
atures, a finite portion of the particles pass to where a0 denotes the Bohr radius of particles. In
BOUNDARY EFFECTS 131
a Bose gas with a given (constant) number of the grain boundary to the point where the diffus-
particles, under a pronounced decrease of tem- ing atom concentration is reduced by the factor of
perature, Bose–Einstein condensation takes place, e = 2.718 from its maximum value. For some met-
while in a Bose gas with a variable number of par- als (e.g., for self-diffusion in nickel) this distance
ticles (quasi-particles) determined by the thermo- is 4 to 5 μm. Diffusion over a phase boundary is
dynamical (thermal) equilibrium condition (e.g., of great importance for phase transition kinetics.
the phonon gas in a crystal or in liquid helium), On a phase boundary, like on a grain boundary, a
the Bose–Einstein condensation does not occur. change of the crystal lattice state, defect cluster-
ing, and chemical segregation are possible factors
BOSE STATISTICS that enhance the atomic mobility compared to that
The same as Bose–Einstein statistics. in the bulk. The principal methods for investigat-
ing grain diffusion are macro-, micro- and electron
BOTTLENECK OF PHONONS microscopic autoradiography, and the method of
See Phonon bottleneck. secondary ion–ion emission.
on the surface, but also the degree of order in the and time coordinates written in a dimensionless
interior. Near the surface the degree of order dif- form (subscripts denote partial differentiation).
fers from that in the bulk (the former being higher, This equation was introduced by Boussinesq
as a rule). It is experimentally possible to form, to describe the propagation of waves of small but
at the boundary of a nematic liquid crystal, a thin finite amplitude when the dissipation is negligi-
layer of smectic liquid crystal that is characterized ble, and characteristic wavelengths significantly
by extra translational ordering of the molecules, exceed the reservoir depth. In solid-state physics,
and by a higher degree of orientational ordering of the Boussinesq equation is also referred to as a
their long axes. The spatial inhomogeneity of the nonlinear string equation in connection with its
degree of ordering near the liquid crystal surface application to the theory of nonlinear waves of
results in a macroscopic electric polarization of the a one-dimensional lattice and a nonlinear string.
subsurface layer which is referred to as order elec- The Boussinesq equation is more general than the
tricity (compare with the exoelectric effect). Korteweg–de Vries equation (see Soliton) because
it takes into account motion in both directions.
BOUND WATER, combined water The Boussinesq equation is completely integrable
Water interacting with the surface of min- by the inverse spectral transformation associated
eral particles, and thereby acquiring some spe- with a scattering problem for a third order oper-
cial properties. It freezes at temperatures below ator. Therefore, it has all attributes pertinent for
0 ◦ C. The “bonding” mechanism of water in rocks integrable systems: an infinite series of polyno-
involves the processes of physical and chemical mial conservation laws, Bäcklund transformation,
adsorption, capillary condensation (see Capillary and many-soliton solutions. The simplest single-
phenomena), and osmosis. The action of surface soliton solution of the Boussinesq equation is
forces in minerals results in the perturbation of
hydrogen bonds between the molecules of water, 1 2 2 x ± ct
U (x, t) = c − 1 Sech ,
which is then transferred in succession from mole- 4 Δ
cule to molecule of water. Bound water is usually where c is the soliton velocity, and Δ =
separated into adsorbed, capillary, and osmotic 2(c2 − 1)−1/2 is its width. It is a single para-
types. Adsorbed water is more strongly held at the meter family of stationary solitary waves. There
surface of minerals, and is therefore considered are several modifications of the Boussinesq equa-
strongly bound, while capillary and osmotic water tions which cannot be integrated. In particular,
are both weakly bound. Strongly bound water is Ut t − Uxx − 6(U 2 )xx − Uxxt t = 0 is the ad-
nonuniform. One may identify “island” water ad- justed Boussinesq equation used to describe lon-
sorption (bonding energy within 40–130 kJ/mol) gitudinal acoustic waves in rods; and Ut t − Uxx −
and multilayer adsorption (bonding energy within 6(U 3 )xx − Uxxt t = 0 is the modified adjusted
0.4–40 kJ/mol). Capillary and osmotic water are Boussinesq equation used in the theory of nonlin-
contained in rocks at humidities exceeding the ear seismic waves.
maximum rock hygroscopicity. They both differ in
their properties from either free or strongly bound
BOUSSINESQ–GALYORKIN SOLUTION
water. Osmotic water may only be contained in
(J. Boussinesq, B.G. Galyorkin)
rocks if some ions are present. It is the type most
A general solution of the equilibrium equations
weakly bound to the surface, with its bonding en-
of an elastic medium in the absence of bulk forces
ergy remaining below 0.4 kJ/mol. The freezing
expressed via an arbitrary biharmonic vector. If
temperature for capillary and osmotic water varies
the equilibrium equation is written in the form
from −0.2 to −(2–3) ◦ C.
(1 − 2ν)∇ 2 u + grad div u = 0, where u is the dis-
placement vector, and ν is the Poisson ratio, then
BOUSSINESQ EQUATION (J. Boussinesq, 1871)
the Boussinesq–Galyorkin solution has the form:
An equation of the form Ut t − Uxx −
6(U 2 )xx − Uxxxx = 0, where U (x, t) is the veloc- 1
ity field; and x and t are, respectively, the space u = ∇2f − (1 − ν) × grad div f ,
2
BRAGG MULTIPLE SCATTERING 133
where f is an arbitrary biharmonic vector, i.e. a effect) and thermal scattering. At small scattering
vector satisfying the equation ∇ 2 ∇ 2 f = 0. angles (ε 1◦ ), the contribution of Bragg double
reflections is significant and exceeds the thermal
BRAGG DIFFRACTION (W.L. Bragg) and the Compton scattering by several orders of
A type of diffraction of X-rays, neutrons, elec- magnitude. The small-angle scattering of X-rays
trons, etc., by crystals when the reflection from due to Bragg double reflection allows one to de-
each plane of atoms is specular, and the rays re- termine the size of grains and blocks in polycrys-
flected from successive planes constructively in- tals, their orientation distribution, and the angu-
terfere with each other. Symmetric and asymmet- lar block disorientation (see Bragg multiple scat-
ric Bragg diffraction are distinguished in which tering).
the reflecting planes (hkl) are parallel or not par-
allel to the entrance plane of the crystal, respec- BRAGG LAW, Bragg relation (W.H. Bragg,
tively. The most important effect in Bragg dif- W.L. Bragg, G.V. Vulf, 1913)
fraction is the total interference reflection of the The main equation of X-ray diffraction. X-ray
transmitted wave in a small angular interval (for diffraction from the atomic lattice of a monocrys-
X-rays, a few seconds of arc) due to the extinction tal can be formally described as reflection from
effect when the wave field penetrates a crystal to atomic planes. According to the Bragg law, the
a depth of the order of the extinction length. The angle of reflection (Bragg angle) θ between the
atomic plane and diffracted beam satisfies the con-
reflectivity of a crystal of finite thickness in Bragg
dition
diffraction has some subsidiary maxima in its an-
gular dependence, which are the analogues of the 2dhkl sin θ = nλ,
pendulum solution in Laue diffraction. Bragg dif-
where dhkl is the interplane distance between the
fraction is mostly employed to check the structural
series of parallel atomic planes with crystallo-
perfection of the near-surface layers of crystals.
graphic indices (hkl), λ is the X-ray wavelength,
Unlike Laue diffraction, Bragg diffraction is not
and n = 1, 2, 3, . . . is the order of the reflection.
limited by sample thickness t ; and for thick sam-
ples t
tn , where tn is the absorption length, the BRAGG MULTIPLE SCATTERING
diffraction pattern is simpler, and it is more readily Scattering of radiation in a crystal under the
interpreted. conditions when the Bragg relation (see Bragg
law) is valid for two or n reflections simultane-
BRAGG DOUBLE REFLECTIONS ously. One can represent the conditions of the
This diffraction occurs when an X-ray, elec- Bragg multiple scattering with the help of an
tron, neutron, etc., beam reflected in accordance Ewald sphere in k -space: the Ewald sphere with
with the Bragg law from a family of parallel planes radius |k 0 | = 1/λ corresponds to the incident ra-
in a crystal reflects again from the planes of the diation with wavelength λ. Vectors k 0 , k 1 , k 2
same family. corresponding to the incident wave and waves re-
The direction of propagation and the intensity flected from reciprocal lattice sites (1) and (2) for
of Bragg double reflections are determined by the the case of triple Bragg scattering are shown in
crystal structure of a crystal, and by the experi- the figure (n = 3, including the primary beam).
mental conditions. While the first and the second Here, |k 0 | = |k 1 | = |k 2 |. By rotating the crystal
reflections occur in crystal regions disoriented by about the scattering vector (straight line H01 con-
the angle δ, the direction of Bragg double reflec- necting sites (0) and (1)), one can remove site (2)
tions differs from that of the primary radiation by from the Ewald sphere, i.e. violate the Bragg con-
the angle ε = δ sin θ where θ is the Bragg an- dition for the site and thus exclude the triple dy-
gle. There are three kinds of Bragg double reflec- namic scattering. At this, site (1) remains at the
tions in polycrystals: intrablock, intragrain and in- sphere and the Bragg condition applies to it (see
tergrain ones. At large scattering angles (ε 10◦ ), Bragg double re ections). The greater the dimen-
these reflections are veiled by the diffuse back- sions of a crystal unit cell and the shorter the wave-
ground due to Compton scattering (see Compton length λ (i.e. the larger the Ewald sphere radius),
134 BRAGG RULE
BRAKING RADIATION
See Bremsstrahlung.
BRANCHING OF FRACTURES
A process of the separation (bifurcation or
polyfurcation) of a main-line fracture into two or
more secondary ones. One can distinguish quasi-
static branching of fractures caused, in particular,
by processes of plastic deformation at its vertex,
and dynamic branching which is observed when
the fracture attains about 60% of the speed of
transverse elastic waves. Due to this the stresses in
Bragg multiple scattering. front of the fast moving fracture stop being a max-
imum, and the failure becomes more probable in
other directions. Dynamic branching of fractures
the greater the number of reciprocal lattice sites has been observed in hardened glasses (see Vitre-
that can be on the sphere. That is, the more diffi- ous state of matter), steels, Plexiglas and epoxy
cult it will be to find a position of the crystal where resins. Branching arises in the core layers of a
there is no Bragg multiple scattering. The eval- sample at the front of the main-line fracture, and
uation of the contribution of the Bragg multiple is directed along the tangent to the curved seg-
scattering to the intensity of the measured reflec- ment of the fracture trajectory. Thus, a secondary
tion is quite a difficult problem to solve, consider- fracture is oriented at an angle of approximately
30◦ with respect to the primary one. After each
ing that it is complicated by the crystal mosaicity,
act of branching the speed of the main-line frac-
the divergence, and the nonmonochromaticity of
ture decreases. Therefore, multiple branching pro-
the incident radiation. Bragg multiple scattering
duces destructions that move in jumps with high
allows one to observe reflections in those direc-
accelerations, up to 108 m/s2 , thus managing to
tions where the Bragg scattering is prohibited
branch at small distances (e.g., in the millimeter
by the crystal symmetry conditions; this is the
range for hardened steel), break loose, pick up crit-
so-called Renninger effect (M. Renninger, 1937).
ical speed, and then branch again. The branching
Bragg multiple scattering in crystals can also play
is stimulated by internal residual stresses that arise
a positive role, e.g., it underlies one of the methods
in the sample being fractured.
of high-precision determination of crystal lattice
parameters.
BRASS
Alloy of copper and zinc belonging to the class
BRAGG RULE of compounds with metallic bonding. It exists over
This rule involves charged particle energy loss a wide range of concentration. At low Zn con-
during the passage through a material with the tent the alpha phase (face-centered cubic lattice)
chemical formula Al Bm Cn . . . and it is expressed is formed. At higher Zn content the stability of the
by the approximate relation new phases (the so-called “electronic phases”) is
determined by the average concentration n of va-
dE dE dE
− =l − +m − lence electrons per atom via the proper matching
dZ dZ A dZ B of the Fermi surface and the Brillouin zone. The
dE three electronic phases are the beta phase (com-
+n − + ···, pound CuZn, body-centered lattice, n = 3/2, be-
dZ C
low 460 ◦ C ordered according to the CsCl type),
where (−dE/dZ)i , i = A, B, C, are the energy gamma phase (compound Cu3 Zn8 , complex cu-
losses per unit length due to the ith component of bic lattice, ordered, contains 52 atoms, n = 21/13)
the material. and epsilon phase (compound CuZn3 , hexagonal
BREAKDOWN 135
close-packed lattice, n = 7/4). Unusual mechan- tices are included in the same Bravais type if their
ical properties are found in β-brass, caused by a coordinates are the same and the parallelepipeds
martensitic transformation (shape memory effect, are centered identically. The Bravais classification
superelasticity, high damping). The figure shows has also a group theory basis: two lattices are in-
the phase diagram. cluded in the same Bravais type if their full sym-
metry groups are isomorphous (see Isomorphism).
BRAVAIS LATTICES (A. Bravais, 1848) These groups (14 out of 73 symmorphic space
The classification of three-dimensional lattices groups) are called Bravais groups and are desig-
into 14 types that takes into account the full point nated by the symbols of the space groups. Each of
symmetry groups (holohedries) and the parallel 14 Bravais types is presented in Table 1 (see next
translation groups. page) by a corresponding Bravais parallelepiped
If a Bravais coordinate system (chosen along with points at the apices, and where appropriate,
the most symmetrical directions of the lattice) co- with centering points indicated (see inside front
incides with the principal lattice system then the cover).
Bravais parallelepiped constructed on it contains
lattice sites only at its apices, i.e. is a primitive par- BREAKDOWN of solids
allelepiped. Such lattices are called primitive or P - A sharp increase of electric current density in
lattices; all other lattices are referred to as centered insulators and semiconductors that occurs when a
lattices. There are three ways to center the Bra- certain critical level of the applied electric field
vais parallelepipeds, I – body-centered, F – face- strength, Ecrit , is reached. The value Ecrit is called
centered, and C – base (or side)-centered. Two lat- the electric strength. The value of Ecrit falls within
136 BREAKDOWN
Lattice type
System Primitive Base-centered Body-centered Face-centered
Triclinic
Monoclinic
Orthorhombic
Trigonal (rhombohedral)
Tetragonal
Hexagonal
Cubic
105 –107 V/cm for various dielectrics and starts current–voltage characteristic has an S-shape. In
from 105 V/cm for semiconductors. During im- the case of an electric breakdown the value of Ecrit
pact ionization it may be as low as 1 to 10 V/cm. only weakly depends on the sample temperature,
One may distinguish between the electric and ther- thickness, and the duration of electric field pulse,
mal mechanisms of breakdown. The mechanism τ , provided τ = 10−7 –10−6 s. Thermal break-
of electric breakdown is typically characterized down results from melting of the material via Joule
by an avalanche increase of the number of free heating when its thermal equilibrium is disturbed
carriers (electrons or electron–hole pairs), for ex- (thermal breakdown without a disturbance of ther-
ample due to ionization by collision. Current la- mal equilibrium is also known, the so-called type
mentation may also occur, i.e. a conducting chan- II thermal breakdown). A combined mechanism of
nel in which the current density is higher than its breakdown is possible, when the Joule heating of
average across the sample may form. This effect a particular region of the sample works to lower
results from an instability of the spatially homo- the value of Ecrit necessary for an electrical break-
geneous distribution of carriers when the sample down in that region. The application of a strong
BREMSSTRAHLUNG 137
field may result in a gradual change of the com- more parts. Breaking stress is defined by the ra-
position and structure of an insulator, so that both tio of the load to the specimen cross-section at
eventually work to lower the value of Ecrit . These the moment of failure. In brittle metals (see Alloy
effects could, for example, lead to the formation brittleness) this stress is achieved via the opening
of color centers in alkali-halide crystals or to ero- of brittle cracks when the external stress remains
sion of polymer surfaces. Such processes are called below the yield limit, and macroscopic plastic de-
electric aging of insulators. In contrast, dielectric formation is absent. When plastic metals subjected
breakdown in semiconductors may not necessar- to stresses exceed the yield limit, there develops a
ily be accompanied by changes in the crystal lat- tough failure or a quasi-brittle failure in them, while
tice. In certain cases, when the conduction elec- the failure stress exceeds the yield limit because
tron concentration is limited and the field strength, of strain hardening. Failure stress in brittle met-
E is moderate, the initial process of breakdown als depends only weakly on the temperature, ap-
does not result in the failure of the material, and proximately as the elastic modulus, while its de-
may be repeatedly reproduced (see Impurity break- pendence in plastic materials is controlled mainly
down, Thermoelectric instability of photocurrent, by the temperature dependence of the yield limit,
Streamer breakdown). and may happen to be stronger.
For different varieties, see also Acoustic break-
down, Electric breakdown, Magnetic breakdown.
BREIT–RABI FORMULA (G. Breit, I. Rabi)
BREAKDOWN NOISE An exact solution of the secular equation for
See Telegraph noise. the spin Hamiltonian of a paramagnetic ion with
electron effective spin S = 1/2, nuclear spin I ,
BREAKING STRAIN and isotropic g, g n and hyperfine (A) tensors.
Strain that results in the break-up of a body The energy levels are as follows:
into several macroscopic parts (failure). The break-
Δ
ing strain produces such a degree of deformation E(F, mF ) = − − gn βn BmF
that it results in failure. This strain is one of the 2(2I + 1)
principal characteristics of a material, delimiting
1 4xmF 1/2
its margin of plasticity. It is determined during the ± Δ 1 + x2 + ,
2 2I + 1
mechanical testing of materials.
Breaking strain depends on the type of state of where F = I + 1/2, mF = m ± 1/2, m = −I,
stress (distension, compression, e xure etc.), i.e. . . . , +I ; Δ = a(I + 1/2) is the splitting in zero
on the “rigidity” of various types of mechanical magnetic field B; a is the isotropic hyper ne inter-
tests, and on the softness factor a = tmax /Smax n ,
action constant, x = (gβ + gn βn )B/Δ, and β and
which is the ratio of the maximum tangential ten- βn are the Bohr and nuclear magnetons, respec-
sion, tmax to the maximum normal reduced stress, tively.
n . With an increase in the material capabil-
Smax
ity to withstand stress the plastic deformation in-
creases. Breaking strain also depends on the type BREMSSTRAHLUNG, braking radiation, collision
of crystal lattice, the structure, testing temperature, radiation
environment, etc. When determining the level of An emission of photons by electrically charged
breaking strain the scheme reproducing both the particles decelerating in an electric field (or in
tensed and the deformed states of the specimen a magnetic field, called magnetic Bremsstrahlung
during tests should be as close as possible to the or synchrotron radiation). The intensity of Brems-
actual operating conditions of the material. strahlung is proportional to the square of the
particle acceleration, i.e. proportional to Z 2 /M 2
BREAKING STRESS, failure stress, failing stress where Z is the atomic number, and M is the
A stress reached during distension tests, which mass of the nucleus being decelerated by the
result in the breakup of the specimen into two or electric field. Owing to the small mass of elec-
138 BREMSSTRAHLUNG, COHERENT
trons and positrons, the intensity of their Brems- BRILLOUIN SCATTERING, also called
strahlung is 3·106 times greater than that for Mandelshtam–Brillouin scattering
protons. This leads to a substantial contribution Inelastic light scattering by thermal acoustic
to the Bremsstrahlung by the energy loss of phonons in condensed media with the light fre-
fast electrons and positrons in a medium (see quency changed by a value equal to a phonon fre-
Charged particle energy loss). The spectrum of quency. (Analogous processes by optical phonons
Bremsstrahlung is continuous. The maximum are called Raman scattering of light). The scatter-
value of the energy of the emitted photon is ing is caused by fluctuating inhomogeneities of
ωM = E − mc2 (here E is the total energy, m is the refractive index, which are related to propaga-
the electron mass, and c is the speed of light). tion of thermal (Debye) excitations of a medium.
In a crystal the constructive interference of From the classical point of view, Brillouin scat-
photons scattered by closely situated atoms of tering can be interpreted as result of the diffrac-
tion of light by thermal elastic waves. By virtue of
the crystal influences some specific features of
the diffracted light interference, only waves with
the Bremsstrahlung, in particular the intensity
wavelength Λ, which satisfies the Bragg condi-
of the Bremsstrahlung increases (see Coherent
tion λl = 2nΛ sin(θ/2) (where n is the refrac-
Bremsstrahlung). These changes are especially
tive index of the medium, λl is wavelength in
significant in the case of the channeling of elec-
vacuum, θ is angle of scattering), are responsi-
trons and positrons (see Quasi-characteristic radi-
ble for scattering in a given direction. Further-
ation). See also Radiation by charged particles.
more, due to the Doppler effect, the light with fre-
quency ωl , which is scattered by elastic waves,
BREMSSTRAHLUNG, COHERENT undergoes a frequency change ωs = ωl ± ω0 , de-
See Coherent Bremsstrahlung. fined by the relationship ω0 /ωl = ±v/(Λωl ) =
±2n(v/c) sin(θ/2), where c is the speed of light in
vacuo, and v is the sound velocity in the medium.
BRILLOUIN FUNCTION (L. Brillouin, 1926)
Elastic waves with frequency ∼1010 Hz (hyper-
Function Bj (x) introduced to describe para- sonic waves) are of importance in the light scatter-
magnetism of noninteracting atoms. For atoms
ing. The line width is determined by hypersound
with the angular momentum j (in units of Planck attenuation.
constant ), the function is The scattering efficiency is of the order of
10−6 scattered photons per incident photon. How-
2j + 1 2j + 1
Bj (x) = coth x ever, if the intensity of the incident light ex-
2j 2j ceeds some threshold value (∼1010 W/cm2 ), an
avalanche-like growth of the amplitudes of sound
1 x and of scattered waves occurs, a process called
− coth .
2j 2j stimulated Brillouin scattering. Brillouin scatter-
ing is used for studying material properties: sound
The Brillouin function determines the mean wave propagation velocity and decay rate, mole-
value of the projection of j on the direction of cular structure, intermolecular interactions, etc.
the magnetic field H (directed along the z axis) Stimulated Brillouin scattering is used for the gen-
jz = j Bj (x), x = gj |μB |H /(k B T ), where μB eration of intense hypersonic waves, and in exper-
is the Bohr magneton and gj is the Landé g -factor. iments on wave front reversal of light waves. Res-
The Brillouin function is the basis for the quan- onance Brillouin scattering is an effective method
tum theory of paramagnetism of substances with for investigating excitonic polaritons.
uncompensated atomic magnetic moments, and of
magnetically ordered materials (when described BRILLOUIN ZONE
in the framework of the molecular eld approxi- A limited region in the reciprocal space
mation; in this case, H is the effective magnetic (k -space) of wave vectors inside which the quasi-
field). particle energies in the crystal take on the entire
BRITTLE FRACTURE 139
spectrum of permitted values. The overall k-space where a fracture of atomic bonds occurs through
can be partitioned into an infinite number of Bril- microspalling (without attendant plastic deforma-
louin zones of the same dimensions, and equal in tion). However, in so far as the origin of microc-
volume to a reciprocal lattice cell. The use of a sin- racks in metals is possible only as a result of the
gle Brillouin zone is due to the crystal periodicity displacement and pile-up of dislocations, so a fail-
since the quasi-particle energy E(k) is periodic in ure in metallic materials can only be of a quasi-
the quasi-momentum space with the periodicity of brittle type (see Quasi-brittle failure) since a condi-
a reciprocal lattice vector G: E(k) = E(k + G). tion of purely elastic deformation does not appear
A region with its center at the point k = 0 is at their failure.
chosen as the first Brillouin zone, or simply re- In practice when a failure of products, con-
ferred to as the Brillouin zone. It is delimited by structions and installations (e.g., hulls of ships)
planes passing through the mid-points of the lines occurs, it is considered to be the brittle type, if
(normally to them) that connect point k = 0 with an overall deformation of the product was elastic
neighboring sites of the reciprocal lattice. A Bril- before its failure, in spite of the fact that a large
louin zone is analogous to a Wigner–Seitz cell in local plastic deformation took place in the focus
real (coordinate) space. Its boundaries are deter- of failure. Hence in practice the term “brittle fail-
mined by the reciprocal space form of the Bragg ure” does not have a precise meaning. However,
equation 2k·G = G2 , involving a reciprocal lat- it can correspond to a complicated type of fail-
tice vector. At the boundaries the quasi-particle ure combining features of overall elastic and lo-
energy E(k) reaches an extremum. To describe cal plastic deformation. The latter forms only in
some physical phenomena it is convenient to use small regions near cracks, or in action zones of
the extended Brillouin zone scheme which makes large residual stresses of a welding or quench-
use of the periodicity of the quasi-particle spec- ing origin. An important feature of such a failure
trum in k-space. is that the average (nominal) failure stress should
be less than the yield limit of the material. Exces-
BRITTLE CRACK sively high strength of the material, low tempera-
A crack, whose propagation is not followed by ture (cold brittleness), high velocity of deforming
a plastic deformation. During brittle crack prop- (a shock load), the presence of cracks or other con-
agation the theoretical strength is reached at its centrators of stresses, the presence of large resid-
vertex. The breaking of interatomic bonds occurs ual macrostresses of the first type; all of these fac-
without activation. Elastic energy, stored in the tors contribute to the significance of brittleness in
material during a deformation process, converts engineering practice.
into surface energy at its propagation. The criti-
cal stress for the brittle crack propagation is de- BRITTLE FRACTURE
termined by the Griffith equation (see Grif th the- A fracture arising from brittle failure. This type
ory). of failure is associated with failure by spalling,
when a progression of the crack is not accom-
BRITTLE FAILURE panied by plastic deformation, and failure occurs
A type of solid failure involving purely elas- over the plane of crystal cleavage through the
tic strains. Ideally only materials with no capa- breaking of atomic bonds. Actual brittle failure
bility for plastic deformation (e.g., glass, ceram- has been found in covalent crystals such as dia-
ics, diamonds, some minerals) undergo brittle fail- mond and silicon carbide. A brittle transcrystalline
ure under conditions of ordinary tension. Metals, failure is possible also in other crystals with a high
due to the presence of dislocations, behave as ide- degree of covalent bonding, where dislocations are
ally brittle only in special circumstances when practically “frozen in”. The typical feature appear-
the failure is initiated in extremely small regions. ing on the surface of a brittle fracture is a step of
These regions have an ideal (defect free) structure spalling arising at the intersection of a screw dislo-
in perfect crystals (whiskers), or in microregions cation and a crack front. A junction of the splitting
near the tip of atomically-sharp submicrocracks, steps makes up a stream-like pattern. The same
140 BRITTLENESS AND TOUGHNESS CRITERION
features of fracture are typical for quasi-brittle fail- the failure is supposed to be tough. In mechan-
ure through splitting accompanied by plastic de- ics, the engineering criterion of brittleness is used,
formation of the region adjacent to the crack. Such whether the product is destroyed by the average
a failure is typical for crystals with body centered (nominal) stress σn below the yield limit σy , i.e.
cubic lattices at ambient temperatures. when σn < σy , which can be satisfied only in the
Intercrystalline brittle failure occurs at sharply presence of stress concentrators (see Stress con-
weakened intergranular linkages. It is observed centration).
along grain boundaries through the catastrophic
crack progress due to the segregation of impuri- BRITTLENESS, BLUE
ties, and also to the presence of internal stresses. See Blue brittleness.
Intercrystalline failures along boundaries of grains
BRITTLENESS OF ALLOYS
and subgrains are distinguished. When a failure
See Alloy brittleness.
runs along grain boundaries a fracture is com-
pletely free from a stream-like pattern, linkage BROADENING OF A LINE
points of three (or more) boundaries of grains ap- See Line (level) width.
pear, and often a “bubble” due to the segregation
of impurities is observed on facets of split grains. BROKEN BOND
The same as Dangling bond.
BRITTLENESS AND TOUGHNESS CRITERION
This is a condition determining the relation BROOKS–HERRING FORMULA (H. Brooks,
of parameters responsible for the type of failure C. Herring)
of a solid under load. There is no unique brit- This expression determines the energy depen-
tleness and toughness criterion, since these cri- dence of the mean free path time τ (E) of charge
teria differ in various areas of science involving carriers of energy E in semiconductors when the
failure problems. In solid state physics, the brit- scattering takes place predominantly from ionized
tleness and toughness criterion is connected with impurities (low temperatures, high impurity con-
the possibility of plastic stress relaxation via the centrations). The Brooks–Herring formula is
generation of dislocations by a vertex of a per- ε2 (2m∗ E 3 )1/2
fectly sharp crack in a crystal. This possibility de- τ (E) = ,
πe4 NΦ(x)
pends on the relation between theoretical strength
against shear τtheor and rupture σtheor and maxi- where ε is the dimensionless dielectric constant;
mum tangential τmax and tensile stress σmax at the m∗ is the effective mass of the carriers, N is
vertex of the crack. If τtheor /σtheor < τmax /σmax , the impurity concentration, Φ(x) = ln(1 + x) −
the crystal is tough (FCC metals: copper, silver, x/(1 + x) where x = 8m∗ E/(2 q 2 ), q is the in-
gold); otherwise, it is brittle (diamond and BCC
verse Debye screening radius. It follows from the
Brooks–Herring formula that scattering by ionized
metals: tungsten, iron, sodium). In physical mate-
impurities becomes more efficient at small carrier
rials science, as a brittleness and toughness crite-
energies and, consequently, at low temperatures.
rion, the one based on the microspalling criterion
The Brooks–Herring formula was derived in the
my be chosen: if j > Kt , then the failure is brittle,
Born approximation of collision theory taking into
if j < Kt , then it is tough (more precisely, a quasi-
account the screening of impurities by free carri-
brittle failure). According to the fractographic clas-
ers. In its derivation the impurities are assumed to
sification of destruction types, one does not suc-
be arranged in a crystal lattice in a disordered way
ceed in giving a clear distinction between brittle-
(see Current carrier scattering).
ness and toughness in metals, and brittleness and
toughness criteria are qualitative: if in a fracture of BUBBLE DOMAIN
brilliant facets of a spalling no appreciable traces See Magnetic bubble domain.
of plastic deformation are seen, then the failure is
supposed to be of a brittle (or quasi-brittle) type; if BUBBLE DOMAIN LATTICE
typical pores or pits are observed (a fibrous break), See Magnetic bubble domain lattice.
BULK PHOTOVOLTAIC EFFECT 141
The first term with the second-rank tensor σij de- can control processes of homogenization, recov-
scribes photoconductivity in an external field Ej . ery, recrystallization, and phase transitions. A spe-
The second term describes the linear bulk pho- cific feature of the action of intensive vibrations
tovoltaic effect (i.e. in linearly polarized light); on a material is the nucleation and growth of mi-
Ej and Ek being the projections of the light polar- crocracks (see Crack) at early stages of loading, a
ization vector. The components of the third-rank fatigue type phenomenon. Depending on the con-
tensor αij k of this effect differ from zero for only ditions of treatment, and the initial state of mate-
20 piezoelectric symmetry groups. The third term rials, acoustic dehardening may occur, e.g., with
describes the circular photovoltaic effect. sound waves incident on initially deformed metals
The microscopic theory of the bulk photo- with a face-centered cubic lattice.
voltaic effect takes into account two basic mech-
anisms. The first, so-called ballistic, mechanism is BURGERS CONTOUR
related to the asymmetry of the distribution func- A contour constructed in a regular crystal lat-
tion of the nonequilibrium carriers in momentum tice by sequentially tracing a closed line from one
space. This asymmetry, in its turn, is caused by atom to another.
the asymmetry of the generation, recombination, It is used to classify defects in crystals, and
and scattering of the nonequilibrium carriers in is of particular importance for the description of
a crystal without a center of symmetry. The sec- linear defects or dislocations (Fig. (a)). If a dis-
ond, so-called shift mechanism, is due to displace- location appears in a crystal enclosed with such
ment of the center of gravity of the wave packet a contour, the contour remains open. In order
to close a Burgers contour, it must be comple-
of an electron during its optical transition between
mented with a vector called a Burgers vector that
bands, or between an impurity center and a band
closes the break (Fig. (b)). A Burgers vector b
in a crystal without a center of symmetry. The two
characterizes a dislocation unambiguously. Linear
mechanisms describe both intrinsic and impurity
needle-shaped defects of different types uncon-
bulk photovoltaic effects, and predict similar mag-
nected with translational slip of a part of a crystal
nitudes of the bulk photovoltaic effect. Therefore,
the experimental differentiation of these mecha-
nisms is a complicated task. At present, the lin-
ear and circular effects are observable in a large
number of ferro- and piezoelectric crystals, includ-
ing ferroelectric niobates and tantalates, Te, GaAs,
GaP, ZnS, and some other piezosemiconductors.
an increase of the free current carrier concentra- where E g is the band gap, mn and mp are the ef-
tion. fective masses of the electron and the hole, respec-
The Burstein–Moss effect is due to filling the tively; and k F is the wave vector corresponding
bottom of the conduction band by electrons in to the Fermi level. At T > 0 K and mn /mp 1,
n-type semiconductors or the top of the valence the magnitude of the Burstein–Moss effect is ap-
band by holes in p-type semiconductors. Then, proximately equal to Δω = E F −4k B T in direct-
only states above the Fermi level E F are available gap semiconductors, and in semiconductors with
for transitions involving photoexcited carriers. As small mn (e.g., InSb), it already appears at carrier
a result, direct optical band-to-band transitions at concentrations 1017 cm−3 . The Burstein–Moss
T = 0 K are possible with the energy of quanta effect is used in studies of the energy band struc-
satisfying the inequality ture of semiconductors.
2 kF2 mn BURSTING
ω ωmin = Eg + 1 + , See Penetration deformation.
2mn mp
Cc
CADMIUM, Cd in cadmium polycrystal at T melting is 1.7·10−12
Chemical element of Group II of the peri- m2 /s. Effective high-energy (3 to 10 MeV) neu-
odic system with atomic number 48 and atomic tron cross-section for absorption and scattering is
mass 112.41. Natural cadmium consists of sta- 4.3 barn, and 2610 barn for thermal (0.025 eV)
ble isotopes with mass numbers: 106 (1.215%), neutrons. Ion-plasma frequency is 8.54 THz; low-
108 (0.875%), 110 (12.39%), 111 (12.75%), 112 temperature linear electronic specific heat (Som-
(24.07%), 113 (12.26%), 114 (28.86%), 116 merfeld coefficient) is 0.70 mJ·mole−1 ·K2 ; elec-
(7.58%). The electronic configuration of the outer trical resistivity is 83.6 n·m (at 293 K) along the
shells is 4d 10 5s 2 . Successive ionization energies principal axis 6 of a monocrystal and 68.7 n·m
are 8.994, 16.908 eV. Atomic radius is 0.148 nm; in a perpendicular direction at room tempera-
radius of Cd2+ ion is 0.097 nm. Oxidation state ture, thermal coefficient of electrical resistance in
is +2. Electronegativity is ∼ =1.5. these directions is 0.0041 and 0.00405 K−1 , re-
In free form, cadmium is a silvery-white forge- spectively; Hall constant is ∼= + 5.7·10−11 m3 /C
able and ductile metal. It has a hexagonal close- at room temperature. Work function of cadmium
packed crystal lattice; a = 0.297912 nm, c = polycrystal is 4.1 eV. Critical temperature of cad-
0.561827 nm (at 298 K). Density is 8.65 g/cm3 mium transition to the superconducting state is
at 293 K; T melting = 594 K; T boiling = 1030 K. 0.517 K; critical magnetic field is 2.8 mT (at 0 K).
Bond energy is 1.16 eV/atom at 0 K. Heat Cadmium is a diamagnetic material with molar
of melting is 6.38 kJ/mole; heat of sublima- magnetic susceptibility −19.7 · 10−6 CGS units
tion is 112.2 kJ/mole; heat of evaporation is (at 300 K); nuclear magnetic moment of 111 Cd
99.9 kJ/mole; specific heat is 0.231 kJ·kg−1 ·K−1 ; isotope is −0.592 nuclear magnetons. Cadmium
Debye temperature is ∼ =200 K; coefficient of is used in nuclear power engineering and as a
the linear thermal expansion of cadmium monocrys- component of alloys for solders, printing blocks,
tal is 54.0·10−6 K−1 along the principal axis 6 jewelry, electrodes.
and +19.6·10−6 K−1 perpendicular to it (at
300 K); coefficient of thermal conductivity is CADMIUM SULFIDE
83.7 W·m−1 ·K−1 along the principal axis 6 and A semiconductor compound of the AII BVI
105 W·m−1 ·K−1 perpendicular to it (at room group (see Semiconductor materials), which crys-
temperature). Adiabatic coefficients of elastic ri- tallizes in the cubic sphalerite structure (zinc
gidity of cadmium monocrystal: c11 = 114.5, blend structure), space group F 4̄3m (Td2 ), and
c12 = 39.5, c13 = 39.9, c33 = 50.85, c44 = 19.85 in the hexagonal wurtzite structure, space group
(in GPa) at 300 K; bulk modulus is ∼ =58 GPa; P 63 mc (C6v 4 ). The unit cell of sphalerite
Young’s modulus is ∼ =65 GPa; shear modulus is contains four molecules with lattice parameter
∼
=25 GPa; Poisson ratio is ∼ =0.31 (at 300 K), a = 0.5835 nm, while that of wurtzite contains
ultimate tensile strength is ∼ =0.08 GPa and rel- two molecules, with lattice parameters a =
ative elongation is ∼ =32%. Brinell hardness is 0.41368 nm, c = 0.6713 nm, and the ratio of an-
∼
=0.18 GPa. Cadmium is easily forged and rolled. ion and cation radii is 0.52. For both structures
Surface tension of liquid cadmium is 0.628 J/m2 the tetrahedral positions of the atoms is charac-
(at 623 K). Self-diffusion coefficient of atoms teristic, the crystals have no center of symmetry,
145
146 CALAMITICS
they are polar; they form polytypes (see Poly- donors; elements of Group I Cu, Ag, replace
typism) as derivatives of basic structures. Below the cations, cation vacancies behave as deep ac-
the pressure 3 GPa a transition to the struc- ceptors. The association of the point defects is
ture of sodium chloride is observed. The cubic characteristic. Conductivity may change from a
modification is stable at considerably lower tem- semimetal to insulator type, no p-type conduc-
peratures than those of growing monocrystals. tivity is observed due to the self-compensation
Data given for the hexagonal wurtzite modifi- phenomenon. Mobility of electrons and holes:
cation are as follows: temperature coefficients
350 cm2 ·W−1 ·s−1 and 30 cm2 ·W−1 ·s−1 . Minor-
of linear expansion with temperature reduction
ity carrier lifetimes (holes) are 10−6 to 10−8 s,
from 600 to 150 K are reduced from 5.9·10−6
diffusion lengths of holes are short (7 to 0.5 μm).
to 0.5·10−6 K−1 , density is 4.825 g/cm3 , melt-
ing temperature is 1748 K, heat of formation The presence of direct band gaps facilitates
◦
Hf,298.15 = −157 kJ·mole−1 , standard entropy effective luminescent and laser radiation. Blue ra-
◦
S298.15 = 71 J·mole−1 ·K−1 , specific heat diation is caused by the annihilation of free and as-
◦ sociated excitons, green, orange, red and infrared
Cp,298.15 = 47.4 J ·mole−1 ·K−1 , heat conduc-
radiations are associated with impurity recombi-
tivity 0.20 W·cm−1 ·K−1 , Debye temperature is nation. At T 15 K energy of excitons Eex A =
219.3 K. Covalent bonding prevails with some 2.550 eV, Eex = 2.564 eV, Eex = 2.632 eV, bond-
B C
admixture of ionic bonding. The phonon spec-
ing energies of excitons: EA = EC = 0.0294 eV,
trum has nine optical and three acoustic branches.
EB = 0.0295 eV. Centers of radiative recombi-
The band gaps are direct, the valence band
nation are centers of photosensitivity. Having the
presents three closely located subbands: Γ9 (the
upper one), Γ8 , Γ7 ; the final conduction band high photosensitivity in a wide spectral range:
has symmetry Γ7 . The band gap width: from UV to near IR, in the X-ray spectral range
Eg,1.8 K = 2.58 eV, Eg,300 K = 2.53 eV. Av- CdS is characterized by high sluggishness, low
erage values of temperature coefficient of Eg temperature and time stability of photocurrent.
variation at constant pressure (dEg /dT )P in the The absence of a center of inversion and a strong
range 90 to 300 K and coefficient (dEg /dP ): electron–phonon interaction determine the piezo-
−5.2·10−4 eV/K and 3.3·10−6 eV/bar. The spin– electric properties. The capability of fabrication by
orbit interaction constant is 0.065 eV, the valence simple methods in the form of powders, tablets,
band splitting due to the crystal e ld action is and caked layers of thin lms with controllable
0.027 eV, electron af nity is 4.73 eV, work func- properties is a pronounced advantage. The unique-
tion is 7.26 eV. Effective masses of electrons, ness and variety of its physical properties allows
holes: m∗n = 0.205, m∗p = 0.7 (⊥ c), m∗p = 5 ( c). its use as a model material of the physics of semi-
Refraction index dispersion and birefringence are conductors and quantum electronics: in photo-,
characterized by the values: ne = 2.726, no = opto-, acoustoelectronics, optics, laser engineer-
2.743 at λ = 515 nm; ne = 2.312, no = 2.996 at ing. Cadmium sulphide is widely used as a crys-
λ = 1500 nm. Static and high-frequency dielec-
tallophor for the creation of the screens of electron
tric constants are: εs = 8.64 ( c), εs = 8.28 (⊥ c),
tubes, including color television, as a photocon-
ε∞ = 5.24.
ductor for the creation of photoresistors, of nu-
The Cd–S system in the phase equilibrium dia-
gram has one compound CdS with a narrow range clear particles radiation sensors, as an element of
of homogeneity. Inherent atomic defects to a large heterojunctions for solar phototransformers.
extent determine the diffusion and solubility of im-
purities, the structure of local centers, optical and
CALAMITICS
electric properties. Elements of Group III Al, Ga,
In, can replace the cations, elements of Group VII Uniaxial liquid crystals with structure units
Cl, Br, I, replace the anions, vacancies of an- (molecules or micelles) of oblong cylindrical
ions, interstitial cations are shallow hydrogen-like shape (cf. Discotics).
CALORIMETRIC COEFFICIENTS 147
diagonal form for noninteracting quasi-particles becomes unstable (imaginary), which corresponds
with a restructured energy spectrum. to an exponential growth of long-wavelength per-
If the initial Hamiltonian has the form turbations in time, and a spontaneous contraction
1 1
(collapse) of the system. For charged bosons (e.g.,
H= Ak ak+ ak + Bk ak a−k + Bk∗ ak∗ a−k
∗ ,
bipolarons) with a long-range Coulomb repulsion
2 2
(1) V (k) = 4πe2 /k 2 (e is the boson charge), the spec-
then by introducing new operators according to the trum (2) of quasi-particles at k = 0 has a finite
relations energy gap E(0) = ωp (ωp is the plasma fre-
+ + + quency) and satisfies the superfluidity criterion,
ak = λk ck + μ∗k c−k , a−k = λ∗k c−k + μk ck which corresponds to superconductivity (see Bipo-
(2) laron mechanism of superconductivity).
the Hamiltonian assumes the form
In the case of electrons (fermions) interact-
H = ε(k)ck+ ck + E0 , ing under the conditions of Cooper pairing (see
1/2 (3) Bardeen–Cooper–Schrieffer theory) one can use
ε(k) = A2k ∓ |Bk |2 . Hamiltonian (1), where k denotes both the mo-
The parameters of the canonical transformation mentum and the spin of the electron, and Ak and
obey the normalization condition Bk are given by
radius (in Gaussian system of units). In a sys- is based on the measurement of the capacitance
tem of conductors, a correspondence between their relaxation kinetics for different combinations of
charges and potentials
is determined by the linear factors applied to fill or ionize the deep lev-
relations qi = j Cij ϕj (Cij is called the capac- els (such as forward and reverse bias pulses of
ity coefficient of the ith and j th conductors). various durations, photoionizations, and temper-
In a system metal–insulator–metal (condenser), atures). Using these methods provides separate
the capacitance is determined by the relation C = measurements of concentrations of deep levels
q/
ϕ where q is the value of charges of op- with different ionization energies, and also deter-
posite sign on every capacitor plate;
ϕ is the mines the cross-sections of charge carrier trapping
potential difference between the plates. For a pla- at the levels. The efficient Lang method measures
nar capacitor C = εA/d where A is the area of the differences between C at two times, t1 and t2 ,
each plate, d is the distance between them, and ε during the course of the relaxation of C with the
is the relative (dimensionless) dielectric constant emission of electrons or holes from individual lev-
of the medium between the plates. For a very els into a corresponding allowed band. There are
small d, the capacity of the system can be very modifications of this method when the relaxation
large. The systems metal–semiconductor, semi- of current rather than capacitance is measured; and
conductor junction, and so on are similar to a taken together they constitute the relaxation spec-
capacitor, and can also have a significant capaci- troscopy of deep levels.
tance (see Barrier capacitance). Variations of the
thickness of the dielectric layer under actions of CAPACITANCE, SURFACE
any kind provide information on the distribution See Surface capacitance.
of electrically active extended defects in a system
(see Capacitance–voltage characteristic, Capaci- CAPACITANCE–VOLTAGE CHARACTERISTIC
tance spectroscopy).
Dependence of the sample capacitance C on
Josephson junctions, with typical capacitances the applied voltage V .
C ∼ 10−15 F, exhibit voltage changes
V =
Measurement of the capacitance–voltage char-
e/C ∼ 0.016 mV for individual electron tunnel-
acteristic is one of the most widespread methods
ing. This is called a Coulomb blocade, and it
of investigating semiconductor electron–hole tran-
appears on a current versus voltage characteristic
sit, heterojunctions, Schottky diodes and metal–
curve (I versus V , or dI versus dV plot) as a se-
insulator–semiconductor structures. From the
quence of steps called a Coulomb staircase.
shape of a C–V characteristic it is possible to de-
CAPACITANCE, BARRIER termine the distribution of doping impurity in the
See Barrier capacitance. structure. In particular, in sharp p–n junctions and
Schottky diodes the capacitance–voltage charac-
CAPACITANCE SPECTROSCOPY teristic plotted as C −2 versus V is a straight line
A collection of methods for determining the with a slope characterizing the doping of the semi-
energy spectrum of deep levels in semiconduc- conductor, and the cutoff voltage is equal to the
tors from changes in their barrier capacitance, C, equilibrium height of the potential barrier at the
during variation of the applied voltage. A simple junction (contact potential difference). The charac-
version of capacitance spectroscopy is based on teristic shape of the C–V characteristic for M–I–S
measuring steady-state capacitance–voltage char- structures is shown in the figure.
acteristics. When a reverse bias (i.e. a bias in the The presence of deep levels within the bulk or
cut-off direction) is applied to a system contain- at the surface of the sample, which recharge dur-
ing a semiconductor junction or a Schottky barrier, ing voltage variations, causes additional features
the thickness L of the semiconductor region with of the capacitance–voltage characteristic. Besides,
depleted current carriers increases, and the ca- the shape of the C–V characteristic has a strong
pacitance of the depletion layer decreases. More dependence on the frequency f of the probing
information can be obtained by methods of non- signal, because at high frequencies there is not
static capacitance spectroscopy of deep levels. It enough time to recharge the deep levels, and to
CARBIDE TRANSFORMATIONS 151
ature of cementite (Fe3 C) formation increases. of the density of the electron clouds and the con-
Redistribution of carbon and alloying elements centration of electron density in certain directions,
between phases leads to alloying of iron carbide. such as along chemical bonds (see Hybridization
With a certain temperature of tempering, degree of orbitals). Depending on the way atomic orbitals
of steel alloying, and ratio of carbide-forming el- are combined, sp-, sp2 - and sp3 -hybridizations
ement content to that of carbon, the formation of are distinguished. Each hybridization corresponds
special carbide is possible. Destabilization of ce- to a particular spatial arrangement of a bond. Ox-
mentite (or special carbide) occurs at complete idation states of carbon in its compounds are −4,
replacement of one atom of iron in cementite (or −2, +2, and +4; electronegativity value is ≈2.58;
the basic element in special carbide) with the al- atomic radius 0.077 nm, ionic radii: 0.016 nm for
loying element. In multicomponent alloying such C4+ and 0.26 nm for C4− .
replacement occurs with several alloying elements There exist five allotropic (see Allotropy) mod-
simultaneously. Carbide transformations that run ifications of solid carbon: diamond (sp3 -hybrid-
according to the mechanism of alloy aging and di- ization of orbitals), graphite (sp2 -hybridization),
rect precipitation of carbides from ferrite, bypass- fullerene (C60 and analogues, sp2 -hybridization),
ing cementite, are used to harden structural, stamp, carbyne (sp-hybridization), and amorphous car-
tool, high-speed and certain austenite steels, as bon (mixed type of bonding). Carbyne is a quasi-
well as to maintain hardening in long-term oper- one-dimensional semiconductor, the width of the
ation of boiler and turbine, motor equipment, of band gap of which is ∼1 eV. Amorphous carbon is
industrial holders and fastening parts, etc., at in- a polyphase structure, which consists of the finest
creased temperatures. inclusions of polymorphic modifications of car-
The specifics of redistribution of carbon and bon. Depending on the mode of preparation, car-
alloying elements between phases in multicompo- bon may also contain chemically bound hydrogen
nent alloying causes the carbide transformations atoms, which saturate the dangling bonds of the
to run at a lower number of alloying elements carbon structure. The principal carbon-containing
than in single or simple alloying (and the same to- minerals are salts (carbonates) of carbonic acid
tal degree of alloying). Here, the ferrite contains H2 CO3 . Carbon atoms are the chief constituents
more alloying elements, which permits a more of coal, oil, peat and other materials. Carbon is a
complete use of their effect upon physical me- nonmetal and behaves as an active reducing agent
chanical, technological and service properties of on heating. It forms carbides with various ele-
steel. The composition and type of the carbide ments, e.g., boron carbide, silicon carbide SiC,
crystal lattice, distribution of carbides within the cementite Fe3 C. Carbon is used as a reducing
structure, and the mechanism of their formation agent, as a component of various alloys, as fuel,
are investigated using chemical phase analysis, for production of electrodes, etc.
X-ray structure analysis, microscopic diffraction
and magnetic analyses, Mössbauer spectroscopy, CARBON, ACTIVE
internal friction, and electron microscopy with the See Active carbons.
help of replicas and foils.
CARBON FILM, DIAMOND-LIKE
CARBON, C See Diamond-like carbon lms.
The chemical element of Group IV of the pe-
riodic table with atomic number 6 and atomic CARBYNE (fr. Lat. carbon , coal)
weight 12.011. The mass numbers of the two sta- The third allotropic modification of carbon,
ble carbon isotopes are 12 (12 C, 98.892%) and 13 which consists of rectilinear carbon chains bound
(13 C, 1.108%). The 14 C isotope is unstable (half- into crystals through van der Waals forces. The
life 5730 years). The wide diversity of properties rectilinear shape of the chains is due to sp-hybrid
of carbon compounds is due to the electronic con- bonding of the carbon atoms. In actual samples the
figuration of the carbon atom: 1s 2 2s 2 2p2 . Inter- individual chains are cross-linked, with perhaps
action with other atoms results in rearrangement some periodicity, and this involves a transition of
CASTING MATERIALS 153
some carbon atoms into sp2 - or sp3 -hybrid type to produce various parts, from time-piece compo-
bonding. The presence of carbyne can be detected nents to heavy-duty constructions for engineering,
by the methods of electron diffraction and X-ray dif- metallurgy, power and other branches of machine
fraction. building. Steels and alloys are cast into molds
Crystalline particles of carbyne are colorless, (plate casts, forging casts) or poured continuously
and no large monocrystals have been prepared. into water-cooled crystallizers (for a continuously
Carbyne is hard to obtain in a pure form (free from operating casting installation). More than 50 types
admixture of laminated modifications of carbon). of casting techniques are available. Traditional
Density and other physical properties depend on technologies include pouring molten metal from
the method of manufacture and purification. Car- dippers or crucibles driven by gravity into sand
byne has a higher speci c heat than graphite and casting forms. Special casting techniques make
diamond, and it surpasses graphite in hardness. it possible to improve the quality of casts, per-
Heat of combustion is 84 to 372 kJ/mole, which form precision casting in both size and shape, and
is significantly lower than that of graphite and achieve special physical and mechanical charac-
diamond, which attests to the thermodynamic sta- teristics. Such technologies involve various exter-
bility of carbyne. nal factors which act upon the crystallizing metal
There are 3 bands in the Raman spectrum of and provide for using special casting forms, mod-
carbyne at 1360, 1580 and 2140 cm−1 , the last of els or casting techniques. Among the latter are
them being the characteristic one related to the vi- casting under high or low pressure, centrifugal
brations of conjugate C–C bonds. Carbyne is an casting, electric slag casting, continuous or sus-
n-type semiconductor with a band gap of 0.64 eV; pension casting, casting into melted-out models,
electrical resistivity is 106 ·cm (at 293 K). Car- casting into chill molds, ceramic and shell forms,
byne was originally obtained and characterized etc. Art casting is done using melted-out models
by the methods of oxidative polycondensation of to produce monumental sculptures, architectural
acetylene. It forms upon condensation from vapor- decorations, and jewelry.
gas phase, plasma, an ion-molecular beam under
the conditions that are softer compared to those CASTING MATERIALS
for forming diamond, and of laminated polymers Macroscopically inhomogeneous substances
of carbon. Natural carbyne was discovered in 1968 obtained by fusion of two and more mutually solu-
in a meteorite crater and received the name chaoite ble metals, nonmetals, or organic substances, and
(white veins and disseminations in graphite). In by subsequent solidi ca tion of the liquid phase in
its chemical inertness it surpasses graphite and ap- a mold of a particular shape. Depending on the
proaches diamond. Defect-free thread-like crystals temperature range of metastability of the liquid
of carbyne are believed to have very high rupture phase, the structure of casting materials may be
strength. crystalline or amorphous. In a general case, the
conditions of solidification in the amorphous or
CASCADE, COLLISION crystalline states may be controlled by the cool-
See Collision cascade. ing rate, by the purity of the materials, by their
solubility, and by their crystallization activity.
CASTING The properties of casting materials crystalliz-
The process of forming pieces (casts) by filling ing during the course of solidification are not
a hollow casting form or mold with melted ma- only determined by their chemical composition,
terial (metals, their alloys, plastics, certain rocks, by the method of fusion, thermal and other types
metallurgy slags). The principal advantages of of deformationless treatment, but are also depen-
casting include the possibility of obtaining casts dent on the thermal conditions of solidification,
of a complicated configuration that is as close in on the nucleation rates of centers, and the linear
size and shape to the final product as possible. growth rates of crystallization nuclei. These factors
Mold casting of steel, cast iron, nonferrous alloys determine the degree of development of physi-
(from 5–10 to 100 g and more in mass) is used cal, chemical and structural inhomogeneities in
154 CAST IRON
casting materials. The processes of their solidifi- monoclinic system and triclinic system. Depending
cation and crystallization, and hence their proper- on the category the nature of the anisotropy, and
ties, may be significantly changed by modi cation, many physical properties of a solid change char-
electromagnetic mixing and other methods (see acteristically (see Anisotropy of crystals).
Casting). This classification term “category” is not com-
monly used in the Western literature.
CAST IRON
Alloy of iron with carbon (usually 2 to 4% C),
also containing (Si, P and S, Mn), and sometimes CATHODE
additional alloying elements (Al, V, Cr, Ni, etc.); See Cold cathode, Field-emission cathode,
it solidifies with the formation of eutectics. Unlike Thermionic cathode, Photocathode.
steel due to its low plasticity, cast iron is machin-
able under pressure. It is characterized by a high CATHODE SPUTTERING
liquid yield. Cast irons are subdivided into the fol- See Sputtering.
lowing classes: pearlite, ferrite, perlite-ferrite, sor-
bite, troostite, bainite, martensite, troostite-ferrite,
CATHODOLUMINESCENCE
and austenite. According to its microstructure the
following types are distinguished: gray cast iron A type of radioluminescence excited by an
(in the form of cementite or iron carbide Fe3 C), electron beam (“cathode rays”). Depending on the
and forgeable cast iron obtained by annealing electron energy E, one distinguishes low-voltage
white cast iron (flocculent graphite). The proper- (E = 10–100 eV) and high-voltage (E > 0.5 keV)
ties of cast iron are improved by alloying, modi - cathodoluminescence. At high-voltage the elec-
cation, thermal treatment. Cast iron is the primary trons overcome the surface barrier, penetrate to a
product obtained from the processing of iron ore depth proportional to E 1.4 , knock out secondary
materials. It is used in chemical, petroleum, and electrons that also ionize atoms, create electron–
agricultural machine building, in tool, car and trac- hole pairs, or directly excite the ions of the ac-
tor fabrication, in civil engineering, etc. tivator (Mn2+ , Tb3+ and others). The minority
carriers migrate across the lattice and are trapped
CATALYSIS, RADIATION at luminescent centers, accompanied by radiative
See Radiation catalysis. recombination. At low voltage the excitation by
primary electrons only covers several near-surface
CATASTROPHE, POLARIZATION atomic layers, and the diffusion of nonequilibrium
See Polarization catastrophe. carriers produces a zone of fluorescence spreading
into the depth of the sample. The forming negative
CATEGORY (fr. Gr. κατ ηγ oρια , accusation)
charge is neutralized at high voltage luminescence
A level of crystallographic classification by the
mainly through the emission of electrons from the
number and order of special (or individual) di-
rections, i.e. directions in the crystal lattice which surface of a luminophor, while at low voltage the
do not repeat under a symmetry operation (see neutralization is due to the conductivity of the lat-
Crystal symmetry). There are three categories: the ter. In its spectrum, cathodoluminescence is simi-
highest one (no special direction and some sym- lar to other types of luminescence. Its brightness
metry axes of order higher than 2); the medium in a certain spectral region is proportional to E
type (one special direction along a unique symme- and the current density in the beam. The energy
try axis of order 3, 4, or 6); the lowest one (no yield is 10–30% for high voltage and 0.5–2% for
special direction and no symmetry axes of order low voltage. Cathodoluminescence is employed in
higher than 2). The highest category involves crys- picture tube screens and in low-voltage indicators
tals of the cubic system, the medium one includes that are used in devices for the visualization of in-
those of the hexagonal system, trigonal system formation, TV sets and other electronic devices.
and tetragonal system, and the lowest category en- Cathodoluminescence underlies the operation of
compasses structures of the orthorhombic system, lasers excited by an electron beam.
CATLOW–DILLER–NORGETT POTENTIAL 155
CATLOW–DILLER–NORGETT POTENTIAL
(C.R. Catlow, K.M. Diller, M.J. Norgett, 1977)
A central pair-potential that describes short-
range ion-to-ion interactions in ionic crystals using
the shell model. An analysis of the relation be-
tween the equilibrium interionic distance in a
crystal and its elastic constants on the one hand,
and the interionic interaction potentials yielded by
this model for alkali-halide crystals (such as LiCl,
NaCl, KCl, RbCl) on the other hand, demonstrated
that the ion-to-ion interaction is attractive at long
distances r and repulsive at short distances. Start-
ing with that observation, Catlow, Diller and Nor-
gett offered the following two proposals for the
short-range potential of (identical) second nearest
neighbors in alkali halides. Schematic represention of the potential described in pro-
I. The Buckingham potential V (r) = posal II for the four regions: (1) r > ra , (2) ra > r > rm ,
A exp(−r/ρ) − Cr −6 contains an exponential (3) rm > r > rb , and (4) rb > r.
repulsive term with parameters A, ρ deter-
mined using an electron gas model, and an first derivatives. The parameters A and ρ
attractive van der Waals potential with a pa- coincide, while C considerably exceeds the
rameter C (see Van der Waals forces). For value expected from proposal I. For example,
example, the F− –F− ionic interaction in a LiF C = 26.8 eV, while ra = 0.3457 nm, rb =
crystal is described by A = 1128 eV, ρ = 0.2 nm, rm = 0.2833 nm for the F− –F− inter-
0.02753 nm, and C = 12.71 eV·nm5 . The action. These parameters are self-consistently
parameters for the anion–anion (C−− ) and adjusted over the entire alkali halide series.
cation–cation (C++ ) interactions are arbitrar-
ily assumed to be equal to each other, C−− = Having fixed the potential parameters for sec-
C++ = C/2. ond nearest neighbors, one may update the respec-
II. A combined (“spline”) potential consists of ex- tive parameters for first neighbors, that is for adja-
pressions used in proposal I for attraction at cent ions, that potential being the common Born–
long distances and repulsion at short ones, plus Mayer potential, V (r) = A exp(−r/ρ). Next the
the third and fifth power polynomials for inter- parameters of the ions are determined using the
mediate distances (see Fig.): shell model. The shell charge and the nucleus–
shell bonding constant are found by comparing
V (r) = −Cr −6 , r > ra , (1) the static and the high frequency dielectric con-
stants and the transverse optical frequencies with
3
experimental data for all the crystals, so that the
V (r) = ai r i , ra > r > rm , (2)
polarization of a crystal may be adequately de-
i=0
scribed by the result.
5 The Catlow–Diller–Norgett potential, as pre-
V (r) = bi r i , rm > r > rb , (3) sented in II above, has a number of advantages
i=0 over other potentials used for ionic crystals. It cov-
ers every individual ion and is easily applied to
r
V (r) = A exp − , rb > r. (4) modeling alkali halides. It offers a consistent pic-
ρ
ture of the ion-to-ion interaction for all the alkali
Matching the functions at points ra , rb , and halides. In combination with the Coulomb poten-
at the minimum point rm is carried out by in- tial it adequately describes the properties not only
voking continuity of the functions and their of an ideal crystal, but of a real one as well. In the
156 CAUCHY RELATIONS
latter the ion-to-ion distances may differ consid- volume of a liquid. Ultrasound is widely applied
erably from the equilibrium ones in the vicinity of in some technological processes where cavitation
defects. Thus it becomes possible to study both in- plays an important role, and this has prompted
trinsic crystalline and impurity defects in various the search for materials with high cavitation re-
ionic crystals (see Interatomic interaction poten- sistance. Its measure is the loss of mass per unit
tials). area during a preset period of time. The cavita-
tion intensity is evaluated by the destruction of a
CAUCHY RELATIONS (A. Cauchy) thin aluminum foil placed in the cavitation area.
Additional relations between the 36 crystal There are also other methods to determine cavita-
elastic moduli cik following from the symme- tion resistance in investigations of hydrodynamic
tries of interatomic interactions, where i, k = cavitation, such as rotational, jet-impact, and flow-
1, 2, . . . , 6. If the interatomic forces are pairwise through ones.
and central then each term is at a center of
symmetry, the undeformed crystal is stress free, CAYLEY TREE (A. Cayley, 1889); also called a
and the harmonic approximation is valid. Then in Bethe lattice
the general case there are 21 independent mod- A branched structure in which each site has f
ulus constants cik (in Voigt notation) by virtue nearest neighbors, and no two branches are per-
of the symmetry condition cik = cki . The six mitted to cross each other. It is cycle-free, and also
Cauchy relations c12 = c66 , c13 = c55 , c23 = c44 , self similar since each branch is itself an entire
c14 = c56 , c25 = c46 , and c36 = c45 reduce the tree. Such graphs are used for percolation stud-
number of independent elastic moduli to 15. For ies, in cluster techniques to provide a convenient
cubic crystals there are three independent elastic boundary condition for a cluster, to calculate the
number of isomers of saturated hydrocarbons, and
constants (c11 , c12 , c44 ), and one Cauchy re-
to describe disordered solids.
lation (c12 = c44 ). These results hold well for
some alkali-halide crystals, but not for most other
CELLULAR STRUCTURE
cubic crystals. In metallic crystals the Cauchy re-
Regions 1 μm in size which are free of disloca-
lations usually do not hold (e.g., in noble metals
tions, or which contain individual dislocations,
c12 /c44 ∼ 2–4). In elastically isotropic crystals dislocation dipoles, multipoles, interlacements,
obeying the Cauchy relations c11 = 3c44 , and networks, etc., and are separated by dislocation in-
the Poisson ratio is ν = 1/4. For orthorhombic terlacements, balls, or rods. The cell boundaries
crystals there are 9 independent constants and 3 can be either dense or spread out, either narrow
Cauchy relations, corresponding to 6 independent or wide, and so on. The cells can be equiaxial
moduli when the Cauchy relations hold. or nonequiaxial. Sometimes the cellular structure
is called block-cellular structure (see Block struc-
CAVITATION RESISTANCE ture).
The ability of a material to resist the effect of
cavitation (formation of vapor bubbles) that ap- CEMENTATION
pears in a liquid at the passage of a powerful sonic Coating a metal, usually iron or steel, by heat-
or ultrasonic wave (acoustic cavitation), and also ing it with another substance so that the added
during high-speed flow past bodies (hydrodynamic material, usually a powder, diffuses into the sur-
cavitation). The collapse of a cavitation bubble at face; it is a type of chemical heat treatment. The
the compression phase produces a local hydraulic coating material might be a metal like aluminum or
shock which affects the surface of a tool, an ar- zinc, or perhaps carbon. Diffusion saturation of a
ticle, or a part. Cavitation erosion is especially steel product surface by carbon typically produces
strong for brittle materials such as rocks, concrete, a saturated layer that is 0.5 to 2 mm thick, with
and ceramics. Metals possess considerably higher a carbon content of 0.8–1.1%. Cementation with
cavitation resistance that reduces noticeably in subsequent quenching and low annealing provides
hostile media and suspensions with abrasive par- articles with a high surface hardness, wear resis-
ticles, and upon increasing the pressure within the tance and increased fatigue strength, leaving their
CENTRAL FORCES 157
core tough. Cementation may be carried out in moves
as if the resultant sum of all applied forces
gaseous, liquid, and solid media containing car- R = i F i were acting at the center
of mass,
bon. It is used in the production of tools. and the total mass of the body M = i mi were
concentrated there. This definition implies that the
CEMENTITE (fr. Lat. caementum , broken stone) center of mass is, in a uniform gravitational field,
A structural component of steel and cast iron the center of gravity, as well as the point where the
containing the elements iron and carbon. Cemen- resultant force of gravity Mg is acting.
tite is the most stable iron carbide (Fe3 C), con-
taining 6.7 wt.% C (see Carbide transformations). CENTER OF PERCUSSION
It is formed upon the cooling of iron–carbon alloys The point of a body about which it starts to ro-
(see Iron alloys) if the conditions are favorable for tate when an impulse (impulsive force) is applied
crystallization to occur in the metastable phase di- at another point in the body.
agram of these alloys (iron–cementite). Cementite
CENTER OF SYMMETRY, center of inversion
is often a component of pearlite, sorbite, troost-
A special point of a figure possessing the prop-
ite and ledeburite. The crystal lattice of cementite
erty that during the inversion operation (r → −r)
is orthorhombic; the unit cell contains 12 atoms
through that point as origin of coordinates, all
of Fe and 4 atoms of C. Lattice constants are:
the elements (points) of the figure transform into
a = 0.4514 nm, b = 0.5079 nm, c = 0.6730 nm.
themselves (r is the coordinate of an arbitrary
According to some indications cementite is close
point of the figure). The presence of a center of
to an interstitial phase (see Interstitial alloys). It
symmetry is equivalent to the presence of an inver-
forms substitutional solid solutions: carbon may be sion axis of the first order. Many properties of the
replaced by nitrogen or oxygen, iron by mangan- solid state depend on the presence or absence of a
ese, chromium or tungsten (alloyed cementite). center of symmetry; see, e.g., Electric e ld effects
Their dispersity, the form of cementite crystals, in radio-frequency spectroscopy.
and their crystal structure associated with a ferrite
environment (see Ferrites), to a large extent de- CENTI. . . (fr. Lat. centum , a hundred)
termine the strength and plasticity of iron–carbon Prefix to the name of a physical value to form
alloys after a heat treatment. Cementite is hard and a fractional unit equal to 1/100 of the source
brittle, and there is practically no elongation if it unit. Abbreviation symbol is c. For instance, 1 cm
fails during the course of tensile tests. The melting (centimeter) = 0.01 m.
temperature is ∼1873 K; at this high temperature
it decomposes into austenite and graphite. It is fer- CENTRAL FORCES
romagnetic with the Curie point of 486 K. In solids these are forces of interaction be-
tween pairs of atoms (ions) acting along the
CENTER straight line connecting them. In such a case these
See Active centers, Amphoteric center, Color atoms are considered to be point masses. Formally
center, Dilatation center, Luminescent centers, the concept “central forces” is based on “rigid”
Paramagnetic center, Positron center, Positronium point atoms (ions). When interatomic interactions
center. in solids are described exclusively by means of
central forces, this constitutes a coarse and phe-
CENTER OF INVERSION nomenological approximation. The central force
See Center of symmetry. is a special case of a pairwise interaction between
atoms, when its adiabatic potential U has the form
CENTER OF MASS, center of inertia 1
A point in a solid body relative to which the to- U= ϕ |R i − R j | ,
2
ij
tal first moment of
the all the components of the
mass vanish, i.e. i mi r i = 0, where r i is the where R i is the radius-vector of the lth atom, and
radius-vector of the ith mass relative to the cen- ϕ is a function which depends on the type of in-
ter of mass. In the absence of torques a solid body teraction. For a gravitational field involving two
158 CENTRAL PEAK
masses ϕ = −Gm1 m2 /|R 1 − R 2 |, corresponding dering point. This central peak is due to the contri-
to an inverse square central force law, where G is bution of nonlinear spin excitations, i.e. magnetic
the gravitational constant. solitons.
to diamond-like crystals with predominantly co- components, but to a substantial degree, by the
valent bonds (carbides, silicon nitride and boron technology of production (Fig. 1). Also widely
nitride), and these substances all have similar employed are the methods of self-propagating
structures constructed on the principle of diversi- high-temperature synthesis, chemical sedimenta-
fied sequential packing of identical layers. These tion, infiltration from the gas phase, etc. The
compounds are characterized by the existence of a stability of the crystal lattice and the low mobil-
substantial number of polytypes (see Polytypism); ity of defects in high-melting compounds results in
the latter have also been found in diamond. low rates of the diffusive and viscous processes re-
Besides covalent and ionic crystals, there are sponsible for mass transport, and for settling at the
many ceramic material components from a large solid-phase sintering. For that reason, additives
group of metal-like high-melting compounds, such to activate the sintering are introduced into the
as interstitial solutions that form easily in transi- batch charge, and this is then subjected to explo-
tion metals by light elements with relatively small sion processing to augment the number of defects,
atomic radii. Most of these phases are compounds or there could be a physical and chemical modi -
of variable proportion having wide ranges of ho- cation of the surface, etc. The activating additives,
mogeneity, and a great variety of structure types usually oxides, secure the liquid-phase sintering
(e.g., CaFe, ZnS, CaB6 , etc.). regime, which not only facilitates the rearrange-
Conventionally the properties of a ceramic ment of component particles, but also results in
are defined not only by the properties of its their recrystallization via the liquid phase. This ac-
160 CERAMICS
An ideal crystal, 50 –
theoretical strength
Whisker thread-like 15 Screw dislocations along
crystals the axis of growth, and
surface defects
Single crystals 7 Surface and bulk defects
of crystal lattice
Hot-pressed 0.8–1.0 Same as above, plus grain
ceramic boundaries and additional
phases
Sintered technical 0.2–0.3 Same as above,
ceramic plus pores
Thermal neutron cross-section is 0.63 barn. De- Cesium is used in the production of photocells
bye temperature is 139 K. At room temperature (it exceeds all the other metals in light sensitiv-
the elastic modulus is 29.99 GPa, shear modulus ity), of gas absorbers, and of low work function
is 11.99 GPa, Poisson ratio is 0.248; ultimate ten- coatings for emitters. The application of “cesium
sile strength of cast sample is 0.1206 GPa, relative plasma” is prospective for ionic reactor motors.
elongation is 32.6%. Pure cerium is a forgeable
and tough metal. At room temperature without CHAIN REACTIONS
preheating it is possible to manufacture cerium Reactions of active particles (free radicals in
sheets by forging, and wire by pressing. It is used chemical reactions, neutrons n in nuclear reac-
as an alloying addition to the light metals, and also tions), when the elementary interaction produces
as a getter. one or a number of additional active particles (rad-
icals, A·). In the first case the reaction is called
CERMET (fr. ceramic metal) an ordinary, non-branched chain reaction, in the
A composite material consisting of ceramic second case it is a branched one. The chlori-
grains dispersed in a metallic matrix. It is formed nation process of H2 : Cl· + H2 → HCl + H·,
by techniques of powder metallurgy. Cl2 + H· → HCl + Cl·, . . . , etc., is an example
of a non-branched chain reactions. The process of
CESIUM, Cs H2 oxidation represents an example of a branched
Chemical element of Group I of the periodic chain reaction: H· + O2 → ·OH + O:, O: + H2 →
table with atomic number 55 and atomic mass ·OH + H·, ·OH + H2 → H2 O + H·, . . . , etc. In
132.9054. There is one stable isotope 133 Cs, and such a manner, the primary species A· may give
22 radioactive isotopes, among which 137 Cs with rise to a large number (from ten to 107 ) of trans-
a half-life 27 years has found the broadest ap- formations of inactive particles (e.g., molecules).
The number of such transformations per primary
plications. Outer shell electronic configuration is
active species A· is called the chain length. The
6s 1 . Ionization energies are 3.893, 25.1, 34.6 eV.
breaking of a chain (chain reaction termination)
Atomic radius is 0.262; radius of Cs+ ion is
may be due to the presence of an impurity which
0.167 nm. Oxidation state is +1. Electronegativ-
interacts with A· so an additional A· is not pro-
ity is 0.68.
duced. Also it may be due to the departure of A·
Cesium is a silvery-white metal. It has a body-
from the reaction region.
centered cubic lattice with a = 0.605 nm (at
In the simplest situation (homogeneous case)
448 K). Density is 1.9039 g/cm3 (at 273 K) and
the chain reaction rate is determined by the equa-
1.880 g/cm3 (at 300 K); Tmelting = 301.75 K;
tion:
Tboiling ≈ 960 K. Heat of melting is 15.8172
kJ/kg, heat of evaporation is 613.2 kJ/kg; specific v = K1 C1 P0
heat at constant pressure is 32.2 J·mole−1 ·K−1 t
at 298 K; thermal conductivity ratio is 16.7 to
× exp (α − 1)K1 C1 − K2 C2 dt − βt ,
27.2 W·m−1 ·K−1 at 301.6 K; linear expansion
coefficient is 9.7·10−5 K−1 (273 to 300 K). Elec- 0
trical resistivity is 0.183 μ·m (at 273 K) and where P0 is the concentration of the primary
0.2125 μ·m (at 300 K). Mohs hardness is 0.2; species A· created by the initiating chain reaction,
Brinell hardness is 0.015. Normal elastic mod- K1 and K2 are the velocity constants of chain re-
ulus is 1.716 GPa at room temperature. Debye actions and the chain breaking reaction; C1 and
temperature is 42 K. Ion-plasma frequency is C2 are the concentrations of initial substances of
1.724 THz; linear low-temperature electronic spe- the chain reactions, and of the chain breaking cen-
cific heat coefficient is ≈3.20 mJ·mole−1 ·K−2 . ters; α is the number of A· species arising in the
Electronic work function of cesium polycrystal is elementary act of chain initiation; β is a constant
≈1.85 eV. Cross-section of thermal neutron trap- related to the probability of A· leaving the reac-
ping is 29 barn. Cesium is paramagnetic with tion region. The exponential term in this equation
magnetic susceptibility 0.1·10−6 (at 291 K). reflects the time change of the radical concentra-
CHALCOGENIDE MATERIALS 163
tion; and the relations between the terms in the in nuclear power plants for heating cities, produc-
exponent index determine the kinetic character of ing electricity, etc. and the number of neutrons
the chain reactions. So, if the first term in the ex- generated is greater than one (i.e. k > 1). The
ponent index turns out to be larger than the sum cross-section of the fission reaction of U-235 in-
of the others, then the chain reactions velocity in- creases with the decrease of the neutron velocity
creases exponentially with respect to time, and a v as 1/v, and there are resonances in the range of
chain explosion is possible. neutron energy from a fraction of an electronvolt
In solids chain reactions have been found in to perhaps a hundred electronvolts. As a result, the
organic matrices. The matrix itself and also the so-called “thermal neutrons” play the main role in
components diluted in it can participate in these fission reactions. At these low energies the cross-
reactions. For example, in organic polymeric ma- section for the U-235 fission reaction is much
trices which contain halide complexes of Au, larger than the cross-sections for other reactions
Pt, Cu, Bi chain reactions can be initiated by with neutrons. Since there is a large probability
the action of light or ionizing radiation on the that many neutrons will escape from a very small
system. As a result, atomic Cl· appears (e.g., sample without reacting, there is a critical mass
[AuCl4 ]− +Cl· → [AuCl2 ]− +Cl2 +Cl·), and the for the chain reaction to take place, and this is
chain length in such systems may reach 104 . The approximately 50 kg for U-235 and 10 kg for
use of chain reactions for the photographic record- Pu-239. If the average neutron multiplication fac-
ing of information provides new opportunities for tor k is too large, the number of neutrons in
the creation of sensitive media with high resolu- the specimen can grow exponentially, leading to
tion. an explosion. To prevent this, neutron absorbing
Free radicals giving rise to chain reactions rods (e.g., graphite) are inserted into the reactor to
may be generated at the heterogeneous burning of remove excess neutrons and keep k ∼ 1 so the
solids and its mixtures in the reaction zone. In such chain reaction will proceed slowly, and generate
a case the very existence these radicals determines heat gradually.
the burning kinetics. There is some evidence that
CHALCOGENIDE MATERIALS (fr. Gr. χαλκoς ,
at burning, chain reactions also occur in the con-
copper and γ ενναω, engender)
densed phase.
Compounds of elements of Group VI of the
Nuclear chain reactions take place in uni-
periodic table (O, S, Se, Te, Po, collectively
form solid specimens of U-235 or Pu-239 (nu- called chalcogens) with other elements. The class
clear bomb), and also in the case of a non- of chalcogenides (oxides, sulfides, selenides, tel-
uniform distribution of U-235 in some other ma- lurides) contains the X−2 anion of Group VI
terial (graphite, heavy or light water, or of course elements. In view of their differences in prop-
U-239) which form the nuclear reactor. The first erties and fields of application, oxides are often
nuclear chain reaction was realized by E. Fermi unfairly excluded from the list of chalcogenide
in 1942 in a uranium-graphite reactor. Up to the materials, and polonium compounds are rarely in-
present time only uranium nuclear reactors are cluded because they are strongly radioactive. In
in use, and they satisfy the following conditions. modern areas of engineering chalcogenide mate-
The elementary act of fission involves a neutron rials are often used in the form of thin lms or
n striking a U-235 nucleus in accordance with the monocrystals, which are obtained through pre-
reaction cipitation of aqueous solutions, decomposition of
U-235 + n → A + B + kn + Q, chelate compounds on a substrate, crystallization
from gaseous (vapor) phase, etc. Typical ionic lat-
where A and B are nuclear fragments (daughter tices of chalcogenides are characteristic of these
nuclei), k is the neutron multiplication factor, and elements, which exhibit very low values of ioniza-
the chain reaction proceeds because the average tion potentials. Many chalcogenide materials form
value of k ∼ 2.5 for the elementary act of fission. lattices of the chalcopyrite type, the spinel type,
The kn secondary neutrons in turn induce further and others. They are often nonstoichiometric com-
fissions The energy Q of the reaction is harnessed pounds. Chalcogenides comprise a broad class of
164 CHANNELED PARTICLE FRICTION
simulation that incident bombarding ions which between the trapping of the ion nucleus by the
have abnormally long paths through a monocrys- crystal atom nucleus and the emission of the reac-
tal when they enter the crystal almost parallel to tion product is long enough for the activated atom
a main crystallographic direction. The channel- to displace noticeably from the blocking chain
ing of relativistic charged particles gives rise to (plane). This allows one to estimate experimen-
an intense electromagnetic quasi-characteristic ra- tally the nuclear reaction time. One can observe
diation in the gamma ray and X-ray bands. See the blocking by the shadow effect resulting from
also Hyperchanneling, Molecule channeling, Neu- the embedding of a radioactive isotope into a
tral particle channeling, Quantum states of chan- monocrystal. Fig. 2 shows the angular distribution
neled particles, Okorokov effect, Channeled parti- of the intensity of emitted α-particles in a tung-
cle self-acceleration, Surface channeling, Surface sten crystal upon the introduction of the 222 Rn
conductive channel, Current lamentation. isotope; the minimum corresponds to the 111 di-
rection. Heating the crystal results in washing out
CHANNELING OF ION BEAMS the shadow due to thermal displacements of the
This takes place along the directions of atomic atoms.
chains and planes in a monocrystal because the
ions scattered at small angles to these directions
CHARACTERISTIC ENERGY LOSS
cannot penetrate deep into the crystal (see Fig. 1).
Discrete energy losses, which are independent
Shadow traces of the atomic chains and planes
of the energy of incident primary electrons, and
appear (the so-called shadow effect) at the tar-
are characteristic of a given solid; they are ob-
get detecting the scattered ions. The emission of
served in reflection and transmission spectra of
products of a nuclear reaction between the inci-
secondary electrons. Characteristic energy loss is
dent ions and the crystal atoms also produces the
shadow effect. The shadow disappears, if the time energy expended on the excitation of (i) electron
transitions from inner core levels or from va-
lence band levels to unoccupied levels, vibrational
energy levels, and (ii) collective longitudinal os-
cillations of a solid-state plasma (plasmons). The
values of characteristic energy losses of the first
kind (30–105 eV) are easily determined from the
binding energies (Eb ) of electrons in atoms, and
are used in ionization spectroscopy for the analysis
of solids. Different solids exhibit different binding
energies because of many-particle effects (relax-
Fig. 1.
ation, multiplet splitting, con guration interaction,
shake-off and shake-up, chemical shifts). Charac-
teristic energy losses of the second kind (1–30 eV)
are more difficult to interpret, since they are due to
processes involving levels of the valence band and
upper unoccupied levels. Vibrational characteris-
tic energy losses are related to transitions between
the vibrational states of surface atoms (molecules)
of a solid, or of adsorbed layers; these characteris-
tic energy losses are a source of information on
phonon involvements. Volume and surface plas-
mon characteristic energy losses (ωP and ωS ,
respectively) are determined by the concentration
of valence electrons
√ n: ωP = (4πe2 n/m)1/2 ,
Fig. 2. ωS = ωp / 2, and they manifest themselves
166 CHARACTERISTIC ENERGY LOSS SPECTROSCOPY
t = T /Tc . In a type II superconductor, the magni- and τQ = [4T /(πΔ)]1/2 τε is the period of the
tude of ξ is related to the value of the upper critical imbalance relaxation. In the neighborhood of Tc ,
e ld Bc2 through the expression Bc2 = Φ0 /(2πξ 2 ) λE varies as λE (T ) ≈ λE (0)/(1 − t)1/4 . The
(Φ0 = h/(2e) is the ux quantum). value of λE determines the characteristic dimen-
The Meissner effect in superconductors arises sions of the region of inhomogeneity (so-called
from the fact that external magnetic field cre- phase slip centers) in the resistive state of super-
ates a current (Meissner current) in the surface conducting films.
layer which screens the bulk of the superconductor The Josephson penetration depth has the value
from the penetration of the field. The characteris- λJ = [Φ0 /(2πμ0 Jc deff )]1/2 where Jc is the criti-
tic thickness of this layer is called the penetration cal current and deff is the effective thickness of the
depth λ. In the case of a pure type I superconduc- junction. A Josephson junction is considered short
tor, the London penetration depth is determined when its length is less than λJ , and it is long when
from the London formula λ0 = [mc2 /(4πne2 )]1/2 it has a greater length.
and is of the order of 10−7 m. In the case of
a superconducting alloy, λ > λ0 , with the value CHARACTERISTIC SURFACE, ellipsoid
λ ≈ λ0 (ξ0 / ll )1/2 at ξ0
ll . In the neighborhood A second-order surface, which represents the
of Tc , λ has the temperature dependence λ(T ) ≈ orientation dependence of a particular physical
λ(0)/(1 − t)1/2 . The penetration depth λ may be property of a crystal, which is described by a
determined experimentally from the change of the second-rank tensor.
superconductor inductance, which is due to partial The equation of the characteristic surface,
penetration of the field into the bulk of the solid. when reduced to the principal axes of the crys-
The imbalance relaxation length λE deter- tallophysical coordinate system (e.g., an orthog-
mines the characteristic distances of nonequi- onal crystal coordinate system compatible with
librium processes in superconductors. The sim- nonorthogonal lattice constant directions) is of the
plest case is the current flow through a galvanic form
contact between a normal metal and a super-
s1 x12 + s2 x22 + s3 x33 = 1,
conductor. In the normal metal, the current is
carried by electrons, whereas in the supercon- where s1 , s2 and s3 are the principal values of the
ductor it is carried by Cooper pairs. The con- tensor sij . If all sii are positive, then the charac-
version of normal excitations into pairs in the teristic surface is (in the general case) a three-axis
normal metal takes place at the distance λE from ellipsoid; with its semiaxes equal, respectively,
the interface. The electric field E differs from to (s1 )−1/2 , (s2 )−1/2 and (s3 )−1/2 , and pro-
zero in this region, therefore λE may be consid- portional to the values of the particular physical
ered as the depth of penetration of a longitudinal property represented by the sij in question along
(in contrast to the Meissner effect) “dissipative” the principal axes. For a random direction, speci-
electric field. The nonequilibrium state of the fied by radius vector r, we have rij = (sij )−1/2 .
superconductor exists within the distance ∼λE In accordance with the Neumann principle
of the interface. This state is characterized by (which states that the symmetry group of a cer-
an imbalance in the occupancy of electron-like tain physical property of a crystal must include
(P > PF , where PF is the Fermi momentum) the symmetry elements of the point group of the
and hole-like (P < PF , where P is excitation crystal), the symmetry of the characteristic sur-
momentum) spectrum band branches. The imbal- face corresponds to the symmetry of the physical
ance relaxation is related to inelastic collisions property. For all positive values of sii the char-
(i.e. electron–phonon types), characterized by a acteristic surface is a sphere for cubic crystals
mean free path length lε = vF τε , which are in- which belong to the highest symmetry category.
efficient at low temperatures, and therefore lead The characteristic surface of the crystals which
to rather high values of λE (e.g., λE ∼ 10 μm). belong to the medium category, those with tetrago-
According to theory, λE = (DτQ )1/2 , where nal, hexagonal or trigonal structures, has the shape
D = vF ll /3 is the electron diffusion coef cient, of spheroid (prolate or oblate ellipsoid), whereas
168 CHARACTER OF GROUP REPRESENTATION MATRIX
crystals of the lowest symmetry category which in random-access memories. The mechanism of
are orthorhombic, monoclinic, or triclinic, are de- charge coupling action can, to some extent, be
scribed by a characteristic surface in the form of a represented by a combination of a bipolar tran-
three-axis ellipsoid. An example of a characteristic sistor (motion of minority current carriers) and a
surface associated with optical anisotropy repre- eld-effect transistor (controls carrier motion by
sents the dependence of the refractive index n on changing displacement voltage at the gate). The
the direction in the crystal; depending on the shape electrodes (gates) have a mosaic structure, and
of this surface crystals are classified as optically the distance between the edges of adjacent el-
isotropic, uniaxial crystals or biaxial crystals. ements does not exceed the diffusion length of
The main property of the characteristic sur- nonequilibrium carriers. The application of a bias
face may be stated as follows: if vectors A and voltage to a single gate element creates a “poten-
B are interconnected by the relation Bi = sij Aj tial well” under this element for minority carriers.
and the radius vector r of the characteristic surface The latter accumulate under the gate in quanti-
tensor sij is parallel to the vector A, then the di- ties proportional to their rate of generation. In a
rection of the vector B coincides with the normal two-dimensional array of such elements, each “op-
to the plane, which is tangent to the characteristic erating element” with the bias that corresponds to
surface at the contact point of the vector r. Char- the accumulation of minority carriers has unbiased
acteristic surfaces apply in the linear or harmonic elements as neighbors on every side. Upon supply-
response range of physical properties. ing the bias to an element adjacent to the operating
Besides the characteristic surface, the concept one, a “potential well” appears under it. If the
of an indicating surface is sometimes introduced: well is deep enough the packet of minority carri-
ers from the “operating element” well moves into
x12 x2 x2
+ 22 + 32 = 1. it. The sequence of voltage pulses at the gate ele-
s12 s2 s3 ments is set so that the packet shifts by the length
Radius vectors, which terminate at this surface, of one element with each pulse, thus the charge of
designate the absolute values of the physical prop- every “operating” element can be led sequentially
erties in a given direction. through to the output element, producing an output
signal proportional to this charge.
CHARACTER OF GROUP REPRESENTATION
MATRIX CHARGE DENSITY OSCILLATIONS
The trace or sum of the diagonal elements of See Plasma oscillations.
a group representation matrix. The character is in-
dependent of the choice of coordinate system, and CHARGE DENSITY WAVE in metals
all irreducible representations of the same class Periodic redistribution of electronic, ionic and
have the same character. The character table of total charge in space, caused by small periodic
the irreducible representations has orthogonality displacements of ions near their equilibrium po-
properties (see Group theory). sitions in the crystal lattice. The charge density
wave (CDW) state is detected via the scatter-
CHARGE CARRIER DRIFT ing of X-rays, fast electrons and neutrons, which
See Current carrier drift. exhibits diffraction peaks from the source lat-
tice and weaker peaks from their “satellites” (see
CHARGE COUPLED DEVICES X-ray structure analysis, Electron diffraction analy-
A broad class of devices (charge transfer de- sis). A charge density wave can appear when the
vices, charge coupling devices, and related charge- metal is cooled below some critical temperature.
injection devices) that allows for the scanning The phase transition to the CDW state is charac-
of an optical image in a manner similar that of terized by changes in the temperature dependence
a cathode ray tube (vidicon). The main purpose of the resistance, Hall constant, magnetic sus-
of such devices is the transmission of TV sig- ceptibility, and a modification of the electronic
nals of various spectral bands: they are also used spectrum of the metal. The charge density wave
CHARGED PARTICLE ENERGY LOSS 169
period can be commensurate with the period of energy gap may occur on the whole Fermi surface.
the source lattice, and in these cases one can talk Then quasi-one-dimensional compounds in a state
about commensurate CDWs which differ from in- with a charge density wave become insulators (see
commensurate ones. As a rule in the cases of Peierls transition).
incommensurability the lattice period depends on
the temperature, and structural transitions to the CHARGE DENSITY WAVE IN
commensurable CDW are possible. SUPERCONDUCTOR
The transition to a charge density wave state See Superconductor with charge (spin) density
is found in metals with strong anisotropy of the waves.
electronic spectrum. This anisotropy can have
CHARGED PARTICLE ENERGY LOSS in a solid
a two-dimensional character, when the electrons
move freely in a plane (their Wannier wave func- This loss occurs during the interaction of in-
coming charged particles with atoms and electrons
tions at different points overlap), but between the
of a target. It is characterized by the average brak-
planes their motion is restricted (due to the weak
ing ability equal to −dE/dx, where E is the
overlapping of the electronic Wannier functions).
particle energy and coordinate x is its initial di-
Related to such compounds are, e.g., the layered
rection, and by straggling of the charged particle
compounds of dichalcogenides of transition met-
energy loss (by the root mean square scattering
als like TaS2 , NbSe2 (see Quasi-two-dimensional
of the energy loss value). This energy loss de-
crystals). Anisotropy of a one-dimensional type is
termines the values of charged particle path in
realized in compounds with a chain structure, e.g.,
the material. Calculation of the energy loss is a
in organic conductors (see Quasi-one-dimensional
complex many-particle problem, and to solve it
crystals).
requires the adoption of a series of approxima-
The origin of transitions to the CDW state tions which differ for high- and low-energy par-
for all these systems is associated with the fea- ticles, for long-range and short-range collisions,
tures of the geometry of the Fermi surface. The and also for the different correlations of masses
theory shows that if sufficiently large sections and charges of interacting particles, taking into ac-
of this surface are nested, meaning that they are count the so-called density effect associated with
can be overlapped with each other by transla- the polarization of the medium by the ionizing
tion along a reciprocal lattice vector Q, then the particle, of the resonance recharging of the mov-
polarizability of the system of electrons in the pe- ing ions, etc. The main mechanisms of charged
riodic electric field of the lattice E = Re (E0 eiQr ) particle energy loss are the nuclear decelleration
(r is the radius-vector of the lattice point) is high or braking caused by elastic collisions (taking
and the lattice becomes unstable with respect to place without variation of internal energy), col-
the appearance of periodic distortions with wave lisions of atoms (ions); electron braking caused
vector Q. The distortions are formed below the by the inelastic collisions leading to the excita-
critical temperature and cause an energy gap to tion and ionization of the atomic electron shells,
appear in the electronic spectrum at the nested sec- and also of the free electrons of the irradiated
tions of the Fermi surface, i.e. with the complete material (see Characteristic energy loss). Besides,
or partial loss of metallic properties. The degree there is the charged particle energy loss from radi-
of nesting, as well as the extent of the metallic ation – Bremsstrahlung and Cherenkov radiation;
property loss grows with increasing anisotropy of the loss is substantial for light particles, especially
the electronic spectrum. In quasi-two-dimensional for electrons and positrons. The value of this loss
(layered) compounds the appearance of an en- depends on the energy, charge, and mass of the
ergy gap at the whole Fermi surface is impossi- moving particles, as well as on the charge, mass,
ble, and they preserve their metallic properties at and density of the target atoms. For motion close
transitions to CDW state, or in case of large dis- to the direction of low-index crystallographic axes
placements of atoms they become semimetals. In and planes, the energy loss strongly depends on
quasi-one-dimensional compounds the region of the direction of the incident beam (see Orientation
Fermi surface nesting is more extensive, and the effects; Channeling); in other cases the target may
170 CHARGED PARTICLE PATHS
be approximately considered as an amorphous one particle momentum (e.g., the latter may be nor-
with isotropic loss. In the approximation of the mal to the bombarded plane, yz). The transverse
Lindhard–Scharf–Schiott theory charged particle path R⊥ characterizes the change of the initial
energy loss of ions in an amorphous medium is coordinate of the particle in the yz plane at the
expressed in terms of dimensionless variables for time the particle stops. Finally, one may consider
the energy ε = aEM2 /[e2 Z1 Z2 (M1 + M2 )], and the total or effective path RL which is equal to
coordinate ρ = 4πxa 2 NM1 M2 /(M1 + M2 )2 , the full length of the distance travelled by the
where M and Z are the masses and atomic particle taking into account all deviations from
numbers of the incident ions (subscript 1) a linear track. The number of particle collisions
and of the atoms of the medium (subscript 2), per unit path length, the impact parameters for
N is the density of the atoms of the medium, each collision, and, hence, the trajectory of each
a = 0.8853aB (Z1 + Z2 )−1/2 , aB is the Bohr
2/3 2/3 individual particle are probability variables, even
radius. Taking into account the structure of the for identical initial conditions of particle motion,
atomic electron shells supplements the results of so there exists a certain distribution of charged
the Lindhard–Scharf–Schiott theory concerning particle paths over their lengths. When bombard-
an oscillating (within limits of tens of percents) ing a sample with a beam of ions of the same
dependence of dE/dx on Z2 caused by the contri- chemical element with equal initial energies the
bution of the valence atomic shells to the charged profile of stopped ions appears to exhibit a certain
particle energy loss of the electrons. broadening. Such a distribution is often charac-
The charged particle energy losses of positrons terized by the moments M1 = Rp = x; M2 =
and electrons of not too high energies E < Ecr
Rp2 = (x − x)2 , . . . , Mk = (x − x)k
when the role of braking radiation is not yet ap- (k = 3, 4, . . .), x k = (r cos θ)k p(r, θ; E) dr,
preciable, are determined by nuclear and electron where p(r, θ; E) is the probability for a particle
braking, with values having qualitative regulari- entering the crystal with energy E at the zero
ties close to the charged particle energy loss of coordinate point along the direction n to stop
ions, but differing from the ion values mainly due at point r, so that θ = arccos(nr/r). The value
to the mass difference between the electrons (tak-
Rp presents the statistical average scatter (strag-
ing into account relativistic corrections) and ions. gling) of particle paths, M3 that is a measure of
The value of Ecr may be evaluated from the cor- distribution asymmetry. The accepted measure of
relation Z2 Ecr [MeV] ≈ 800. For different energy asymmetry is the so-called distribution skewness,
ranges, masses, particle charges, and target para- 3/2
Sk = M3 /M2 . In the case of Sk < 0, the path
meters, there are different approximations for the distribution falls off more sharply from its max-
calculation of the charged particle energy loss, and imum in the bulk sample, and when Sk > 0 a
there are also many available experimental data. sharper fall-off is at the surface of the sample. The
pathways become longer in the case of channeling.
CHARGED PARTICLE PATHS in condensed Electron pathways in the medium typically fea-
media ture much sharper trajectory breaks compared to
Characteristics of charged particle trajectories the pathways of heavy particles. Because of mul-
that are limited by the charged particle energy tiple collisions, the electron quite soon “forgets”
loss during collisions with atoms and electrons its initial direction of propagation. Moreover, elec-
of the medium. The particle path in a medium trons penetrate the medium in a diffusion manner.
is a broken line since each collision results in a To retrieve the pathways experimentally electrons
change of particle momentum and direction. The of varying initial energies are shot through thin
total vector path, R, is the distance between the plates. The maximum plate thickness at which a
particle entry point and its terminal point where throughput electron current is still recorded corre-
it has already expended most of its energy and sponds to the maximum path length for an elec-
stopped. The projective path Rp presents a pro- tron with a given energy penetrating the material.
jection of R onto the direction x of the initial A normal pathway is one with a plate thickness
CHEMICAL BONDS 171
such that the initial electron beam intensity de- CHEMICAL BOND PREDISSOCIATION
cays by a factor of 6 when passing through it. The state of a chemical bond, usually an ex-
An empirical dependence for electrons with ini- cited one, having an energy that exceeds the en-
tial energies E 0.6 MeV is Rp [mg/cm2 ] = ergy of atoms in the dissociated state. The latter
526E[MeV] − 94. is separated from the initial state by a barrier
which may be overcome by tunneling (see Tunnel-
CHARGE, MAGNETIC ing phenomena in solids). The simplest example
See Magnetic charge. of chemical bond predissociation is the excited
hydrogen molecule H2 with a dissociation en-
CHARGE-TRANSFER COMPLEXES ergy ≈4.4 eV, whereas the excitation energy upon
Organic compounds consisting of molecules transition of one of the atoms to the level 2p
of two types: one acts as an electron donor and is ∼10 eV. Another example is the state of an
the other as an electron acceptor, the charge be- α-particle in a radioactive nucleus, e.g., of Ra
ing redistributed between the neighboring donors or U, whose average lifetime is, respectively, 1600
and acceptors. Charge-transfer complexes have an and 4.4·109 years.
electric dipole moment even if they are created
from nonpolar molecules, and they exhibit para- CHEMICAL BONDS in solids
magnetic properties, whereas the initial donor and Bonds between atoms of a solid; isolated atoms
acceptor molecules are diamagnetic. If an electron decrease their free energy during the redistribution
in the ground state of the organic solid is com- of their electron densities when they join together
pletely transferred from donor to acceptor, such a through chemical bonding. The nature of a chem-
compound is called an ion-radical salt (limiting ical bond is determined by the nature of the con-
bond case in charge-transfer complexes). Besides stituent atoms, and by the type of structure adapted
weak charge-transfer complexes of ion-radical by the solid. There exist molecular crystals, whose
salts, there are many such compounds with inter- structural elements are molecules; under evapora-
mediate strength bonds. The properties of charge- tion this crystal type decomposes into molecules.
transfer complexes are strongly dependent on the Crystals can be formed from monatomic mole-
extent of charge transfer, and on the magnitude cules of noble gases (He, Ne, Ar, Kr, Xe, Rn),
of the donor–acceptor interaction. The electrical diatomic molecules (H2 , O2 , NO, CO), triatomic
conductivity of these complexes at room temper- molecules (H2 O, H2 S, NO2 ), and molecules of a
ature varies over a wide range: from 10−15 up more complex composition (NH3 , CH4 , etc.). In
to 102 –103 −1 ·cm−1 . The temperature depen- these cases every molecule has saturated bonds,
dence of the electrical conductivity is of the semi- and the attraction of molecules toward each other
conductor type; hence, most charge-transfer com- is through van der Waals forces or hydrogen
plexes are organic semiconductors. These are ex- bonds, as in H2 O, H2 S, NH3 . A certain (addi-
emplified by compounds of different molecules tional) role is played by Coulomb forces, if the
with the molecular acceptor tetracyanoquinodi- atoms have a partial charge, as, e.g., in CO2 and
methane (TCNQ). Because of the strong acceptor H2 O molecules. In cases when the molecules con-
properties, the planar structure, and the high sym- tain atoms of Groups I, II, III, together with those
metry of the TCNQ molecule, one can obtain of Groups V, VI, VII of the periodic table, i.e. they
charge-transfer complexes with a rather high con- are polar molecules, then Coulomb forces become
ductivity (as high as 104 −1 ·cm−1 ). A large dominant to the extent that individual molecules
variety of organic charge-transfer complexes have are not observed in the crystal, and its structural el-
been synthesized that exhibit properties ranging ements are ions. Typical examples are provided by
from those of semiconductors to metals and super- alkali-halide crystals, and oxides of alkaline earth
conductors (see Superconductivity). The latter two metals. In these compounds every ion is usually
categories include mainly ion-radical salts. Mole- symmetrically surrounded by ions of opposite sign
cular charge-transfer complexes are of interest as (six for FCC NaCl, eight for BCC CsCl); and the
photosemiconductors (see Photoconductivity). stronger ionic bond is formed. For these cases the
172 CHEMICAL DIFFUSION
action of van der Waals forces between ions be- exhibit several types of chemical bonds, usually
comes of secondary importance. with one or two bonding types dominant.
The opposite case is provided by crystals of el-
ements of Group IV: diamond, silicon, germanium, CHEMICAL DIFFUSION
gray tin. There are no charges on the atoms of these The same as Heterodiffusion.
crystals, and the formation of covalent bonds, or
homopolar bonds, is due to the sp3 -hybridization CHEMICAL HEAT TREATMENT
of orbitals (atomic orbitals), and to the attraction Heat treatment in a chemically active medium
of electrons to the nuclei (skeletons) of neighbor- involves the diffusive saturation of a metal surface
ing atoms. As only four hybrid orbitals may be with chemical elements of the medium. Both non-
formed by a combination of s- and p-orbitals, the metals (B, C, N, etc.) and metals (Al, Ti, Cr, Zn,
number of closest neighbors (coordination num- etc.) are used as diffusible elements. In a num-
ber) for these crystals is limited to four. Each ber of cases simultaneous saturation with metals
atom of one sublattice is tetrahedrally surrounded and nonmetals is used, including multicomponent
by four atoms of the other sublattice. The binary saturation. During chemical heat treatment a diffu-
A(n) B(8−n) type crystallization of elements (spha- sion layer is formed as a result of prior processes
lerite lattice) is formed analogously, e.g., BN, GaP, of dissociation in the active external medium, ad-
ZnS, HgTe, CuCl. However, the atoms of the metal sorption of dissociated atoms on the surface of
and the metalloid in these compounds possess the base metal, and diffusion of the saturating
fractional charges of opposite sign, with the metal element. Chemical reactions bring about dissoci-
atom positively charged. This is a combined (hy- ation with the generation of active atomic states of
brid) polar-covalent bond (ionic-covalent bond). chemical elements (e.g., ammonia decomposition
The smaller the number n of the periodic table under nitriding). Adsorption of dissociated atoms
group, the more ionic the bond. may take place either through the action of van
In metals and alloys the so-called metallic bond der Waals forces of attraction, with the genera-
also occurs: it is formed through delocalization of tion of an adsorption layer (physical adsorption),
valence electrons. At the formation of the crystal or because of chemical bonding with atoms of
the outer (valence) s- and p-electrons (elements a metal surface (chemisorption). Simultaneously
of Groups I to III, and transition elements) leave with these processes, the main chemical heat treat-
their atomic nuclei and migrate between ions over ment process goes on, namely the diffusion of
the entire lattice. As this takes place, their kinetic adsorbed atoms into the depth of the metal. The
energy decreases (in accordance with the uncer- result is the formation of a diffusion layer with a
tainty relation). It is this energy decrease that is certain distribution of diffusible elements relative
responsible for the formation of a metallic bond; to the depth. The concentration changes monoton-
alkali metals and alkaline earth metals possess rel- ically with depth, so a continuous sequence of
atively weak metallic bonds (1 to 2 free electrons solid solutions forms with the base metal. The
per atom), whereas elements such as Fe, Ti, Pt ex- chemical composition of the layer is more compli-
hibit rather strong metallic bonds. cated if the solid solutions have limited solubility,
For a fairly broad class of hydrogen-containing or if chemical compounds form. In this case the
organic compounds there is a noticeable contribu- layer includes different phases, in accordance with
tion to the crystal formation energy from hydrogen the phase diagram of the saturating element with
bonds. Typical examples are provided by ice, solid the base metal at the diffusive saturation temper-
ammonia, and the ferroelectric crystal KH2 PO4 . ature, and the concentration dependence on the
Because of its comparatively small mass, a pos- depth can undergo abrupt changes. The duration
itively charged hydrogen atom exhibits quantum of diffusive saturation is determined by the depth
behaviour and spends some time near an elec- of the diffusion layer, which increases quadrati-
tronegative atom (see Electronegativity) of one of cally with time. The main methods of chemical
the neighboring molecules, which results in a de- heat treatment with respect to the type of ac-
crease of the overall system energy. Many solids tive medium are: saturation from powders and
CHEMICAL SHIFT 173
pastes, gaseous media, melts of metals and salts, component equals the partial derivative of every
as well as evaporation of the diffusible element in thermodynamic potential (e.g., Helmholtz free en-
vacuo. Chemical heat treatment involving a single ergy F ) with respect to Ni while holding other
element includes nitriding, aluminizing, boroniz- thermodynamic quantities constant (e.g., μi =
ing, siliconizing, titanizing, zinc plating, cemen- ∂F /∂Ni at T = const, V = const and Nj = const
tation. Saturation with two elements takes place for j = i). The chemical potential for constant
at aluminoboronizing, aluminosiliconizing, chro- pressure conditions is closely related to Gibbs free
mosiliconizing, nitrocementation, cyaniding. This energy G = i μi Ni , and in the case of a single-
treatment involving several diffusible elements in- component system μ = G/N , i.e. the chemical
cludes saturation with boron and silicon together potential is equal to Gibbs energy per particle.
with other metals (Ti, Cr, Mo, W, etc.), as well as For quantum systems of bosons, which obey
saturation with aluminum and chromium together Bose–Einstein statistics, the chemical potential is
with other metals (Ti, V, Mn, Zn, etc.). negative at temperatures higher than the Bose–
Chemical heat treatment is used in industry to Einstein condensation temperature T0 , and is equal
increase the reliability and service life of metallic to zero below T0 , and in systems with varying
working parts of devices and engines. Depending numbers of particles (phonons in solids, pho-
on the chemical composition of the metallic work tons in equilibrium with thermal radiation). For a
piece and the nature of the saturating elements, system of fermions, which obey Fermi–Dirac sta-
treated articles may acquire increased hardness, tistics, the chemical potential is positive at temper-
strength, corrosion resistance and cavitation resis- atures below the degeneracy temperature (Fermi
tance, wear resistance (see Wear), etc. temperature) TF , it coincides with the Fermi en-
ergy EF at absolute zero of temperature; and for
CHEMICAL METHODS OF ANALYSIS T > TF the chemical potential satisfies the in-
Methods of analysis based on the use of chem- equality μ < EF .
ical reactions. Methods of quantitative analysis In external potential fields (gravitational, elec-
involve the determination of the masses of the tric) the role of the chemical potential is played
reaction products (gravimetry); determination of by the quantity μi = μ0i + Ui (r) = const, where
the volume of the solution of one substance re-
μ0i is the chemical potential in the absence of the
quired to react quantitatively with a given solution
field, and Ui (r) is the potential energy of particles
volume of another substance (titration analysis),
of the ith type in the external field. For charged
determination of thermal effects of the reaction
(calorimetry), of its rate (kinetic methods), etc. particles in an electrostatic field E = −∇ϕ(r)
Methods of qualitative analysis are based on car- the expression for the electrochemical potential
rying out chemical reactions, which indicate either μi = μ0i +ei ϕ(r) = const holds true. Equal values
the presence of ions of certain type in the solution, of chemical potential in different phases define
or the presence of certain atoms in organic com- equilibrium conditions for phase equilibria (the
pounds. Chemical methods of analysis have been so-called Gibbs’ phase rule) and chemical reac-
largely superseded by more accurate physical and tions. The chemical potential in a solid determines
physicochemical methods of analysis. the electronic work function of a crystal for e ld
electron emission, thermionic emission and photo-
CHEMICAL POTENTIAL (introduced by electron emission, electron af nity, contact poten-
J.W. Gibbs, 1873–1878) tial difference, and other properties.
Thermodynamic state function, which defines
the change of thermodynamic potential for the CHEMICAL SHIFT
change in the number of particles of the ith type Term used to describe changes (depending on
Ni in open multicomponent systems with vary- the specific nature of the material under investiga-
ing numbers of particles, which are described by tion) of the positions of energy levels or spectral
the grand canonical distribution (see Gibbs dis- lines which can arise from the local electronic
tributions). The chemical potential μ of the ith or chemical environment. The values of chemical
174 CHEMILUMINESCENCE
shifts are usually given as parameters correspond- chemical shifts in their ground state energies due
ing to the spectrum (energy state) of a given to correlation and exchange effects involving the
material. Generally the chemical shift is measured donor and core electrons.
either in the same units as the basic quantities
which characterize the spectrum, or in dimension- CHEMILUMINESCENCE
less units. Luminescent excitation that results from chem-
The term is most widely used in high resolution ical reactions. The process of chemiluminescence
nuclear magnetic resonance (NMR) spectroscopy,
involves the emission of radiation either by prod-
ucts of a chemical reaction, or by other com-
in Mössbauer spectroscopy, in the physics of
ponents which become excited by absorbing en-
semiconductors to describe the location of impu-
ergy transferred to them from reaction products.
rity energy levels, etc.
Chemiluminescence in solids arises at a bound-
In high-resolution NMR the chemical shift is
ary with a gaseous or liquid phase as a result of
the shift of the resonance signal. The mechanism
heterogeneous reactions (see Heterogeneous sys-
producing this shift is as follows: an atom (mole- tem), which proceed either spontaneously (intrin-
cule) is placed in an applied magnetic field B 0 ; sic chemiluminescence) or under action of various
and as a result of Larmor precession of elec- external factors: e.g., electrical, electrolysis (elec-
trons oriented parallel to and antiparallel to B 0 the troluminescence). A specific case of chemilumi-
field B ∗ arises (see Diamagnetism), so the nucleus nescence is bioluminescence (e.g., luminescence
experiences the action of an effective magnetic of rotten wood). Chemiluminescence constitutes
field B = B0 − B ∗ = B0 (1 − σ ), where σ is the basis of action of chemical lasers, and is used
called the screening constant; usually σ ∼ 10−6 . for the analysis of gases and solid surfaces.
For nuclei of a given type, which are in differ-
ent electronic surroundings, i.e. either in different CHEMISORPTION, chemical adsorption
molecules or at nonequivalent sites of one mole- Chemical binding of particles of a substance
cule, the resonance will be observed for different (atoms, molecules, radicals) by a solid surface. It
values of B0 (at a given frequency). In practice the differs from physical adsorption by its stronger
NMR chemical shift is measured in relative units bonding, that is by a higher value of the bind-
δ = (B − Bst )·106 /B0 , where B and Bst are the ing energy (0.5–10 eV). The bond between the
respective resonance values of the external field adsorbed particle and the solid can have both co-
for the sample under investigation and a standard valent and ionic components; the former due to
substance. The chemical shift is one of the main overlap of the electronic wave functions of the ad-
characteristics of high resolution NMR spectra, sorbed particle and the substrate atoms, and the
which are observed for liquids (and occasionally latter due to the transfer of charge from the ad-
gases). The value of this shift provides informa- sorbed particle to the solid material. An adsorbed
tion on the electronic structure of molecules, and ion in conjunction with charges which screen it
the nature of chemical bonds. The development can form a dipole, which results in the adsorbed
film changing the electronic work function of the
of a recent trend in spectroscopy, namely high-
solid. Particles chemically adsorbed on the surface
resolution NMR spectroscopy of solids, permits
of a solid interact with one another, and some-
the measurement of chemical shifts in solids.
times generate an ordered two-dimensional lattice
In the Mössbauer effect the chemical shift (also
of adatoms (see Monolayer). Chemisorption plays
called isomer shift) makes its appearance if the
an important role in the oxidation of metals, het-
source and absorber differ chemically. erogeneous catalysis, crystal growth, electronics,
In the physics of semiconductors a chemical etc.
shift manifests itself in the difference between
the energy states of impurity levels, expressed in CHEMISTRY
terms of the effective mass, and the energy states See Crystal chemistry, High-energy chemistry,
calculated from theory, taking into account correc- Mechanochemistry, Mesonic chemistry, Quantum
tions to the potential due to the specific atom (ion). chemistry, Radiation chemistry, Solid state chem-
For example, some donor levels exhibit fairly large istry.
CHIRALITY OF MAGNETIC BUBBLE DOMAINS 175
CHERENKOV RADIATION (S.P. Vavilov, in the system Mo–Re and at 2150 ◦ C in the sys-
P.A. Cherenkov, 1934); referred to as tem W–Re; and this reaction only takes place
Vavilov–Cherenkov radiation in the Russian for a narrow region of homogeneity. In the sys-
literature tems Nb–Re and Ta–Re the χ -phase melts without
Photons emitted by charged particles moving decomposition and exhibits a broad range of ho-
in a medium with a speed v which exceeds the mogeneity. In systems formed by Re with Ti, Zr
speed of light u = c/n in this medium, where n and Hf the composition of the χ -phase is approx-
is the refractive index. In condensed matter with imately Me5 Re24 ; it is formed by the peritectic
high values of n Cherenkov radiation is observed reaction, and has a moderate range of homogene-
at electron energies of about 105 eV, and at protons ity (Me–Ti, Zr or Hf). The chi phase has also
energies of 108 eV. This radiation is explained via been observed in binary systems Al–Re, Nb–Os,
Huygens principle by noting that waves which are Tc–Zr. Chi phases are assumed to be “electron
emitted at points along the charged particle trajec- compounds” (see Hume-Rothery phases). A num-
tory for the condition v > u will overtake waves ber of chi phases transform to the superconduct-
emitted earlier by the particle. As a result of their ing state, with the highest transition temperature
interference, a light cone is formed having a ver- Tc = 12.9 K for Nb0.24 Tc0.76 (see Superconduct-
tex that coincides with the instantaneous position ing intermetallic compounds).
of the emitting particle, with θ the angle between
the direction of the particle’s movement and the CHIRALITY (fr. Gr. χειρ , hand)
direction of propagation of the outgoing radiated Handedness, or belonging to one of two (“left-
wave satisfying the relation hand” or “right-hand”) non-superimposable mir-
1c ror image configurations. See also Enantiomorph-
cos θ = . ism.
nv
The spectral distribution of energy E(ω) for Che- CHIRALITY IN LIQUID CRYSTALS
renkov radiation is determined by the relation Absence of planes of symmetry in the mole-
cules of liquid crystals. Chirality in liquid crystals
dE e2 c2
= 2ω 1− 2 2 leads to instability of a uniform (homogeneous)
dω c v n state, and the formation of spatially-periodic (spi-
for the condition v > c/n(ω). There is no radi- ral) structures (right-handed or left-handed) in
ation for frequencies ω which do not satisfy the some liquid crystal phases (cholesteric liquid crys-
above condition. Cherenkov radiation has found tals, blue phases, smectic liquid crystals of the
wide application in Cherenkov counters used for C type); see Chirality, Enantiomorphism.
determining the speeds, charges, and masses of
fast charged particles (electrons, positrons, mes- CHIRALITY OF MAGNETIC BUBBLE DOMAINS
ons, protons, and antiprotons); and for study- The sign of the relative rotation of the mag-
ing optically anisotropic media, including wave- netization in the center of a domain wall for a
guides. magnetic bubble domain with topological index
S = 1, which does not contain Bloch lines. The
CHI PHASE (χ -phase) magnetization in a magnetic domain wall is ei-
An intermetallic compound with a cubic lattice ther aligned along the vector tangent to the lateral
isomorphic to that of cubic α-Mn. It is formed surface of a domain wall, or is opposed to it.
mainly by transition metals. The best known chi Hence two possible states of a magnetic bubble
phases are those involving rhenium with refractory domain, labeled χ ± , are more favorable (in the
materials. In alloys of rhenium with molybdenum, context of magnetostatics) than all other states. By
tungsten, niobium and tantalum the chi phase con- convention the label χ + is assigned to a clock-
tains about 75 at.% Re. In the systems Mo–Re wise rotation of the magnetization when observing
and W–Re the chi phase is obtained by the peri- along the rotation axis direction. The twisting de-
tectic reaction (see Peritectic alloys) at 1850 ◦ C flections of χ + and χ − are equal in magnitude.
176 CHOLESTERIC LIQUID CRYSTAL
It should be noted that, by virtue of domain wall flat space the Christoffel symbols are used to form
twisting, the above definition is valid only for the covariant derivatives of tensors. These symbols are
cross-section of the magnetic bubble domain by used in elasticity theory to characterize changes of
a plane that is at equal distances from the mag- the basis vectors ei of the local coordinate system
netic lm surfaces. Transitions between the χ + deformed together with the body: dei = Γkil el dx k .
and χ − states of magnetization rotation usually Christoffel symbols and gik are both expressed
occur when horizontal Bloch lines are present. through the strain tensor uik : gik = δik + 2uik
(δik is the unit tensor), Γi,kl = ∂k uil + ∂l uki −
CHOLESTERIC LIQUID CRYSTAL, cholesteric, ∂i ukl . To take into account dislocations and inter-
chiral nematic, cholesteric phase (fr. Gr. χoλη , nal stresses, the Christoffel symbols are replaced,
bile, σ τ ερεoς , solid, and χειρ , hand) within the framework of the mechanics of solids,
An analog of a nematic phase (see Nematic liq- by the coefficients of linear (affinity) coherence,
uid crystal) for a material containing chiral mole-
Γikl . The number of independent values of Γikl
cules, that is molecules which differ from their
is equal to 15 (in contrast to 9 for the previ-
mirror image structures (see Chirality in liquid crys-
ous case). Γikl may be separated into symmetric
tals). When chiral molecules are dissolved in a ne-
(Γi(kl) ) and antisymmetric (Γi[kl] ) parts with re-
matic liquid the structure undergoes a helical dis-
spect to the indices k, l: Γikl = Γi(kl) + Γi[kl] . The
tortion to form a helical phase called a cholesteric.
quantity Γi(kl) characterizes the pure deformation
This phase was originally observed in the ethers
of cholesterol, hence its name. A cholesteric, like a of the trihedron defined by the basis vectors ei ,
nematic, lacks long-range order, and the molecular and Γi[kl] characterizes the rotation of the trihe-
orientation has a preferred axis denoted by a direc- dron during its transition from point to point. The
tor n which is perpendicular to the helical z axis. quantity Γi[kl] is directly linked to the density of
The local symmetry is close to ∞22 (or C∞h ), dislocations in the material; the sum of the Burg-
with the spatial distribution of the apex positions ers vectors of dislocations passing through a small
of the director vector n tracing out a spiral or helix. area S is proportional to the convolution of the ten-
The spatial period or step along the helical axis is sor of the area S and Γi[kl] .
one half the pitch of the spiral because director ori-
entations n and −n are equivalent, as is the case CHROME PLATING
in nematics. The step of the spiral L depends on
Application of chromium coatings to the sur-
the chemical composition and on the temperature,
face of materials to increase their hardness, wear
and typical values lie in the range from 350 nm
resistance (see Wear), corrosion resistance and
to 650 nm, corresponding to visible and near UV
friction resistance, and also for decorative pur-
wavelengths. Cholesteric liquid crystals possess
unique optical properties such as Bragg diffraction poses. The thickness of the coatings varies from
of light beams at the spiral structure, enhanced op- 1 μm to 3 mm, their hardness ranges from 500
tical activity, etc. These properties are widely used to 1200 kg/mm2 . Electrolytic chrome plating is
in various fields of engineering, in particular in the carried out from aqueous chromium-oxide solu-
medical area. tions either with a sublayer of copper and nickel,
or without it. The appearance, structure, physic-
CHRISTOFFEL SYMBOLS (E.B. Christoffel, ochemical and mechanical properties of the coat-
1869) ings may be changed by variation of the electrolyte
The Christoffel symbol is Γkli = g in Γn,kl , composition, and the method of electrolysis. Also
where the metric tensor gik defines Γi,k,l = in some areas of engineering diffusion chrome
(∂k gil + ∂l gki − ∂i gkl )/2, and ∂i = ∂/∂x i . The plating is used, preferably from the lowest halide
Christoffel symbols do not form a tensor; in partic- compounds of chromium, by means of thermal
ular by choosing a particular system of coordinates decomposition or reduction in a gaseous, solid
one may reduce all the Christoffel symbols to zero or liquid medium, with successive diffusion of
at any given point. In curvilinear coordinates in a chromium into the depth of the parts being coated.
CINNABAR 177
CIRCULAR PHOTOVOLTAIC EFFECT sign of the circular polarization changes, but also
Photovoltaic effect observed in doubly refract- when the light propagation direction changes. The
ing crystals (see Gyrotropic media) illuminated by circular photovoltaic effect may used in inertialess
circularly or elliptically polarized light (see Po- receivers of polarized laser radiation, in elements
larization of light). There is a difference between of optical memory on photorefractive crystals (see
the circular photovoltaic effect and other photo- Photorefraction), and also as a mechanism of four-
voltaic effects: the resulting current is proportional wave mixing of isofrequency waves in various
to the degree of circular polarization of the radia- nonlinear optical frequency converters.
tion Pcirc and changes its direction in accordance
with the sign of the polarization. The “circular” CLATHRATE COMPOUNDS (fr. Gr. κληθρα ,
photovoltaic effect for elliptically polarized light adler-tree)
is described by the following relation: Phases comprising at least two compounds,
with the molecules of one enclosed in the molecu-
jα = γαβ i[E × E ∗ ]β , lar voids or lattice cavities of the other. They bear
a close similarity to interstitial solid solutions.
where j is the current density, and E is the am-
As a rule, only van der Waals forces act be-
plitude of the light electric field (see also Bulk
tween the molecules of a clathrate compound. The
photovoltaic effect). For circular polarization this
binding energy of each enclosed molecule is about
becomes
≈50 kJ/mol. This relatively high value for van
i[E × E ∗ ] = |E|2 Pcirc q/q, der Waals interaction is explained by the mole-
cule and void geometry fitting well into one an-
where q is the wave vector of the light. Initially other, and by the many points of bonding by the
the circular photovoltaic effect was observed in tel- molecule. Depending on the void type, clathrate
lurium, then it was studied in a variety of crystals compounds are categorized as lattice or molecu-
including a number of ferroelectrics. The origin of lar types. Lattice clathrate compounds can only be
this effect arises from the asymmetry of the distri- solids, and are further classified by the shape of
bution of the electrons excited by the light. This their voids as channel, cage (originally only this
asymmetry is due to the correlation between the type was termed clathrate), and laminated vari-
electron angular momentum (spin) and its linear eties. Examples of “host” compounds that form
momentum in gyrotropic crystals. The inverse ef- cage clathrates are uria, thiouria, hydroquinone,
fect, the so-called optical activity (see Rotation of phtalocyanines, etc. There is a wide range of
light polarization plane) induced by the current and corresponding “guest” compounds: hydrocarbons,
resulting from the spin orientation during the cur- alcohols, many gases. Cyclic molecules usually
rent flow, has also been observed in tellurium. In form the molecular clathrate compounds, such as
nongyrotropic crystals, which possess no center cyclodextrin and other macrocyclic compounds.
of symmetry, the circular photovoltaic effect does Compounds of relatively small dimensions can
not occur, while the phenomenon of optical activ- be enclosed in the voids of such “hosts”. The
ity may appear from uniaxial deformation or an clathrate compounds exhibit, in the general case, a
applied magnetic field. Such a magneto-induced variable composition. They are employed mainly
circular galvanomagnetic effect was observed in to isolate and purify “guest” molecules.
p-type GaAs. In crystals of low symmetry, includ-
ing those possessing a center of symmetry, the CLAUSIUS–MOSSOTTI EQUATION (R. Clausius,
circular drag effect described by the relation O. Mossotti, 1879)
Relation between the relative (dimensionless)
jα = Tαβγ i[E × E ∗ ]β qγ dielectric constant of isotropic nonpolar insulators,
may be observed. ε(ω), at a frequency ω (see Temporal dielectric dis-
In contrast to the circular photovoltaic effect, persion) and the polarizability, α(ω), of the atoms
the circular drag effect is due to the transfer of that make up the crystal lattice:
phonon momentum to the electrons. The current α(ω)N ε(ω) − 1
in this effect changes its sign, not only when the = , (1)
3ε0 ε(ω) + 2
CLOSE-PACKED STRUCTURES 179
density hexagonal packing or hexagonal close- aggregates that comprise a small particle can be
packed (HCP). The three-layered ABCABC. . . formed during the coagulation of particles, e.g.,
arrangement is given the name highest-density cu- those dispersed in air (aerosol particles), during
bic packing or cubic close packing (CCP or FCC). the sputtering of materials under ion bombardment
The centers of the spheres of the latter structure or other high-energy factors, during evaporation,
form a face-centered cubic lattice. The tendency etc. Surface clusters can be formed by deposition
of atoms to attain closest packing results in the of particles on the surface of a solid. The sta-
fact that most metallic atoms crystallize in one of bility of a cluster can depend on the number n
the close-packed arrays which is almost always of constituent particles, and sometimes it depends
the hexagonal (e.g., Mg, Nd, Zn) or the face- on whether n is odd or even. Thus it was estab-
centered cubic (e.g., Ag, Al, Cu) one. In addition lished by means of mass spectrometry that when
bulky components (e.g., anions oxygen or sulfur) Li atoms are evaporated into an inert gas an ap-
of various widely occurring chemical compounds preciable amount of Li2 , Li6 and Li8 appears,
also form close-packed arrays, and the diversity of whereas Li3 , Li5 , Li7 , and Li9 are either entirely
structures stems from the fact that there are many absent, or are present in only trace amounts. Dur-
ways of arranging the smaller cations throughout ing the ion sputtering of Ag crystals three times
the smaller tetrahedral and larger octahedral cavi- more Agn clusters form with odd n compared to
ties of the closest-packed structure. the number with even n.
A method termed the cluster approximation is
CLOSE PACKING, RANDOM
often employed in numerical computations of the
energy spectra of ideal crystals, or of crystals with
See Random close-packing model.
defects, with a cluster defined here as a lattice of a
finite number of atoms. Atom–atom potentials and
CLUSTER cluster boundary conditions are chosen to provide
A stable aggregation or group of atoms or for a high cluster symmetry, for correct values of
molecules distinguished by certain physical or the crystal lattice constants, and for stability. Sys-
chemical features from other similar stable micro- tems of particles can also have fractal properties
objects. Several particles constitute the lower limit (see Fractal) arising from special rules followed
on the size of a cluster, while the upper limit in building the system. For instance, a percolation
can often be deduced from one of the following cluster is a set of certain lattice sites obtained by
two traits: (1) the addition of one more particle successive translations of a given length from an
to the aggregation appreciably affects the system arbitrarily chosen starting site to other randomly
properties; or, (2) the number of particles at the chosen sites; the dimensions of such a cluster are
surface, i.e. those which differ in properties from not limited to integers (see Percolation theory).
the bulk due to their proximity to the surface, is In quantum scattering theory coupled states of
comparable to the total number of internal parti- several particles are termed clusters; with every
cles in the cluster. These conditions usually limit cluster considered as a separate scattering chan-
the number of particles in a cluster to the range nel.
102 –103 . Larger agglomerates of particles in a
material are called phase precipitates. Examples CLUSTER APPROXIMATION
of clusters are: agglomerations of impurity atoms The breaking down of a system into subsys-
or intrinsic defects (see pile-up of dislocations and tems (see Cluster) that interact weakly with each
point defect clustering) arising in the bulk of a other. This approximation is employed in thermo-
solid due to their mutual attraction and to the diffu- dynamics, statistical physics, and quantum scatter-
sion of particles (during the growth of crystals or ing theory. It involves expressing the wave func-
during heat treatment), under the action of exter- tion of a multiparticle system far away from the
nal energetic particles (see Collision cascade and scattering region as a superposition of states cor-
Subcascades), under the action of light (in photo- responding to individual isolated particles and/or
graphic emulsions), etc. The atomic or molecular groups of them. The cluster approximation in the
COADSORPTION 181
theory of lattice defects in crystals is based on sin- clusters, but also positive and negative ions, poly-
gling out a region in the crystal that incorporates atomic and molecular ions and their fragments,
the defect and several coordination spheres sur- are found in the sputtered flux during the cluster
rounding it. The local defect properties (binding sputtering of metals, semiconductors, insulators,
energy, energy level position, wave function of the organic compounds, polymers and biomolecules.
localized state, lattice configuration) are then com- The mass-spectra of sputtered clusters range from
puted over this region using one of the standard dimers through aggregates of tens and hundreds
molecular theory methods. The interaction of the of atoms, up to large bioorganic molecules and
cluster with the rest of the crystal is accounted for their clusters. The energy distribution of sputtered
by using a self-consistent eld approach. clusters reaches 100–200 eV, the average being
Sometimes the term cluster approximation is 2–10 eV per cluster of ions and 10–20 eV for
employed to designate an approximate method for neutral clusters. The sputtering probability sharply
computing the parameters of an ideal crystal by drops with an increase in the number of atoms in
representing the crystal as a cluster consisting of a the cluster, and it can undergo oscillations associ-
relatively small number of atoms. Certain bound- ated with the binding energy. In particular, there
ary conditions are imposed upon the outermost may be less charged metallic clusters having an
atoms to ensure the stability of such a configura- even number of atoms than those having an odd
tion. See also Watson sphere. one. The mechanisms of the cluster sputtering of
solids are not yet understood. Theories of direct
CLUSTER, CYBOTAXIC knock-out (for small clusters), of recombination,
See Cybotaxic clusters. and of shock waves, have been proposed to ex-
plain the phenomenon. The latter theory allows
CLUSTER DEPOSITION for the emergence of large clusters and of com-
Formation of heterogeneous lm s composed of plete bioorganic molecules (such as insulin) under
clusters of particles. The films form via the depo- the impact of a single heavy accelerated ion, or
sition from a flow of clusters, e.g., produced by ion of a fragment from nuclear fission. Cluster sput-
sputtering of different targets, or by homogeneous tering plays an important role in secondary ion
gas phase nucleation, or due to a heterogeneous mass-spectrometry, and it also finds applications
nucleation at the substrate surface. A cluster may in designing sources for accelerated cluster and
contain from several to a few thousand particles. macromolecular ion formation, including ionized
An assortment of cluster deposition technology bioorganic molecules.
methods provides a range of structures and clus-
ter properties which determine the film formation COADSORPTION
kinetics and film properties. Using the method of A cooperative adsorption of adsorbates of dif-
cluster deposition at low temperature one can ob- fering chemical composition. In the case of physi-
tain the epitaxial deposition of multicomponent sorption the mutual influence of these adsorbates
compounds such as semiconductors, or one can is associated with the difference in the heats of ad-
grow diamond-like carbon lms, oxides, nitrides, sorption, which may result in the particles with
carbides, etc. a greater heat of adsorption driving out those
with lesser heats of adsorption. In the case of
CLUSTER, NONCRYSTALLINE chemisorption, the mutual effects of the adsorption
See Noncrystalline clusters. of different species is associated with the forma-
tion of chemical bonds of the adsorbates with the
CLUSTER SPUTTERING substrate; this can result in a promotion or inhi-
Sputtering of a solid under the action of ion bition of the sorptional activities of neighboring
bombardment during conditions when clusters of centers. Such an influence may stretch over the
different size composed of the solid’s own atoms distance of several lattice constants, and may re-
plus impurity atoms are predominant in the flux sult in the emergence of not only a short-range,
or flow of sputtered material. Not only neutral but also a long-range order. The phenomenon of
182 COAGULATION
heterogeneous catalysis is based on the process of COALESCENCE (fr. Lat. coalesco, grow together,
coadsorption of molecules of differing chemical be coupled)
composition. The spontaneous merger of small particles of
a solid that results in the formation of a single
larger particle whose shape minimizes its free en-
COAGULATION ergy. The phenomenon of the formation of pores
See Coalescence. in a solid is also known. The coalescence of parti-
cles may occur in sols, aerosols, and on the surface
of a solid. Coalescence is distinct from coagula-
COAL tion, the latter being the mere sticking together of
Carbonaceous porous combustible stratified colliding particles. Coalescence is associated with
rock; with a surface of fracture which is either diffusion processes that take place under the influ-
lustrous or lusterless. Coal contains a maximum ence of surface tension forces. It will be slowed
of 30% of incombustible components (ash). Coal down by charged particles. Surface coalescence
is formed by carbonization (metamorphic trans- is conditioned by strains in the substrate, by its
surface roughness, and by its polycrystalline struc-
formation) of the remains of decomposed plants
ture.
under prolonged exposure to temperature, ionizing
radiation of natural radionuclides, etc. According
COATING
to the degree of carbonization, brown coal (lig-
Application of a layer, which has protecting,
nite) and black coal are distinguished. The carbon decorating, conducting, etc., properties, to the sur-
content increases from ∼65% (by mass of organic face of a material. Metals and alloys, as well as
components of coal) for soft lignites to >91% for nonmetallic substances, are used as coating mate-
anthracites; accordingly, the oxygen content de- rials. Application of a metallic coat is performed
creases from 30 to 2%, and the hydrogen content through hard facing with a thin layer of salts in oil
decreases from 8 to <4%. The percentages of sul- (baking coating), chemical deposition, galvaniza-
fur and nitrogen are little affected by the extent tion (electrolytic deposition, see Electroplating),
of carbonization, and are respectively 0.5–3% and dip coating (dipping into melted metal or into
0.5–2%. Coal can also include P, Ge, Ga, B, V, aqueous salt solutions in the presence of oxygen),
W, U, and other elements. The X-ray diffraction vacuum evaporation, cathode sputtering (sprayed
pattern is indicative of the presence of both the metal coating), etc. Nonmetallic coats (varnishes,
amorphous (see Amorphous state) and the crys- enamel and oil paints) are applied to surfaces
tal structures of coal. The carbonization process through brushing, spraying under pressure, fusion.
involves a structural-molecular rearrangement of Colloid mixtures of graphite with varnish, or of a
the organic material; this rearrangement being metal with varnish, are used as nonmetallic con-
ducting coatings. Anodic oxidation is used in order
responsible for the extreme values of the follow-
to coat the surface with an oxide lm ; carbon films
ing characteristics of coal at intermediate stages
are obtained through decomposition of a dried
of carbonization (carbon content ∼85%): elastic
layer of carbonic acid through heating to 473 K,
modulus, microhardness, yield, density, heat of
or high vacuum evaporation.
combustion, caking ability, volume and radius of
pores, extracting characteristics, etc. At these in- COATING, ANTICORROSIVE
termediate stages coals gain the capacity to form See Anticorrosive coatings.
coke on heating. The electrical conductivity of
coal ranges from 10−2 to 10−10 (·cm)−1 . Coals COBALT (Lat. cobaltum ), Co
which contain >91% C exhibit an abrupt increase Chemical element of Group VIII of the peri-
of aromaticity, and an abrupt decrease of electric odic system with atomic number 27 and atomic
resistance. Coals contain paramagnetic centers, mass 58.9332. The naturally occurring stable iso-
with concentration reaching 1020 spin/g. tope 59 Co is very close to 100% abundant. The
COERCIVE FORCE 183
most important artificial isotope is radioactive 0.36·109 m3 /C. The absolute thermal electro-
60 Co with a half-life of 5.24 years. The outer motive force coefficient equals 18.5·10−6 V/K.
shell electronic configuration is 3d 7 4s 2 . Succes- The optical reflectivity for 5.0 μm wavelength is
sive ionization energies are 7.865, 17.06, and 92.9%. The work function for a polycrystal equals
33.50 eV; electronegativity is 1.7. Atomic radius 4.41 eV. Cobalt is a ferromagnet with Curie point
equals 0.125 nm; the Co2+ ionic radius equals T C = 1393 K. The nuclear magnetic moment
0.072 nm, that of Co3+ is 0.063 nm. The oxida- of 59 Co equals 4.639 nuclear magnetons, and
tion state is +2, +3, rarely, +1 and +4. the atomic magnetic moment equals 1.715 Bohr
Cobalt in the free state is a metal of silver-steel magnetons. The thermal neutron absorption cross-
color. It exists in two polymorphic configurations: section is 34.8 barn.
α-Co with a hexagonal close-packed lattice (space Cobalt finds its use as a component of hard
4 ), a = 0.2507 nm and c = magnetic, corrosion resistant and high-temperature
group P 63 /mmc (D6h
alloys and coatings, as well for glass dyeing; the
0.40686 nm at 298 K), and β-Co with a face-
intermetallic compound SmCo5 is used for the fab-
centered cubic lattice (space group F m3m (Oh5 ),
rication of permanent magnets (see Samarium–
a = 0.3548 nm). The temperature of the start of
cobalt magnet), and radioactive 60 Co is employed
the transformation of α-Co into β-Co by heating
in medicine as a source of γ -radiation.
is 750 K; the temperature of the start of the reverse
transformation by cooling is 676 K. The density COERCITIMETER
of α-Co is 8,183.6 g/cm3 (at 293 K), that of Device for measuring the coercive force or co-
β-Co is 8,739 g/cm3 (at 298 K), the binding en- ercivity Bc of magnetic materials. The most com-
ergy is −4.387 eV/atom at 0 K and −4.4 eV/atom mon types are coercitimeters for coercive force
at room temperature; T melting = 1768 K, measurement by magnetization MBc or by induc-
T boiling = 3203 K. The heat of fusion is tion BBc . For an MBc measurement the sample is
16.2 kJ/mol. The specific heat is 414 J·kg−1 ·K−1 magnetized to saturation and then demagnetized
at standard pressure and temperature. The De- up to the point of zero magnetization, and the
bye temperature is 385 K. The coefficient of lin- magnetic field strength that brings this about cor-
ear thermal expansion parallel to the c axis is responds to MBc . For measuring Bc by induction
14.62·10−6 K−1 , and perpendicular to the c axis, the sample is made part of a closed magnetic cir-
it is 10.96·10−6 K−1 . The thermal conductivity is cuit, and the strength of the demagnetization field
142 W·m−1 ·K−1 at 163 K and 70.9 W·m−1 ·K−1 at which the induction B in the sample goes to
at 290 K; the adiabatic elastic rigidity coeffi- zero is accepted as the value Bc (i.e., BBc ). For
cients of α-Co crystal are: c11 = 306.28, c12 = small values of the coercive force MBc ≈ BBc . See
Magnetometry, Magnetic structure analysis.
165.12, c13 = 101.9, c44 = 75.33, and c33 =
357.4 GPa at 298 K. Young’s modulus of α-Co
COERCIVE FORCE, coercive e ld, coercivity
is 220.9 GPa, that of β-Co is 199.7 GPa; the
(fr. Lat. coercitio, con nement)
modulus of rigidity of α-Co is 84.5 GPa, that of Value of an applied magnetic field Bc which re-
β-Co is 74.7 GPa; Poisson ratio of α-Co is 0.307, duces the magnetization of a magnetized magnetic
and that of β-Co is 0.337 (at 298 K). The ul- substance to zero. In many magnetic materials
timate tensile strength is 260 MPa (at 298 K), the magnetization process is not reversible, but
and the relative extension is 5% (at 293 K). The takes place around a hysteresis loop. An applied
Brinell hardness is 124 MPa for cobalt subjected magnetic field Bm produces a state of magneti-
to 1.2–3 GPa treatment. The self-diffusion coeffi- zation Mm in the material. The removal of the
cient equals 5.91·10−26 m2 /s. The Sommerfeld applied field does not reduce the magnetization
low-temperature electronic specific heat coeffi- to zero, but rather to the remanent magnetization
cient (linear term) equals 4.73 mJ·mol−1 ·K−2 . value Mrem . A reverse direction field −Bc must
Electrical conductivity is 55.7 n·m (at 293 K); be applied to reduce the magnetization to zero.
the temperature coefficient of electric resistance Increasing the reversed applied field to −Bm in-
equals 604·10−5 K−1 . The Hall coefficient is duces the reversed magnetization −Mm , and most
184 COHERENCE
of a hysteresis loop is thereby traced out. When latter appear in a superconductor under the effect
Bm is small the coercive field Bc is also small, of an external perturbation with corresponding
where of course Bc < Bm , and increases in Bm matrix elements B0 (k , s | k, s) for the single-
cause Bc to increase. Very large applied fields Bm electron transitions k, s → k , s in the normal
induce a saturation magnetization Ms , so further state of a metal: B1,2 = F1,2 B0 .
increases in Bm have no effect. The coercive field According to the BCS theory,
Bc measured at saturation is called the coerciv-
ity, and the associated hysteresis loop is called an F1 (k, k ) = Uk Uk ± Vk Vk ,
extremal hysteresis loop. The coercivity of vari- F2 (k, k ) = Uk Vk ± Uk Vk ,
ous materials varies over a very wide range from
10−4 mT to 10 T. (Some authors define Hc = where the Uk , Vk are the Bogolyubov transfor-
Bc /μ0 , where 1 mT ∼ = 795.8 A/m.) The coer- mation coefficients (see Canonical transformation
civity arises from the back magnetization process, method), with the sign determined by the nature of
and involves the magnetic anisotropy energy, the the external perturbation.
magnetostriction, and the saturation magnetiza- Coherence factors are needed for theoretical
tion. In a particular material the coercivity can dif- descriptions of the different kinetics in supercon-
fer considerably depending on its structural state; ductors associated with transitions in the system
it depends also on the direction in anisotropic ma- of Bogolyubov quasi-particles. This provides, e.g.,
terials, on the temperature, and on the presence in the BCS theory, an explanation of the differ-
of external stresses. The extremal coercivity value ence between the type of temperature dependence
for a given material equals its anisotropy field and of the nuclear spin relaxation time and that of the
can be achieved, e.g., in fine monocrystal ferro- coefficient of ultrasound adsorption in a supercon-
magnetic single-domain particles, in which back ductor.
magnetization is brought about by irreversible ro-
tation of the spontaneous magnetization vector. COHERENCE LENGTH in superconductors
High coercivity values can be obtained in close to Length scale for the variation of the order
perfect crystals where, near saturation, the forma- parameter (see Ginzburg–Landau theory of super-
tion and growth of back magnetization nuclei are conductivity)
impeded. In most ferromagnets the coercivity is ξ0 vF
determined by the critical field for irreversible dis- ξ(T ) = √ , ξ0 = ,
placement of magnetic domain walls, and this can 2(1 − T /Tc )1/2 πΔ0
be impeded by various inhomogeneities such as where v F is the Fermi velocity, Δ0 = E g /2 is
internal stress gradients, foreign inclusions, struc- the superconducting energy gap at T = 0, and
tural defects, and so on. T c is the critical temperature of the second-order
phase transition from the normal to the supercon-
COHERENCE ducting state. According to the scale invariance
See Macroscopic quantum coherence, Struc- hypothesis, at the point T = T c the coherence
tural coherence. length ξ goes to infinity as ξ(T ) ∼ (1 − T /T c )−ν .
The critical index is ν = 1/2 in the self-consistent
COHERENCE FACTORS e ld approximation. The coherence length also de-
Renormalization coefficients F 1,2 (k, k ) ap- termines the effective size of Cooper pairs with
pearing in the Bardeen–Cooper–Schrieffer the- the binding energy 2Δ. In pure superconductors
ory (BCS theory) which connect the matrix ele- with T c ≈ 1–10 K, for T T c , ξ is equal to
ments B1 (k , s | k, s) for single-particle transi- ξ0 ≈ 10−5 –10−4 cm in order of magnitude. In
tions k, s → k , s in the system of non-interacting “dirty” superconductors with short electron mean
Bogolyubov quasi-particles (see Bogolyubov meth- free paths l ξ(T ) the effective coherence length
od), and the matrix elements B2 (k , s | −k, −s) is ξ eff ∼
= (ξ l)1/2 . In highly disordered supercon-
for the processes of the annihilation of quasi- ducting alloys (compounds) and in amorphous
particle pairs in the states k , s and −k, −s. The superconductors ξ 10−6 cm, which provides
COHERENT BREMSSTRAHLUNG 185
COHERENT BREMSSTRAHLUNG
A type of Bremsstrahlung (braking radiation)
specific for monocrystals. The Bremsstrahlung of
charged particles in single crystals at high par-
ticle energies effectively differs from that of an
isolated atom. For relativistic particles (total en-
ergy E far exceeds rest energy, Lorentz factor
γ = [1 − (v/c)2 ]−1/2 E/(m0 c2 )
1), there is
a formation zone one radiation length l long in Intensity I of Bremsstrahlung of electrons in a diamond
which the Bremsstrahlung process takes place single crystal as a function of the energy ω of the emit-
which can substantially exceed an atomic size, ted quanta at the incidence angle θ = 4.6 mrad relative to
i.e. the creation of a radiated photon takes a long the (110) axis (an electron beam of initial energy 1 GeV
enough a time so that a particle can traverse a is directed along the (001) plane); shown are theoretical
macroscopic distance. The time of photon genera- calculations (solid lines) and experimental data (dots).
tion is long compared with the duration of a parti-
cle’s impact upon a nucleus, therefore the process
Bremsstrahlung emission at the second atom will
characteristics are determined mainly by interfer- occur at the time t0 + a/(vθ). As the first emitted
ence during the radiation process, rather than by quantum reaches the second atom at the moment
the nature of the collision with an individual nu- t0 + a/(cθ), the phase difference between the two
cleus. The formation zone in monocrystals may be must be equal to 2πn (where n is an integer) for
substantially longer than the separations between the interference to amplify the radiation:
the atoms of the crystal lattice. In this case all the
atoms in the zone take part in the process of coher- a 1 1
ω − = 2πn.
ent radiation, with the Bremsstrahlung intensity in θ v c
monocrystal increasing, relative to that of an an
With a fixed particle energy, E, and a fixed mag-
amorphous substance, proportional to the number
nitude of quantum energy, ω, there is a particular
of atoms located in this zone:
angle θmax at which constructive interference is
l 2E(E − ω) observed:
N eff ∝ ∝ .
a m2 c3 ωa a
θmax = ,
The presence of interference in the monocrystal 2πnl
Bremsstrahlung arises from simple kinematic con- where l is the radiation length. There may be
siderations. Suppose a particle entering a crystal different numbers of lattice sites inside the Brems-
at a small angle, θ , relative to a crystallographic strahlung region, depending on the angle of in-
axis encounters atoms along its path spaced a dis- cidence of the charged particle relative to the
tance a/θ apart (a is the lattice constant), and crystallographic planes or axes, which results in
subsequent acts of Bremsstrahlung emission take an appearance of maxima and minima in the
place. Suppose the energy of the radiation of Bremsstrahlung spectrum presented as a func-
quanta is small, and the radiation only weakly tion of the angle at which fast particles en-
affects the particle motion (its momentum is vir- ter the monocrystal. The intensity of coherent
tually unchanged at radiation). The quanta are Bremsstrahlung in the Born approximation is pro-
emitted mainly in the forward direction of the portional to the square of the number of atoms
particle motion (within a cone whose angle is in- within the radiation length. If the Born approxi-
versely proportional to the Lorentz factor γ ). If mation is inapplicable, the dependence of coherent
t0 is the time of emission at the first atom, the emission intensity I on the number of atoms
186 COHERENT CHANNELING
within the formation zone is weaker, having a This implicit equation defines σ (E) ≡ k (E),
logarithmic dependence (see Fig.). The coherent and permits the determination of the coherent
Bremsstrahlung exhibits a rather high degree of potential as a function of a spectral variable z.
polarization, and this results in intense beams of This method is believed to provide one of the
polarized γ -quanta being produced by electron ac- best single-site approximations for obtaining the
celerators. spectral characteristics of disordered systems. The
small parameter of the method for the calculation
COHERENT CHANNELING of the energy spectrum is the quantity (a/R0 )3 ,
See Okorokov effect. where a is a lattice constant, and R0 is a char-
acteristic damping or jump length of the hopping
COHERENT CONDENSATE integral. A similar small parameter appears when
See Bose condensate. computing corrections to the expression for the
electrical conductivity of a disordered binary al-
COHERENT POTENTIAL METHOD (P. Soven, loy within the formalism of the coherent potential
D. Taylor, 1967) method. The physical sense of this parameter can
An analogue of the mean field method used for be explained as follows: if a sphere of radius R0
computing the electronic structure and electronic encloses a large number of sites then one can ne-
properties of disordered alloys (see Disordered glect the fluctuations of the number of atoms of
solids) based on the formalism of multiple scat- both types A and B inside this sphere.
tering theory. The applicability of this method is
not related to limitations on the magnitude of the COHERENT PRECIPITATES
impurity scattering potential, or of the alloy com- New phases formed at phase boundaries which
ponent ratios. A scattering matrix (t -matrix) is display a structural coherence there. Coherent
assigned to each individual atom in this approach. precipitates arise as a result of the decomposi-
Systems of atoms are described on the basis of tion of supersaturated solid solutions (see Alloy
the scattering characteristics of individual atoms decomposition), or from other phase transitions:
using a Green’s function that depends upon the po- a martensitic transformation, a eutectic (see Eu-
sitions of the scattering atoms. It is assumed that tectic alloys), etc. Coherent precipitates may be
the energy spectrum of the ensemble as a whole wholly or partially coherent, depending on the
can be “macroscopically” described with the mean type of structural coherence. They form during
Green’s function the early stages of phase transformations; later on,
−1
the structural coherence will be destroyed due to
G(z) = z − ε(k) − (z) , structural dislocations which form at the phase
k interface to make the precipitates non-coherent.
The appearance of coherent precipitates with an
where k (z) is a self-energy term (see Mass op- appreciable mismatch of crystal lattice constants
erator of quantum mechanics), and ε(k) is the δ = (a p − a m )/a m , where a p and a m are lattice
dispersion law for an electron in an unperturbed constants of the precipitate and matrix, respec-
translationally-invariant crystal. To complete the tively, gives rise to coherent elastic stresses, and to
Green’s function definition, the system is pre- an increase of the elastic energy of the system by
sumed to behave as if one and the same coherent a magnitude εe ∝ δ 2 . That results in an increase
potential, σ (z), is assigned to each lattice site, of the strength of an alloy, of its coercive force,
and an effective potential, wn = wn − σ (E), is and of the magnetostatic energy (see Magnetosta-
introduced for every site to replace the random tics). In aging alloys (see Alloy aging), the elastic
potential energy of an electron wn (see Random interaction of coherent precipitates may produce
potential). The self-consistency condition is taken modulated structures. Coherence disruption, and
as follows: an appearance of non-coherent precipitates, will
n
w usually result in deterioration of the alloy strength
tn = = 0. characteristics.
1−wn n|G(z)|n
COLD BRITTLENESS 187
COHERENT SCATTERING REGIONS the work required to move far apart the units of
Largest regions of a crystal from which elastic a bound structure. It can be experimentally found
scattering of radiation (X-rays, neutrons, etc.) by by evaporation or sublimation. The largest cohe-
individual atoms or unit cells is coherent. The di- sion energy values are found in condensed phases
mension D of a coherent scattering region can be (liquids and solids) where the chemical bonds are
defined as the distance at which the phase shift of mainly ionic, covalent or metallic in character. The
the waves scattered by atoms (cells) reaches π as cohesion energy in real gases is several orders of
compared to an ideal crystal. The value of D de- magnitude smaller.
pends on the reflection indices hkl of the planes. The cohesion strength depends on the mutual
In the case of an ideal mosaic crystal, D coin- orientations of the crystal lattices of the bodies in
cides with the size l of mosaic blocks. Provided contact, on their geometrical perfection, and on
the disorientation of adjacent blocks is ψ
λ/ l the degree of their surface cleanliness. The co-
(λ is the radiation wavelength), the value of D hesion strength is calculated from the expression
may be estimated from the width δθ of the X-ray P = (α10 + α20 − α12 )/a where α10 and α20 are
line at its half-height: D ≈ λ/(
θ · sin θ B ), where the specific energies of the surfaces in contact, α12
θ B is the Bragg angle. In the opposite case (when is the specific energy of the grain boundary in-
ψ λ/ l), the regions of coherent scattering may terface, and a is an interatomic distance. If the
cover several blocks (D ∝ λ/ϕ > l). The blocks of contacting surfaces are aligned relative to the sym-
the mosaics do not exclusively determine the range metry elements then α10 = α20 = α, α12 = 0, and
of coherent scattering in a real crystal. Stacking Pmax = 2α/a. The actual strength is smaller than
faults (leading to an effective reduction of the co- Pmax due to the lack of smoothness of an actual
herent scattering regions), small-angle boundaries, contact. Showing up in the cohesion linkage zone
degree of uniform bending of crystallites, and in- are stresses arising from mismatches in the crystal
ternal microstresses in crystallites also play roles. geometry of the surfaces in contact.
To separate the effect of mosaics from that of mi-
crostresses, one may apply harmonic analysis to
COINCIDENT-SITE LATTICE
calculate the normalized Fourier coefficients An
An auxiliary lattice considered in the descrip-
in the reflection curve (the profile of the X-ray
tion of particular grain boundaries. A coincident-
line), P (2θ). The values of An can be written
site lattice is that sublattice of the complete super-
as the product An = An An , where An , An are
position lattice whose sites are occupied by atoms
the contributions from regions of coherent scat-
of both lattices of a bicrystal, i.e. the sites coin-
tering and of microstresses, respectively. In the
cide in each of the crystals that are separated by
xyz coordinate system, with xy planes coinciding
the boundary. The reciprocal relative fraction of
with reflecting planes, one may define the depen-
dence of An on the reflection index 00l and then, coincident sites is designated by Σ . The symme-
extrapolating An (l) to l = 0, find An = An (0). try of the coincident-site lattice is determined by
The derivatives of An characterize the average the symmetry of the pair of conjugated crystals,
size D0 of the coherent scattering region, and the by the axis of their relative orientation, and by
distribution Pn of the regions of coherent scat- the value of the corresponding rotation angle. See
tering transverse and along the z axis direction: Grain boundary, Grain boundary lattice.
−(dAn /dn)n=0 = c/D0 (c is the unit period along
the z axis), d2 An /dn2 = Pn . COLD BRITTLENESS
The change of the failure mechanism from
COHESION (fr. Lat. cohaesus , stuck) tough failure to brittle failure (or quasi-brittle fail-
Linkage (binding) between two solid bodies ure), which occurs with a decrease of temperature
of similar chemical composition in contact with within a comparatively narrow temperature range.
each other. It results from a balance between long- This change of failure mechanism involves abrupt
range repulsive and short-range attractive forces. reduction of plasticity-related characteristics and
The cohesion energy is an important characteris- strength. According to the Ioffe theory, cold brit-
tic associated with chemical bonds. It is defined as tleness is the consequence of an abrupt increase of
188 COLD CATHODE
the yield limit σs with decreasing temperature, the films, semiconductor films, p–n junctions biased
temperature dependence of the breaking stress σf in the gate direction (see Semiconductor junc-
being comparatively weak. Cold brittleness oc- tion), transistor structures, M–I–S structures (see
curs below the temperature at which the curves of Metal–insulator–semiconductor structures), M–S
σs (T ) and σf (T ) versus T intersect. Modern mi- structures (M is metal, I is insulator, S is semicon-
croscopic dislocation theories attribute the sharp ductor), semiconductor structures with a negative
temperature dependence of the yield point to the electron af nity, etc.
presence of a covalent component in the inter-
atomic bond (see Covalent bond), and to the high COLD ELECTRONS
value of the Peierls–Nabarro stress (see Peierls– An electron gas in the conduction band of
Nabarro model) at the plastic deformation of co- semiconductors (holes in the valence band) with
valent crystals. Hence the cold brittleness arises an average energy ε appreciably less than the
due to the abrupt decrease of the mobility of dis- thermodynamic equilibrium value ε0 . This phe-
locations with decreased temperature, and to the nomenon of electron cooling is due to various
decrease of the rate of stress relaxation at the mechanisms related to the motion of the current
apexes of cracks and at other stress concentrators carriers in an electric field, or to optical transi-
(see Stress concentration). tions of the charge carriers. For instance, elec-
The reciprocal of cold brittleness is given the tron cooling will occur if the overall current in
name cold reluctance. This term is descriptive of a semiconductor is directed opposite to the in-
the ability of materials to retain plasticity and fail- ternal electric field in a space charge region, or
ure toughness with decreasing temperature. because of another inhomogeneity of the semi-
The susceptibility of a material to cold brit- conductor. In electric fields, where the current is
tleness is determined by the cold brittleness tem- directed along the field, a cooling of the carriers
perature (or cold brittleness threshold) Tc . Upper is possible when the dominant scattering of the
and the lower cold brittleness points (Tcb u and T l ) electron energy occurs by optical phonons, and
cb the momentum scattering is by ionized impurities
are commonly distinguished. The lower point Tcb l
under conditions when the electron–electron inter-
is the minimal deformation temperature at which action is negligible. While acted on by an electric
plastic deformation precedes the failure, and the field an electron is able to emit an optical phonon,
failure is totally tough in its nature. The value of and thereby diminish its energy (i.e. cool), and the
u depends heavily on the deformation rate ε̇,
Tcb scattering by ionized impurities promotes the ac-
increasing with its increase. Various methods of cumulation of such electrons. Hence it is possible
hardening of materials (raising the dislocation den- to cool carriers by light through the agency of di-
sity): solid-solution hardening, dispersion hard- rect band-to-band transitions. For this purpose it is
ening (see Precipitation-hardened materials), etc. necessary (when the energy of scattering via inter-
result in an increase of the value of Tcb l . Only a
electron collisions is negligible compared to such
reduction of the crystal grain size or the formation scattering of photocarriers by optical phonons) to
of a disoriented dislocation cellular structure can excite the photoelectrons with a fixed initial en-
lead to the reduction of Tcb and the improvement ergy εi greater than ε0 by utilizing light of such
of strength properties. As the failure mechanism a wavelength that, after the emission of n optical
changes from transcrystalline failure to intercrys- phonons, the photocarriers should have a final en-
talline failure, the cold brittleness temperature may ergy εf < εi .
abruptly increase.
COLD WORKING
COLD CATHODE Industrial term no longer used much. It means
Unheated cathode emitting electrons by means hardening of a metal under the action of cold plas-
of eld emission of electrons. These are e ld- tic deformation, but not strain hardening (see also
emission cathodes, and there are also various Work hardening). Cold working in optimal mode
types of cathodes emitting nonequilibrium elec- increases hardness and yield limit, resistance to fa-
trons excited by an internal field: dispersed metal tigue deformations, and a decrease of plasticity and
COLLECTIVE EXCITATIONS IN SUPERCONDUCTORS 189
superconductor, along with currents of normal ex- sity of the number of particles, density of electric
citations and of the superconducting condensate charge, spin density, etc. In the simplest case of a
directed opposite to one another, so that there is single-component system, the collective variables
no net current. The phase velocity of such ex- will be defined by means of the Fourier compo-
citations equals: c = [2ΔDχ(2πT τ )]1/2 , where nents of the operator of the number density of
D = v F l/3 is the electron diffusivity (diffusion particles:
coefficient), Δ is the energy gap in the supercon-
1
N
ductor, τ is a characteristic time of scattering from
k = √
ρ exp(ik·r j ), k = 0, (1)
impurities, T is the temperature, χ is the Gorkov N j =1
function introduced in the theory of superconduct-
ing alloys, v F is the Fermi velocity, and l is the where r j are the coordinates of the j th particle,
electron mean free path. and N is their total number. The condition
√ k = 0
Another type of oscillation existing in super- excludes the constant term ρ k=0 = N δk,0 . The
conductors is an oscillation of the order parame- transition to an expanded phase space that includes
ter amplitude which has a resonance character individual coordinates, {r} = (r 1 , . . . , r N ), and
(Schmid mode, A. Schmid, 1975). This is as- collective variables, {ρk }, is effected via a trans-
sociated with the possibility of a virtual decay fer function which is a product of corresponding
of Cooper pairs, which requires an energy input δ-functions. Hence, e.g., the system configuration
of 2Δ. This process involves violation of the en- integral can be represented as follows:
ergy conservation law and, in accordance with the
quantum-mechanical uncertainty principle relat- QN = Q0N (dρ)J (ρ) exp −βθ(ρ) , (2)
ing energy and time, must proceed for times of
the order of /(2Δ). The amplitude of the order where Q0N is the configuration integral of the sub-
parameter perturbation oscillates with a frequency
system with short-range interactions, the function
ω = 2Δ/, and tends toward its steady-state value
with a power law dependence on the time. In- 1
J (ρ) = (dr) exp −βU0 (ρ)
elastic scattering mechanisms can bring about an V N Q0N
exponential decay of such oscillations.
In a hypothetical superconductor treated as a × δ(ρk − ρk ), (3)
superfluid Fermi liquid there are weakly damped k=0
oscillations of the phase of the order parame-
is the Jacobian of the transition to collective coor-
ter which have an acoustic spectrum (Anderson–
dinates, the differential factors have the form
Bogolyubov mode). However, an excitation of the
electron density accompanies these oscillations,
N
hence, their frequency is increased up toward the (dr) = dr j , (dρ) = dρk ,
plasma frequency in real superconductors because j =1 k=0
of Coulomb interactions.
where β = 1/(k B T ) is the reciprocal temperature
COLLECTIVE VARIABLES METHOD of the system, U0 (r) is the potential energy of
A method of statistical physics that intro- the short-range interactions, and the function θ(ρ),
duces long-range interactions into the thermody- given by
namic and structural properties of multiparticle N N2
systems. The description of interacting particles θ(ρ) = ν(k)(ρk ρ−k − 1) + ν(k = 0),
V 2V
(e.g., atoms, ions or molecules) using collective k=0
variables is based on an extended phase space (4)
consisting of individual particle coordinates and is the representation in collective variables of the
collective variables associated with the waves of potential energy of long-range
pair inter-particle
density fluctuations of a certain system parameter interactions, ν(k) = dr e−irk Φ(r) is a Fourier
(see Fluctuations of atomic positions). Depending component of the long-range potential Φ(r), and
on the specific problem, the latter may be the den- V is the system volume.
COLOR CENTERS 191
A similar transformation to collective variables interstitial atoms. The primary relaxation of a cas-
also can be brought about by means of an integral cade takes ∼10−13 –10−11 s, with some intersti-
transform of the Gaussian factor exp[−βθ(ρ)]. tial atoms returning to their initial sites. Then the
The method of collective variables is used to secondary stage of relaxation occurs, controlled
describe the properties of a variety of real systems. by thermally activated processes (primarily by dif-
Thus, an ion-molecular approach in the theory of fusion). On irradiation with dense beams or with
electrolytic solutions, based on an equivalent ac- radioactive decay, spatially uniform collision cas-
count of the interactions of all the particles in solu- cades can arise, their only common quantitative
tion, has been developed within the framework of characteristic being the cascade function or the ra-
collective variables. This method, supplemented dioactive damage function ν(ε). It equals the total
by the density matrix shift transformation, has number of atoms (ions) displaced from the sites
been successfully employed in the theory of quan- of the solid body lattice by the initial atom with
tum Bose- and Fermi-systems, in particular for energy ε.
constructing a theory of liquid 4 He and describ- Low-energy cascade: ε < εc , where εc is the
ing the electron gas in metals (see Bose–Einstein critical energy, with negligible retardation of the
statistics and Fermi–Dirac statistics). moving atom by electrons below it. All the energy
Variables associated with the order parameter of the atom goes to displace atoms from lattice
of the system have been introduced into the set sites, so that
of collective variables in the theory of second- αε
order phase transitions, and the explicit form of ν(ε) = (ε < εc ),
εd
the transfer Jacobian, J (ρ), is of a non-Gaussian
character for three-dimensional systems. In an- where εd is the energy for creating a stable
other approach, the collective motions of systems vacancy–interstitial atom pair, and α is a factor
of interacting particles are treated by the canoni- on the order of unity. As commonly assumed,
cal transformation method, whereby a transition is εd ∼ 25–80 eV, α ∼ 0.4–1. Theoretical values
made to collective coordinates and momenta that of εc are 0.1 MeV for Fe and 3 MeV for U .
correspond to collective oscillations, or to the ele- For a high-energy cascade: ε > ε c . The retar-
mental excitations associated with them. Since the dation of atoms down to energy εc occurs mainly
collective coordinates and momenta are not con- due to electrons of the medium, i.e. without dis-
ventional variables, a transition to them involves lodgment of new atoms. Therefore,
additional conditions that conserve the total num- α εc
ber of degrees of freedom in the system. ν(ε) = (ε > εc ),
εd
COLLISION CASCADE where α is a factor on the order of unity (it can
The avalanche-like process of dislodging (dis- differ somewhat from α). See also Seeger zone.
placement of) atoms (ions) from the sites of a
crystal lattice (or of an amorphous solid). The col- COLLISION IONIZATION
lision cascade arises after energy transfer to the See Impact ionization.
target atom by an external particle as a result of
sequential collisions of atoms in the vicinity of the COLLISION RADIATION
initially displaced one. See Bremsstrahlung.
If the energy of the initially displaced atom
is ε, the collision cascade occupies the spatial do- COLOR CENTERS, F -centers
main of size l(ε), where l(ε) is the mean free Point defects with optical absorption in the
path as a function of its energy. The cascade transparent region of a defect-free crystal. Color
time is 10−14 –10−13 s; where the cascade over centers are related to either impurities (impurity
a domain of enhanced concentration of radiative color centers), or intrinsic structural defects in
defects arises with its central part enriched with crystals (intrinsic color centers). The main meth-
vacancies, and the peripheral one enriched with ods to stimulate the formation of color centers are
192 COLORED SYMMETRY
COLORED SYMMETRY
A theory of generalized symmetry of material
objects (media and structures), which combines
the global symmetry of the system as a whole
Color centers.
with the local symmetry of its structural elements.
Depending on the method of combination of the
doping, deviation from stoichiometric composition, groups of global (external) G- and local (internal)
and irradiation by ionizing radiation. Electronic P -symmetries, the following colored symmetry
transitions that determine the color center ab- groups are distinguished: (1) P -symmetry groups,
sorption are shown schematically in the figure. which are subgroups G(p) ⊆ P × G of the di-
Transitions 2 and 3 produce, as a rule, wide ab- rect products of groups P and G; (2) Q-symmetry
sorption bands; transitions 1 give either narrow groups, which are subgroups G(q) ⊆ P λG of the
or broad bands, depending on the magnitude of semidirect products of P and G; (3) Wp -sym-
the electron–phonon interaction. Transitions 4 are metry groups G(Wp ) ⊆ Pwr G = P G × G, which
observed only under high-power pulse excitation are subgroups of the wreath product (for groups
with the aid of time-resolved spectroscopy. Re-
G(Wq ) – of twisted wreath product) of P and G,
verse transitions may proceed with photon emis-
where P G is the direct product of groups P , taken
sion; in this case a color center is simultaneously
nG times, n = |G| is the order (or power) of the
a luminescent center. Many color centers involve
group G: P G = P g1 × P g2 × · · · × P gn , and the
a point defect that has trapped an electron (elec-
elements gi ∈ G indicate the position of each indi-
tron color centers) or a hole (hole color centers).
vidual group P in the n-fold direct product. These
The association of elementary defects (vacancies,
groups differ in the manner of their effect on a
interstitial atoms, impurity atoms, or impurity ions)
results in the formation of aggregate color centers. mass point, and in the multiplication law of their
An increase in temperature and illumination in the elements.
absorption bands of color centers may lead to the In terms of physical interpretation, the low-
release of trapped electrons or holes, i.e. to the de- est groups G(p) and G(q) , isomorphic to the
struction (discoloration) of color centers. The most groups G, describe the magnetic exchange sym-
widely studied color centers are in ionic crystals, metry of the structures of magnetically ordered
in particular, in alkali-halide crystals. For example, crystals. Groups G(Wp ) and G(Wq ) ↔ G, involv-
F -centers (anionic vacancy with a trapped elec- ing the effect of translation on a mass point (de-
tron, named from the German word “Farbe” for pending on its coordinate), describe the symmetry
“color”) produce an intense absorption band in of real crystals, including quasi-crystals (icosahe-
the visible region, which corresponds to a transi- dral phases), and polysystem molecular crystals
tion similar to the 1s–2p transition in a hydrogen with so-called “super-Fedorov” symmetry.
atom (F -band), and weaker K-, L-bands refer to
transitions to higher F -center excited states. Also COLORED SYMMETRY GROUP
H -centers (interstitial atoms of halide X0 ), and In the general case the subgroups of the di-
Vk -centers (self-localized holes X− 2 ), produce
rect products (P × G) and the semidirect products
broad bands in the ultraviolet region. In covalent (P λG) of groups of external (global) G and in-
crystals, the absorption of color centers typically ternal (local) P symmetries, as well as subgroups
occurs as narrow lines, e.g., GR1 and ND1 cen- of the wreath product Pwr G = P G λG, which op-
ters in diamond (neutral and negatively charged erate over subspaces of geometrical coordinates
COMMENSURATE STRUCTURE 193
transition. In a number of cases the system un- COMPENSATION RULE, Meyer–Neldel rule
dergoes a sequence of phase transitions to various (W. Meyer, H. Neldel)
commensurate states (with sequential variations of A relationship between the activation energy
the superlattice constant), until reaching the final
E and the preexponential factor σ0 in the tem-
commensurate phase which has a period a that perature dependence of the electrical conductivity
is a multiple of the initial lattice constant a, i.e σ of semiconductor:
a = na, where n is an integer (so-called “devil’s
E
staircase”). σ (T ) = σ0 exp − . (1)
2kB T
The transition from the initial structure to a
commensurate one proceeds in certain materials The analysis of many experimental data obtained
through an intermediate temperature incommen- for semiconductors possessing current carriers
surate structure, with a superstructure constant
with small mobilities has shown that
incommensurate with the lattice constant of the lg σ0 = α
E + β. (2)
initial phase (see incommensurate–commensurate
phase transition). The commensurate structure Here α and β are parameters characteristic of the
may be represented as modulated by shear waves, material. Eq. (2) represents the compensation rule
e.g., charge density waves (spin density waves), which has been widely investigated for polymeric
in one, two or three directions; with the mod- semiconductors. This is explained with the help of
ulation period determining the displacements of the charge transport tunnel model (see Tunnel ef-
atoms from their initial equilibrium positions. The fect).
periodicity of the new structure is fully commen-
surate with that of the initial structure: Q = nq, COMPLEX ELASTIC MODULI
where Q and q are wave vectors of the com- Tensor factors in Hooke’s law written for har-
mensurate and the original structures, respectively, monically time-varying stresses and strains in dis-
ω =
sipative media (see Dissipative structures): σik
and n is an integer. The translational symmetry is
ω , where σ ω and ε ω are Fourier com-
ciklm (ω)εlm
not violated in the generation of the commensu-
rate structure. The commensurate structure can be ponents of stresses and strains, respectively. The
detected by diffraction methods, e.g., through an imaginary part of the complex elastic modulus,
electron diffraction observation of extra reflections called the loss modulus, is responsible for the
corresponding to basic structure reflections for damping of elastic waves in a medium.
multiple distances, and by changes of other physi-
cal characteristics: electrical resistance, magnetic COMPLEXY
susceptibility, Young’s modulus, etc. The commen- A thermodynamic equilibrium defect in a solid
surate structure occurs in dichalcogenides of tran- solution consisting of an impurity atom complex
located at a group of lattice sites whose num-
sition elements MX2 (M = Ta, Nb; X = Se, Te,
ber differs from the number of complexy atoms.
S), in crystals of the A15 and cubic Cu2 Mg types,
The concept of complexy was introduced by
in the premartensitic region of temperatures in a
M.A. Krivoglaz (1970). It can be energetically ad-
number of compounds based on equiatomic TiNi,
vantageous to form a complexy from atoms with
in Cr, etc.
strongly differentiated atomic radii if this results in
a partial removal of stresses and a decrease of the
COMPACTION overall elastic energy. A complexy is less advan-
See Pressing. tageous from the entropy standpoint than isolated
impurity atoms, and the complexy concentration
in a solid solution decreases with temperature in
COMPENSATION contrast to the behaviour of point defects. Com-
See Conductivity self-compensation, Ionic con- plexy formation radically affects the temperature
ductivity compensation, Magnetic compensation. and concentration dependences of crystal lattice
COMPTON EFFECT 195
photons scattering off particles with large masses. electronics, providing visualization and identifica-
If the electron involved in the Compton scatter- tion of the subsurface micro-structure and defects.
ing has a high velocity then an increase in photon See also Tomography of micro e lds.
energy is possible at scattering (inverse Compton
effect). The Compton effect gives a strong contri- COMPUTER ALGEBRA
bution to the generation of fast electrons in gamma A process in applied mathematics and com-
ray irradiated solids, and these generated electrons puter engineering, that uses computers for the
create radiation defects through impacts with tar- automation of symbolic derivations. Computer al-
get atoms. gebra is designed for the execution of cumbersome
(routine) analytic manipulations that cannot be
COMPUTATIONAL MICROTOMOGRAPHY readily carried out manually: multiplication (di-
A method of scanning an object with a probe vision) of polynomials, cancellation of similar
of penetrating radiation and, with the aid of terms, formation of an arbitrary algebraic expres-
a computer, subsequently reconstructing an im- sion with respect to a dependent variable, substi-
age which contains details of the internal mi- tution of variables, integration, differentiation, etc.
crostructure of the object in a series of thin layer This approach can facilitate and improve the accu-
cross-sections. Ordinary transmission images, be-
racy of subsequent numerical calculations, provide
ing two-dimensional shadow projections of the
analytical relationships between various parame-
actual three-dimensional object, do not give sig-
ters, and overcome difficulties involved in the
nificant information about its internal structure,
subtraction of nearly equal large magnitude num-
because they display information averaged ac-
bers.
cording to the thickness. In contrast to this the
digitized images called tomograms (see Tomogra-
COMPUTERS in solid state physics
phy) provide considerably sharper differentiation
Various methods for applying computers to
of areas with differing absorption coef cients, and
better separation of the images of overlaying struc- solve problems in the field of solid state physics.
tures. In experimental solid state physics a computer per-
To apply computational microtomography the forms the following three functions: control of
object is transilluminated with a narrow beam of apparatus (e.g., the control of devices designed
the penetrating radiation (e.g., X-rays, electrons, for molecular beam epitaxy when synthesizing
ultrasound, or IR) and the radiation that passes superlattices); data analysis (such as deciphering
through it is recorded by corresponding detec- various spectra); planning experiments. In an opti-
tors. By sequentially scanning the beam and/or mal situation, all these functions are coordinated,
turning the sample, a set of projection data neces- so that the computer performs quite a complicated
sary for reconstructing the tomogram is obtained. range of research tasks. In analytical solid state
The result is a sequence of two-dimensional half- physics most of the important applied problems, as
tone cross-section images, and the data from these well as an ever-growing share of the fundamental
cross-sections can be used to reconstruct a three- problems, are solved using computers. Two ap-
dimensional image which reflects the true mi- proaches are dominant in this field: the numerical
crostructure of the object. solution of equations, and the direct modeling of
Microtomography has been successfully used observed solid state phenomena with the help of
in transmission electron microscopy for the inves- computers (see also Computer simulation).
tigation of molecules, bacteriophages and mem- The first approach of solving systems of equa-
branes (with spacial resolutions of angstroms), tions is not specific for solid state physics, and its
in scanning electron microscopy, and in optical efficiency is determined, to a great extent, by the
and X-ray microscopy for investigating the prop- capabilities of the computer. For single-processor
erties and structure of solids with dimensions of computers the absolute or theoretical limit of their
about 1 mm and resolutions of micrometers. The productivity is about 109 operations/s, but in prac-
technique is also used in the non-destructive de- tice the limit is lower by one or two orders of mag-
fectoscopy of components of solid-state quantum nitude. For a relatively simple problem, such as a
COMPUTER SIMULATION 197
study of the initial stages of monocrystal growth to 10−9 s, which permit the solution of a large part
from an atomic beam at a perfect monocrystal of the problems of solid state physics in real time.
face, it is necessary to analyze a Markovian chain Further progress depends, in principle, on creating
consisting of hundreds or (at low saturation in the new solid and molecular modeling systems that
vicinity of the critical point) a few thousand non- implement the probability functions at the quan-
linear differential equations. The determination of tum level.
the distribution function of the atomic groupings In terms of the complexity of the modeling, the
for any given set of initial conditions requires problems of solid state physics could be classi-
the performance of about 108 operations/s. For a fied as follows: (i) crystallization and recrystalliza-
reasonably complete analysis involving the entire tion in the absence of continuous defects; (ii) the
range of available growth conditions one needs to same processes with the participation of dislo-
perform 1010 or more operations/s. The problem cations and other continuous imperfect defects;
becomes too difficult for multicomponent crystals, (iii) electronic, thermal, and diffusion (see Diffu-
and indeed insoluble for single-processor comput- sion) processes with changes in the location of in-
ers when the final growth stages are included in ternal boundaries (e.g., semiconductor junctions);
the calculation. Multiprocessor computers are also (iv) non-structural phase transitions; (v) first-
inefficient in the framework of the first approach, order structural phase transitions with alteration of
because extensive parallelism of the data process- the lattice symmetry; (vi) second-order structural
ing is hardly achievable. phase transitions; (vii) modeling processes on a
In contrast, the second approach of model- “rigid” lattice with a quantum-mechanics analysis
ing observed phenomena is quite compatible with of elementary events; (viii) subsequent quantum-
the architecture of multiprocessor computers. The mechanical analysis of processes in solids with
real space of the crystal can be divided into a changes of lattice symmetry. At present, models
network of conventional cells (hence the term of the former four types are really quite well-
network models) and a step-by-step analysis of developed, and the efficacy of the latter four types
changes in the states of cells and their interac- depends on future progress.
tions can be carried out. In the extreme case, a cell
is represented by an individual atom (e.g., Ising COMPUTER SIMULATION, computer modeling
model). The organization of this modeling system Investigation of structures and phenomena by
can be either linear, two-dimensional, or three- their information simulation on computers. Di-
dimensional. A conformity between the compu- rect simulation is achieved by an organization of
tational process and the processes occurring in elements of the computer and the topology of
solids is provided by means of both the programs communications between them such that the prop-
and the apparatus. The former are universal (be- erties of data-regulated processes are similar to
cause the program approach allows one to model the corresponding characteristics of the object. In
any physical process using any computer) but they this case, the maximum performance of the simu-
do increase the duration of each step of the model- lation process is achieved, but each structure and
ing. The latter (i.e. the apparatus part) specifies the each type of phenomena need a special machine
computer for each individual problem, and pro- (or its complicated adjustment). Alternatively sim-
vides for its rapid solution. In the extreme case, ulation is carried out by numerical treatment of
the relations between the processors of the mod- mathematical equations describing the object at
eling system are isomorphous to the relations in various boundary and initial conditions. The simi-
the modeled solid; and the productivity increases larity to the object is achieved only by means of
as the number of processors increases. There ex- software. The hardware of simulation is univer-
ist modeling systems (so-called distributed matrix sal but its efficiency (even at the maximum speed
processors, “hypercube”) consisting of 105 (in or- of the elements) is rather limited. Historically,
der of magnitude) processors. Theoretically it is the first simulation was performed by analog ma-
possible to produce such systems containing up to chines for physical processes formally described
1012 modeling processor units with a clock period by the same equations as the phenomena under
198 CONCENTRATION
study. Now there are analog machines based on can be expressed as fractions, as mass or volume
various physical phenomena (transients in electric percentages (%), or for high dilutions as, e.g., parts
circuits, optical and acoustic waves, processes in per million (ppm) or parts per billion (ppb).
solids and on their surface, etc.). They allow one Other ways for expressing the amount of a
to carry out a parallel multichannel simulation in component and the amount of a total system are:
the real-time mode of a physical process achiev- mass concentration – the relation of the compo-
ing an effective capacity up to and even more nent mass to the system volume (kg/m3 , more
than 1010 –1011 operations per second. The phys- often g/l); molar concentration (molarity) – the
ical structure and manufacturing techniques of relation of the component quantity in moles to the
such devices are relatively simple, and the energy system volume (mol/m3 , more often mol/l); mo-
consumption is minimal. However, besides the lal concentration (molality) which is the relation
specialization, their shortcoming is low accuracy of the component quantity in moles to the solvent
of calculations and instability of the simulation mass (mol/kg); normal concentration (normal-
process. Computer simulation by numerical solu- ity, n) – the number of gram-equivalents of the
tion of the equations describing a physical object substance in one liter of solution; titer – the num-
possesses quite opposite features. There are two ber grams of the component dissolved in 1 ml of
approaches to computer simulation combining the the solution. The last two quantities are common
advantages of digital and analog principles. The in analytical chemistry. Furthermore, in the litera-
first is based on multiprocessor computers (see ture concentration is sometimes expressed as mass
Computers in solid state physics). In this case, of the component being dissolved in 100 g of sol-
a similarity is reached between the organization vent, or as number of moles of the component
of flows of the internal information exchange in in 1000 moles of solvent. The SI system recom-
a computer and mass–energy flows in a modeled mends using only mass and molar concentrations.
system. The performance of computer simulation
increases by several orders; under optimal condi-
CONCENTRATION-DEPENDENT PHASE
tions, proportional to the number of processors in
TRANSITION
the system. Another approach is based on join-
A phase transition characterized by the spon-
ing digital and analog processors in one device
taneous onset of long-range order at T = 0 (see
(so-called hybrid devices) in which, at each stage
Long-range and short-range order) at a change
of simulation, the optimal distribution of data
in concentration of the repetative units forming
fluxes between digital and analog subsystems is
the crystal. Depending on the physical nature of
realized. The analog subsystem provides the maxi-
the associated order parameter that arises, such a
mum speed of analysis of special problems arising
during simulation, and the digital one provides sta- phase transition can be ferromagnetic, ferroelec-
bility and universality of simulation, and the accu- tric, structural, etc. Let us consider for definiteness
racy of the calculated data. Both approaches (digi- the ferromagnetic concentration-dependent phase
tal multiprocessing and digital–analog hybrid) are transition in a system in which magnetic atoms re-
based on the development of microelectronics and place nonmagnetic ones in a random fashion. Be-
an increased integration of computer elements. cause the exchange interaction rapidly decreases
In prospect, supermultiprocessor (up to 109 –1011 with distance it is possible to consider as a first ap-
modeling cells) hybrid devices will unite the ad- proximation that only nearest neighbor magnetic
vantages of all the above-mentioned approaches to atoms are interacting. At absolute zero tempera-
computer simulation. ture all magnetic moments coupled by exchange
are oriented parallel. If the concentration x of
CONCENTRATION magnetic atoms is small they form clusters iso-
The relative content of a component in a sys- lated from each other with dimensions which do
tem (solution, mixture, melt). There are many not vary with an increase of the crystal volume
different ways of expressing this: mass, moles, (finite clusters). The average magnetic moment
volume fraction, relation of individual mass or in- of the whole crystal is equal to zero in this case
dividual moles to the total. These concentrations since various clusters have oppositely directed
CONDENSED MATTER INTERFACE 199
densed matter interfaces. The concept of a geo- CONDENSON (fr. Lat. condensatio , densi cation,
metrical interface is used in describing the phys- thickening)
ical properties of two bodies in contact when the A self-localized state of an electron in an elas-
layer thickness is very small compared to the bulk tically deformed medium, or a hypothetical quasi-
dimensions of these bodies so it has a negligi- particle in a covalent crystal in which all atomic
ble influence on the bulk properties. A physical bonds are covalent. An electron induces dipoles
condensed matter interface is a transition layer in neutral surrounding atoms of a medium, and
of finite thickness between bulk materials whose then by its electric field attracts them. As a re-
chemical composition differs from the chemical sult of small atom displacements, an increased
composition of the media in contact, and hence it density region is formed which represents a po-
can influence their properties. One can distinguish tential well for the electron (see Electrostriction)
sharp and smooth interfaces. A transition layer be- where it is localized. A macroscopic considera-
tween contacting media with a thickness the order tion of a condenson which takes into account a
of magnitude of an interatomic distance is called strain potential shows that if self-localized conden-
sharp, and it can be obtained by the molecular epi- son states are at all possible in homopolar crystals,
taxy method. A transition layer much thicker than they will have a small radius. A macroscopic
the atomic scale is called smooth. condenson can be realized in three-dimensional
homopolar crystals with a “loose” crystal lattice
CONDENSED STATE OF MATTER
due to weak medium anharmonicities. In one-
Solid and liquid states of matter. The con-
dimensional crystals a condenson can also exist in
densed state is characterized by high matter den-
an harmonic elastic medium. A similar situation
sity in comparison with a gas, and by a rather
is realized, in particular, with the application of a
low speed of motion of its component atoms.
quantizing magnetic eld to a crystal. In crystals
The regular arrangement of atoms or molecules
with negative electron af nity energy (He, Ne) a
on a three-dimensional lattice with long-range
reciprocal type of electron self-localization is pos-
order (see Long-range and short-range order) de-
sible with the formation of a lattice disperse region
termines the crystal structure that corresponds to
or even the appearance of a small cavity inside
an equilibrium state. There is only short-range or-
which the electron cloud (|Ψ |2 ) is mainly local-
dering in the arrangement of atoms in a liquid. The
ized.
structure of amorphous solid bodies is equivalent
to that of a liquid; and from the thermodynamic
point of view the amorphous state represents a CONDON APPROXIMATION
supercooled liquid with a very long relaxation See Franck–Condon principle.
time for transition to the equilibrium crystal state.
Macroscopic liquids and amorphous materials are CONDUCTANCE, ELECTRICAL
isotropic. Materials intermediate between crystal See Electrical conductance.
and amorphous solids are polymers, liquid crys-
tals, quasi-crystals, and the gaseous crystalline CONDUCTANCE, HIGH-FREQUENCY
state. Polymers consist of periodic linear chains See High-frequency conductance.
at sites where particular atom groupings are allo-
cated. The chain molecules can be bound together CONDUCTION BAND
in different ways. The parallel arrangement of An energy band of a crystal comprising elec-
polymeric chains defines the polymer anisotropy tronic states with energies higher than those of
(see Anisotropic medium). The properties of liquid filled valence bands. Weak external effects on the
crystals are determined by the ordered orienta- conduction electrons can cause a change in their
tional arrangement of their molecules which are average linear momentum, i.e. induce the appear-
often elongated or disk like in shape, and this re- ance of current.
sults in the anisotropy of liquid crystals. There is To create current carriers in the conduction
only short-range order in the spatial arrangement band of intrinsic semiconductors and insulators it
of the molecule centers. is necessary to overcome an energy barrier, with
CONDUCTIVITY SELF-COMPENSATION 201
(high temperatures, pressures), then vacancies or from single-particle states (see Slater determi-
impurity interstitial atoms may be the compensat- nant). The matrix elements of this interaction op-
ing defects. erator were found by J.C. Slater (1929) for orthog-
onal, and by P.O. Löwdin (1955) for nonorthog-
CONDUCTIVITY, THERMAL onal single-electron functions. The configuration
interaction decreases the total energy of a fermion
See Thermal conductivity and Thermal conduc-
system by the so-called correlation energy. It is
tivity coef cie nt.
necessary to take the configuration interaction into
account for systems with uncompensated elec-
CONDUCTIVITY, THERMALLY STIMULATED tronic spin (atoms, ions, radicals, impurities with
See Thermally stimulated conductivity. partially filled d- and f -shells in crystals, etc.),
that is systems with unfilled shells. From consid-
CONDUCTORS, electrical conductors ering a set of multiparticle configurations as the
Materials with low specific electrical resistance basis for solving the multielectronic Schrödinger
(resistivity, usually below 10−6 ·m). Conductors equation, and utilizing its linearity, one arrives at
include, for example, metals, metal alloys, semi- the method of second quantization which is a more
metals (including graphite). The current carriers convenient one for treating a large number of par-
(charge carriers) in solid conductors are ordinarily ticles. Applying the configuration interaction in
conduction electrons. Typical for a conductor in solids results in the narrowing of the electronic
its normal (not superconducting) state is the linear bands obtained from solutions of the Hartree–
dependence of the current density on the strength Fock equations (see Hartree–Fock method).
of the applied electric field (Ohm’s law). Identi-
CONFOCAL DOMAIN, focal conic domain
fying conductors as a separate class of materials
A domain defect in a structural liquid crys-
can be somewhat subjective since conductivity de-
tal (smectic liquid crystals A, C, cholesteric liquid
pends on many factors, including temperature. For
crystals) which has the form of a cone of rev-
example, many (poor) conductors transform to a
olution. Because of the different orientations of
superconducting state at extremely low tempera-
molecules on the liquid crystal boundary the layers
tures.
inside the confocal domain are curved with their
shapes determined by the condition of maintain-
CONFIGURATION INTERACTION ing an invariable width of each layer. The layers
Interaction of multiparticle states (configura- with Dupin cyclides fulfill this condition, as shown
tions) described by wave functions constructed in Figs. 1 and 2. This results in the appearance in
Fig. 1. Configuration of smectic layers in a confocal domain of a general form (left) and of a toroidal form (right).
CONFORMATIONS 203
under consideration laying below the lowest adja- a full cone whose vertex is located at the point
cent first-order saddle point. of the beam incidence on the face. Thus linearly
polarized waves with continuously varying (in the
CONICAL REFRACTION OF ACOUSTIC WAVES range of 180◦ ) polarization azimuth propagate in
A special type of acoustic wave refraction on directions of cone generators. Conical refraction
a flat crystal face oriented normally to an acoustic was predicted by W. Hamilton (1832) and ob-
axis, i.e. in a direction along which a great number served by H. Lloyd (1833). Internal and external
of quasi-transverse waves with identical velocity conical refraction can be distinguished. Internal
can propagate. An exception is the acoustic axis conical refraction occurs in light beam transmis-
being coincident with a 4th or 6th order symme- sion in the binormal direction, in this case the
try axis. Conical refraction was observed for the vertex is located on the forward lamina face; on
first time during the propagation of acoustic waves emerging from the lamina the beam makes up a
in nickel crystals (J. de Klerk, M.J.P. Musgrave, hollow cylinder. External conical refraction oc-
1955) and in dihydroammonium phosphate crys- curs for light transmission in the biradial direction,
tals (K.S. Aleksandrov et al., 1963). in this case the cone vertex is located on the out-
Interior and exterior conical refraction are dis- put lamina face. If the light beam direction does
tinguished. For the interior conical refraction the not coincide with the optical axis direction then
whole cone of acoustic beams corresponds to ordinary birefringence occurs.
one wave normal, while for the external coni-
cal refraction the whole cone of wave normals CONOSCOPY (fr. Gr. κωνoς , a cone, and
corresponds to one beam direction. The angu- σ κoπεω, am looking, am observing)
lar opening is determined by the crystal elastic A method of investigating optical properties of
moduli and can reach tens of degrees as distinct oriented anisotropic media (birefringent crystals)
from conical refraction of light waves. For exam- with the help of interference figures (so-called
ple, it is 61◦ 30 for a calcite crystal. The special conoscopy figures) observed in convergent polar-
case of conical refraction in the direction of an ized light beams. In conoscopy polarized micro-
acoustic axis located in the plane of elastic sym- scopes, conoscopes, and gas lasers with polarized
metry of a hexagonal crystal, when the refracted devices are used.
beams are distributed in a fan with a flat section, A conoscopy figure is obtained for light beams
was predicted by A.G. Khatkevich (1962). Under convergent on a birefringent crystal (uniaxial crys-
the conditions of conical refraction several effects tal or biaxial crystal) using a crossed polarizer and
have been observed: the reflection of a linearly analyzer. The type of the figure depends on the
polarized acoustic wave from a free crystal face number of optical axes of a crystal, the angles be-
parallel to the wave vector of an incident wave tween them, the presence or absence of optical
accompanied by a 90◦ rotation of its polariza- activity (see Rotation of light polarization plane),
tion plane, acoustic wave refraction on the internal crystal lamina orientation or their combination,
surface of a twinned crystal (see Twinning of crys-
tals), in quartz, and in a number of other ways.
At ultrahigh frequencies the observation of coni-
cal refraction is complicated by acoustical activity
phenomena (see Rotation of acoustic wave polar-
ization plane).
scribe physically equivalent states, if k − k = where ε(ki ) and ε(ki ) are energies of quasi-
2πb, where b is a vector of the reciprocal lattice). particles before and after a collision, respectively.
In the interaction of elementary excitations (the Related to the law of conservation of angular mo-
collision of quasi-particles), the law of conserva- mentum is the symmetry of the stress tensor in
tion of quasi-momentum, similar to that of linear nonpolarized media.
momentum, holds: The essential information on the interaction
of quasi-particles (as well as ordinary particles)
k i − k i = 2πb. can be obtained on the basis of a conservation
i i law analysis of their kinematics. In particular,
this is how selection rules are established that
Here k i and k i are the quasi-momenta of quasi- forbid processes where the conservation laws do
particles before and after a collision. To avoid am- not hold. The complexity of the dispersion laws
biguity, the quasi-momenta of all colliding quasi- of quasi-particles and the presence of Umklapp
particles (before and after a collision) should be processes make the kinematics of quasi-particle
“located” in the same Brillouin zone. Therefore, collisions very complicated: mutual particle trans-
if the number of colliding quasi-particles is not formations, creation and annihilation, as well as
large, the vector of the reciprocal lattice, involved breakup and fusion of particles, are possible. Even
in the conservation law of quasi-momentum, does in the case of an elastic collision of quasi-particles
not exceed one or two elementary vectors of the (e.g., electron scattering by an impurity atom in
reciprocal lattice. Collisions with b = 0 are re- a crystal), some unusual (for particles) situations
ferred to as normal collisions (emphasizing the are possible, e.g., the transition from one energy
satisfaction of the law of linear momentum con- band (minimum) to another band (minimum), or
206 CONSTRUCTION MATERIALS
Criteria Materials
scattering when the particles scattered in a given to 14 MeV and power up to 2.5 MW/m2 , (2) heat
direction have different quasi-momenta. flux with power up to 0.25 MW/m2 , (3) flux of
neutral atoms of hydrogen isotopes D and T gener-
CONSTRUCTION MATERIALS ated at the ion recharge, j ∼ 1020 –1021 m−2 ·s−1 .
Materials for production of engine parts, build- The main factors which affect the aperture and
ing constructions, etc., for which shape mainte- divertor receiving plates are (1) flow of D+ and
nance under mechanical forces is necessary. Met- T+ ions and heavy impurities with energies up to
als and their alloys such as steel, natural and 20–50 eV, j ∼ 1023 m−2 ·s−1 , (2) heat flux equal
synthetic polymers, and ceramics are used as con- to 10–20 MW/m2 .
struction materials. Composite construction mate- In addition in a contingency the first wall,
rials play a special role, and for usage in aviation aperture, and divertor should be able to sustain a
and astronautics a reduction in weight without sac- discharge disruption and release of all the energy
rificing strength is important. The main quality accumulated in the plasma (1–2 MJ/m2 ) during a
factors of these materials are strength, elastic mod- time of 20 ms. For uncontrollable modes of TR op-
ulus, hardness, impact strength, and temperature eration the aperture and divertor can erode because
dependence of physico-mechanical properties. of the development of unipolar arcs, i.e. electri-
cal discharges between the plasma and plasma trap
CONSTRUCTION MATERIALS FOR chamber.
THERMONUCLEAR REACTORS For design and economy reasons the first wall
Materials of thermonuclear reactor (TR) el- cannot be changed, and should operate in an indus-
ements in contact with the plasma: first wall, trial reactor for ∼30 years. Aperture and divertor
aperture, and divertor. The other parts of the re- receiving plates are changeable. Table 1 cites main
actor construction are exposed mainly to neutron criteria for choice of materials for the first wall of
irradiation so the choice of materials for them a tokamak listed according to their priority. The
is less complicated. In the advanced TR design thermal stress parameter M
called the tokamak the main factors which affect 2σT k(1 − ν)
the first wall are: (1) neutron flux with energy up M= ,
αE
CONTACT ANGLE 207
where k is the thermal conductivity, α is the ther- Constructional strength is appreciably de-
mal expansion, E is Young’s modulus, and ν is graded by the decline of performance factors aris-
the Poisson ratio, characterizes the thermal stress ing from exposure to repeatedly changing loads.
from the temperature gradient: the greater the M There are always a number of factors under actual
value the greater is the allowable wall thickness. working conditions which cause the concentration
The tokamak operates in the pulsed mode, and of stress to be the main origin of the appearance of
the international reactor INTOR pulse lasts 200 s fatigue cracks on exposure to repeatedly changing
with a 20 s pause between. Thermal load cycling stresses.
promotes the aging of materials. The most im-
portant features for the material choice for the
aperture and divertor receiving plates are thermal CONSTRUCTION PLASTICS
and chemical properties, and surface erosion. For a Plastics (see Polymeric materials) designed for
hybrid reactor combining a thermonuclear and fis- the production of parts and components capa-
sion reactor which uses thermonuclear neutrons, ble of keeping their shape during exposure to
construction materials will have less plasma ap- mechanical loads. Polyolefines, polyformaldehy-
plied loads because the high power of thermonu- des, polyamides, fluoroplasts, polyarylates, etc.
clear energy is not required, but the materials are are used in various industries as constructional
subjected to radiation exposures typical of those plastics. The main processing methods for con-
from atomic fission reactors. verting these plastics into components are pressure
die casting, hydroextrusion, and pressing. In most
CONSTRUCTION MATERIAL STRENGTH cases constructional plastics are multicomponent
The strength of construction materials or their systems: antioxidants, lighting stabilizing agents,
models having design shapes and geometrical plasticizing agents, oilers, bulking agents, etc.,
sizes typical of actual construction elements. In which are loaded as additional binders that impart
contrast to the material strength which is evaluated specific properties (e.g., ber glass, carbon, boron
experimentally by testing relatively small material nitride, fire-retardant additives, etc.).
samples on devices of simplified shape, the con-
structional strength is evaluated by testing material
CONTACT ANGLE
samples with geometrical shapes and sizes that are
typical of the majority of actual parts of engines Angle between surfaces of a fluid and a solid
or other devices. Values of constructional strength body (or another fluid) in contact. The contact
are much lower than those of material strength angle θ is counted from the fluid. The equilib-
which is generally obtained for simplified labora- rium contact angle is referred to as the edge angle
tory samples. This difference arises mainly from and is denoted by θ0 . During contact formation
the design shapes which often involve sharp transi- the angle varies from θ = π up to θ ≈ θ0 during
tions from one section size to another, the presence 0.02–0.03 s. On rough solid surfaces the contact
of holes which induce some stress concentration, angle usually decreases, so that for θ < π/2, cos
and other factors. In addition, the surface quality θ = γ cos θ0 , where γ is the surface roughness co-
of actual samples can differ considerably from the efficient. The contact angles for fluid inflow on a
surface of laboratory ones because the handling substrate θin , and for outflow θex , differ from θ0 .
technology of actual device production, as a rule, This is caused by the roughness and heterogeneity
differs from the technology used for preparing of the actual surface. For substrate backings with
simplified laboratory samples. The strength in- a high surface energy the property of “autopho-
dices of actual samples are strongly influenced by bia” is possible in contact with polar fluids: the
their production technology, especially by welding fluids on the substrate generate adsorption films
which causes inhomogeneities in the material, and on which they form a contact angle θ > 0. It is
also the tangent stress due to so-called welding known that θin > θ0 > θex , and this effect is re-
stresses. ferred to as wetting hysteresis. With increasing
208 CONTACT CORROSION
temperature θin → θ0 and θex → θ0 , so the hys- CONTACT PROBLEM OF ELASTICITY THEORY
teresis decreases (see Wetting). Determination of the strain and stress distrib-
ution in a system of solid bodies having a general
CONTACT CORROSION, galvanic corrosion contact surface. For the case of two elastic bodies
A type of electrochemical local corrosion re- in three-dimensional space this problem was set up
sulting from the contact of two or more metallic and solved first by H. Hertz (1882) by considering
materials with different initial corrosion potentials, that the contact area consists of small, strain-free
situated in a conductive corrosion environment. surfaces around the contact point referred to as
The contact corrosion intensity is determined by second-order surfaces. With the help of an electro-
a polarization cathode and anode resistance, the static analogy he determined the pressure function
ratio of their dimensions, and decreases with in- at the contact point, and derived explicit formulae
creased distance from the contact point. The metal relating the magnitude of the impression h, the im-
quantity being dissolved because of contact cor- pressing force F , and the main radii of curvature
rosion is equivalent to the current. Methods for of strain-free surfaces at the contact point, from
protecting against contact corrosion consist of which a relationship of the type h = const · F 2/3
followed. Among contact problems of the theory
sensible equipment designing, eliminating or iso-
of elasticity are stamp problems in which one of
lating contacts of metals having much different
contacting bodies is an elastic halfplane, and the
corrosion potentials, using protectors, etc.
other is an perfectly rigid body impressed into the
halfplane by fixed forces. In the simplest case of a
CONTACT POTENTIAL DIFFERENCE flat die uniformly pressed onto the boundary of a
Difference of electric potentials between sur- half-plane on the segment (−a, a) of the Ox axis
faces of conductors brought into contact at ther- the contact pressure distribution P (x) is deter-
modynamic equilibrium. It is defined by the differ- mined by the formula P (x) = F /[π(a 2 − x 2 )1/2 ]
ence in the work functions of the media in contact where F is the pressing force. One approach to
divided by the electron charge. The contact po- solving such a problem involves reducing it to a
tential difference arises from the requirement of boundary value problem of the theory of analyti-
electrochemical potential equality for media in cal functions.
contact; it is established because of electron ex-
change between the conductors. Thus the surfaces CONTACT STRESSES, mechanical
of both media become charged: negative for the Stresses which appear at mechanically inter-
conductor with the greater work function, and pos- acting solid deformable bodies, or around them
itive for the other. The width of the boundary (e.g., at the compression of bodies in contact).
(contact) layer in which the electric field differs Knowledge of contact stresses is important for the
from zero is defined by the shielding distance strength calculation of bearings, gear and worm
in the conductor, with values 10−10 –10−9 m for transmissions, ball and cylinder compactors, col-
metals and up to 10−6 m for semiconductors. liding bodies, etc. The contact surface size is often
The potential difference can reach several volts, small in comparison with the body dimensions,
and depends on the physico-chemical condition of with the contact stresses rapidly decreasing with
the boundary region. In multiple layer heterophase distance from the initial contact point. The stress
structures representing a chain of conductor con- distribution at the contact area and in its neigh-
tacts the overall potential difference does not de- borhood is nonuniform, and the peak shearing
pend on the intermediate links, but equals the stresses which characterize the strength of com-
potential difference of the extreme outer conduc- pressed bodies can appear at a particular distance
tors. This potential difference is important for the from the contact point. In Fig., the upper part
formation of electrovacuum devices and solid con- shows two spheres in contact, subject to the com-
tact structures: semiconductor junctions, heteroge- pression force P , with A indicating the point of
neous structures, metal–semiconductor junctions, peak stress. The lower part shows how the stress p
etc. in the interface decreases with the radial distance r
CONTINUUM MECHANICS 209
CONTINUUM INTEGRAL
See Path integral.
i.e. l mic l mac , where l mac is the characteris- the field of the stress tensor σik (r, t). Evaluat-
tic length of variations of averaged macroscopic ing the system of equations of the mechanics of
quantities. One can obtain in this manner, e.g., continua requires specifying the constitutive rela-
hydrodynamic equations from quantum equations tions (rheological equations) that establish a link
of motion of the density matrix, or equations of between σik and u, v. In the case of a solid un-
elasticity theory from classical equations of mo- der smooth and rather small strain, this link may
tion of crystal ions. Thus the validity criterion for be considered linear and local (i.e. σik (r, t) de-
the present description of a dynamical system of pending on the values of u, v, taken at the same
a large number of particles is of a self-consistent point r and at the same instant of time t ). More
nature, and is stated in the form of an inequality complicated relationships between σik and u, v
l part l mic l mac . This criterion corresponds are also possible (linear, nonlocal, with allowance
to sufficiently homogeneous systems (gases, liq- for the history of deformation of the medium), rep-
uids). In the case of inhomogeneous systems that resented by complicated functions σik = fik (u, v)
contain many (macroscopic) structural elements or by functionals σik = &{u(r, t), v(r, t)}. For in-
this description may prove to be too detailed, and stance, we have for a linear nonlocal medium
it may be necessary to perform a second averaging
of macroscopic equations of motion over volumes σik (r, t) = Alm
ik (r, t; r , t )ulm (r , t ) dr dt .
of a larger characteristic size l that satisfies the
inequality l str l lm . Here, l
str is the char- Equations of continuum mechanics in the lin-
acteristic size of structural elements or the mean ear local approximation often proceed from gen-
distance between them, and lm is the characteristic
eral considerations (law of conservation of particle
length of variations of doubly averaged quantities. number, symmetry conditions, etc.). They contain
Typical examples are the averaged description of a number of unknown coefficients (elastic mod-
the motion of multiphase media with a heavy mix-
uli, coefficient of viscosity, etc.) that should be
ing of phases (i.e., attenuation of sound in solids
determined experimentally or calculated from a
with many pores, or the elastic properties of a
microscopic theory.
composite material), plastic flow of solids under
The above procedure for obtaining macro-
conditions of a high density of dislocations and
scopic equations from microscopic ones involves
dispersed particles, and the possible propagation
not only purely spatial averaging, but also sta-
of a main crack in a material with a number of mi-
tistical averaging (for quantum dynamic systems
crocracks and cavities in the zone of failure.
Two equivalent methods are used for describ- both kinds of averaging are performed simulta-
ing the motion of macroscopic particles of con- neously through the density matrix). Statistical
tinuous media. The Lagrange method involves averaging may fail to produce closed mechani-
tracking the location of a particle r = r(a, t) at cal equations, since the continuum equations may
every instant of time t ; vector a specifies the parti- contain thermodynamic and chemical quantities
cle, and is usually taken to be the particle location (temperature, concentration, chemical potentials,
at the initial instant of time t0 . The Euler method etc.) along with parameters characterizing internal
involves tracking the velocities, accelerations, etc. degrees of freedom of microparticles or other sub-
of various particles that pass through a certain systems of the body (electric, magnetic, and oth-
point r of the space. Ordinarily the Eulerian de- ers). In this case, the mechanical problem should
scription is used when the mechanical state of be solved together with the thermal, chemical,
the medium is characterized by fields of various etc. ones by adding equations of thermal con-
types: displacement field u(r, t), velocity field ductivity, diffusion, etc. to continuum mechanics
v(r, t), density field ρ(r, t), etc. The fields usu- equations. Typical examples are diffusive-viscous
ally satisfy second-order partial differential equa- creep of polycrystals (see Diffusion mechanism of
tions (Navier–Stokes equations of hydrodynamics, crystal creep), dynamics of ferroelectrics and fer-
equations of elasticity theory, etc.). Along with u romagnets, and propagation of shock waves in
and v, these equations involve the force charac- crystals. Besides that, empirical or calculated re-
teristics of the medium, described most often by lationships may be available, which describe the
COOLING DUCT 211
dependence of mechanical parameters (e.g., elas- derived, apparently, within the framework of the
tic constants) on temperature, concentration, mag- continuum approximation.
netization, etc. The most consistent description of
related mechanical, thermal, electrical, and mag-
CONTRACTION, TRANSVERSE
netic processes can be made within the framework
of the thermodynamics of irreversible processes, See Transverse contraction.
which includes much of continuum mechanics as
a special case. In certain situations, however, one CONVOLUTION OF LINE SHAPE
can confine oneself to the solution of a purely me-
See Line shape convolution.
chanical problem. For instance, in a static state
during a very slow deformation of a solid, the
temperature has time to equilibrate and the de- COOLING AGENT, refrigerant
formation process is described by equations of A material (working medium) used to cool
elasticity theory with isothermal elasticity moduli. a substance (cooling agent) or maintain it at a
Also, under very rapid deformation (e.g., propa- low temperature (cryo-reservoir). The most widely
gation of high frequency sound), heat transport used refrigerants are liquids and gases with low
between neighboring parts of a body does not take condensation and crystallization temperatures
place, and the process is described by the same (3 He, 4 He, H2 , Ne, N2 , CO2 /acetone, etc.), para-
equations, but with adiabatic elastic moduli.
magnetic salts (magnesium cerium nitrate
Ce3 Mg3 (NO3 )12 ·24H2 O, ferric chromium alum
CONTINUUM THEORY OF DEFECTS
and ferric aluminum alum), van Vleck paramag-
Theory of defects in a broad sense based on
nets (PrNi5 , PrPt5 , PrTl5 , PtCu6 , etc.), metals
the continuum approximation in analysis. From
with paramagnetic nuclei (Cu, Ag, In, etc.).
a more specialized viewpoint the continuum ap-
proach to defects is founded on the assumption
that they fill the physical space of the crystal. The COOLING DUCT
subjects of this analysis are usually dislocations, A heat exchanging device to provide the ther-
disclinations, dispyrations, and planar defects, al- mal connection of an object being cooled with
though the methodology can be extended to other a cooling agent (refrigerant), or with a cryostat
problems which are not closely related to the (low-temperature reservoir). In the simplest case a
presence of defects: e.g., this theory is used for cooling duct is a rod (tube, stripe, etc.) of a mater-
the calculating thermal stresses, electrostrictive
ial with good thermal conductivity, which provides
or magnetostrictive mechanical effects (see Elec-
a reliable thermal connection with the parts of an
trostriction, Magnetostriction), elasto-plastic fields
installation containing the object being cooled, as
generated by phase transitions or twinning of crys-
well as the cooling agent, or the low temperature
tals, etc. This continuum theory is constructed on
reservoir. Often (especially at low temperatures)
rather general physico-mechanical principles.
It includes the determination of the tensor of the cooling duct is in the form of a large number
densities of defects, the balance conditions for of thin wires (usually made of copper) which im-
these densities, and the continuity condition (ex- proves the heat transfer efficiency by increasing
istence of solutions). In this theory the connection the contact area. To arrange for the adjustment of
between defect motion and the plastic fields gener- the temperature the cooling duct can be provided
ated by this motion (see Plastic deformation) is an- with a heater, and when it is necessary to rapidly
alyzed. The forces acting in this process and other change the thermal connection then heat switches
factors are evaluated. The system of equations in are used. The associated properties of a thermal
the continuum theory involves a fundamental rela- switch and a cooling duct are combined in a ther-
tion that connects the mobility of defects and the mosiphon, which is an hermetically sealed tube
forces acting on them. This fundamental equation filled with a gas under pressure, providing for its
is considered to be known because it cannot be liquefaction in the desired temperature range.
212 COOPERATIVE LUMINESCENCE
is an integer (0 or 1) it is a Boson particle, and a (+) or anion (−), and counterions (monatomic or
Bose–Einstein condensation called a Cooper pair complex) comprise the outer coordination sphere.
condensate can form in which all of the pairs are in To separate the complex ion from its component
+
the same quantum-mechanical ground state. How- ions such as SO2− 4 or NH4 one can assume that
ever, Cooper pairs have been called “bad” bosons the reaction of complex formation from the cen-
since their creation Bp+ = apσ+ +
a−p,−σ and annihi- tral atom (ion) and ligands can occur under normal
lation Bp = a−p,−σ apσ operators are more com- chemical conditions, and that the central ion is ca-
+ pable of existing independently in solution in a
plicated than the usual creation apσ (annihilation
apσ ) operators of a fermion with momentum p solvated form. This limitation is of a conditional
and spin s = ±1/2 (see Second quantization). character since there is no essential difference be-
Cooper pairs have the more complex commmuta- tween such ions and, for example, PF− 6 and VF6
−
amorphous solids (see Amorphous state), under- resistivity is 15.5 n·m. Electric resistance tem-
standing as the nth coordination sphere the spheri- perature coefficient is 440·10−5 K−1 (at 273–
cal layer with atom positions within the range
rn 373 K). Hall constant is −5.2·10−11 m3 /C. Ab-
of radius values rn corresponding to the magni- solute thermal electromotive force coefficient
tude of nth RDF peak, and considering the number +1.7·10−6 V/K. Reflection factor for 5 μm op-
of atoms in the nth sphere as 4πrn2 Sn ρ, where ρ is tical wave is 98.9%. Work function of a copper
the average atom density. For an amorphous solid, polycrystal is 4.4 eV. Copper is diamagnetic with
with increasing r the RDF peaks rapidly broaden magnetic susceptibility (5.4–10)·10−6 CGS units.
and become flattened, so the coordination sphere Ion plasma frequency is 7.655 THz. About 50% of
concept is used only for small n (usually n = 1–3). copper produced is used in the electrical power in-
Thus, in virtue of the statistical nature of the RDF dustry, and over 30% is used for copper alloys. It
the number of atoms in the sphere (the coordina- is widely used for chemical equipment.
tion number) can be nonintegral. The coordination
sphere concept is useful in the analysis of struc- COPPER ALLOYS
tures with short-range order (see Long-range and Copper forms the following systems with other
short-range order), for self-ordered alloys, for the elements: (a) infinite solubility in solid and liquid
approximate construction of a crystal potential for states: with Ni, Rh, Pt, Pd, Mn, Au; (b) infi-
band structure calculations, and so on. nite solubility in liquid and limited solubility in
solid state: with Li, B, Ag and Bi; (c) peritec-
COPPER (Lat. cuprum ), Cu tic type of interaction (see Peritectic alloys): with
Chemical element of Group IB of the peri- Ir, Nb, Co, Fe; (d) limited solubility in liquid
odic table with atomic number 29 and atomic and solid phases: with Pb, Tl, Cr, V); (e) in-
weight 63.546. Stable isotopes 63 Cu (69.1%) and termediate compounds with many elements (for
65 Cu (30.9%). Nine radioactive isotopes are
dissimilar electrochemical properties Cu forms
known. Outer shell electronic configuration chemical compounds; for the same properties, an
3d 10 4s 1 . Successive ionization energies (eV): electron compound is formed). Cu dissolves up
7.72, 20.29, 36.83. Atomic radius is 0.127 nm. to 31.9 Zn, 20 Li, 16.6 Be, 16.4 Ga, 15.8 Al,
Ionic radii are (nm): Cu+ 0.096, Cu2+ 0.072. Co- 12 Ge, 11.15 Si, 10 In, 8 Ir, 8 Ti, 8 Sb, 7.7
valent radius 0.117 nm. Oxidation states +1, +2, Sn, 7 Mg, 6.85 As, 4.5 Fe, 3.4 P, 2.1 Cd (mo-
more rarely +3. Electronegativity is 1.75. lar content, %). Some copper alloys have special
Copper is a red-tinted metal of face-centered names: bronzes (with Al, Sn and some others),
cubic structure, space group F m3̄m (Oh5 ); a = brasses (with Zn), German silver (with Ni). On
0.361479 nm. Density is 8.932 g/cm3 ; Tmelting = addition of Zn, Sn or Al to Cu, α-solid solutions
1356 K, Tboiling = 2843 K, heat of melting appear first; then, as the concentration increases,
13.0 kJ/mole, heat of sublimation 340 kJ/mole, β-, γ -, δ-phases appear, and so on. All indus-
heat of evaporation 307 kJ/mole (under standard trial brasses and bronzes are of two types differing
conditions). Specific heat is 390 J·kg−1 ·K−1 at in structure and properties: single-phase α-alloys
298 K. Debye temperature is 339 K; linear thermal (plastic, easily deformable, strength and hardness
expansion coefficient 16.8·10−8 K−1 . Coefficient slightly exceeding those of copper) and two-phase
of thermal conductivity is 385–402 W·m−1 ·K−1 alloys (α-solid solution + another phase). The lat-
(at 273 K) and 5 W·m−1 ·K−1 (at 20 K). Adia- ter exhibit higher strength and hardness but lower
batic elastic moduli of crystal: c11 = 168.8, c12 = plasticity. Of practical use are brasses that contain
121.8, c44 = 75.5 GPa at 298 K. Young’s modu- up to 43% Zn, tin bronzes with up to 22% Sn,
lus is 110–145 GPa, shear modulus is 42–55 GPa, aluminum bronzes that contain up to 11% Al. Tin
Poisson ratio is ≈0.33. Rupture strength is bronzes exhibit enhanced strength, hardness, cor-
≈230 MPa, relative elongation 60%. Brinell hard- rosion resistance (α-bronzes that contain up to 7–
ness is 35 HB (at 293 K). Coefficient of self- 8% Sn). Aluminum bronzes are stronger and more
diffusion is 9.1·10−14 m2 /s (at 1233 K). Low- plastic, corrosion- and wear-resistant; two-phase
temperature linear electronic heat capacity co- alloys are commonly used (≈ 10% Al). Quench-
efficient 0.687–0.695 mJ·mole−1 ·K−2 . Electric ing and tempering (because of phase transitions
CORRELATION ENERGY 215
or annihilation of field quanta (see Second quan- the quantities V (r 1 ) and V (r 2 ) decreases. This
tization), or of particles (quasi-particles), or the provides an opportunity for the factorization of
quantum system as a whole in a certain state: the average of the types V (R 1 ) . . . V (R n )V (r 1 )
ξi = ai+ or ai . In solid state physics these are . . . V (r m ) = ψm (r 1 , . . . , r m )ψn (r 1 , . . . , r n ), if
creation and annihilation operators of electrons R i = r i + α and |α| → ∞. In particular, ψ2 (r) →
(holes), excitons, phonons, magnons, photons, or 0 as |r| → ∞. The distance |r 1 − r 2 | at which
states of impurity centers (see Impurity atoms). the pair correlation function ψ2 (r 1 , r 2 ) becomes
In solid state quantum theory pair correlations and remains a negligible quantity is called the
ξ1 (t1 )ξ2 (t2 ), or second-order correlation func- correlation radius or correlation length. This ra-
tions ξ1 (X1 )ξ2 (X2 ), are the most widely used. dius is an important characteristic length scale of
When considering the pair correlator it is con- a random field, which determines the behavior of
venient to use average and difference times: t = macroscopic quantities that depend on the field.
(t1 + t2 )/2, τ = t1 − t2 , hence ξ1 (t1 )ξ2 (t2 ) = See Correlation length.
f (τ, t). Of fundamental importance
is the spectral
representation f (ω, t) = eiωτ f (τ, t) dτ which CORRELATION LENGTH, correlation radius
defines the system spectral characteristics. In the The characteristic dimension of the region
stationary case the dependence on t is absent. The within which correlation between physical vari-
pairwise temporal correlator f = a + (t1 )a(t2 ) in ables is appreciable. At distances exceeding the
its spectral representation describes, through the correlation radius these variables become statis-
frequency or the energy E = ω, the occupation of tically independent of each other, and their cor-
the state to which the operators a + , a are related. relation function becomes equal to the product of
To evaluate the correlation function it is usually their averages. Consider, e.g., a ferromagnet that
more effective to use the related Green’s func- is far from its critical point (transition tempera-
tions. ture). The correlation radius is of the order of an
interatomic distance. Thermal motions and quan-
CORRELATION FUNCTION OF RANDOM FIELD tum uctuations destroy the correlation of spin
A statistical characteristic of a random field orientations at distances noticeably larger than this
(see Random potential) in a disordered system de- radius. As the critical point Tc is approached from
fined by the equality above, the correlation radius grows, accompanied
by the stronger and stronger cooperative effects
ψn (r 1 , . . . , r n ) = V (r 1 ) . . . V (r n ),
that eventually result in the formation of a new
where n 2 is the order of these functions. ordered phase at T = Tc , and the material is fer-
Here 1 . . . n means averaging over the random romagnetic below Tc .
field Vr , and the origin of the energy scale is
selected so that V = 0. Most widely used is CORRELATION TIME
the pair correlation function of a random field Time interval τ during the course of which the
ψ2 in terms of which many experimentally mea- value of the second-order correlation function de-
sured quantities can be expressed, such as the creases by the factor of e = 2.718. The stationary
square of a random field fluctuation V 2 (r). In random quantity u(t1 ) (e.g., thermal displacement
the Gaussian random field case all functions ψn of the crystal atom) correlates with its value u(t2 )
(n 3) are expressed in terms of ψ2 . In a macro- at a later time t2 if the ensemble average value
scopically homogeneous disordered system the of the product c(t1 , t2 ) = u(t1 )u(t2 ) = 0. The
functions ψn (r 1 , . . . , r n ) depend only on (n − 1) quantity c(t1 , t2 ) is called the second-order cor-
differences r 1 − r n , r 2 − r n , . . .; and in particular, relation function. For stationary processes, e.g.,
ψ2 (r 1 , r 2 ) = ψ2 (r 1 − r 2 ). lattice vibrations, c(t1 , t2 ) depends only on the
An important property of random fields in real time difference τ = t1 − t2 , and the dependence
systems is the attenuation of correlations in sys- c(τ ) is such that |c(τ )| |c(0)|. For some classes
tems at infinity. With the distancing of the points of random quantities c(τ ) has an exponential de-
r 1 and r 2 from each other the correlation of pendence on τ : c(τ ) = c(0) exp(−τ/τc ), where
218 CORROSION
τ c is the correlation time, or the time constant conditions the oxidizing agent diffuses through the
associated with the relaxation decay of correla- scale to the solid surface, and solid components
tions. The quantity c(τ ) does not always have an diffuse in the reverse direction. If the transport of
exponential dependence on τ . For example, the components through the scale is impeded its pri-
single-phonon correlation function ui (t1 )uj (t2 ) mary layers show a protective, anticorrosion effect
is not exponential in the low-temperature region at the subsequent contact with the medium. The
where the anharmonicity of vibrations is negligi- corrosion rate can be varied by introducing into
ble (see Anharmonic vibrations), and its correla- the corrosion active medium some chemical com-
tion time corresponds to the reciprocal width at pounds, e.g., inhibitors which decrease the rate of
half-maximum
∞ of the Fourier transform c(ω) = corrosion, or promoters which increase it. The pro-
−∞ c(τ ) exp(iωτ ) dτ . tective mechanism of inhibitor molecules usually
consists in their ability to diminish the limiting
CORROSION (fr. Lat. corrodo, gnaw away) step rate of the corrosion process at adsorption
Physicochemical interaction of a solid and a on a solid surface, or to promote the formation
medium (liquid, gas) resulting in the degrada- of a protective coating which prevents corro-
tion of the material operating properties. During sion (see corrosion protection). Volatile inhibitors
corrosion the oxidation of source material or its in- which protect engineering industry products from
dividual components (anode process) occurs with atmospheric corrosion have gained wide-spread
a simultaneous cathode process, i.e. the reduction acceptance: corrosion inhibitors are added to lu-
of medium components by the uptake of electrons bricants, oils, fuel, brake fluids, and emulsions that
released at the anode reaction. Corrosion takes considerably decrease material losses from cor-
place irregularly on the surface due to the chemi- rosion processes. The formation of anticorrosive
cal and structural inhomogeneity of an actual solid coatings can also be promoted by alloying mate-
body, and it is accompanied by product formation rials (steels) with special additives (Cr, Al, Si).
(including compact non-conducting depositions)
from the elements present in the solid composition CORROSION, CONTACT
and surrounding medium. The local corrosion rate See Contact corrosion.
is limited by the rate of the slowest process, in ac-
cordance with the mixed electron potential on the CORROSION CRACKING
surface (corrosion potential) which is being gener- Brittle failure of a metal in a corrosive medium
ated for conditions during which the rates of anode with a static field of tensile stresses present in
and cathode processes are equal. On the basis of the lattice. The onset of corrosion cracking re-
the type of external exposure on the solid, which quires the presence in the medium of a component
accelerates the breakdown process, we can distin- which creates the conditions for sharp localiza-
guish electrochemical corrosion (during electric tion of the corrosion process on the metal surface,
current flow), stress corrosion, and photocorro- and at structural defects. Corrosion cracking takes
sion (during illumination). In the latter case, for place in solutions of electrolytes (sulfide, chloride
oxidation–reduction processes charge carriers are cracking, caustic embrittlement of steel, hydrogen
excited in the solid, and corrosion products par- cracking of steel in acid, etc.); in gases (hydro-
ticipate in generating a new phase on its surface. gen, ammonia), and in molten metals (mercury).
There are external and internal photoeffects, such A measure of a metal stability against corrosion
as the photoactivation of anode dissolution of cracking is its limiting tension loading with a par-
n-type silicon, at a rate which is limited by the gen- ticular base test, or more precisely, the critical
eration of holes at the Si surface layer. coefficient of stress intensity (cracking stability)
The nature of the surrounding medium deter- KSI in the corrosion crack vertex below which the
mines different corrosion types: electrochemical, crack does not grow. The rate of crack growth is
gas, atmospheric, soil, biological, sea, and so on. determined by the anode (active) metal dissolving
During high-temperature gas corrosion solid oxi- rate in the crack vertex, the possible decrease of
dation products (scale) are usually formed. In such adsorption strength, hydrogen saturation of the
CORROSION OF METALS 219
metal volume before the crack, and the value of is a chemical oxidizing reaction, e.g., by dry ox-
the stress intensity. The relative role of the cor- idizing gases in accordance with the following
rosion factors differs for different metal–solution reaction:
systems. In a hydrogen atmosphere at high temper-
nM + mO2 = Mn O2m . (1)
ature with cathode hydrogen saturation the main
role is played by the hydrogen saturation of the During the corrosion of metals in electrolytic me-
metal, i.e. hydrogen brittleness. dia (electrolytic solutions, including those impreg-
nated by porous bodies (e.g., soils), electrolytic
CORROSION FATIGUE melts) Mz+ cations transfer into the medium to
The degradation of a metallic crystal lattice in form compounds with different solubilities. There
a corrosive medium at exposure to a variable field are corresponding characteristic reactions in acidic
of tension stresses. Corrosion fatigue is a measure and neutral aqueous solutions:
of the limit of permissible cyclic metal loads in a M + zH3 O+ = Mz+ + 0.5zH2 + zH2 O, (2)
given corrosive medium based on the number of
selected load cycles, or on the critical intensity co- M + 0.25zO2 + 0.5zH2 O = Mz+ + zOH− . (3)
efficient of cyclic tension stresses below which the A metal–medium system, as a rule, is thermody-
growth of cracks is not observed. namically unstable, and hence it is an environment
The corrosion fatigue limit is much lower than for the spread of corrosion. Despite successes
that of metal fatigue in an inert medium. An in- in the development of materials with increased
crease of metal strength through a long-term base corrosion resistance, anticorrosive coatings, and
test, as a rule, does not cause an increase in other corrosion protection methods, the problem
corrosion fatigue in a given metal–medium sys- of metal corrosion persists because of the defi-
tem. Depending on the frequency and loading ciencies of materials, the greater severity of the
level, we can distinguish low-cycle and high-cycle conditions imposed on metal usages, and other
corrosion fatigue. The types of loading are the reasons. Economic losses from the corrosion of
stretching from zero, expansion–contraction, arch- metals result from the decreased machinery and
ing (pure and cantilever), torsion with different communications productivity, direct metal losses,
forms of loading cycles (sinusoidal, rectangular, and so on. In the case of the formation of thin
trapezoidal, symmetric, asymmetric). Taking into (<5 nm) surface lm s Mn O2m reaction (1) is
account the presence of residual technology load- associated with the tunneling of metal electrons
ing in real systems, the most general is asymmetric through the boundary film–gas, and is limited by
loading, which is also the most dangerous for a the tunneling itself or by cation Mz+ migration
metal–medium system that is susceptible to corro- arising from the appearance of an electric field in
sion cracking. Methods for limiting the increase in the film. In the case of a thick film (scale), be-
corrosion fatigue are the creation of compression cause of its density the reaction rate is more often
stresses in the surface layers of a construction, sur- determined by cation Mz+ diffusion to the bound-
face alloying or use of protecting coatings, change ary film–gas, the diffusion of O2− ions to the
of corrosion medium composition, lowering of the metal–film boundary, or their cross-diffusion. Re-
actions (2) and (3) usually go like an electrochemi-
temperature, application of inhibitors, and electro-
cal process, i.e. without any direct contact between
chemical protection.
the metal atom being ionized and the ion or oxi-
dizer molecule being adsorbed at another surface
CORROSION, INTERCRYSTALLINE location, with the electron transfer from metal to
See Intercrystalline corrosion. oxidizer taking place at the metal conduction band.
Such a process includes two reactions associated
CORROSION OF METALS with the condition of system electroneutrality, but
Metal destruction due to chemical or chemico- which are kinetically independent, namely anode
physical interaction with a corrosive medium. metal ionization (e.g., M = M2+ + 2e) and cath-
A common cause of the corrosion of a metal (M) ode reduction of oxidizer particles on the surface
220 CORROSION PROTECTION
are neutralized in insulators and semiconductors hopping conductivity versus hopping distance in
with their low concentration of current carriers are regimes where that the principal contribution to
much slower, and the influence of the Coulomb ex- the electrical conductivity arises from states in the
plosion on the formation of defects may be much energy band that are within the energy range Δ
more effective. from the Fermi level. A technique to observe the
Coulomb gap directly is tunneling spectroscopy.
COULOMB GAP
A gap in the density of states g(E) of electron
COULOMB PSEUDOPOTENTIAL
excitations around the Fermi level μ in a disor-
See Morel–Anderson Coulomb pseudopoten-
dered system with localized electrons, e.g., the
tial.
gap in weakly doped compensated or amorphous
semiconductors that results from the long range
Coulomb interaction. An important characteristic COVALENT BOND, homopolar bond
of the Coulomb gap is its energy Δ. Provided that A particular type of chemical bond formed be-
the Coulomb interaction energy between two lo- tween similar atoms or atoms whose properties
calized current carriers separated from each other differ very little from each other, when the pair-
by an average distance (averaged over all local- ing (overlap) of the electron shells of neighboring
ized carriers) remains small in comparison with atoms takes place without any appreciable ex-
the characteristic spread of the energy levels, the change of the electric charges of the two atoms.
1/2
value of Δ is of the order of e3 g0 /ε3/2 in a A typical example of a covalent bond is the bond
three-dimensional system. Here ε is the dielec- in a hydrogen molecule. The two-electron wave
tric constant, and g0 is the density of states at the function of such a molecule can be written down in
Fermi level in the absence of the electron–electron the simplest Heitler–London approximation as fol-
Coulomb interaction (see Fig.). Such a situation lows:
is found in a crystalline semiconductor for both
strong and weak compensation, and also in amor- ψ(r1 τ1 , r2 τ2 ) = ψa (r1 )ψb (r2 ) + ψa (r2 )ψb (r1 )
phous semiconductors. In the case of medium
compensation in a weakly doped semiconductor Δ × χa (τ1 τ2 ), (1)
is comparable to the overall width of the impurity
where ψa and ψb are the 1s-type one-electron
band, and the Coulomb gap is more strongly ex-
wave functions centered on the a and b atoms,
pressed. The density of states at the Fermi level
vanishes only for zero temperature. With increas- respectively, and χa is the antisymmetric wave
ing temperature the Coulomb gap “closes” and function of the spins. The difference between the
totally vanishes as soon as the thermal energy kB T eigenenergy obtained from solving the Hamil-
reaches several tenths of the value of Δ. This gap tonian of the molecule constructed using the wave
manifests itself in experiments that measure the functions (1) and twice the electron energy in an
isolated atom corresponds to the covalent bond en-
ergy. A similar construction is used for the wave
function of two electrons of a σ -bond between
nearest neighbor atoms in homopolar crystals (di-
amond, gray tin, germanium, silicon) as well as
in the AIII BV and AII BVI compounds (ψa and
ψb are distinguishable in this case, appearing as
sp3 hybrid functions of the valence electrons of
the atoms involved). If the crystal is formed from
atoms belonging to different groups of the periodic
system then there can be a greater or lesser admix-
Density of states g(E) of an amorphous semiconductor ture of ionic bond character in the bond, and this
near the Fermi level μ. will be termed an ionic-covalent bond.
CRACK 223
√
proportional to 1/ r where r is the distance from in the crack of dislocations from the slip bands
the crack tip. Depending on the shape of the dis- stopped by boundaries. At this stage of failure
placement between the edges, cracks are classified a substantial role is played by rotation processes
as rupture cracks, and longitudinal and transverse when individual grains and their groups are able
shear cracks. In respect to stability, cracks are to rotate freely about the crack tip (see Plastic
divided into equilibrium (i.e. subcritical which twisting deformation). Subcritical growth of the
do not grow spontaneously) and nonequilibrium macroscopic crack proceeds via the same mech-
(i.e. overcritical which can propagate under ac- anisms with the added feature that the plastic
tion of the elastic potential accumulated by the deformation involving a large volume leads to the
system). In respect to the involvement of plastic appearance of a system of secondary microcracks
deformation, cracks are divided into brittle (i.e. not and pores prior to the main crack formation. The
connected with material yield) and tough (see Brit- failure proper is then effected through a merging
tle failure, Tough failure) with a certain portion of of the crack system into a common one by the
plasticity. There are microscopic and macroscopic tough rupture of bridges.
cracks. Classed among the former are sizes of The third stage of crack development, involv-
cracks from the atomic scale to the subgrain scale. ing overcritical growth, proceeds with completely
Referred to the latter are cracks whose dimensions or almost completely suppressed plasticity at the
are greater, by at least an order of magnitude, com- vertex. Here, the spreading of the crack exhibits a
pared to the structural elements of the solid such as distinctly elastic and overstructural character, with
subgrains, crystallites, etc. the limiting velocities of growth determined by the
The first stage of failure is microcrack nucle- elastic constants of the medium, and by the sam-
ation. There are two kinds of microcracks differing ple geometry. The maximum possible velocities
in the shape: elastic (force) cracks and dislocation of failure, according to different evaluations, vary
cracks. The former are like an incision along an from 0.53 of the transverse elastic wave velocity
atomic plane with broken molecular bonds from to 0.38 of the longitudinal wave velocity. The lim-
one tip to another, and the surfaces are opened iting velocity of the crack, equal to the velocity
by the applied stresses. A particular feature of of surface Rayleigh waves (see Surface acoustic
such a crack is its instability. It can spontaneously waves) is considered to be the preferable index.
collapse after the stress release, so that it cannot In the case when additional feeding of the crack
exist in an unloaded body. A dislocation micro- vertex by elastic energy is not mediated by the
crack is an empty nucleus of a superdislocation elastic waves that spread through the destructible
with a large Burgers vector. It is mechanically material, but is performed directly, e.g., by plasma
stable, and can occur in an unloaded body. The expansion at the vertex of crack under the effect of
nucleation of this type of crack is determined by a laser beam, the velocities of crack propagation
collective dislocation processes. There is a se- may be considerably higher, and may even exceed
ries of crack nucleation mechanisms: dislocation the sound velocity.
junction, braked shear, opening of bent slip band, See also Brittle crack, Nucleating crack, Submi-
interaction of dislocations at crossing slip planes crocracks, Tough crack, Wedge-shaped crack.
and at broken walls of subgrains, generation ac-
cording to the scheme of opposite clusters and CRACKING
interacting twin layers, and so on. The process of extraction of products with
The second stage of microcrack development lower molecular mass from hydrocarbons, mainly
(subcritical growth) is also fully determined and of petroleum origin. The first and foremost high
controlled by dislocation phenomena. Its essence tonnage industrial application of the process is ex-
is the “falling” of dislocations in the crack that traction of motor fuels. Many processes of crack-
induces the growth of its Burgers vector. There ing are run in the presence of solid catalysts
are several mechanisms for this process: the crack (catalytic cracking). The catalysts most often used
is initiated by the slip band approaching a tip; are highly porous synthetic aluminosilicates con-
two slip bundles adjoin the crack; an accumulation taining Al2 O3 and SiO2 in various proportions.
CREEP 225
At the second stage of creep the duration and strain is achieved in a given time. Creep resis-
rate of deformation depend on the stress and tem- tance depends on the type of interatomic bond,
perature. The equation K = A exp[H /(k B T )] × type of crystal lattice, stacking fault energy, grain
sinh(Bσ )n , where T is temperature, σ is stress, structure, and other characteristics of the material.
A, B, n are constants, describes the temperature High creep resistance is the main factor determin-
dependence and stress dependence of K. The acti- ing the high-temperature strength of materials. See
vation energy H monotonically increases with the also Athermic creep, Diffusion mechanism of crys-
increase in temperature; for T > T G it is a con- tal creep, Radiation-induced creep.
stant equal to the self-diffusion activation energy.
This means that at low temperatures diffusion- CREEP RESISTANCE
related mechanisms of plastic deformation are ac- See Creep.
tivated. It has been found that slippage along the
grain boundaries contributes significantly to the CRITERION OF LOCALIZATION
creep at the second stage. The dependence of ε̇ See Anderson localization.
is ∝ d 2 at low temperatures and ∝ 1/d at high
temperatures, while at intermediate temperatures CRITICAL CURRENTS in superconductors
it takes the form k/d + Kd 2 , where k and K are The minimum value of the current Ic or cur-
constants, and d is the grain size. The influence of rent density jc at which the superconducting state
grain boundaries at elevated temperatures may be is destroyed. Several types of critical currents are
explained by localization of plastic deformation in identified. In type I superconductors the value of
border zones. Ic depends on the shape (geometry) of the sam-
The third stage of creep is prepared by struc- ple. In particular, for a long enough cylindrical
tural changes, which take place at preceding superconductor it is obtained from the condition
stages. It is thought that acceleration of creep oc- that the azimuthal component of the magnetic
curs because of an increase of stress that is related field of the longitudinal current, Bf = I /(2πa),
to the generation of a neck, the decrease of the reaches the value of the thermodynamic critical
actual cross-sectional area caused by accumula- magnetic eld Bc (T ) at the surface of a sample
tion of microcracks (see Cracks) and pores, or of radius a. When I > Ic = 2πaBc (T ) the su-
processes of recovery and recrystallization. perconductor transforms to the intermediate state
(4) The fourth kind of creep, which is called with a finite resistance. In type II superconduc-
diffusive creep, is observed at low temperatures tors the critical current density is determined by
and low stresses, and does not involve the trans- the pinning strength of the Abrikosov vortices, or
fer of dislocations, but is a result of directional by the fluidity limit of the vortex lattice (see Vor-
diffusive mass transport. The rate of such creep, tex pinning). There is also a decoupling critical
due to the self-diffusion or flux of vacancies be- current at which the Cooper pairs are broken and
tween external surfaces of the sample, has been the supercurrent flow is terminated. The current
calculated in works of F.R.N. Nabarro, C. Herring velocity v = j/(eN0 ) (here j is the current den-
(1950) and R.L. Coble (1963). Taking into account sity, N0 is the concentration of electrons in the
the increase of creep rate caused by grain bound- metal) exceeds the critical velocity vc = Δ/pF ,
ary diffusion, we obtain the equation: where Δ is the superconductor energy gap, and
pF is the Fermi momentum (see Kinetic decoupling
5D σ Ω πδDg
ε̇ ≈ 2 1+ , of Cooper pairs). Such currents may be reached
d kBT Dd
in very thin superconducting films and filaments,
where D and Dg are the coefficients of bulk and the magnetic field of the current at their surface
grain boundary self-diffusion, respectively, δ is the being lower than Bc (T ) in type I superconduc-
effective thickness of a grain boundary, and Ω is tors, or such that their thickness is smaller than
the atomic volume. the London penetration depth of the magnetic eld
Creep resistance of materials is characterized of a type II superconductor. However, recalling the
by the creep limit, i.e. the stress at which a certain presence of inhomogeneities, it becomes clear that
CRITICAL EXPONENTS 227
the destruction of the superconducting state, and CRITICAL EMBRYO, critical bubble
the recovery of the resistive state in thin films and Seed or embryo of a stable thermodynamic
filaments are usually controlled by vortex pinning phase in unstable equilibrium with its surround-
at these inhomogeneities. Finding ways to increase ing metastable phase (see Metastable state). The
critical currents is one of the most important tasks radius of a spherical critical embryo is given by
of applied superconductivity.
2γ v 2γ Tc (vM − v)
rE = = .
CRITICAL DISORIENTATION OF DISLOCATION μM (P ) − μ(P ) q(T − Tc )
SUBBOUNDARY Here γ is the coefficient of surface tension; v,
The threshold value of the disorientation an- v M are the atomic volumes; μ and μM are the
gle in a cellular structure at which, similar to chemical potentials (see Thermodynamic poten-
grain boundaries, cell boundaries start to resist tials) for the embryo and the metastable phase,
the movement of dislocations. The concept of crit- respectively; P is the external pressure, Tc is
ical disorientation of dislocation subboundaries the phase equilibrium temperature; q is the latent
resulted from comparing the mechanical proper- heat of the phase transition from the metastable
ties of metals during the formation of the sublat-
phase to the embryo phase. Only embryos above
tice dislocation structure (either cellular or frag-
the critical size continue to grow. The process of
mented) that develops in the process of plastic
growth, stable with respect to fluctuations, sets in
deformation (see Fragmentation). While the over-
at r > r E + [k B T /(4πγ )]1/2 . For a supercritical
all density of dislocations remains unchanged with
embryo (r > r E ) to form it has to penetrate the
growing plastic deformation, they undergo redis-
thermodynamic potential barrier 4πrE2 γ /3. See
tribution in the process. The number of disloca-
also Nucleation at phase transitions, Crystalliza-
tions in the body of the subgrain diminishes, their
tion.
density at subboundaries increases and the an-
gle of disorientation of the adjacent cells grows,
so that dislocation slip to large distances becomes CRITICAL EXPONENTS, critical indices
more difficult. The critical value of the angle of Power law indices or exponents defining the
disorientation, θcr , usually remains within several dependency of dynamic and thermodynamic pa-
degrees, and may be estimated from the expression rameters such as the external field h, etc., on the
reduced temperature, τ = T /Tc − 1, etc. around a
Ky 2 (1 − ν)2 2πn second-order phase transition point (Tc ). Critical
θcr = ,
G 0.45b exponents only characterize the principal, leading
where G is the shear modulus, ν is the Poisson asymptotes of these quantities. When τ, h change
ratio, Ky is a constant value from the Hall–Petch the system may pass through several regimes or
equation (see Petch relation), b is the Burgers vec- phases, each of them described by its specific
tor, n is the ratio of the effective slip length L to set of critical exponents. For example, when ap-
cell size l. Following the growing plastic deforma- proaching the triple critical point the system passes
tion the number of subboundaries in the already through a transition region in which its behavior
formed dislocation cells increases, such that their changes from that typical of a second-order phase
disorientation angle is either equal to or already transitions to a tricritical (around that point) be-
exceeds the value of θcr . This results in a sig- haviour.
nificant strengthening of the material, while the Critical exponents used to describe thermody-
temperature of cold brittleness simultaneously de- namic (statistical) properties of the system define
creases, which means that the formation of a dis- the temperature dependences of the speci c heat
ordered cellular dislocaton structure significantly cp ∝ |τ |−α , the order parameter M ∝ (−τ )β ,
improves the overall mechanical properties of the the susceptibility χ ∝ |τ |−γ , and the correlation
material. For such a structure to form certain con- length rc ∝ |r|−ν , where the exponents β, γ ,
ditions have to be met, and the strain should take ν > 0. The critical exponents of anomalous di-
place within the temperature range of warm defor- mensions for fluctuations of the order parameter
mations. characterize the law by which the correlation func-
228 CRITICAL FLUCTUATIONS
tion decreases with distance r for r rc , G(r) ∝ the crystal. Much effort has been expended to-
r −(d−2+η) , where d is the dimensionality. For ward explaining the strong nonlinear interactions
T = Tc the critical exponents describe the behav- of critical fluctuations (see Renormalization group
ior of the system in the external field h: cp ∝ h−ε , method, Epsilon expansion). Experimental studies
M ∝ h1/δ , rc ∝ h−μ (δ, μ > 0). Critical expo- of critical fluctuations by neutron diffraction, opti-
nents are related to each other via certain algebraic cal, radiospectroscopic, and other techniques yield
expressions called scaling relations: α + 2β + γ = rich data on the behavior of materials in the critical
2, βδ = β + γ , ε(β + γ ) = α, μ(β + γ ) = ν, region.
ν(2 − η) = γ , γ d = 2 − α (see Scaling invariance
hypothesis). Critical exponents also describe the CRITICAL MAGNETIC FIELDS in
kinetic (dynamic) properties of the system around superconductors
Tc . For example, the relaxation time t (k, T ) of Values of applied magnetic field strength (that
fluctuations of the order parameter with wave vec- depend on superconductor shape) above which
tor k has the form t (0, T ) ∝ |τ |−y and t (k, Tc ) ∝ the character of the penetration of the magnetic
field into the superconducting material changes,
k −z in the critical region. Scaling relates these ex-
or the sample reverts to the normal (nonsuper-
ponents to each other: y = zν. Critical exponents
conducting) state. One distinguishes between the
are universal values, i.e. they do not depend on
thermodynamic critical magnetic eld Bc in type I
the nature of the particular phase transition, or
superconductors (see Meissner effect), and the
on the specifics of the physical system. They de-
lower critical eld, Bc1 , the upper critical eld, Bc2 ,
pend, instead, only on the number of components
and the third critical eld Bc3 in type II supercon-
of the order parameter, and on the dimensionality
ductors. See also Mixed states.
d of the space. Critical uctuations become in-
significant for d > dc , where dc is the so-called CRITICAL PHENOMENA
upper critical dimension, so that the critical ex- Specific effects observed in the neighborhood
ponents coincide with those provided by the Lan- of second-order phase transition points. One dis-
dau theory of second-order phase transitions. At tinguishes between static and dynamic critical
d = dc fluctuations generate multiplicative loga- phenomena. Among the first are the anomalies of
rithmic corrections to the results from the Landau the thermodynamics of crystals, and features in the
theory. If d < dc the corresponding critical expo- behavior of static kinetic coef cients. Examples of
nents differ from those prescribed by the phenom- static critical phenomena are: sharp increase in
enological theory of phase transitions. Their val- the magnetic permeability of ferromagnets as the
ues may be (approximately) calculated using the phase transition point Tc is approached; anom-
techniques of high temperature expansions within alous temperature dependence of the speci c heat,
the renormalization group method, in particular elastic moduli, etc.; increase in the electrical con-
through expansions in powers of ε = dc − d (ep- ductivity of normal metals near superconducting
silon expansion), in terms of the spherical model, phase transitions; and temperature anomalies in
employing the 1/n expansion approach, and by the specific heat of ferroelectrics. Dynamic criti-
other techniques. cal phenomena involve the dynamic susceptibility,
a sharp increase in relaxation times as Tc is ap-
CRITICAL FLUCTUATIONS proached (critical slowing down), anomalies in
Particularly strong thermodynamic fluctuations Raman scattering of light, in inelastic neutron scat-
of the order parameter and other physical variables tering, in X-ray scattering, and in Mössbauer spec-
observed near second-order phase transitions. The tra (see Mössbauer effect).
correlation length of these fluctuations and their re- Sometimes critical phenomena are understood
spective relaxation times grow without limit when in a narrower sense, as effects produced specifi-
approaching the transition point, thus reflecting cally by critical uctuations of the order parameter.
destabilization of the corresponding phase, and The susceptibility χ , the order parameter M, and
the stronger role of the cooperative effects in the specific heat C, all behave as power law func-
forming the dynamics and thermodynamics of tions of the reduced temperature τ = (T − Tc )/Tc
CRITICAL STATE 229
around the critical point Tc : χ ∝ |τ |−γ , M ∝ the diminishing field energy as the lines of force
|t|β , c ∝ |t|−α . Second-order phenomenological flatten, so that ratio grows for larger pore vol-
theory of phase transitions predicts the values umes. As the pore enlarges it sucks in more
α = 0, β = 1/2, γ = 1 for these power law indices and more energy from the field, and it even-
called critical exponents, but it does not take fluc- tually reaches the critical size V = V cr , where
1/3
tuations into account. Experiments usually yield Vcr = 0.476γ (ln(0.309/N 7/6 ))1/6 /(M 2 N 7/6 ).
the ranges −0.1 < α < 0.1, 0.3 < β < 0.4, and The principal axes of the demagnetization factor
1.2 < γ < 1.4 for three-dimensional (non-layered, tensors of the sample and pore (see Demagnetiz-
non-quasi-one-dimensional) systems. Specific val- ation) are assumed to coincide. Here M is the
ues of the critical indices depend on the number of moment per unit volume, and N is the principal
components of the order parameter. value of the sample demagnetization tensor along
the pore axis. As a rule, the segregation of impu-
CRITICAL POINT rities in the pore surface layer lowers the surface
A point associated with a second-order phase tension and results in a lower r cr .
transition. When the transition occurs during a
temperature change the critical point corresponds CRITICAL REGION, critical temperature range
to a critical temperature. Examples are the Curie A fluctuational range in the neighborhood of
point of a ferromagnet, and the Néel point of an a phase transition, located in the phase diagram
antiferromagnet. The critical point can also be the adjacent to the second-order phase transition line,
value of an electric or magnetic field which in- or to the critical point, where the thermodynamics
duces a change in phase, such as the upper and and dynamics of the system are strongly affected
lower critical magnetic fields of type II supercon- by critical uctuations of the order parameter. The
ductors. size of the critical region is characterized by the
Ginzburg number, and depends on the type of
CRITICAL POINT, TRIPLE
interatomic (electron, spin) interaction potential
associated with the phase. The behavior of the
See Triple critical point.
system in the critical region may differ both quan-
titatively and qualitatively from that predicted by
CRITICAL PORE SIZE
the Landau theory of second-order phase transi-
Size of a pore in unstable equilibrium with
tions.
its surrounding environment. Smaller pores shrink
and heal (see Healing of defects), while larger CRITICAL STATE
sizes grow. The critical radius of a spherical pore The limiting equilibrium state of a two-phase
r cr depends on the coefficient of surface tension system at which the physical differences between
γ , the density of point defects, their mobility, the two phases vanish. For example, the α- and
the nature of their elastic interaction with the γ -phases of cerium with the same face-centered
pore (see Elastic interaction of defects), the gas cubic lattice become identical in the critical state.
pressure inside the pore and outside the sample, The “liquid–vapor” case is quite well known. The
and other factors. For example, we have r cr = critical state on a phase diagram corresponds to
2γ /[P + k B T N0 ln(1 + c/c0 )] in the presence of a critical point or critical line. When approach-
extra vacancies (vacancy concentration c, equi- ing that point critical uctuations of the density
librium concentration c0 ). Here P is the differ- of the substance become very much larger, while
ence between the pressures inside the pore and the latent heat of a rst-order phase transition ap-
outside the sample, k B is the Boltzmann con- proaches zero. At the critical point itself the phase
stant, N0 is the number of atoms per unit vol- transformation occurs simultaneously throughout
ume of material. In polarized substances (with the entire system, similar to a second-order phase
electric or magnetic moments) the pore tends to transition. The values of the state parameters for
stretch along the moment direction. The equilib- the critical state are called critical parameters,
rium ratio of pore axes is determined by com- e.g., critical temperature, critical pressure, critical
petition between the growing surface energy and volume.
230 CRITICAL STRAIN
CRITICAL STRAIN, critical degree of deformation by power law functions (see Critical phenomena,
The extent of strain that results in a qualitative Scaling invariance hypothesis). Several examples
change in the behavior of a material under a given of critical temperatures are: transformation of a
external action (e.g., deformation which produces metal to a superconducting phase, transitions to
sample failure), or an irreversible change of the magnetically ordered or ferroelectric states (Curie
sample shape (i.e. plastic deformation). Sometimes point), and transformation of a liquid to a super-
one may speak about a critical deformation with fluid state.
respect to certain physico-chemical characteristics
of a solid, in particular about a critical plastic de- CRITICAL TEMPERATURE OF
formation resulting in recrystallization, etc. SUPERCONDUCTORS
Transition temperature Tc from the normal
CRITICAL SUPERCOOLING FIELD state of finite resistance R to the superconduct-
The value of the magnetic eld Bs at which ing state of zero resistance, which is specific for
a sample in a normal (i.e. nonsuperconducting) each superconductor (see Fig.). In the absence of
metastable state transforms into a superconduct- a magnetic field the transformation corresponds to
ing state. A type I superconductor can be super- a second-order phase transition. For the phonon
cooled, meaning that it can remain in the normal mechanism of superconductivity the value of Tc
state when the temperature is below the supercon- depends on the magnitude of the electron–phonon
ducting transition temperature T c , and the applied interaction constant λ. This interaction involves
magnetic field is also below the thermodynamic the attraction between pairs of electrons in the
critical magnetic eld B c . A further lowering of vicinity of the Fermi surface with the forma-
the applied field to the value B s brings about a tion of Cooper pairs, despite repulsion from the
phase transition to the superconducting state. The Coulomb interaction. The calculation of Tc from
effect is only observed in pure (type I) supercon- first principles is so complex that Tc is usually
ductors, and its magnitude is characterized by the estimated from its dependence on various parame-
ratio α = Bs /Bc < 1. In some superconductors ters in terms of simplified models. The most useful
(Al, In, Sn, etc.) the extent of the overcooling may approach involves the Bardeen–Cooper–Schrieffer
be quite significant, and indium has been super- theory (BCS) which yields the following expres-
cooled down to α < 0.1. sion for Tc :
matrix elements of the electron–phonon interac- much weaker than deformation scattering, mean-
tion (second order in perturbation theory) and the ing |
ik |/λik ∝ (eF τ )−1/2 for eF τ
1, where
Coulomb repulsion, respectively. Note that g > 0 τ −1 is the frequency of electron–impurity scat-
for Vph > Vc , corresponding to the BCS crite- tering. However, since the deformation potential
rion for superconductivity. A more accurate ac- in one-dimensional metals does not depend on the
counting for the Coulomb repulsion in the energy quasi-momentum of the electron at the Fermi sur-
range near EF results in its significant weakening face, and the electron energy bands do not over-
(the logarithm ln[EF /(ωD )] becomes large (see lap, the Coulomb interaction between the electrons
Bogolyubov–Tolmachyov logarithm) and a soften- almost completely screens the potential λik (to the
ing of the superconductivity criterion accuracy of terms in (qa)2 of the order of mag-
nitude (qa)2 1, where q is the wave number
Vc
Vph > Vc∗ ≡ . of sound, and a is the crystal lattice constant).
1 + N(EF )Vc ln[EF /(ωD )] Therefore, alongside the inertial interaction re-
McMillan (1968) was the first to make approxi- sulting from the Stewart–Tolman effect, the cross-
mate calculations of Tc for the cases of intermedi- deformation interaction is the principal type of in-
ate and strong coupling in the limit λ 1. He used teraction of electrons with long-wavelength sound
the integral Eliashberg equations and obtained: waves in one-dimensional metals (see Electron–
phonon interaction).
ΘD 1.04(1 + λ)
Tc = exp − ,
1.45 1 − μ∗ (1 + 0.62λ) CROSS-LINKING OF POLYMERS
where μ∗ = N(EF )Vc∗ is the Morel–Anderson See Polymer cross-linking.
Coulomb pseudopotential.
CROSSOVER TRANSITIONS
CROSS-DEFORMATION INTERACTION, See Direct transitions.
electron–phonon interaction in one-dimensional
metals CROSS-RELAXATION
Interaction between conduction electrons and Relaxation processes in spin and other quan-
long-wave phonons in conductors exhibiting one- tum systems with non-equidistant energy levels
dimensional electrical conductivity (one-dimen- that control the reestablishment of equilibrium be-
sional conductors). It is produced by modula- tween these levels after excitation by a radiation
tion of a random static scattering field of im- source. For example, consider a spin system with
purities or other defects by a sound wave. It two resonant frequencies ωa and ωb , so that the
is non-diagonal in the electron momenta at the difference between them does not exceed their
Fermi surface of a one-dimensional metal, i.e. the line widths. The process of cross-relaxation de-
sign of the electron velocity can change due to velops via mutual spin flips, the so-called spin-flip
this interaction. It is similar to the common de- process, resulting from the magnetic dipole inter-
formation interaction (A.I. Akhiezer, 1938) that action between the spins. It can also involve spin
is described by introducing an additional term exchange. The extra energy (ωb − ωa ) released
Hd = λik uik into the Hamiltonian of an electron, during cross-relaxation is compensated by the in-
where the strain potential tensor λik is of the or- crease of the dipole–dipole interaction energy of
der of the Fermi energy, eF , and uik is the strain the spin system.
tensor. The cross-deformation interaction results Cross-relaxation processes may involve two
in the appearance of a new term in the Hamil- qualitatively different types of spin systems: res-
tonian, its structure schematically represented by onant frequencies ωa and ωb for different types
Hcd = Δik uik , where Δik is the complex cross- of spins (Fig. (a)); and frequencies ωa and ωb for
deformation potential tensor. In the case of the one and the same spin type (Fig. (b)). The latter
weak scattering of electrons at impurities (de- case can occur in a system with S > 1/2 under
fects), the cross-deformation interaction appears the influence of a crystalline electric field. The rate
232 CROSS-SECTION
CROSS SLIP
A transition of helical sections of dislocations
from one slip plane with the normal n, to another
one with normal m, if the line of the intersection
of these planes is parallel to the Burgers vector
of the dislocation b, i.e. [n × m]||b. A slip plane Crowdion.
is the closest packed plane of a crystal on which
dislocations glide. The cross slip appears, first of
A dynamic crowdion appears when a pulse suffi-
all, near the obstacles for the usual dislocation slip.
cient to displace a neighboring atom is transmitted
If such an obstacle is local, then after the motion
along a fully packed atomic row to one of the
within the cross slip plane the dislocation may re-
atoms of the crystal. A relay displacement then
turn to the initial plane, i.e. the process of double
takes place, and the crowdion propagates along
cross slip takes place. It is experimentally estab-
the direction of the transferred pulse. A site va-
lished that in face-centered cubic single crystals
cancy remains at the site of the initially dislodged
the beginning of the third stage of hardening is as-
atom. A dynamic crowdion is able to propagate
sociated to the fact that as a result of the stress
a distance of the order of 103 a from the position
concentration by single or multiple steps there ap-
pear areas of dislocation lines which may perform of the initially dislodged atom. A static crow-
cross slip. The process of cross slip determines the dion may remain stable in some crystals within
creep of metals in the interval of moderate tem- a certain range of temperature. It propagates as a
peratures (e.g., for aluminum single crystals within whole, and the process involves one-dimensional
the temperature range 0 to 100 ◦ C). diffusion along the atomic line. Although a crow-
dion can travel a long distance, the displacement
CROWDION (fr. crowd) of an individual atom at each step in the dis-
A specific quasi-one-dimensional point defect placement process is the same (equal to a). In its
of the type of an interstitial atom such that adja- quantum-mechanical description the crowdion is a
cent atoms are preferentially displaced along one quasi-particle (a crowdion wave).
of the directions of dense packing. The extra atom
in a crowdion in a fully packed row displaces 5 CRYOCRYSTALS
to 10 atoms along the line by individual minimal Solidified gases, i.e. crystals formed by atoms
distances (much shorter than the interatomic dis- or small molecules with weak intermolecular in-
tance a, see Fig.). H. Paneth (1950) introduced the teraction forces, which therefore feature low crys-
concept of a crowdion to describe self-diffusion tallization temperatures. Their relative simplicity is
in alkali metals with BCC lattices (with crowdion convenient for studies of the general properties of
direction 111). Experimental data are available the crystalline state of matter.
proving the existence of crowdions in FCC metals Atomic cryocrystals – solid Ne, Ar, Kr, Xe are
(Cu, Ni) at low temperatures (along the 110 di- closest to theoretical models of crystals with cen-
rection). A crowdion may form during radiation. tral force interactions. They form face-centered
234 CRYOGENIC TECHNOLOGY
cubic lattices containing identical atoms inter- gases. Industrial applications of cryogenic tech-
acting via valence forces of attraction, and dis- nology are numerous, since pure oxygen obtained
persion forces of repulsion. Their low-frequency by liquefying air and separating its gases at low
collective excitations are phonons, and their high- temperature is widely used in metallurgy and other
frequency ones are excitons. industries. Liquid nitrogen, also obtained from liq-
Solid N2 , O2 , CO, CO2 , CD4 , etc., are mole- uid air, is more and more widely used for the long-
cular cryocrystals. Their molecules may be con- term preservation of food products and biological
sidered nondeformable since their intramolecular materials, to operate transport refrigerators, for
forces are much stronger than the intermolecular the needs of cryofracturing of various substances,
ones. Due to their rigidity these molecules rotate etc. Various options of hydrogen liquefiers (up to
as whole units. Non-central interaction forces (dis- 104 l/h), helium liquefiers (100 l/h), and refriger-
persion, exchange, and multipole) result in certain ators which reach millikelvin temperatures using
orientations of these molecules in the crystal lat- 3 He dissolved in superfluid 4 He are manufactured.
tice. Phase transitions mainly arise from changes Cryogenic technology applies low-temperature
in that orientational order. Their collective exci- methods to other technical areas, such as super-
tations are phonons (both acoustic and optical); conductivity. Using the phenomenon of supercon-
librons (which appear only when some long range ductivity one may develop economically efficient
orientational order exists); bound intramolecular systems (e.g., large-scale magnetic devices with
vibrations; excitons, and their combinations. If superconducting magnets for nuclear magnetic
the central interaction dominates over the non- resonance imaging, accelerators, and thermonu-
central variety (e.g., in solid CD4 ), the breakdown clear installations). One may attain high specific
of long-range orientation order (orientation melt- power output for electric motors and generators,
ing) occurs within the solid phase. Regions of achieve very high sensitivities in measuring de-
short-range orientational order may still persist in vices using the Josephson effect, etc. Projects
orientationally disordered crystals. With increas- are in progress to build superconducting electric
ing temperature these regions shrink, and librons
power lines and transportation systems.
transform into weakly perturbed rotations. Of par-
ticular interest is solid oxygen, a molecular mag-
CRYOGENIC TEMPERATURES
netic substance with particular structural features
Low temperatures, in the range from 120 K to
and collective excitations (magnons, biexcitons)
below 1 K (see Low-temperature physics).
that result from the uncompensated electron spin
There exist various methods of deep freez-
of its paramagnetic O2 molecule.
ing, or reaching these temperatures: cooling a gas
Quantum crystals form a separate group of
cryocrystals (3 He, 4 He, H2 , HD, D2 and CH4 ). when throttling (Joule–Thompson effect), isoen-
The dynamics of the translational and rotational tropic gas expansion in a gas expander, desorption
motions of these molecules defies a classical de- cooling, expansion liquefaction, vapor pumping
scription, being controlled by the laws of quantum over a cooling agent, and so on. Combinations of
mechanics instead. two or more methods are often used.
The weak intermolecular interactions, small The liquefaction of nitrogen, hydrogen and
size, and light mass of these molecules all combine helium provides the principal refrigerants in tech-
to produce properties typical of cryocrystals: low nical and scientific research. Vapor pumping over
crystallization temperature, strength and thermal these liquids produces low temperatures in the
conductivity, and high plasticity, compressibility, range of 120–0.25 K (N2 ∼ 120–63 K; H2 ∼
vapor pressure, and thermal expansion coefficient. 20–10 K; 4 He ∼ 4–0.7 K; 3 He ∼ 3–0.25 K).
These refrigerants are employed in practice for
CRYOGENIC TECHNOLOGY cooling various electrical engineering devices (su-
Branch of technology involving the develop- perconducting magnets, power transmission lines,
ment, manufacture and use of instruments and electric generators, electric motors, transformers,
machines that produce and maintain low temper- etc.), electronic devices (radiation detectors, first
atures, and that can liquefy and separate various stages of amplifiers, computer elements, etc.) and
CRYOSTAT 235
other units. Deep freezing is used for gas blend CRYOSTAT (fr. Gr. κρυoς , cold, and σ τ ατ oς ,
separation, gas purification, cooling of food, and standing, unmoving)
for other purposes in medicine, construction, en- A device for cryostatting, i.e. for maintain-
gineering, and so on. ing a constant cryogenic temperature between 0
and 120 K. Cryostats are used to study physical
properties of materials, the phenomenon of super-
CRYOSORPTION (fr. Gr. κρυoς , cold, and Lat. conductivity, to cool photodetectors, and to pursue
sorbeo , absorb) other tasks at low temperatures. The simplest cryo-
The process of absorption of gas from the stat consists of two vacuum concentric bottles. The
environment by a strongly cooled solid or a cryo- inner bottle, where the specimen under study is
genic liquid called a cryosorbent. The removal placed, is filled with a cryogenic liquid boiling at
from the gas phase by the liquid cryosorbent is low temperature, e.g., with liquid helium (4.2 K).
called absorption. This process is explained by The outer bottle playing the role of thermal iso-
the finite solubility of certain gases in cryogenic lation is filled with a liquid boiling at a higher
liquids. A solubility coefficient α (in m3 per ton temperature, e.g., liquid nitrogen (77 K). Cryostats
of cryosorbent at 0 ◦ C and 760 mm Hg) is intro- are often used to cool and store the specimen at
duced. The value of α is a function of the partial a specific temperature within a wide range (from
room to a very low value, e.g., 4.2 K). Such a de-
pressure, P , and the amount of the absorbed gas is
vice is not a cryostat in a strict sense, or is not
given by the expression σ = αP . The coefficient
one at every temperature. However, if its minimum
α has a different dependence on the temperature
attainable temperature falls within the cryogenic
for different materials. It decreases with the tem-
range, it is commonly called a cryostat.
perature if gaseous H2 or He is absorbed by liquid In most cryostats the sources of cold that ab-
nitrogen; while for most gases absorbed by liquid sorb the heat removed from the specimen are
methanol it grows with the drop in temperature. liquefied gases. The most widely used are liq-
The absorption process is used in several sepa- uid nitrogen (N2 ), helium (3 He, 4 He), hydrogen
ration cycles. During cryosorption of vapors by (H2 ) and, less commonly, oxygen (O2 ). Special
porous solids capillary condensation may occur. cryostats used to cool photosensors sometimes
Cryosorption of vapors from certain multiphase contain solid noble gases. More recently compact
gas systems may proceed by way of freezing the refrigerators have been developed that can con-
atoms of highly volatile fractions into a solid tinuously pump a liquid or gaseous cooling agent
layer of a weakly volatile fraction precipitated at into a cryostat. Together the two units form a joint
a cooled surface. cryostatting system. To reach superlow tempera-
A special case of cryosorption is adsorption of tures (T < 1 K) complex multistage devices are
a material from the gas phase by a surface layer employed which achieve cryogenic temperatures
of a cooled adsorbent. The amount of gas, V , ad- of a specimen using liquid gases (3 He–4 He solu-
sorbed by a unit mass of adsorbent depends on the tion refrigerators), or through adiabatic demagne-
tization (see Adiabatic demagnetization cooling).
nature of the adsorbent, and that of the adsorbed
Depending on the type of study, the cryostat
substance, and it grows for lower temperatures and
may include optical windows (optical cryostat), a
higher gas pressures. The curve describing the de-
superconducting solenoid, a very high frequency
pendence of V on P at T = const is called the
lead, or other devices to probe the specimen in var-
adsorption isotherm. The amount of adsorbed gas ious ways.
is linear with pressure at low pressures, following Many types of cryostats are available, varying
Henry’s law: V = kP . in purpose, cooling technique, temperature range
As a rule, several processes proceed in parallel covered, permissible thermal load, accuracy, relia-
during cryosorption. Cryosorption is widely used bility, stability, character of probing, etc.
in systems of vacuum cryogenic evacuation (see Two main approaches have been followed in
Low-temperature pumping methods). building cryostats:
236 CRYOTRON
one may proceed from a physical model of a The structural unit in molecular crystals is
crystal envisaged as an atomic electron-dominated a molecule. Crystal chemistry of organic com-
system, to a geometrical model that looks upon a pounds considers the rules of dense packing of
crystal as an ordered arrangement of incompress- molecules, the relation between the symmetry of
ible balls. molecules and the symmetry of the crystal, and
A full crystal chemistry description includes the types of organic structures involved. Specific
specifying the size of the unit cell, the space group crystal patterns are identified in the structures of
of the crystal, its individual atom coordinates, dis- polymers, liquid crystals, and biological crystals.
tances between them, and the types of chemical The coordination number K (see Coordination
bonds that are present. In addition, the near en- sphere) and the type of coordination polyhedron
vironment of separate atoms, the characteristic characterize the chemical bonding of a given atom,
atomic groupings, thermal vibrations of the atoms, as well as the structure of a given crystal as a
etc., have to be taken into account. whole. For example, Be (with a few exceptions)
In terms of the type of their chemical bonding, and Ge form a tetrahedral environment (K = 4),
crystals may be divided into four major groups: Al and Cr have an octahedron (K = 6) for their
ionic crystals (e.g., NaCl), covalent crystals (e.g., coordination polyhedron, and Pd and Pt are often
diamond, silicon), metallic crystals (metals and square planar coordinated (K = 4). Low coordi-
intermetallic compounds), and molecular crystals nation numbers point to a significant role of di-
(e.g., naphthalene). rectional covalent bonds, and large K values (e.g.,
A certain structure is typical for each of these 6, 8, 12) indicate an appreciable role of ionic and
types of crystals. However, when thermodynamic metallic bonding. Bonds of various types coexist
conditions change these may sometimes alter in many crystal structures (graphite, MoS2 , etc.).
(polymorphism). As a rule, the simpler the chemi- Such structures are called heterodesmic structures
cal formula of the compound, the more symmetric in contrast to homodesmic ones, the latter fea-
its structure. Crystals with identical chemical for- turing uniform bonding (diamond, metals, NaCl,
mulae (in the ratios of constituent atoms) may crystals of inert elements), and the former mixed
have identical crystal structures (identical struc- bonding.
tural types), despite possible differences in their The laws of thermodynamics determine
bonding type (isostructure). Alkali halides of the whether or not a particular crystal structure will
NaCl type and certain oxides (e.g., MgO), as well form. The structure most stable at a given temper-
as several alloys (e.g., Ti–Ni), are isostructural. ature T is the one that has the lowest Helmholtz
Isostructural crystals with the same kind of bond- free energy F = U − T S, where U is the bond-
ing are called isomorphic (see Isomorphism). In ing energy of the crystal (energy needed to break
many cases a continuous sequence of solid solu- up the crystal into separate atoms or molecules)
tions exists between isomorphic crystals. at T = 0 K, and S is its entropy. The higher
The geometric model of crystals involves the the free energy the stronger the bonding in the
concept of effective radii of atoms, molecules and crystal. It is usually within 40–80 kJ/mol for crys-
ions (crystal chemical radii), so that their bonding tals with covalent bonding, is somewhat lower
distance is equal to the sum of such radii (additiv- for the molecular crystals, and is lowest for van
ity of crystal chemical radii). Using experimental der Waals bonding (4–15 kJ/mol). In practice,
data one can tabulate crystal chemical radii for all theoretically calculating the free energy and pre-
the types of bonding (see Atomic radii, Ionic ra- dicting the structure of crystals is only possible for
dius). The principal geometric concept in crystal comparatively simple cases. Such calculations can
chemistry is the theory of dense packing that ex- be carried out within the framework of the band
plains the positioning of atoms in certain metallic theory of solids. In some cases semiempirical ex-
and ionic structures graphically. For ionic struc- pressions for the potential energy of interaction of
tures one may talk about “holes” in the packing of atoms in crystals can yield fairly accurate results
larger anions that are occupied by smaller cations for certain types of bonding (see Interatomic inter-
of similar ionic radii. action potentials).
238 CRYSTAL CLASSES
Table 1. Meaning of the first (I), second (II) and third (III) terms in the International Symbols of the 32 crystallographic
point groups of the seven crystal systems (see figure).
I II III
into bars along cleavage plane), average degree CRYSTAL DETECTOR, crystal counter
of cleavage (cleavage facets accompanied by ran- An instrument to record and analyze spectro-
domly shaped facets), imperfect cleavage (frac- scopic properties of nuclear particles and photons.
tography patterns exhibit mostly macroscopically
The crystal counter is an electric capacitor (usually
irrational fracture regions), completely imperfect
parallel plate type), filled with a single crystal in-
cleavage (practically no cleavage: fracture is con-
sulator (e.g., diamond or CdS). A charged particle
choidal).
entering the dielectric produces ionization along
Cleavage is one of the characteristic proper-
ties of a crystalline material. It is helpful for the its path, and free current carriers appear which
determination of a crystal disorientation. Surfaces are driven to the capacitor electrodes by the ex-
of high perfection, exposed on splitting along the ternal electric field. A short electric pulse results,
crystal cleavage plane, are convenient for study. its amplitude proportional to the energy released
The phenomenon of crystal cleavage is used in by the particle in the crystal. After amplification
jewelry trades to prepare a gem for its crystal it is recorded by a counting device or amplitude
faceting. analyzer.
240 CRYSTAL DIFFRACTION SPECTROMETERS
Simplicity of design and maintenance, small clearly preferred directions form, as a rule, uni-
size (below 1 mm2 ) and the fact that certain crys- formly developed isometric polyhedra. The acicu-
tals, e.g., diamond, can function at high temper- lar crystals are typical of materials with needlelike
atures, make crystal counters convenient nuclear or chain structures; plate-like crystals are typical
radiation detectors that are widely used in physics, of materials with laminated structures, and so on.
dosimetry, various industrial applications, etc. In most cases, the only crystal faces that partic-
ipate in the crystal faceting are those parallel to
CRYSTAL DIFFRACTION SPECTROMETERS crystal lattice planes of the structure. The latter
A class of X-ray spectral instruments for study- possesses high reticular density, i.e. their unit area
ing absorption and emission spectra of matter maximizes the number of particles (Bravais law).
and the structural features of monocrystals. Crys- In an ideal case the faces of each simple form must
tal diffraction spectrometers may be single-axis have the same configuration and dimensions. The
(sample crystal), double-axis (collimator crystal shape of a real individual crystal is determined by
plus sample), and triple-axis (collimator, sample, the specific conditions of its growth: temperature,
and analyzer crystals) types. As the number of pressure, concentration, impurity atoms content,
etc. When characterizing the crystal faceting, the
elements in the instrument grows, its spectral dis-
terms form and habit are commonly used. The for-
persion and angular divergence of the radiation
mer term concerns a general view of a polyhedron
diminish, so that it becomes possible to better
(isometric, acicular, fibriform, etc.), while the lat-
expand the diffracted intensity over the angle.
ter serves to describe the outward appearance of
Single-axis diffraction spectrometers are only ca-
the individual crystal with the help of dominant
pable of measuring the integrated intensity of the
simple forms (octahedral, prismatic, dipyramidal,
diffraction. Double-axis instruments can record
and so on).
X-ray spectra or rocking curves for monochro-
matic radiation (i.e. dependence of diffracted in-
CRYSTAL FIELD, internal crystal eld, crystalline
tensity on X-ray angle of incidence). They also
electric eld
provide for antiparallel and parallel positioning of
The electric field at a point within a crystal,
the crystals so that the resulting dispersion is either
produced by its atoms or ions. This field deter-
doubled after the beam passes through the sam-
mines energy level splittings of atomic electronic
ple crystal, or it reverts to zero. The first position
configurations, and the periodic potential of the
is used for spectral studies, and the second (zero
crystal field determines the band structure of the
dispersion) for structural studies during which the
crystal for an ideal lattice corresponding to the
width and intensity of the diffraction maximum
crystallographic space group. An impurity ion (or
are measured. Triple-axis spectrometers permit the
local ionic center) has an energy spectrum deter-
separation of the coherent component of the in- mined by the local crystal field symmetry. The
tegrated intensity from the diffuse one. See also internal electric fields of defects bring about an
X-ray spectrometer. inhomogeneous broadening of the spectral lines
of resonant transitions. The line shape (profile)
CRYSTAL FACETING, crystal cut depends on the nature of the defects and their dis-
Outward shape of crystals specific for a given tribution within the crystal. To compute the shape
crystal class. Faceting conforms strictly to the of the resonance lines one can use statistical theory
point group symmetry relationships of a crystal of line shape, or possibly the method of moments.
(see Crystal symmetry), and only includes faces of Crystal eld theory is applied in various
certain categories, the so-called simple forms (see branches of physical science (optics, radiospec-
Crystal polyhedron). As a rule, the simple forms troscopy, etc.) to calculate the energy transitions
that are possible within a given point group sym- responsible for spectra. It considers various mod-
metry tend to display various combinations one els of the field, and basically accounts for the spa-
with another. The shapes of individual crystals are tial motions of electrons and nuclei. Energy level
closely related to their structure. Materials without splittings produced by the crystal field generally
CRYSTAL FIELD THEORY 241
lie in the optical frequency range (1013 –1015 Hz). where q is the quadrupole moment of the nu-
Accounting for the spin variables along with the cleus with a spin I 1. The crystal field gradient
spatial degrees of freedom results in an additional for cubic symmetry is zero. When calculating a
splitting or shifting of the levels. The crystal field crystal field gradient one has to take into consider-
affects the spin through the spin–orbit interaction. ation the fields of point charges (lattice summation
Spin level splittings may be treated using a frame- carried out by the Ewald method), and the multi-
work similar to crystal field theory called the spin pole moments of ions and conduction electrons.
Hamiltonian technique. The energy level spacings To take into account the polarization of the full
associated with electron spin interactions lie in the electron shells of an ion in the crystal e ld, its
frequency bands of electron paramagnetic reso- field gradient at the nucleus is written in the form
nance (108 –1011 Hz), while those resulting from (1 − γ )Vαβ , where γ is the antiscreening factor.
the nuclear spin are 2 to 3 orders of magnitude For most ions the value of γ is within the range
smaller, the same as the frequency range of nu- from −10 to −100.
clear magnetic resonance.
CRYSTAL FIELD SPLITTING
CRYSTAL FIELD GRADIENT See Bethe splitting.
A symmetric second-rank zero trace tensor
with components equal to the second derivatives CRYSTAL FIELD THEORY (H. Bethe, 1929)
of the potential V (r) of the electric field at Theory describing electronic states of ions with
the site of an atom or ion in a crystal: Vαβ = unfilled d- and f -shells in crystals. The scope
involves such areas as optical spectroscopy, mag-
(∂ 2 V /∂xα ∂xβ )|r=0 . The crystal field gradient de-
netic properties, and mechanisms of the electron–
fines the energy H of the quadrupole moment
phonon interaction, in crystals containing ions of
Qαβ of a nucleus (see Quadrupole), located at the
the three transition metal groups, the lanthanides,
site:
(2) and the actinides. Crystal field theory treats the
H= Qαβ Vαβ = Qm V−m , splitting of electronic energy levels of localized d-
αβ m and f -electrons by an effective internal crystalline
field. The Hamiltonian is solved using perturba-
(2)
where Qm are the components of the tensor op- tion techniques for a single configuration, with the
erator of the nuclear quadrupole moment: states of a single electron in the central field of
a free ion taken as the zero-order approximation.
1
V0 = Vzz , The perturbation operator defined in the space of
2 the wave functions of the (nl)N -shell equals the
1/2
1 sum of the energies of the electrostatic interaction,
V±1 = ± (Vxz ± iVyz ), Hee , the spin–orbit interaction, H so , and the en-
6
1/2 ergy of the electron in the crystal field, Hc . The
1 Hamiltonian, Hc is averaged over the radial distri-
V±2 = (Vxx − Vyy ± 2iVxy ).
24 bution of electron density, and has the form:
The crystal field gradient of the tensor Vαβ in a
N
(k)
system with two principal axes is characterized Hc = Bqk Cq (i),
by two independent parameters eq = Vzz and η = i kq
(Vxx − Vyy )/Vzz , where |Vzz | |Vyy | |Vxx |, (k)
where Cq (i) are spherical operators in the space
the dimensionless asymmetry parameter η is lim-
of angular variables of the ith nl-electron, and
ited to the range 0 η 1, and the Hamiltonian
Bqk are the crystal field parameters, quantitative
of the quadrupole interaction assumes the form
characteristics of the potential of the electric field
−1 at the paramagnetic ion site in an equilibrium
H = e2 qQ 4I (2I − 1)
configuration. They have dimensions of energy,
× 3Iz2 − I (I + 1) + η Ix2 − Iy2 , are generally complex, and satisfy the relations
242 CRYSTAL FIELD THEORY
Bqk ∗ = (−1)q B−qk . The number of independent for constructing the matrix of the Hc operator
parameters depends on the symmetry of the crystal depend on the choice of the basis for the zeroth ap-
field, while the structure of the operator Hc is de- proximation (e.g., single electron wave functions,
termined by the choice of local coordinate system. functions of terms |γ LSML MS , or functions
For example, a cubic crystal field has the octahe- of multiplets |γ LSJ MJ ). Piezospectroscopic or
dral point group of symmetry Oh . In the system of electric field effects in the spectra of paramag-
Cartesian coordinate axes coinciding with the C4 netic crystals can result from changes in the crystal
symmetry axes the Hamiltonian of the electron in- field during corresponding external actions on the
teraction with the cubic crystal field is given by lattice. Modulation of the crystal field by lat-
tice vibrations is the principal mechanism of the
(4) 5 (4) (4)
H = B4 C 0 + C + C−4 electron–phonon interaction.
14 4 Crystal field models help with computational
(6) 7 (6) (6)
techniques. The simplest and historically the first
+ B6 C 0 + C4 + C−4 . point charge model envisaged the crystal lattice as
2
a collection of point charges Qλ e located at lattice
The even component of the crystal field potential sites (λ labels the site), with the charges them-
is the same in the case of tetrahedral (Th , Td ) co- selves prescribed by the valences of the respective
ordination. Often the cubic symmetry component ions. To determine the field of point charges one
dominates at a low-symmetry site. The parameter has to calculate the lattice sums:
10Dq = (10/21)B4 equals the difference between 1/2
the two d-electron energies in a crystal field of cu- 4π
k
Bq = −(−1) q e2 r k
bic symmetry (doublet eg and triplet t2g states), 2k + 1
and provides a measure of the field “strength” for
ions with partially filled d n -shells. The f n -shell Qλ Y−q
k (θ ϕ )
λ λ
× ,
multiplet splittings in cubic symmetry are defined
λ Rλk+1
by the parameters B4 and B6 , and the structure
of the wave functions of the respective states de- where Yqk are spherical harmonics, rλ , θλ , ϕλ are
pends on only a single parameter (x), proportional the spherical coordinates of site λ (the paramag-
to the ratio B4 /B6 . The effects produced by the netic ion is at the coordinate origin), and r k
spatial correlation of electrons, and the change of are the moments of the electron radial density
the electrostatic interaction parameters in the crys- in the unfilled shell. Discrepancies of an order
tal field, are described by a two-term operator: of magnitude often exist between calculated and
(2)
(k) (k k )k measured crystal field parameters. To account
Hc C (k1 ) (i) × C (k2 ) (j ) Bq Bq 1 2 ,
for this one may consider changes in the radial
i>j
density of the d- and f -electrons in the crystal
that has a complex tensor structure. The parame- field, fields of point multipole moments (dipoles,
(k k )k
ters Bqk , Bq 1 2 can be evaluated from optical quadrupoles), the distribution of charges of lattice
or magnetic measurements, or computed within ions, screening of f -electrons by the outer filled
various crystal field models. Quantum-mechanical electron shells, etc. The electronic structure of se-
calculations taking into account the electron struc- lected complexes that include the paramagnetic
ture of the nearest neighbors of a paramagnetic ion and its nearest neighbors (ligands) immersed
ion are sometimes called ligand field theory. When in the crystal is treated by cluster models (see
calculating the energy spectrum of the (nl)N elec- Cluster). Electron energies in the complex are
tron shell one should distinguish between strong calculated using a one-electron Hamiltonian that
crystal fields (ions of 4d and 5d transition groups, operates in the space of molecular orbitals (LCAO
Hc > Hee > Hso ), intermediate fields (ions of method, see Molecular orbital), or in the space of
the 3d iron group, Hc ∼ Hee > Hso ), average mixed configurations. Charge transfer from the
fields (Hee > Hc > Hso ), and weak fields (rare- ligands to the paramagnetic ion is also taken into
earth elements, Hee > Hso > Hc ). Various options account (the Heitler–London approximation). To
CRYSTAL LATTICE 243
simplify estimates of matrix elements of the one- electron in an odd crystalline field corresponds
2p+1 2p+1
electron Hamiltonian (see Single-electron approx- to the Hamiltonian H = Bk Ck , de-
imation) one uses either the Mulliken–Wolfsberg– fined in the subspaces of one-electron functions
n l ml |, |nlml with their parameters Bk
Helmholtz method (R.S. Mulliken, M. Wolfsberg, 2p+1
G. Helmholtz, 1952) or the local exchange poten-
(where 0 p (l + l − 1)/2) different from zero
tial (Xα -technique). The dominant role in forming when the point group of symmetry of the ion does
the energy spectrum of the cluster and the cor- not include inversion. The odd components control
responding potential of the effective crystal field the intensities of the dipole transitions between the
involves effects produced by the overlapping of states of an unfilled (nl)N -shell (forbidden for free
the spatial distributions of charges of the ligands atoms), as well as the linear electric eld effects
and the paramagnetic ion. Among such effects are in optical spectra and the spectra of paramagnetic
the Kleiner correction to the Coulomb field of the crystals.
ligands, the exchange interaction, the mutual non- When perturbation theory is applicable one can
orthogonality of wave functions of the ligands and introduce an effective operator (H A + AH )/Δ in-
the metal ion, covalency, virtual excitations of the stead of the odd single particle operator A in the
cluster accompanied by charge transfer from the space of states of the basic electron configuration.
ligands to the paramagnetic ion, followed by the Here Δ is the average energy difference between
formation of holes in its filled electronic shells. the excited and ground state configurations. Addi-
One may explicitly determine the interrela- tional terms in the effective operator of the dipole
tions between the composition and structure of the moment proportional to the squared electron ra-
crystal lattice, on the one hand, and the energy dius (r 2 ) are called pseudo-quadrupole moments.
spectrum of the paramagnetic ions on the other, Calculating the Stark structure of the spectrum of
using semiphenomenological models that involve the unfilled shell, the effective even field H 2 /Δ
a limited number of parameters determined from is allowed along with the even component of the
experimental data. The energy of an nl-electron crystal field. The odd components, induced by odd
in an axially symmetric field of the ligand λ is
k (k)
vibrations, define the intensities of the electron–
equal to Hc(λ) = k B 0 (Rλ )C0 (quantization vibrational transitions between the states of the
axis parallel to r λ ). Assuming that individual lig- same parity for ions in the crystal lattice sites, the
and fields may be summed, and using the power latter also serving as the centers of symmetry.
k −t
series approximation B 0 ∝ Rλ k , one may ex-
press the terms of the crystal field through four CRYSTAL GRAINS
(k = 2, 4 for l = 2) or six (k = 2, 4, 6 for See Polycrystal.
l = 3) parameters of the “superposition model”
k
B 0 , tk . To parameterize the spectra of paramag- CRYSTAL, IDEAL
netic ions using the model of angular overlap and See Ideal crystal, Perfect crystal.
the model of exchange charges it is sufficient to
k
represent the functions B 0 (Rλ ) by linear combi- CRYSTAL LATTICE
nations of squared overlap integrals of the wave A geometrically ordered arrangement of atoms
functions of the d- (or f -) electron with the ligand (ions, molecules) typical of the crystalline state of
wave functions. Satisfactory models are still lack- matter (see also Lattice). The geometry of an ideal
ing for rare-earth metals and actinide compounds. crystal involves a spatial periodicity (translational
The so-called odd components of the crystal field symmetry, see Crystal symmetry) characterized by
are treated separately since they mix together the three noncoplanar basis vectors a j (j = 1, 2, 3)
states of different parity that correspond to the such that the displacement of the lattice by a vec-
ground state ((nl)N , 1 = N < 2(2l + 1)), and tor n1 a 1 + n2 a 2 + n3 a 3 (n1 , n2 , n3 are integers)
to the excited ((n l )4l +1 (nl)N+1 , (nl)N−1 (n l )) brings it into itself. If every a j is selected to be
electronic con gurations of the ions of rare-earth the shortest possible displacement in its respective
and transition metals. The energy of a localized direction, the parallelepiped formed by the three
244 CRYSTAL LATTICE BASIS
basis vectors is called the unit cell or the primi- • a = b = c, α = β = γ = 90◦ , orthorhombic sys-
tive parallelepiped. By simply replicating the unit tem;
cell in space one may generate the entire crystal • a = b = c, α = β = 90◦ , γ = 90◦ , monoclinic
structure. The choice of the set of a j vectors is system; and
not unique; yet the unit cell volume is independent • a = b = c, α = β = γ , triclinic system.
of such a choice. If a unit cell can be chosen so
that it contains only a single atom then it is called CRYSTAL LATTICE DYNAMICS
a primitive cell. All other unit cells are called The branch of solid-state mechanics, which
nonprimitive cells. A complex crystal lattice may studies the internal movements within a crystal,
be envisaged to consist of one or several Bravais taking into account the discreteness of its atomic
lattices fitted into each other, each such Bravais structure. It includes the classical and quantum
lattice being built of atoms of a single type. The mechanics of collective movements of atoms in
presence of a crystal lattice explains the phenom- an ideal crystal, the theory of interactions of the
enon of anisotropy of crystals and crystal faceting. crystal with penetrating radiation, and also the dy-
Experimentally, crystal lattices can be identified namics of imperfect crystals.
and classified by X-ray, electron, and neutron dif- Due to the strong chemical bonding the internal
fraction techniques. motion cannot be described as the superposition
of independent displacements of separate atoms.
CRYSTAL LATTICE BASIS, unit cell basis If the displacements are small and the harmonic
Set of atoms and their coordinates with the approximation applies then normal vibrations oc-
same location at each point of a Bravais lattice. By cur, each of which involves the participation of all
adding the basis to the Bravais lattice, one obtains the atoms. The normal vibration has a form of a
a crystal lattice structure. plane wave characterized by the wave vector k,
expressed within the first Brillouin zone, which de-
CRYSTAL LATTICE CONSTANTS termines the direction of propagation of the wave
The magnitudes of three noncoplanar transla- front, with wave length λ = 2π/|k|, and a polar-
tions (a, b, c), by which the Bravais lattice may ization vector which indicates the direction of the
be constructed for a given crystal. The lattice is displacement of atoms within the wave. During
to satisfy three main conditions: (1) the symme- the course of a normal vibration all the atoms of
try of the lattice unit cell must correspond to the the crystal vibrate harmonically near their equilib-
microscopic crystal symmetry; (2) the unit cell is rium positions at the same frequency ω which is
to possess the highest possible number of right determined by the dispersion law that connects the
angles (or the highest possible number of equal an- values of ω and k:
gles and equal sides); (3) the unit cell is to have a ω2 = fα (k), α = 1, 2, . . . , 3q,
minimal possible volume (with due regard for the
conditions (1) and (2)). The set of crystal lattice where α denotes the branch of the dispersion law
constants (a, b, c) and the angles between them: and q is the number of atoms in the crystallo-
α (between b and c), β (between a and c) and graphic unit cell.
γ (between a and b), is called the crystallographic In three dimensions there are at least three
basis. There are seven types of crystallographic branches of vibrations. In the long-wave approx-
basis sets that correspond to the seven crystallo- imation (λ
a, a is the interatomic distance)
graphic systems, as follows: all the atoms in the unit cell vibrate with the
same phase, corresponding to acoustic vibrations
• a = b = c, α = β = γ = 90◦ , cubic system; (Fig. 1(a)). In this limit λ
a they are trans-
• a = b = c, α = β = γ = 90◦ , tetragonal sys- formed into the usual sound waves in a solid
tem; with the linear dispersion law ω = cs k, s = 1,
• a = b = c, α = β = 90◦ , γ = 120◦ , hexagonal 2, 3. Acoustic vibrations cover the frequency band
system; from 0 to 1013 Hz. At high frequencies the dis-
• a = b = c, α = β = γ = 90◦ , trigonal system; persion law is no longer linear. However, in the
CRYSTAL LATTICE DYNAMICS 245
Fig. 1. Diagram of a long-wave vibration of a one-dimensional crystal: (a) acoustic mode; (b) optical mode.
Debye model it is assumed that acoustic vibrations Many lattice modes can be excited simultane-
satisfy the linear dispersion law at all frequen- ously with different amplitudes. The total num-
cies within the range 0 < ω < ωD , where the ber of possible independent vibrations equals the
Debye frequency ωD is the maximum possible vi- number of mechanical degrees of freedom of all
brational frequency (1013 Hz), and serves as the the atoms in the crystal, and their distribution in
most important parameter of the spectrum of crys- frequency is determined by the frequency distri-
tal vibrations. In a complex crystal lattice (q > 1), bution function ν(ω) (phonon density of states).
3q − 3 branches of optical vibrations are possible, The form of ν(ω) strongly depends on the di-
and at k = 0 the center of mass of the unit cell mensionality of the crystal. In a three-dimensional
remains stationary with the motion reduced to rel- lattice at low frequencies (ω ωD ) for each
ative displacements of atoms within the unit cell acoustic branch ν3 (ω) ∝ ω2 /ωD 3 . For increasing
(Fig. 1(b)). In the limit k → 0 the optical vibration frequency the behavior of ν(ω) changes: it tends to
frequencies ω(0) are different from zero. Ordinar- zero at the band edges, and within the bands it ex-
ily the frequency bands of optical vibrations are hibits Van Hove singularities (Fig. 2). The overall
located above those of acoustic vibrations, with a vibrational density of states is obtained by sum-
forbidden region, or band gap, separating them. It ming the distributions of the individual branches.
is also possible for acoustic and optical frequency The quantum dynamics of a crystal lattice
bands to overlap. is well described by the harmonic approxima-
There are crystals in which some optical fre- tion: to each wave of normal vibration with fre-
quencies strongly depend on external parameters quency ω and wave vector k there is a corre-
(temperature, pressure, magnetic field, etc.) and sponding quasi-particle with the energy ε = ω
at specific values of these parameters they may and quasi-momentum p = k, with the number of
tend to zero (soft modes). As a result, static rel- quasi-particles proportional to the wave intensity.
ative displacements of atoms can occur, i.e. the These quasi-particles are the elementary excita-
unit cell is restructured by undergoing a structural tions of crystals and are called phonons; they
phase transition. Optical vibrations of ionic crys- obey Bose–Einstein statistics and form an ideal
tals can strongly interact with an electromagnetic gas in the harmonic approximation. At sufficiently
field producing coupled polarized vibrations of the low temperatures when the crystal is weakly ex-
lattice and the field. This phenomenon permits the cited mechanically, its thermodynamic properties
excitation of optical vibrations of ionic crystals arise from these elementary excitations, and in
by incident oscillating electromagnetic fields (e.g., particular the lattice portion of the crystal energy
light waves), generally in the IR frequency range, coincides with energy of the phonon gas.
which is the reason for calling these frequencies The quantum nature of the crystal motions
optical. is manifested through the presence of so-called
246 CRYSTAL LATTICE VIBRATIONS
(1)
Fig. 2. Plot of the frequency distribution function in a three-dimensional crystal for the acoustic (0 < ω < ωm ) and
(2) (3)
optical (ωm < ω < ωm ) branches, where ω1 , ω2 , ω3 , ω4 are the frequencies of Van Hove singularities.
in a unit volume, n. The rate of formation of pre- granularity roughens for larger I and smaller v.
cipitates I , i.e. the number of precipitates formed See also Monocrystal growth.
in a unit volume per unit time, may be determined
within the formalism of the thermodynamic theory CRYSTALLIZATION CENTER
of their formation. One has to multiply the density See Crystal nuclei.
of precipitates of a critical radius by the number
of atoms (molecules) striking their surface per unit CRYSTALLIZATION FROM THE GAS (VAPOR)
time: PHASE
⎧
⎪
⎪ 4πrc2 pn(2πmk B T )−1/2 , The formation of crystals by the transformation
⎪
⎪ of a material from its gas (vapor) phase to a solid.
⎨ precipitatesform from gas phase,
I= E When crystals grow from the vapor an adsorption
⎪ 2
4πrc nνa exp − ,
⎪
⎪ layer differing from the crystal in composition and
⎪
⎩ kBT
precipitates form from condensed phase, properties appears at the crystal–medium interface
(see Adsorption). Moreover, surface diffusion of
where m is the mass of the molecule, a is the atom- the adsorbed particles at the growing faces plays
to-atom distance, ν is the frequency of atomic a decisive role there.
oscillations, and E is the activation energy for par- Crystallization from the gas phase has been
ticles adding to the precipitate. To account for the used to produce massive monocrystals, plates,
evolution of the system of precipitates of radii be- bands, acicular crystals, thread-like crystals, and
low rc in the initial phase one adds an extra term epitaxial lm s. Condensation techniques and
in the expression for I , namely: chemical reactions are also employed to grow
crystals. When using the first it is assumed that
4rc α 1/2
z= , material reaches the growing crystal as an atomic
9k B T nc or molecular vapor, or as associations of such
where nc is the number of particles in a precipi- particles. Depending on the technique used to
tate of a critical radius. With decreasing temper- bring the material to the crystallization zone one
ature (at relatively small overcooling) I grows as may speak about molecular beams in vacuo (see
precipitates grow, since their formation energy di- Molecular beam epitaxy), cathode sputtering (see
minishes; on the other hand, the energy of thermal Sputtering), volume vapor phase in an isolated
motion and the mobility of particles in the ini- system, and crystallization in an inertial gas flow.
tial phase simultaneously decrease, so that the rate To obtain high quality crystals and films chemical
at which they add to the surface of the precipi- reaction techniques (e.g., as chemical transport,
tate decreases as well. Competition between these decomposition of compounds, chemical synthesis
two factors produces a maximum of I at a cer- in the vapor phase) appear to be quite efficient.
tain temperature, after which it starts to decrease In certain cases when filaments or massive crys-
at lower temperatures. The value of I is remains tals are grown (as well as epitaxial films), the
negligibly small at comparatively small supercool- vapor–liquid–crystal transport mechanism is em-
ing (or supersaturating) of the initial phase, so ployed. Vapor material reaches the growing crystal
that very few precipitates form in volumes of ap- through a liquid layer of different composition
preciable size. In other words, the initial phase formed either unintentionally or intentionally on
remains in a metastable state. The range of re- the growing surface by introducing impurities into
spective supercooling (supersaturating) is called the crystallization zone. It may result from devia-
the metastability range. tions from a stoichiometric composition, or present
As a result of the mass crystallization a poly- a liquid phase intentionally formed at the substrate
crystal forms, with fine crystals growing from as a smelt droplet of a prescribed composition.
separate precipitates (crystallites) that fill the vol- The phase-to-phase energy of the crystal–vapor
ume of the initial phase with their crystallographic interface is almost an order of magnitude less
axes randomly oriented. The structure of the poly- than the respective energy of the crystal–vapor in-
crystal depends on the values of I and v. Its terface. Two-dimensional precipitates (embryos)
CRYSTALLOGRAPHIC INDICES 249
Triclinic (anorthic) 2 1 Ci
Monoclinic 13 2/m C2h
Orthorhombic 59 mmm D2h
Trigonal 25 3m D3d
Hexagonal 27 6/mmm D6h
Tetragonal 68 4/mmm D4h
Cubic (isometric) 36 m3m Oh
a vector to that site is denoted by [xyz]. A vec- coincides with the concept of syngony for every
tor parallel to the x coordinate axis is in the [100] crystallographic system, with the exception of the
direction, etc. Crystallographic planes are denoted hexagonal and the trigonal systems which together
by what are called Miller indices (hkl). Consider comprise the hexagonal syngony or family. Ta-
a plane which intersects the a, b, c axes at the ble 1 outlines the crystallographic classification
points h a, k b, l c, and write down the recip- scheme, and lists a high-symmetry point group for
rocals of these intercepts 1/ h , 1/k , 1/ l . The each system.
Miller indices are the smallest three integers h, k, l
having these same ratios. For example, if h = 4, CRYSTALLOGRAPHIC TEXTURE
k = 1 and l = 2 then (hkl) = (142). The three See Texture.
integers which define orientation directions in a
crystal, and the positions of its verges, are called CRYSTALLOGRAPHY (fr. crystal, and Gr.
the Weiss indices, and they are written in the form γραϕω, am writing)
[p1 p2 p3 ]. As for the hexagonal system and the The science describing crystals and the crys-
trigonal system, their directions and planes are talline state of matter. Six basic branches are
sometimes presented in a four axes system (three recognized within crystallography: geometric
equivalent directions are then positioned at 120◦ crystallography (crystal symmetry, crystal struc-
to each other in the basal plane of the crystal). The tures, and crystal morphology); crystallogenesis
Bravais indices [mnrp] and (hkil) satisfy the con- (crystal formation and growth, see Crystallization);
ditions r = −(m + n), i = −(h + k). When using crystal physics; crystal chemistry; applied crys-
the Bravais indices one may clearly see the equiva- tallography (methods and techniques for crystal
lence of the respective directions and planes in the growth and use in technical applications); space
crystal. crystallography or growing crystals in zero gravity
(in orbit) and studying their properties. Geomet-
CRYSTALLOGRAPHIC SYSTEM ric crystallography gave birth to the hypothesis
Crystallographic classification scheme for the of ordered three dimensional positioning of parti-
230 space groups that takes into account their cles (atoms and molecules) composing the crystal
characteristic symmetries (see Crystal symmetry), which form the crystal lattice. The mathematical
and uses the ratios of their symmetry axes. There apparatus of crystallography is based on Euclid-
exist seven independent combinations of the ro- ean geometry, group theory, and tensor calculus.
tation symmetry axes of the nth order (n = 1, Structural crystallography studies the atomic
2, 3, 4, 6), and seven crystallographic systems and molecular composition of crystals using X-ray
corresponding to them: cubic system, hexagonal structure analysis, electron diffraction analysis,
system, trigonal system, tetragonal system, or- and neutron diffractometry, all based on the dif-
thorhombic system, monoclinic system and triclinic fraction of waves and particles within the crystals.
system. The concept of crystallographic system Optical spectroscopy, electron microscopy, and
CRYSTAL MORPHOLOGY 251
resonance methods, etc., are sometimes used. As CRYSTAL MORPHOLOGY (fr. Gr. μoρϕη ,
a result, the crystal structures of over a hundred appearance, shape, and λoγ oς , word)
thousand chemical substances have been deter- Outward shape of crystal that grows from a
mined. When studying the processes of formation solid, liquid or gaseous phase. It depends on the
and growth of crystals one uses the basic prin- nature of the substance, on growth conditions, and
ciples of thermodynamics, the laws of r st-order other factors. During crystallization from the melt
phase transitions, and of surface phenomena. One with a small deviation from equilibrium, crystals
also accounts for the interaction of crystals with of some materials grow in the shape of regular
their ambient environment, and for anisotropies. crystal polyhedra (Fig. 1), while crystals of other
Industrial processes used to grow diamonds, ru- materials grow in a shape approximating a sphere
bies, and other crystals have been perfected (see (Fig. 2(a)). A faceted shape (see Crystal faceting)
Synthetic monocrystals), and provide a basis for is typical of materials that exhibit a high entropy
quantum and semiconductor electronics, optics, of fusion
S = L/T > 4R (L is the latent heat
acoustics, etc. Crystallography also studies vari- of melting, R is the gas constant), and a high
ous perturbations in an ideal crystal lattice, such as anisotropy of surface energy. A rounded shape
defects and dislocations, that originate during the is typical of metals and other materials that ex-
course of crystal growth due to various external hibit a low entropy of fusion (
S < 4R) and a
influences, and influence many properties of crys- low surface energy anisotropy. Metals grown from
tals. The structure and properties of various ag- a gaseous phase or from solution may exhibit a
gregates formed from microcrystals (polycrystals, faceted shape. A smooth shape loses stability and
textures, ceramics), as well as substances with develops bulges (Fig. 2(b)) at increased overcool-
atomic ordering similar to that found in crystals ing of the melt (in metals, this process occurs
(liquid crystals, polymers), are also studied. Con- at smaller values of overcooling, than in crystals
with a high entropy of fusion). The presence of
sistent symmetrical and structural patterns stud-
impurities intensifies this process owing to emerg-
ied within the framework of crystallography find
ing structural overcooling. Thermal and structural
their application when treating general structural
overcooling first results in the formation of bulges
features and properties of various non-crystalline
(cellular morphology), which then evolve into
states of condensed matter, such as amorphous
dendrites (Fig. 2(c, d)). Under sufficiently high
solids and liquids, polymers, macromolecules, su-
overcooling, the dendritic shape degenerates into
permolecular structures, etc. (generalized crystal-
a needle one without lateral branches. When crys-
lography). Studies of biological crystals have been
tallization takes place from a viscous medium,
successful in determining the structures of large
protein and nucleic acid molecules.
Modern crystallography typically concentrates
on the atomic and defect structure of crystals,
processes of their formation and growth, and
searches for new properties of crystals. These
problems involve comprehensive tasks aiming at
discovering new materials with important physical
properties. Results obtained from crystallographic
studies are routinely applied to physics, mineral-
ogy, chemistry, molecular biology, etc.
CRYSTALLOGRAPHY, INTERNATIONAL
TABLES
See International Tables for Crystallography.
CRYSTALLOGRAPHY, TWO-DIMENSIONAL
See Two-dimensional crystallography. Fig. 1. Crystal polyhedron.
252 CRYSTAL NUCLEI
culus; it is based on the fact that the period of equation together with the tensorial relation be-
a crystal lattice (∼1 nm) is much shorter than tween the vectors D and E. Thus, we get the
the wavelength of light (∼500 nm), and there- dispersion equation, which corresponds to a ho-
fore, a crystal can be considered as a homogeneous mogeneous system of three linear equations for the
anisotropic medium. This is expressed by a tensor components of the vector E:
form of material equations connecting the electric
ux density Di and magnetic ux density Bi with
εij (ω)ω2 2 δ + k k E (ω) = 0,
− k ij i j j
the corresponding fields Ej and Hj : Di = εij Ej c2
and Bi = μij Hj (here εij and μij denote the where k, ki , and kj are the magnitudes of the wave
polar tensors of rank two of dielectric constant vector and its corresponding components, δij = 0
and magnetic permeability). In most problems of for i = j , and δij = 1 for i = j . This system has
crystal optics, one considers nonmagnetic crystals a nontrivial solution E = 0 if and only if the de-
(μij = 1) and concentrates upon the properties of terminant in the parentheses is not zero. The latter
the tensor εij . In the absence of spatial dielectric condition leads to the Fresnel equation
dispersion, the dissipation of energy, and an ex-
ternal magnetic field, by the symmetry principles kx2 ky2 kz2
+ + = 0,
of kinetic coefficients, the tensor εij is symmet- v 2 − vx2 v 2 − vy2 v 2 − vz2
ric and real. Therefore, for a fixed frequency of
light ω, in anisotropic media, the directions of where v is the magnitude of the phase velocity, the
vectors D and E do not coincide. This deter- direction of which coincides with the direction of
mines the anisotropy of the optical properties of the wave vector k; vx , vy , and vz are the principal
crystals, i.e. the dependence of the phase veloc- phase velocities of the wave for the corresponding
√ directions of the vector D when the light propa-
ity v = c/n and of refractive index n = ε on the
direction. For the totality of all directions at an ar- gates along one of the principal directions.
bitrary point of a crystal, the magnitudes of n form The dispersion and Fresnel equations are qua-
an ellipsoidal surface, which is called the optical dratic with respect to k and v, i.e. at light propaga-
indicatrix. The symmetry axes of the optical in- tion in anisotropic crystals, there arise two waves
dicatrix form three mutually orthogonal principal (see Fig.) with different values of the magnitude
directions in the crystal (crystallophysical axes), of the wave vector and different phase velocities
along which the directions of vectors D and E (refractive indices) v1 and v2 coincide only along
coincide. In the rectangular Cartesian coordinate the directions of the optical axes. They correspond
system with axes coinciding with the crystallo- to two mutually orthogonal values of the vectors
physical axes, the equation of the optical indicatrix D and E.
is of the form If from some point of the crystal one locates all
the vectors corresponding to v1 and v2 in all the
x2 y2 z2 directions m, then the endpoints of these vectors
+ 2 + 2 = 1,
n2x ny nz form two surfaces called the surfaces of normals.
In uniaxial crystals, one of these surfaces is a
where nx , ny , and nz are the principal values of n. sphere (characterizing the ordinary ray v0 ), and
Crystals of a cubic system are optically isotropic, the other one is an ellipsoid of revolution (char-
i.e. their ellipsoid degenerates into a sphere; the acterizing the extraordinary ray ve ) tangent to the
crystals of medium systems are optically uniax- sphere at its intersection points with the optical
ial (the indicatrix is an ellipsoid of revolution); the axis. In biaxial crystals, both surfaces are ellip-
crystals of lower systems are optically biaxial (the soids intersecting at 4 points lying pairwise on
indicatrix is a triaxial ellipsoid). The optical axis two optical axes (binormals). The ray vectors form
of a crystal corresponds to the normal of a circular similar surfaces called ray surfaces (or wave sur-
section of the optical indicatrix. faces). Thus, in anisotropic crystals, in an arbitrary
The laws of propagation of plane harmonic direction m, there may propagate two plane waves
waves in dielectric crystals follow from the wave polarized in two mutually orthogonal planes. The
254 CRYSTAL, PERFECT
Crystal optics.
directions of vectors D 1 and D 2 of these waves different velocities. Moreover, in uniaxial crys-
coincide with axes of the ellipse of the central tals, the surfaces of normals do not osculate at the
section of the optical indicatrix by the plane or- points of intersection with the optical axis. The
thogonal to m, and the lengths of the ellipse axes ellipticity of waves brings about optical activity
determine their refractive indices. Consequently, of crystals, which is expressed, e.g., as gyration,
the optical indicatrix and surfaces of normals are electrogyration, Faraday effect, etc. The complex-
mutually dependent. ity of the tensor εij and the elliptical two-ray
The birefringence in the crystals because of refraction are also caused by light absorption in
the different refraction indices of two polarized crystals. The directions of by-pass of ellipses in
waves propagating in the same direction implies this case coincide. In such crystals, the absorp-
a path difference. If, in a convergent light, one tion of differently polarized waves is different, and
reduces the polarization of these waves into one occasionally circular dichroism takes place in op-
plane (using some polarization devices), this re- tically active crystals.
sults in the interference of the light. Such inter- Crystal optics adjoins crystallography, solid
fering conoscopic images allow one to investigate state physics, and mineralogy. With the appear-
crystal-optical properties of crystals. ance of laser technology, new branches of physical
In the presence of a spatial dispersion of first optics and optical tool making arose that are fun-
order with respect to k or an external magnetic damentally based on crystal optics: coherent and
nonlinear optics, holography, optoelectronics, in-
field, and also in magnetic crystals, the tensor εij
tegrated optics, etc. The control of optical (in
is complex and, in general, nonsymmetric; more-
particular, laser) radiation is based on the phenom-
over, εij (ω, k, H ) = εij (ω, −k, −H ). In partic-
ena of crystal optics.
ular, in such crystals two elliptically polarized
waves propagate with orthogonal axes of the el- CRYSTAL, PERFECT
lipse coinciding with the axes of the section ellipse See Perfect crystal, Ideal crystal.
of the optical indicatrix. The directions of by-pass
of vectors E are opposite in both waves. Along the CRYSTAL PHYSICS, physical crystallography
direction of the optical axis, and in cubic crystals, Branch of crystallography involving the phys-
the polarization ellipses degenerate into circles, ical properties of crystals and crystalline aggre-
i.e. the waves are circularly polarized and have gates of polycrystals, as well as changes in their
CRYSTAL PHYSICS 255
properties arising from external influences. Crys- directionally independent or isotropic. Crystals of
tal physics explains anisotropy of crystals, and intermediate symmetry (tetragonal, trigonal and
basic variations in the properties of crystals due hexagonal) have these properties characterized by
to the regular distribution of particles (molecules, the symmetry of their rotation ellipsoid, i.e. their
atoms, and ions) in the crystal lattice, and the types second-rank tensors have two independent compo-
of their interrelations. The mathematical apparatus nents, one of which describes the property along
of matrix algebra, tensor calculus, and group the- the principal axis, and the other along any di-
ory provides a quantitative description of the phys- rection orthogonal to this axis. To describe the
ical properties of crystals. Directionally indepen- property along an arbitrary direction one needs to
dent properties such as density are characterized know these two principal components plus the di-
by scalar variables. To describe physical proper- rection cosines of the direction.
ties involving relationships between two vectors Physical properties described by higher-rank
(e.g., between either polarization or electric cur- tensors are characterized by a larger number of
rent density and electric fields, etc.) or between parameters: e.g., elasticity properties, described
pseudo-vectors (e.g., magnetic induction and mag- by a fourth-rank tensor, need three such variables
netic field strength) one uses second-rank tensors in a cubic crystal, and two independent variables
(e.g., dielectric constant and magnetic permeabil- in an isotropic medium. The elastic properties of
ity tensors). Many physical fields in crystals (e.g., a triclinic crystal require 21 independent tensor
electric and magnetic fields, mechanical strains) components for their description. To determine the
are themselves tensor fields. The relation between number of independent components of tensors of
the physical fields and the properties of crys- higher rank (5th, 6th, etc.) one can make use of
tals may be described by tensors of higher rank,
the corresponding point group of symmetry. To
as is the case with the piezoelectric effect (see
obtain a complete description of the physical prop-
Piezoelectricity), electrostriction, magnetostriction,
erties of crystals and textures one may make use of
elasticity, photoelastic tensor, etc.
radiofrequency, acoustical, optical, and other tech-
It is convenient to represent dielectric, mag-
niques.
netic, elastic, and other physical properties of crys-
Crystal physics studies both phenomena spe-
tals in the form of characteristic surfaces such as
cific for anisotropic media (birefringence, rotation
ellipsoids. The radius-vector that traces out such a
of light polarization plane, direct and inverse piezo-
surface indicates the value of a particular property
along its given direction (see Crystallographic in- effects, electro-optic, magneto-optic, and piezo-
dices). The symmetry of any property of a crystal optic effects, generation of optical harmonics,
may not be lower than the symmetry of its ex- etc.), as well as properties more often associated
ternal shape (Neumann principle, F. Neumann). with isotropic media (electric conductivity, elas-
In other words, a symmetry group G1 that de- ticity, etc.). In crystals these latter properties may
scribes any physical property of a crystal includes also exhibit features resulting from the crystal
necessarily the symmetry elements of its associ- anisotropy.
ated point group G, that serves as its supergroup: Many phenomena within the scope of crys-
G1 ⊃ G. The symmetry elements determine the tal physics are related to changes in the crystal
orientations of the principal axes of characteristic symmetry under various thermodynamic condi-
surfaces, as well as the number of tensor compo- tions. The Curie principle may be used to pre-
nents that determine a given physical property. In dict changes in both the point and the space
many cases these are surfaces of ellipsoids with symmetry groups of crystals undergoing phase
their principal axes related to the symmetry axes transitions, e.g., entering ferromagnetic or ferro-
of the crystal (see Crystal classes). Crystals of low electric states (see Ferromagnetism, Ferroelectric-
symmetry have their physical properties described ity). Crystal physics studies various defects of the
by second-rank tensors with three principal val- crystal lattice (color centers, vacancies, disloca-
ues (plus the orientation of the tensor principal tions, grain boundaries, stacking faults, domains,
axes). Physical properties of cubic symmetry crys- etc.) and their effects upon the physical proper-
tals that are described by second-rank tensors are ties of crystals (plasticity, strength, luminescence,
256 CRYSTAL POLYHEDRON
electrical conductivity, mechanical Q-factor, etc.). the intercepts on the axes for different tilted faces
Among the tasks of crystal physics is the search of a crystal can always be expressed by rational
for new crystalline materials with specific prospec- numbers such as 1 : 2 √ or 2 : 3, and never by irra-
tive properties needed for practical technological tional ones such as 1 : 2 (law of rational indices).
applications. Other general macroscopic properties of crys-
talline matter include crystal homogeneity (inde-
CRYSTAL POLYHEDRON pendence of the choice of measurement point),
A monocrystal grain under equilibrium con- anisotropy of crystals (dependence on direction),
ditions (see Monocrystal growth) consisting of a and crystal symmetry (spatial transformations that
certain combination of faces, edges and vertices. bring the crystal into itself).
The number of faces, edges and vertices should There are 32 classes of morphological symme-
follow the Euler theorem (L. Euler): the sum of the try of crystals, each corresponding to its specific
number of faces and vertices equals the number set of symmetry elements: axes, planes, center of
of edges plus two (F + V = E + 2). The so- symmetry, etc. Three-dimensional atomic struc-
called simple forms participate in crystal faceting. ture periodicities are described in terms of a three-
These are polyhedra whose faces can be obtained dimensional space lattice (the crystal lattice), a
from one another via symmetry operations that are three-dimensional periodic system of points that
specific for the crystal class of the given crystal; may be formed starting from an arbitrary point by
they usually form various combinations with each simply repeating the lattice by a series of parallel
other. There are 47 geometrically distinct simple translations. There exist a total of 14 types of such
forms known; this number increases to 193 when lattices called Bravais lattices.
taking into account enantiomorphous (mirror re- The crystal appearance (crystal morphology) is
flection) pairs (see Enantiomorphism), or to 146 in characterized by the crystal shape and faceting, i.e.
case one accounts for differences in point symme- the set of faces bounding the crystal, and by the ra-
try. Only certain simple forms are permissible in tio of the sizes of these faces (see crystal faceting).
each point group. Specific forms of crystal poly- Such facets and their relative sizes characterize the
hedra depend on their structure and conditions of crystal morphology or crystal habit.
crystallization. The shapes of crystals reflect their crystal struc-
ture, and the nature of the interactions between
CRYSTAL, QUANTUM their atoms. In particular, these factors affect the
See Quantum crystals. shape of the forming crystal through kinetic phe-
nomena occurring at the surface of a growing
CRYSTALS (fr. Gr. κρυσ τ αλλoς , ice) crystal. Such phenomena depend on the level of
Solids with ordered three-dimensional periodic supersaturation, temperature, composition of the
spatial atomic structure. The formation medium ambient medium, thermal conductivity, diffusion,
may be an overcooled vapor or liquid of a pure and the presence of impurities. The shape ac-
substance, a supersaturated solution or flux of quired by the crystal during its growth (the growth
one material in another (the solvent), or a crystal shape) is very sensitive to the conditions of crys-
medium of a polymorphous modification or com- tallization, and reflects the mechanism of growth.
position different from that of the newly formed Therefore, the growth shape implies certain con-
crystal. Some features of crystals depend on their clusions beyond the conditions of growth. During
formation process. If the material is supplied the artificial growth of crystals the crystallization
evenly and uniformly from the ambient environ- parameters may be chosen accordingly. Although
ment then a polyhedral shape results, limited by a typical crystal shape is a polyhedron, rounded
flat faces intersecting each other along straight growth shapes or strongly perturbed shapes may
edges. These faces, often mirror smooth in per- form under certain ambient conditions (thread-like
fect crystals, are tilted toward each other at certain crystals, plates, dendrites, etc.). Some crystals can
angles characteristic for the particular crystal type form in two different modifications, a right-hand
(law of constancy of angles). The ratios between and left-hand one (enantiomorphism). Substances
CRYSTAL STABILITY CONDITIONS 257
are found that may crystallize in the shape of other structures. Substituting crystal atoms (sub-
twins (twinning of crystals). In cases when crystal stitutional impurities) or imploding between the
faceting is suppressed by the limited availability atoms of the crystal (implosion impurities) can
of feeding material, forced growth shapes develop. produce point defects (see Impurity atoms). Other
The crystal can then acquire the shape of the point defects perturbing the lattice periodicity are
surface limiting its growth (e.g., crucible walls, vacancies and interstitial atoms.
smelting isotherm in a smelt, an isoconcentration The level of deficiency of the crystal structure
saturation surface in a gas phase). Most monocrys- affects both crystal growth (e.g., the dislocation
tals are the result of forced growth. Technologies mechanism of growth) and various crystal proper-
for growing crystals with forced growth shapes ties that are structure sensitive (plasticity, strength,
find wider and wider applications. Almost all electrical conductivity, photoconductivity, light ab-
semiconductor crystals are produced that way, as sorption capacity, etc.). Therefore, to grow crystals
well as most piezoelectric, ferroelectric, and opti- with prescribed properties, one must intentionally
cal crystals (see Monocrystal growth). introduce desired defect structures. Examples are
If crystal generation and growth occur simul- donor and acceptor impurities which determine
taneously throughout the whole volume a poly- the electrical properties of semiconductors. Added
crystal forms, that is an aggregate of randomly impurities (doping) are widely used to produce
oriented fine crystals of various sizes and shapes semiconducting crystals with various types and
called crystallites or crystal grains. Sometimes levels of conductivity, to produce laser crystals
crystallites follow a certain predominant orienta- with a prescribed spectral transmission band, etc.
tion indicating the presence of a texture. During The development of electronics, quantum elec-
mass crystallization when embryos appear only in tronics, computer technology, and various other
certain regions of the generating medium, the col- industries is directly linked to using crystals and
lective growth of crystals ensues, characterized by complex crystal structures based on them (e.g.,
the interaction of the growing crystals with each epitaxial layers, heterostructures, microchips, su-
other, to produce a geometric selection. Crystals
perlattices).
having an optimal orientation for the given am-
bient conditions then survive. Such a competition
CRYSTAL STABILITY CONDITIONS
results in the formation of a pole structure, typ-
Conditions imposed on the coefficients of the
ical for metal ingots, in which all the crystals are
expansion of the crystal potential energy U in
elongated and almost parallel to each other. Exam-
terms of atomic displacements usi :
ples of such collective growth are druses (crystal
agglomerations in which the vertices of separate U = U0 + Asi usi
crystals face in more or less the same direction) s,i
and parallel inshoots (crystals growing together to
merge in a parallel orientation). 1
+ Ast ij usi utj + · · · ,
The ideal structure of a crystal always under- 2
s,t,i,j
goes certain perturbations (accumulates defects)
because growth conditions change, impurities are where U0 is the crystal energy at equilibrium
trapped, and various external influences are always (the indices s and t are used for labeling atoms;
at work. Real crystals often display macrodefects i, j = x, y, z). From the condition that the resul-
(gas–liquid, solid inclusions, cracks, etc.) which tant of all forces acting on the atoms at equilibrium
worsen their properties. A typical structural defect vanishes, it follows that Asl = 0. From the in-
is mosaicity (see Mosaic crystals) which results variance of U under translation and
rotation of
from various defects (internal stress, block bound- the
crystal as a whole it follows that t Ast ij =
aries, twins (see Twinning of crystals), stacking 0, t (Ast ij Xt k − Ast ik Xtj ) = 0 (Xsi desig-
faults, dislocations, etc.) that appear during crystal nates the sth atom coordinates). Crystal symmetry
growth. The presence of foreign atoms and ions properties impose additional restrictions on the
(impurities) in a crystal during its growth may re- harmonic force constants Ast ij , and the applica-
sult in the development of zonal, sectional, and tion of these relationships considerably reduces
258 CRYSTAL STRUCTURE
the number of independent force constants. Be- The main techniques for studying crystal struc-
sides that, some restrictions on the force constants tures are X-ray diffraction structural analysis, neu-
follow from the condition of crystal equilibrium tron diffraction, and electron diffraction.
(U = min at equilibrium), and these are in the
form of inequalities arising from the requirement CRYSTAL SYMMETRY, microscopic crystal
of positive definiteness for the quadratic form symmetry
(third term) of the expansion of U . Similar con- Geometrical regularity in the spatial arrange-
ditions may be derived for the anharmonic force ment of particles (atoms, ions, molecules, com-
constants, which serve as the coefficients of addi- plexes), composing a crystal. One can characterize
tional terms of the expansion of U . the symmetry of a crystal by the totality of all
spatial transformations (x , y , z ) = f (x, y, z),
CRYSTAL STRUCTURE which bring the crystal into coincidence with it-
The structure of an actual crystalline solid. The self. From all conceivable transformations it is suf-
concept of crystal structure starts with the con- ficient to restrict attention to linear ones, and not
cept of crystal lattice, i.e. of the parameters of a necessarily to all of these, but only to deformation-
unit cell, such that any ideal crystal (monocrys- free types, i.e. to linear orthogonal transformations
tal) may be reproduced by repeating its unit cell that preserve the lengths of segments and an-
in space. A full description of a crystal structure gles between them: (x , y , z ) = α(x, y, z) +
also includes specifying the positions of atoms, (x0 , y0 , z0 ), where α is the transformation ma-
ions and molecules inside the cell (including the trix. The determinant D = |α| in this case can have
orientations of molecules). Beside the locations of only two values D = ±1. The transformations
the atoms, other important factors are the distrib- with D = 1 are called proper transformations,
ution of the electron density in the unit cell, the and all of them involve rotational spatial displace-
atomic and ionic radii, chemical bonds (see Crystal ments (i.e. rotations) around some axis through
chemistry), coordination numbers (see Coordina- an angle ϕ, while transformations with D = −1
tion sphere), etc. are called improper; the unit improper transforma-
Ferroelectric and magnetic crystals are also tion being inversion (x , y , z ) = (−x, −y, −z).
characterized by the direction of their electric and All improper transformations are combinations of
magnetic moments, respectively. The correspond- proper transformations with inversion. In general a
ing lattices may not necessarily coincide with the linear transformation can involve a shift or trans-
crystallographic ones, and they may possibly be lation, designated by the term (x0 , y0 , z0 ) through
mutually incommensurate. a distance appropriate to the crystal structure lat-
The crystal structure of an actual crystal is tice spacings. If the term (x0 , y0 , z0 ) = 0 there is
quite variable, thus opening ways for numerous no translation, and the transformation is either a
practical applications. Specific branches of the proper rotation for D = +1, or an improper rota-
physics of solids are dedicated to their descrip- tion for D = −1.
tion (see Crystal physics). The concept of crys- The three-dimensional periodicity of a crystal
tal structure also includes various lattice defects: automatically sharply restricts the possible an-
point defects (such as vacancies, interstitional or gles ϕ of allowed proper (and improper) rotations
substitutional impurity atoms, atomic clusters); to five values: 0◦ , 60◦ , 90◦ , 120◦ and 180◦ . It
linear defects (dislocations); two-dimensional de- is customary to express the allowed angles ϕ by
fects (stacking faults, twins, etc.). When studying the expression ϕ = 2π/n, and in the international
the crystal structure of solid solutions one has notation to designate proper rotations by the five
to know the distribution of atoms (ions, mole- integers: n = 1, 2, 3, 4, 6. Improper rotations are
cules) of various types in their sublattices. Crys- designated by 1̄, 2̄ = m, 3̄, 4̄, 6̄, where m is the
tal structure also includes describing the physical symbol for a mirror reflection in a plane. In the
composition and parameters of structural elements older Schönflies notation the proper and improper
(grains, cells, and blocks) and various other fac- rotations are designated, respectively, by C1 , C2 ,
tors. C3 , C4 , C6 , I , S2 = σ , S3 , S4 , S6 . There are
CRYSTAL SYSTEM 259
two special combined rotation-translation symme- groups, 122 point groups and 36 translational mag-
try operations: (a) a screw rotation which is a netic symmetry groups (see also Antisymmetry).
rotation by 2π/n combined with a translation by A further generalization brought about the ap-
half a lattice spacing along the rotation axis, and pearance of the theory of colored symmetry, and
(b) a glide reflection which is a reflection in a plane the creation of many additional colored symmetry
combined with a translation by half a lattice spac- groups. All these types of groups find use in solid
ing parallel to the plane. state physics.
The complete set of all available symmetry
transformations of a crystal form the space group Concerning macroscopic physical properties
of the crystal (called Fedorov group in the Russian (optical, electrical, mechanical, magnetic, etc.),
literature). It is discrete, but with an infinite set of crystals behave as media that are homogeneous,
elements, because translation through all possible but anisotropic (see Anisotropy of crystals); and the
periods is endless. There exist 230 possible space discreteness of their atomic structure is not signifi-
groups, and they are described explicitly in the In- cant. Because of this only the rotational symmetry
ternational Tables for Crystallography. of a crystal, i.e. its point group, is of impor-
Also important in the physics of crystals are fi- tance for the macroscopic directionally dependent
nite groups containing only proper (and possibly properties described by tensors of different ranks.
improper) rotational elements (without transla- Then the distinction between crystals by their dif-
tions) which are called crystal classes or point ferences in symmetry finds its expression in the
groups, and their number is only 32. In three di- intrinsic symmetry of tensors for different physical
mensional space there are only 14 translational properties. The literature provides information on
groups which contain no rotational elements. A set the specific forms of various tensors of many ranks
of points invariant under translation form a crystal
for all 32 of the crystallographic point groups.
lattice (Bravais lattice). Each space group has as-
sociated with it a point group and a crystal lattice.
When describing the symmetry of magnetic CRYSTAL SYSTEM
(magnetically ordered) crystals (see Magnetism),
The classification of crystals in accordance
the situation becomes complicated, since trans-
with their crystallographic coordinate system, i.e.
forming a crystal into itself requires matching not
only coordinates of individual identical atoms, but by the metric represented by the set of independent
also the directions of their atomic magnetic mo- linear (a, b, c) and angular (α, β, γ ) crystal lattice
ments (spins). The Shubnikov extension of sym- constants. It is possible to choose seven different
metry theory to the symmetry of magnetic crystals unit cells with the same number of crystal systems
involves the addition of the non-spatial opera- corresponding to them (see Table 1). The concept
tion R of spin reversal (see Shubnikov groups). of syngony (in common use in the Russian litera-
The addition of this new symmetry element per- ture) coincides with the concept of crystal system
mits one to construct 1651 space (Shubnikov) for all crystal systems except the hexagonal.
Table 1. Unit cell shapes and metric parameters for the seven crystal systems
In the free form, curium is a soft silvery-white appearance of bulk charges whose field should
metal. It exists in two allotropic modifications: be taken into account in addition to the external
a low-temperature α-Cm below 873 K, and a field and the recombination of current carriers. The
high-temperature form β-Cm. The α-Cm variety motion of injected nonequilibrium carriers (see In-
has a double hexagonal close-packed crystal lat- jection of current carriers) in the external electric
tice with alternation of layers according to the field is described by so-called ambipolar mobility
ABAC. . . scheme; space group P 63 /mmc (D6h 4 ),
μe μh |n − p|
a = 0.344 nm, c ≈ 1.16 nm. The β-Cm form μa = ,
μe n + μh p
has a face-centered cubic lattice, space group
F m3m (Oh5 ), a = 0.5039 nm upon hardening. which, in general, differs from μe and μh . For an
Density of α-Cm is 13.51 g/cm3 ; T melting = intrinsic semiconductor n = p and μa = 0; for an
1620 K, T boiling is about 3380 K; specific heat n-type semiconductor n
p, μa = μh ; and for
a p-type semiconductor p
n, μa = μe , which
cp = 27.6 J·mole−1 ·K−1 ; heat of melting is 14.64
means that in an impurity semiconductor μa co-
kJ/mole, heat of sublimation is 372.6 kJ/mole.
incides closely with the mobility of the minority
Below the Néel point T N = 52 K, α-Cm trans- carriers. The velocity of the grouping of nonequi-
forms from the paramagnetic state to the antifer- librium carriers in an external field E is equal to
romagnetic one (see Antiferromagnetism); below μa E.
the critical temperature T c = 205 K, β-Cm is a An important characteristic of charge carrier
ferrimagnet, or has a noncollinear magnetic struc-
drift is its drift length l, which is the distance
ture. Targets from isotopes 246 Cm and 248 Cm are that the carriers traverse in going from their point
used in the synthesis of the heavier transuranium of generation (see Generation of current carriers)
elements by bombardment with the multicharged to their recombination point. This length has the
ions. High heat release in the preparations from value l = μEτ , where τ is the lifetime of the non-
the isotopes 244 Cm and 242 Cm caused by their equilibrium carriers. The method for measuring
α-decay suggests the possibility of using these the drift length is the same as that for measuring
nuclides to produce small-size nuclear electric the diffusion length.
current sources that persist for several years (e.g., In anisotropic crystals the direction of the drift
in onboard space systems). can differ from that of the electric field (since mo-
bilities are tensors). In strong electric fields the
CURRENT CARRIER DRIFT, charge carrier drift drift can be anisotropic even in isotropic (cubic)
A directed movement of charge carriers in many-valley semiconductors. In the presence of a
semiconductors under the action of external fields transverse magnetic field the drift direction does
imposed on their disordered (thermal) motion. not coincide with the electric field direction.
The current density resulting from the charge car- In a strong magnetic field B normal to the
rier drift in the electric field E (drift current) is external electric field E satisfying the condition
j = σ E, where the electrical conductivity σ = cB
E, the charge carriers drift in the direction
e(μe n + μh p), n and p are the concentrations of normal to both E and B with the velocity v =
the conduction electrons and holes, respectively, E/B that does not depend on the carrier mobility.
and μe and μh are their mobilities (see Mobility A circular motion of the carriers at the cyclotron
of current carriers). The total conduction current frequency is superimposed on this drift.
density is the sum of the drift current, the diffu-
sion current, and the thermoelectric current arising CURRENT CARRIER GENERATION
from the presence of a temperature gradient. The See Generation of current carriers.
charge carrier drift can also result from the carry-
ing along of carriers by either an ultrasonic wave CURRENT CARRIER MOBILITY
(see Acousto-electric effect) or an electromagnetic See Mobility of current carriers.
wave (radio electric effect, light electric effect). In
this event the drift involves nonequilibrium carri- CURRENT CARRIER RECOMBINATION
ers, and therefore it becomes complicated by the See Recombination.
CURRENT CARRIER SCATTERING 263
summation
process extends to 1/μ = i (1/μi ) Boltzmann constant, the scattering is quasi-elastic,
and ρ = i ρi (Matthiessen rule for resistivi- so that only the phonons with q < 2k take part
ties ρI ), which hold for τpi independent of k. in it. The Planck distribution function for them
Provided W (k, k ) depends on the angle between may be approximated by the classical expression
k and k only, the microparticle scattering is called kB T / that corresponds to an equilibrium energy
isotropic scattering, and when it also depends on distribution over the degrees of freedom, so that
the directions of k and k it is anisotropic scat- these are equidistributed phonons. In the simplest
tering. The relaxation time is a tensor in the latter case of an isotropic parabolic dispersion law for
case. ε(k) we have s = −1/2, α = −3/2, r = 3π/8 =
Current carrier scattering in semiconductors. 1.18.
Besides the mechanism of scattering proper, this Optical vibrations of atoms produce a scatter-
effect is controlled by the dispersion law for the ing optical strain potential (scattering by nonpolar
current carriers. When k and k belong to the same optical phonons). When kB T
ω0 , where ω0 is
valley the process is called intravalley scattering, the frequency of an optical phonon, the scattering
otherwise it is designated intervalley scattering. If is elastic, s = −1/2, α = −3/2, r = 3π/8 = 1.18.
one may express τp ∝ es and μ ∝ T α then the At all other temperatures the scattering becomes
quantities s and α control the scattering process. inelastic. Provided, however, that the dispersion
During the scattering of electrons by lattice law is isotropic (or close to isotropic and av-
vibrations (phonons) a phonon with an energy eraged over directions), the equality W (k, k ) =
ω(q) and a quasi-wave number q (ω is the W (k, −k ) makes it possible to introduce τp . If
frequency) is either emitted or absorbed in an ele- kB T ω0 , μ ∝ τp ∝ exp[ω0 /(kB T )], r = 1.
mentary scattering event. The conservation laws of Only these two cases of scattering by phonons
energy ε(k) = ε(k ) ± ω(q) and quasi-momen- are possible in nonpolar crystals such as Ge
tum k = k ± q + b are then obeyed. The and Si. In polar crystals (e.g., AIII BV ) scatter-
“minus” sign here corresponds to the absorption ing by the polar optical potential resulting from
of a phonon, while “plus” corresponds to its emis- the lattice polarization (see Polarization of insu-
sion, and b is a reciprocal lattice vector. Processes lator) contributes to the optical vibrations of the
with b = 0 are the so-called flip-over or Umk- atoms (scattering by polar optical phonons). Such
lapp processes. It is important to account for scattering may be stronger than scattering from
them when (k − k ) lies outside the first Bril- acoustic and optical deformation potentials. It is
louin zone, where b brings q back into the elastic for kB T
ω0 and s = 1/2, α = −1/2,
first zone. Umklapp processes are insignificant for r = 45π/128 = 1.105, and becomes inelastic at
valleys in the center of the first Brillouin zone other temperatures. However, when kB T ω0
(see Many-valley semiconductors). Acoustic vibra- one may introduce τp since electrons can only ab-
tions of atoms deform the crystal, changing the sorb an optical phonons in a weak electric field.
energy of the bottom of the conduction band, Moreover, having absorbed the optical phonon, the
and produce a scattering acoustic strain potential electron emits it almost instantaneously (the prob-
Hac = ε Δv/v , where ε is the strain poten- ability ratio then becomes exp[ω0 /(kB T )]
1.
1 0 1
tial constant (sometimes the strain potential itself), The electron energy undergoes almost no change
and Δv/v0 is the relative change of a given semi- during such a composite scattering event, while k
conductor volume, e.g., a unit cell. does change. We have μ ∝ τp ∝ exp[ω0 /(kB T )],
The above expression for Hac holds for an r = 1 in the absence of degeneracy.
isotropic electron dispersion law only, and for an In crystals with partial ionic bonding acoustic
extremum at the center of curvature of the first vibrations may also produce, beside the deforma-
Brillouin zone. In the general case the strain po- tion potential, a piezoelectric scattering potential,
tential constant is a tensor (strain potential tensor) provided the crystal possesses no center of sym-
that relates Hac to the strain. When T > T = metry (e.g., in AIII BV , AII BVI semiconductors,
0
2
mn u /kB ≈ 1 K, where mn is the effective mass see Piezoelectricity). A piezoelectric constant en-
of the electron, u is the sound velocity, kB is the ters the formulae describing scattering by that
CURRENT CARRIER SCATTERING 265
potential, and s = 1/2, α = −1/2, r = 45π/128 = electron–electron scattering time, τee . When the
1.105 for the case. electrons collide with each other, a noticeable ex-
At low temperatures the phonon gas is rarefied change of both quasi-momentum and energy takes
and the main effect may consist in scattering by place, because of the equality of their masses, and
impurity atoms; the scattering in that case is elastic τee controls the rate of both processes. In the ab-
because of the great difference between the masses sence of degeneracy we have τee ∝ n−1 , where
of the electron and the atom. Scattering from n is the electron concentration. The electron–hole
ionized impurities is similar to the Rutherford scattering is significant for high concentrations of
scattering of α-particles. The Conwell–Weisskopf current carriers (high temperatures, high levels of
formula (E.M. Conwell, V.F. Weisskopf, 1950) ob- injection of electrons). It is similar to scattering by
tained for μ for that kind of scattering does not ionized impurities when the ionic charge is oppo-
take into account the screening by free electrons site to that of the scattered carriers.
of the field produced by the impurity ions (see Scattering by dislocations occurs due to the
Electric charge screening), while the Brooks– interaction of current carriers with charged dis-
Herring formula does. In both cases, and to within location filaments, and the deformation potential
the accuracy of certain factors slowly changing produced by the field of strained dislocations.
with temperature T and the impurity concentra- When the scattering center has a magnetic moment
tion N , we have τp ∝ e3/2 /N , μ ∝ T 3/2 /N , then electron spin- ip scattering becomes possible,
r = 315π/512 = 1.93. Scattering by neutral im- which leads to the Kondo effect. The energy re-
purities is similar to electron scattering by atoms laxation time, τe , may often be represented in the
of hydrogen: s = 0, α = 0, r = 1. The correspond- form te ∝ el , where the value of l depends on the
ing formula for τp is called the Erginsoy formula scattering mechanism. We have l = −1/2 for the
(C. Erginsoy, 1950). In the case of scattering acoustic deformation potential, l = 1/2 for the op-
from point defects with a short-range δ-function tical, l = 3/2 for the polar optical, and l = 1/2 for
potential α = −1.2, s = −1.2. In all the above the- the acoustic piezoelectric potential. The inequality
ories the scattering of current carriers is assumed τp τe holds for the averages of τp and τe in
to be incoherent scattering, with the assumption the case of quasi-elastic scattering.
of no interference between electron waves scat- Scattering of current carriers in metals. This
tered by different impurity atoms. This is generally manifests itself in the principal kinetic properties
true for a random distribution of impurity atoms. of the metal, its electrical conductivity and thermal
The above dependences of the mobility μ on the conductivity. The electrical resistance is related to
temperature T are usually not observed experi- the relaxation of the electron quasi-momentum,
mentally, since they were obtained from idealized while the thermal resistance arises from both mo-
models. The reason for such a pronounced dis- mentum and energy relaxation. When the momen-
agreement between theory and experiment may lie tum relaxation prevails, the Wiedemann–Franz law
in a dependence of mn and of the deformation po- holds: κ/(σ T ) = π 2 kB 2 /(3e2 ), where κ is the elec-
tential on the temperature T . Also, the band may tronic thermal conductivity and σ is the electrical
deviate from a parabolic shape, the electrons may conductivity. The scattering of current carriers
exhibit partial degeneracy, etc. For example, in in metals is also influenced by the shape of the
the case of scattering by phonons in germanium, Fermi surface. In the case of electron–phonon
α = −1.66 for electrons and α = −2.33 for holes. scattering, determining the explicit form of the
Electron–electron scattering does not change Hamiltonian H is a difficult problem, since the
the total quasi-momentum of the electron gas, and displacement of ions leads to a significant restruc-
hence has not affect on either the current or μ. turing of the electronic wave functions. At the
However, provided some other scattering mecha- same time semiempirical techniques are available
nism is simultaneously at work, characterized, as that make it possible to relate matrix elements
it were, by a τp (ε), both the current and μ will of the Hamiltonian to the structure of the Fermi
change because of energy redistribution among surface. The most well-known one is based on
the electrons. This effect is characterized by the combining the model of rigid (undeformable) ions
266 CURRENT CARRIER SCATTERING
with the pseudopotential method. It is not diffi- holds. Strictly speaking, the above refers to com-
cult to demonstrate, without falling back to model pensated metals (with equal numbers of electrons
considerations, that the squared matrix element of and holes), or to metals with open Fermi surfaces.
the Hamiltonian H is proportional to the phonon For an uncompensated metal the relaxation of to-
momentum to first order for the case of scattering tal momentum at low temperatures is controlled by
from long-wave vibrations. electron–phonon collisions, accompanied by Umk-
The mechanism of electron–phonon relaxation lapp processes. Moreover, such processes are only
depends qualitatively on the relation between the possible in the regions where closed Fermi sur-
momentum of thermal phonons q and the typical faces come closest to each other, the so-called hot
size PF of the Fermi surface (note that an inter- spots or hollows. The transport relaxation time,
mediate range PF1 < q < PF2 is also realized roughly speaking, is equal to the sum of the time
in Bi, PF1 and PF2 being the major and the mi- span of transitions across the Fermi surface to
nor semiaxes of the ellipsoidal Fermi surface). Let one of the hollows by diffusion, plus the time an
us introduce the maximum energy of the phonons electron takes to flip over or undergo an Umk-
that can interact with electrons: θ = uPF . The ra- lapp process in a hollow. Usually the first time
tio θ/kB in typical metals is of the order of the span significantly exceeds the second (Umklapp
Debye temperature. At high temperatures we have processes proceed relatively readily), so that the
kB T
θ and the phonons with energy of the or- Bloch law remains valid. Umklapp processes may
der of θ and momentum q ≈ PF are important. control momentum relaxation only in the range of
The number of such phonons and, accordingly, extremely low temperatures (about 1 K), resulting
the probability of electron–phonon collisions, are in an exponential temperature dependence of the
proportional to the temperature T . The electri- electrical resistivity against the background of the
cal resistance is, correspondingly, ρ ∝ T . Since predominant influence of electron–impurity scat-
the characteristic energy scale of the change in tering.
the Fermi distribution function for electrons (see Electron–electron collisions. Only electrons
Fermi–Dirac statistics) kB T is much larger than the from regions where the Fermi distribution is
phonon energy, such collisions are quasi-elastic, blurred or diffuse may take part in such colli-
and the Wiedemann–Franz law holds. At low tem- sions, due to the Pauli principle and the conser-
peratures kB T θ and phonons with energy of vation of energy law. In other words, a given
the order of kB T and momentum q = kB T /u = electron may only collide with a small frac-
PF kB T /θ start to play the decisive role. In this tion of the conduction electrons, of the order of
particular case energy relaxation occurs during a kB T /εF (εF is the Fermi energy), since the num-
single collision event. A significant change in the ber of available states is proportional to kB T /εF .
electron momentum ω requires [PF /(q)]2 = (After one of the electrons enters the diffuse
[θ/(kB T )]2 collisions. Such a relaxation is es- range, another appears automatically, due to en-
sentially a diffusion of electrons over the Fermi ergy conservation during their collisions.) Thus,
surface that proceeds with a typical step of q taking into account the Pauli principle lowers
PF . Due to the laws of conservation of energy the probability of an electron–electron collision
and momentum such an electron may only inter- by a factor of [εF /(kB T )]−2 (see Fermi liquid).
act with those phonons that have their momenta One should also take account of the fact that the
normal to the electron velocity vector, the num- screening radius in a typical metal is of the or-
ber of such phonons being proportional to T 2 . der of the lattice constant, and as a result, the
In addition, the squared matrix element of the mean free path for electron–electron collisions is
electron–phonon interaction is proportional to the lee ≈ a[εF /(kB T )]2 . Umklapp processes during
first power, q ∝ T . As a result, the energy relax- electron–electron collisions are possible when the
ation time at low temperatures is te ∝ T −3 , while Fermi surface crosses a plane normal to a certain
the transport (pulse) time is proportional to τp ≈ vector b of the reciprocal lattice located a distance
τe [θ/(kB T )]2 ≈ T −5 . The corresponding electri- b/4 from the center of the Brillouin zone. The law
cal resistivity follows the Bloch law, ρ ∝ T 5 . Nat- ρ ∝ T 2 might manifest itself at low enough tem-
urally, the Wiedemann–Franz law now no longer peratures when one could neglect the input from
CURRENT STATES IN METALS 267
electron–phonon scattering, but there is no reli- can increase still further, with a further rise in the
able experimental evidence for this. Meanwhile, lattice temperature. The channeling often brings
electron–electron collisions play a role during the about irreversible processes in the current channel,
absorption of electromagnetic waves in metals in which could put the device out of operation. If the
the near-IR spectral range. Having absorbed an channeling associated with an S-shaped character-
electromagnetic quantum with energy ω
kB T istic happens at high frequencies it may be used
the electron may interact with other electrons in for the generation of super-high frequency oscil-
a layer of thickness ω. Calculations result in a lations (harmonics), as occurs in various types
substitution in the expression for the probability of diodes (e.g., trapped plasma avalanche trig-
of electron–electron collisions: the kB2 T 2 factor is
gered transit diode, tunnel diode, impact ionization
2 2 2
replaced by kB T + [ω/(2π)] . This does not avalanche transit time diode). Besides, the pinch
apply at T = 0 where everything is in the ground effect may be responsible for current channeling
state. in semiconductors (see Magnetic pressure). Also
Scattering from static defects is usually elastic, this channeling may be observed in pure metals
hence the collision probability does not depend on at low temperatures through the magnetodynamic
the temperature, and the Wiedemann–Franz law mechanism of nonlinearity (see Magnetodynamic
holds. Local defects such as impurity atoms signif- nonlinearity).
icantly perturb the crystal field at distances of the
order of a lattice constant a. The field of a charged CURRENT, PERSISTENT
impurity is screened by electrons, the screening ra- See Persistent current.
dius (Debye radius) in a typical metal also being
of the order of a. Therefore, the scattering cross-
CURRENT STATES IN METALS
section is of the order of a 2 , and depends on the
relatively weak influence of the local defects. The A hysteresis effect (in an external magnetic
mean free path is led ∝ aC −1 , where C is the field B0 ) of electric current rectification, and the
number of local defects divided by the number of excitation of a static magnetic moment in the sam-
cells in the crystal lattice. ple when irradiating the latter by an rf (radio wave)
of a high amplitude Brf . This takes place at the
CURRENT, CRITICAL quasi-static limit ωτ 1 (here ω is the rf fre-
See Critical currents. quency; τ is the electron characteristic lifetime)
of the anomalous skin-effect, with B0 parallel to
CURRENT, ELECTRIC the rf magnetic field vector Brf . The current states
See Electric current. are excited when Brf exceeds a threshold value
Bcr ∼ 0.1–1 mT that is determined by ω, the
CURRENT FILAMENTATION, current channeling mean free path l, and the character of the elec-
Electric current flow in a sample with a nonuni- tron surface scattering. For diffusion scattering
form distribution over the cross-section, i.e. high Bcr ∝ ω−1/3 l −2 . In the ideal case of specular re-
densities along some channels (filaments), and low flection the current states do not emerge. At the
or zero densities along neighboring parallel paths. regime of strong nonlinearity Brf > Bcr , the mag-
The phenomenon of current filamentation may netic moment does not vanish even at B0 = 0.
result from an S-shaped current–voltage character- For growing Brf , the hysteresis loop approaches
istic. Such a characteristic can arise from certain a universal curve which is determined only by
mechanisms of current carrier scattering in solids. amplitude Brf , and does not depend on the elec-
In addition an S-shaped characteristic in a uni- trodynamic properties of the metal. The excitation
form semiconductor may arise from the release of of current states is stimulated by sharp conductiv-
Joule heat (overheating instability); when a tem- ity jumps during the rf period which are due to the
perature increase caused by the current flow leads appearance and disappearance of trapped electrons
to additional electron transitions from local elec- in the skin-layer (see Magnetodynamic nonlinear-
tron levels in the conduction band then the current ity).
268 CURRENT STATES IN SUPERCONDUCTORS
CYCLIC HEAT TREATMENT amplitude. The greater the average cyclic stress,
A heat treatment of metals and alloys involv- the lesser the stress amplitude needed for mater-
ing periodic heating and cooling of the material. ial destruction at the same test base. The character
This is employed to induce structural changes that of the stress change between σmax and σmin has
lead to the improvement of mechanical properties. a negligible effect on the cyclic strength. That
One heating and one cooling event constitute a is why cycles of complicated shapes may always
unit cycle of a cyclic heat treatment. The cycle be reduced to ordinary ones, and standard fatigue
configuration (i.e. temperature versus time depen- tests are carried out with the use of the simplest
dence), its parameters (maximum and minimum cycle configuration.
temperatures of the cycle, heating and cooling An increase of the cycle frequency, other things
rates, exposure time at the maximum temperature being equal, usually gives rise to an increase in
of the cycle) and the number of cycles can vary, the cyclic strength, especially at high tempera-
depending on the type of material being treated, tures. So long as fatigue cracks are forming from
its original condition, and the needed structural regions of stress concentration on the surface,
changes. their state strongly influences the resulting cyclic
A cyclic heat treatment can be carried out so strength. As the temperature increases, a shift of
that no phase transitions take place in the tem- fatigue curves towards lower stresses is observed.
perature range of the cycle. In this case structural When temperature changes take place, fatigue dis-
changes in the cyclic process account for the plas- turbances may also arise because of the onset of
tic deformation due to the thermal stresses which thermal stresses. Cyclic strength rises with the in-
arise as a result of temperature gradients induced crease of metal purity with respect to interstitial
by heating and cooling the material. Substantial elements, with the decrease of grain size, and with
changes take place in the intragrain dislocation the formation of a cellular dislocation structure
structure, grain interfaces (see Grain boundaries), (see Fragmentation). It also depends on the pres-
regions of grains adjacent to interfaces, and the ence and size of the second phase of dispersed
quantity and morphology of particles of phases particles.
which gain strength. In particular, the durability of The environment in contact with the surface of
metals and alloys under the conditions of high- the crack affects the crack growth due to physico-
temperature creep increases significantly. chemical processes taking place at its peak (see
If a cyclic heat treatment involves a phase Fatigue crack growth). The reliable determination
transition within its temperature range of thermal of the cyclic strength is only possible when one
cycling (e.g., the α ↔ γ transformation in iron takes into account the nature of the product, and
and its alloys), the structural changes of a cyclic the details of the fatigue.
heat treatment arise as a result of both the plas-
tic deformation caused by thermal stresses and the CYCLOTRON FREQUENCY, gyromagnetic
phase transitions. At low rates of heating and cool- frequency
ing (or for a small difference between Tmax and Gyromagnetic frequency Ω of charged parti-
Tmin ) phase transitions play the main role, and cle rotation in a constant magnetic field B in a
the presence of these transitions can result in the plane perpendicular to the field. For a free charged
production of tiny uniform sized grains which in- particle the cyclotron frequency Ω is determined
crease the plasticity and strength of the material. from the balance of the Lorentz force ev × B by
the centrifugal force mv 2 /r, and it is given by
CYCLIC STRENGTH Ω = eB/m, where e and m are, respectively, the
Resistance to fatigue failure of materials un- charge and mass of the free particle, and v = Ωr.
der the application of cyclic loads. Quantitatively The cyclotron frequency determines the energy
cyclic strength may be estimated by the endurance difference
E between the energy levels of a par-
limit. As a rule, cyclic strength is somewhat lower ticle in a magnetic field:
E = Ω. In crystals the
than static strength. First of all, cyclic strength is motion of particles is more complicated since they
determined by the maximum cyclic stress and its interact with the ions of the lattice. In a constant
CYCLOTRON–PHONON RESONANCE 271
magnetic field the energy of the electron (or hole) increase of the resonance number n (Ω = nω/2,
E and the projection pB of its quasi-momentum p n = 1, 2, 3, . . .). This resonance can be observed
along the direction B are conserved. Hence in mo- when the electrons make many revolutions in B0
mentum space the motion takes place along the before they collide with other particles and scat-
intersection of the isoenergetic surface E(p) = E ter, i.e. Ω0
νl , where νl is the a relaxation rate
with the plane pB = const. If this curve is closed of the electron pulse. In solids collisions of the
then the motion is periodic with the frequency conduction electrons with crystal lattice defects
Ω = eB/m∗ , where m∗ is the charge carrier cy- (νl = 109 –1011 s−1 ), and scattering by thermal
clotron effective mass. vibrations (electron–phonon interaction) are the
most important. The last process limits the obser-
CYCLOTRON PARAMETRIC RESONANCE vation region of cyclotron parametric resonance: it
Selective absorption of electromagnetic radia- is at low temperatures (1 to 10 K), where collisions
tion by a semiconductor located at the maximum with thermal phonons become rather infrequent.
magnetic field of a stationary wave (frequency ω) The actually achieved relaxation times provide the
inside a resonator, when a longitudinal constant lower limit of the frequency region (ω > 109 Hz)
magnetic field B0 is applied, and the cyclotron fre- where cyclotron parametric resonance can be de-
quency Ω is a multiple of half the pump wave
tected at the field value B0 = 0.1 to 1 T. This
frequency ω. Cyclotron parametric resonance is
resonance is a threshold-type effect which can be
due to instability, arising from time modulation
observed when the pump levels are rather high, i.e.
of the electron cyclotron rotation frequency (see
Ω1 > (νe νl )1/2 , Ω1 = eB1 /m, where B1 is the
Cyclotron resonance). The stabilization of this in-
amplitude of the alternating magnetic field, and
stability results from nonlinear processes, arising
νe is the relaxation rate of the electron energy,
at large electron energies, that lead to a highly
νe ≈ (10−4 –10−3 )νl . This condition is valid in
nonequilibrium and anisotropic distribution of car-
fields B1 ∼ 0.01–0.1 mT. The energy of the reso-
riers formed at the cyclotron parametric reso-
nance particles increases exponentially with time,
nance. The presence of this distribution results in
and can attain large values (electron gas overheat-
the generation and amplification of electromag-
ing can reach 100 K). It results from the fact that
netic and acoustic vibrations in semiconductors
over a broad frequency range. In a thin sample at the electron distribution becomes rather nonequi-
cyclotron parametric resonance the electron distri- librium and anisotropic. The energy increase is
bution function is independent of the coordinates. not infinite, but is limited by nonlinear processes.
In a bulk semiconductor, when the sample size ex- These are processes such as the nonquadratic spec-
ceeds the wavelength of the pumping wave in the trum of conduction electrons, which results in the
resonator; then the heterogeneity of the wave mag- dependence of the mass (and the cyclotron fre-
netic field plays an important role. The emerging quency) on the energy, and also affects the energy
nonequilibrium distribution becomes rather het- dependence of the electron relaxation rates.
erogeneous. It results in gradients of the average Cyclotron parametric resonance can be used as
energy and concentration, and finally it gives rise a method of high-frequency resonance pumping
to a static electric field Eind (hence a static electro- of electrons in conductors to produce an inverted
motive force) and a heterogeneous Hall effect with population distribution of charge carrier energies.
a resonance character. The main cyclotron para-
metric resonance appears when the cyclotron fre- CYCLOTRON–PHONON RESONANCE
quency Ω0 = eB0 /m (e and m are particle charge A sharp increase of the electromagnetic wave
and mass) is half the modulation frequency ω/2 absorption in semiconductors, when the condition
(Ω0 ∼ ω/2 for linear polarization of the pump ω = ω0 + nωL is satisfied, where ω is the electro-
field in a rectangular resonator, and Ω0 ∼ ω for magnetic wave frequency, ω0 is the optical phonon
circular polarization). Cyclotron parametric res- frequency, ωL is the Larmor precession frequency,
onance, like any other parametric resonance of and n is an integer. The resonance is observed
mechanical or electrical vibrations, occurs in reso- at high frequencies (ωτ
1) and in a quantizing
nance zones of instability, and diminishes with the magnetic eld (ωL τ
1, ωL
kB T ); τ is the
272 CYCLOTRON RESONANCE
time of the electron mean free path. The resonance on the Luttinger Hamiltonian matrix. In addition
is associated with transitions of electrons between to cyclotron resonance there are spin resonances
Landau levels with the simultaneous absorption of of free carriers (combined resonance and para-
a photon and an optical phonon. The singularity of magnetic resonance). These types of absorption,
the absorption coefficient at its approach to reso- together with the investigation of magnetooptical
nance has a logarithmic nature. Cyclotron–phonon interband transitions, provide a reliable starting
resonance has often been studied experimentally. point for understanding energy band structures
It is used to determine semiconductor parame- (see Band theory).
ters, and in particular the frequency of the optical
phonon. CYCLOTRON RESONANCE, ACOUSTIC
See Acoustic cyclotron resonance.
CYCLOTRON RESONANCE
Resonant absorption of radiofrequency or mi- CYCLOTRON RESONANCE IN METALS,
crowave energy by an electron or other charged Azbel–Kaner resonance
particle in a static magnetic field when its gyrofre- In metals cyclotron resonance amounts to an
quency equals the frequency of a perpendicular increase in the high-frequency conductance at fre-
electric field. It corresponds to the diamagnetic quencies ω which are multiples of the electron
resonance absorption of an electromagnetic wave, cyclotron resonance frequency in a constant mag-
accompanied by the transition of an electron (or netic field parallel to the sample surface. This
hole) from one stationary orbit to another (mainly, phenomenon is due to the repeated synchronous
to a neighboring one) in a crystal situated in an ex- acceleration of electrons during the part of the
ternal constant magnetic field B. By its nature this orbit located within the skin depth δ (see Skin-
is an electric dipole resonance. In a solid a charge effect). In 1956 M.Ya. Azbel and E.A. Kaner the-
carrier can move freely longitudinally along B.
oretically predicted cyclotron resonance in metals,
The transverse motion takes place in a fixed or-
and E. Fawcett discovered it in Sn and Cu. It has
bit with energy equal to ωc n (in an elementary
been observed in over 32 metals.
Brillouin zone with energy E(k) = 2 k 2 /(2m∗ )),
The electron acceleration mechanism for cy-
where ωc = eB/m∗ is the cyclotron frequency and
clotron resonance in metals is as follows (see Fig.).
n is a positive integer (Landau levels). In the case
For a magnetic field B parallel to the metal surface
of absorption, the change of carrier energy is ωc
all electrons move in circular orbits, and periodi-
for the ground state absorption, and a multiple of
cally with the cyclotron frequency Ω they return to
ωc for harmonics arising when the dependence
and pass through the skin δ where they are acceler-
E(k) is more complex. The nature of E(k) can
ated by the electromagnetic wave. If the resonance
be deduced by methods of group theory in solid
condition ω = nΩ, n = 1, 2, 3, . . . , is satisfied
state physics. A main feature of cyclotron reso-
then the acceleration of the electrons takes place
nance is the possibility of determining the effective
on each cycle of the trajectory, just as it occurs in
mass m∗ directly through a measurement of the
absorption frequency. When E(k) is anisotropic a cyclotron with a single accelerating region. The
then ωc depends on the angle between B and skin depth δ in a metal plays the role of the latter.
the symmetry axes of the crystal. The main con- The metal region outside the skin acts as the dees
dition for observing cyclotron resonance is the
requirement that the conduction electron be able
to complete an orbit before colliding (ωτ > 1),
where impurities, heterogeneities, and crystal lat-
tice vibrations determine the scattering time τ . For
complicated forms of the energy E(k), in particu-
lar for the presence of both light and heavy holes
in cubic semiconductors, the equations of motion
have an obvious classical interpretation. The the-
ory uses a quantum-mechanical calculation based Cyclotron resonance in metals.
CYCLOTRON WAVES 273
of the cyclotron. Cyclotron resonance is observed firstly to the splitting of the resonance peaks, then
in pure single crystal metals at helium tempera- to the doubling of their number (Ω = nΩ/2),
tures when the rate of electron collisions is less and to the inversion of the lines. In numerous
than Ω. It is observed only for the conditions of experiments on cyclotron resonance in metals an
the anomalous skin-effect, when the radius of the important fact has been established: electron scat-
orbit R and the electron mean free path l are much tering on the surface heterogeneities of the sample
larger than the skin depth δ. is almost of the mirror-type. It results in a consid-
Cyclotron resonance in metals differs signif- erable decrease of the amplitude of the cyclotron
icantly from diamagnetic cyclotron resonance in resonance in metals, since there arises a new group
semiconductors. The latter is also called cyclotron of electrons sliding along the surface near the mir-
resonance when it takes place in a gas-discharge ror boundary of the skin depth. Their contribution
plasma. Diamagnetic resonance occurs in a ho- to the current is rather more than the contribution
mogeneous radio-wave field. Therefore: (a) it is of resonance particles, and cyclotron resonance
observed at the fundamental frequency (first har- in metals arises from corrections to the smooth
monic) ω = Ω, but not at multiple frequencies, as nonresonance part of the surface impedance of
is the case of cyclotron resonance in metals; (b) it metals. A complete self-consistent theory of cy-
is a maximum when the polarization of the rf elec- clotron resonance in metals at arbitrary reflections
tric field is perpendicular to the vector B, whereas of electrons from the metal surface over a wide
cyclotron resonance in metals exists at any polar- range of frequencies and magnetic fields has been
ization; (c) it is characterized by a sharp increase developed.
of the electromagnetic absorption at resonance,
while absorption decreases at cyclotron resonance CYCLOTRON RESONANCE, QUANTUM
in metals. The last circumstance is associated with See Quantum cyclotron resonance.
the fact that a sharp increase of the current at cy-
clotron resonance in metals results in its resonance CYCLOTRON WAVES
screening from the the external wave, in the in- Intrinsic electromagnetic excitations of degen-
crease of the reflection coefficient, and, therefore, erate electron–hole solid-state plasmas in a mag-
in a decrease of the energy absorption. The line netic field B. The spectrum of cyclotron waves
shape of cyclotron resonance in metals is rather is localized near the cyclotron resonance frequen-
asymmetric due to the presence of intrinsic cy- cies ω = nΩ (n = 1, 2, 3, . . .), for a wave vector
clotron waves, which have their spectrum localized k⊥B. Its origin is the collective precession of
near harmonics nΩ of cyclotron frequencies. In electrons with the cyclotron frequency Ω, in the
the case of a nonquadratic electron dispersion law, absence of magnetic collision-free damping. The
which is typical for most metals, the cyclotron distinguishing features of this spectrum are due to
frequency Ω depends on the value of the elec- the nonlocal properties of the conductivity oper-
tron momentum projection along the direction B, ator, and depend on the electron dispersion law.
hence it differs for different electrons. In this case For the square law of dispersion there are three in-
cyclotron resonance in metals occurs at the ex- dependent waves in the vicinity of each cyclotron
treme frequencies nΩext , since for electrons with resonance, namely the ordinary, extraordinary, and
Ωext there is a singularity in the density of states. longitudinal ones, differing in their spectrum and
Owing to the decrease of the resonance electron polarization. The spectrum of the transverse cy-
number the amplitude of the cyclotron resonance clotron waves begins at the point k = 0 from
in metals decreases, and the lines broaden and be- the frequency nΩ, reaches a minimum (about
come more asymmetric in comparison with the nΩ − Ω/2) at k ∼ R −1 (R is the cyclotron ra-
case Ω = const. In so doing the resonance of elec- dius), and then asymptotically approaches the ini-
trons from the central cross-section, and from the tial value nΩ. The attenuation of cyclotron waves
vicinity of special points of the Fermi surface, be- is determined by the rate of electron relaxation.
comes sensitive to the wave polarization. A small Parameters of the Fermi liquid interaction in alkali
slope of the field B relative to the surface leads halides have been measured by cyclotron waves.
This Page Intentionally Left Blank
Dd
DAMPING, COLLISION-FREE with different orientation in a unit cell. It is observ-
See Magnetic collision-free damping. able by the splitting of bands of absorption and lu-
minescence, the components of which appear to
DAMPING, LANDAU be polarized in orthogonal directions.
See Landau damping. Davydov splitting happens in the spectra of in-
organic crystals with ionic or mixed-type bonds
DANGLING BOND, broken bond in which one can distinguish groups of strongly
An unsaturated covalent bond; a bond from bound atoms with different orientation in the lat-
one atom without a terminus, a kind of structural tice (K2 CrO4 , CaWO4 , etc.). Davydov splitting is
defect in a solid involving a covalent or ionic-
especially apparent in semiconductors with a lay-
covalent bond. Broken bonds are inherent in both
ered and chain-like structure (As2 S3 , CdP2 , ZnP2 ,
crystalline and amorphous solids (see Amorphous etc.) where the unit cell is composed of atoms be-
state), and can be arranged in a type of random longing to two or more translationally inequivalent
continuous network; this is common in organic chains or layers, between which, like in molecu-
and polymeric glasses (see Vitreous state of mat- lar crystals, there is a weak van der Waals bond
ter). Associated with broken bonds in amorphous (see also Fermi–Davydov combined resonance,
semiconductors and insulators are features of the Van der Waals forces).
electronic spectrum that significantly affect the Davydov splitting in liquid crystals is close in
processes of electron transfer and photostimulated nature to Davydov splitting in crystals with sev-
phenomena. In many cases broken bonds suffi- eral molecules in the unit cell and is of the order
ciently high in concentration can be observed with of
νe ∼ 10−13 s−1 for electronic transitions or
the help of electron paramagnetic resonance since
νv ∼ 10−11 s−1 for vibrational transitions. In
they involve unpaired electrons. The density of liquid crystals, there appears an additional con-
broken bonds in amorphous solids can reach 1020 tribution to the splitting, which is associated with
to 1021 cm−3 . Examples of broken bonds are un- the orientation dependence of the dispersion inter-
saturated bonds of surface atoms. Since the energy action. It is comparable in its magnitude with the
can be lowered by saturation (closure) of broken Davydov splitting, and observable in impurity ab-
bonds, efforts are made toward surface reconstruc- sorption bands.
tion of covalent crystals, and this is an important
factor in chemisorption on crystals. DE BOER PARAMETER (J. de Boer, 1948)
Ratio of the amplitude of the zero-point vibra-
DAVYDOV SPLITTING (A.S. Davydov, 1948) tions of the atoms in a crystal to the crystal lattice
Splitting of electronic, electronic-vibrational spacing, denoted by ΛB . This provides the stabil-
and vibrational states in molecular crystals caused ity criterion of the classical crystal state, namely,
by the presence of inductive coupling between ΛB 1. However, there are crystals for which
molecules and by the migration of excitation en- ΛB ∼ 1. Since the existence of the crystalline state
ergy in a crystal. Davydov splitting is possible in such a situation is itself problematic, the defin-
when there are two or more identical molecules ition of the de Boer parameter should not contain
275
276 DE BROGLIE WAVE
the lattice period. It is determined through the pa- case the components of the pair provide the mu-
rameters of the pairwise interaction of atoms, pa- tual screening). The relation between the concen-
rameters of the Lennard-Jones potential. The de tration of pairs and the concentration of individual
Boer parameter for a series of elements has the screened ions is determined by the total concentra-
values: tion of ions in the solution. This theory is used for
3 He 4 He
the description of defects in nonmetallic crystals,
H2 D2 Ne Ar K Xe e.g., for the calculation of diffusion coef cients, of
0.48 0.43 0.28 0.20 0.09 0.03 0.02 0.01 ionic conductivity, etc.
Crystals for which ΛB ∼ 1 are called quantum
DEBYE–SCHERRER METHOD, powder X-ray
crystals.
method, polycrystal method (P. Debye,
P. Scherrer)
DE BROGLIE WAVE (L. de Broglie, 1924) Method of obtaining the diffraction diagram
The quantum mechanical attribution of wave of a polycrystalline material (see Polycrystal) by
properties to microparticles (electrons, atoms, X-ray scattering. In this method a monochromatic
molecules, etc.). According to the assumption of narrow beam of X-rays is incident on a sample in
L. de Broglie a moving particle with the energy E the form of a cylinder or a polished section from
and momentum p has associated with it a wave a polycrystal. All the crystallites in the sample are
with a frequency ν = E/ h and a wavelength λ = assumed to have random spatial orientations, and
h/p. The de Broglie wave length corresponding to the Bragg condition is satisfied for each set of suit-
the thermal motion of the particles increases with ably oriented crystallographic planes. For each al-
the decrease in temperature T . At a sufficiently lowed crystallite orientation diffracted X-rays will
low temperature when λ becomes comparable to emerge along a cone coaxial with the incident
the distance between the particles of a system, then X-ray beam with an aperture angle 4θ , where θ is a
quantum effects begin to play a role (see Quantum Bragg angle (see Bragg law). The result is a family
liquid, Quantum crystal). of cones corresponding to diffracted X-rays with
various allowed aperture angles 4θ for the vari-
DEBYE FREQUENCY ous allowed reflections. The diffraction picture in
See Speci c heat. the Debye–Scherrer method may be recorded on a
cylindrical or flat photographic film (debyegram).
DEBYE–HÜCKEL THEORY (P. Debye, E. Hückel, In the X-ray diffractometry of polycrystalline sam-
1923)
ples the photographic film is replaced by a rotating
detector of X-rays.
Thermodynamic theory of electrolytes taking
into account the screening of the electrostatic po-
DEBYE TEMPERATURE (P. Debye, 1912)
tential (see Electric charge screening) by the mo- The characteristic crystal temperature Θ D cor-
bile ions of the solution. It follows from Debye– responding to the Debye frequency ωD (k B Θ D =
Hückel theory that an ion located within the so- ωD ), where ωD is the maximum frequency of the
lution is surrounded by a coordination sphere crystal lattice vibrations of the crystal (ωD ∝ vs /a,
of ions of opposite sign, its radius being l D =
# where vs is the sound velocity and a is the crys-
εkB T /(8πe2 I ), where ε is the dielectric con- tal lattice constant). In the Debye model the crys-
stant,
I is “ionic force” of the solution, I = tal vibration frequencies ω are limited by the De-
(1/2) i ni Zi2 where ni is the concentration of bye frequency (ω ωD ), and the linear dispersion
ions of the ith type, and Zi is the valence of the law ω = vs |k| is assumed to be valid, where k is
ith ion. The coefficient of activity of the ith ion the phonon wave vector which takes on 3N dis-
is equal to exp[−Z+ Z− e2 /(2k B T εl D )], where crete values (N is the number of atoms in the crys-
Z+ , Z− are valences of ions of the corresponding tal), the largest value of k being π/a. The numer-
sign of the charge. In the Debye–Hückel theory the ical values of the Debye temperature depend on
possibility of forming a bound pair of oppositely the atomic mass, and on the strength of the elas-
charged ions is also taken into account (in this tic bond between atoms (e.g., for Pb Θ D ∼ 90 K,
DECHANNELING 277
for diamond Θ D ∼ 2340 K). Knowledge of the crystal, on the atomic masses and on tempera-
value of Θ D allows one to obtain important in- ture. The static Debye–Waller factor depends on
formation about the vibrational spectra of a crys- the elastic constants of the crystal, and on the
tal, and about the characteristics of its chemical characteristics of defects (concentration, dimen-
bonds. The notion of a Debye temperature is es- sions, correlation in their location, etc.). Usually
sential for treating the thermodynamics of crys- Wth 1, and for room temperature W th ∼ 0.01 to
tal; it is the approximate temperature boundary 0.1. For impurity atoms, small clusters and dislo-
which separates the low-temperature (T Θ D ) cation loops W st 1; for dislocations W st
1
and high-temperature (T
Θ D ) lattice properties and the effects of coherent dissipation disappear
of the crystal, where T is the crystal temperature. in the kinematic approximation; but taking into ac-
Using only the single parameter Θ D together with count the dynamic effects of scattering W st in the
some approximation expressions it is possible to case of dislocations requires renormalizing, while,
describe the behavior of a whole series of observed e.g., for the scattering of X-rays the renormalized
physical characteristics of crystals over a broad Debye–Waller factor W st < 1 for a dislocation
temperature range below as well as above the density less than 105 to 107 cm−3 .
Debye temperature. Examples of these are: spe- The Debye–Waller factor determines the re-
ci c heat, electrical resistance (Bloch–Grüneisen duction of the coherent scattering intensity I (Q)
formula), elastic moduli, Debye–Waller factor for of X-rays, neutrons and other radiations in crys-
the scattering of X-rays, neutrons, electrons, etc.; tals: I (Q) = I0 (Q) e−2W , where I0 is the in-
properties involving the Mössbauer effect, IR ab- tensity of scattering by the fixed ideal (average)
sorption, etc. lattice. A Debye–Waller factor also describes the
efficiency of gamma ray radiation in a crystal
DEBYE THEORY OF SPECIFIC HEAT without recoil (see Mössbauer effect); in this case
See Speci c heat. it is sometimes called the Lamb–Mössbauer fac-
tor.
DEBYE–WALLER FACTOR (P. Debye, I. Waller)
DECA. . . (fr. Gr. δεκα , ten), da
The temperature factor in the expression for
the probability of the radiation of photons from Decimal prefix which denotes a 10-fold in-
crystals, or of elastic processes for the scatter- crease of the value of the basic unit. For instance,
1 dal (dekaliter) = 10 l.
ing of particles or photons, which takes into ac-
count thermal and static displacements of atoms
DECAY, ISOSTRUCTURAL
from their sites on a crystal lattice. The Debye–
See Isostructural decay.
Waller factor is written in the form e−2W , where
e−W = exp[iQ(uth + ust )], . . . means DECHANNELING
double (thermodynamic and configurational) av- Exit from the channelling mode; influenced by
eraging, and Q is the scattering vector or the the excitation of phonons, electrons, plasmons,
wave vector of the radiated photons. If the dis- etc., by the channeling particles. Motion of the
placements are counted with respect to the sites channeled particles in the transverse direction can
of the “average” lattice, W assumes the approxi- be considered as a drift in the field of a ran-
mate form W ≈ (1/2)[Q(uth + ust )]2 . In the dom force, caused by deviations from the aver-
isotropic approximation 2Wα = (Q2 /3)u2α . For age continuous potential. In this case the energy
a monatomic crystal, assuming the Debye model, of the transverse motion of the channelized ions
it has the temperature dependence E⊥ increases in proportion with their movement
along the channels (upon reduction of their to-
3Q2 2π 2 T 2
2Wth = 1+ + ··· , tal kinetic energy), and upon reaching the criti-
8MkB ΘD 3 ΘD
cal value E⊥c = E0 Ψk the dechanneling occurs;
where Θ D is Debye temperature, and T is ab- E0 is the initial channelized particle energy, Ψk
solute temperature of the crystal. In the general is the Lindhard angle (see Channeling). The dis-
case Wth . depends on the force constants of the tance at which the density of channeling particles
278 DECI. . .
is reduced by one half is called the dechannel- the electrons of the defects and the valence elec-
ing length. This length for protons with energies trons of neighboring atoms). For calculating lo-
from 1 to 10 MeV in silicon lies within the range cal level parameters given in the literature, various
from 2 to 18 μm. Dechanneling is enhanced by the quantum chemical methods are used (see Quan-
presence of defects in crystals, especially by im- tum chemistry of solids), in particular, a Lifshits
purity or inherent interstitial atoms, located near method of regular perturbations (Green’s func-
the axes of the channels, and also by dislocations; tions), a modified method of effective mass, and
in alloys it depends on the degree of ordering, etc. some others. These methods are arbitrary in their
choice of the periodic field potential perturbed by
DECI. . . (fr. Lat. decem, ten), d a defect in an ideal crystal, and as a rule are not
The decimal fraction prefix meaning 1/10 of very accurate.
a particular unit. For instance, 1 dg (decigram) =
0.1 g. DEFECT
For different varieties see Defects in crystals,
DECOMPOSITION OF ALLOYS Defects in liquid crystals, Anti-Schottky defect,
See Alloy decomposition. Antisite defect, Antistructural defect, Electrically
active extended defects, Frenkel defect, Grain-
DECOMPOSITION, SPINODAL boundary defects, Intrinsic defects, Macroscopic
See Spinodal decomposition. defects, Monopole-type defect, Needle-shaped de-
fect, Planar defect, Rod-shaped defects, Schottky
DECORATION defect, Stacking faults, Stoichiometric defects, Sur-
See Surface decoration. face defects, Swirl defects, Thermal defects.
a random local temperature, depending on their critical indices. The reorienting random local field
influence upon the coefficients of the density η(r) defects are a special class of defects which, unlike
or the thermodynamic potential Φ(r) describing frozen defects, bring about an increase in the tem-
the phase transition. Within the phenomenology of perature of pure crystal phase transitions.
the Landau theory, the density in the absence of
defects has the form: DEFECT STRESS SCALE
A B D 2 See Stress scales of defect structures.
Φ(r) = η2 (r) + η4 (r) + · · · + ∇η(r)
2 4 2
+ · · · − h0 η, (1) DEFECT STRUCTURE ANALYSIS
See Microprobe X-ray defect structure analysis.
where for simplicity the long-range forces are not
taken into account, and the one-component order
DEFECT SYMMETRY GROUP
parameter A = A0 (T −T0 ) is included; B, D, A0
This involves both the symmetry of a defect
are medium constants, and h0 is the external field
which is conjugated to the order parameter. Point itself as well as the symmetry group of the crys-
defects of the random local field type add the fol- tal where the defect appears. Translational sym-
lowing contribution to Φ(r): metry (see Translation) disappears when a single
defect is introduced into a crystal. The point group
Φh = hi (r − ri )η(r), (2) of symmetry of a crystal with a defect reflects the
i symmetry of the point defect, and is a subgroup of
the point group of the ideal crystal. If we disregard
and defects of the random local temperature type
the crystal distortion induced by a defect then the
lead to the additional contribution
symmetry of this defect is the same as the sym-
1 metry of the site that it occupies. The symmetry
ΦT = A(r − ri )η2 (r). (3)
2 groups of the various points of the unit cells for all
The origin of the terms, defects of random lo- 230 space groups are given in the International Ta-
cal field and random local temperature, is asso- bles for Crystallography. For an impurity atom (or
ciated with the fact that the “averaged over con- a vacancy) at a site of a Bravais lattice the defect
figuration” values
Φ h (r) = hη(r),
Φ T (r) = point symmetry is the same as the crystal point
(1/2)Aη2 (r) describe renormalization by the de- symmetry. The presence of a pair of point defects,
fects of the value of the external field h0 and of dislocation loops in cubic crystals, or particles of
the phase transition temperature T0 . a new phase of different symmetry all result in a
The local field defects bring about strong lo- lowering of the symmetry compared with that of
cal distortions of the crystal matrix, because it the host crystal.
follows from Eq. (1) that (not too close to T0 )
η(r) ∝ (1/r) exp(−r/r c ) with r 2 c = D/A → ∞ DEFECT THRESHOLD ENERGY
at T → T0 . These defects eliminate the long-range See Threshold energy of defect formation.
order (see Long-range and short-range order),
leading to a complex domain structure in systems
DEFORMATION AT LOW TEMPERATURES
with infinitely degenerate η, or to long-term relax-
See Low-temperature deformation.
ation in systems with finite degenerate η. Such po-
larized defects create an average field within the
crystal, and induce η = 0 at any temperature T , DEFORMATION BY EXPLOSION
i.e. the phase transition, speaking strictly, is absent See Explosion deformation.
(blurred).
The random local field defects bring about a DEFORMATION DIAGRAM, stress–strain curve
shift of T0 and lead to anomalies of different phys- Graphic representation of the dependence of
ical values in the vicinity of T0 . In the area of sim- the deforming force (stress) on the material re-
ilarity theory these defects cause a variation of the sponse (strain). The deformation diagram reflects
DEGENERACY 283
the results of testing solids by different methods tion of temperature, strain, and rate of deforma-
such as stretching, compressing, e xure and tor- tion. The simplest kinetic equation presenting a
sion. As deformation parameters one can select microscopic description of plastic deformation in
absolute values of elongation, of shearing angle, terms of the dislocation model of plastic deforma-
of torsion angle (see Rod twisting), of deflection, tion takes the form: ε̇ = bρv, where ε̇ is the rate
or preferably one can work with relative deforma- of deformation, b is the Burgers vector, ρ is the
tions which are ratios of sample dimensions during density of mobile dislocations, and v is the mean
loading to those before the test. Relative deforma- velocity of their motion. The deformation kinet-
tions are usually subdivided into two types: a con- ics is accompanied by an irreversible displacement
ditional deformation which is the the sum of suc- of dislocations. In addition to the displacement of
cessive variations of size all relative to the same dislocations, one has to consider the multiplication
initial dimension before the test; and a true de- of dislocations and interactions between them, as
formation which is the sum of small variations of well as the effects of other lattice defects, in par-
size, with each of them relative to the current value ticular, of disclinations. Two competing processes
of the dimension. The deforming effort is charac- proceed simultaneously when a crystal is undergo-
terized also by either the absolute value of the pro- ing a deformation: strain hardening and recovery.
jection of the force or torsion along some direction During hardening the stresses required to main-
(direction of deformation), or by the specific para- tain a constant rate of deformation increase, while
meter of stress, i.e. by the force effecting a unit the rate of deformation decreases, if a constant
external stress is applied. Recovery decreases the
area of some cross-section of the body. The nor-
rate of hardening during active deformation (ε̇ =
mal stress, namely the component of force perpen-
const), and produces a steady-state creep under a
dicular to the cross-section area, and the tangential
constant stress. Existing theories describe the de-
stress, due to the projection of the force along (par-
formation kinetics at the first stages of deforma-
allel to) the area of the cross-section, are singled
tion, in particular, the appearance of a yield cusp
out. The stresses are usually divided into condi-
(upper and lower yield limits on the deformation di-
tional and true stresses: the former determined by
agram, or stress–strain curve). Still, the deforma-
the relation of the actual force to the initial area of
tion kinetics at large plastic deformations cannot
the cross-section, and the latter by the same rela- yet be unambiguously described in terms of exist-
tion to the current area. To plot a deformation di- ing theories.
agram it is possible to use different combinations
of the above mentioned deformation and deform- DEFORMATION, PLASTIC
ing force parameters. For the analysis of physical See Plastic deformation, Discontinuous plastic
dependencies of the deformation process the dia- deformation, Localization of plastic deformation,
gram using the coordinates “true deformation–true Plastic twisting deformation, Temperature condi-
stress” is the most significant. This diagram pro- tions of plastic deformation.
vides data for the calculation of many basic me-
chanical parameters of the material (e.g., elastic DEFORMATION RATE
moduli, yield limit, ultimate strength, strain harden- See Plastic deformation rate.
ing factor, margin of plasticity). The analysis of a
DEFORMATION TENSOR
deformation diagram also gives information about
See Strain tensor.
the physical mechanisms of plastic deformation,
of hardening and of degeneracy. Alternate termi- DEFORMATION WORK
nology for a deformation diagram is deformation See Work of deformation.
curve, hardening curve and diagram of loading.
DEGENERACY
DEFORMATION KINETICS There are two main uses for this term.
The development and interplay of elemental 1. The first use is the presence in a quantum-
acts of deformation that produce strain in a mater- mechanical system of several different states
ial. Generally, the strain-inducing stress is a func- which have the same value of some physical quan-
284 DEGENERACY, SPIN
tity, such as energy, angular momentum, etc. The then one or both types of carrier is already degen-
number of such states is called the multiplicity or erate at room temperature. In doped semiconduc-
degeneracy factor. Usually degeneracy is associ- tors the conduction electrons (holes) become de-
ated with the symmetry of a system; lowering by generate at high donor (acceptor) concentrations.
external factors (fields) brings about the partial or For high optical excitation or strong injection of
complete removal of degeneracy, e.g., it leads to charge carriers the degeneracy of nonequilibrium
the splitting of energy levels (see, e.g., Zeeman ef- carriers is possible.
fect, Stark effect). For an arbitrary degree of degeneracy or multi-
2. The second use of this term concerns the plicity the thermodynamic and kinetic characteris-
state of a gas of identical particles, whereby tics of an equilibrium electron–hole semiconduc-
their quantum properties become revealed below tor system are expressed through the Fermi–Dirac
what is called the degeneracy temperature T0 . For integral:
Bose particles (bosons) degeneracy corresponds to ∞
x n dx
Bose–Einstein condensation, and in the case of an Fn (z) = ,
ideal Bose gas 0 exp(x − z) + 1
where z = ξ/(k B T ), and ξ is the chemical po-
3.32 N 2/3
T0 = , tential. For strong degeneracy, exp[ξ/(k B T )]
1,
mk B (gV )2/3 these formulae are considerably simplified, and for
where N is the number of gas particles, V is the doped degenerate semiconductors they have the
volume, m is the particle mass, and g is the mul- same form as for metals.
tiplicity. For Fermi particles (fermions) the state The degeneracy of charge carriers is especially
of degeneracy corresponds to the almost complete noticeable in those kinetic effects, which arise
filling by the particles of states with energies less from the thermal spread in the energy distribu-
than the Fermi energy, and to the very low occu- tion of the carriers. These include the magne-
pancy of higher energy states. For an ideal Fermi toresistance effect, electronic thermal conductiv-
gas ity, Peltier effect, Ettingshausen effect, and other
effects in semiconductors with isotropic energy
(6π 2 N)2/3 2
T0 = . spectra (see Thermomagnetic phenomena, Ther-
2mk B (gV )2/3 moelectric phenomena). In completely degener-
Conduction electrons in metals (N/V ∝ 1022 to ate semiconductors at T = 0 K these effects are
1023 cm−3 ) are degenerate up to temperatures of absent, because, according to the Pauli principle,
about 104 K. In semiconductors with high dop- in transport phenomena only the charge carri-
ing almost complete degeneracy of electrons in the ers located at or near the Fermi surface and pos-
conduction band (or holes in the valence band) can sessing the same energy take part. At T = 0 K
take place at room temperature or below. these effects are present but they are not apprecia-
ble, their value being approximately ξ/(k B T ) or
DEGENERACY, SPIN [ξ/(k B T )]2 times (depending on the effect under
See Spin degeneracy. consideration) less than in non-degenerate semi-
conductors.
DEGENERATE SEMICONDUCTOR The features of degenerate semiconductors are
A semiconductor with the charge carrier en- most clearly evident in the presence of quantized
ergy distribution described by Fermi–Dirac statis- magnetic fields (see De Haas–van Alphen effect,
tics. The Fermi level is located either within the Shubnikov–de Haas effect, Quantum Hall effect,
conduction band or valence band, or is situated in etc.).
the band gap within ∼k B T from the edge of one
of these bands. Inherent semiconductors become DEGREE OF ORDER of liquid crystals
degenerate at high temperatures when k B T be- (V.N. Tsvetkov, 1942)
comes comparable to the gap width εg , and when The quantity which designates the fraction
these materials have narrow gaps (HgSe, HgTe) of liquid crystal molecules with their long axes
DE HAAS–VAN ALPHEN EFFECT 285
DE HAAS–VAN ALPHEN EFFECT (W. de Haas, cross-section of the Fermi surface ε(p) = εF (the
P. van Alphen, 1931; predicted by L.D. Landau, points of extrema En (pz ) and S(εF , pz ) coincide
1930) at En = εF ). If the Fermi surface has several ex-
The oscillating dependence of thermodynamic tremal cross-sections, then the oscillating parts
characteristics of metals, semimetals and degen- of the thermodynamic characteristics contain har-
erate semiconductors on the magnetic field B, monics with periods
(1/B), responding to all the
caused by the quantization of energy levels of the extremal areas. The experimental study of the de-
conduction electrons, with a fixed projection of pendence of
(1/B) on the direction B gives de-
quasi-momentum (pz ) along the direction B = tailed information about the Fermi surface geom-
(0, 0, B) (Landau quantization). etry, and in some cases provides data sufficient for
The de Haas–van Alphen effect is explained in its complete reconstruction.
terms of the presence of a selected Fermi energy Since the quasi-classic parameter N0−1 is neg-
εF in the spectrum of excitations of an electron ligible in magnitude the quantum oscillations of
Fermi liquid. When the magnetic field B is scanned the thermodynamic potential of the electronic sys-
the various n terms of magnetic quantization with −5/2
tem have the relative amplitude A ∝ N0 . As
the energy En (pz ) and momentum pz in turn in- a rule it is not the potentials themselves that are
tersect εF , and the density of one-electronic states determined, but rather their derivatives (magnetic
having the energy εF oscillates and passes through susceptibility, compressibility, etc.). In this case A
a sharp peak each time one of the n curves E = may grow by several orders. At finite tempera-
En (pz ) is tangential to the dashed horizontal line tures, T = 0, the boundary of the energy dis-
E = εF (see Fig.). It is these oscillations which tribution of the electrons is blurred, which re-
constitute the de Haas–van Alphen effect. Usu- duces the value of A. In the temperature region
ally the number of terms (N0 ) with energy En < k B T
Ω (Ω is the cyclotron frequency corre-
εF is very high: N0 = pF2 /eB (pF is the elec- sponding to the given extremal cross-section of
tron momentum at the energy εF ) and the struc- the Fermi surface) the relative amplitude of the
ture of the de Haas–van Alphen oscillations is de- lth harmonic is proportional to the exponential
termined by the quasi-classical Lifshits–Onsager factor AT = exp[−2π 2 lk B T /(Ω)]. In addition,
quantization rules for electrons with an arbitrary the broadening of the electronic terms caused by
dispersion law ε(p) (p is the quasi-momentum) in the electron scattering further reduces A. This ef-
the magnetic field: S(E, pz )/eB = 2π(n + 1/2), fect is taken into account by the so-called Din-
n
1. Here S is the cross-section area of the gle factor AD = exp(−2πlν/Ω), where ν is the
isoenergetic surface ε(p) = E at the plane pz = reciprocal lifetime of the electronic state, averaged
const. The oscillations are periodic with respect over the extremal cross-section of the Fermi sur-
to the reciprocal magnetic field with the period face. In order to make the factors AT , AD of the
(1/B) = 2πe/S ext , where S ext is the extremal order of unity, low temperatures (He range) and
286 DEHARDENING
pure samples are needed. At T → 0 not only do room and lower temperatures when the thermally
point defects (impurity atoms, vacancies) influence activated dislocations penetrate the Peierls barriers
the value of AD , but also the weak deformational (see Peierls relief), thereby, to a large extent, deter-
fields of dislocations, usually unimportant for the mining the value of the critical shear stress. One
kinetics of the conduction electrons. In some cases may relate dehardening by alloying with the low-
the contribution of dislocations to AD may domi- ering of the Peierls–Nabarro strength (see Peierls–
nate. Nabarro model). The latter may be explained by
A systematic theory of the de Haas–van Alphen the easier formation of double kinks on disloca-
effect was formulated by I.M. Lifshits and A.M. tions at those points where the atoms of the al-
Kosevich (1955). In the presence of magnetic loying element are located. Dehardening by alloy-
breakdown the de Haas–van Alphen effect is qual- ing is also observed when highly active elements
itatively more complicated in comparison with the are added to body-centered cubic metals. Such ele-
quasi-classical one, but the Lifshits–Kosevich the- ments bond the doping atoms into a second phase,
ory can be generalized to this case. clean the metal matrix, and thus lower the impurity
hardening. In that case, the effect of lowering the
DEHARDENING, softening critical shear stress is observed over a wide tem-
Decrease in strength of a material that had been perature range.
brought into a hardened, thermodynamically non-
equilibrium state by the effect of some external DELAY LINE, ULTRASONIC
field (temperature, gravitational, mechanical, elec-
See Ultrasonic delay lines.
tric, magnetic, etc.). Such a softening or dehard-
ening proceeds by the transformation of the ma-
terial into a more stable state with lower energy. DELTA PHASE (δ -phase)
Dehardening may occur either during the process The fourth allotropic modification (see Al-
of the dehardening action itself (dynamic dehard- lotropy) of polymorphous metals (see Polymor-
ening) or subsequent to it (static dehardening). To phism) (e.g., δ-Mn, δ-Ce), and also solid solutions
activate the latter in practice the material is usually based on them. Iron is an exception, because the
heated. solid solution on its third high-temperature mod-
In the case of crystalline solids dehardening ification (see Gamma phase) is called the delta
proceeds through the formation, displacement (re- phase or δ-Fe. Like α-Fe it has a body-centered
distribution) and annihilation of point defects, of cubic lattice and exists from 1400 ◦ C up to the
linear, surface and volume crystal defects, and melting temperature 1540 ◦ C. Alloying iron with
their complexes. Depending on the type of defects elements of Groups III, IV, V and VI broadens the
responsible for the dehardening the latter may ex- temperature range of the delta phase existence. In
hibit different kinetics, which variously affect the some systems (e.g., Fe–Si, Fe–Cr, Fe–P) the stabi-
physical and mechanical properties of the mater- lization of the δ- and α-phases (see Alpha phase)
ial. In most cases dynamic dehardening improves brings about a merging of their phase regions in
the material plasticity, and only weakly affects the their phase diagram. As a result, such phases over
specimen shape as the latter changes under the ef- their entire temperature range of existence only
fect of the external field. Dehardening that is re- have a body-centered cubic lattice. This phenom-
lated to the formation and spread of cracks lowers enon is used to create soft magnetic materials
the material plasticity. (e.g., silicon steels, high-chromium steels, stable
against corrosion under stress, etc.). In phase di-
DEHARDENING BY ALLOYING, softening by agrams of other systems, the fourth intermediate
alloying phase (counting from the first toward the second
A decrease of the critical shear strength in met- component) is called the delta phase. In systems
als upon the addition of a small amount of an al- with interstitial elements (C, N, etc.) the mono-
loying element (e.g., when alloying tantalum with carbides with the NaCl type crystal structure are
rhenium). Dehardening by alloying is observed at designated as delta phase.
DEMAGNETIZATION FACTOR TENSOR 287
The internal fields are related to the applied field A generalization similar to (4) may also be intro-
through the expression duced for the demagnetization factor tensor with
allowance made for the above change in notation.
NBi + (1 − N)μ0 Hi = B0 (2)
with the following individual values: DEMAGNETIZATION FIELDS
Magnetic elds within a magnetic substance
1+χ
Bi = B0 , (that is either magnetized or placed in an exter-
1 + Nχ
nal magnetic field), which seek to lower its mag-
H0 netic moment. The sources of demagnetization
Hi = , (3)
1 + Nχ fields are either surface or bulk (ficticious) mag-
netic charges. To describe demagnetization fields
χH0
M= , one traditionally uses the coefficients of the de-
1 + Nχ that relates the de-
magnetization factor tensor N
where B 0 = μ0 H 0 . From symmetry considera- magnetization field H D with a specimen magne-
tions it is clear that N = 1/3 for a sphere. In the tization M, so that H D = −N M. The value of
limiting cases of an infinitely oblong and oblate the tensor N significantly depends on the speci-
ellipsoids of revolution (a long cylinder and a men shape and may be calculated exactly for an
flat disc (plate)), the same considerations yield ellipsoidal shape only. In the latter case, both H D
N = 1/2 for the case when B 0 is normal to the and N remain uniform within the specimen bulk
cylinder axis, N = 0 for B 0 parallel to that axis, and the tensor N is diagonal with the normaliza-
while N = 1 for B 0 normal to the plate plane. tion condition Nx + Ny + Nz = 1. For a spec-
The demagnetization factor tensor is used to imen of an arbitrary shape, its demagnetization
calculate internal magnetostatic fields and also ex- field is nonuniform. Either empirical or approx-
ternal fields produced by permanent magnets of imate relationships are then used to describe the
various shapes, and likewise for electrostatic cal- tensor N. When exciting a nonuniform precession
culations. The demagnetization factor tensor plays of the magnetization M in a variable magnetic
a significant role when determining frequencies field charges form that result in the appearance of
of ferromagnetic resonance and ferrimagnetic res- variable demagnetization fields. Such fields appre-
onance, and it is important for electron para- ciably affect the spectrum of magnetostatic vibra-
magnetic resonance for gμB B/(kB T ) 1, where tions and waves.
gμB B is the Zeeman splitting (see Zeeman ef-
fect). The demagnetization factor tensor also plays DENDRITES (fr. Gr. δενδρoν , tree)
an important role in the theory of the intermedi- Crystals (or aggregations of crystals) of a
ate state of superconductors and magnetic materi- branched tree-like configuration. They are formed
als in an external magnetic field, while the dielec- during crystallization from the melt, solution, va-
tric analogue theory is used to explain the interac- por, and some metallic glasses. The formation of
tion of electromagnetic waves with small particles dendrites is explained by the instability of the
of various shapes in the Rayleigh approximation source crystal convex shape under the conditions
(see Rayleigh scattering of light). In the latter case of oversaturation or overcooling of the medium.
they often introduce a generalization of the con- At the first stage projections appear which develop
cept of depolarization factors, considering parti- into axes (trunks, first-order branches) at which
cles placed in a medium with the dielectric con- branches of second and higher orders form (see
stant εm > ε0 so that besides the purely geomet- Crystal morphology). The directions of dendrite
ric factor (1) there appears a dependence on the growth usually coincide with the axis of a pyramid
dielectric properties of both the particle and the with its sides formed by planes of greatest reticular
medium. If the principal axes of tensors ε̂ and N density: {100} in a face-centered or body-centered
coincide, the generalized factor has the form: cubic lattice, {1010} in a hexagonal close-packed
lattice, and {110} in a body-centered tetragonal
i = εi − εm Ni .
N (4)
εi − ε0 εm /ε0 lattice. In the structures formed by these growths
DENSITY FUNCTIONAL THEORY 289
(dendrite structure) rounded and faced dendrites particle system with strong interactions to a single-
are singled out. The former are characteristic of particle problem. Along with Fermi liquid theory
most crystals of pure metals, and of some alloys and many-particle perturbation theory (method of
and organic materials which grow from the alloy Green’s functions), density functional theory pro-
and have low melting entropies. Edged dendrites vides a formalism for explaining many-particle
have been found during crystallization from the systems. The basis for density functional theory
melt of many nonmetallic materials (semiconduc- is the Hohenberg–Kohn theorem (P. Hohenberg,
tors, oxides, semimetals), and also upon crystal- W. Kohn, 1964), which states that the energy of a
lization from vapor and solution. system of interacting fermions (usually electrons)
in the ground state is a function of the total parti-
DENSITY cle density only; an external field appears in this
One of the fundamental physical characteris- functional only as the term Vext (r)ρ(r) dr, and
tics of a body, which is numerically equal to the the functional minimum in the given field corre-
mass of a unit volume: ρ = dm/dV , where dm sponds to the true ground state in this field. Usu-
is the mass of a differential element of this sub- ally the following energy contributions are explic-
stance of volume dV . The mean density of a body itly separated out in the density functional: the
is defined as the ratio of the mass of body m to kinetic energy of noninteracting fermions of the
its volume V , i.e. ρ m = m/V . The ratio of the same density ρ(r), the energy of interaction with
densities of two materials under certain standard the external field, and the Coulomb (Hartree) en-
physical conditions is referred to as relative den- ergy. The remaining contributions to the energy
sity. In the case of porous and loose materials, are called exchange-correlation (Exc ) ones (in the
two types of density are distinguished: true den- context of many-particle theory this term has a
sity (not including cavities present in the body) somewhat different meaning; the difference has
and apparent density (ratio of mass of body to to- to do with introduction of part of the kinetic en-
tal volume occupied by it). The SI unit for density ergy into Exc ). To determine the density distribu-
is kg/m3 , the CGS unit is g/cm3 . As a rule the tion, which minimizes the above functional, in the
density increases with increasing pressure and de- form of the density of a certain system of nonin-
creases with increasing temperature (due to ther- teracting particles under the influence of the ef-
mal expansion), except near phase transitions. The fective potential, the effective potential itself is
value of the density undergoes an abrupt change as expressed through an externally applied potential,
a material undergoes a change of state (see States the Hartree potential and the exchange-correlation
of matter): it sharply decreases upon the transition potential Vxc (r) = δExc {ρ}/δρ(r). The resulting
into the gaseous state, and increases on solidi ca - set of self-consistent equations is referred to as
tion (there exists an anomaly in the behavior of the the Kohn–Sham equations (W. Kohn, L. Sham,
density of water and cast iron: it decreases at the 1965). In the limit of nearly constant density, the
liquid–solid transition). The value of the density of local density approximation
is valid, according to
a body is determined by precise determination of which Exc = ρ(r)εxc (ρ(r)) dr , where εxc (ρ)
the mass and volume of this body by application of is the density of the exchange-correlation energy
various types of densimeters. Also used in density (see Correlation energy) of the uniform electron
determinations is the dependence of density on the gas of density ρ. Most density functional theory
velocity of sound propagation, the intensity of γ - calculations are carried out in the context of this
and β-radiation passing through a substance, etc. approximation. The accuracy of this approxima-
tion is quite high for the calculation of observ-
DENSITY FUNCTIONAL, NONLOCAL able properties of real solids (equilibrium volume,
See Nonlocal density functional. compressibility, etc.), but there exists a number
of problems (e.g., description of the interaction
DENSITY FUNCTIONAL THEORY of an electron with a metal surface), which re-
Method involving a formal, accurate reduction quire greater accuracy. Available for application
of the problem of the ground state of a many- are also nonlocal density functionals, in which
290 DENSITY MATRIX
Exc is an integral function of the electron den- of Exc . The errors involved in calculations of vari-
sity in a certain region (mean density approxima- ous quantities, based on the local density approxi-
tion, weighted density approximation), and gra- mation theory, are (compared to experiment) from
dient density functionals. The simple expansion, several percent (equilibrium lattice constant, mag-
with the gradient-dependent part of Exc propor- netic moment of ferromagnets, etc.) to tens of per-
tional to the square of the density gradient, is now cents for elastic moduli, etc.
rarely used. Widely accepted are Langreth–Mehl
processes (D.C. Langreth, M. Mehl, 1983) and the DENSITY MATRIX, statistical operator (J. von
generalized gradient expansion, with its compli- Neumann, L.D. Landau, 1927)
cated dependence of Exc on the density gradient. Operator ρ useful for the calculation of the
Density functional theory may be generalized to mean value of any physical quantity in quantum
treat magnetic systems (spin density functional) at statistical mechanics. The density matrix plays
finite temperatures, and in time-independent exter- the same role in quantum statistical physics as
nal fields. There exists a relativistic formulation of the distribution function does in classic statisti-
density functional theory. cal mechanics. Ordinary quantum mechanics deals
Density functional theory provides a formal with systems characterized by a definite state
basis for the band theory of solids, since it per- vector |ψ. The mean value A of any physical
quantity, represented by an operator A, in this
mits the treatment of the band structure of nonin-
state is A = ψ|A|ψ. Quantum-statistical me-
teracting particles under the influence of an effec-
tive periodic potential; the properties of the ground chanics describes systems that cannot be repre-
sented by a state vector, but may be described
states of this calculated band structure are in full
by giving probabilities W1 , W2 , . . . for existing in
agreement with the properties of the ground states
states |ϕ1 , |ϕ2 , . . . , respectively. These systems
of the crystal under consideration. In combina-
are called statistical mixtures. The mean value in
tion with band structure computation methods, the
such systems is defined by the formula
density functional theory allows the calculation of
the following parameters: equilibrium crystal lat- A= Wk ψk |A|ψ k , Wk = 1,
tice constants (to an accuracy of several percent), k k
elastic moduli and phonon frequencies (accurate
which may be written in the following contracted
within 10–20%), magnetic moment in magnetic
form:
substances (correct to several percent), distribu-
tion of electron and spin densities, relative stabil- A = Tr(
ρ A), =
ρ Wk |ψk ψk |,
ity of crystal structures, etc. k
The formulation of density functional theory where Tr is the trace of the operator (i.e. sum of its
using the Kohn–Sham equation results in the diagonal elements). In order to write the operator
generation of an additional spectrum which, in in the form of a matrix ρmn in some representa-
ρ
general, does not coincide with the actual spec- tion, one has to select an appropriate complete set
trum of single-particle excitations of the system of wave functions {|ϕi }. Then
|ψk =∗ i aki |ϕi
(since density functional theory is a ground state and ρmn = ϕm | ρ |ϕn = k Wk akm akn . The
theory). At the same time, the assumption of probability of finding a system in a (“pure”) state
a close relationship between these spectra has |ϕm from the overall set {|ϕi } is given by the di-
become the usual practice; this assumption al- agonal matrix element ρmm . In the general case,
lows one to calculate observed X-ray, optical the system is in the “pure” state, if its matrix
and other spectra, kinetic characteristics, electron– operator satisfies the condition ρ2 = ρ . To find
phonon interaction, etc. Thus, density functional the operator ρ , nonequilibrium quantum statisti-
theory provides an opportunity to derive system cal mechanics uses the quantum Liouville equa-
response functions; in the context of this theory tion (Neumann equation) which plays the same
these functions are adequately represented by the role as the Schrödinger equation for the time evo-
above-mentioned additional single-particle spec- lution of “pure” states (see Quantum kinetic equa-
trum, and the second-order variational derivative tion). When studying systems that interact with the
DENSITY OF STATES 291
environment the reduced density matrix is used. Of particular importance for metals is the den-
It is obtained through averaging the total den- sity of electron states at the Fermi surface. Meth-
sity matrix of the whole system (quantum sys- ods of determining the density of states are highly
tem plus environment) over variables of the lat- diversified, and depend on the type of spectrum.
ter. The reduced density matrix plays an important Thus, the best results for the phonon and magnon
role in investigating the system response to exter- cases of different materials over a wide range of
nal disturbances, as well as in studies of such ir- energies are obtained by application of inelastic
reversible processes, as transport phenomena, re- neutron scattering, whereas the electronic density
laxation processes, etc. of states may be determined by employing X-ray
photoemission (see X-ray emission spectroscopy).
DENSITY OF ELECTRON STATES In periodic structures (crystals) the states are
One of the major characteristics of the elec- described by the quasi-wave vector k, and the
energy spectrum is determined by the dispersion
tron spectrum of solids and other many-electron
law ελ (k), where λ is the number of the spectral
systems. The density of electronic states g(E)
band or branch. For these structures, the density
equals the number of electronic states within a
of states per unit cell may be represented as
unit interval of energy. If the state characterized
by quasi-wave vector k in the band λ has the 1
g(E) = δ E − ελ (k)
energy Ekλ , then the density of electronic states N
λk
per unit
cell is given by the expression g(E) =
(1/N) kλ δ(E − Ekλ ) (see Density of states). If ν0 dS
= ,
the electron wave functions are expanded in terms (2π)3 νλ
λ
of a certain set of basis functions, then the partial
densities of electronic states may be introduced, where N is the total number of cells, δ(x) is the
which define the relative contribution of a given Dirac delta function, ν0 is the unit cell volume,
basis function (e.g., orbital state) over a certain en- and integration with respect to dS is taken over
the surface in k -space (within the Brillouin zone),
ergy interval. Of particular importance is the elec-
on which ελ (k) = E, and νλ = (1/)|∇k ελ (k)| is
tronic density of states at the Fermi level. It deter-
the group velocity at this surface.
mines kinetic, superconducting, and other proper-
For the case of the phonon spectrum in three-
ties of a material; and it bears a direct relation to
dimensional crystals, the spectral density of states
the stability of a crystal lattice and to the thermo-
behaves as g(E) ∝ E 2 at low values of the en-
dynamic properties of electrons.
ergy due to the linear form of the dispersion law
(for two-dimensional systems g(E) ∝ E, whereas
DENSITY OF STATES, spectral density of states for one-dimensional ones the density of states is
Number of the energy levels of a system per constant in this energy region). Van Hove singu-
unit energy interval (in more exact terms, the ratio larities occur in the density of states in the neigh-
of the number of levels within a certain energy in- borhood of edges of phonon, electron, and other
terval to the value of the said interval, when the lat- types of spectra, as well as within the allowed
ter tends to zero). Phonon, electron, magnon, exci- spectral regions, and they feature zero group ve-
ton and other densities of states are distinguished locity. In the case of three-dimensional systems,
in solids. These different types of densities of g(E) ∝ (E − Ec )1/2 in the neighborhood of these
states correspond to different types of elementary points, whereas one- and two-dimensional sys-
excitation spectra of disordered solids. The density tems show sharper singularities.
of states determines the spectral (cross-section of In imperfect crystals, on the one hand, the
absorption and scattering of radio frequency and density of states peaks in the neighborhood of
X-ray radiation, light, neutrons, gamma quanta, energies of impurity localized states (see Local
ultrasound), thermodynamic (speci c heat, ther- electronic levels), which become more and more
mal expansion, etc.), and a number of kinetic prop- broadened with increasing concentration of de-
erties of a material. fects, and on the other hand, the smearing of Van
292 DEPLETION
and they are also widely used in electrooptical paramagnetism is either absent or is weaker than
modulators, shutters and deflectors of different diamagnetism. The magnetic susceptibility of dia-
laser devices. The complete deuteration of crys- magnets is negative and low, and the molar mag-
tals of triglycinesulfate increases the Curie tem- netic susceptibility χ is of the order of −10−5 to
perature from 322 K to 333 K. Simultaneously −10−6 . Many liquids and gases, the molecules
with the increase of T C the light absorption of (or atoms) of which do not have inherent mag-
the crystals D-KDP and cesium dideuteroarsenate netic moments (inert gases, H2 , N2 , F2 , H2 O,
(D-CDA) decreases to λ = 1.06 μm, which pro- some organic compounds) are diamagnets, and the
motes their application in the wide-aperture el- strongest diamagnetism appears in aromatic com-
ements of neodymium lasers at installations for pounds. Many molecular crystals formed by such
controllable laser thermonuclear synthesis. Some molecules are diamagnets, as well as the series of
ferroelectrics with hydrogen bonds, e.g., Rochelle covalent and ionic crystals, containing no atoms of
salt, exhibit only a small shift of T C upon deuter- the transition and rare-earth elements (NaCl, CsCl,
ation. Ge, Si). Metals are diamagnets when Landau dia-
magnetism and diamagnetism of the ionic cores
DEVICES, SEMICONDUCTOR prevail over the Pauli paramagnetism (e.g., cop-
See Semiconductor devices, Microwave semi- per, gold, zinc, graphite). Type I superconductors
conductor devices, Photocontrolled microwave with χ = −1 in mks units (χ = −1/(4π) cgs) are
semiconductor devices. considered as ideal diamagnets (see Absolute dia-
magnetism). Anisotropy (tensor character) of the
magnetic susceptibility, which is especially strong
“DEVIL’S STAIRCASE”
for molecular crystals of aromatic hydrocarbons
A temperature- or pressure-stimulated step-
and metals with an anisotropic electronic spec-
wise change of the wavevector k s of a superlat-
trum (graphite), is characteristic of many crys-
tice in a crystal whose compositional phase dia-
talline diamagnets.
gram exhibits a sequence of thermodynamically
stable phases with various integral (in regard to DIAMAGNETIC DOMAINS
reciprocal lattice vectors) values k s . The possibil- Uniformly magnetized macroscopic regions of
ity of a stepwise change of the vector k s that de- a diamagnetic metal into which it spontaneously
termines the period of a superlattice modulated decomposes in an external magnetic field. The ap-
phase with a gradual monotonic change of the ex- pearance of the diamagnetic domains is caused by
ternal thermodynamic parameters was proposed the fact, that at low temperatures the relationship
by I.E. Dzyaloshinskii (1964). For such a step- between the applied magnetic field intensity H
wise change the vector k s runs through a series and the internal magnetic field B involves the
of sequential magnitudes corresponding to non- de Haas–van Alphen oscillations (see de Haas–
symmetric (“random”) points of the Brillouin zone. van Alphen effect) (see Fig.). The sections of the
In principle there are no limits to restrict the curve H (B) with ∂H /∂B < 0 reflect a thermody-
number of steps of a “devil’s staircase”, however, namically unstable state (see Schoenberg effect)
for stability with regard to thermal and quantum which brings about second-order phase transitions
fluctuations only those phases appear which have with a jumpwise variation of the magnetization M.
small translational periods. As the period grows, The magnetization jump
M = M2 − M1 oc-
the region of existence of the corresponding phase curs at H = H c and has the magnitude
M =
shrinks. Stepwise changes of the vector k s were (B2 − B1 )/μ0 as a result of the equality of the
observed in the crystal CeSb at temperatures be- areas from B1 to B2 that are shaded in Fig.
low TN = 16 K. This equality gives for the value of the integral
2
1 H (B) dB = H c (B2 − B1 ). For a sample of el-
DIAMAGNET lipsoidal shape with the dimensionless demagneti-
A material which becomes magnetized in the zation factor N (0 N 1), placed in the uniform
direction opposite to that of an external magnetic magnetic field B0 = μ0 H0 , the jumps of magne-
e ld; a material without magnetic ordering, where tization are absent, and upon the variation of the
294 DIAMAGNETISM
decreases with increasing porosity. Also ob- Diamond-like films as a rule are a heterophase,
tained artificially. containing regions of short-range order (clusters)
(4) Impact diamonds, found in meteorites and with sp3 -, sp2 - and sp-hybridization C–C and
in meteorite craters, are cryptocrystalline tex- C–H bonds, i.e. the films may contain the phases
tured aggregates containing considerable of diamond, graphite, and carbyne, respectively.
(sometimes up to 50%) hexagonal phase or The phase C8 has been also discovered. The for-
lonsdalite. Density is 3.3 to 3.5 g/cm3 . They mation of other metastable allotropic modifica-
have a high specific surface area (20 to tions of carbon (see Allotropy) is possible as well.
40 m2 /g), high abrasive capacity, and en- The dimensions of clusters, their composition and
hanced birefringence (0.0005 to 0.0100). Ar-
correlation with different kinds of bonds hybridiza-
tificial analogues include explosive diamond
tion are determined by the contents, average en-
obtained by exposing graphite to a shock
ergy and density of the particle flux, from which
wave.
the formation of film occurs, and also by the na-
Synthesis of diamonds. (1) At high pressure by ture of the substrate and its temperature. Hydro-
the direct transformation of graphite to diamond gen content may reach 25%. The density of films
above 12 GPa and 1300 K. (2) Transformation of lies within the range 1.0 to 3.5 g/cm3 ; refraction
graphite to diamond via the melt of a metal (Mn, index is 1.6 to 2.7, optical width of the band gap is
Fe, Co, Ni, etc., and their alloys). Crystallization of 0.8 to 2.6 eV. Diamond-like films may be used as
the dissolved carbon takes place above 4 GPa and durable, masking, passivating, anti-reflection coat-
1500 K. (3) Explosive method exposes graphite ings of optical elements for different ranges of
to a shock wave with P > 49 GPa. (4) Detona-
wavelengths, as coatings for optical storage disks,
tion method using explosive gas containing extra
UV range lasers, and also for the design of fun-
carbon, at pressure 13 to 25 GPa and tempera-
damentally new active elements for microelectron-
ture 3000 to 4000 K. Rapid adiabatic expansion
ics and optoelectronics. Density of states at the
of explosion followed by rapid cooling yields 20
to 90% diamond. (5) Synthesis from atomic car- boundary of a semiconductor upon application of
bon (via epitaxial growth, deposition from ionic a diamond-like film can be 1010 to 5·1011 cm−2 .
beam, condensation of vapors under laser heat-
ing, and ionized carbon deposition from subnor- DIAZOTYPE MATERIALS
mal discharge plasma).
See Nonsilver photography.
Both natural and synthetic diamonds are used
as extremely hard abrasive material for manufac-
turing of grinding, polishing, cutting, drilling and DICHROISM OF CRYSTALS (fr. Gr. δικρoνoς ,
other tools; in electronic engineering as heat elim- two-colored)
inating elements, luminescent radiation sources, Dependence of light absorption on its polariza-
temperature sensors, photoreceivers, nuclear radi- tion (see also Pleochroism); in the narrower sense
ation detectors. The semiconducting properties of
the term is sometimes applied to uniaxial crystals.
diamond are used in solid-state physics; its hard-
The latter, due to the anisotropy of absorption,
ness and transparency are utilized in high-pressure
have two “main” colorations: the “basic” one and
apparatus.
the “axial” one for white light propagating, respec-
DIAMOND-LIKE CARBON FILMS tively, along and perpendicular to the optical axis.
Metastable carbon or hydrogen-containing car- Linear dichroism (difference of absorption of radi-
bon lm s with hardness, density, optical and elec- ation linearly polarized in mutually perpendicular
trical properties, and chemical stability close to di- directions) and circular dichroism (difference of
amond, whereas the structure of films differs from absorption of right-and left-hand circularly polar-
that of diamond. The structure of diamond-like ized radiation) are singled out. Circular dichroism
films is amorphous, though in some cases the pres- of crystals is associated with spatial dielectric dis-
ence of the diamond phase crystals is observed. persion.
DIELECTRIC CONSTANT 297
where Π(k) is the static polarizability (polariza- vicinity of a resonance the dielectric loss exhibits
tion operator) of the free charge carriers (electrons, a maximum. In the case of inertial (delayed re-
holes) and εi is the relative dielectric constant of sponse) mechanisms of polarization, the value of
the lattice caused by the polarization of the bound the dielectric loss is determined by the relaxation
electrons in the shells of atoms, and by the dis- time τ of the medium. The Debye theory of dielec-
placement of ions from their equilibrium positions tric relaxation applied to relaxation processes act-
in ionic crystals (or by the orientation of molecu- ing when an applied electric field is switched on
lar dipole moments in polar crystals). Eq. (7) de- provides the expression
scribes the effects of screening the electric fields
(εs − ε∞ )ωτ
and the Coulomb interaction of the charges (see tan δ = ,
Electric charge screening). εs + ε∞ ω 2 τ 2
In dielectrics (insulators) the difference be- where εs and ε∞ are respectively the static
tween the local electric field at the point where an (ω → 0) and high-frequency dielectric constants,
atom (ion) is located and the average macroscopic and tan δ has a maximum at the electromag-
field leads to the Clausius–Mossotti equation, and netic field frequency ωM = (1/τ )(εs /ε∞ )1/2 . The
in crystals with molecules possessing electric di- small dielectric loss at low frequencies ω ωM
pole moments it leads to the Onsager–Kirkwood is explained by the slow response of the medium,
relations and to the Langevin–Debye equation. and at high frequencies it is explained by the fact
In the CGS system the relative and absolute that the inertial response does not have time to
dielectric constants or permittivities are equal to follow the changes in the field when ω
ωM .
each other and are all dimensionless. In the SI sys- For relaxation processes involving the reorienta-
tem the absolute εa and relative (dimensionless) ε tion or displacement of ions, the temperature de-
are related by the expression εa = εε0 where ε0 is pendences of τ (T ) and ωM (T ) are significant.
is the dielectric constant of vacuum with the value Some relaxation mechanisms result in deviations
8.8542·10−12 F/m. from the Debye theory. The dielectric loss caused
by the electrical conductivity, σ , of a dielectric in
DIELECTRIC DISPERSION
the absence of other types of loss is determined
See Spatial dielectric dispersion.
by the relation tan δ = σ/(ε0 ω) (mks units) and
grows in proportion to σ as the temperature in-
DIELECTRIC DISPERSION, TEMPORAL
creases. For ferroelectrics, specific loss associated
See Temporal dielectric dispersion.
with the reorientation of domains in the electric
DIELECTRIC FUNCTION field, as well as the presence of anomalies in the
See Dielectric constant. vicinity of the phase transition temperature, can
play a significant role.
DIELECTRIC LOSS
A part of the electric field power transferred DIELECTRIC MEASUREMENTS, dielectrometry
to an insulator (dielectric) located in an electric Measurements of electric polarization and of
field, and dissipated in the form of heat. The value technical parameters of insulators at frequencies
of the dielectric loss is characterized by the di- below the optical range. Between 0 to 108 Hz
electric loss tangent angle δ in the expression capacitor methods of measurement are used: the
tan δ = ε /ε , where ε and ε are the real and relative (dimensionless) dielectric constant ε =
imaginary parts of the dielectric constant (or per- Cd/A is determined from the geometric dimen-
mittivity), respectively. There are various mecha- sions of a parallel plate capacitor with capaci-
nisms of dielectric loss in solid dielectrics which tance C, area A and thickness d, filled with the
lead to specific frequency and temperature depen- dielectric under study. The dielectric loss (loss
dences of tan δ. Losses via the resonance excita- tangent, or tan δ) is found from the variations of
tion of bound electrons of dielectrics or of oscil- the phase ϕ of alternating current passing through
lating ions occur at a series of resonance frequen- the capacitor. Typical equipment includes bridges,
cies with a weak temperature dependence. In the Q-meters, impedance measuring devices, phase
DIFFRACTION GRATING 299
meters and dielectrometers. Within the range 108 where t is time and P (0) = 0. The value of β de-
to 1011 Hz there are waveguide (microwave) pends on the temperature of the ferroelectric, and
methods of determining ε and tan δ, as well as in the region of high fields it begins to depend
resonator ones. The waveguide methods are based on E.
on measuring the standing wave ratio, the phase
of the reflected signal, or that of the signal pass- DIELECTRONICS, dielectric electronics
ing through the investigated dielectric sample. In A branch of physics and engineering which
resonator methods the change in the frequency deals with transformation of electric, acoustical,
and the Q (loss or quality) factor of a resonator mechanical and electromagnetic waves by the uti-
are measured when the sample of dielectric is in- lization of particular attributes of insulators (di-
troduced. Measuring devices include generators electrics) such as piezoelectric, pyroelectric, elec-
and lines, frequency meters, reflectometers, phase trooptic, and other properties. The response of a
meters, etc. At the frequencies 1011 to 1015 Hz dielectric to an external action or perturbation can
quasi-optical (beam) methods are employed, which entail either a transformation of the perturbing en-
permit the determination of the wavelength of ergy into another kind of energy, such as elec-
electromagnetic radiation in the dielectric and its trical or optical energy (or in the reverse direc-
attenuation there: Fabry–Perot interferometers, go- tion), and so on, or it can involve the nonlinear
niometers (see Goniometry), refractometers (see optical, electric, magnetic, or acoustical proper-
Refractometry). Fourier spectrometers (see Fourier ties of dielectrics. Quite a number of discrete and
spectroscopy) can analyze the response of di- integrated devices of microelectronics, acousto-
electrics to a pulsed electric field. electronics, and optoelectronics are widely used in
data processing systems, and in information trans-
DIELECTRIC RELAXATION fer systems. Dielectronic-based devices are em-
See Maxwell relaxation. ployed in the energetics for, e.g., the transforma-
tion of nuclear radiation energy or solar energy
DIELECTRICS into electricity, etc.
See Insulators. In the narrow sense, the word “dielectronics”
is the name of a branch of solid state physics
DIELECTRIC VISCOSITY
that deals with “semi-insulators”. These materi-
Phenomenon of a time lag in the variation of
als have properties intermediate between insula-
electric parameters (polarization of insulator, di-
tors and highly conductive semiconductors, with
electric constant, etc.) in ferroelectric materials
their specific electrical resistivity in the range
depending on the intensity of the external elec-
103 –1010 ·cm. Classified among them are stron-
tric field. Viscous phenomena in ferroelectrics are
gly compensated semiconductors, materials which
similar to the those of magnetic viscosity and su-
have a relatively wide band gap. The nonlinear
perviscosity in ferromagnets. As a parameter of
electrical properties of semi-insulators are con-
the switching properties of ferroelectric crystals
trolled through the injection of current carriers by
during the rearrangement of the magnetic domain
means of contacts; their current–voltage character-
structure the dielectric viscosity coefficient β is in-
istics are sensitive to the presence of intrinsic and
troduced, and it determines the relationships be-
extrinsic defects in the sample.
tween the value of the external electric field and
the velocity of processes involving the variation DIFFRACTION
of dielectric parameters under the effect of this See also Acoustic diffraction, Acousto-optic dif-
field. The equation describing the transition of fer- fraction, Electron diffraction, Laue diffraction, X-ray
roelectricity in the equilibrium state with polariza- diffraction.
tion P0 under the effect of the constant electric
field with intensity E reflects the role of the dielec- DIFFRACTION GRATING
tric viscosity during the processes of polarization: Periodic structure of many regularly arranged
Et elements at which the diffraction of light takes
P (t) = P0 1 − exp − , place. Diffraction gratings may be flat or concave,
β
300 DIFFRACTION METHODS OF INVESTIGATION
reflecting or transmitting. The grating decom- The first three methods have a sufficiently com-
poses the incident light beam into its component mon basis. During elastic scattering of incident ra-
wavelengths according to the equation d(sin α + diation by periodically spaced atoms of the crys-
sin β) = mλ, where d is the grating period, α and tal lattice there appear sharp maxima of the scat-
β are the angles of incidence and diffraction; λ is tered radiation and the diffuse background, caused
the wavelength, and m = 0, ±1, ±2 . . . is the or- by inelastic (Compton) scattering. By the analy-
der of diffraction. The basic parameters of a dif- sis of the positions and intensities of the diffrac-
fraction grating are: angular dispersion
β/
λ = tion maxima the type of crystal lattice, unit cell
m/(d cos β); resolution R = λ/
λ = mN (N is size and atomic coordinates may be determined.
number of lines of the diffraction grating); free Crystal lattice defects change the positions, in-
spectral region Fλ = λ/m. The diffraction effi- tensities and shapes of the profiles of the diffrac-
ciency is the fraction of the intensity of a mono- tion maxima; and can add extra diffuse scattering
chromatic light beam diffracting in the required which helps to provide information about the de-
order. To increase the number of lines of a diffrac- fects in the crystal. Indirect methods are also used
tion grating a step like structure of grooves can for investigating the structure of amorphous and
be used (echelette). Holographic gratings, which liquid substances. With the help of electron dif-
record in photosensitive media upon the interfer- fractometry it is convenient to study the thin sur-
ence of laser beams, may be thin (thickness l d) face layers of a material, and with the help of neu-
or bulk (l > d), and are of high quality. The dif- tron diffraction the compositions of elements with
fraction efficiency may be close to 100%. Gratings close atomic numbers, and the ordering of alloys.
are used in spectral devices, dispersion laser res- A method using the orientation effects of charged
onators, as sensors of linear and angular motions, particle scattering has found wide application. For
as beam dividers, etc. the analysis by the method of orientation effects
particles with energy from several electronvolts to
DIFFRACTION METHODS OF INVESTIGATION kilo- and megaelectronvolts are used. Proton mi-
Methods of investigating materials based on croscopes are employed for investigation of var-
the scattering (diffraction) of X-rays, electrons, ious structural defects. The method is especially
neutrons and other microparticles by the atoms effective for the study of the localization of im-
of the object under investigation. They are subdi- purities within a unit cell of the lattice, as well at
vided into indirect and direct ones. With the help a small distance below the crystal surface. Differ-
of indirect methods based on the measurement ent modifications of this method are based on the
of the parameters of the scattered radiation inten- channeling of particles, on the effect of shadows,
sity, the atomic structure of crystals, the character and on the double orientation.
and density of defects in crystals are investi- The direct methods are used for the investiga-
gated, a phase analysis is performed; the structure tions of defects of crystalline material (e.g., for
of amorphous and liquid substances is studied. dislocation structure studies). Some examples of
The indirect diffraction investigations are car- such methods: transmission electron microscopy
ried out using special equipment – diffractometers and X-ray topography. Upon using these methods
(X-ray and neutron types), electronographs, etc. the local variations in the intensity of the diffrac-
For recording the intensity of the scattered radi- tion maxima, caused by the defects of the crys-
ation the following equipment and materials are talline structure, are analyzed. The direct diffrac-
used: photographic films and plates, and also ion- tion methods are also used for investigations of
ization and scintillation counters (detectors) with magnetic domain structures in ferromagnets and
the intensity recorded on ruled paper, or with the ferroelectrics.
output of the counter signals to printing devices.
X-ray structure analysis, electron diffraction analy- DIFFRACTION OF LIGHT
sis, neutron diffractometry and methods of analy- Deviation of the light waves in media with op-
sis using the orientation effects on the scattering of tical non-uniformities of the order of a wavelength
charged particles are related to indirect methods. and more (e.g., at the boundaries of bodies, slits,
DIFFRACTION OF WAVES 301
etc.); a particular case of the diffraction of elec- pendular oscillations of intensity of the diffracted
tromagnetic waves (see Diffraction of waves). The beams due to the scattering at the ultrasonic super-
approximate theory of diffraction of light was for- lattice. In the case of neutron diffraction it is nec-
mulated by A. J. Fresnel (1816). Diffraction man- essary to take into account not only the exchange
ifests itself in the formation of a diffraction pic- of momentum, but also of energy (about 10−7 eV)
ture (alternation of intensity maxima and minima). of the ultrasonic wave. The high sensitivity of scat-
The Fresnel band (zλ)1/2 (λ is the wave length, tering of ultrasound is used for the detection of
z the distance from the object of size a from which the structure of vibrational modes in piezoacoustic
diffraction takes place to the observation point) is converters (see Acousto-electronics).
an important notion in the theory of diffraction.
For a ∼ (zλ)1/2 diffraction occurs for converging DIFFRACTION OF WAVES
beams (Fresnel diffraction, A.J. Fresnel), and for The deviation from rectilinear propagation of
a (zλ)1/2 it occurs for approximately parallel waves in a medium containing non-uniformities
beams (Fraunhofer diffraction, J. von Fraunhofer). or obstacles. The mathematical description of the
Diffraction determines the resolution of optical de- wave fields (sound, X-rays, neutrons, electrons)
vices. The properties of diffraction from volume follows from the Huygens principle (see Acoustic
diffraction gratings form the basis of holography. diffraction), and in this sense it is similar to the
In solid-state physics there is an important role of description of the diffraction of light. Wave dif-
the diffraction of light from dynamic optical non- fraction is usually discussed for the case of an
uniformities, including those caused by ultrasound appreciable deviation from the laws of geometri-
(acousto-optical diffraction) by the self-dispersal cal optics. For example, upon the scattering of a
of radiation within the medium (self-effect of light plane wave of monochromatic light from the ob-
waves in non-linear optics, self-diffraction of light ject (or from a non-uniformity of a medium) with
in real-time holography), and by the spontaneous the characteristic dimension a the angular devia-
and stimulated inherent elastic vibrations of solids tion of the beams is θ ∼ λ/a (λ is the wavelength).
and liquids (see Brillouin scattering). In lasers dif- Therefore, at a screen located at the distance l
fraction determines the minimum (diffraction) di- from the object there appear linear displacements
vergence of beams of the stimulated radiation ϕ ∝ of beams in the transverse direction of the order
λ/a, and many properties of optical resonators. x ∼ lθ ∼ λl/a. For the formation of a sharp geo-
Operation of a diffraction grating, of a zone plate, metrical image of the object the condition x a
of acoustooptical modulators and deflectors (see or l a 2 /λ should be satisfied, which is the cri-
Light de ectors) is based on the diffraction of terion for the applicability of geometrical optics.
light. For l ∼ a 2 /λ (Fresnel diffraction) the linear dis-
placements of the diffracted beams at the screen
DIFFRACTION OF RADIATIONS in ultrasonic are of the order of a. At the screen different beams,
e ld deviated from rectilinear propagation, may con-
This is usually investigated in perfect mono- verge and interfere, so that a complicated distribu-
crystals with application to X-rays and neutrons. tion of diffraction radiation intensity is observed
The effect of ultrasound depends on the corre- at the screen, that is not unambiguously associ-
lation between the ultrasound wavelength λ and ated with the optical properties of the scattering
the extinction length Λ of radiation in crystal. For object (e.g., at the image center of a nontranspar-
λ
Λ ultrasound increases the intensity of the ent object a bright spot (Arago spot) may be ob-
diffracted beams due to the widening of the an- served; the Arago–Poisson–Fresnel effect, 1811).
gular interval of the rays, satisfying the Bragg law, For l
a 2 /λ (Fraunhofer diffraction) at every
and it also suppresses the anomalous passage of point of the screen the beams, deviated approxi-
X-rays. The suppression is sharply enhanced under mately in the same direction, converge and inter-
conditions of X-ray acoustic resonance at λ ≈ Λ. fere; and hence (at least for small deviation an-
If λ < Λ ultrasound causes transitions between the gles) the intensity distribution at the screen repro-
branches of the dispersion surfaces which deter- duces a two-dimensional Fourier-image (of obsta-
mine the dynamics of radiation in the crystal, the cles or non-uniformities), because the coordinates
302 DIFFRACTION POLARIZATION
DIFFRACTOMETRY, NEUTRON
See Magnetic neutron diffractometry, Neutron
diffractometry.
DIFFRACTOMETRY, X-RAY
See X-ray diffractometry, Energy-dispersion
X-ray diffractometry.
fluxes. The following types of diffusion are sin- The kinetics of diffusion is determined by the
gled out: self-diffusion (in a medium of similar second Fick law
diffusing particles, e.g., tracer atoms), heterodif- ∂n
fusion, and interdiffusion. The rate of diffusion is = − div j.
∂t
provided by the diffusion coef cient D, which lin-
The diffusion flux also appears under the influence
early relates the flux density j of diffusing parti-
of any external or internal fields which determine
cles to the gradient of their concentration through
the gradient of the chemical potential μ of the par-
the first Fick law
ticles – the “motive force” of diffusion:
j = −D grad n, (1) D
j = nv = −n ∇μ. (2)
kB T
where n is the number of particles per unit volume.
In gases diffusion occurs through random colli- It can be seen that the drift velocity v of the par-
sions of diffusing particles with the gas molecules, ticles is proportional to the motive force (−∇μ)
and D is determined by the expression and to the mobility b which is related to D through
the Einstein relation b = D/(k B T ). In gases or di-
1 1 λ2 lute solutions ∇μ = k B T ∇n/n, and Eq. (1) fol-
D = λv = ,
3 3 τc lows from (2) as a particular case. In the general
case the motive force of diffusion is determined
where λ is the mean free path, v is the average
by the concentration gradients as well as by the
speed of the thermal motion, τ c is the average
types of interactions of the diffusing particles with
time between collisions; and D is inversely pro-
the fields. For example, for diffusing particles with
portional to the gas pressure with a weak temper- charge q in an electric field E the diffusion flux is
ature dependence (∝T 1/2 ).
In condensed systems (liquids, solids) diffu- D
j = −D∇n + nqE. (3)
sion occurs via random jumps of atoms between kBT
states corresponding to absolute or relative min- The second term in Eq. (3) determines the flux of
ima of their potential energy (e.g., sites and inter- particles under the influence of the electric field. In
stices of the crystal lattice). The corresponding co- many cases (e.g., gases, liquids, solid electrolytes)
efficient of diffusion is the directed diffusion drift of charged particles in
a2 the electric field determines the electrical conduc-
D≈ , tivity of the medium, and this conductivity σ is pro-
τ
portional to D through the Nernst–Einstein rela-
where a is the characteristic length of an elemen- tion:
tary jump of a particle (of the order of the aver-
Dnq 2
age distance between the molecules), τ is the av- σ= .
erage time between jumps which at high temper- kBT
atures depends exponentially on the temperature: This expression permits the determination of D
τ = τ0 exp[Q/(k B T )], τ0 ≈ 10−12 to 10−13 s from the known value of σ (and vice versa). If the
is the characteristic period of the particle vibra- temperature varies throughout the medium then
tions, and Q is the activation energy. Thus, D ∇(μ/T ) serves as the motive force of particle dif-
depends exponentially on the temperature: D = fusion. The current density j is the sum of the
D0 exp[−Q/(k B T )]; D0 ≈ a 2 /τ0 , which is con- usual flux and the thermodiffusion flux (Soret ef-
sistent with the fluctuation nature of the particle fect)
jumps. At low temperatures the jump between ad-
DQ∗ n
jacent energy minima may be performed by quan- j = −D∇n − ∇T ,
tum tunneling; and the temperature dependence kBT 2
of D is different (see Quantum diffusion). In crys- where the dimensionless quantity kT = Q∗ /(k B T )
tals diffusion is a second-rank tensor Dik , and in is called the thermodiffusion relation, and Q∗ is
cubic crystals this tensor Dik reduces to a scalar. the heat transferred by the particles per unit act of
306 DIFFUSION ANNEALING
DIFFUSION LENGTH
A mean distance LD reached by a diffusing
particle or quasi-particle away from its initial po-
sition within its lifetime τ ; here LD = (Dτ )1/2 ,
and D is the diffusion coef cient. Owing to the
random nature of the diffusion motion and its wan-
dering (zig-zagging) path, the diffusion length is
much shorter than the overall trajectory length of
a diffusing particle. Charged particles, e.g., current
carriers in semiconductors in an electric field E,
drift and diffuse at the same time. The distance tra-
versed by these drifting particles during a time τ is
specified by a delay length which can be longer or
shorter than LD , and in a strong electric field the
drift can dominate. For this case the diffusion is
Sketch of the diffusion flux of material (shown by the
mainly transverse, and within a time τ the current
arrows) arising from single-axis loading.
carriers drift over the distance LE = μEτ called
the drift length (here μ is the ambipolar mobility)
(see Ambipolar diffusion).
DIFFUSIVITY, temperature conductivity factor For elastically isotropic bodies, taking into ac-
The parameter that determines the dependence count only the terms that are linear in uik , we
of the temperature of a material on the coordi- have:
nates and the time during nonstationary thermal 1
δ= σii ,
processes; it serves as a measure of the rate of 3K
i
heat diffusion. In the model of an isotropic ho-
mogeneous body with thermal characteristics in- where σik is the deformation tensor, and K is
dependent of temperature, the diffusivity is a con- the modulus of uniform compression. In isotropic
stant a which enters the thermal conductivity equa- crystals with point defects and dislocations the di-
latation is everywhere zero in the linear deforma-
tion: a∇ 2 T = ∂T /∂t . A formal expression for the
tion tensor approximation except at places where
diffusivity is a = λ/(Cρ) where λ is the thermal
defects are located. Dilatations different from zero
conductivity coef cient, C is the speci c heat (at
can occur for such defects when account is taken
constant pressure), ρ is the density of the material.
of the crystal anisotropy, or when deformation ten-
In the presence of appreciable temperature gra- sor terms which are second order in smallness
dients in a body it is necessary to take into ac- make a contribution.
count the dependence of the thermal characteris-
tics of the material on the temperature. The ther- DILATATION CENTER
mal conductivity equation conserves its form only A point defect which inserts an “extra” infini-
in the case of the replacement tesimal volume
V into a crystal. A macroscopic
of the temperature
by new variable Θ = (1/λ0 ) λ dT , and we have analog of a dilatation center is a spherical cavity
a∇ 2 Θ = ∂Θ/∂t , where a is expressed through the in an elastic medium into which an elastic glob-
variable Θ, and is a function of the temperature. ule of larger volume is inserted. Its elastic field
The dimensionality of a is [a] = [L2 ][t −1 ]; the is equivalent to the field of three pairs of forces
acting along the coordinate axes and concentrated
unit of measurement is m2 /s.
in the center. Elastic strains and stresses decrease
inversely with the cube of the distance from the
DILATATION (fr. Lat. dilato, making wider) center. A dilatation center simulates interstitial im-
The relative local elastic change of volume purities (
V > 0) (see Impurity atoms), substitu-
upon deformation. The element of volume dv after tional impurities (any sign of
V is possible) and
deformation becomes dV : vacancies within the crystal (
V < 0).
dV δXi ∂Ui DILATOMETER (fr. Lat. dilato, extend, and
=
= δik +
dv δxk ∂xk Gr. μετρεω , am measuring)
A device for measuring variations of the di-
(Xk and xk are the Cartesian coordinates of the mensions of bodies (e.g., length) under the influ-
body points before and after its deformation, and ence of external actions, internal physico-chemical
Uk is the displacement vector). When account is processes, or both. The most wide spread types are
taken to second order of the smallness of the strain optico-mechanical (sensitivity 10−6 to 10−7 cm),
tensor interference (10−7 to 10−8 cm), electronic with
inductive, capacitive, electro-mechanical, galvano-
1 ∂ui ∂uk magnetic (see Galvanomagnetic effects), or mag-
uik = + ,
2 ∂Xk ∂Xi netostriction sensors (10−7 to 10−9 cm), radio-
resonance (10−12 cm), and X-ray (10−5 to
we obtain the following expression for the dilata- 10−6 cm). Dilatometers are used for determining
tion: the coefficient of thermal expansion of materials,
dV − dv and for dilatometric analysis.
δ=
dv 2 DILATOMETRIC ANALYSIS
1 1
= uii − uik uki + uii . The analysis of physical and physico-chemical
2 2 properties of solids according to the variation of
i ik i
310 DILATON
their volume or sizes. Dilatometric analysis is and kinetic parameters of metals due to conduc-
based on the tendency of a solid to change its tion electron scattering (see de Haas–van Alphen
volume (dimensions) under the influence of heat, effect, Shubnikov–de Haas effect).
pressure, electric and magnetic fields, ionizing ra-
diation and other external effects, and also due DIODE
to internal physico-chemical processes. The varia- A two-electrode electronic gas-discharge, vac-
tions of volume are characterized by the bulk ex- uum tube or semiconductor device with the prop-
pansion coefficient. Dilatometric analysis is most erty of unipolarity (unidirectionality) of its electri-
widely used for investigations of thermal expan- cal conductivity. Nowadays semiconductor diodes
sion. Processes taking place in solids are charac- play a dominant role in science and engineering;
terized by variations of volume (bulk effects) and and other (non-linear) properties not associated
of the temperature coefficient of bulk expansion. with their unipolarity have found various practical
Absolute measurements (with respect to a fixed applications. A semiconductor junction or metal–
standard) and differential ones (the standard un- semiconductor junction forming a Schottky bar-
dergoes changes similar to the test body) are per- rier transmits current well in one direction and
formed during the course of dilatometric analysis. has a high electrical resistance in the opposite or
This analysis is used to determine the true and av- blocking direction. Because of this, the rectify-
erage temperature coefficients of expansion, their ing action a diode can be used as a semiconduc-
phase composition at critical points, the evaluation tor gate to rectify low-frequency current up to
of their structural state, and for the identification 1 kA (power gate) and for voltage stabilization
of ongoing physico-chemical processes and their (supporting diode). Pulse diodes perform trans-
rates. formations of ultrahigh frequency radiation, e.g.,
for switching, or for detection associated with the
DILATON variation of the radiation frequency. Demodulation
A non-linear excitation that occurs under the or the removal of a modulated low-frequency sig-
action of a stretching load in anharmonic atomic nal from a high frequency carrier wave is based on
chains, or in structures fabricated from such chains this process. Diodes with negative resistance, or
(e.g., in fibrillar crystals of polymers). A dila- generation type semiconductor diodes, have been
ton is a dynamic structural defect which can al- used for generating, amplifying, switching, multi-
leviate failure. Two types of dilatons are distin- plying (increasing the frequency) and other func-
guished: a subcritical dilaton and supercritical tions in the ultra-high frequency range, up to hun-
dilaton. The first is a deformation fluctuation cre- dreds of gigahertz. Gunn diodes which have no
ated by a group of interatomic bonds, each of p–n junction, or Schottky barrier have found wide
which is stretched by an amount εI which lies application; their action is based on the presence
within the range ε < εI < εcr , where ε is the of a negative resistance arising from the charac-
stretching strain, and εcr is the strain correspond- teristics of the energy band structure of the many-
ing to the ultimate strength of the bond. The su- valley semiconductor. The action of a varactor
percritical dilation has in the center of the group (also called varicond) is based on the dependence
an anomalously strained interatomic bond with the of the capacitance, present in the semiconductor
strain εII > εcr . Both types of dilatons are partic- diode due to a p–n junction or Schottky barrier,
ular solutions of the same equation of motion in on the value of the applied external voltage. The
which the cubic and quartic anharmonic terms are non-linear dependence of the capacitance on the
retained in the expansion of the potential energy. voltage forms the basis of the action of parametric
diodes which are used for the amplification and
DINGLE FACTOR (R.B. Dingle, 1952) conversion of ultra-high frequencies. In photodi-
An exponential multiplier of the form odes the separation of non-equilibrium electrons
[−2πm∗ /(eτ B)], (where τ is the average time of and holes by the field of a p–n junction causes
collision, B is the magnetic field; m∗ is the ef- the appearance of a photo-electromotive force; and
fective mass), which describes the amplitude re- this is utilized in highly sensitive detectors of ra-
duction of quantum oscillations of thermodynamic diation in the infrared, visible and near ultraviolet
DIPOLE GLASS 311
to those of spin glasses, and they are related to of elementary dipoles. The dipole moment of two
dipole glasses. For example, in the mixed crys- charges q equal in magnitude and opposite in sign
tals Rb1−x (NH4 )x H2 PO4 the spin-glass like be- separated by the distance l is given by p = ql;
havior is caused by the random freezing of pro- where the vector p is directed from the negative
tons at positions along hydrogen bonds. The term to the positive charge.
structural glass is often used, because the forces Variations of the electrical dipole moment of
of interaction of atoms in such systems are dif- a system system of charges, e.g., brought about
ferent from dipolar interaction forces. Also the by their motion, results in the emission of elec-
orientational or quadrupole glasses of the type tromagnetic waves. The value of a dipole moment
(KCN)x (KBr)1−x , where the random freezing of is measured in debyes, where one debye (1 D) is
the quadrupole moment of CN− occurs, are re- the dipole moment of two unit electric charges +e
lated to the structural glasses. and −e separated by the distance of one angstrom
(e = 1.6022·10−19 C).
DIPOLE, INDUCED
See Induced dipoles. DIPOLE RELAXATION of an insulator
Return of a polarized insulator to its equilib-
DIPOLE LAYER on a surface rium non-polar state after switching off an applied
A system of electric dipoles on a surface due electric field. In the electric field the randomness
to the adsorption of atoms or molecules. A di- of the heat induced motion of weakly-bound elec-
pole moment per unit area np appears (n is con- trons, ions, and dipoles is weakened, and some
centration of adsorbed particles per unit area, of them appear to be “fixed” by the electric field
p is their dipole moment) as a result of the ori- in positions corresponding to the polarized state.
entation of the intrinsic dipoles of the adsorbed Unlike liquids and gases where a dipole is a ro-
particles, of their polarization, or of charge trans- tator with unlimited rotation angles, in a crystal
fer between the adsorbed particles and the surface. the dipole reorientation capability is restrictive: to
A dipole layer alters the work function ϕ of the sur- only a limited number of stable orientations sepa-
face:
ϕ = 4πnp. rated by potential barriers with heights that depend
on the applied field. The following events may
DIPOLE, MAGNETIC bring about dipole relaxation: jumps of weakly-
See Magnetic dipole. bound ions in the vicinity of a defect with oppo-
site charge; redistribution of the electron density
DIPOLE MOMENT in the vicinity of an anion vacancy (assuming de-
A characteristic of a system of generate electron states); jumps of protons in po-
electrically
charged particles is the vector p = ei r i , called lar crystals between local energy minima along
the dipole moment, where ei is charge of the ith hydrogen bonds, and other possible mechanisms.
particle, r i is its coordinate, and the sum is taken The relaxation time τ usually (for weak dipolar in-
over all the particles of the system. For a charge teractions) depends exponentially on the tempera-
of density ρ(r ) the dipole moment
distribution ture. At 300 K for various dielectrics τ = 10−4 to
is p = r ρ(r ) d3 r . If the total charge of the 10−10 s.
system is zero then its electric field at a distance
which considerably exceeds the dimensions of the DIRAC MONOPOLE (P. Dirac, 1931)
system is approximately determined by the dipole See Magnetic monopole.
moment. The value of the dipole moment p does
not depend on the choice of origin of the coordi- DIRECTOR
nates. The electric dipole moment per unit vol- A unit vector n, whose directions +n and −n
ume is called the electric polarization, with the are equivalent. The term is used for the description
symbol P (see Polarization of insulator). It may of the molecular ordering in liquid crystals, where
be formed by the spatial separation of opposite n points along the preferred direction of the axes
polarity mobile charges, as well by the presence of the molecules. For the description of the most
DISCLINATION MODELS OF PLASTIC DEFORMATION 313
typical liquid crystal, which is the single-axis ne- with an even number of electrons, and for transi-
matic liquid crystal type, it is sufficient to introduce tions between Kramers doublets where the num-
only one director, because in this medium only ber of electrons is odd we have a ∝ B 4 . Direct
one of the three molecular axes is oriented along a relaxation usually dominates at low temperatures
common direction, the other two being randomly (T < 10 K).
oriented. In the two-axis nematics it is necessary to
specify the triplet of directors n, l, and n × l be- DIRECT TRANSITIONS, crossover transitions,
cause the three molecular axes are orientationally vertical transitions
ordered. To describe C-type smectic liquid crys- Band-to-band electron (hole) transitions in
tals the notion of a c-director is introduced. Smec- crystals resulting from the absorption of light
tic C has a lamellar structure and the director n quanta, such that they occur without any appre-
does not point along the normal to the layers. The ciable change in the electron quasi-momentum (to
c-director is introduced to designate the direction the accuracy of the photon momentum, the latter
of the common slope of the molecules in the plane being negligibly small within the scale of the Bril-
of the layers. It is a regular vector because the louin zone). See Semiconductors.
states with +c and −c are non-equivalent.
Through the components of the director and its DISCLINATION MODELS OF PLASTIC
coordinate derivatives it is possible to express all DEFORMATION
the characteristic quantities of a liquid crystal: its They regard plastic deformation as a con-
dielectric constant tensor, its elastic energy, its mo- sequence of the motion of disclinations. Plas-
ment of forces, etc. tic deformation may be realized by three main
mechanisms: diffusion, shear (translational), and
DIRECT RELAXATION PROCESS rotational. Point defects, mainly vacancies (see
A relaxation transition between two levels with Diffusion mechanism of crystal creep) are the car-
a splitting of ω accompanied by either emis- riers of diffusion plasticity, dislocations are the
sion or absorption of a single resonance frequency carriers of shear plasticity, and disclinations are
phonon (a single-phonon process). For a direct re- the carriers of rotational plasticity. Disclination
laxation process occurring by way of transitions models of plastic deformation describe the genera-
between the spin sublevels the rate of spin–lattice tion and development of plastic rotations in solids.
relaxation is τ1−1 = a coth[ω/(2kB T )], where a A.F. Ioffe found at the beginning of the 1920s that
is a coefficient depending on the value of the during the course of plastic deformation mutual
matrix element of the spin–phonon Hamiltonian, rotations of crystal microvolumes take place. We
the resonance frequency (or the external mag- now know that these rotations not only accompany
netic field) and certain constants of the material deformation, but constitute an essential and often
(material density and the speed of sound). When determining factor. The rotation of a microvolume
ω/(2kB T ) 1 we have τ1−1 ∝ T , an important cannot be smooth because of its interaction with
feature of the direct relaxation process that may be surrounding volumes. Like shear strain it leads
tested experimentally. We have ω/(2kB T )
1 to the generation and motion of linear defects or
in the low-temperature limit, so that τ1−1 does not disclinations; see Fig. (b). Disclinations with rat-
depend on T , but rather it is controlled by the rate ings (Frank vectors) |ω| = 0.1 to 5◦ are observed
of spontaneous emission of phonons from the up- in solids. After the rotation of the adjacent micro-
per spin state. Another important aspect of the di- volumes to such angles their ideal conjugation is
rect relaxation process that makes it possible to impossible. A flat defect formed along the plane
clarify the mechanism of spin–lattice relaxation is of a turn together with a linear defect – the line
the magnetic field B dependence of τ1−1 . Consider of the rotation jump – is called a partial discli-
the case of the Kronig–Van Vleck mechanism (see nation. The stresses from a single disclination do
Van Vleck mechanism) where the electric crystal not attenuate, and so screened disclination struc-
e ld is modulated by the lattice vibrations. In the tures – disclination loops and disclination multi-
high-temperature limit we have a ∝ B 2 for ions poles – are usually encountered. A dipole is an
314 DISCLINATIONS
Fig. 1. Torsional disclination (a, b) and wedge disclination (c) along the axis of a continuous cylinder.
edges of a cut of an initially continuous cylin- elastic one, and its dependence on the distance is
der to some pregiven vector angle (Frank vector; similar to that of the field of stress of a dipole of
see Somigliana–Volterra dislocations) as shown in dislocations (see Elastic eld of dislocations). In
Fig. 1. Negative and positive disclinations are dis- crystals of macroscopic size R isolated rectilinear
tinguished, respectively, by the rotation of the disclinations are seldom found because their elas-
planes toward each other and in the opposite di- tic energy is proportional to R 3 . However, discli-
rection. If a Frank vector is parallel to the disclina- nations may be present in crystals in the form of
tion line, it is called a wedge disclination (disclina- small loops, dipoles, or networks, forming grain
tion of slope), while if it is perpendicular to it then boundaries, whose energy is proportional to r,
it is a torsional disclination. The space contain- where r is the radius of the loop or the distance
ing a disclination acquires the Riemann–Cristoffel between adjacent disclinations. The other possi-
curvature (see Curvature tensor), and space with bility for reducing the high disclination elastic en-
translational dislocations has a Cartan torsion. In ergy is by limiting the crystal size, e.g., using two-
crystalline media complete and partial disclina- dimensional crystals or liquid crystals.
tions are singled out depending on whether the
rotation angle satisfies the angular symmetry of
the medium or not. Partial disclinations are not
isolated defects, but serve as the edges of defect
planes. Partial disclinations violate the stacking
regularity of the sites of the crystal lattice in one
or several planes, adjacent to the disclination line.
A torn slope boundary is an example of a partial
disclination. In crystals with helical axes of sym-
metry complete disclinations are combined with
translational dislocations having a Burgers vec-
tor equal to the pitch of a screw. Such a defect
is called a dispyration. Any complete disclination
violates the natural crystal symmetry. Thus posi-
tive wedge 60-degree disclinations transform six-
fold rotation axes into pseudo-fivefold ones, and
the negative ones bring about a transformation
to pseudo-sevenfold axes. Disclinations move ei-
ther conservatively, or non-conservatively. In the
latter case they emit (absorb) lattice dislocations.
Disclinations appearing as linear singularities are
sources of stress. The stress field near rectilinear Fig. 2. Disclinations in nematics: (a, b) distribution of
disclinations are at the line of the disclination as director near disclinations with m = 1 and m = −1, re-
well as at infinity. The field of stress of disclina- spectively, (c, d) disclination with m = 2 which is unsta-
tional dipoles at large distances from them is an ble with respect to a transition to the uniform state.
316 DISCONTINUOUS PLASTIC DEFORMATION
In liquid crystals disclinations have the form croscope. They are present in biological objects,
of ruptures of vector fields, which describe the small crystals, and thin crystalline films. Disclina-
orientational ordering of molecules, as shown for tions are used to explain the structure of glasses
nematics (see Nematic liquid crystals) in Fig. 2. (see Vitreous state of matter), and sometimes also
These disclinations appear in the form of flex- for interpretating processes of plastic deformation
ible mobile filaments within the volume and at and fracture of materials.
the surface of practically any sample, a circum-
stance which led to the introduction of term “ne-
DISCONTINUOUS PLASTIC DEFORMATION
matic” itself. Due to the fact that nematics have
a second-order symmetry axis, the rotation angle Process of irregular, unstable plastic flow (see
of a disclination is restricted to the set mπ , where Plastic deformation) of crystalline materials. Un-
m is an integer. All disclinations with odd m are der conditions of active deformation (extension
topologically stable, so to bring about their transi- or contraction of samples with constant velocity)
tion to the uniform state requires an energy that is spiked or jumpwise strain appears on the deforma-
considerably in excess of their self-energy. Those tion diagram (stress versus strain plot, τ (ε)) as a
with even m are unstable and readily transform more or less irregular alternation of spontaneous
to the uniform state, as shown in Fig. 2. Discli- increases and decreases of deforming stress taking
nations are the dominating defects in liquid crys- place at a constant average deformation rate, so the
tals, where they are observable by an optical mi- diagram acquires a spiked nature (Fig. 1).
Fig. 1. Stress τ versus strain (deformation) ε plot showing regions of deformation of aluminum monocrystals with various
purity values 99.5% (1), 99.99% (2) and 99.9997% (3), illustrating spiked (jumpwise) stress (1, 2) and its disappearance
at high purity (3). The temperature of the experiment was 1.5 K.
DISCOTICS 317
DISLOCATION ANNIHILATION
See Annihilation of dislocations.
DISLOCATION CONTRAST
Local variation of the intensity on an electron-
microscope or X-ray topogram (see X-ray topog-
raphy), caused by the field of dislocation distor-
Fig. 2. Ordering of molecules in the orthogonal column- tions. The value of the local reciprocal lattice vec-
wise phase of a discotic. tor bh (r) = bh − ∇(bh u) (bh is a reciprocal lattice
vector of the ideal crystal, u is the atomic displace-
ment) changes due to the distortions, which leads
The columnwise phases of discotics, or the dis- to the variation of local conditions for Bragg re-
cotics themselves, are two-dimensional transla- flection (see Bragg law), and therefore to the vari-
tionally ordered lattices of long columns formed ation of the intensity of transmitted and diffracted
by the stacks of disc shaped molecules, as shown radiation. As a result, the intensity distribution has
in Fig. 2. Along the axes of the columns there is no an interference nature typical of dislocation con-
long-range order in the location of the mass cen- trast; from the number and configuration of the
ters of the molecules, and the system is liquid in light and dark interference fringes it is possible
one dimension and solid in two others (compare to determine the dislocation orientation within the
with smectic liquid crystals). This is because the crystal and the direction of Burgers vector. Images
interaction of the central parts of the molecules of rectilinear dislocations have the form of thin
(usually rigid) is considerably weaker than the in- fringes, whose width is inversely proportional to
teractions of the lateral parts which contain flex- the half-width of the Bragg reflection, of the order
ible hydrocarbon chains. If the molecular planes of an extinction length.
are perpendicular to the plane of the column the
discotic is called orthogonal discotic, and if they DISLOCATION DENSITY TENSOR
subtend some angle to this plane it is a slope Second-rank tensor which characterizes the
discotic. Depending on the two-dimensional lat- distribution of dislocation lines and their Burgers
tice type, hexagonal discotics and rectangular dis- vectors in a crystal. In the case of a continuous
cotics are distinguished. medium, it describes the continuum distribution of
DISLOCATION–DISLOCATION INTERACTION 319
dislocations of infinitely small Burgers vectors of of the deviator tensor (see Stress deviator); εilm
the form are the components of the so-called Levi–Civita
tensor among which only six differ from zero,
db = dU = β dr = α ds, namely ε123 = ε231 = ε312 = +1 and ε132 =
dc dc ε213 = ε321 = −1; τ l is the lth component of the
where dc is the infinitely small Burgers contour, unit vector of the normal to the dislocation line,
ds is the infinitely small surface which is bounded bk is the kth component of the Burgers vector of
by dc, β is the distortion tensor having components the dislocation.
βik = ∂Ui /∂xk , and the tensor = ∇ × β
α = rot β The force of interaction between two parallel
is referred to as the dislocation density tensor. By rectilinear edge dislocations with Burgers vectors
definition the divergence of a curl is zero, hence b1 , b2 is not central; and its projection on the di-
divα = ∇· α = 0, i.e. the dislocation lines cannot rection of the Burgers vector b1 is equal to
end inside the crystal. An isolated dislocation line μ(b1 , b2 ) cos ϕ cos 2ϕ
may be characterized in terms of δ-functions in the Fb = ,
2π(1 − ν)r
x and y coordinate directions, the dislocation den-
sity tensor being in this case ∇· α = τ · b δ(x)δ(y), where r, ϕ are the polar coordinates of the sec-
where τ is the unit vector tangent to the disloca- ond dislocation with respect to the first, in which
tion line, and b is the Burgers vector. The state the axis from which the angle ϕ is measured lies
of internal stresses is defined by the incompati- along the direction b1 . Fb is equal to 0 at ϕ = π/2
bility tensor η, which is expressed through α as (stable equilibrium for the case b1 b2 > 0 corre-
η = (1/2)(
α × ∇ × ∇ × α ), where
α is the trans- sponds to this) and ϕ = π/4 (this corresponds
posed dislocation density tensor. Thus, the specifi- to stable equilibrium for dislocations of differ-
cation of the dislocation density tensor completely ent signs, b1 b2 < 0). The field of characteristic
defines the field of internal stresses. The disloca- elastic stresses arising upon the bending of a dis-
tion density ρ is experimentally determined as the location line induces the appearance of the dis-
overall length of dislocation lines in a unit volume: location self-action force. The presence of bro-
ρ = L/V . The value of ρ is commonly expressed ken or uncompensated interatomic bonds means
in cm−2 . that close to dislocation lines, within a region
several angstroms in size, electrical space charge
DISLOCATION–DISLOCATION INTERACTION can accumulate. In metals this charge is strongly
The forces which affect a dislocation arising screened, and in insulators and semiconductors
from other dislocations. a stable redistribution of space charge can oc-
The dislocation–dislocation interaction can be cur. The electrostatic field associated with this
long-range (through the fields of elastic, elec- charge takes part in the electrostatic dislocation–
tric or magnetic stresses) or short-range (accom- dislocation interaction. In ionic crystals contain-
panied by reconstructions in the cores of dislo- ing dislocations, ions with charges of different
cations at their intersection). The strength of a signs contribute to this interaction. It is conve-
dislocation–dislocation interaction depends on the nient to determine the effective charges at dislo-
form of dislocation lines and their spatial distrib- cations by counting the dangling bonds. For ex-
ution. The elastic dislocation–dislocation interac- ample, elementary steps at a pure edge disloca-
tion involving distances between them which are tion have a charge of ±e/2. For a medium with
larger than the core radius r0 are determined by dielectric constant ε < 6.12 the electrostatic inter-
the field of elastic stresses created by these dis- action of steps is stronger than their elastic interac-
locations. The stresses are described by the elas- tion. In crystals with magnetostriction properties
tic stress tensor
σ , with components σij which (see Magnetostriction) a magnetic dislocation–
determine the force per unit length F acting on dislocation interaction is possible. At high tem-
the dislocation. The components of the vector F peratures the diffusion dislocation–dislocation in-
b , where
are written in the form Fi = εilm τl σmk teraction can arise through the flux of vacancies
k
σmk = σmk − (1/3)δmk σll are the components or interstitial atoms between the dislocations.
320 DISLOCATION ELASTIC FIELD
DISLOCATION ELASTIC FIELD a lower rate, and the temperature range where
See Elastic eld of dislocation. the thermoelastic stresses induce the nucleation
and the multiplication of the dislocations is much
DISLOCATION EXCITON wider. Nevertheless, one can obtain dislocation-
An exciton localized near a dislocation line free crystals of Cu up to 6 mm in diameter and
by the elastic eld of dislocations, which creates 60 mm in length by the Czochralski method, pro-
the potential well for the exciton. The electronic vided the temperature regime is maintained with
excitations (excitons) in such states may travel high precision, a thin (1 mm) neck is created,
along the dislocation line, undergoing small trans- and a heater of special design and a system of heat
verse motions in the plane perpendicular to it. screens are used. These are to be combined with
The quantum state of a rectilinear dislocation is other technological procedures to provide for the
characterized by a wave vector directed along the necessary conditions for break-off of the crystal
dislocation line, and by a discrete set of quantum from the melt and for its subsequent cooling.
numbers, related to the transverse motion. In semi- So-called thread-like crystals (cat’s whiskers,
conductors there are excitons of several varieties, fibers) up to 20 μm in diameter, with a ratio of
which differ by the type of bonding of the elec- length to diameter not less than 103 , have also
tron and hole forming the exciton with the disloca- been obtained free of dislocations (see Quasi-
tion core: excitons localized within the dislocation one-dimensional crystals). Such crystals possess a
core, or far from it in the field of the strain po- number of interesting physical properties, in par-
tential; excitons of a mixed type with one charge ticular their strength being close to the theoretical
within the dislocation core and the other in the value.
deformation potential field. Kinks on dislocations
(dislocation steps) on the dislocation line create DISLOCATION GRIDS
traps for the dislocation exciton and disrupt the ex- A system of mutually intersecting dislocation
tension of the longitudinal motion. The exciton in- lines. At the point of intersection, a reorganiza-
troduces specific features into optical and kinetic tion can be observed resulting in the appearance of
properties of semiconductors. sites whose equilibrium position can
be approxi-
mately determined by the relation ni=1 τi bi2 = 0.
DISLOCATION-FREE CRYSTALS Here τi is the unit vector tangent to the ith dislo-
Highly perfect monocrystalline samples that cation line, entering the site; bi is the magnitude
contain no dislocations. Dislocation-free crystals of its Burgers vector (summation carried out over
can be obtained by eliminating the causes that all dislocations passing through the given site). If
induce mechanisms of nucleation and multiplica- the attractive (or repulsive) forces act between the
tion of dislocations in the growing and subse- different dislocations near the site, the dislocation
quent cooling of a monocrystal. When producing reactions may result in the appearance of grids
dislocation-free crystals, the important factors are of dislocations with hexagonal cells. Within the
the material purity, the crystal thickness, and the boundaries of purely planar torsion (110) in body-
degree of oxidation of its surface; the most sig- √ cubic crystals, hexagonal grids with τ =
centered
nificant feature is the thermal stress induced due (1/ 6)112 and b = (1/2)111 appear. In face-
to a nonuniform temperature distribution in the centered cubic crystals two types of lattice points
crystal. When priming is used the subboundaries are distinguished. The lattice points are called
and dislocations could grow from the priming. The K-sites, if they are formed by three dislocations
most advanced results are achieved in obtaining so that one character A, B or C in notation based
the dislocation-free crystals of Si and Ge. With the on the Thompson tetrahedron is located in the sec-
help of the Czochralski method (see Monocrystal tor. If a pair of characters A, B or C separated
growth), the dislocation-free crystals of Si and Ge by an extension of opposite dislocations is located
80 mm and 35 mm in diameter, respectively, were in each sector, then the lattice points are called
obtained. The growth of perfect metallic crystals P-sites. The symmetric K-sites, denoted by Ks ,
exerts severe requirements on the heating condi- are characterized by a clockwise sequence of char-
tions, since in metals the plasticity decreases at acters A, B, C, and the asymmetric Ka sites have
DISLOCATION MODEL OF PLASTIC DEFORMATION 321
a counterclockwise sequence. The lattice points may form dislocation tangles, mixed up disloca-
with a counterclockwise sequence A, B, C, A, B, tion lines. Burgers vectors of dislocations in the
C are called Ps , and the ones with a clockwise tangle are usually of the same sign, and thus the
sequence are called Pa . If only subtraction stack- long-range elastic stress far from the aggregation
ing faults are possible, then K-sites are collapsed may be described as the field of a superdislocation
and P -sites are split. The K-sites can split if it is with the total Burgers vector of the tangle. How-
possible to form defects of implantation stacking ever, at the head of the relaxed aggregation the
faults. In annealed crystals (see Annealing) stable elastic stress is lower due to its “bluntness”. The
three-dimensional Frank networks of dislocations interlacements may consist of a system of disloca-
are possible. tion multipoles (dipoles, quadrupoles, octupoles,
etc.), formed from equal numbers of dislocations
DISLOCATION INTERACTIONS with mutually opposite Burgers vectors. There are
See Microcrack–dislocation interaction, Point the following special types of dislocation inter-
defect–dislocation interaction, Vacancy–dislocation lacements:
interaction.
(1) bulk dislocation grids;
DISLOCATION INTERLACEMENTS (2) walls of dislocation loops in a layer much thin-
A relatively isolated group of dislocations, with ner than the other two dimensions (all the
lines forming a more or less complex interlac- loops need not belong to one system);
ing. The following types of dislocation interlace- (3) dipole dislocation walls.
ments can be singled out: those containing an
excess of dislocations of the same sign, which DISLOCATION KINK
involve fields of elastic stresses decreasing with See Kinks on dislocations.
the distance r as r −1 (see Elastic eld of dislo-
cations), and those with their total Burgers vec-
DISLOCATION LOOP
tor equal to zero. The latter elastic stresses de-
crease with distance at least as fast as r −2 . The A dislocation whose line forms a closed curve.
boundaries of disoriented block structures may If all the points of this line are confined to a plane
consist of closed dislocation entanglements of the then the dislocation loop is called planar. We can
same sign, located in a relatively broad band. distinguish a planar slipping dislocation loop with
Local non-uniformities of the dislocation density its Burgers vector b lying in the same plane, and a
(dislocation interlacements, tangles and bundles) planar prismatic dislocation loop with its Burgers
form during early stages of plastic deformation at vector perpendicular to this plane.
the second stage of the evolution of dislocation At large distances r
r L from a dislocation
structures, with a dislocation density of 108 to loop of size r L the displacement vector ui and
1011 cm−2 . the stress tensor σik decrease with distance in ac-
A special type of dislocation interlacement cordance with the laws |ui | ∼ b/r and |σik | ∼
consisting of a relatively isolated series of paral- Gb/r 2 , where G is the shear modulus, and b
lel dislocations belonging to one system of dis- is the magnitude of the Burgers vector. The en-
location slip and located in adjacent slip planes, ergy W of the dislocation loop is of the order
constitutes a pile-up of dislocations, which under W ∼ Gb2 r L ln(r L /r C ), where r C is the radius of
the effect of an applied stress become grouped the dislocation core.
in front of some obstacle. Such aggregations are
concentrators of stresses (see Stress concentra- DISLOCATION MODEL OF PLASTIC
tion), creating elastic stresses which are propor- DEFORMATION
tional to the applied stress, and to the number of Explains plastic deformation caused by the mo-
dislocations near the obstacle. As a result of plastic tion of dislocations within a crystal, including
creep (transverse slip and climb of dislocations), at conservative motion (slipping) in the slip plane,
the head of an aggregation near an obstacle there and non-conservative motion of edge dislocations
322 DISLOCATION PILE-UP
along the normal to the slip plane (climb). A dis- of double bends. Quantum-mechanical effects ap-
location can be considered as the boundary be- pear necessary for overcoming the Peierls barriers
tween regions shifted during the course of defor- by dislocations at temperatures close to 0 K (see
mation, and the unshifted (undisplaced) regions of Low-temperature yield limit anomalies). Screw sec-
the crystal, so the motion of dislocations in the slip tions of dislocations can easily change their slip
plane brings about plastic deformation by shear- plane, which causes the transfer of dislocations to
ing. Upon the application of an external stress τ other planes via double transverse slipping. The
due to the presence of an elastic eld of disloca- multiplication of dislocations during the course of
tion, there arises a force which affects each seg- plastic deformation is performed via the action of
ment of dislocation length: F = τ b, where b is Frank–Read sources, multiple slipping, and other
the Burgers vector. During the dislocation mo- mechanisms. The climb of dislocations occurs as
tion in the slip plane interatomic bonds are bro- a result of the diffusion of vacancies or intersti-
ken and rebound only near the dislocation core, tial atoms to the extra semiplane at the dislocation
but not along the entire slip plane (Fig.). The shear core (or due to their outflow), and this becomes
in the slip plane occurs sequentially in propor- significant at high temperatures and long loading
tion to the dislocation motion. During the passage hold times (see Diffusion). The processes of crys-
of the dislocation through the crystal the magni- tal hardening by doping of the solid solution, by
tude of the displacement is equal to the Burgers the introduction of dispersed particles of a second
phase (see Precipitation-hardened materials), and
vector b. The atoms near the dislocation core are
also by strain hardening, are explained by the dis-
almost symmetrically displaced to opposite sides
location model of plastic deformation. This model
from the extra semiplane, which involves equal
has been experimentally confirmed by transmis-
and oppositely directed forces for the shear. For
sion electron microscopy, by X-ray topography, by
some crystals this produces the minor stress (sev-
the method of etching of pits, and others. See also
eral orders below the theoretical strength) neces-
Disclination models of plastic deformation.
sary for moving dislocations in a perfect crystal
lattice at 0 K, the so-called Peierls–Nabarro stress DISLOCATION PILE-UP
τp . This stress is τp ≈ G exp[−2πw/b], where See Pile-up of dislocations.
G is the shear modulus and w is the dislocation
width (see Peierls–Nabarro model). Upon the slip- DISLOCATIONS (fr. Lat. dislocatio,
ping of dislocations, rotation (twisting) and bend- displacement, movement)
ing of interatomic bonds take place, so the Peierls– Defects of a crystal lattice which distort the
Nabarro stress is increased in covalent crystals correct arrangement of atomic planes. Disloca-
with directed interatomic bonds (small w), where tions differ from other defects in crystals because
dislocations may be almost immobile at room tem- they involve a considerable disruption of the reg-
perature (see Plastic deformation of covalent crys- ular interlacing of atoms concentrated within a
tals). At temperatures above 0 K the motion of close vicinity of some line passing through the
dislocations may be accelerated through the ther- crystal, and this line is the edge of an incomplete
mally activated overcoming of Peierls barriers (see displacement of one part of the crystal with respect
Peierls relief), with the accumulation of disloca- to another part through one of the lattice transla-
tions lying at the energy minima (Peierls valleys) tion vectors.
DISLOCATIONS 323
Fig. 3. Illustration of a plastic deformation that results from sequential dislocation slip.
in appreciably distorted (cold-rolled) metals. If the stresses at the onset of slipping determine the
the dislocation core is enclosed by some tube, microscopic elastic limit σs of a monocrystal. The
then outside this tube the crystal may be con- value of σs is 102 to 104 times smaller than the
sidered ideal, subjected only to elastic deforma- shear modulus of a monocrystal. This small value
tion. Therefore, distortions far away may be an- explains why the shear strength of real crystals
alyzed by the methods of elasticity theory. Dislo- with dislocations is several orders of magnitude
cations generate elastic deformations and stresses smaller than that of monocrystals without disloca-
in their neighborhood so they are sources of elas- tions. Some dislocation motions (climb) displace
tic elds of dislocations within the crystal. These an edge dislocation to a direction perpendicular
fields are reminiscent of the magnetic field lines to the slip plane, and the growth or dissolution of
encircling an electric current. The stress field de- atomic rows at the edge of an “extra” semiplane
termines the inherent elastic dislocation energy is accompanied by the formation or disappear-
proportional to b2 with a value of about 10−4 erg ance of vacancies in the vicinity of the dislocation
per cm of its length, and it also involves inter- core. This climb provides the mechanism for dis-
actions between dislocations. A dislocation as a locational diffusion (see Diffusion mechanism of
source of an elastic field is subjected to forces aris- crystal creep, Dislocation model of plastic defor-
ing from the shear stresses within the crystal, re- mation).
minding one of the Lorentz force of a magnetic The development of dislocational plasticity is
field on a conductor carrying electric current. The determined by the mobility of dislocations, and
value of the force f = bσ applied to a unit length the intensity of their generation. The mobility of
of a dislocation involves a component of the strain isolated dislocations in a very pure monocrystal
tensor σik . For example, for an edge dislocation depends on the interatomic bonding forces, and
with its axis lying parallel to the z axis and the vec- on the interaction with phonons and electrons (in
tor b directed along the x axis the force component metals). The mobility of dislocations in non-ideal
is fx = bσxy . The external force fairly easily sets crystals is reduced by their interaction with each
dislocations into motion. A moving dislocation is other, and with other defects which can retard or
capable of emitting elastic (sonic) waves, which stop them (dislocation barriers). Impurities can
explains the emission of sound during plastic de- block dislocation motion and pin a dislocation line
formation (see Acoustic radiation by dislocations). at some points. At high temperatures the barriers
Shear forces can move dislocations and thereby can be overcome by thermal activation, while at
cause plastic deformation. Plastic deformation low temperatures quantum tunneling is possible
brought about by a sequence of individual disloca- (see Tunnel effect).
tion slips is sketched in Fig. 3. Plasticity is associ- Dislocations mainly affect the mechanical
ated with different types of this motion, as well as properties of solids (elasticity, plasticity, strength,
with the motion of pile-up of dislocations. Slipping internal friction), for which their presence is often
occurs at comparably low external loadings, and the determining factor. The elastic fields induce an
DISLOCATION SLIP 325
increased concentration of impurities near the axis to the dislocation line. In the case of linear disloca-
and the formation of a Cottrell atmosphere, which tions lying in parallel slip planes and having paral-
may bring about the coagulation of impurities at lel lines, the dislocation sign is associated with the
the dislocations (in transparent crystals it “deco- direction of the Burgers vector. To determine the
rates” dislocations). Dislocations change the op- direction of b a rule is used which, not withstand-
tical properties of the crystals, which is the basis ing a certain arbitrariness, gives an unambiguous
of the method of observing isolated dislocations in result in successive applications. Initially a posi-
transparent materials. At points where dislocation tive direction is chosen (arbitrarily) along the dis-
lines emerge at the external surface of the body the location line; i.e. the direction of a unit vector τ
chemical strength of the crystal is weakened, and tangent to this line is assigned at a point, and then
special reagents can destroy the neighborhood of a closed Burgers contour in a crystal with dislo-
a dislocation axis, forming visible pits. This is the cations is drawn so that, when sighting along the
basis of the selective etching method which is the positive τ direction, the tracing of the boundary
the main method for observing individual disloca- is in the clockwise sense. The resulting Burgers
tions in bulk samples of non-transparent materials. vector b connects the end with the beginning of
A system of dangling bonds at the dislocation core the gap in the Burgers contour traced in the initial
distinguishes the dislocation line from electrical, ideal lattice. This method of defining b, suggested
magnetic, and optical points of view. Dislocations by Frank and Bilby, is called the rule of “end to be-
may carry or capture electrical charge, and possess ginning along$ a right-hand screw”. As a result, we
a magnetization that differs from that of the bulk have b = c (∂u/∂l) dl, where the integral along
crystal. They can also increase the electrical resis- the closed contour c that encloses the dislocation
tance of conductors and change the concentration line is taken in the direction determined by a right-
of free electrons in semiconductors. They play an hand screw relative to τ . If the positive direction
important role in magnetic crystals by determining along this line is changed, then the Burgers vec-
different relaxation phenomena. While moving, a tor reverses its sign. Sometimes a definition of b
is used whose sign is changed to the opposite, so
dislocation can emit or adsorb vacancies, thereby
it is necessary to check the rule for assigning the
changing their total number within the crystal. The
sign to b. It is important that the Burgers vector
dynamic formation of charged vacancies in ionic
remain constant along the dislocation. In the sim-
crystals and semiconductors may be accompanied
plest case of a linear edge dislocation, a positive
by a glow (luminescence) of the crystal. The rate
(+) sign is assigned to the dislocation, for which
of diffusion of point defects along the dislocation
the extra half-plane is situated in the upper part of
axis is, as a rule, greater than the rate of their dif-
the crystal, and the negative (−) sign corresponds
fusion through the volume of the regular crystal.
to the extra half-plane located below. “Upper” and
The coefficient of linear diffusion along a dislo-
“lower” are relative concepts that are determined
cation can exceed the coefficient of bulk diffusion
by the crystal orientation relative to an observer.
by several orders of magnitude. Therefore, dislo- The sign of elements of a closed dislocation loop,
cations play the role of “drain tubes” along which having a unique Burgers vector, is determined by
point defects can easily travel relatively long dis- the direction of the vector tangent to a given ele-
tances within the crystal. ment.
See also Edge dislocation, Epitaxial disloca-
tions, Frank dislocation, Helical dislocation, Partial DISLOCATION SLIP, dislocation glide
dislocation, Perfect dislocation, Prismatic disloca- Dislocation motion along the plane, determined
tions, Screw dislocation, Somigliana–Volterra dis- by a line of dislocations and its Burgers vector b
locations, Surface dislocations. (slip surface). The slip of dislocations is accompa-
nied by the shear of one part of the crystal relative
DISLOCATION SIGN to another through b, i.e. the slip of dislocations
Characteristic (sign) attributed to elementary causes the shear strain of the crystal. All possible
(linear) dislocations, determined by the orientation slip planes and their directions form the slip sys-
of the Burgers vector b, and of the vector tangent tem. The presence of pencil slip causes twisting
326 DISLOCATION STEP
slip lines to appear on surfaces of deformed body- the dislocation slip as a whole can occur only dur-
centered cubic metals, in contrast to the long lin- ing non-conservative motion (climb) of the step,
ear slip lines in face-centered cubic and hexagonal accompanied by the appearance within the crystal
close-packed crystals. Screw dislocation compo- of chains of vacancies or interstitial atoms. Such a
nents (see Screw dislocations) in inhomogeneous step motion may only occur when thermal activa-
internal stress fields can exhibit double cross slip. tion strongly contributes to the motion of the dis-
locations with steps. Due to the diffusion flow (see
DISLOCATION STEP Diffusion) of vacancies (or of interstitial atoms) the
A dislocation segment connecting two sections step may move along the dislocation line, i.e. the
of an edge dislocation lying in different parallel dislocation step is a source (or sink) for point de-
slip planes. If the distance between the slip planes fects of the crystal lattice. When the Burgers vec-
equals one lattice parameter the dislocation step tor of the dislocation being intersected is coplanar
is called a unit step, and for larger separations it with the slip plane of the initial dislocation then
is called a superstep. The step at a screw disloca- the step located at it will have the same slip plane,
tion sector coincides with the kink on dislocation and such a dislocation with its step may move
because in this case the plane where the disloca- conservatively. At the intersection of two disloca-
tion is located, is its slip plane. It is possible to as- tions, each of which is split into partial disloca-
cribe a sign (positive or negative) to a dislocation tions, there appears a constriction at the stacking
step to distinguish steps which transfer the disloca- fault connecting the partial dislocations, because
tion line from the given slip plane to the plane lo- the formation of the step brings about the merg-
cated above, and to that below. Between two steps ing of the partial dislocations accompanied by the
at the dislocation line there is a Coulomb type annihilation of the stacking fault at the step. The
interaction whereby steps of the same sign repel intersection of split dislocations involves a stress
and those of opposite sign attract. In the elastic- that is considerably higher than that at the inter-
isotropic continuum approximation (see Isotropy section of unsplit dislocations. In an ionic crys-
of elasticity) the modulus of the interaction energy tal, at a dislocation with a Burgers vector of the
of two steps, spaced a distance L apart, is given by 011 type and a line along 100, at a dislocation
the expression step with a bend height of one interatomic distance
Gb2 a 2 there appears an effective (either positive, or neg-
W =A , ative) charge, equal to the half of the charge of the
8πL
boundary ions. A neutral step appears at such a
where b is the magnitude of Burgers vector, a is
dislocation, if the step length is equal to two inter-
the size of the dislocation step in the direction per-
planar distances. Dislocation steps play a very im-
pendicular to the slip planes, G is the shear mod-
portant role in the kinetics of crystal strains since
ulus, and A is a constant of the order of unity, the
their appearance and motion to a large extent de-
exact value of which depends on the type of dis-
termine the aggregation of point defects (vacan-
location. There are several mechanisms for form-
cies and interstitial atoms) during the process of
ing a dislocation step. They may occur at the in-
plastic deformation.
tersection of two dislocations moving in intersect-
ing slip planes, so the steps on one dislocation are
formed parallel to the Burgers vector of the other DISLOCATION STRING
one. The magnitude and direction of the Burgers Mechanical model which treats a dislocation in
vector of the dislocation step are the same as those terms of a flexible elastic string with a linear mass
of dislocation sections located above (or below) density M and intrinsic linear energy density (lin-
the slip plane. A pair of steps of opposite signs ear stretching) C. This model is used for the de-
appears in the so-called cross slip of edge dislo- scription of the conservative movement (slip) of
cations. Since in the general case the slip plane weakly bent dislocations. If the linear deviation of
of the dislocation step does not coincide with the the dislocation line u (x, t) from some static con-
slip plane of the main part of the dislocation line, figuration u0 (x) projected along the x axis is cho-
DISLOCATION WALLS 327
Static u0 (x) and instantaneous u0 (x) + u(x, t) configuration of a weakly bent dislocation line.
sen as the dynamic variable (see figure), then the DISLOCATION SUPERCONDUCTIVITY
equation for the dislocation motion in this string Non-uniform superconducting state which oc-
approximation has the form curs near a dislocation network at a temperature
and magnetic field exceeding the corresponding
∂2u ∂2u ∂u critical values for the same material without dis-
M − C +B = F (u, x, t). locations. The supercurrent that flows within such
∂t 2 ∂x 2 ∂t
a structure is anisotropic with respect to the axes
The first term in the left represents the force of of the dislocations. This superconductivity, which
inertia, the second represents the stretching force, occurs in the vicinity of individual dislocations,
and the third is the force arising from viscous fric- and also of their aggregates, has a localized nature.
tion. The total force per unit length F (u, x, t) on The critical magnetic eld in this case is strongly
the right-hand side includes external mechanical anisotropic with respect to the dislocation axes.
stresses, internal stresses, and the Peierls poten-
Dislocations in superconductors very strongly
affect the value of the critical current, which due to
tial (see Peierls relief). The parameters M and C
the vortex pinning at the dislocations considerably
provide the approximate effective mass and self-
exceeds its value in “pure” superconductors.
energy of the dislocation, with values determined
The critical magnetic field of superconductors
by the Burgers vector of the dislocation b, and by
with dislocations is significantly increased by the
the material constants of the crystal; they can be
variation of the properties of the material, as well
estimated from the expressions M ∝ ρb2 and C ∝
as due to the variation of the mean free path of the
Gb2 , where ρ is the density of the crystal, and G is electrons, which becomes anisotropic. The pres-
the shear modulus along the direction of slipping. ence of a gap in the electron spectrum changes the
The coefficient B of the equation characterizes the force resisting dislocation motion, and helps to in-
viscous retardation of the dislocation by elemen- crease the plasticity of the material in the super-
tary excitations of the crystal (by phonons, con- conducting state.
duction electrons, etc.) or by mobile point defects.
The string equation replaces the more complex DISLOCATION TENSION, LINEAR
rigorous equation of dislocation motion where the See Linear dislocation tension.
inertial force and the self-action force have a non-
local nature. The condition for the applicability DISLOCATION TRANSITION RADIATION
of the string model is limited by the inequalities See Transition radiation by dislocations.
∂u/∂x 1 and ∂u/∂t c, where c is the speed
of sound. The string model is widely used for the DISLOCATION WALLS, dislocation boundaries
theoretical analysis of the effect of dislocations on One-dimensional ensembles of straight, paral-
the speci c heat, thermal conductivity, and electri- lel, regularly arranged dislocations of the same
cal conductivity, the acoustical properties of crys- type with parallel slip planes. Regular equilib-
tals, and also for investigating the processes of rium dislocation walls of edge dislocations do not
dislocation motion through Peierls barriers, and create long-range stress fields since the latter ex-
through point defects of the crystal structure. ponentially attenuate at distances of the order of
328 DISORDERED SOLIDS
the average spacing between dislocations within in solids. See also Long-range and short-range or-
the wall. Non-equilibrium dislocation walls, how- der, Phase transitions.
ever, may serve as a source of considerable stress.
A wall formed by parallel screw dislocations cre- DISORDERED SOLIDS, SPECTRA
ates stresses that extend over distances comparable See Elementary excitation spectra of disor-
with the characteristic sizes of dislocation walls. If dered solids.
the wall is torn it becomes the source of a stress
field that is similar to the stress field from a discli- DISORDER, ELECTRON-INDUCED
nation, corresponding to the disorientation of a See Electron-induced disordering.
dislocation wall. These walls play an important
role in the strain hardening of monocrystals be- DISORDERING OF ALLOYS
cause they are sources of both dislocations and the See Alloy disordering.
barriers to them. Dislocation walls have photore-
sistive (see Photoconductivity) and semiconduct- DISPERSE STRUCTURE
ing properties in semiconductor crystals. The structure of liquid or solid systems con-
sisting of dispersed phase particles distributed in
DISORDERED SOLIDS various ways within the volume of a matrix phase.
A generic name for solids lacking any type of The inclusions may be solid, liquid, gaseous. They
either short-range or long-range order. Routinely, are introduced into the matrix as a result of phase
one has to specify which particular type of disor- transitions (alloy aging, martensitic transforma-
der is implied. The destruction of long-range order tion, eutectoid alloy decomposition (see Eutectoid
in the positions of atoms in a solid, regardless of alloys), eutectic crystallization, devitrification of
the types of atoms in particular positions, is called amorphous metals and metallic alloys, etc.), by in-
a topological disorder. When the term “disordered ternal oxidation, by introduction of the high-melt
solid” is used without specifying the type of disor- particles into the melt (composite materials), and
der, a topological disorder is usually implied. Dis- as a result of the action of nuclear radiation or ion
ordered solids featuring this disorder are classified implantation. The properties of the inclusions de-
as amorphous solids (see Amorphous state). The pend on the degree of dispersity of particles, on
breakdown of order in the relative positioning of the energy of their interaction with the medium,
atoms of differing types is called compositional on shape, state of matter and structure (e.g., pres-
disorder. Topologically ordered binary alloys may ence of atomic ordering of the inclusion lattice).
exhibit this type of disorder, with each type of Depending on their size, the inclusions are clas-
atom randomly occupying the sites of a regular sified as rough (low) dispersed (>0.1 μm) and
crystal lattice. The breakdown of the relative or- fine (highly) dispersed (0.001 to 0.1 μm). Their
dering of magnetic moments of different atoms, or shape may be threadlike (needle-like), plate-like,
atomic complexes that display certain short-range or spherical. Disperse particles have excess sur-
order, represents a magnetic disorder, while that face energy, which implies thermodynamic insta-
of the relative orientation of molecules or atomic bility and a disposition toward coalescence (co-
complexes is an orientational disorder. An exam- agulation). Inclusions formed as a result of phase
ple of solids featuring orientational order while transformations have a shape and orientation with
having no topological order or translational in- respect to the crystal lattice or matrix that is con-
variance of their structure is represented by quasi- ditioned by the ratio of the surface energy to the
crystals. The breaking of topological, composi- energy of the internal stresses, which arise dur-
tional, magnetic, orientational, and other types of ing the formation of a disperse structure. The ma-
symmetry or order may arise from thermal u c- trix containing the disperse structure is character-
tuations or by external actions, such as electro- ized by increased strength and high-temperature
magnetic fields, mechanical strains, etc. The emer- strength (see Alloy aging, Precipitation-hardened
gence of order, its breakdown and change in type, materials). The distribution of particles over the
all correspond to order–disorder phase transitions volume of the matrix may be classified as regular
DISPERSION GRINDING 329
or irregular. A regular, quasi-periodic distribution cluding sonic) and light waves in dispersive me-
of particles is called a modulated structure. The dia depend on ω and k (e.g., susceptibility, elastic
disperse particles create a macrolattice, with a unit constants, etc). Spatial dispersion is caused by the
cell which may be cubic, tetragonal etc., depend- fact that the response of the dispersive medium to
ing on the symmetry of the elastic fields around an external disturbance at an arbitrary point r de-
the particles (see Crystal symmetry) within the ma- pends on the value of the disturbing field not only
trix volume. For a low bond energy between the at this point, but also at surrounding points (non-
particles (less than their thermal energy, in order local response). Temporal dispersion is caused by
of magnitude), the so-called thixotropic disperse the time lag of the response, i.e. by its dependence
structures form, which have low thermodynamic on the value of the disturbing field not only at the
stability, but recover in time (e.g., disperse struc- moment of observation, but also at earlier times.
tures formed in aqueous solutions of aluminum In addition, the dependence of the frequency ω
or iron hydroxide). (Thixotropic gels liquify when or the energy E of a quasi-particle in a solid on
shaken and solidify when left standing.) Non- its quasi-momentum k is called dispersion, or
thixotropic disperse structures are more stable, but more accurately, a dispersion law. The dispersion
can be destroyed irreversibly. These are related to law E(k) determines the structure of the energy
disperse structures of ceramics, binding materials bands of electrons, holes, excitons, and the spec-
(cements) and of the dispersion-hardening alloys. trum of phonons, plasmons, magnons, and other
quasi-particles.
DISPERSION (fr. Lat. dispersus, scattering,
dispersion) DISPERSION, ANOMALOUS
In different areas of the natural sciences dis- See Anomalous dispersion of light.
persion determines the characteristics of the scat-
ter of values of a physical parameter. Its meaning DISPERSION DIELECTRIC CONSTANT
varies somewhat with the parameter under con- ANALYSIS
sideration, and some examples will be given. In One of the methods for determining the fre-
mathematical statistics the dispersion σ of a ran- quency dependence of the dielectric constant of
dom variable parameter x determines the inten- solids from experimental data about the response
sity of uctuations: σ 2 = x − x2 , where the an- of the body to an applied alternating electric field.
gular brackets . . . denote an average value. The The simplest model to explain the frequency de-
wave dispersion relation is the dependence of the pendence of the reflectivity coefficient is that of a
frequency ω(k) on the wavenumber k which pro- solid consisting of a set of the non-interacting dis-
vides the magnitude k = 2π/λ and the direction persion oscillators. By variation of oscillator para-
of motion of the wave, where λ is the wavelength. meters (frequencies, attenuation coefficients, and
The phase velocity v ph = ω/k and the group ve- also their number), it is possible to attain good
locity v g = dω/dk depend on the dispersion law agreement between model calculations and exper-
(see Sound velocity dispersion, Dispersion of light). iment.
In the presence of dispersion the propagation of an
anharmonic wave is accompanied by a variation DISPERSION GRINDING
of its shape. A dispersive medium is one whose Fine grinding of a solid to form disperse struc-
dielectric constant ε = ε + iε , magnetic perme- tures. The specific amount of work spent for the
ability μ = μ + iμ (or susceptibility χ = χ + dispersion depends on the cohesion parameters
iχ ) depend on the frequency (wavelength). When (see Cohesion), on the structural characteristics of
these quantities are complex, the real (primed) the pulverized body, on the surface energy, and on
part is called dispersion and the imaginary (dou- the extent of the grinding. Introduction of surface-
ble primed) part is called absorption. The real part active agents into the system reduces the energy
determines the frequency and the imaginary part expenditure during the course of dispersion, and
determines the losses in a resonator. The singular- increases the dispersity of the ground phase. De-
ities of propagation and absorption of elastic (in- pending on the composition and properties of the
330 DISPERSION HARDENING
DISPERSION HARDENING
Improving the strength (hardness) of metallic
oversaturated solid solutions by their decomposi-
tion (see Alloy decomposition), and by the sep-
aration of the dispersed phase enriched by the
alloying element (see Alloying). The degree of
dispersion hardening depends on the bulk share
(quantity) of the separation phase, on the size
and shape of its particles, on the character of the
interface between the phases (coherent or non-
coherent), on the value of coherent stresses, on
the internal structure (atomically ordered or dis-
ordered), on the value of the elastic moduli of the
matrix, and on the extent of the separation. Sev- Fig. 1. Dispersion laser resonators.
eral theories describe the hardening of alloys and
the yield limit at different stages of decomposition
(see Alloy aging). Dispersion hardening at the early laser resonator is carried out by moving the curve
stages is well described by the Mott–Nabarro the- of losses γ (ν) along the amplification curve α(ν)
ory, which explains it in terms of the interaction of the active laser element (Fig. 2), for δνγ < δνα ,
of internal coherent stresses around the particles where δνγ and δνα are the half-widths of the γ (ν)
and moving dislocations. The hardening at later and α(ν) curves, respectively. For δνγ δνα the
stages of decomposition is satisfactorily described generation may be of a single frequency. The ad-
by the Orowan theory of precipitation strengthen- justment of the resonator during the power gener-
ing (E. Orowan, 1948), where the yield limit of the ation provides for the control the spectral-kinetic
aged alloy depends only on the distance between parameters of the radiation (see Tunable lasers).
its non-coherent particles. Using the principle of a dispersion laser resonator,
refractive lasers of the solid and liquid active me-
DISPERSION LASER RESONATOR dia types (crystals with color centers, organic com-
An optical resonator with losses or Q-values
which depend on the frequency ν. This depen-
dence is achieved by the introduction of disper-
sion elements into the resonator (Fig. 1): refract-
ing prism (a), holographic diffraction grating (b),
Fabry–Perot interferometer (c), interference-polar-
ization filters (d). The selection of modes is deter-
mined by the propagation of the radiation beams at
different angles relative to the axis of the resonator
(see Fig. 1(a, b)), or by the spectral dependence of
the transparency of the dispersion elements (see
Fig. 1(c, d)). The adjustment of the radiation spec-
trum ν L of the laser with the help of a dispersion Fig. 2. Amplification curve of the laser active element.
DISPERSION OF LIGHT 331
DISPERSION LAW
Relationship either between energy ε and mo-
mentum p (quasi-momentum in the periodic crys-
tal lattice potential) of elementary excitations
(quasi-particles), or between the frequency ω and
the wave vector k of waves (sonic, electromag-
netic, spin, etc.) propagating in many-particle sys- Dispersion of light.
tems (e.g., solids). The dependence of ε(p) on p,
or ω(k) on k, can be rather involved, and some-
times it is multivalued (see Semiconductors). ε0 is the dielectric constant of vacuum. If several
electrons, bound by different elastic forces, take
DISPERSION LAW, KANE part in the process, the right-hand sides of Eqs. (1)
See Kane’s dispersion law. and (2) are replaced by summations.
Typical dispersion n and adsorption κ curves
DISPERSION OF LIGHT for an individual oscillator are shown in the figure.
Decomposition of a light beam into its com- In the region between the peaks n decreases with
ponent frequencies during the refraction of light, increasing ω (region of anomalous dispersion of
interference of light or diffraction of light. It re- light), and outside this region n grows with the fre-
sults from the dependence of the refractive index quency. The absorption is strong in the anomalous
of a material on the wavelength (frequency) of dispersion region near ω0 . The frequency range
the light. The dispersion of light was discovered (ω02 − ω2 )2
γ 2 ω2 corresponds to the region of
by T. Hariot (approx. 1605), then rediscovered by transparency (κ ≈ 0). In this region
J.M. Marci (1658), and experimentally studied by Ne2 1
I. Newton (1672). In the classic electronic the- n2 = 1 + . (3)
mε0 ω2 − ω2
ory of light dispersion introduced by H. Lorentz 0
an electron bound to an atom, ion or molecule in For metals the classic Drude–Zener–Kronig the-
the crystal lattice is modeled by a harmonic os- ory (P. Drude, C. Zener, R. Kronig) of light dis-
cillator. In the field of the light wave Eeiωt with persion is based on the free electron model. For
frequency ω the oscillator performs forced vibra- this model Eqs. (1) and (2) reduce to the form
tions near its equilibrium position, and this in-
duces an ac polarization of the crystal at this fre- Ne2 1
n2 − κ 2 = 1 − , (4)
quency. When the frequency ω approaches the mε0 ω2 + γ 2
natural frequency ω0 of the electron’s vibration,
Ne2 γ
there is a resonance that enhances the light ab- 2nκ = . (5)
sorption. The frequency dependences of the refrac- mε0 ω ω2 + γ 2
tive index n and the absorption index κ of non- Ideal metals with infinitely long average electron
conducting crystals are determined by the rela- mean free paths should not absorb light, so γ = 0
tions and
Ne2 ω02 − ω2 Ne2
n2 − κ 2 = 1 + , (1) n2 = 1 − . (6)
mε0 (ω2 − ω2 )2 + γ 2 ω2 mε0 ω2
0
(mε0 ω2 ) > 1, n has an imaginary value, and to- displacement of layers to alternate structure po-
tal internal reflection occurs for all angles of in- sitions (e.g., from layers A or B of the wurtzite
cidence. As the free motion of conduction elec- hexagonal type structure to layers A, B or C of the
trons, limited by their mean free path, only lasts sphalerite FCC structure), and turbostratum stack-
for very short time intervals, the above extreme ing faults where the layers are displaced to random
cases are not easily observed. Nevertheless, the locations. The first type of disordering is typical of
experiments of R. Wood with thin layers of alkali SiC, ZnS, and other tetrahedral structures tending
metals demonstrated their transparency in the UV toward polytypism. In this case no turbostratum de-
spectral range and their strong absorption in the fects are formed. The largest amount of layer dis-
visible and infrared. The quantum theory of light ordering in tetrahedral structures was detected in
dispersion leads to similar results. lonsdalite, a hexagonal modification of diamond.
The second kind of disordering takes place exclu-
DISPERSION OF MAGNETIC ANISOTROPY in sively in strongly anisotropic laminated structures,
magnetic materials such as graphite, graphite-like boron nitride, and
Dependence of the constants of the magnetic some clay minerals. The concentration of poly-
anisotropy, and the direction of the easy magne- type and turbostratum stacking faults is deter-
tization axis on the crystal structure spatial coor- mined from characteristic features in diffraction
dinates. The dispersion of anisotropy constants is patterns of layer-disordered structures.
called amplitude dispersion, and that of the easy
magnetization axes is called angular dispersion. DISPLACEMENT, ELECTRIC
In magnetic lms the angular dispersion leads to See Electric ux density.
the appearance of the so-called magnetization rip-
ples (or ripple structures).
DISPLACEMENT SPIKE
The region of a solid, which contains atoms
DISPERSITY
subjected to inelastic displacements and vacant
A parameter of a disperse structure associated
lattice sites, which are generated as a result of
with the dimensions of inclusions (e.g., precip-
propagation of the collision cascade of atoms pro-
itated particles) in a medium. The term disper-
duced by irradiation with energetic particles. Two
sity often denotes the average size of scattered
material. More precisely, dispersity is character- zones are distinguished within the displacement
ized by the distribution function of the dimen- spike: the depletion zone, which is located at the
sions of dispersed inclusions, or by the distribution center of the cascade and contains an excess quan-
function of the dispersive phase volume according tity of vacant sites, and the enriched zone, which
the sizes of particles F (V ). The probability dW occupies the peripheral region of cascade and con-
to find an inclusion with a volume between V tains an excess quantity of interstitial atoms. See
and V + dV is given by the expression dW = also Seeger zone.
F (V ) dV . A monodispersed system is character-
ized by the presence of one narrow maximum DISPLACEMENT VECTOR, vector of
F (V ), corresponding to the most probable dimen- displacement, vector of a deformation
sion of the inclusions. For polydispersive systems, The difference u of radius vectors of a point in
where the function F (V ) has the form of a diffuse the solid without (x) and with (x ) a strain; u =
curve, perhaps with a succession of local maxima, x − x.
it is more correct to characterize dispersity by the
specific surface area of the dispersed inclusions. DISPYRATIONS
The arrangement of Volterra dislocations (see
DISPLACED LAYER STRUCTURES Somigliana–Volterra dislocations) which simulta-
Crystals with high concentrations of parallel neously combine disclinations and translational
stacking faults of various kinds, which break the dislocations subject to the condition that the Frank
regular alternation pattern of atomic layers in a lat- dislocation vector of disclination is parallel to
tice. These include polytype defects formed by the the Burgers vector of the dislocation. Perfect
DISSIPATIVE STRUCTURES 333
DOMAIN, MAGNETIC
See Magnetic domain.
have energetically less favorable magnetizing ori- of the Walker solution, we have η = λvσ (0)/Δ,
entations. In this case the domains have magne- where Δ is the wall thickness, an extended phe-
tized directions close to that of the overall magne- nomenological description gives λ → λeff , where
tizing field. One can distinguish reversible and ir- λeff involves exchange coupling, and relaxation
reversible displacement processes. At low applied arising from the spin orbit and dipole–dipole inter-
fields the initial gently sloping part of the mag- actions (see Landau–Lifshits equation). When tak-
netization curve (region of initial magnetic sus- ing into account the domain wall twisting or the
ceptibility) corresponds to quasi-elastic reversible presence of Bloch lines, the value of η increases,
displacement processes. On this portion of the in some cases, quite considerably.
curve magnetic hysteresis is almost absent. At In materials which contain magnetic ions with
higher fields irreversible displacement processes strong spin–orbit coupling (e.g., in rare-earth iron
become prominent in the next part of the magne- garnets and orthoferrites) the domain wall brak-
tization curve associated with a maximum mag- ing results from the wall energy transfer to rare-
netic susceptibility. In this region intermitent irre- earth ions. In this case, an ordinary phenomeno-
versible displacements of domain walls take place logical description based on the relaxation term of
(see Barkhausen effect), which constitute the main the Landau–Lifshits equation is in good agreement
mechanism responsible for magnetic hysteresis in with experiment. These materials exhibit, as a rule,
ferromagnets. The displacement process contribu- large magnetic losses and the value λ = 1 to 10−2 .
tion to the magnetization is determined by the size In magnetic materials with small losses, and lack-
and form of the magnetic texture, by characteris- ing the ions indicated above, the main contribution
tics of defects in crystals, by domain wall struc-
to the domain wall drag arises from the interaction
ture.
of the wall with thermal quasi-particles (especially
with magnons). In this case an ordinary phenom-
DOMAIN WALL DRAG in a magnetic material,
enological description does not agree with, exper-
domain wall braking
iment: e.g., the values of the relaxation constant λ
This braking or frictional drag is determined
determined from measurements of the wall mobil-
by the energy dissipation of the moving domain
ity, and from the width of the magnetic resonance
wall, and is described by the frictional force which
line, can differ considerably from each other. Us-
acts on and retards the motion of the wall: v·F b =
ing a microscopic description, the drag force of
−dσ/dt , where σ is the wall energy. The domain
the wall is expressed in terms of the probability
wall drag is caused by: defects in crystals (disloca-
of processes involving the interaction of thermal
tions, impurities); the presence of dynamic braking
magnons or photons with the moving wall.
(viscous friction) due to the transfer of the domain
wall energy to various subsystems of the magnetic For magnon relaxation processes there is a
system which are close to the state of thermody- reduction of the domain wall drag or, which is
namic equilibrium (to magnons, phonons, quasi- equivalent, a rise of the domain wall mobility
particles, magnetic impurity atoms, and so on). with decreasing temperature; as a rule η ∝ T n
The domain wall drag determines the charac- or μ ∼ T −n , where n = 1 to 4. Such behavior is
ter of the magnetic domain wall dynamics: the ve- readily observed in magnetic materials with small
locity of its stationary uniform motion under ac- losses. Relaxation from impurity ions grows with
tion of the external force (magnetic pressure Pm ), the decrease in temperature, and becomes substan-
or the dissipative characteristics at its nonuniform tial at low temperatures (see Fig.).
motion. At small velocities v, the force is F b = Upon increasing the domain wall velocity, the
F0 v/|v| + ηv, where F 0 is the force of friction at dependence F (v) becomes nonlinear, resulting in
rest relative to the coercive force of the magnetic a nonlinear dependence v(Pm ). If the wall veloc-
material, η is the viscosity coefficient. Therefore, ity is close to the sound velocity vs then phonon
the velocity is v ∝ (P m − P 0m ), where P m is re- Cherenkov radiation emerges, and induces sub-
lated to F 0 (see Domain wall mobility). When de- stantial growth of Fb (v) at v ∼ = vs . In this case,
scribing domain wall dynamics in the framework there appears an anomaly in the dependence v(B)
DOMAIN WALL TWISTING 337
Electrical conductivity at room temperature (left), and conduction activation energy (right), of amorphous Si as functions
of composition of the gas mixture used to form the films.
have fourfold coordination in Si. Boron impurities (p–n junction in amorphous Si), solid-state lasers,
result in hole conduction. Ion implantation is used etc.
to dope amorphous silicon with H as a donor im-
purity. The conductivity of amorphous silicon in- DOPING
creases from 10−17 –10−18 to 10−11 (·cm)−1 Introducing impurity atoms into semiconduc-
when doped (in a melt) with 2·10−3 % of O as tors to increase the electrical conductivity of the
an acceptor impurity, and up to 10−9 (·cm)−1 material, i.e. to produce donor centers in elec-
when 5·10−2 % of donor Cl is added. Chalco- tron semiconductors and acceptor ones in hole
genide glasses significantly increase their con- semiconductors. The corresponding elements are
ductivity when As2 S3 , As2 Se3 , and As2 Te3 are called doping impurities. Pairs of elements form-
doped with Cu, Ag, In, Ti impurities. GeS reacts ing centers of opposite types are called compen-
similarly to Ca and Ag impurities. The mecha- sating impurities. Both doping and compensat-
nism for the increase in the conductivity of doped ing impurities also affect the photoelectric, op-
amorphous semiconductors consists in reconstruc- tical, magnetic and other properties of semicon-
tion of their atomic structure and accompanying ductors. A particular feature of semiconductors is
changes in the ratio of their negatively and pos- the high sensitivity of their properties (especially
itively charged free bonds. Transition series and electric) to the introduction of impurities. As for
rare earth ions added to oxide glasses produce pure materials, their electric properties start to al-
new local electronic levels in the forbidden gap of ter noticeably at impurity concentrations of 1010 –
the latter. The levels influence optical absorption, 1011 cm−3 . Monocrystals may be doped during
luminescence, and electron paramagnetic reso- the growth process in the following ways; add ele-
nance spectra. ments to the melt from which the solid crystallizes;
Due to low production costs, and the possibil- stimulate thermally activated diffusion from a solid
ity of varying their physical properties over wide surface coated with a doping element beforehand;
limits, doped amorphous semiconductors present perform ion bombardment (see also Ion implanta-
a promising class of materials for various engi- tion, Ionic synthesis); subject the doped material
neering applications. These doped semiconduc- to nuclear radiations that precipitate nuclear reac-
tors have been successfully used for solar cells tions in it (see Nuclear doping). Diffusion from the
340 DOPING ATOM
surface and ion implantation are used to dope sur- emission of photons):
face layers so as to achieve surface hardening of ⎧ #
metal parts, and to change the distribution of elec- ⎪
⎪ ω0 1 − β 2
⎪
⎨ 1 − βn(ω) cos θ , βn(ω) cos θ < 1,
trical resistance in electronic semiconductor de-
ω= #
vices. ⎪
⎪ ω0 1 − β 2
⎪
⎩ , βn(ω) cos θ > 1,
βn(ω) cos θ − 1
DOPING ATOM
where ω0 is the intrinsic frequency of the radia-
An impurity atom occupying a lattice site (sub-
tor at rest in the fixed coordinate system, β = v/c
stitutional), or at an off-site position (interstitial) in
(v is the radiator speed), and n(ω) is the refrac-
a crystal lattice, e.g., in the empty space between
tive index (the medium is assumed to be trans-
surrounding atoms located at the neighboring sites
parent and isotropic). For particles channeling in
of a crystal lattice (see Interstitial atom, Substi-
a crystal, ω0 is determined by the difference in
tutional atom, Dumb-bell interstitial defect, Crow-
the energies of the transverse motion. In the mo-
dion).
tion of a radiator with a velocity whose projec-
tion on the direction of propagation of the pho-
DOPING, ION IMPLANTATION ton is smaller that the phase velocity of light in
See Ion implantation doping. the medium (e.g., v cos θ < v ph ), the photon emis-
sion event takes place at the transition of the sys-
DOPING, NUCLEAR tem from a more excited state to a less excited
See Nuclear doping. state (normal Doppler effect). In the opposite case
(v cos θ > v ph ), the photon emission is accom-
DOPING, SURFACE
panied by excitation of the system (anomalous
Doppler effect). The equations for ω in a medium
See Surface doping.
with dispersion are not linear with respect to ω.
As a result, for a given ω0 , v, and θ the emis-
DOPPLER EFFECT in solid state physics sion of quanta with different frequencies is possi-
(Ch. Doppler, 1842) ble (complex Doppler effect). If the radiator moves
A change of frequency ω perceived by an ob- in a vacuum there is no dispersion, and a quantum
server when moving toward or away from an oscil- of a single frequency is emitted (simple Doppler
lating source. The frequency of the waves emitted effect). A source with an arbitrary intrinsic fre-
by a radiation source which moves in a medium quency cannot overtake the irradiated energy at
depends significantly on the velocity v of the radi- any speed and in any refracting medium. It is
ator and the phase velocity v ph of the waves. Any for this reason that there is always a wave in the
system can serve as a radiation source if it pos- emission spectrum for which v cos θ/W (ωθ ) < 1,
sesses its own inherent oscillation frequency, e.g., where W (ωθ ) is the group velocity of the waves
atoms, molecules, electrons oscillating under the radiated at the angle θ . The inverse condition
action of alternating external fields, and channeled v cos θ/W (ωθ ) 1 determines the range of exis-
(see Channeling) relativistic particles in a crystal. tence of a complex emission spectrum. Therefore,
Not only can light waves be emitted in a medium, the anomalous Doppler effect is always compli-
but also plasma (longitudinal) waves (see Plasma cated since, in addition to a frequency for which
oscillations), sound waves, etc., that is all possi- v cos θ/W (ωθ ) > 1, there is always a normal com-
ble types of waves that can exist in a given crys- ponent for which v cos θ/W (ωθ ) < 1. In this case
tal. From the conservation laws of energy and mo- the frequency of the normal component is higher
mentum for the source before and after emission than the anomalous one. As a result, the normal
the following equations are obtained for the fre- component that overtakes the radiator carries away
quency ω of a quantum emitted at an angle θ rel- more energy per unit frequency than the anom-
ative to the direction of motion of the source (in alous component. The expressions obtained are
what follows, for definiteness we will speak of the valid for an optically transparent body. Taking into
DOPPLERON–PHONON RESONANCE 341
account absorption, photon emission accompanied collision-free types (see Magnetic collision-free
by system excitation also occurs in a medium with damping). The latter is determined by the reso-
n < 1. Both complex and anomalous Doppler ef- nance cyclotron absorption of the wave by other
fects are observable in the X-ray range. In view groups of carriers at the Fermi surface, and by
of a small departure of the X-ray refractive index the magnetic Landau damping that suppresses the
from unity, to observe the effect in this case one dopplerons when B is bent away from c. The
should use channeled relativistic particles For a propagation of dopplerons is also possible along
non-stationary medium (whose parameters change open Fermi surfaces in a magnetic field parallel to
with time) a frequency change can occur even the metal boundary.
when both the radiator and the acceptor are mo- Specific features of dopplerons are the circular
tionless, and this is called the parametric Doppler polarization of the wave in the plane of B, the de-
effect. pendence of the wavelength on the magnetic field,
the threshold (in regard to frequency and magnetic
DOPPLERON field) nature of the spectrum, and, as a rule, anom-
A low-frequency, weakly decaying, slow elec- alous dispersion (dω/dk < 0). These features can
tromagnetic wave that can appear and propagate appear, for example, in compensated metals, in the
in a metal in the presence of a permanent mag- long-wave limit, or at certain singularities at the
netic field B. A doppleron is one of five known Fermi surface such as saddle points and local min-
intrinsic electromagnetic excitations in a dense ima of curvature.
electron metal plasma (see Solid-state plasma).
Its existence is caused by the Doppler-shifted cy- DOPPLERON–PHONON RESONANCE
clotron resonance which is a consequence of the A resonance in an applied magnetic field in-
Fermi degeneracy of conduction electrons in met- volving the interaction of dopplerons with sound.
als. Therefore, a doppleron as a particular feature It is similar to a helicon–phonon resonance. The
of metals, in contrast to a helicon, has no ana- doppleron–phonon resonance exists in metals un-
logues in an equilibrium gas plasma. Being a col- der the conditions needed for the existence of a
lective excitation of a metal plasma, a doppleron doppleron; involving the same circular polariza-
nevertheless is closely connected with the local tion in the range of intersection of the doppleron
characteristics of the Fermi surface, and this de- wave and sound wave spectra (assuming no cou-
termines the diversity of information contained in pling between elastic and electromagnetic subsys-
it. On the one hand, there is a possibility of restor- tems), i.e. for coinciding phase velocities of the ul-
ing the local conduction in a magnetic field, and on trasonic and doppleron waves. The wave coupling
the other hand, one can extract accurate informa- results in the removal of degeneracies in the res-
tion about the Fermi surface curvature, the mean onance region. However, in the case of the weak
free path of resonance groups of carriers, the state bond that is specific for the doppleron–phonon
of the metal surface, and so on. In the case of com- resonance, a mode repulsion does not take place:
plicated Fermi surfaces, the existence of several the sound wave brings elastic energy to the res-
doppleron modes is possible, including multiple onance range, while the doppleron wave brings
dopplerons for a non-axial Fermi surface. electromagnetic energy. The damping of both the
Doppleron propagation in pure metals is pre- waves in the resonance gives rise to a resonance
dominantly along the axis of high crystal sym- maximum. Since the doppleron is circularly po-
metry c (k||B||n||c) in strong magnetic fields B larized the resonance does not exist for a longitu-
at low temperatures when ω ν Ω, where dinal ultrasonic wave, appearing only for circular
ν is the collision frequency of carriers, Ω is the polarization of the transverse wave. As a result,
cyclotron frequency, n is a unit vector normal for linearly polarized transverse elastic waves in
to the metal surface, ω and k are the frequency the region of resonance a rotation of light polariza-
and the wavevector of the doppleron wave, re- tion plane and ellipticity may arise. The transfor-
spectively. Attenuation is characteristic of both mation of electromagnetic waves into ultrasonic
collision (with photons and lattice defects) and waves and vice versa can occur in the resonance
342 DOPPLER-SHIFTED CYCLOTRON RESONANCE
DOPPLER-SHIFTED CYCLOTRON
RESONANCE
A resonance interaction of the conduction elec-
trons that are drifting along a magnetic field B
with an electromagnetic wave propagating in a Doppler-shifted cyclotron resonance.
metal. This Doppler-shifted cyclotron resonance
appears when the electron velocity ṽ along the
wave vector k averaged over the cyclotron pe- usual temporal cyclotron resonance, is, in fact, a
riod differs from zero, i.e. if k is not perpendic- space resonance.
ular to B, or in the presence of open cyclotron or- The spectrum of this resonance (see Fig.) in the
bits on the Fermi surface, and kl
1 , where l is coordinates (k, B) represents a straight line that
the mean free path of the free carriers. The prop- passes through the origin, and separates the re-
agation of an ultrasonic wave in a metal is known gions (shaded) where the collision-free Doppler-
as a Doppler-shifted acoustic cyclotron resonance, shifted cyclotron absorption is present, and where
and it satisfies the following condition in the free- it is absent. The latter is caused by electrons with
electron model: values of ṽB varying from the Fermi velocity v F at
the base point to zero at the central cross-section of
|ω ± kB ṽB | = ωc , (1) the Fermi surface. The Doppler-shifted cyclotron
resonance results from the electron Fermi–Dirac
called the Kjeldaas edge (1959). This means that
statistics and the presence of a sharp velocity dis-
electrons drifting toward the wave shift their fre-
tribution of electrons.
quency to the cyclotron value ωc owing to the
In the case of an anisotropic Fermi surface,
Doppler effect. In connection with the high elec-
the Doppler-shifted cyclotron resonance can ex-
tron velocities (≈108 cm/s) on the Fermi surface,
ist for a few groups of carriers. Distinguished (or
the Doppler shift of the frequency is very large,
resonant) are the groups with extremal values of
ω kB ṽB , and the resonance is observed at fre-
quencies ω ωc . The condition (1) takes the form u(pB ), because in this case a large number of car-
riers is formed which move in phase, and at the
ωc = |kB ṽB | (2) same time do not contribute to the wave absorp-
tion. The contribution of other carriers is insignif-
or icant due to their out-of-phase motion, and can be
u(pB ) = λB , regarded as providing a background signal.
(3) For a non-axial Fermi surfaces the condition
∂s
u(pB ) = eB , for the Doppler-shifted cyclotron resonance takes
∂pB
the form
i.e. those electrons are involved in the resonance
whose displacement along B within a cyclotron |kB ṽB | = nωc , (4)
period, u(pB ), is equal to the wavelength λ, where u(pB ) = nλ, (5)
p is the electron momentum, and s is the area of
the cross-section of the Fermi surface defined by where the integer n is determined by selection
the plane pB = const. Thus, the Doppler-shifted rules. For example, for the resonance with longitu-
cyclotron resonance, similar in many ways to the dinal sound n = mα; and in the case of transverse
DOSIMETRY 343
sound n = mα ± 1, where m = 0, ±1, and α is the ionization, excitation, and elastic interactions. If
order of the crystal symmetry axis. nuclear reactions and ionizing radiation sources
The nature, shape, and magnitude characteriz- are present in volume
V then
E should take
ing the resonance as a function of the sound ab- into account the total energies from the nuclear
sorption coefficient Γ versus B are dependent on transformations E n and the sources of radiation
the type of ultrasonic wave (longitudinal or trans- energy E s :
verse), its frequency and polarization, the multi-
plicity of the resonance n, the type of resonance
E = E in − E out + E n + E s . (2)
singularity ∂s/∂pB , and so on. Doppler-shifted
The macroscopic values determined in this fash-
cyclotron resonance can manifest itself (converted
ion, i.e. the absorbed energy and the dose, do not
to a function of ω, B, etc.) in the form of either
an edge, an absorption step, or a resonance or reflect the stochastic nature of the interaction of
antiresonance. At sufficiently high ω and l, the ionizing radiations with matter. Therefore, when
line of the Doppler-shifted cyclotron resonance determining the dose,
m and
t should be, on
can be split (see Eq. (1)). This resonance allows the one hand, sufficiently large (to exclude the ef-
one to determine the curvature of the Fermi sur- fect of fluctuations), and, on the other hand, suffi-
face (∂s/∂pB )extr , the effective mass of the carri- ciently small so as to be able to disregard macro-
ers of resonance groups, the mean free path, and scopic quantitative and qualitative changes in the
so on. See also Doppleron. incoming radiation (i.e. active spectrum) and in
the radiation-damaged material under considera-
DOSE, SMALL
tion.
See Small-dose effect. A determination of the absorbed radiation dose
is related to the measurement of any dose-de-
DOSIMETRY
pendent radiation effect in matter. However, a
A determination of the effects of nuclear ra- radiation-stimulated process convenient for mea-
diation on irradiated objects. A quantitative mea- surement may be lacking in the irradiated mater-
sure of these effects depends on the case (solids ial for which the absorbed dose has to be deter-
experience a change of resistance, optical absorp- mined. In such cases a standard material is intro-
tion, strength, etc.). However, independently of duced at the detection point, and the absorbed dose
secondary effects, the primary cause of radiation- is deduced from the radiative reactions in the ex-
induced effects and the most general physical tra material The Bragg–Grey probe method which
value that characterizes the interaction of radia- is widely used to investigate irradiation damage
tion with the matter is the absorbed energy. The by both non-charged and charged high-energy par-
absorbed energy dose D is the macroscopic value ticles is based on this principle. A probe, that
that best characterizes the absorbed energy: is a material introduced at the detection point,
must satisfy the Bragg–Grey conditions: (i) the ac-
E tive spectrum at the probe should be the same at
D= , (1)
m that detection point in the absence of the probe;
where
E is the energy absorbed in the bulk, and (ii) charged particles generated in the probe by
m is the mass of the substance. In SI units the indirectly ionizing non-charged particles should
absorbed radiation dose in joules per kilogram is not contribute significantly to the irradiation dose
called a grey (Gy): 1 Gy = 1 J/kg. Let Ein be absorbed in the probe. These conditions are suf-
the sum of the energies of all the particles that ficiently satisfied if the linear dimensions of the
during time
t enter volume
V which contains probe are small compared to the mean free path R
the mass
m, and E out be the energy sum of all of the charged particles (e.g., 0.1R). If the Bragg–
particles that leave this volume during the same Grey conditions are satisfied, and the mean energy
time. If there are no sources of ionizing radia- ε of the active spectrum of charged particles (elec-
tion, and no reactions, then the absorbed energy trons) at the detection point is known, then with
E = E in − E out will be expended entirely for the help of the Bragg–Grey equation one can find
344 DOUBLE LAYER, ELECTRIC
the dose D m in the material with a good accuracy: or frequencies in different ranges of the electro-
magnetic spectrum, those that belong to spins of
Sm( ε )
Dm = Dp , (3) different types (electrons and atomic nuclei), or
S p (ε) to identical particles which differ by some pa-
where D p is the experimentally found dose in rameter. Double resonances may arise from the
the probe; S m ( ε ) and S p ( ε ) are respectively the same type of interaction (e.g., electromagnetic
mean braking abilities of the material and of the wave) or from different types (e.g., electromag-
probe, which represent the mean charged particle netic and acoustic waves). The most common are
energy loss per unit of path divided by the ma- magnetic in nature: electron–nuclear double reso-
terial density. A more exact characteristic of the nance (ENDOR), acoustic double magnetic reso-
radiation in the medium (in comparison with the nance, and optical–magnetic resonance, and these
absorbed dose) can be obtained from the deter- are all various methods of magnetic resonance
mination of the active spectrum. Knowledge of spectroscopy.
this spectrum allows one to calculate the absorbed In double magnetic resonance there is selective
dose and the yield of various radiation-stimulated absorption (radiation) at two resonance frequen-
reactions. cies of electromagnetic energy by a spin located
in a constant magnetic field. Observation of double
DOUBLE LAYER, ELECTRIC magnetic resonance simplifies the spectrum of one
See Electric double layer. resonance by suppressing the interaction between
the spins through high power irradiation at the sec-
DOUBLE RESONANCES ond resonance frequency, and it also increases the
Resonances in systems oscillating at two fre- sensitivity or resolution of the low-frequency reso-
quencies arising from external effects. In solids nance by recording it as part of the high frequency
where the constituent microparticles have a set of spectrum.
discrete energy levels εi , εj , εk , . . . the two res- In nuclear–nuclear double resonance (NNDR)
onances are due to transitions (indicated in Fig. the nuclear magnetic resonance (NMR) experi-
by arrows) between pairs of levels with different ment is carried out simultaneously at two frequen-
energy separations corresponding to the frequen- cies which belong to atomic nuclei with different
cies νij and νj k . Usually in double resonance ex- energy spacings between their spin sublevels. It is
periments there is a coupling between the transi- used to investigate substances with nuclei (1 and
tions through interactions which permit one reso- 2) coupled together though the spin–spin interac-
nance to be recorded through the other one. The tion, which causes each individual resonance to in-
two pairs of levels may have similar or differ- fluence the other. Homonuclear NNDR involves
ent natures (e.g., magnetic sublevels with one pair two nuclei of the same type and heteronuclear
involving an electron flip and the other involv- NNDR involves nuclei of different types. In the
ing the flip of a nuclear spin), close frequencies former case the resonance at nucleus 1 is detected
at low power, and then the irradiation of resonance
2 at high power suppresses their interaction. This
simplifies the spectrum of nucleus 1 and enhances
the amplitude of the detected NMR signal, a phe-
nomenon called the Overhauser effect. A second
experiment, called internuclear double resonance
(INDOR), is a method of recording the NMR of
nucleus 2 in a manner similar to double electron–
nuclear resonance. The technique of NNDR per-
mits one to study the NMR of rare isotopes or of
nuclei which provide weak signals disturbed by in-
Double resonances. teraction with other nuclei, to determine the signs
DROP-IMPACT EROSION 345
of the spin–spin interaction constants, and to ob- significant when E reaches a value of 106 to
tain information about processes of nuclear spin 107 V/cm. The Dow–Redfield model was applied
relaxation. to explain the Urbach rule that involves an ex-
In electron–electron double resonance ponential law for the α(ω) dependence near the
(ELDOR) the electron paramagnetic resonance light absorption edge in amorphous semiconduc-
(EPR) is performed simultaneously at two mi- tors. It is assumed that there exist strong local (on
crowave frequencies (of observation and pump- the scale of atomic dimensions) electric fields in
ing) corresponding to the resonance lines of two amorphous semiconductors. These fields are due
pairs of spin levels. These lines may involve tran- to a local distortion of the atomic structure of the
sitions between the spin sublevels of one paramag- short-range order inherent in amorphous materi-
netic center, or between two different centers cou- als.
pled by the spin–spin interaction. In ELDOR the
investigated substance is placed into a special bi- DRAG OF DOMAIN WALL
modal resonator where it is subjected to the ef- See Domain wall drag.
fect of two microwave fields of different frequen-
cies. Variations are observed in the EPR signal at DRAWING in metallurgy
the observation frequency when the power level A method based on metal pressure treatment,
is changed at the pumping signal frequency. This whereby stock of fixed cross-section is introduced
experiment provides information about spin–spin into the channel of the drawing instrument (die)
interactions, cross-relaxation, and mechanisms of and is pulled (drawn) through it. The shape of the
EPR line broadening. channel cross-section is the same or close to that of
the metal being pulled, and it smoothly decreases
DOUBLET
from the point of metal entry to its exit. During
The term is employed in two senses: a dou-
the process of drawing the stock is deformed and
ble degeneracy of levels (e.g., a Kramers doublet),
its cross-section varies, at the end attaining the
and two closely situated lines in the spectrum of
shape and dimensions of the narrowest section of
a resonance absorption which are usually of the
the channel.
same origin (e.g., hyper ne structure in an elec-
tron paramagnetic resonance spectrum when the
DRIFT, AMBIPOLAR
nuclear spin I = 1/2). In a similar way, one can
See Ambipolar drift.
speak about a triplet (three), a quadruplet or quar-
tet (four), and a general expression for such terms
DRIFT, MAGNETIC
is multiplet (many). For a single line or an absence
of degeneracy it is common to use the term sin- See Magnetic drift.
glet.
DRIFT OF CURRENT CARRIERS
DOW–REDFIELD MODEL (J. Dow, D. Red eld, See Current carrier drift.
1972)
A description of the light absorption in the case DROP-IMPACT EROSION
of direct (i.e. without emission and absorption of Failure of a solid surface under the action of
phonons) exciton transitions in a strong uniform nonstationary fields of stresses arising at impact
electric field E. It was shown that the light absorp- from high-speed drops of a liquid. The threshold
tion coefficient α(ω) near the absorption edge speed vk , below which no failure occurs, is related
energy ε0 depends on the photon energy ω ac- to the drop diameter d by vk d = const. The dam-
cording to an exponential law, i.e. as age results from the stress wave propagation, jet-
type spreading of a drop, and penetration of a liq-
β(ε0 − ω)
α(ω) ∼ exp − , uid inside the defects of a solid. The mean value
E of the pressure in the contact zone is given by
where β is a constant, and E is the electric field P = ρcv, where ρ is the liquid density and c is the
strength. The light absorption at ω < ε0 becomes speed of sound in the liquid. The maximal pressure
346 DRUDE–LORENTZ–SOMMERFELD MODEL
3P is observed after a time t = 2dv/c2 from the a dumb-bell domain undergoes a rotation whose
instant of contact, and is concentrated on the con- direction is determined by the sign of the vertical
tact zone boundary. Corresponding to this instant Bloch lines. See also Magnetic bubble domains.
is the beginning of the drop spreading over the sur-
face at the speed (3–4)v. Under a single impact at DUMB-BELL INTERSTITIAL DEFECT
transonic or supersonic speed depressions appear A configuration of an interstitial atom in a
in metals and alloys, sometimes with a crater at the monatomic crystal where the interstitial atom dis-
center. In nonmetals, circular cracks arise with a places one of the adjacent lattice atoms from its
subsurface zone of failure at the central axis. Un- lattice position position O and forms with it a pair
der multiple impacts we define drop-impact ero- with the center of gravity at O. The energy Ei+
sion at the stage of steady-state failure by the ratio of a dumb-bell depends on the orientation ij k
of mass loss to the area and the amount of liquid of its axis. In many face-centered cubic crystals
striking the solid surface. The dependence of ero- this is a minimum for orientations 100 while in
sion E on the drop speed v and the angle θ of the body centered cubic crystals the minimal value is
drop speed direction to the normal to the surface attained for orientations 110. Computer calcula-
is given by the relation E ∝ (v cos θ − vk )n . The tions gave the following results: Cu: Ei+ 3 eV,
value of n depends on the properties of the mater- α-Fe: Ei+ = 3–4 eV. A migration of a dumb-bell
ial and the method of testing, and varies within the involves the displacement of a neighboring atom
range 2.5 to 6. See also Surface erosion. from its lattice position followed by the formation
with it a new dumb-bell with another orientation.
DRUDE–LORENTZ–SOMMERFELD MODEL The other atom of the initial dumb-bell returns si-
See Macroscopic description. multaneously to its original lattice position. With
the help of computer calculations, using various
DUMB-BELL DOMAIN interatomic potentials (see Interatomic interaction
A kind of rigid magnetic bubble domain, oval- potentials) one can estimate the dumb-bell migra-
shaped with a neck. Its specific feature is the great tion energy, which corresponds to the migration
number (>100) of vertical Bloch lines in a mag- energy of an interstitial atom; for Cu Eim = 0.05–
netic domain wall giving the predominant contri- 0.13 eV, in agreement with the experimental value
bution to the wall energy. In the case of a high den- 0.12 eV. The turning energy of a dumb-bell with-
sity of vertical Bloch lines, a corresponding diam- out a displacement of its center of gravity is 3 to
eter of a cylindrical magnetic domain can become 4 times higher. In substitutional alloys interstitial
greater than the diameter of the elliptic instabil- atoms can form mixed dumb-bells consisting of
ity that is the cause of the formation of an “oval” atoms of two kinds.
domain. The appearance of a neck on the oval is
due to the dependence of the equilibrium distance DURABILITY
between the vertical Bloch lines on the curvature Time during which a solid preserves its prop-
of the domain wall. A dumb-bell domain is sta- erties. In the case of strength, durability is an im-
ble over a certain range of magnetizing field varia- portant characteristic for long-term loading when
tion, at the lower boundary of which the dumb-bell a creep develops in a solid. Here durability is the
domain transforms into a strip domain; and at the time before fracture. The higher the temperature
upper boundary it either collapses (in the case of and the higher stress, the shorter the time t f before
very high density of vertical Bloch lines), or trans- fracture:
forms into an elliptic domain. The field strength U −γσ
of the dumb-bell domain collapse (unlike the field t f = A exp 0 .
kB T
strength of the collapse of circular rigid cylindrical
magnetic domains with a small vertical Bloch line On the other hand, the rate of stationary creep ε̇
density) decreases with an increase in the number increases with increasing stress and temperature:
of vertical Bloch lines on its surface. Under a pe- Q − γ1 σ
riodic pulsed modulation of the magnetizing field, ε̇ = B exp − 0 .
kB T
DYNAMIC HOLOGRAPHY 347
In these dependences A and B are constants; γ character of a process is not specified by its u c-
and γ1 determine the measure of the stress action tuations (noise), but rather by its inherent com-
on the height of the barrier (activation volumes), plex dynamics. Dynamic chaos, or chaos for short,
and U0 and Q0 are the activation energies of frac- possesses the properties of a random process: the
ture and quasi-viscous flow, respectively. It has autocorrelation function is reduced (see Correla-
been shown that U0 = Q0 and γ = γ1 . Therefore, tion function) and the power spectrum is contin-
processes that develop at the second stage of creep uous. The instability of trajectories in the phase
and at fracture have the same nature, i.e. fracture space of the dynamical system serves as the crite-
occurs at the second stage of creep, which con- rion of chaos. The Lyapunov exponent, often taken
firms the earlier finding for the index of disorder, is the quantitative in-
dex of this instability. In Hamilton’s dynamical
t f ε̇ = AB = const.
systems chaos occurs as a set of non-zero mea-
The validity of this relation was established using sure within the phase space where the motion has
alloys with various compositions and structures. the property of ergodicity. There is, in particu-
However, it should be realized that this expression lar, a strict proof of chaos in the system of elas-
starts to be violated at elevated temperatures and tically colliding spheres, which permits substan-
small σ values due to the change of the fracture tiation of the ergodic hypothesis for this system.
mechanism. This relation suggests the possibility In typical situations, within the phase space there
of predicting the durability of a loaded solid long are regions of chaos which coexist with regions of
before it begins to fracture from short-time tests regular (quasi-periodic) behavior. Hamilton’s dy-
of creep. There are also other methods of durabil- namic chaos appears in the dynamics of non-linear
ity prediction. Examples are parametric methods chains (lattices) of atoms, during the movement of
based on semi-logarithmic diagrams of durability particles in external periodic fields, and in prob-
versus T or 1/T dependences. However, in order lems of celestial mechanics.
to ensure the validity of the extrapolation it is im- In dissipative dynamical systems chaos corre-
portant not to try and predict durability beyond an sponds to a set of non-zero measure in phase space
order of magnitude of the time for a short term referred to as the strange attractor. As a rule,
test. the transition from regular behaviour to chaos in-
volves changes in one or more control parameters.
DYES Some scenarios for this transition (e.g., through an
Organic compounds of low molecular weight infinite succession of bifurcations or period dou-
with semiconductor-type electrical conductivity blings) involve universal numerical constants, and
and photoconductivity, capable of dyeing other they are described with the help of renormalization
materials (see Organic semiconductors). The clas- group methods. Chaos has been experimentally
sic representatives of dyes with semiconductor observed in such systems as lasers, spin-wave (see
properties are phthalocyanines, molecules which Spin waves) turbulence, non-equilibrium electron-
are planar, rather large, and contain a system of hole plasmas in solids (see Electron–hole liq-
conjugated bonds. They are close in structure to uid), non-linear electronic chains, etc. In quantum-
the porphyrin molecule. In terms of conductiv- mechanical systems the local instability of motion
ity, dyes fall into electron (crystal fuchsin – vi- is absent, thus, strictly speaking, there is no dy-
olet, aqueous fuchsin – blue) and hole (uranine, namic chaos. However, quantum systems that are
erythrosin, phosphine, etc.) types. Dyes are used stochastic in the classical limit do possess some
to build solar energy transformers. They are also features of dynamic chaos.
used as photosensitizers of organic and inorganic
materials, and as active laser media (dye lasers). DYNAMIC COOLING
See Dynamic nuclear polarization.
DYNAMIC CHAOS
Irregular, quasi-random vibrations in otherwise DYNAMIC HOLOGRAPHY
deterministic dynamical systems. The stochastic See Real-time holography.
348 DYNAMIC MIXED STATE
where SG is the coordinate along K + G, and the the amplitudes, which is especially convenient for
coefficient BGG contains information about the consideration of diffraction of radiation distortions
distortions. These equations are solved by com- in crystals with randomly distributed defects. The
puter or analytically using successive approxima- dependences of integrated reflections of X-rays
tions, depending on the distortions. In particular on the thickness of the plane-parallel sample,
cases Eq. (1) transforms to the system of equations calculated by this method, correlate well with ex-
for the many-wave diffraction of electrons, and to periment. According to the experimental data it is
the system of Takagi equations (S. Takagi, 1962) possible to determine the statistical Debye–Waller
for the two-wave diffraction of X-rays (G assumes factor L (with accuracy ∼10−3 ) and the effective
two values: 0, H ). These equations provide prac- addition
μ to the adsorption coefficient due to
tically complete solutions for electron-microscope the diffuse scattering, and these provide such para-
or X-ray contrast of defects (clusters, dislocation meters of defects as their concentrations, and the
loops, etc.) or their complexes. For very smooth average size of clusters or dislocation loops.
distortions the approximation of geometrical op-
tics applies, and Eqs. (1) may be transformed into DYNAMIC RECOVERY
the usual equations for beam trajectories. The recovery of the properties of crystals in the
The exact solution of system (1) in the two- course of plastic deformation. This includes the dy-
wave approximation is obtained for the important namic restoration of crystals, dynamic polygoniza-
case of X-ray diffraction in a uniformly bent crys- tion and dynamic recovery with reorientation. The
tal. This allows one to follow the variation of the initial stage of dynamic recovery involves heal-
mode of X-ray scattering with the growth of the ing of defects of a point type (interstitial atom,
degree of distortions (the transition from purely vacancy) formed during the course of plastic de-
dynamic scattering to the mode characteristic for formation, and it is accompanied by anomalous
a mosaic crystal, and from this to the kinematic mass transfer. The partial annihilation of disloca-
mode), to clarify the limits of applicability of the tions and their redistribution during the course of
geometric optics approximation, to investigate the deformation bring about dynamic polygonization.
focusing of plane wave X-rays, or of point sources These processes play an important role in creep
by bent crystals, etc. The other case of regular dis- with recovery, and also at the third stage of the
tortions involves deformation fields created by ul- strain hardening of the crystals. The dynamic re-
trasonic waves (see Diffraction of radiations in ul- covery with reorientation occurs in crystals with
trasonic field). For the case of statistical distor- a high degree of plastic deformation under condi-
tions the statistical dynamic theory of two-wave tions which bring about dynamic polygonization.
diffraction of X-rays in the limit lc Λ (lc is Reorientation of separate sections of a deformed
the correlation length of the distortion field) is crystal accomplishes their dehardening, and pre-
applicable, and closed form equations of energy vents the formation of cracks. Dynamic recovery
transfer of the type (1) for the coherent and inco- is observed also at temperatures of 0.1 times the
herent components of the scattered intensity are absolute melting temperature, and is enhanced by
obtained. These equations describe the following increasing the strain temperature.
effects: weakening of the reflecting ability of the
distorted atomic surfaces; increasing of the ab- DYNAMIC SCALING, dynamic similarity
sorption coefficient of coherent and non-coherent hypothesis, dynamic scale invariance
radiation due to the scattering from the random In the theory of critical phenomena it is a gen-
distortions, and the redistribution of the coherent eralization of the scaling invariance hypothesis to
component intensity to an incoherent one. Besides the function, which determines the critical behav-
this, the Takagi equations have been successfully ior of kinetic (dynamic) characteristics of the sys-
applied for the determination of deformations in tem: thermal conductivity, coefficient of sound ab-
thin (<Λ) disrupted surface layers, and in perfect sorption, etc., which have anomalies at the point
crystals from X-ray diffraction data. of a second-order phase transition. It was initially
The equations of dynamic radiation scattering introduced in the works of R.A. Ferrell, N. Meny-
may be written in the Fourier representation for hard et al. (1968), and also of B.I. Halperin,
350 DYNAMICS OF IMPERFECT CRYSTALS
P.C. Hohenberg (1969). By analogy with the static Tunnel effect) from some position to an adjacent
similarity hypothesis, wherein the critical behavior one. As a result of this latter possibility the crystal
of the system is determined by a single parame- can transform to having the quasi-particle, called
ter with the dimensions of length, namely the cor- a defecton, moving slowly through it. An intersti-
relation length, in the case of dynamic scaling it tial atom migrates with mechanical motion even
is the characteristic parameter with the dimension in a classic crystal if it is in the so-called crow-
of frequency which determines the dynamics of a dion configuration (see Crowdion). Often the me-
system in the vicinity of the phase transition point. chanical movement (slip) is characteristic of the
In particular, this involves the so-called critical re- specific linear defect called a dislocation. The dis-
tardation of the kinetic processes. placement of its line along the slip plane does not
violate the continuity of the crystal and thus it oc-
DYNAMICS OF IMPERFECT CRYSTALS curs comparatively easily. It appears that the dislo-
This field of crystal lattice dynamics is devoted cation motion is always associated with an uncon-
to the study of internal motions of real crystals (i.e. trollable variation of the shape of the crystalline
those containing defects). It includes the theory of sample, therefore the dislocation is the unit carrier
lattice vibrations with defects, and the dynamics of crystal plasticity. The atomic rearrangements
of defects. that accompany dislocation motion do not require
Vibrations of a crystal with defects. The nor- very high loads, and this is why plastic deforma-
mal vibrations of a real crystal are not plane waves tion of a crystal starts at stresses that are far be-
as they are in an ideal crystal. New vibrations may low the theoretical strength of the crystal. The mo-
appear that are completely localized in the vicinity tion of dislocations is accompanied by the emis-
of defects (local vibrations). Frequencies which lie sion of elastic (sound) waves, which is similar to
above the limiting frequency of an ideal crystal or the manner in which moving (accelerated) electric
are found in a band gap between bands correspond charges bring about the emission of electromag-
to these vibrations. If there are many point de- netic waves. Besides this, while interacting with
fects of the same type then resonance transitions of the lattice vibrations the dislocation takes part in
the local vibrations from one defect to another are the oscillator dissipative motion, and thereby con-
possible. In this case the defective crystal has an tributes to the internal friction.
impurity band of vibrational frequencies. The local Two-dimensional defects of the twin type (see
vibrations of extended defects (e.g., dislocations or Twinning of crystals), cracks or martensite inclu-
stacking faults) spread along them without pene- sions may manifest themselves as dynamic forma-
trating into the crystal volume, corresponding to tions. Together with dislocations, they play some
an inherent spectrum of dislocational or surface vi- role in the plasticity and strength of crystals.
brations, different from those of the bulk. In addi-
tion to these local vibrations so-called quasi-local DYNAMICS OF LIQUID CRYSTALS
vibrations are possible which embrace the crystal Hydrodynamics of anisotropic liquids. In the
as a whole, but with amplitudes which consider- case of nematic liquid crystals, besides the com-
ably exceed those of the regular lattice vibrations. mon hydrodynamic variables like speed v(r, t)
The frequencies of such vibrations are found in the and pressure p(r, t) there is also the director
continuous spectrum of the ideal crystal, but most n(r, t). Therefore, the usual hydrodynamic equa-
tions should be supplemented with the equation of
often in the vicinity of the edges of these bands.
motion for the director. In addition it is necessary
Local and quasi-local vibrations show themselves
to take into account that the shear stresses within
also in the appearance of additional lines in in-
the nematic (stress tensor components) may be of
frared adsorption spectra, at singularities of neu-
three types:
tron scattering, and of Mössbauer spectra.
Dynamics of defects. Point defects such as im- (1) purely fluid stresses associated with the gradi-
purity atoms, vacancies and interstitial atoms are ents of the velocity of the liquid ∇v,
capable of moving through the crystal by means (2) stresses not depending on the orientation or
of diffusion, and also via quantum tunneling (see speed of the rotation of n, and
DYSON EQUATION 351
(3) stresses involving the interaction of the ori- upon as corresponding to the dc Josephson effect.
entation of molecules n with the non-uniform The interaction of vortices with the fields within
flow (∇v). a superconductor resonator (in the high density
limit they are considered as a “density wave of
To treat these shear stresses coefficients of viscos-
Josephson current”) brings about the appearance
ity are introduced, called Leslie coefficients. The
of Fiske steps (or Shapiro steps) on the current–
viscosity (or friction) associated with the motion
voltage characteristic of the junction. The interac-
of the director may be of two types: purely orien-
tion of two fluxons of opposite polarity (“vortex–
tational and hydrodynamic, corresponding to the
antivortex” system) can involve the formation of a
second and third types of shear stresses. The dy-
dynamically bound state, etc.
namics of cholesteric liquid crystals and smectic
liquid crystals is similar to that of nematics (ne-
DYNAMIC SPIN HAMILTONIAN
matodynamics), but its formulation appears to be
more complex (see also Nematic liquid crystals). See Spin–phonon Hamiltonian.
function, M is the self energy operator (sometimes is hexagonal close packed with parameters a =
called mass operator). Using the diagram tech- 0.3591 nm, c = 0.5649 nm at room temperature
nique (see Feynman diagrams) the Dyson equation (space group P 63 /mmc, D6h 4 ). Above 1661 K
is obtained by summing an infinite series of dia- up to T melting = 1681 K β-Dy is stable with
grams. The solution of this equation with respect body centered lattice and parameter of cubic con-
to the full Green’s function is ventional unit cell a = 0.398 nm (extrapolated
1 value) after hardening (space group I m3m, Oh9 ).
G = −1 . The density is 8.54 g/cm3 , T boiling ≈ 2700 K.
G0 − M
Heat of melting is 16.6 kJ/mole, heat of subli-
This solution is formal as the self energy operator mation 280.5 kJ/mole and heat of evaporation is
in its turn contains Green’s functions, but it is use- 283 kJ/mole, specific heat is 0.170 kJ·kg−1 ·K−1
ful for investigating general properties of Green at room temperature; Debye temperature ≈185 K;
functions, as well as for their approximate calcu- linear thermal expansion coefficient of a monocrys-
lation. In the spectral representation G(ω) is given tal is 16.6·10−6 K−1 at 400 K along the main
by axis 6 and 6.0·10−6 K−1 in the perpendicu-
1 lar direction, and of the dysprosium polycrys-
G(ω) = , tal 8.6·10−6 K−1 (at room temperature); coeffi-
ω − Ω − M(ω)
cient of thermal conductivity is 10.1 W·m−1 ·K−1 .
where Ω is determined by the zero-order Hamil- Adiabatic coefficients of elastic rigidity of an
tonian. α-Dy monocrystal: c11 = 74.66, c12 = 26.16,
c13 = 22.33, c33 = 78.71, c44 = 24.27 GPa at
DYSON–MALEYEV REPRESENTATION 298 K; adiabatic modulus of uniform compres-
See Spin operator representations. sion is 41.1 GPa, isothermal modulus of uni-
form compression is 39 GPa; Young’s modulus is
DYSON SHAPE (F. Dyson, 1955) 63.08 GPa; tensile strength is 0.263 GPa, elon-
The line shape of electron paramagnetic reso- gation is 6%; Vickers hardness is 42 HV. Dys-
nance arising from conduction electrons in metals. prosium is easily subjected to mechanical treat-
It is determined by the following parameters: elec- ment. Effective thermal neutron cross-section is
tron mean free path l, thickness of skin-depth δ 1050 barn. Resistivity of a monocrystal at 298 K is
(see Skin-effect), time of electron passage through of 774 n·m along the main axis 6 and 1003 n·m
it, and times of spin–lattice relaxation and spin– in the perpendicular direction; temperature coef-
spin relaxation. The shape differs for regions of ficient of electrical resistance of polycrystalline
normal (l < δ) anomalous (l > δ) skin effect, and dysprosium is 0.00119 K−1 . Work function of
in the general case the shape is asymmetric. a dysprosium polycrystal is 3.09 eV. In dyspro-
sium two temperature regions of magnetic order-
DYSPROSIUM, Dy ing are observed: in the temperature range T c
A chemical element of Group III of the pe- to TN – helicoidal antiferromagnetism with screw
riodic table; atomic number 66, atomic mass axis along the main axis 6, and in the range 0 K
162.50; it is a lanthanide. 21 isotopes are known to T c – collinear ferromagnetism, and in both
with mass numbers from 147 to 167. Natural magnetic structures the magnetic momenta are lo-
dysprosium consists of the mixture of 7 sta- cated in the basal planes of α-Dy. Transition from
ble isotopes with mass numbers 156, 158, 160 the paramagnetic to the anti-ferromagnetic state at
to 164. Outer electronic shell configuration is the Néel point TN = 178.5 K is a second-order
4f 10 5d 0 6s 2 . Successive ionization energies are phase transition (according to the other data, a
5.93, 11.67, 22.8 eV. Atomic radius is 0.177 nm; “weak” rst-order phase transition). At the Curie
radius of the ion Dy3+ is 0.091 nm. Oxidation point T C = 86 K there is a first-order phase trans-
states are +3 (the most characteristic), +4. Elec- formation to the low-temperature ferromagnetic
tronegativity is ≈1.19. state, and when the α-Dy crystal is subjected to
In the free form dysprosium is a silvery-gray jumpwise small distortions its symmetry is low-
metal. There are α and β modifications. α-Dy ered to orthorhombic. According to other data, the
DZIALOSHINSKII INTERACTION 353
dysprosium transition from the antiferromagnetic and L is the antiferromagnetism vector. The form
state to the ferromagnetic state upon cooling is ob- of the tensor Dik is determined by the symmetry
served at 88.0 K, and the reciprocal transformation of the crystal which enters the expression for the
upon heating at 90.6 K. Magnetic susceptibility is magnetic energy. The added terms are introduced
0.102 CGS units, nuclear magnetic moment of the for the explanation of the weak ferromagnetism of
24.97% abundant isotope 163 Dy is −0.53 nuclear antiferromagnets (see Antiferromagnetism). As a
magnetons. Dysprosium is a component of some rule, the Dzialoshinskii interaction is small due to
magnetic alloys (with iron, nickel, copper and alu- the smallness of the spin–orbit interaction in com-
minum), of phosphors, fireproof ceramic materi- parison with the exchange interaction. The case
als, et al. when wD = D[M × L]z is in practice the most
important, where z is the symmetry axis direc-
DZIALOSHINSKII INTERACTION tion, because in this case the Dzialoshinskii in-
(I.E. Dzialoshinskii, 1957) teraction is enhanced due to the exchange inter-
A part of the spin–spin interactions and spin– action (T. Moriya, 1960), an enhancement which
orbit interactions in two-sublattice antiferromag- is also referred to as antisymmetric exchange. The
nets, leading through a phenomenological de- Dzialoshinskii interaction is found only in crystals
scription to the appearance of terms of the form whose magnetic symmetry group allows the pres-
wD = Dik Mi Lk , where M is the magnetization, ence of the tensor Dik .
This Page Intentionally Left Blank
Ee
EASY MAGNETIZATION AXIS
The direction of the magnetization vector of a
ferromagnet, M, which corresponds to the mini-
mum of the magnetic anisotropy energy, or the di-
rection of possible alignment of the vector M in
an infinite magnetic material with neither an ex-
ternal magnetic field nor an elastic strain present.
If the magnetic anisotropy is not unidirectional, Edge angle.
the easy magnetization axes are axial directions,
so their number is always even, depending on the
liquids (see Wetting). When the surface of the solid
symmetry and nature of the magnetic anisotropy.
is deformed, the equilibrium of a droplet is char-
In the case of unidirectional anisotropy, the easy
acterized by two angles, θ0 and α, as shown in the
magnetization axes represent polar directions. The figure. These angles are related to surface tensions
direction corresponding to the absolute minimum σ L , σ S , and σ SL of the liquid, solid, and solid–
of the magnetic anisotropy energy is sometimes liquid interface, respectively (see Surface tension),
referred to as the easiest magnetization axis. See by the following expressions:
also Hard magnetization axis.
σ S = σ L cos θ0 + σ SL cos α,
ECHELETTE σ L sin θ0 = σ SL sin α.
See Diffraction grating. By measuring θ0 and α and knowing σ L , one may
calculate σ S and σ SL .
ECHO, ELECTROACOUSTIC
See Electroacoustic echo. EDGE DISLOCATION
A dislocation line which is normal to a Burg-
ers vector. In an infinite isotropic elastic medium
ECHO, PHONON (see Isotropy of elasticity) a linear edge disloca-
See Electroacoustic echo. tion creates a plane strain inversely proportional
to the distance from it, and independent of the
ECHO, SPIN position along the dislocation. The Burgers vec-
See Spin echo. tor in a crystal coincides with one of the lat-
tice vectors. An edge dislocation can be repre-
sented as an edge of an incomplete (removed)
EDGE ANGLE, wetting angle lattice plane normal to the Burgers vector. The
The angle θ0 between the contacting surfaces plane surface parallel to the Burgers vector and
of a liquid and a solid (or another liquid), measured passing through an edge dislocation line is re-
from the side of the liquid that is in a state of ther- ferred to as a glide plane. An edge dislocation
modynamic equilibrium (see Fig.). Depending on can move along this plane, remaining parallel
the value of θ0 , there are wetting and non-wetting to itself without any diffusion of atoms (con-
355
356 EFFECTIVE ATOMIC CHARGE
ferroelectric phase transitions are studied, due to system the tensor m−1ij has three components mi
−1
the effective enhancement of the space dimension- (mi > 0 is associated with electrons, and mi < 0
ality, uctuations near the point of the phase tran- corresponds to holes). The anisotropy of the dis-
sition appear to be damped. This alters the local- persion law (1) as well as the presence of a few
ization laws of elementary excitations at point de- maxima result in a great diversity of galvanomag-
fects, at plane stacking faults, and at composition netic effects in semiconductors and metals. There
fluctuations in solid solutions. also exist other kinds of effective masses, e.g., a
cyclotron effective mass which is measured with
EFFECTIVE IMPURITY-ATOM CHARGE the help of a cyclotron resonance experiment, and
Difference between the electric charge of an the effective masses of rotons, excitons, etc.
impurity atom and an intrinsic atom in a solid. The
value of the effective impurity-atom charge arises EFFECTIVE RYDBERG (after J.R. Rydberg)
from the redistribution of the electron density of a Ionization energy of a small Coulomb impurity
chemically foreign atom when it is embedded in a center with a hydrogen-like energy spectrum in
solid. The same difference with respect to an el- a semiconductor. For a nondegenerate band with
ementary charge is called impurity valence. This an isotropic parabolic dispersion law in the vicin-
differs from the valence of an isolated atom, and ity of a minimum, the effective Rydberg Ry∗ =
from a difference in valences of impurity and in- [m∗ /(mε 2 )]Ry, where m∗ is the effective mass,
trinsic isolated atoms. In the general case, this is m is the mass of a free electron, ε is the dimen-
specified by a non-integral or fractional number sionless static dielectric constant of the crystal, Ry
(so-called fractional valence). is a Rydberg with the value Ry = me4 /(8ε02 h2 ) ≈
13.60 eV corresponding to the ionization energy
EFFECTIVE MASS
of a hydrogen atom. Values of the effective Ryd-
A value that characterizes quasi-particles with
berg in germanium and silicon are of the order
a quadratic energy dispersion law E(p) where p is
of 0.01 eV. The expression “effective Rydberg”
the quasi-momentum. Formally, it is the quantity
is also used in the broader sense to designate
m∗ present in the formula E(p) = p 2 /(2m∗ ) that
the binding energy of a system with an electro-
is called the effective mass. A differential defini-
static interaction (sometimes with a more gen-
tion is m∗ = 2 (∂ 2 E/∂k 2 )−1 , where p = k. The
eral type of interaction), in particular to describe
concept of effective mass results from solving dy-
Wannier–Mott excitons. For such excitons, Ry∗ =
namical problems (Newton equation, Schrödinger 2 )]Ry, where μ is the electron and hole
[μ/(mε∞
equation, or Euler–Lagrange equation) in which
reduced effective mass, and ε∞ is the dimension-
the process of taking into account the interaction
of a given particle with the environment is reduced less optical dielectric constant.
to the substitution of a certain effective mass for its
EFFECTS OF RETARDATION
ordinary mass. For example, the effective mass of
a polaron is related to the electron–phonon inter- Effects stemming from the finite velocity of
action. In the band theory of charge carriers in a propagation (light velocity) of an interaction im-
crystal, the concept of effective mass is introduced plemented through the electromagnetic field. Un-
near the extrema of the band energy En (p) where like the instantaneous Coulomb interaction which
n is the band number. If there is a minimum (elec- depends only on the positions of charges, the re-
trons) or a maximum (holes) at a particular point tarded interaction depends also on the distribution
p 0 , En (p), then in the vicinity of p0 = k 0 we of their velocities. It is caused by the solenoidal
have the expression portion of the electromagnetic field, and corre-
sponds to the exchange of virtual transverse pho-
2 m−1
ij (ki − ki0 )(kj − kj 0 ) tons. The effects of retardation can be described by
En (k) = En (k0 ) + , the dimensionless parameter α = τ/T , where T is
2
(1) the characteristic time of the motion of the inter-
where the second-rank tensor m−1 ij is called the acting particles, and τ is the time of the field prop-
inverse effective mass tensor. In its principal axes agating between them. For instance, if applied to
358 EINSTEIN–DE HAAS EFFECT
atomic electrons, 1/T ∼ e2 /(4πε0 a) is the Bohr group F m3m (Oh5 ); a ≈ 0.558 nm at room tem-
frequency of the electron; e is its charge; a is the perature. The existence of a metastable modifi-
radius of its orbit; τ ≈ a/c, and α = e2 /(4πε0 c) cation of einsteinium is possible with a hexago-
is the fine structure constant. In the case of a crys- nal close-packed lattice, space group P 63 /mmc
tal with excitonic excitation there is an interaction 4 ), having a two atom basis and a primitive
(D6h
between electrons that are much farther from one unit cell volume 0.08916 nm3 (at room temper-
another than in an atom, at a distance on the order ature); with T melting = 1130 K. Einsteinium is
of an exciton wavelength λ, i.e. τ ≈ λ/c. Also, the paramagnetic, with no known magnetic ordering.
time for an appreciable displacement of charge is Targets of isotope 253 Es with 20.47 day half-life
T ≈ 1/ω, where ω is the frequency of the exciton are used for synthesis of heavier transplutonium
wave. The result is α ≈ 1/n, where n is the re-
elements.
fractive index of the electromagnetic wave accom-
panying the exciton. Thus, the effects of retarda-
tion are significant as long as n (or the wave vec- EINSTEIN RELATION (A. Einstein, 1905)
tor k = nω/c) is not too large. The latter holds in An expression which relates the diffusion co-
the region of the intersection of the spectral curves ef cient Di of particles i (electrons, holes, ions,
of photons and excitons, where effects of retarda- etc.) possessing an electric charge ei and the mo-
tion are appreciable, and a mixed electromagnetic bility μi in an electric field E: ei Di = k B T μi .
wave or polariton (photoexciton) appears. The Einstein relation is valid when the Boltzmann
distribution is satisfied by the particles. In the case
EINSTEIN–DE HAAS EFFECT (A. Einstein, of electric field heating of current carriers then the
W.J. de Haas, 1915) Boltzmann distribution applies using the electron
Variation of the mechanical angular momen- (hole) temperature T ∗ , so the Einstein relation re-
tum of a magnetic material as a result of its mag- mains valid when T ∗ is substituted for T .
netization in an external magnetic field. For a mag-
net, the induced mechanical momentum is pro-
ELASTIC ENERGY, strain energy
portional (neglecting magnetic anisotropy) to the
variation of its magnetic moment, and it is di- The energy due to the elastic strain of a solid
rected along the magnetization axis. Experimen- body (see Elasticity theory). The elastic free en-
tally a twisting motion is observed for a sample ergy density F can be expressed in terms of the
freely suspended in an alternating magnetic field distortion tensor uik :
with a frequency close to the frequency of the in-
1
trinsic oscillations of the sample. The Einstein–de F= cik,lm uik ulm dV ,
Haas effect is a complement of the the Barnett ef- 2
fect (slight magnetization induced by rapid rota-
where cik,lm is the fourth-rank elastic modulus
tion), and it is an example of a magnetomechani-
tensor of the crystal,
cal phenomenon.
1 ∂ui ∂u
EINSTEINIUM, Es uik = + k ,
2 ∂xk ∂xi
Chemical element of the actinide group of
the periodic system with atomic number 99 and and ui = ui (x, t) is a displacement vector at the
atomic mass 252.0828. There are 14 isotopes point x. This free energy can also be presented
known with mass numbers from 243 to 256, none
in terms of the strain tensor σik as these quanti-
of which are stable. Outer shell electronic con-
ties are connected with each other by Hooke’s law:
figuration is 5f 11 6d 0 7s 2 . Radius of Es3+ ion is
σik = cik,lm ulm .
0.0953 nm. Oxidation state is +3, more rarely +2.
Electronegativity is 1.27.
Einsteinium is a metal characterized by high ELASTIC ENERGY, STORED
volatility, it has a face-centered cubic lattice, space See Stored elastic energy.
ELASTICITY 359
ELASTICITY ANISOTROPY
See Anisotropy of elasticity.
ELASTICITY ISOTROPY
See Isotropy of elasticity.
is similar to Young’s modulus, and characterizes deformation tensor uik (deformations are coordi-
the change of interlayer distance under the action nate-dependent), and hence the free energy of a
of an external force. In this case, the free-energy deformed solid is a functional of the deformations.
density is given by The free energy of a unit volume of the deformed
2 solid may be represented as a series in terms of
1 1 ∂u
f = K1 (∇n)2 + B , (2) powers of uik , where only quadratic terms (har-
2 2 ∂z monic approximation) of this series are employed
in linear elasticity theory:
where B ∼ 106 N/m2 , and u is the displacement
of the smectic layers. In the case of smectic C the 1 T
F = F0 (T ) + ciklm uik ulm , (1)
free energy (2) must include a term for the en- 2
ergy of the C-director elastic distortions, as well where F 0 (T ) is the free energy of the unde-
as terms to take into account the interactions of T
formed solid, and the terms ciklm are temperature-
the deformations of layers and the C-director (see dependent isothermal elastic moduli. The remain-
Wave-like modulation). In the presence of an elec- ing (anharmonic) terms of the series are used for
tric field E and a magnetic field B the energy describing effects that are left out in the linear
terms Π e and Π m are added to the free energy f : approximation (e.g., volume effects in a crystal
εa with dislocations, nonlinear effects occurring at
Π e = − (n · E)2 ,
2 the propagation of elastic waves, etc.).
χa Thermodynamic definitions of stresses and
Π m = − (n · B)2 . elastic moduli in linear elasticity theory are as fol-
2
lows:
These terms depend on the dielectric constant εa
and the magnetic susceptibility χ a (see Anisotropy ∂F
σik = ,
of liquid crystals). ∂uik T
∂2F
ELASTICITY THEORY ciklm = .
Branch of continuum mechanics concerned ∂uik ∂ulm T
with reversible strains of solids. Elasticity theory The dynamic equation of elasticity theory is of
is used in solid state physics for describing struc- the form:
tural distortions which vary only slightly over dis-
ρ∂ 2 ui ∂σ
tances on the order of interatomic spacings. This = ik + fi , (2)
theory provides the basis for a continuum theory ∂t 2 ∂xk
of dislocations, grain boundaries, cracks, pores, where ρ is the density of the material, and fi is
point defects, for physical theories of plasticity, the force acting on a unit volume of the solid. This
strength, failure and acoustical phenomena. equation presumes the applicability of Hooke’s
The basic concepts of elasticity theory are re- law with certain elastic moduli, and it involves adi-
lated to internal stresses and deformations, de- abatic elastic moduli in the case of high-frequency
scribed respectively by the stress tensor σik and (sound) vibrations. The boundary conditions for
the strain tensor uik . In linear elasticity the- Eq. (2) are established either by specified surface
ory the deformation tensor is given by uik = forces σik ni = fis (ni is the unit vector of the
(1/2)(∂ui /∂xk + ∂uk /∂xi ), where ui is a dis- outward-directed normal to the surface, fis is the
placement vector, and the stress tensor σik is re- force acting on a unit surface area of the deformed
lated to uik by Hooke’s law: σik = ciklm ulm , solid), or by specified shifts of the outer surface of
where ciklm is the elastic modulus tensor. The a solid. The applied force fi may be that of gravity
condition for applicability of linear elasticity the- ρgi (where gi is the gravitational acceleration), or
ory is that the stresses be small compared with the force arising from the difference between the
elastic moduli. Elasticity theory assumes that the local temperature and the volume-averaged tem-
equilibrium state of an elastic solid at given tem- perature of the body under nonuniform heating.
perature T is unambiguously determined by the The distribution of forces fi may be generated
362 ELASTICITY THEORY CONTACT PROBLEM
p ∈ P . For instance, the number of linearly inde- posite materials, etc. The units of elastic moduli
pendent second order elastic moduli equals 21 for are gigapascal, GPa.
the triclinic system, 13 for the monoclinic system,
and 9 for the orthorhombic system; for the tetrag- ELASTIC MODULI, COMPLEX
onal system, the number of moduli is 7 for sym- See Complex elastic moduli.
metry classes 4, 4 and 4/m, and 6 for classes 422,
42m, 4mm and 4/mmm; for the trigonal system: ELASTIC MODULI, RELAXED AND
7 for classes 3 and 3, and 6 for classes 32, 3m, UNRELAXED
3m; the number of linearly independent second- Factors of proportionality between stress and
order elastic moduli is 5 for the hexagonal sys- deformation under an infinitely slow (rapid) de-
tem and 3 for the cubic system. The number of formation process when relaxation processes in
linearly independent elastic moduli of third order the medium have sufficient (inadequate) time to
equals 56 for triclinic, 32 for monoclinic, and 20 take place. A relaxed elastic modulus is an isother-
for orthorhombic; for the tetragonal system: 16 for mal elastic modulus, used if the relaxation of the
classes 4, 4, 4/m and 12 for classes 422, 42m, medium has time to take place, and, therefore,
equilibrium with respect to all degrees of free-
4mm, 4/mmm; for the trigonal system: 20 for
dom is established. An unrelaxed elastic modu-
classes 3, 3 and 14 for classes 32, 3m, 3m; for the
lus is an adiabatic elastic modulus, used if the
hexagonal system: 12 for classes 6, 6, 6/m and 10
system undergoes no irreversible processes un-
for classes 622, 6m2, 6mm, 6/mmm; for the cu-
der deformation. To describe the relation be-
bic system: 8 for classes 23, m3̄ and 6 for classes
tween stress and deformation in partial relax-
432, 43m, m3m. The components of higher-order
ation, the rheological equation of the medium
elastic modulus tensors of crystals are determined
is used. For a standard linear material (Zener
in a similar manner.
model), the rheological equation is of the form
If the deformation occurs so slowly that the
σ + τε (dσ/dt) = M R (ε + τσ (dε/dt)), where σ
material temperature remains unchanged, the mod-
and ε are the rates of variation of stress and
uli are called isothermal elastic moduli (see Elastic
strain, and τσ and τε are the relaxation times at
moduli, relaxed), and if it is so rapid that no heat
constant stress and constant strain, respectively.
exchanges between nearby regions of the body When we expand stress and deformation into
then the moduli are called adiabatic elastic mod- Fourier integrals, we obtain the following equation
uli. Isothermal λ(i) , μ(i) and adiabatic λ(a) , μ(a) for the ratio of their complex Fourier amplitudes
Lamé coefficients for an isotropic body are re- (this ratio is the complex elastic modulus M ∗ ):
lated as follows: λ(i) < λ(a) , μ(i) = μ(a) . If the M ∗ = σ (ω)/ε(ω) = M R (1 + iωτσ )/(1 + iωτε ).
rate of deformation is comparable to that of re- For ω = 0 we have M R = σ (0)/ε(0), where M R
laxation processes (see Stress relaxation) that take is the relaxed elastic modulus, and for ω = ∞ we
place in the body, the relation connecting stress have M UR = σ (∞)/ε(∞) = M R τσ /τε , where
and strains at a given instant of time is replaced M UR is the unrelaxed elastic modulus. Typically,
by the rheological equation whereby stresses and M R < M UR .
strains are related not only at a given time, but also
at prior times. Elastic moduli of a polycrystal de- ELASTIC MODULUS TENSOR
pend on the location in the medium. However, if Tensor of fourth rank cij kl , which relates the
the characteristic deformation length significantly stress tensor σij and the strain tensor uki in
exceeds the sizes of crystallites, then it is conve- the linear Hooke’s law σij = cij kl ukl . The elastic
nient to treat the polycrystal as a homogeneous modulus tensor possesses the following symmetry
medium with constant effective elastic moduli, de- properties with respect to the permutation of in-
rived from elastic moduli of individual crystallites dices: cij kl = cj ikl = cij lk = cklij . Values of ten-
by averaging over various orientations (e.g., Voigt sor components for crystals often lie in the range
averaging, Reuss–Voigt–Hill approximation). The 1011 –1013 dyn/cm2 .
above effective medium method is used for calcu- The number of linearly independent compo-
lating effective characteristics of mixtures, com- nents of this tensor is determined by the crystal
ELASTIC SCATTERING 365
system and cannot exceed 21. An isotropic med- the coefficients of the elastic rigidity tensor, and
ium (see Isotropy) is described by two elastic mod- of the tensor of elastic moduli cij kl with respect
uli, a cubic crystal by three, and so on. Sometimes, to permutations of the indices i, j, k, l, and hence
the Voigt notation is used, where the first pair of it has the same number of independent compo-
tensor indices is replaced with one index accord- nents. It is related to the change
cij kl of the
ing to the scheme: 11 → 1; 22 → 2; 33 → 3; cij kl moduli. For the case of N d uniformly dis-
23, 32 → 4; 31, 13 → 5; 12, 21 → 6. A similar tributed point defects, with a small relative con-
replacement is also done for the second pair; as a centration c = N d /N (c 1), we have
cij kl ≈
result, the matrix cik , i, k = 1, 2, 3, . . . , 6, is intro-
Vc−1 cαij kl , where V c is the primitive unit cell vol-
duced instead of the fourth-rank tensor; this matrix
ume, and N is the number of such cells in the
has the same number of independent elements, as
crystal. The components of α̂ are of the order of
does the tensor ciklm .
1–10 eV, and higher.
The tensor sij kl , that is reciprocal to cij kl , is
called the compliance tensor, uij = sij kl σkl . The
concept of elastic modulus tensor is important not ELASTIC POTENTIAL OF A DEFECT
only in static elasticity theory, but also in dynamic A phenomenological parameter that character-
elasticity theory, since the elastic modulus tensor izes the displacement field in a medium containing
enters the wave equation to describe the propaga- a point defect. In an unbounded isotropic medium
tion of elastic waves in crystals: the elastic potential determines the displacement
2 u(r) produced by a single point defect at the point
∂ 2 ui ∂ ui
ρ= = cij kl , characterized by radius vector r, where u(r) =
∂t 2 ∂xj ∂xk
−A∇(1/r), and A is called the elastic potential of
where ρ is the crystal density. When the non- the defect. This elastic potential is analogous to the
linear Hooke’s law applies (σij = cij kl ukl + value of an electrostatic point charge that produces
cij klqr ukl uqr + · · ·), the nonlinear acoustic equa- around itself an electric field E = −∇V , where
tions (see Nonlinear acoustic effects) involve a ten- V (r) = k/r is the Coulomb potential. In contrast
sor of sixth rank cij klqr , the so-called third-order to the electrostatic charge case, A is determined by
elastic modulus tensor. the elastic properties of the medium, and the force
constants of the defect.
ELASTIC POLARIZABILITY OF DEFECTS
The ability of defects to produce additional
mechanical stresses
σ̂ d (ε̂) in a crystal which ELASTIC RECOVERY
has been exposed to external mechanical stresses Spontaneous decrease of elastic macrostresses
and related strains ε̂. The initiation of
σ̂ d (ε̂) in that occurs in a solid on the removal of an external
the presence of defects is related to the fact that load. After being exposed to elastic deformation
the magnitudes of interatomic interaction forces (see Elasticity), the material completely resumes
in a crystal undergo additional changes when the its original shape during the course of elastic re-
ε̂ field is established in this crystal. The induced covery (e.g., a spring returns to its original shape).
stresses
σ̂ d (ε̂) die out as the field of external de- If the applied load exceeds the yield limit, then only
formations is turned off. In the field of small crys- partial recovery of the original shape takes place,
tal deformations (smooth field of atomic displace- the degree of recovery being lower, the higher the
ments), the ε̂-dependence of
σ̂ d (ε̂) for a single degree of plastic deformation. In such cases, elas-
defect is linear: tic recovery can be observed by cutting the de-
σ̂ijd (ε̂) ≈ V −1 αij kl εkl , |εkl | 1, formed solid into parts, or by partial dissolution
of the solid (see Recovery).
where V is the crystal volume, and indices i, j, k, l
represent the coordinate axes x, y, z. The fourth-
rank elastic polarizability tensor of the defects α̂ ELASTIC SCATTERING
exhibits permutation symmetry identical to that of See Elastic scattering of radiation.
366 ELASTIC SCATTERING OF RADIATION
to reduce its length L, and the decelerating forces electret along its center (neutral) line, two electrets
(similar to dry friction forces) which retard the mo- are formed, as is the case of magnets. The name
tion of the twin boundary. “electret” originally appeared in formal analogy
The equilibrium value of L monotonically in- with the term “magnet”. One should differenti-
creases with increased external load. In the case ate electrets from ferroelectrics which are electric
of a crystal of finite dimensions under consider- analogues of ferromagnets. In ferroelectrics the
able load, the twin becomes unstable and “slips macroscopic electric moment is a result of a struc-
through” the crystal, extending all the way through tural phase transition accompanied by the forma-
it, forming a plane-parallel residual twin inter- tion of regions of spontaneous polarization (see
layer. If the concentrated applied load provides the Ferroelectricity), namely domains, whose orienta-
twin with some stability, then the twin may remain tion induced by the electric field results in the
in the crystal after the load has been completely residual polarization.
removed (wedged twin). This is related to the fact The electret state of an insulator is determined
that the total surface tension force, which acts as by the long-term persistence of an electric dipole
a buoyant one, is proportional to L−1/2 , while the moment per unit volume in it. Electrets bring about
retarding force is L-independent. The presence of
an electric field in their surroundings. A dipole
the dry friction forces is the reason for a hysteretic
moment can be induced by positioning an insula-
dependence of L on the external load (see Hystere-
tor which contains polar molecules (i.e. molecules
sis). The temperature dependences of the friction
with intrinsic dipole moments) in an electric field
and the surface tension γ determine the conditions
at an elevated temperature, thereby orienting the
under which elastic twinning and allied phenom-
initially randomly directed molecules, with subse-
ena are realized. The dynamic motion of elastic
twins is accompanied by acoustic emission. quent cooling. The partially oriented polar mole-
The quantitative description of elastic twin- cules become quenched (frozen in position); then
ning, which permits detailed correlation with ex- the external field is switched off. The electron–
perimental results, is given in the context of dislo- hole mechanism for the appearance of the electret
cation theory, the latter being extended to related state is the spatial redistribution of charge carriers
phenomena: thermoelastic martensitic transforma- that leads to their separation. Electrons and holes
tion, superelasticity effects, behavior of domains in captured at a deep local electron level aligned in
ferroelastics. accordance with the direction of the polarizing
electric field, i.e. nonuniformly over the sample,
ELASTOMERS with sufficient cooling conserve their induced di-
Polymers and polymer-based materials which pole moment for a long time. In crystals with a lat-
possess highly-elastic properties over a wide tem- tice possessing a spontaneous dipole moment, the
perature range of their application. Typical elas- presence of free charges results in electric charge
tomers are rubbers, or materials with rubber-like screening. The lack of free electrons (either ex-
properties. See also Polymeric materials. ternal or internal ones) or other mechanisms of
shielding (e.g., the formation of oppositely di-
ELASTOOPTIC TENSOR
rected domains) leads to the electret state in such
See Photoelastic tensor.
crystals. In the case of orientational polarization as
ELECTRETS well as at the displacement of charge carriers from
Amorphous and crystalline insulators with per- inside the sample to the surface, the direction of
manent electric polarization P ; electrical analogue the induced field is opposite to the inducing field
of a permanent magnet. In other words, the electric direction (electrets with heterocharge). In the case
polarization (so-called residual polarization; see of the external injection of charge carriers, both di-
Polarization of insulator) remains a long time af- rections coincide (electrets with homocharge). The
ter removing the external action (electric field, me- electret can be either electrically neutral overall, or
chanical stress, and so on) that produced it. Elec- it can have a certain charge. Depending on their
trets in the polarized state possess all the prop- method of formation, such as heating and cool-
erties of pyroelectric materials. When cutting an ing, strong electric field, illumination, nuclear ra-
368 ELECTRIC-ACOUSTIC INTERACTION
diation, etc., one can distinguish thermoelectrets, sponse to the application of an electric field; a co-
electroelectrets, photoelectrets and radioelectrets. efficient that specifies this property. The electric
A number of methods are known for the prepa- conduction current is a result of the current car-
ration of electrets. Thermoelectrets are obtained rier drift in an electric field. In an alternating elec-
by cooling a melt of dielectric (such as wax, paraf- tric field, a displacement current which is not con-
fin wax, resins) in a strong electric field. This nected with the carrier drift arises. In weak elec-
was the method used by the Japanese physicist tric fields of strength E, current density J and E
M. Eguchi who had made the first electret (1929). are related linearly in accordance with Ohm’s law:
Highly mobile polar molecules of the melt quite Ji = σik E k , where σik is the (electrical) conduc-
easily acquire a predominant alignment in an elec- tivity tensor; i, k = x, y, z are Cartesian coordi-
tric field. This orientation sometimes remains sta- nates. Ohm’s law can be rewritten in the inverted
ble for several years after the solidification, i.e. form Ei = ρik Jk where ρik is the resistivity ten-
the electret state is metastable but quite persistent. sor. In a magnetic field B we have σik (B) =
Residual polarization of the lattice can appear as σki (−B) by the Onsager principle of the sym-
a result of the orientation in the electric field of metry of kinetic coefficients (1931). More gener-
so-called quasi-dipoles: two vacancies of different ally, in a magnetic field the tensor σik is divisible
sign, or an impurity ion with a vacancy of oppo- into symmetric and asymmetric parts Sik and Aik ,
site sign (e.g., in alkali-halide crystals, in rutile). respectively: σik = Sik + Aik , where Sik (B) =
When evolving toward a thermodynamic equilib- Sik (−B), Aik (−B) = −Aik (B). The tensor Sik
rium state, the electret gradually loses its induced describes the effect of magnetoresistance, Aik is
dipole moment. This process is accompanied by a related to the Hall effect (see Galvanomagnetic ef-
depolarization current (or, at heating, by a ther- fects). Crystal symmetry reduces the number of
modepolarization current). independent tensor components. In particular, in
a cubic crystal with B = 0 the conductivity is a
Electret properties are well pronounced in wax,
scalar, i.e. isotropic: J = σ E, E = ρJ where σ is
paraffin wax, glass ceramics, corundum Al2 O3 ,
the electrical conductivity, ρ = 1/σ is the specific
rutile TiO2 , and so on. In good electrets, such as
resistance (resistivity).
carnauba wax, the polarization persists for sev-
Electronic conductivity by the transport of elec-
eral decades. An attainable potential difference be-
trons (or holes) has been explained in terms of
tween opposite surfaces of polymer films a few
band theory. In an electric field the electrons are
microns thick is a few hundred volts. Electrets
accelerated, thereby passing to higher energy lev-
are applied as membranes in microphones, tele-
els in the band. According to the Pauli principle,
phones, vibration and piezoelectric transducers,
this is possible only if the band is not completely
and in the area of electric photography (photoelec-
filled with electrons, as is the case in metals. In
trets).
metals free current carriers exist at any tempera-
ELECTRIC-ACOUSTIC INTERACTION ture, and their concentration does not depend on
See Acousto-electronic interaction. the temperature. If the upper filled (valence) band
(at temperature T → 0 K) has no unoccupied lev-
ELECTRICAL CONDUCTANCE els then the crystal is a semiconductor or an insu-
Reciprocal G = 1/R of the electrical resis- lator. An equilibrium conduction in such a crystal
tance of a material. Sometimes the dimensionless arises through the thermal excitation of electrons,
variable G/G0 is called the conductance, where e.g., from the valence band to the conduction
R0 = h/(4e2 ) is the quantum of resistance, and band, in this case σ = eμn n + eμp p where e is
G0 = 1/R0 . the electron charge, μn and μp are the mobilities,
and n and p are the concentrations of electrons
ELECTRICAL CONDUCTIVITY, electric and holes in the conduction and valence bands, re-
conduction, conductivity spectively. In an intrinsic semiconductor (n = p),
Measure of the ability a material to allow the the conduction results from the intrinsic conduc-
passage of (to transmit) an electric current in re- tivity. In doped semiconductors the excitation of
ELECTRICAL PROBES 369
electrons and holes to bands from local electronic on them. At the same time, the degree of filling
levels of defects takes place and brings about an the levels of extended defects remains low due
impurity conductivity. If the excitation is induced to the Coulomb repulsion of charges localized at
by an external action, the resulting conduction them. When the energy bonding the carrier with
is called nonequilibrium conduction. Depending the defect, E
k B T , does not depend on the tem-
on the kind of action, one distinguishes photo-, perature T or the density of levels localized on
X-ray, gamma conduction, and so on. Electron the defect, then the critical density is given by
(hole) injection also induces a nonequilibrium ndc = [3ne2 /(εE)] ln[0.163εE/(e2 n1/3 )] where
conduction. For μn n
μp p the conduction is n is the current carrier concentration, e is the
called (n-type) electronic conductivity in contrast electron charge, and ε is the dielectric constant.
to (p-type) hole conductivity at μn n μp p. If Typical values of ndc are 1011 –1012 m−2 . The
the carriers of one sign contribute to the conduc- charge per unit length of a dislocation is q L =
tion much more than the carriers of the opposite en/nd for nd ndc (nd is the dislocation den-
sign, the conduction is predominately monopolar sity per unit area), and q L = en/ndc for nd ndc .
conduction; in the case when both contributions Complete localization of carriers at grain bound-
are comparable then one speaks about bipolar con- aries corresponds to a specific defect size l c =
duction. (6εE/πe2 n)1/2 . Typical values for l c are 1 to
In highly-doped semiconductors (see Doping) 10 μm. The charge per unit boundary area q A =
and amorphous semiconductors, a type of hopping 0.577enl c at l
l c (here l is the grain size), and
conductivity might predominate. In cases when the q A = enl/3 at l l c . In the case of nd = ndc or
percolation properties of the conductive medium l = l c in n-type semiconductors, then all the elec-
play the dominant role then the electrical con- trons in the conduction band are localized at de-
ductivity is described in terms of the percolation fects, and are capable of transferring to defects in
mechanism. the valence band as their density increases; this
If the period of an alternating electric field is leads to an inversion of the conductivity type. In
of the order or less than the quasi-momentum re- the case of localization of all the carriers at dislo-
laxation time, then the current transport is charac- cations and grain boundaries, the Mott law of con-
terized by a high-frequency conductance that has a ductivity is valid. The electrostatic field around the
complex value and depends on the frequency (see charged dislocation line causes the band bending
High-frequency conductance). that makes possible the inversion of the conduc-
tion type in the range adjacent to the dislocation
ELECTRICAL CONDUCTIVITY, SURFACE axis. Thus, a semiconductor junction can arise. In
See Surface electrical conductivity. the case of amphoteric extended defects (see Am-
photeric center), a pinning of the Fermi level be-
ELECTRICAL CONDUCTORS tween the donor and acceptor levels is possible
See Conductors. (e.g., for dislocations in p-type silicon and germa-
nium).
ELECTRICALLY ACTIVE EXTENDED DEFECTS
Dislocations and grain boundaries which pos- ELECTRICAL PROBES
sess local electronic levels. Such defects can have Contacts attached to a sample for determining
a dominant effect on the concentration of current the potentials of various points at its surface. In
carriers, and the electrical properties of a material. many cases the electrical probes are prepared in
In a monopolar semiconductor with shallow impu- the form of thin pointed wires made of a hard con-
rity levels of point defects and deep levels at dis- ducting material. Thin electrical probes branch-
locations and grain boundaries, the number of lo- ing from the surface, manufactured by a common
calized current carriers increases as the density nd technological process and made of the same semi-
(per unit area) of extended defects increases, and conductor material as the sample, are also used.
when a particular (critical) density ndc is reached The probes terminate at large contact areas to
all the existing current carriers become localized which the wires of the control circuit are attached.
370 ELECTRIC BREAKDOWN
Samples of these probes are made in the form owing to the polarization (displacement) of bound
of dumb-bells, and in the form of a clover leaf. charges, i.e. electrons and ions.
The input resistances of the devices connected to In any material (metal, semimetal, semicon-
probes are chosen so that the current uptake by ductor, electrolyte, and gas-discharge plasma) with
them does not distort the electric field in the sam- a finite concentration of free current carriers, the
ple, and there is a negligibly small potential drop electric charge screening is determined by the re-
at the transition resistance between the sample and distribution of the current carrier density in the
the probe. Probe methods used for specific resis- field of a given charge. In the simplest approxi-
tance measurements of semiconductors include: mation of a constant (independent of energy and
the two-probe method, the four-probe method, and momentum) linear response of the medium this is
the van der Pauw method (1958). In these methods described by the (Yukawa) exponential law as
the specific resistance value is determined from e2 r
V (r) = exp − , (1)
the potential drop between two probes at a known r re
current. The four-probe method is applicable for
where V (r), called the Yukawa potential, is the po-
determining the specific resistance of semiconduc-
tential energy of two charges of value e separated
tor bars; the van der Pauw method is useful for
by a distance r; here r e is the screening radius
plates of arbitrary shape. If a contact between the
that is equal to either the Debye–Hückel shield-
probe and the sample is undesirable, one can use a ing radius (see Debye–Hückel theory) in a clas-
vibrating capacitive probe which utilizes an elec- sical (nondegenerate) plasma of a semiconductor
trode vibrating above some point of the sample or gas discharge and in weak solutions of elec-
surface. If the potentials of this point and the probe trolytes, or it is the Thomas–Fermi shielding ra-
are different, an alternating current runs in the cir- dius in a degenerate electron plasma (Fermi gas or
cuit between the point and the probe. The lack of Fermi liquid) in metals, doped semiconductors, and
current flow means that the potential of the point semimetals.
is equal to the potential of the probe. A vibrating A more detailed quantum-mechanical calcula-
capacitive current is used when measuring the con- tion demonstrates that the static linear response
tact potential difference. In the case of an alternat- (polarization operator) of the ideal Fermi gas de-
ing electric field in a sample (e.g., the Gunn effect), pends on the transfer of momentum q, and has a
immobile capacitive probes are employed. singularity at the point q = 2k F (k F is the Fermi
momentum, setting = 1) which corresponds to
ELECTRIC BREAKDOWN in a solid a sharp discontinuity of the fermion distribution
Sharp increase of electric current through a function at the Fermi surface (see Fermi–Dirac sta-
non-conducting (insulator) or weakly conduct- tistics), and depends on the type and dimensional-
ing (semiconducting) crystal in a strong elec- ity of the electronic spectrum. So for an isotropic
tric field exceeding a certain critical field E cr three-dimensional (3D) system with a quadratic
(see Breakdown of solids, Impurity breakdown, dispersion law and an isotropic (spherical) Fermi
Streamer breakdown). This disruptive process in- surface, the static polarization operator Π in mo-
volves complete failure under electrostatic stress, mentum space takes the form:
and can do irreversible damage. 4k 2 − q 2 q + 2kF
Π3D (q) = N3D (0) 1 + F ln ,
4kF q q − 2kF
ELECTRIC CHARGE SCREENING, electric (2)
charge shielding where N 3D (0) is the density of electron states
Effective reduction of the Coulomb interaction at the Fermi level. As follows from Eq. (1),
of charges in a medium as a result of its polar- dΠ 3D (q)/dq at q = 2k F has a logarithmic singu-
ization. In insulators the electric charge screening larity, which leads to the fact that the Fourier com-
is reduced by the factor ε0 /ε(here ε is the sta- ponent of the shielded potential of the Coulomb
tic dielectric constant of the material, and ε0 is interaction
its free space value) in the electrostatic interaction 4πe2
for distances exceeding the lattice constant value V (q) = 2 , (3)
q + 4πe2 Π3D (q)
ELECTRIC CURRENT 371
holes), but can also involve jumps of current carri- have the same parity, and due to the polarity of the
ers between localized states (e.g., electronic hop- D vector operator we have (D β )ij = 0 for any i
ping conductivity and ionic conductivity). In this and j . As a result, linear electric effects do not oc-
case the current carrier spends most of its time as cur. If the states ψi of the main term are pure spin
a stationary particle localized at a defect, but from states then only diagonal matrix elements of oper-
time to time this carrier makes a jump to a simi- ator D should exist, and they are all equal to each
lar neighboring site. In an electric field the jump other so electric effects are still lacking. Therefore,
probability (frequency) along the field direction is for the appearance of an electric effect it is neces-
higher than in the opposite direction; and this de- sary to have an admixture of orbital states in the
termines the magnitude of the current. main term. In other words, two circumstances play
Quantitatively the current is characterized by a significant role in the appearance of the electric
a current density j , i.e. by the charge transported effects: the spin–orbit interaction, and an interac-
per unit time unit through a unit area normal to tion with the crystal resulting in the contribution
the current flow. The direction of positive charge of an admixture of orbital states and states with
flow is accepted to be the direction of the current. different parity.
The value of the current I = J S (S is the traverse The matrix elements (D β )ij are related to the
area) is measured in amperes, 1 A = 1 C/s, and parameters of the of the electric–spin interaction
the current density J is measured in A/m2 . The Hamiltonian, and they differ (in contrast with the
current density includes an applied electric field spin Hamiltonian method) in their description of
term J drift = σ E, where σ = enμ is the specific electric effects in terms of the perturbation matrix
conductivity (reciprocal to the resistivity or spe- method.
cific resistance); e is the electron charge, the con-
centration of current carriers n is averaged over ELECTRIC DOUBLE LAYER
the length dx, μ is the mobility; and E(x) is the Electric charges of opposite signs (ions, polar
electric field strength. The action of concentration molecules), distributed along the boundary at the
gradients, mobility, and temperature leads to a sec- interface of two phases. An electric double layer
ond term J diff . These terms are referred to as drift appears at the surface of a solid (electrode), sub-
current and diffusion current, respectively. merged in a solution of electrolyte. We can distin-
guish a dense part of this double layer [Helmholtz
ELECTRIC DIPOLE MOMENT OF A layer (G. Helmholtz, 1879)] located directly at
PARAMAGNETIC CENTER the surface of the electrode and a diffusion part
Matrix element of the electric dipole moment [Gouy–Chapman layer (C. Gouy, 1910; D. Chap-
operator D evaluated using wave functions ψi of man, 1913)] located within the solution behind
the main electronic term, as follows: the Helmholtz layer (see also Semiconductor–
electrolyte boundary). The effective distance be-
(Dβ )ij = Ψi∗ D β Ψj dτ. (1) tween the layers is the order of a molecular di-
mension so this double layer has a high electrical
Electric dipole moments of paramagnetic centers capacitance (about 10 to 100 μF/cm2 ), it is elec-
are used for describing electric eld effects in elec- trically neutral, and within the layers the electric
tron paramagnetic resonance since the initial op- field intensity may reach very high values. The
erator for the interaction of paramagnetic centers formation of an electric double layer influences
with the electric field E has the form (E · D). electrokinetic phenomena, as well as the course
Taking into account the symmetry of the local- of electrochemical reactions (e.g., in the chemi-
ized paramagnetic center and its behavior under cal sources of current, in electrolysis, etc.). For in-
time inversion makes it possible to identify ma- vestigations of the double layer structure different
trix elements which vanish, and to establish inter- methods are used: measuring the capacitance with
relationships between non-zero matrix elements an alternating current, adsorption methods includ-
without actually calculating them. In particular, ing radioisotope investigations, optical methods
for paramagnetic centers with a center of symme- (e.g., electrore ection, Raman scattering of light,
try the wave functions of the main electronic term etc.).
ELECTRIC FIELD EFFECTS 373
energy or the intracrystalline electric field energy. symmetry. (An exception is when the impurity ion
In this case, in addition to first-order perturbation nucleus belongs to a paramagnetic ion at an inver-
theory terms (linear in Ei ), nonlinear effects are sion center.)
also possible: a quadratic effect, and even a cu-
bic electric field effect, despite the linearity in the
field of the original operator WE . In contrast to ELECTRIC FIELD GRADIENT
this situation, there are possible cases when the A tensor of the second rank with components
original operator already contains terms ∼Ei Ej qj k ≡ ∂Ej /∂xk characterizing the spatial nonuni-
(or Ei Ej Ek ), which appear in EPR spectra, and formity of an electric field, E. The electric field
are specific for strongly polarized crystals such as vector can be written as the gradient of a scalar
ferroelectrics. As a result, the observation of elec- potential E = −∇V , or in component form Ej =
tric field effects for them is available even when −∂V /∂xj . In the absence of charges Laplace
the site of the paramagnetic center is a center of equation holds ∇ 2 V = 0, and there is a system
inversion. of principal axes of a tensor where the qj k ten-
The most important applications of electric sor is diagonal and traceless, i.e. qxx + qyy +
field effects are the following: unambiguous deter- qzz = 0. If the choice of axes is made so that
mination of the symmetry group of paramagnetic |qxx | |qyy | |qzz | then we can define qzz ≡ q
centers, and uncovering the nature of defects and and (qxx − qyy )/qzz ≡ η, where the quantity η
the character of their distribution over a crystal.
is called the asymmetry parameter of the electric
Electric field effects in nuclear magnetic reso-
field gradient. For axial symmetry qxx = qyy and
nance and electron–nuclear double resonance are
η = 0.
similar to those of electron paramagnetic reso-
nance, consistent with the use of the same phe-
nomenological description for EPR and NMR ELECTRIC FIELD, LOCAL
spectra: the nuclear Zeeman energy corresponds See Local electric eld.
to the electron Zeeman energy, and the quadrupole
interaction energy corresponds to the intracrys-
talline electric field. The energy operator for the ELECTRIC FIELD MODULATION METHOD
interaction of the nucleus with the electric field One of the methods employed to observe elec-
differs from the electronic operator only by the in- tric e ld effects in electron paramagnetic reso-
terchange S ↔ I where S and I are the electron nance (EPR) spectroscopy. It is often used in cases
and nuclear spins, respectively. The most specific when electric effects are small, or other special
features of the effects are the shift (splitting) of conditions arise to hinder the direct observation
lines in the NMR and ENDOR spectra under the of the shift of EPR spectral lines in an electric
action of a constant electric field, and the quan- field. The method consists in comparing EPR lines
tum transitions between nuclear magnetic levels that are recorded with the help of two different
under the action of an alternating electric field. To techniques: an ordinary technique (using magnetic
observe the linear electric field effect, the site of field modulation) and a special technique (using
the nucleus should not be a center of inversion. In electric field modulation). The comparison pro-
the presence of inversion-nonequivalent positions,
vides the shift of lines due to the applied electric
the shift of lines of the NMR (ENDOR) spec-
field.
trum appears as a splitting. The observation of the
electric field effect in ENDOR provides new ap-
proaches for studying paramagnetic centers. Thus, ELECTRIC FIELD PENETRATION DEPTH
ENDOR expands significantly the class of objects See Penetration depth of electric eld.
that might be investigated, since nuclei of para-
magnetic centers are generally situated at sites that
are not centers of inversion; this is true whether ELECTRIC FIELD, QUANTIZED
or not the paramagnetic center possesses inversion See Quantizing electric eld.
ELECTROCAPILLARY PHENOMENA 375
ELECTRIC FLUX DENSITY, electric between the first and the second pulses. Hence at
displacement t = 2τ the averaged electric field produces the sig-
A vector D which characterizes the electric nal of a two-pulse echo. The vibrational phases of
field, and equals a sum of electric field strength E a third pulse applied at moment T become coin-
(the main field characteristic) and dielectric po- cided at T + τ , and the signal of a three-pulse
larization P (see Polarization of insulator). In the echo appears. Electroacoustic and other nonlinear-
CGS system of units D = E + 4πP = εE; in SI ities of different orders take part in the parametric
units D = ε0 E + P = ε0 (1 + χ e )E = εE, where processes of the birth of reflected vibrations oc-
χ e = P /(ε0 E) is the electric susceptibility, ε0 is curring in the dielectric material; in piezoelectric
the dielectric constant of free space with the value powders where the echo appears at the acoustic
ε0 = 1/μ0 c2 = 8.8542·10−12 F/m, and μ0 is the resonance of powder particles, the main contri-
magnetic permeability of vacuum. bution is often provided by elastic nonlinearities.
The relation between D and E is called a mate- The relaxation time of the two-pulse echo T2 is in-
rial equation. In a static electric field ε is a tensor versely proportional to the absorption coef cient
for anisotropic media, in particular, crystals. For of the acoustical vibrations. It has been found that
an alternating field in a medium with space and in piezoelectric crystals and powders the three-
time dispersion, the relation between D and E is pulse echo might be observed when applying the
not local, and ε is a function of the frequency ω third pulse after attenuation of the sound induced
and wave vector k. by the first and the second pulses, i.e. there ex-
ists a long-term persistent memory with a storage
ELECTRIC PULSE MACHINING time of a few months. The long-term storage is due
One of the technological methods of mater- to nonuniform distribution of electric charge and
ial processing based on using an electric pulse residual strain in the bulk crystal; the latter takes
discharge between the processed item (as a rule, place under the action of a resulting stress induced
anode) and the other electrode (i.e. instrument, by the nonlinear effects from the pair of recording
cathode) separated from it by a high electric po- pulses. In powders, a contribution to a record can
tential. The electric discharge is accompanied by also be provided by particle orientation processes
heat generation and a directed explosion which stimulated by ponderomotive forces.
produces surface erosion of the material. Electric The electroacoustic echo is used for studying
pulse machining is employed mainly for electri- phase transitions and sound attenuation in solids.
cally conducting materials (for cutting, hardening The method may be promising for storage and cor-
the instrument, alloying the surface, and so on). relation processing of radio-frequency signals.
value Uc = 5–10 V. They are seen under a micro- field, e.g., ZnS:Cu, Cl (Destriau effect); thin-film
scope as alternating dark and light bands caused structures as metal–semiconductor–metal, metal–
by the periodic variation of the orientations of liq- semiconductor–insulator–metal, and metal–insu-
uid crystal molecules, and hence variations in the lator–semiconductor–insulator–metal, where the
distribution of liquid crystal optic axes (director) n-layer of high-ohmic (105 –1012 ·cm) semi-
(see Fig.). For a further rise in the applied poten- conductor or semi-insulator contains luminescent
tial the motion in the liquid crystal takes on a tur- centers, e.g., ZnS:Mn, CdF2 :Eu, SrS:Ce, Cu, Br,
bulent character, with a sharp increase of the light etc.; moving high-field domains, acoustic-electric
scattering at spatial distortions of the optical axis. domains, Gunn domains (see Gunn effect) etc.
This effect, called dynamic light scattering, is uti- (domain electroluminescence). Depending on the
lized in the production of liquid-crystal displays. sample design, prebreakdown electroluminescence
can be excited by either a constant or an alternat-
ELECTROLUMINESCENCE ing field.
Luminescence excited by an electric field capa-
Impact electroluminescence is the most wide-
ble of changing the potential and/or kinetic energy
spread and important type of prebreakdown elec-
of electrons and holes. The mechanism of exci-
troluminescence. It includes the following main
tation is divided into injection and prebreakdown
types of electroluminescence. stages (see the figure):
Injection electroluminescence arises when ap- (a) either the injection of free carriers into strong
plying the electric potential to an semiconduc- fields from electrodes (1, 1 ) or surface
tor junction, Schottky barrier, or metal–insulator–
states (2), or their generation by the impact or
semiconductor structure which lowers the poten-
tunnel ionization of deep donors, traps, and so
tial barrier and stimulates the injection of minority
on (3, 3 );
carriers, whose recombination with majority carri-
(b) acceleration of carriers. Effective electrolumi-
ers leads to the emission of photons. Electrolumi-
nescence is possible if every electron is capa-
nescence was first observed by O.V. Lossev (1923)
ble of being accelerated to the required energy,
in crystals of SiC (see Lossev effect). More re-
cently the effect has been observed in many semi- i.e. its energy increment during a mean mean
conductors (Si, Ge, AIII BV , AII BVI , and so on). free path exceeds losses due to the scattering
Its specific features are the following: low excit- on phonons;
ing electric potential 1.5 to 5 V; proportionality (c) excitation (4) or ionization (4 ) of lumines-
of the radiation intensity to between the first and cence centers directly at an inelastic collision
second power of the current over a broad region; or through some intermediate stages, e.g., the
internal quantum yield close to unity; fast response impact ionization of the lattice atom (5) with
(<10−7 s). Injection electroluminescence is the subsequent trapping of the hole onto the cen-
basis for the action of light emitting diodes and ter (6);
semiconductor lasers. (d) radiation (7, 7 , 8). Particularly effective are
Prebreakdown electroluminescence arises in intra-center radiative transitions (7 ), e.g., in
strong electric fields (>105 V/cm) as a result transition-metal and rare-earth ions (Mn2+ ,
of: (a) impact ionization (excitations) of lattice Tb3+ , and some others) since in a strong elec-
or impurity atoms by the free carriers, in most tric field the probability of radiative recombi-
cases by electrons which have obtained suffi- nation is sharply reduced due to forcing the
cient kinetic energy from the field (impact elec- electrons and holes to separate at opposite
troluminescence); (b) eld ionization or thermal sample boundaries.
field ionization (tunneling electroluminescence).
It is observed in inversely shifted p–n junc- The strong field can cause: an increase of the
tions and Schottky barriers, capacitors filled by probability of radiative dipole-forbidden transi-
crystal-phosphor powder dispersed in insulating tions, a change of the band shape and structure
powder and subjected to an alternating electric due to the Franz–Keldysh effect and Stark effect,
ELECTROMAGNETIC GENERATION OF SOUND 379
ELECTROLYTE, SOLID
See Solid electrolytes.
ELECTROLYTIC POLISHING
Phenomenon of smoothing an anode surface
during the electrolysis of electrically conducting
media which causes them to shine. The elec-
trolytic polishing process provides a luster to the
surface without deforming the adjacent layer be-
low; in addition it reduces the friction coefficient
and the electronic emission, increases the corro-
sion resistance and, consequently, improves sev-
eral important working characteristics of hard-
ware. Electrolytic polishing is also widely utilized
in structure studies (see also Polishing).
ELECTROMAGNETIC-ACOUSTIC EFFECT
See Acousto-electromagnetic effect.
ELECTROMAGNETIC GENERATION OF
SOUND
Exciting acoustic vibrations by electromag-
netic (EM) waves incident on a metal surface. The
EM generation of sound in metals involves forces
exerted on a crystal lattice by conduction elec-
trons. Along with the direct action of the electro-
magnetic field on lattice ions, an essential con-
tribution comes from the action on the lattice
Scheme of the main electron transitions associated with of forces exerted by electrons whose equilibrium
impact electroluminescence in a reversed biased metal– has been disturbed by the electric field E(r, t)
semiconductor junction (barrier). of the EM wave. A complete set of equations
that describes the EM generation of sound com-
prises Maxwell’s equations, the kinetic equation
the appearance of broad-band radiation caused by for the electron distribution function, and the elas-
intraband transitions of hot electrons (8). The in- ticity equation. The latter has the form of the equa-
tensity of impact electroluminescence is propor- tion of forced vibrations:
tional to the carrier current density, the lumines-
∂ 2 ui ∂ 2 um
cence center concentration, and the quantum yield ρ 2
− Kiklm = Fi ,
of the radiative transition. Its internal yield can be ∂t ∂xk ∂xl
greater than unity. Impact electroluminescence is where u is the ion displacement, ρ is the metal
the basis for the operation of various indicators mass density, Kiklm is the adiabatic elastic mod-
and screens used in devices to display informa- ulus tensor (without taking into account electron
tion. Especially promising are thin-film electro- renormalization), F is the force exerted by the
luminescence devices based on metal–insulator– electrons on the lattice (a functional of the field
380 ELECTROMAGNETIC WAVES WITH DISCRETE SPECTRUM IN A METAL
E(r, t)). Various mechanisms of the electron– the manner of a resonance phenomenon when the
lattice interaction make additive contributions sound interacts with collective vibrations of the
to F . For non-magnetic metals these are the iner- metal electron–hole plasma: helicons, dopplerons,
tia mechanism, the induction mechanism, and the and cyclotron waves.
strain mechanism. The contribution of the inertia In magnetic metals, in addition to the above
force term F 1 = −(m0 /e)∂J /∂t is usually small mentioned mechanisms, the magnetostriction elec-
in comparison with other mechanisms, where −e tromagnetic generation of sound occurs (see
and m0 are the charge and the mass of a free elec- Magnetostriction). This is caused by forces that act
tron, respectively; and J is the total current den- on the lattice due to the magnetoelastic interaction
sity taking into account the electrical neutrality. of the magnetization with the lattice. The induc-
The induction mechanism is determined by the tion mechanism of sound generation is modified
average Lorentz force F 2 = [j × B 0 ] where B 0 is by magnetization currents. The magnetoelastic in-
the external uniform magnetic field. teraction can be enhanced by ferromagnetic reso-
The strain mechanism is due to the electron– nance.
lattice interaction through a strain potential λ̂(p): The highest intensity of EM sound genera-
tion is reached in lm s by exciting standing sound
∂
F 3i = Λik χ, waves. This generation method is employed in
∂xk practical applications in laboratory practice and
λ̂ engineering. It does not require direct contact with
Λ(p) ≡ λ̂(p) − ,
g the material, in contrast to routine transformers,
and it has no major restrictions in frequency and
where χ is the nonequilibrium addition to the elec-
temperature.
tron distribution function; the angular brackets de-
note averaging over the Fermi surface; and g is ELECTROMAGNETIC WAVES WITH DISCRETE
the density of states. The role of various compo- SPECTRUM IN A METAL
nents of the force F depends on external para- Electromagnetic waves which are weakly at-
meters: the wave frequency, the temperature, the tenuated only in the neighborhood of discrete val-
magnetic field, and the metal characteristics. For ues of the wave vectors kn . Such waves are ob-
semimetals the EM generation of sound is due to served in metals in magnetic fields satisfying the
a strain force F 3 . For typical non-magnetic metals condition kR
1, where R is the cyclotron ra-
with a normal (classical) skin-effect, the EM gen- dius of the electron orbits, and kn is given by
eration of sound is caused by the induction mecha- the expression kn = π(n + 1/4)/R, where n =
nism. At low temperatures under the conditions of 0, 1, 2 . . . . The physical cause of the weak attenu-
the anomalous skin-effect, this generation of sound ation at these discrete kn values is associated with
takes place due. mainly, to the strain force F 3 . The oscillations of the dissipative conductivity σ diss as
efficiency of this generation depends essentially a function of the magnetic field B. There are min-
on the character of the electron reflection from the ima at k = kn where σ diss becomes small in com-
metal surface: for boundaries with diffusive scat- parison with the Hall conductivity. The effective
tering the EM generation is more significant than dielectric constant of the electron gas in this case
it is for mirror reflecting boundaries. The trans- becomes mainly real-valued, which makes possi-
formation of a metal to the superconducting state ble the propagation of weakly decaying electro-
strongly reduces the intensity of the sound genera- magnetic waves.
tion since this transformation causes a pronounced
decrease of the penetration depth of electromag- ELECTROMECHANICAL EFFECTS
netic e ld into the metal. A particular feature of the Reciprocal causality and correspondence be-
EM generation is the strain mechanism which in- tween electrical and mechanical states (properties)
volves various resonance effects intrinsic in elec- in solids. In ferroelectrics the electromechanical
tromagnetic and magnetoacoustic effects in met- effects are much more pronounced compared to
als: cyclotron resonance, geometrical oscillations, ordinary insulators, piezoelectrics, and semicon-
and so on. The intensity of the sound increases in ductors; this is characterized by abnormally large
ELECTRON AFFINITY 381
values of piezoelectric constants (e.g., piezomod- these expressions, the series is limited to terms of
uli dik and coefficients of electromechanical re- order D 4 . In the vicinity of a ferroelectric phase
lations kik , see Piezoelectricity) and strong de- transition, i.e. when approaching the Curie point
pendence of material parameters (dielectric con- T → θ(σk ), some diagonal components of the in-
stant εij , elastic compliances Skk , and piezomod- verse susceptibility αij , or principal minors of the
uli dik ) on temperature T , electric field E, and me- determinant of the set of equations vanish. This is
chanical stress σ (or uniform pressure). The main equivalent to the loss of thermodynamic stability
cause of strong electromechanical effects is the ex- of the crystal lattice, and is related to the sharp
istence of a ferroelectric phase transition that, in increase under action of external electric and me-
a certain range of variation of T and σk , stim- chanical forces of those components of electric
ulates the appearance (disappearance) or change flux density Di (T , E, σ ) and strain uk (T , E, σ )
of magnitude or direction of the spontaneous po- which are the solutions of the set, i.e. the di-
larization vector P s which is accompanied by a electric constant εij , piezomoduli dik , and so on.
significant deformation of the crystal lattice (see At the same time, the fundamental equation and
Ferroelectricity). In the vicinity of a phase transi- corresponding properties of the ferroelectric be-
tion, i.e. in the neighborhood of the loss of ferro- come essentially nonlinear. Outside the neighbor-
electric thermodynamic stability, one can observe hood of the phase transition, due to low values
an abnormally high pliability (strong response) to of some constants αij (low stability) and a high
an external electric and mechanical action, a gain electrostriction (i.e. constants κij k ), the above-
of reciprocity of thermal, elastic (mechanical) and indicated effects in ferroelectrics are also abnor-
electric properties, and their nonlinearity. Outside mally large. In ferroelectrics which lack the piezo-
the phase transition region these phenomena are effect (gik = 0) in the para-range, i.e. at T > θ and
less pronounced, however, both the pliability and P s = 0, in the ferro-range (T < θ and P s = 0)
the electromechanical effects in ferroelectrics re- the effect is caused entirely by the so-called lin-
main greater than they are in ordinary crystals. earized electrostriction, i.e. it arises from a term
Similar phenomena can be observed in ferroelas- κij k Di Dj σk in the thermodynamic potential; in
tics and, in general, in the majority of cases of other words, the effect is due to the presence of
structural phase transitions which are related to a relation uk (E) = κij k Dis Dj (E) = dik Ei where
a symmetry change of a crystal lattice. A quan- Dis = 4πPis is the spontaneous electric flux den-
titative description of the indicated effects can be sity. Strong electromechanical effects in ferro-
obtained by differentiating the thermodynamic po- electrics determine a number of specific complex
tential Φ(T , Di , σk ) with respect to Di and σk nonlinear dependences P s (T , σk ), εij (T , σk , Ei ),
(here Di is the electric ux density (polarization E (T , σ , E ), d (T , σ , E ); the phenomenon
Skk k i ik k i
Pi ), and σk are the mechanical stresses) to obtain of electromechanical hysteresis, dependences of
the fundamental equations: dielectric loss and mechanical loss on values of
static and dynamic electric and mechanical fields,
∂Φ
= αij Dj + ηij l Dj Dl + βij lm Dj Dl Dm as well as anomalies of the piezoresistive ef-
∂Di fect and electrical conductivity in ferroelectrics-
Ei semiconductors (see Photoferroelectrics).
− gik σk − κij k Dj σk = ,
4π
∂Φ ELECTRON AFFINITY
= Skk σk + gik Di + κij k Di Dj = uk , A measure of the tendency of an atom (or
∂σk
molecule) in its ground state to bind to a free
where uk is the strain tensor. The summation is electron, thereby becoming a negative ion. For an
carried out over repeated indices i, j, l, m = 1, energy affinity χ > 0, i.e. the generation of a nega-
2, 3; k, k = 1, 2, . . . , 6; the quantities αij , Skk , tive ion involves the release of energy and is there-
gik , κij k are, respectively, the tensors of reciprocal fore energetically favorable, then the correspond-
electric susceptibility, elastic compliance, recip- ing atom is said to have a positive electron affinity.
rocal piezoeffect, and electrostriction. To simplify Positive values of χ are exhibited by atoms, e.g.,
382 ELECTRON BEAM PUMPED LASER
hydrogen, carbon, oxygen, halogens, and by vari- applicable in a wide range of active media in the
ous molecules. The order of χ ranges from 0.1 to wavelength range from 0.3 to 30 μm. There are
1 eV. also heterostructure lasers. Advantages of elec-
In the context of the physics of semiconduc- tron beam pumped lasers include rapid scan of the
tors and other condensed media, the term “elec- emitting spot, and high peak output power den-
tron affinity” is sometimes used in reference to the sities. Laser cathode-ray tubes are used for pro-
height of the potential barrier to electron transfer jection television, data presentation, and optical
into the vacuum (an alternative term is “outer work memory systems.
function”). The quantity χ may assume various
numerical values, depending on whether the elec-
ELECTRON BEAM TECHNOLOGY
tron is removed from the semiconductor to infin-
Technological processes for the treatment of
ity, or transferred directly into the region near the
semiconductor surface where image forces play a
solids by an electron beam in vacuo. Included
role. The value of χ also depends on whether the among methods of electron beam technology are
electron is removed from a uniform neutral semi- welding, evaporation, fusion, and thermal treat-
conductor, or from one with a electric double layer ment (including cutting, milling and drilling of
or charged surface. Special treatment of the semi- materials). An electron beam is formed in a vac-
conductor surface allows establishing an electric uum of 100 –10−3 Pa in a device called an elec-
field in the near-surface layer, which lowers the χ tron gun; the beam is accelerated in an electrosta-
barrier, thus promoting electron escape from the tic field, focused, bent (deflected) with the help of
crystal. The potential barrier height may be low- magnetic fields when needed, and directed into a
ered to such an extent that the χ level will be lo- vacuum working chamber. It is in the chamber that
cated below the bottom of the conduction band, objects involved in the electron-beam process are
in the quasi-neutral region, deep within the bulk arranged: in the case of welding, heat treatment
semiconductor (negative electron affinity). This and trimming these are the parts to be welded or
negative value of the electron affinity is character- treated; in the case of fusion these constitute a cru-
istic of certain condensed media, e.g., liquid he- cible and the material to be melted in it; in the case
lium. A negative electron affinity is favorable for of evaporation these are the material to be evapo-
high emissive capacity of materials under condi- rated and those parts (substrates) at whose surface
tions of the X-ray photoeffect, electronic excita- the sputtering is performed. The vacuum pumping
tion, etc. system is an important part of any device of elec-
tron beam technology. In terms of the density of
ELECTRON BEAM PUMPED LASER gas flow during the heating and melting of a solid,
Solid-state lasers (mainly, semiconductor la- the output of the vacuum pumping systems in the
sers) in which bombarding the semiconductor range of pressure 10−2 –10−3 Pa is 102 –105 l/s.
with fast electrons produces high concentrations Typical electron beam parameters are the follow-
of excess carriers and, hence, conditions suitable
ing:
for generating coherent radiation. Electron beam
pumped lasers contain an active element – a semi- (i) for welding: power W ∼ 103 –105 W, energy
conductor target with an optical resonator cav- U ∼ 20–150 keV;
ity, a source of electrons, and a system for their (ii) for evaporating W ∼ 103 –104 W, energy
acceleration and focusing. During the interaction U ∼ 10–20 keV;
of electrons with the solid medium, up to 30% (iii) for fusion: W ∼ 103 –106 W, energy U ∼ 20–
of their initial energy is spent on internal ioniza- 30 keV;
tion, i.e. on the generation of electron–hole pairs. (iv) for trimming W ∼ 102 –103 W, energy U ∼
A high electron energy is selected below the radi- 50–150 keV.
ation defect formation threshold to obviate degra-
dation from cathodoluminescence. Electron beam Electron-beam welding is employed in the
pumping is an important all-purpose way to pro- instrument-making industry, for electrovacuum
duce laser emission in a semiconductor, and it is production, energy engineering, and the aircraft
ELECTRONEGATIVITY 383
ELECTRON COMPOUNDS
See Hume-Rothery phases.
Electron diffraction pattern obtained from a thin foil of
ELECTRON DIFFRACTION stainless steel.
Elastic (without energy loss) scattering of elec-
trons by individual atoms (molecules) or in media result of its scattering, a diffraction pattern (Fig.)
(gases, liquids, amorphous and crystalline solids). is produced. One can differentiate the diffraction
For the structural analysis of solids collimated from macroscopic regions of the sample which are
beams of electrons are used after passage through hundreds of micrometers in size, and microdiffrac-
a specific drop in potential V , and the directions tion with a resolution down to 5 nm. From the na-
and intensities of the scattered beams are recorded ture of the pattern, which is a section of the recip-
far from the object under investigation (Fraun- rocal lattice, one can determine the crystal struc-
hofer diffraction of waves). Depending on the value ture, the phase composition, the shape and dimen-
of V the following types are distinguished: low- sions of particles, and other structural features of a
energy electron diffraction (LEED, V ∼ 10 to sample under investigation. This technique is often
102 V), diffraction of electrons of medium ener- used in conjunction with the electron-microscopic
gies (V ∼ 102 to 103 V), and high-energy elec- observation of surface features (electron micro-
tron diffraction (V ∼ 104 to 106 V). In the LEED graph). The main applications of electron diffrac-
method the depth of penetration of electrons into tion analysis are the investigation of the structure
the solid is several angstroms so this method is of materials containing light elements, thin films,
used for investigations of the structure of the sur- ultradispersed bodies, and processes in solids that
face layers of solids. Using high-energy electron develop at the submicroscopic level. A particu-
diffraction it is possible to study the structure of lar branch of electron diffraction analysis is low-
samples with thickness 10−7 to 10−4 cm, and (by energy electron diffraction (LEED) which utilizes
reflection) surface layers of about 10 nm. Since for an accelerating potential ∼10–100 V, in sharp
high-energy electron diffraction the wave length contrast to ordinary electron diffraction analysis
is fractions of an angstrom this is a highly sensi- where the accelerating potential is ∼100 keV. This
tive method for investigating solids – the type of low-energy version provides information about the
unit cell, mosaic crystals, texture, phase composi- structure of monomolecular layers, adsorbed lay-
tion, defects, etc. Electron diffraction is the basis ers, etc.
of electron diffraction analysis.
ELECTRONEGATIVITY
ELECTRON DIFFRACTION ANALYSIS Measure of ability of an atom in a molecule
A method for investigating the structure of or condensed matter to attract electrons involved
solids based on electron diffraction. In this method in chemical bond formation. Robert Mulliken de-
the accelerated and focused electron beam from fined the electronegativity of an atom in terms
an electron gun is directed onto a sample, and as a of its ionization energy and its electron af nity,
384 ELECTRON EMISSION, SECONDARY
and Linus Pauling used bond dissociation ener- the electron characteristics do not depend on the
gies to devise an electronegativity scale. Tabu- field itself. In contrast to this, when the field is
lated values are arrived at using averages of data strong it results in certain specific effects among
from many compounds, and they range from 0.79 which are the dependence of the electron mobility
for Cs to 3.98 for F. Atoms with a high elec- on the field strength, saturation of electron drift
tronegativity easily add electrons to form anions, velocity, hot electron generation, the appearance
and atoms with low electronegativity easily re- of a thermal electromotive force for nonuniform
lease a valence electron to form a cation. The elec- field electron heating, ionization of the impurity
tronegativity value is of significance when deter- states by electrons, avalanche ionization, N- and
mining the type of chemical bond that exists be- S-shaped current–voltage characteristics, regions
tween atoms which form molecules, amorphous of negative differential conductivity, domains of
and crystalline solids. strong field, current lamentation, the Gunn effect,
the Sasaki effect, a many-valued electron distribu-
ELECTRON EMISSION, SECONDARY tion, and sign change of the bipolar drift velocity.
See Secondary electron emission. Electron heating by a field occurs in semicon-
ductors, but it plays no role in metals because of
ELECTRON GAS, TWO-DIMENSIONAL the high value of the Fermi energy of the elec-
See Two-dimensional electron gas. trons that provide the transport phenomena. Under
certain conditions optical transitions of electrons,
ELECTRON HEATING or the electric field itself, may produce cold elec-
Build-up of electron kinetic energy above its trons, with an energy ε̄ < ε̄0 .
thermodynamic equilibrium value as a result of
pumping additional energy into the electron sub- ELECTRON–HOLE LIQUID
system of a crystal by means of an electric cur- A dense electron–hole phase in which a gas
rent, laser irradiation, etc. Electron heating by a of high-density excitons in semiconductors is con-
field consists in increasing the kinetic energy dur- densed at fairly low temperatures. In an electron–
ing the electron drift in an electric field E (see hole liquid the excitons loose their individuality,
Current carrier drift). When ε̄ − ε̄0 ε̄0 , where ε̄ and these almost free current carriers (electrons
is the average electron energy, and ε̄0 is its value and holes) persist by means of exchange and corre-
in thermodynamic equilibrium, the electrons are lation forces. The mean distance between the car-
called warm electrons, while for ε̄ − ε̄0
ε̄0 , they riers is close in order of magnitude to the Bohr
are called hot electrons. Electrons acquiring addi- exciton radius a B , and their equilibrium concen-
−3
tional energy from the heating electric field within tration is n0 ∼ aB . For example, in Ge and Si the
their mean free path pass it on it to the crystal lat- corresponding free carrier equilibrium concentra-
tice during their collisions with phonons, impurity tion is n0 ∼ 1017 cm−3 and n0 ∼ 1018 , respec-
atoms and other defects of the lattice. Conditions tively. The electron–hole liquid differs from an
favorable for electron heating by a field include ordinary electron–hole plasma in semiconductors
a long mean free path, low temperature, elastic analogous to the way a liquid metal differs from
scattering of electrons by crystal defects, and high an electron–ion plasma: the liquid is sustained
strength of the electric field E. To avoid raising by the internal Coulomb forces and possesses a
both the lattice temperature T and ε̄0 the elec- certain equilibrium density n0 (T ) dependent on
tric field it is often applied to the material in short the temperature. Therefore, the electron–hole liq-
pulses. uid, in contrast to its plasma counterpart, does not
The degree of electron heating may be charac- spread over the whole sample, but rather it occu-
terized by the electron temperature. When the field pies only that portion of the bulk which can be
is weak, the average electron momentum p ∼ E filled uniformly with the temperature-dependent
and ε̄ − −ε̄0 ∝ E 2 , so that the change in elec- density n0 (T ) consistent with the number of non-
tron energy during the course of the transport phe- equilibrium electrons and holes that had been in-
nomena may be neglected as of second order, and troduced into the sample.
ELECTRONIC CONFIGURATION 385
The condensation of the free exciton gas into hole plasma can be formed by the action of strong
the electron–hole liquid is a rst-order phase tran- very-high-frequency fields, and by electron injec-
sition. Upon reaching the mean concentration of tion through a contact in a strong electric field (see
excitons equal to the temperature-dependent value Solid-state plasma).
nc (T ) (at low temperatures nc (T ) n0 (T )), a
layering into two phases takes place in the bulk, ELECTRONIC CONDUCTIVITY
i.e. a separation into portions filled by the exciton This nonstandard term is somewhat arbitrar-
gas, and others filled by drops of electron–hole liq- ily defined as referring to the electrical conduc-
uid (electron–hole drops) with the density n0 . At tivity of a material caused by (i) electron and (or)
further increasing the average bulk concentration hole transport, as in metals, as distinct from ionic
of nonequilibrium electrons and holes, the vol- conductivity where the current carriers are ions, as
ume occupied by the electron–hole liquid grows, takes place in liquids, (ii) jump or hopping con-
but its density at a given temperature n0 (T ) does ductivity when charges hop from ionic site to ionic
not change until the electron–hole liquid fills the site, and (iii) in semiconductors there is n-type
whole sample. The condensation of an exciton gas conductivity when the majority current carriers are
into an electron–hole liquid was first observed in conduction electrons, as distinct from p-type con-
Ge crystals. The presence of this liquid in crystals ductivity when the majority carriers are holes. The
can be established from radiative recombination n-type occurs in semiconductors when the donor
spectra, from plasma and magneto-plasma reso- concentration exceeds the acceptor concentration,
nance absorption, and from light scattering. The and p-type occurs when the acceptors dominate.
electron–hole liquid is stable in semiconductors
with a strong effective mass anisotropy and a high ELECTRONIC CONFIGURATION
degree of orbital degeneracy of the energy spec- Complex of occupied one-electron states of a
trum (many-valley type) (see Many-valley semi- system with designated quantum numbers which
conductors). determine the electron energies. The electron con-
figuration of an atom (n1 l1 )N1 . . . (ni li )Ni is given
ELECTRON–HOLE PAIR by the occupation numbers of Ni of electron
An electron in the conduction band associated shells with the principal quantum number ni and
with a hole in the valence band. It can exist ei- orbital angular momentum li . As a rule, only
ther as separated particles (the Coulomb attrac- partially occupied, non-inner core, shells with
tion is negligibly small) or as a bound pair in 1 Ni < 2(2li + 1) are designated. The energy
the form of a Wannier–Mott exciton. Electron–hole levels (terms) of the system with a given elec-
pairs are formed as a result of either thermal exci- tron configuration are determined by inter-electron
tation of the electron subsystem, or under the ac- (Coulomb and exchange interactions) and other
tion of light, nuclear radiation, electron (hole) in- interactions (in particular, spin–orbit interactions).
jection, shock waves, etc. Single-electron states (orbitals) in molecules and
isolated complexes in condensed media (solutions,
ELECTRON–HOLE PLASMA glasses, crystals) are classified by irreducible rep-
A system of free electrons and holes at suf- resentations of their symmetry point group. An
ficiently high densities and temperatures where electron configuration of a molecule is given in
there is a dissociation of the high-density excitons the form (n1 Γ1 )N1 . . . (ni Γi )Ni where ni 1 is
produced in semiconductors by light. Unlike an the occupation number of the state of a given
electron–hole liquid which appears at high densi- symmetry listed in the order of increasing energy;
ties and low temperatures due to the Mott metal– the symbol $i denotes the irreducible representa-
insulator transition in the exciton system, a gas-like tion with the degeneracy (multiplicity) gΓ , where
electron–hole plasma fills the entire crystal bulk. 1 Ni 2gΓ . In crystal eld theory the elec-
The behaviour of the charged gas particles in a tron configurations of complexes of d-state tran-
plasma is dominated by the electromagnetic inter- sition metals with cubic symmetry are constructed
actions between them. A nonequilibrium electron– from orbitals e(dz2 , dx 2 −y 2 ) and t2 (dxz , dyz , dxy )
386 ELECTRONIC LEVEL, LOCAL
fact that the inefficiency of ionization mechanisms systems the sample should be thin enough to per-
in the bulk of alkali-halide crystals is due to the mit the passage of electrons.
high symmetry of their crystal lattice, and the rel- Several varieties of electron microscopes will
atively compact arrangement of atoms in such lat- be described.
tices. It had been shown by C.C. Parks et al. (1983) The working principle of a reflection electron
that the electron-ion emission at the (100) surface microscope involves the reflectance of the incident
of NaF crystals is caused by ionization of the K- electrons, δ = 30–35 nm.
shell of Na. In the case of alkali-halide crystals, In an emission electron microscope the image
this emission is due to the interatomic interac- is formed by electron emerging from the sample
tion forces, and involves the spatial arrangement surface, δ = 20 nm.
of an overcharged ion. In the case of a surface with In a scanning electron microscope an electron
a prevalence of covalent bonds, the electron–ion probe scans over the sample surface, thereby form-
emission occurs either as a result of local excita- ing its image, δ = 3 nm. This device can work
tion of chemical bonds with the potential barrier by means of secondary reflections and transmitted
reduced to the point where thermal fluctuations electrons at U < 50 keV. The analyzers provide
can eject an electron to the vacuum state outside, the determination of the elemental composition of
or the emission results from the disappearance of a sample.
the barrier when an electron changes to an an- A specular electron microscope furnishes an
tibonding orbital state. Mechanisms of electron– image of a potential relief (i.e. the distribution of
ion emission can also induce stimulated desorp- potential near the sample surface), δ = 8–15 nm
tion from a solid surface of any type. (in practice this value may be closer to 100 nm).
A shadow electron microscope is designed for
ELECTRON MICROSCOPE studying surface relief.
An electronic optical device in which, as a
ELECTRON MICROSCOPE, SCANNING
result of the interaction of a primary beam of
See Scanning electron microscope.
accelerated electrons with a sample, the image
of its surface or its internal structure is formed ELECTRON MICROSCOPY
with a resolution δ down to 0.1 nm. Electron mi- Set of methods for investigating the structure
croscopes are divided into (i) magnetic (electro- of solids (from the mesoscopic to the atomic-
magnetic and magnetostatic), (ii) electrostatic, and molecular level) and their local chemical com-
(iii) combined magnetic/electric types; into de- position with the help of electron-optical devices
vices with resolution δ = 0.1–1.0 nm, δ = 1–3 nm, called electron microscopes.
and δ > 3 nm; and into (i) transmission, (ii) reflec- Transmission electron microscopy can attain a
tion, (iii) emission, (iv) scanning, (v) specular, and resolution up to 0.1 nm when passing the elec-
(vi) shadow apparatus depending on the kind of tron beam through a sample 10 nm thick. With
radiation that forms the image (see also Electron the help of high-resolution electron microscopy,
microscopy). the atomic structure of solids becomes available
A transmission electron microscope allows one for observation. Using thicker sample-foils (0.1 to
to study the internal structure of objects with a res- 1.0 μm thick) one can study imperfections of the
olution down to 1 nm. This class of electron mi- crystalline structure of solids which become de-
croscopes is divided into three subgroups accord- tectable through their phase contrast. The use of
ing to the energy value U of the applied acceler- a narrow electron beam (about 1 nm in diame-
ating potential: (i) U < 100 keV, (ii) U = 100– ter) scanning over the foil surface makes it pos-
500 keV, and (iii) U > 500 keV. An electron mi- sible to combine the electron-microscopic study
croscope equipped with attachments can be used of the solid bulk structure with the analysis of its
for crystallographic and phase analysis carried out local chemical composition (see Local analysis).
at various temperatures, for studying samples un- Spectrometers of various kinds are used for in-
der strain, and for the determination of their ele- vestigating this chemical composition: X-ray spec-
mental composition. To carry out research on these trometer with energy dispersion, X-ray absorption
388 ELECTRON–NUCLEAR DOUBLE RESONANCE
Energy levels of the system of an electron and a nucleus with spins S = I = 1/2 in a constant applied magnetic field B
taking into account the Zeeman and hyperfine interactions: (a) levels in thermodynamic equilibrium; (b,c) variation of
populations under effect of the microwave frequency and the radio-frequency acting in the absence of spin relaxation;
(d) ENDOR in the presence of relaxation, the fastest processes being associated with levels, affected by the input power,
as shown by the dashed line; Ts and Tx are characteristic relaxation times.
power microwave input signal which saturates and nucleus), as well as from the nuclei of atoms (lig-
thereby reduces the EPR signal to zero ampli- ands) adjacent to the paramagnetic center (ligand
tude. The subsequent application of the radiofre- ENDOR) and from nuclei distant from the para-
quency and an adiabatic fast passage through magnetic center (distant ENDOR). There is no di-
resonance cause the inversion of the nuclear sub- rect interaction of the electrons of the paramag-
level populations which induces the nonstation- netic center with such nuclei. Its mechanism is
ary (dnM,m /dt = 0) emission of the EPR signal; associated with the space diffusion of nuclear po-
M, m are the projections of the spins on the di- larization from adjacent nuclei, polarized through
rection of B, nM,m are the relative populations of the solid effect or through the dipole–dipole reser-
levels (hνn hνe and ε 1, ε = hνe /(2k B T )), voir, to the distant nuclei. The ENDOR of this
hνe and hνn are the energies of the microwave type has a signal at the nuclear Larmor fre-
and radio-frequency quanta which cause the cor- quency νn∗ which, after turning off the incident
responding electronic and nuclear transitions. In radio-frequency, falls off with a time scale char-
stationary ENDOR (Fig. (d)) the intensity of the acteristic of nuclear spin–lattice relaxation. The
partially saturated EPR signal is determined by ENDOR of non-ordered systems (polycrystals,
the competition between the comparatively fast amorphous bodies, glasses) is also observed at
processes of relaxation with induced transitions. frequencies close to νn∗ , but its mechanism is
The passage through nuclear resonance is per- the same as for the adjacent nuclei. If the radio-
formed sufficiently slowly, so that the system re- frequency field increases the intensity of mi-
mains close to equilibrium: dnM,m /dt = 0. The crowave adsorption, the ENDOR is called posi-
enhancement of the EPR signal results from the tive, and in the case of reduction the ENDOR is
shortening of the spin relaxation time due to called negative. Radio-frequency selective satura-
switching from the Tx process to the Ts process tion is a specific kind of ENDOR using pulsed
by the applied radio-frequency, and as a result, the methods of detection.
extent of the EPR saturation is reduced. In ENDOR the replacement of the microwave
Electron–nuclear double resonance may arise or radio-frequency electromagnetic waves by sonic
from the nucleus of an impurity atom which consti- vibrations is possible, to provide acoustic-magne-
tutes the paramagnetic center (ENDOR at doping tic (or magnetoacoustic) ENDOR, and when the
390 ELECTRON ORBIT COLLAPSE
signal is recorded through optical transitions it is a constant magnetic field forms a system of en-
called optically detected ENDOR. In ENDOR in- ergy levels Ei which differ in the magnetic quan-
formation is extracted from the frequency, width, tum number M of its magnetic moment. Under
shape and amplitude of the resonance lines, from the action of electromagnetic waves, magnetic di-
their angular dependences on the B field, from pole transitions can be induced between these lev-
ENDOR dynamics (dependences on intensities of els, called magnetic or spin sublevels, with the ab-
microwave or radio-frequency fields), and also sorption (emission) of an energy quantum of fre-
from the variations of signal as a result of exter- quency νij = |Ei − Ej |/ h. In the particular (but
nal effects. quite common) case of a single electron with spin
ENDOR is one of the most effective and accu- s = 1/2, located in a magnetic field B, the follow-
rate methods of measuring hyper ne interactions ing relation is satisfied:
and of clarifying the structure of defects (impu- hν = gβB, (1)
rity, radiative, etc.) in solids. It allows one to de-
termine the spatial distribution of spin densities where g is the electron g-factor, β is the Bohr
e|ψ|2 of electrons at paramagnetic centers (ψ is magneton. The frequency ν defined by Eq. (1)
the unpaired electron wave function). It is used is called the Zeeman frequency, energy levels
also for determining electric quadrupole and mag- E±1/2 = ±gβB/2 responding Eq. (1) are called
netic dipole moments of atomic nuclei, and spin re- Zeeman energies, and the spacing between these
laxation parameters in electronic–nuclear systems. levels E±1/2 is called the Zeeman splitting. As
ENDOR provides information about fundamen- a rule, B = 0.1 to 5 T and g ≈ 2, which yields
tal parameters of crystals: their band structure, ν = 3·109 –1.5·1011 Hz.
intracrystalline fields, strain potentials, and local The EPR experiments are carried out with the
properties in the vicinity of paramagnetic centers. help of a special apparatus called a magnetic reso-
In the presence of two radio-frequency sources nance spectrometer which can operate in the cen-
a triple electron–nuclear–nuclear resonance can timeter or millimeter ranges of wavelengths. Mi-
be carried out. It is similar to ENDOR in its crowave (radar) instrumentation is used, systems
mechanism and properties, the difference being of waveguides and resonators with generators
the presence of two radiofrequency sources. The (klystron) and detectors (crystal diode). A sample
extra radiofrequency permits one to determine the several cubic millimeters in volume is located in a
relative signs of hyperfine coupling constants, to resonator where a component of the electromag-
separate different processes of spin relaxation, and netic wave (usually a magnetic one), which in-
to determine their characteristic times. duces the transitions, has a maximum value. The
resonator is situated between the pole pieces of
an electromagnet (ordinary or superconducting)
ELECTRON ORBIT COLLAPSE
which provides the magnetic field that produces
See Collapse of electron orbits.
the Zeeman splitting of the levels. The resonance
condition (1) is satisfied by scanning the magnetic
ELECTRON PAIR, UNSHARED field strength B while keeping the frequency at
See Unshared electron pairs. a constant value ν. The magnitude of the mag-
netic field at resonance depends, in general, on
ELECTRON PARAMAGNETIC RESONANCE the orientation of the vector B with respect to the
(EPR), also called electron spin resonance (ESR) sample. A signal with a bell shape or its deriva-
Resonance absorption (emission) of electro- tive is observed with the help of an oscilloscope
magnetic waves of the radio-frequency or micro- or recorder. In a crystal positioned in the external
wave (superhigh-frequency) ranges (109 –1012 Hz) magnetic field, a paramagnetic particle with spin
by paramagnets whose paramagnetism is caused s = 1/2 has two levels (E1/2 and E−1/2 ). These
by unpaired electrons; this is a particular case of levels vary in proportion to the value of the mag-
magnetic resonance. The effect was discovered netic field strength B, and in many single crystals
by E.K. Zavoisky (1944). A paramagnetic spin in they are dependent on the orientation of vector B
ELECTRON PARAMAGNETIC RESONANCE 391
arise from transitions between the levels which momentum are specified by different temperature
compete with the microwave-induced transitions. dependences. The strong electron–phonon interac-
If these are dominant then saturation (equaliz- tion in ionic and polar crystals can lead to electron
ing of level populations) occurs, which results in (hole) self-localization, i.e. to the formation of po-
a broadening and decrease in amplitude of the larons (electron–phonon interaction with optical
EPR signal. The relaxation processes are speci- phonons) or condensons (interaction with acousti-
fied by relaxation times and described by kinetic cal phonons). In metals an important role involves
equations and equations for energy level popula- the particular case of the electron–phonon inter-
tions. action, that is the acousto-electronic interaction.
The phenomenon of the EPR has found a wide In second-order perturbation theory the electron–
application in the physics of solids as a method phonon interaction leads to an effective attraction
for studying crystal lattice defects, and site sym- between electrons. Such an attraction in metals
metries. Owing to the existence of low-symmetry in the vicinity of the Fermi surface, arising due
effects and electric eld effects, the determina- to the exchange of virtual phonons, can lead to
tion of the local symmetry of a defect is possible the Cooper pair coupling of electrons (see Cooper
from EPR data. A combination of the EPR method pairs), and the transition from a normal state to a
with the nuclear magnetic resonance method, i.e. superconducting state (see Phonon mechanism of
electron–nuclear double resonance, allows one to superconductivity). The electron–phonon interac-
probe nearest neighbor paramagnetic particles and tion in metals is specified by the electron–phonon
their localization in the lattice. Thus the nature of interaction constant λ as
defects becomes available for study. EPR is also
applied in studies of solid surfaces, phase transi- N(EF )|p|∇Vei |p |2
λ= ,
tions, and disordered systems. An important tech- Mω2
nical application is the design and production of where N(E F ) is the density of states at the Fermi
masers (see Quantum ampli er).
level E F per unit spin, p|∇V ei |p is the matrix
element of the gradient of the shielded pseudopo-
ELECTRON PARAMAGNETIC RESONANCE
tential of the electron–ion interaction calculated
FINE STRUCTURE
using Bloch functions (see Bloch theorem) of elec-
See Fine structure of electron paramagnetic
resonance (EPR) spectra.
trons in the conduction band, M is the ion mass,
and ω2 is the square of the photon frequency av-
ELECTRON–PHONON INTERACTION eraged over the spectrum. The expression in the
Interaction of free charge carriers (conduction numerator is sometimes referred to as the Hop-
electrons and holes) with quanta of crystal lat- field parameter (J.J. Hopfield, 1969). A specific
tice vibrations. In insulators and semiconductors (average) transferred momentum of the electron–
with predominately covalent bonding in the long- phonon interaction (q = p − p ) is of the order of
wave limit the electron–phonon interaction with the Fermi momentum pF , i.e. the effective wave
acoustical phonons can be described in terms of length of the virtual phonons is of the order of
a strain potential, whereas in ionic crystals and h/pF ≈ 10−8 cm. For such short-wave phonons
in metals the basis of the electron–phonon in- the approximation of uncorrelated local vibrations
teraction is the Coulomb (electric) interaction of of ions (atoms) at crystal lattice sites is valid with a
electrons with ions. The electron–phonon interac- high accuracy. In superconducting alloys and com-
tion determines many kinetic and thermodynamic pounds consisting of atoms of different kinds, an
properties of solids, and this is described by the important role is played by the interstitial corre-
Fröhlich Hamiltonian. This interaction leads to cur- lations of ion displacements which are responsi-
rent carrier scattering in crystals. In the case of ble, in particular, for the concentration dependence
semiconductors one should distinguish the mecha- of λ (and, consequently, the critical temperature
nisms of scattering off acoustical phonons and op- T c of the transition to the superconducting state)
tical (polar and nonpolar) phonons. Correspond- for variations of the alloy composition, or for de-
ingly, the relaxation times in regard to energy and viations of the compound from stoichiometry. In
ELECTRON-PLASTIC EFFECT 393
this case, an overall electron–phonon interaction takes place in the presence of two factors: (i) me-
constant cannot be represented by a sum of partial chanical stress that exceeds the yield limit; (ii) the
constants for individual components, but rather it drift of free electrons inside the deforming metal.
has a more complicated nature that is, e.g., clear In its turn, the electron drift can be produced by
from the nonlinear dependence of T c and elec- (1) the application of a potential difference across
tronic speci c heat γ on the composition of solid a sample to induce an electric current; (2) the injec-
solutions of transition metals. The difficulties in- tion of accelerated electrons into a sample; (3) ir-
volved in calculating this interaction constant for radiation by UV, X-ray or γ -radiation to bring
complex compounds from first principles became about the photoeffect and Compton effect, as well
evident with the discovery of high-temperature su- as the formation of electron–positron pairs; (4) ex-
perconductivity in the cuprates. posure of the sample to high power light radiation
(subsurface drift); (5) the Stewart–Tolman effect.
ELECTRON–PHONON INTERACTION IN ONE In these cases one can distinguish the electron-
DIMENSION plastic effects of current, radiative, electromag-
See Cross-deformation interaction. netic, light and inertial origin. The basis of the
electron-plastic effect is the transfer of momen-
ELECTRON–PHOTON SPECTROSCOPY tum and energy from electrons to the deforming
Method of investigation of solid surface prop- crystal lattice, mainly at defect points such as dis-
erties based on measurements of the spectral, locations, dislocation stops, and point defects. The
polarization, angular dependence, etc., character- electron-plastic effect depends linearly on the cur-
istics of photons emitted during electron bom- rent density; the effect is polar, and in this respect
bardment (see Inverse photoelectron emission). it differs in principle from the Joule–Lenz effect.
Electron-photon spectroscopy is utilized for stud- The force of drifting electrons acting on a unit
ies of the energy band structure of the surface lay- dislocation length is determined by the expression
ers of solids (bulk samples and lm s, including in- F = (j/e)(1 − (v h /v e ))mv F b where e and m are
sular lms); this provides, in particular, data on the the electron charge and mass, respectively; v F is
density of electron states, the frequencies of sur- the electron velocity on the Fermi surface; j is the
face plasma oscillations, and exciton transitions. current density; b is the Burgers vector. The con-
This spectroscopy is employed to study lumi- dition for the appearance of the electron-plastic
nescent centers, adsorption–desorption processes, effect is the excess of the electron drift velocity
mechanisms involving the interaction of electrons v e over the phase velocity of elastic dislocation
with the surface, as well as to determine the degree waves v d , or simply over the dislocation velocity
of surface roughness, and so on. (in the case of uniform strain), i.e. v e > v d . Since
In electron–photon spectroscopy the electrons v e = j/(en) (n is electron concentration per unit
with the energies exceeding 103 eV can probe volume), to produce the effect it is necessary to
the surface under study in a superhigh vacuum provide powerful electron fluxes (i.e. a huge j ).
(needed to ensure purity). The emission character- In practice, during rolling, drawing, punching and
istics are analyzed with the help of optical meth- flattening one can use short-time powerful pulses
ods using spectral devices. with magnitude j = 1000 A/mm2 and duration
10−5 –10−2 s. This facilitates the metal treatment
ELECTRON-PLASTIC EFFECT (discovered in with pressure (the treatment effort is reduced by
1963) 35–50%); after the treatment the metal plasticity
Characterizes the effect of a directed electron increases considerably; the physical and mechan-
flux (electron drift) on the plastic flow of a mate- ical properties of the material, such as residual
rial (see Plastic deformation). In the case of met- plasticity and degree of texture perfection, become
als, the electron-plastic effect causes repeated ac- improved, and some undesirable phase transitions
celeration of the flow rate and reduction of the (e.g., γ → α transformation of steel) become sup-
material resistance to deformation. This effect is pressed. An alternative view of the mechanism of
opposite to the electron-strain effect. This effect the electron-plastic effect regards current pulses
394 ELECTRON–POSITRON ANNIHILATION METHOD
gas and the lattice is described by the energy bal- and spin as electrons are called conduction elec-
ance equation which is the energy conservation trons. The Fermi liquid in crystals can also pos-
law for a gas of electrons drifting in the electri- sess elementary excitations with a positive charge,
cal field. For a uniform gas under stationary con- i.e. holes. The main characteristics of metals can
ditions eμE 2 = (ε − ε0 )/τε , where τ (ε) is be described by the dispersion law of conduc-
the relaxation time of the electron energy, also tion electrons (or simply, electrons) εc (p) in close
called the energy relaxation time, ε0 is the equi- proximity with the Fermi surface. In this spectral
librium value of ε; in the absence of degeneracy range (only in it) the attenuation of electron states
ε0 = (3/2)k B T , where T is the lattice tempera- due to the electron–electron interaction is insignif-
ture. The left-hand side of this equation is the en- icant, and disappears completely at εc (p) = εF ,
ergy obtained from the field by one electron per which provides some justification for the widely
unit time, and the right-hand side shows the en- accepted representation of the system of electrons
ergy transferred to the lattice per unit time. For the in a metal as an ideal gas. At present, with the help
quasi-elastic collision of electrons τp τε , of resonance, galvanomagnetic, and other experi-
and often τp ∼ εs , τε ∼ εl , where the exponents s ments, the Fermi surface topology with its main
and l are numerical coefficients. Upon scattering features, the velocity, and the effective mass ten-
on the acoustic strain potential we have s = −1/2 sor of electrons at the Fermi surface, has been de-
and l = −1/2; on the optical deformation potential termined for almost all metals.
(see Acoustic vibrations) −1/2, 1/2; on the polar The system of mobile conduction electrons
optical potential 1/2, 3/2; on the acoustic piezo- plus the background of compensating station-
electric potential (see Piezoelectricity) 1/2, 1/2; ary positive charge forms a so-called solid-state
on ionized impurities s = 3/2; and on neutral im- plasma that exhibits a set of related collective
purities s = 0 (see Impurity atoms). properties: weakly damped high-frequency vibra-
The relaxation times τee , τp and τε play an tions (plasmons, helicons, etc.) can be excited in
important role in the theory of hot electrons. The it. A static charge is screened the more effec-
relaxation of the non-stationary space charge in an tively the greater the density of the electron spec-
electron gas is characterized by the Maxwell relax- trum at the Fermi surface, and in good metals the
ation time, and by the frequency of the plasma os-
screening radius is of the order of the lattice con-
cillations.
stant. In some cases correlation effects induced by
ELECTRONS in a crystal
mutual electron interactions prevent an interpre-
The electrons of atoms which compose a crys- tation in terms of band theory (e.g., Wigner crys-
tal. The states of the inner atomic shell electrons tal). Fermi-liquid correlations between the elec-
in crystals deviate only slightly from their states tron spins sometimes lead to the appearance of
in gaseous atoms. In contrast to this, the valence Bose-type excitations in a metal which have a spin
shell electrons of atoms in a crystal determine wave nature.
the character and the magnitude of the chemical The interaction of electrons with crystal lat-
bonding forces as well as the principal physical tice vibrations (see Electron–phonon interaction) in
properties, i.e. the electric, magnetic, optical, ther- metals is insignificant because the specific phonon
mal and mechanical properties. The basis for de- energy is small compared to εF . Phonon exchange
scribing the structure of their energy spectrum and causes the attraction between electrons that in-
the nature of the quantum states of electrons in duces the formation of Cooper pairs and the tran-
crystals is band theory that explains the division sition of a metal to the superconducting state (see
of solids into metals, insulators and semiconduc- Superconductivity) if this attraction exceeds the
tors. The band electrons interact with each other, Coulomb repulsion.
with phonons, and with various lattice defects. In insulators and semiconductors at T = 0, the
In metals the conduction band is partially full, valence band is fully occupied, and the conduc-
and the electron system behaves as a degenerate tion band is empty. The band gap width E g in in-
Fermi liquid even at room temperature. Elemen- sulators has specific atomic values (∼10 eV), with
tary excitations of this liquid with the same charge much smaller values in semiconductors (E g = 0
396 ELECTRON SCATTERING
to 3.5 eV) (see Gapless semiconductors). Of great whereas in homopolar (covalently bonded) crys-
significance for the electronic properties of these tals this is due to acoustical bending vibrations
crystals are the states corresponding to the bot- (condenson).
tom (electrons) of the conduction band and the High-energy electrons which penetrate into a
top (holes) of the valence band, as well as lo- crystal as a result of external irradiation are braked
cal electronic levels in the forbidden gap (donor (decellerated) by their interaction with the target
or acceptor ones) which are, as a rule, associated atoms. The energy transferred to the crystal in this
with either impurity atoms (donors or acceptors) manner causes ionization, and the excitation and
or intrinsic defects in the crystal. The density of displacements of atoms from their equilibrium po-
charge carriers in semiconductors is several orders sitions. A subsequent relaxation leads to a par-
of magnitude less than it is in metals, the frequen- tial return of the excited electrons and atoms to
cies of plasma oscillations in them are much lower their equilibrium states; however, some of the re-
than in metals, and the shielding radius is greater. laxation can be to metastable states, e.g., atoms
An electron and a hole attract each other and can occupy interstitial positions while their orig-
can form a bound state of the hydrogen-like type, inal lattice sites remain unoccupied (vacancies),
the so-called Wannier–Mott exciton. A related exci- and the electrons can populate local electronic lev-
tation in a molecular crystal is the Frenkel exciton. els of defects. Permanently displaced atoms con-
The Wannier–Mott exciton radius is usually one stitute what might be called electron-induced dis-
to two orders of magnitude larger than the lattice ordering. The paths of high-energy electrons in
constant, so at a certain exciton density attainable crystals are characterized by charged particle en-
using laser irradiation (see Laser) their wave func- ergy losses and mean free paths which depend on
tions overlap, thereby making it possible to form both the incoming electron energy, and the crystal
either an electron–hole plasma or a new phase, that characteristics. In conducting materials the excess
is an electron–hole liquid with metallic properties. electrons are easy emitted from the irradiated sam-
In semiconductor heterostructures the motion ple, which thereby remains electrically neutral. In
of electrons is quasi-two-dimensional. Such sys- insulators there can be an accumulation of a no-
tems located in a strong quantizing magnetic eld ticeably negative charge due to electrons trapped
exhibit a series of unusual properties such as, e.g., in the sample if the sample thickness exceeds the
the quantum Hall effect. electron mean free path.
There is a small group of crystals, called semi-
metals, in which the conduction band weakly
ELECTRON SCATTERING
overlaps the valence band. The electrical conduc-
See Current carrier scattering.
tivity of such systems is of a metallic character, but
much less so than is the case for a true metal, be-
cause the number of charge carriers is 3–4 orders ELECTRONS, HOT
of magnitude less than that in typical metals. See Hot electrons.
In the case of strong scattering of the conduc-
tion electrons the band theory approximation no ELECTRON SPECTROSCOPY
longer holds. Transits of electrons occur via occa- Study of the energy distribution (spectra) of
sional jumps between localized states of the po- electrons emitted by atoms, molecules, and solids
laron type, or states localized at either impurity under action of incident electromagnetic radiation,
atoms or intrinsic defects. The interaction of con- primary electrons or ions that strike the target,
duction electrons with phonons (which is quite or electric fields. The methods of electron spec-
strong) results in the formation of autolocalized troscopy are used for investigating the energy lev-
states (see Self-localization) in which the electron els of atoms, molecules and macroscopic systems
deforms the lattice and becomes confined in a composed of them, as well as the quantum tran-
newly-formed potential well. In ionic crystals this sitions between these levels, to obtain informa-
species (polaron) is produced by the interaction tion about the structure and the properties of ma-
of electrons with the polarized optical vibrations, terials. Core electron levels and valence electron
ELECTRON SPECTROSCOPY 397
levels, collective excitations (plasmons) and vibra- the coincidence technique is used for identifica-
tional spectra are also studied by electron spec- tion of two emitted electrons with particular en-
troscopy. The limiting positions in the observed ergies and momenta excited by the primary elec-
spectra provide the energies of individual groups tron. A comparison of experimental distributions
of emitted electrons, whereas the spectral line in- of emitter electrons with calculated distributions
tensities furnish the relative numbers of these elec- provides the dispersion dependence of the elec-
trons. trons in a solid, information about the orientation
Electron spectroscopy originated from the clas- of adsorbed molecules with respect to the surface,
sical experiments of J. Franck and G. Hertz (1912) and about their molecular orbitals involved in the
where the discrete atomic energy levels were de- bonding to the surface. The most important ar-
termined from the energy loss of electrons passed eas of application of electron spectroscopy are the
through a low-density gas. These data in combi- determination of the cross-sections of excitation
nation with optical spectroscopy data provided ex- and ionization of atoms and molecules, the energy
perimental justification for the Bohr theory (1913), structure of core and valence levels in solids, the
then the experimental underpinning for quantum energy structure of surface and adsorbed layers,
mechanics which, in its turn, became a theoretical the qualitative analysis and the quantitative analy-
basis of modern electron spectroscopy. sis of the material and its surface. Especially wide-
Depending on the type of excitation, one can spread is Auger spectroscopy which has high sen-
distinguish electron-electron, ion-electron, tunnel- sitivity (1010 to 1012 atoms/cm2 ).
ing (autoelectron), photoelectron, and beta spec- The wide application of photoelectron and
troscopy. The β-spectroscopy deals with the spec- X-ray-electron spectroscopy is stimulated by the
tra of electrons and positrons ejected during a availability of synchrotron radiation, i.e. a contin-
nuclear beta-decay, as well as conversion elec- uous photon source with quantum energies from
trons (emitted during the energy transfer from an 10 eV to 104 eV. Tunneling spectroscopy and vi-
excited atomic nucleus to a surrounding atomic brational electron spectroscopy allow one to study
electron shell), and electrons generated during the vibrational terms of adsorbed molecules in
the interaction of gamma rays, X-rays, and other the region of wavelengths λ 0.3 mm which
high-energy radiations with matter. Depending on is inaccessible for optical spectroscopy, and to
the kind of emission, electron spectroscopy is gather information about the microscopic struc-
divided into tunneling spectroscopy (or electron ture of adsorbates. Tunneling spectroscopy is also
spectroscopy of inelastic tunneling), characteristic used in studies of the distinct features of the elec-
energy loss spectroscopy (EELS), Auger electron tronic structure of superconductors (see Tunnel-
spectroscopy, Auger ion-electron spectroscopy, ing spectroscopy of superconductors). When us-
electron spectroscopy of the appearance and dis- ing the photon or electron excitation in electron
appearance of threshold potentials, photoelectron spectroscopy, the depth of the layer under study
spectroscopy, X-ray-electron spectroscopy (elec- depends on the conditions of the experiment and
tron spectroscopy for chemical analysis), and the emitter electron energies. For energies below
the “e–2e experiment”. Characteristic energy loss ∼100 eV there is a minimum of a few atomic
spectroscopy is divided into ionization spectros- layers, i.e. the method is applicable for surface
copy which deals with energy loss spectra for core analysis. Combining electron spectroscopy with
level excitation, and high-resolution characteris- ion etching allows one to carry out a layer-by-layer
tic energy loss spectroscopy (vibrational electron analysis. The fine structure of the electron spec-
spectroscopy) where the vibrational states of the tra contains information on the density of free and
surface and the molecules adsorbed at it are stud- occupied states (see Density of states), the struc-
ied. There are integral and angle-resolved types ture of atomic and molecular orbitals, the spin–
of electron spectroscopy which differ in the col- orbit and multiplet splittings, the Jahn–Teller ef-
lection angle of the emission. In some cases, the fect, as well as multielectron processes like shake-
scattered emitter electrons are divided according off, shake-up, and so on. Elastically and inelasti-
to their spin polarization. In the e–2e experiment, cally scattered electrons also provide information
398 ELECTRON SPIN-FLIP SCATTERING
on the atomic structure of condensed systems and formation of a material. It was discovered in 1977.
surface defects. The electron spectroscopy of ap- In the case of metals this phenomenon is oppo-
pearance potentials and of surfaces, and extended site to an electron-plastic effect. The mechanism
X-ray absorption fine structure (EXAFS) (see Fine of the electron-strain effect, in contrast to that of
structure) allow one to determine radial distribu- the electron-plastic effect, consists of the carrying
tion functions (see also Emission spectral analy- along of free electrons by dislocations and other
sis). movable defects in regions of plastic flow. The
electron-strain effect takes place under the con-
ELECTRON SPIN-FLIP SCATTERING ditions of a high-rate directed strain of a metal
Quantum transitions of electrons in crystals (with the rate 102 –103 % per second). It can be ob-
served when the twinning of crystals or the motion
between two states differing in the direction of
of slip bands take place, as well as when moving
their spin that occur during the interaction of elec-
a neck or a strain region over the extended billet.
trons with quasi-particles of other types (phonons,
The electron-strain effect is detected either by the
holes), impurities or defects. This spin-flip scatter-
presence of microscopic currents in the strain re-
ing with an exchange of spin orientations occurs
gion, or from the potential difference that compen-
either due to the exchange interaction of electrons
sates these currents. This effect can appear at mod-
with magnetic impurities (see Exchange scatter-
ern drawing and rolling machines. In particular it
ing), or current carriers of other groups, or un-
can cause breakage of steel and copper wire under
der the effect of impurities or crystal lattice vi-
high-rate, very fine drawing.
brations when the spin–orbit interaction is suffi-
ciently strong. This type of scattering of elec-
ELECTRON TEMPERATURE
trons controls the intensity and shape of electron
A value of temperature which specifies the
paramagnetic resonance and combined resonance
mean energy of free electrons (holes) when they
lines, the degree of polarization of particles dur-
are overcooled or are heated up (see also Elec-
ing optical orientation of free current carriers in
tron heating). Its significance and properties de-
semiconductors, and the scale of spin-dependent pend on the ratio between the mean relaxation time
kinetic effects. The resonance growth of the scat- and those of the quasi-momentum τp , energy τE ,
tering cross-section in the vicinity of the Fermi en- electron–phonon scattering τ ph-e , and electron–
ergy level, produced by the exchange interaction electron scattering τ ee (see Current carrier scat-
of conduction electrons with magnetic impurities tering). Under field-stimulated warming-up, while
in metals (the s–d interaction), is the source of the drifting in the electric field with the mean veloc-
Kondo effect. ity v d , the electrons obtain from the field quasi-
momentum p and energy E which they redistrib-
ELECTRON SPIN RESONANCE (ESR) ute between themselves and then pass on to the
The term “electron spin resonance” is widely lattice. For the usual case τ ee τ ph-e the in-
used in the Western, but not in the Russian litera- ternal interaction within the electron subsystem
ture. See Electron paramagnetic resonance. dominates over its interaction with the lattice.
Therefore, in view of the lack of the equilibrium
ELECTRON STATE DENSITY between the electrons and the lattice, a state of
See Density of electron states. equilibrium is established within the electron sub-
system over the momentum p and the energy E
characterized by an average electron velocity u,
ELECTRON STATE, SURFACE
and the electron temperature T e which is higher
See Surface electron states. than the lattice temperature. This approximation
is often called the hydrodynamic approximation.
ELECTRON-STRAIN EFFECT For this case τp < τE < τ ee . This is often in-
Appearance of electron drift (see Current car- troduced with the following relation: mean value
rier drift) under the action of high-rate plastic de- E = (3/2)k B T e + E d where E d = mn vd2 /2 is
ELECTROOPTICAL MODULATORS 399
ELECTROOPTICAL EFFECTS
Variation
n of the refractive index n of light in ELECTROOPTICAL EFFECTS IN LIQUID
crystals located in an electric field of strength E. CRYSTALS
It arises from the orientation in the field of the Variation of the passage of light through a sam-
atoms or polar molecules that compose these crys- ple of a liquid crystal to which a constant or alter-
tals. In crystals without a center of symmetry a nating electric field is applied, due to the orienting
linear electrooptical effect (in respect to the field action of the external field on the director (optic
strength) takes place:
n ∼ E (Pockels effect). axis) (see Twist effect, Guest–host effect, Frederiks
The applied electric field transforms an optically transition).
isotropic piezocrystal into an uniaxial one, and an A device for observing these effects is sketched
uniaxial crystal transforms into a biaxial crystal. in the figure. The unit consists of glass plates 1
Therefore, a consequence of electrooptical effects with transparent electrodes 2 (e.g., made of SnO2 )
is the appearance of induced birefringence. and spacers 3 (made of an insulator such as mica
Electrooptical effects are the physical basis for or teflon) which match the thickness to the liquid
the action of an electrooptic modulator of coher- crystal layer 4. As a rule, this thickness d = 10 to
ent radiation (Pockels cell) with a very short re- 100 μm. The observations can be performed using
laxation time (about 10−13 s). Electrooptical ef- either transmitted or reflected light (lower elec-
fects are used also for the modulation of laser trode can be reflecting). These effects are widely
resonators designed to provide single pulses of used in practice (see Liquid-crystal displays, De-
high-power coherent radiation (giant pulses). A dis- fectoscopy).
advantage of modulators based on electrooptical
effects is the high working voltage (tens to hun- ELECTROOPTICAL MODULATORS
dreds of kilovolts). In centrally symmetric crystals Bulk or integral-optic devices employed in
a quadratic electrooptical effect can appear, with quantum electronics and power optics up to ther-
n ∼ E 2 (Kerr effect). monuclear synthesis for the purposes of amplitude
and phase modulating laser beams; their action is
Half-wave voltage, V (at λ = 0.63 μm) 730 370 185 3000 (at λ = 10.6 μm)
Transmission band, μm 0.35–1.1 0.35–1.1 0.5–4.5 1.0–20
Band width, MHz 0–100 0–100 0–200 0–25
Aperture, mm 3 2.5 2 3
Contrast 100 135 120 135:1
Normalized control power, W/MHz 3.68 1.26 0.19 51.8
400 ELECTROOPTIC MATERIALS
a result of the influence of the electric field on the lenite (nBi2 O3 ·mR) where R is an oxide of ger-
optical properties of the material. The main char- manium, titanium, silicon, aluminum, zinc, lead,
acteristics of some bulk electrooptical modulators etc., and n and m are integers. The molecular crys-
are presented in Table 1. Nonlinear optical materi- tals (CH2 )6 N4 , ammonium chloride, and sodium
als have also been used. chlorate are also electrooptic.
Integral-optic modulation is mainly carried out Non-oxide type ferroelectrics and antiferro-
on plates made of lithium niobate with waveguides electrics include the well-known electrooptic
formed by titanium diffusion or cold welding; the materials Rochelle salt and triglycinesulfate. A nu-
controlling power decreases to 1 μW/MHz as the merous group of electrooptic materials is com-
upper limit of the frequency range is extended to posed of octahedral-oxygen cooordinated ferro-
tens of GHz. This completely satisfies the require-
electrics and antiferroelectrics: crystals of the
ments of multichannel fiber-optic communication
perovskite type (CaTiO3 , BaTiO2 ), double ox-
and control systems, including the integration of
ides like PbO·Nb2 O5 , lithium niobate and tanta-
user terminals with national and international net-
late (LiNbO3 and LiTaO3 ). Lithium niobate (see
works of databases. See also Electrooptic materi-
Niobates) is used in nonlinear optics as a frequency
als.
transformer of coherent radiation, and for the pro-
ELECTROOPTIC MATERIALS
duction of modulators.
Crystals with comparatively strong electroopti-
cal effects. There are linear and quadratic (with ELECTROPLATING
respect to the electric field) electrooptical effects A technique for depositing metals (more rarely
(see Pockels effect, Kerr effect). Of particular prac- non-metals) on the surfaces of solids with the help
tical significance for the transformation and mod- of electrolysis. There are two approaches, adding
ulation of radiation is the Pockels effect which a thin metal protective or decorative coating to
is inherent in piezoelectric materials. The most a surface, and obtaining precise, easily detach-
important electrooptic materials are the so-called able metallic replicas from a host surface. A sim-
KDP compounds of the group potassium dihydro- ple electrodeposition system consists of an elec-
genphosphate KH2 PO4 . There are several isomor-
tric circuit with two electrodes (a cathode and an
phic room temperature crystals containing Rb, Cs
anode) and an electrolyte containing the ions of
and NH4 instead of K, arsenates containing AsO4
the metal (or metals) to be discharged at the cath-
instead of PO4 , and deuterium-substituted com-
ode. The structure and the properties of electrolyt-
pounds such as, e.g., KDHPO4 . At low tempera-
ically deposited metals depend on the crystalliza-
tures several of them are ferroelectrics, and others
tion conditions, and these are mainly determined
are antiferroelectrics. Concerning their structure,
all of them belong to a tetragonal-scalenohedral by the electrolyte composition and the electro-
type of symmetry with space group I 42d, D2d 12 lysis regime. Solutions of complex, in particular
12 cyanic, salts that produce a fine-structure deposit
(Vd ).
are widely used. The loss of ions discharged at
Cubic sphalerite-type crystals can also possess
the cathode is compensated by ions dissolving and
electrooptical properties. Classed among them are
semiconductors of the types AII BVI (ZnS, ZnSe,
thereby entering the electrolytic solution at the
ZnTe, and CdTe) and AIII BV (GaAs and GaP), anode. For applications using anodes that do not
copper and silver halides. Technological problems readily dissolve, the ion loss at the cathode is com-
make it difficult to obtain large, defect-free crys- pensated for by a periodic introduction into the
tals of the majority of these compounds, and this electrolyte of salts containing these lost ions.
significantly restricts their application as basic el- The electrodeposition method is important in
ements of electrooptical modulators. To the class the chemical and electronics industries, and is
of electrooptic materials there also belong cu- used for putting coatings on machinery, as well as
bic system crystals like eulytite [Bi4 (SiO4 )3 ], so- for adding conduction contacts to insulating sub-
dalite [Na4 Al3 (SiO4 )3 Cl], and pleomorphic sil- strates.
ELECTROTHERMAL GRADIENT EFFECT 401
The electrothermal gradient effect is caused from that of a single-particle pattern: the so-called
by a nonuniformity of the thermal expansion, to- Coulomb gap appears in the spectrum. Of crucial
gether with a contribution from phonon states. As importance may also be excitations that are due
a result, this effect is observable not only in ferro- to the rearrangement of the particles which con-
electrics where it reaches a relatively large magni- stitute the framework. These excitations are re-
tude in the vicinity of the Curie point T C , but also lated to quantum transitions of the whole system
in centrally symmetric crystals. among the configurations, which differ from one
another only by the positions of a small number of
ELEMENTARY EXCITATION SPECTRA OF inner core particles. If the energies of these con-
DISORDERED SOLIDS figurations are close together and the energy bar-
Energy spectra of low-lying states of disor- rier separating them is not too high, then the time
dered solids, which do not possess translational of a transition is short compared with the charac-
symmetry. Elementary excitation spectra of dis- teristic observation period, and hence the excita-
ordered solids cannot be described in terms of tions of this type are in equilibrium. The existence
quasi-momentum, and hence are more compli- of the simplest excitations of this type, which are
cated in nature than corresponding spectra of or- called two-level excitations, has been established
dered solids. First of all, there are branches of for several classes of disordered solids (in insu-
the spectrum which correspond to a single-particle
lating and metallic glasses, semiconductors, etc.),
pattern, i.e. the analogues of quasi-particles of an
and results in a variety of thermodynamic and ki-
ordered solid. However, apart from those states
netic features of such systems (see Two-level sys-
with an amplitude, which is of the same order
tems).
of magnitude over the entire body of the solid
The simplest quantity which is characteristic of
(analogues of Bloch functions of an ordered sys-
the spectrum of single-particle excitations in dis-
tem), there also exists a macroscopic number of
ordered solids is the density of states ρ(E). This
localized states in a disordered solid (see An-
is a self-averaging quantity in disordered systems,
derson localization). If the degree of disorder is
considerable, then the contribution of localized which are spatially uniform only on the average,
states may be rather high (it should be mentioned and it exhibits a decrease of correlations at dis-
that all states of one- and two-dimensional sys- tant points. Therefore, spectra of all forms of a
tems are localized, see One-dimensional models macroscopic system coincide, since they consist
of disordered structures, Scaling theory of local- of points at which the nonrandom function ρ(E)
ization), and in this case localized states have a differs from zero.
pronounced effect on the mechanisms of kinetic The spectral limits consist of nonrandom points
phenomena. Thus, the static conductivity of a Eg ; and the density of states goes to zero on cross-
three-dimensional macroscopic disordered sample ing these limits. The spectral limits are differen-
is zero at zero temperature if the Fermi level falls tiated into fluctuation limits and stable ones. The
within the region of localized states (it is always spectral region in the neighborhood of a fluctua-
zero if the sample is one-dimensional or two- tion limit (fluctuation region of the spectrum) owes
dimensional). Such a sample is the so-called An- its existence to improbable fluctuations, i.e. to
derson insulator. Its static conductivity can arise non-typical arrangements of impurities; the spec-
only through a thermal (or other, e.g., tunnel- trum in these regions arising from states that are
ing) activation mechanism at T = 0. Beside quasi- localized at those fluctuations. The location of the
particles, other types of excitations may exist in fluctuation limit depends on the specific form of
disordered solids. For example, correlation effects disorder (type of impurities, correlations in their
were found to play an important role in the sys- arrangement, etc.). The spectrum of energy states
tem of interacting electrons in a random field of in the neighborhood of stable limits arises from
impurity atoms. In the latter case, two neighbor- all kinds of factors, and the density of states con-
ing energy states of the system correspond to dif- verges asymptotically to that of a certain effective
ferent arrangements of electrons over the impu- ordered system. The location of a stable limit, and
rity sites, and the density of states is different the energy dependence of the asymptotic density
ELEMENTARY EXCITATION SPECTRA OF DISORDERED SOLIDS 403
of states in its neighborhood, are insensitive to the +∞), the motion of a particle is quasi-classical;
particular type of disorder. Examples of stable lim- it may be scattered by individual impurities al-
its are the high-energy region of the electron spec- though, as a rule, only through small angles, so
trum, and the bottom of the acoustic band of the a large number of collisions is needed to substan-
phonon spectrum. The spectral limits are in many tially change the particle quasi-momentum. As a
cases singular points with respect to the density result, the initial phase is completely erased from
of states. Other singular points appear only in the the system memory, and hence the particle mo-
limit of minor disorder (the local level E0 (see tion may be described in terms of classical theory.
Local electronic levels), the spectrum limit of the Since the energy is assumed to be much higher
initial disordered system). The rearrangement of than the potential variation relief heights, such
quantum states and the systematics of these states a “diffusion” motion takes place in the classi-
take place in the neighborhood of singular points. cally allowed region and, consequently, the parti-
The disturbance due to disorder in the neighbor- cle is in an infinite or delocalized state, i.e. it be-
hood of these points is not small in terms of the longs to the continuum spectrum. Furthermore, all
effect that it exerts on the states and on the spec- states in the proximity of the fluctuation limit are
trum. The presence of small parameters, in addi- highly localized. Thus, the spectrum in this case
tion to those relative to |E − Eg | and |E − E0 |, necessarily contains both continuous and discrete
often allows the study of these new systematics, components, which are separated by the mobility
and permits one to obtain the asymptotic den- limit Ec . The relative contributions of these com-
sity of states (see Lifshits model, Fluctuation lev- ponents to the spectrum are determined by the de-
els). gree of disorder of the system.
The localized states of a macroscopic disor- The qualitative pattern of states in the neigh-
dered system correspond to a high-density dis- borhood of the mobility limit Ec becomes clear
crete spectrum, whereas the delocalized states when described for the sake of simplicity in quasi-
correspond to a continuous energy spectrum. Mod- classical terms. If a random field due to disor-
erately disordered three-dimensional systems ex- der is at the same time sufficiently smooth, then
hibit a region of delocalized states, which is ei- the quasi-classical approximation is also applica-
ther adjacent to the stable limit, or located at the ble at energies below the peak of the random po-
center of the band. The probability of such states tential amplitude. The regions of classically al-
existing with localized particles of equal energy lowed motion are in this case sufficiently high, and
is rather low, because the generation of discrete hence the particle motion at this energy may be
levels in the continuous spectrum takes place as a described as largely classical in nature. Then the
result of fine interference effects, and is therefore question of whether the quantum state of a certain
highly unstable. Consequently, there exist nonran- energy is localized or delocalized, and whether
dom states Ec (mobility limits, see Mobility edge) the corresponding particle motion is finite or in-
along the energy axis, which separate the region finite, reduces to a question of the geometry of the
of localized states from that of delocalized ones. classically allowed region: whether it is divided
The latter have, in order of magnitude, one and into individual parts of finite dimensions, which
the same amplitude on a set that is a macroscopic are separated from one another, or take up the
part of the whole volume. Such a set, commonly whole space of the system. The value of the en-
called an infinite cluster, may possess a multiply- ergy, at which the transition between the above-
connected, highly branched (fractal) structure (see described geometric patterns takes place, is called
Fractal). Therefore, the delocalized states of dis- the percolation level (in the quasi-classical case it
ordered systems bear little resemblance to Bloch corresponds to the mobility limit) and the broad
states of ordered systems, at least in the neighbor- range of problems, which arise in this connec-
hood of the mobility limit. tion, is the subject matter of percolation theory.
In the three-dimensional case of a system with When the band is finite in width and both limits
a fluctuation limit Eg at sufficiently high energy are fluctuation limits, then there exist two mobil-
(in the neighborhood of the stable boundary, E → ity limits Ec1 < Ec2 , the positions of which will
404 ELEMENTARY PARALLELEPIPED
the electron velocity at the Fermi surface (inte- The classical configuration of an ellipsome-
grated or averaged over the entire Fermi surface ter, an instrument for the analysis of elliptically
S F ). The Eliashberg function g(Ω) is often rep- polarized light, includes a light source (mercury
resented as g(Ω) = α 2 (Ω)F (Ω), where α(Ω) is lamp, laser, or monochromator), polarizer, table
the electron–phonon coupling strength, and F (Ω) with sample under study, compensator (quarter-
is the phonon density of states. In the vicinity of wave plate), and analyzer. Incident linearly polar-
the critical temperature (T → T c ), when the su- ized light (see Polarization of light) reflected by a
perconductor energy gap parameter diminishes to sample becomes, in general, elliptically polarized.
zero (Δ → 0), the linearized Eliashberg equations This is characterized by the polarization angles Ψ
in the framework of Green’s function temperature and Δ which describe the amplitude ratios and the
technique, with the Coulomb repulsion taken into phase difference of the parallel (p) and perpendic-
account, assume the following form: ular (s) components. The main equation of ellip-
Z(iωn )C(iωn ) sometry establishes the relation between angles Ψ
C(iωm ) and Δ, and the optical properties of the reflect-
= πTc Q(iωm − iωn ) − μ∗ ,
ωm
|ωm | ing system. This equation is based on the Fresnel
equations; specific terms depend on the choice of
ωn 1 − Z(iωn )
the model for the surface. It is generally solved
= −πTc Q(iωm − iωn ) sgn ωm , by numerical methods with the aid of a computer.
ωm Since each ellipsometric measurement yields two
where magnitudes Ψ and Δ, one can find two unknown
parameters. In the case of a multilayer reflection
Q(iωm − iωn ) system, when it is necessary to determine a large
∞ number of parameters (refractive indices and ab-
2Ω sorption coef cients of the substrate and film situ-
= dΩ α 2 (Ω)F (Ω) ,
Ω 2 + (ωm − ωn )2 ated on it, film thickness, and so on), one should
0
perform many measurements under various con-
where μ∗ is the Morel–Anderson Coulomb pseudo- ditions: at several angles of incidence (multian-
potential, ωn = (2n − 1)πT c are the Matsubara gle ellipsometry); for several thicknesses of a ho-
“frequencies”, and n = 0, ±1, ±2, . . . . mogeneous film (multithickness ellipsometry); in
A precise numerical solution of the Eliashberg various ambient media (immersion ellipsometry).
equations was first obtained by D.J. Scalapino, To obtain the optical parameters of the system
J.W. Wilkins, and J.R. Schrieffer (1963) who un- from measured angles Ψ and Δ it is necessary
covered an essential dispersion (energy depen- to solve a reciprocal problem. The use of mod-
dence) of both real and imaginary parts of the ern high-speed computers removed the ambiguity
complex gap parameter C(ω), and called atten- problem involved in the interpretation of ellipso-
tion to related anomalies of the tunneling density metric data, thereby putting on a firm basis the
of states. Further, with the help of a numerical so- investigation of the surface state (roughness, ad-
lution of the inverse problem, the method of re- sorption coatings and coatings of other kinds, sub-
constructing the Eliashberg function g(ω) from surface disturbed layers, etc.), the determination
the tunneling current–voltage characteristics of a of the thickness and the optical parameters of thin
superconductor was successfully developed (see (1 nm) films, substrate parameters, and direct
Tunneling spectroscopy of superconductors). studies of various physicochemical processes at
the solid surface. The sensitivity of the method in-
ELLIPSOMETRY, optical polarization method creases sharply under resonance conditions when
Branch of optics which studies solid surfaces surface polaritons are excited with the enhance-
and various multilayer coatings on solids by ana- ment of the electromagnetic wave field near the
lyzing the polarization of a reflected or transmitted surface. This is achieved by using the attenuated
light beam (reflection ellipsometry and transmis- total internal re ection method, or nonplanar (ran-
sion ellipsometry). domly rough or regularly profiled) surfaces. Ad-
406 ELLIPTICITY OF ACOUSTIC WAVE
vantages of ellipsometry are its lack of direct con- EMERSLEBEN TRANSFORM (O. Emersleben,
tact with the surfaces, and its unique sensitivity to 1952)
superthin layers situated at the surface (to ∼0.1 of
Mode of transformation of lattice sums such as
n (here l, l are cell numbers,
a monolayer). l 1/|rsl − rs l |
s, s are the numbers of atoms in a cell) into two
ELLIPTICITY OF ACOUSTIC WAVE rapidly convergent series in terms of the direct and
See Acoustic wave ellipticity. reciprocal lattices; a method similar to the Ewald
method for integers n 1. This is applied, in par-
ELONGATION MODULUS
ticular, for the calculation of the crystal lattice en-
See Young’s modulus.
ergy related to van der Waals forces for n equal to
EMBRITTLEMENT 6, 8, 10, or the lattice energy associated with the
A sharp reduction of plasticity of metals and potential A/r m − B/r n , m > n (m, n are positive
alloys. Embrittlement can be a consequence of a integers; m = 12, n = 6 for the Lennard-Jones po-
high level of impurities (see Impurity atoms) and tential).
nonmetallic inclusions, as well as a result of plas-
tic deformation or heat treatment of metals and al- EMISSION ELECTRONICS
loys. The most common cause of embrittlement is See Vacuum electronics.
the segregation of impurities at grain boundaries,
and the weakening of the grain-to-grain bonding EMISSION OF HOT ELECTRONS
strength. In this case, the failure is of an intercrys- See Hot electron emission.
tallite character (see Crystallites). A major haz-
ard is the segregation of impurities with low sur- EMISSION OF PARTICLES
face energy at grain boundaries, since this leads See Particle emission.
to the lowering of the failing stress. To lessen the
segregation of harmful impurities at grain bound- EMISSION SPECTRAL ANALYSIS
aries, an appropriate heat treatment can be carried Totality of methods for determining the chemi-
out. Such a treatment either stimulates the trans- cal composition of a substance by studying its op-
fer of the impurity to the grain body, binds the tical radiation (emission) spectrum. Qualitatively,
impurity to chemical compounds, or else leads to emission spectral analysis involves the detection
the competitive segregation of elements, unable and identification of the spectral lines of the ele-
to reduce the surface energy, at the grain bound- ment of interest in the emission spectrum of a ma-
aries. Embrittlement in alloys may be a conse- terial. For this purpose, there exist special tabula-
quence of solid solution decomposition (see Al-
tions and atlases of the spectral lines of elements.
loy decomposition) with the formation of Guinier–
This method permits the determination of almost
Preston zones (clusters of precipitate without def-
all the elements of the Mendeleev periodic system.
inite crystalline structure) followed by the precip-
In practice for the emission spectral range (160 to
itation of intermetallic compounds. In metals with
1000 nm) in and near the visible region the ele-
a body-centered cubic lattice embrittlement can
ments of Groups I–II with low atomic number Z
result from the blocking of dislocations by impuri-
are found with the best sensitivity, while halogens
ties with the establishment of atmospheres of vari-
ous kinds (see, e.g., Cottrell atmosphere). Embrit- have the lowest sensitivity. Quantitatively emis-
tlement can be caused by the formation of unfavor- sion spectral analysis is based on the relation-
able texture and macrodefects such as slivers, laps, ship between the intensity of the analytical line
pits, etc. during the course of plastic deformation. IAi of the defined element A and its content CA
In some cases, the embrittlement can be explained in a sample, established according to the semiem-
as a rise of the cold brittleness temperature above pirical Lomakin–Scheibe relation (B.A. Lomakin,
room temperature as a result of structural factors. G. Scheibe, 1932) on reference samples of a com-
position similar to that of the one under analysis:
EMBRITTLEMENT, IRRADIATION IAi = aCA b (a and b are constants depending on
See Irradiation embrittlement. the properties of the analytical line and the light
ENDURANCE LIMIT 407
source). A comparison line is usually chosen from tance, thermal-environment resistance, color re-
an element already in the sample, or the line of an sistance, and weathering resistance. These coat-
element artificially introduced into the sample to ings are single or multiple layers 0.07–0.2 mm
serve as an internal standard. The values IA and thick applied by pouring, spraying, and other
I stand are selected such that the ratio of their in- methods, followed by further sintering and glazing
tensities should be insensitive to a variation of the in special furnaces. Sometimes a paint or varnish
discharge conditions. The relative error of emis- is used that produces an enamel-like coating.
sion spectral analysis for contents 10−2 % is 1
to 10%. ENANTIOMORPHISM, also called enantiomerism
Devices for emission spectral analysis are dif- (fr. Gr. εναντ ιoς , opposite and μoρϕη , form)
ferentiated by their spectral wavelength band, the Property of molecules and crystals to form mir-
type of dispersing element, and the light detection ror image structural units called enantiomers. One
method. Spectrographs with interchangeable dif- of the enantiomers is called left, with the other be-
fraction gratings and photoelectric quantometers, ing designated right. Enantiomorphism influences
provided with microcomputers, have particular ad- morphological, optical, and other structural char-
vantages. acteristics. Thus the left and right modifications of
In emission spectrum analysis, about ten types the structure of quartz rotate the polarization plane
of light sources and their modifications are used. (see Rotation of light polarization plane) of light
The best sources for the analysis of solid non- passing along the optical axis in opposite direc-
conducting materials are arc discharges; for the tions. Enantiomorphism is also inherent in many
analysis of monolithic metals and alloys – glow organic materials. As a rule, living systems are
discharges in a hollow cathode after Grimm, constructed of only one of the two enantiomor-
and low-voltage discharges with the frequency phisms of biological molecules, such as the ex-
500 Hz in an argon jet; for the analysis of partic- clusive use of left-handed aminoacids by proteins.
ularly pure substances (after dissolving them and See also Crystal symmetry, Chirality.
concentrating the analyzed impurities) – a high-
frequency induction-bound plasma; for the analy- ENDOSCOPY
sis of the barely-excitable elements – glow dis- Methods for studying the internal structure and
charges in hollow cathodes and high-frequency properties of opaque objects. X-ray, gamma ray
induction-bound plasmas. In use for layer-by-layer and ultrasonic methods, as well as tomography and
analysis of solid materials, and for the analysis of magnetic resonance imaging (MRI), provide infor-
their inclusions with a volume 200 to 106 μm3 , are mation on internal structure. However, the term
the corresponding pulsed discharges in a cooled endoscopy generally refers to the use of endo-
hollow cathode and lasers. The method of laser scopes which are inserted into, and permit the vi-
fluorescent analysis, where a laser with a tun- sual examination of, parts of the interior of a body.
able frequency serves as the light source, is the
most prospective and efficient method of emis- ENDOSCOPY, EPR
sion spectral analysis. This method possesses ex- See EPR imaging.
tremely high selectivity and sensitivity, unachiev-
able by pre-laser analytical methods, which might ENDOSCOPY, NMR
possibly permit the detection of single atoms (in See Magnetic resonance imaging.
a pregiven quantum state) in bulk material illumi-
nated with the laser radiation. ENDURANCE LIMIT
Maximum value of a stress cycle which a sam-
ENAMELING ple endures without failure, or without intolera-
Application of a thin glassy, opaque ceramic ble strain loading during the course of a preset
coating to a metal surface. Enamel coatings are number of cycles (usually 106 cycles; see Cyclic
notable for hardness, wear resistance, abrasive re- strength). To define the endurance limit a series
sistance (see Abrasive materials), corrosion resis- of tests is performed with different loads, and the
408 ENERGY BANDS
number of cycles of loading until failure is deter- number of equivalent atoms in the crystal; it means
mined. The stress values are plotted on a diagram that in practice the band is occupied continuously
that shows their dependence on the number of cy- (quasi-continuously). A band spectrum of an elec-
cles N which at a given stress level the sample tron in a crystal is represented by alternating bands
had endured until failure (fatigue curve or Wöhler of allowed and forbidden energies (allowed band
curve). For pure metals, or polymers, or in case and forbidden band, or band gap). Even in the case
of corrosion fatigue, the Wöhler curve monotoni- of strong localization of the atomic wave functions
cally decreases with the number of cycles; for less one can speak about narrow bands. For exam-
pure samples and for alloys the Wöhler curve ap- ple, in rare-earth elements atomic f -levels form
proaches a straight line, parallel to the abscissa. a narrow f -band. The quantum theory of metals is
The endurance limit is determined by the nature of based on the single-electron approximation which
the interatomic bonds, by the crystal lattice type, postulates that the electron moves in the periodic
by the structural state, and by other parameters potential arising from nuclei and other electrons in
of the material; it also depends on the conditions the crystal. Stationary states of such electrons are
of testing (amplitude and frequency of loading, described by Bloch functions ψnk (r) (see Bloch
sample size, testing medium). Fatigue tests exhibit theorem), and the energy spectrum (or dispersion
considerable scatter of the results, so it is impor- law) εn (k) where k is the wave vector, and n is the
tant to use statistical methods in evaluating them. band number. The periodicity of the crystal poten-
The endurance limit of the sample in the presence tial causes the periodicity of the energy εn (k) in
of a crack may be characterized by the smallest k -space corresponding to εn (k + b) = εn (k),
value of the cyclic stress intensity ratio that brings where b is a reciprocal lattice vector. Many phys-
about a further spreading of the crack. The en- ically different values of k are situated in the first
durance limit in the presence of a crack may be Brillouin zone. To every value of k there corre-
determined with acceptable accuracy by retesting sponds an infinite number of discrete energy lev-
the same sample. The ability of a metal to function els denoted by the index n. For every n the elec-
after the appearance of a fatigue crack is called its tron energy assumes values within a certain finite
viability. range. The form of the dispersion law depends
upon the crystal symmetry, and the symmetry of
ENERGY BANDS in crystals the wave functions of the individual atoms that
Energy spectrum which results from the quan- compose the crystal. The bands are often sepa-
tum equations of motion which are invariant under rated by energy gaps, but they can also partially
crystal lattice translations. Band spectra are inher- overlap. In the case of band overlapping a super-
ent in electrons, phonons, magnons, excitons, etc. position (degeneracy) of values εn (k) and εn (k)
However, as a rule, the terms energy bands and at the same points of space is possible. The loca-
band theory are used when describing electrons in tions of such points are related to the crystal sym-
a crystal. The formation of electron energy bands metry. The occupation of energy levels by elec-
is connected with the Bragg reflection of electrons trons obeys Fermi–Dirac statistics. In pure semi-
in crystals (see Bragg law). conductors and insulators the Fermi level falls
Electronic energy bands represent the overall within the forbidden gap that separates the va-
spectrum of solutions to the Schrödinger equation lence band from the conduction band. In metals
for an electron moving in the periodic potential of and doped semiconductors the Fermi level is in an
a crystal lattice. The electron bands “genetically” allowed band. In the vicinity of a band minimum
originate from the energy levels of the free atoms or maximum the isoenergetic surfaces of electrons
that come together to form a crystal. In the crystal, (εn (k) = const) assume the shape of spheres or el-
as a result of the interaction between the atomic lipsoids, which means that the dispersion law of
electronic shells, each atomic level broadens and electrons is quadratic with respect to the quasi-
transforms into a band, consisting of a great num- momentum k. Hence, near the band extrema a
ber of levels. This number is equal to the degree band electron is similar to a free electron, but with
of degeneracy of the atomic level multiplied by the an effective mass m∗ defined by the expression
ENERGY TRANSFER 409
εn (k) = 2 k 2 /2m∗ . In diamond (Ge) and zinc- the value Δ(T c ) = 0 at the transition tempera-
blende (GaAs, ZnS, etc.) type tetrahedral semi- ture to the superconducting state T c to its final
conductors the effective mass ratio m∗ /m0 < 1 value Δ(0) at T = 0. In the BCS model this has
for electrons and holes (m0 is the free electron the value Δ(0) = 1.76k B T c . In a superconduc-
mass). In the general case the effective mass is tor with strong electron–phonon coupling (Pb, Nb,
anisotropic, or more precisely, the dispersion law etc.) the ratio 2Δ(0)/(k B T c ) can exceed the the-
is characterized by the inverse effective mass ten- oretical value. A consequence of the existence of
sor. the energy gap is the exponential dependence of
the specific heat and the electronic absorption co-
ENERGY-DISPERSION X-RAY efficient of ultrasound, as well as some other char-
DIFFRACTOMETRY acteristics of superconductors, on the temperature
Method for investigating the structural state of at T < T c . The most direct method of measuring
materials. Energy-dispersion X-ray diffractometry the energy gap utilizes the tunnel effect. At T = 0,
is similar to the Laue method, and consists of irra- the contact tunnel current between a superconduc-
diating the sample by an X-ray beam with a con- tor and a normal metal appears only at the volt-
tinuous spectrum followed by detection of the dif- age eV = 2Δ. In the case of a contact between
fracted radiation with wavelengths satisfying the different superconductors with gaps Δ1 , Δ2 , the
Bragg law by a semiconductor detector (Si(Li) or current produces a jump (at zero temperature and
Ge(Li)). The exposure is carried out with a station- T > 0) at eV = −Δ1 + Δ2 .
ary radiation source, detector, and sample.
The semiconductor detector produces electri- ENERGY LOSS, CHARACTERISTIC
cal signals with amplitudes proportional to the See Characteristic energy loss, Characteristic
quantum energy of the diffracted radiation. Af- energy loss spectroscopy.
ter an analog–digital transformation, the signals
ENERGY LOSS OF CHARGED PARTICLES
are accumulated by a multichannel pulse analyzer,
See Charged particle energy loss.
and reproduce the picture of the diffraction spec-
trum. The interplanar distance d (nm), γ -quantum ENERGY OF SURFACE
energy E (keV) and diffraction angle θ are related See Surface energy.
by the expression d = 0.6199/(E sin θ). Energy-
dispersion X-ray diffractometry allows one to per- ENERGY TRANSFER in crystals
form a simultaneous synthesis of the data from the The vector of an energy flux density (EFD)
required spectral range within a few minutes at the S is numerically equal to the quantity of energy
resolution
d/d ≈ 10−4 . It can be used for the transferred per unit time through a unit surface
purpose of a rapid analysis, for studying the inten- area which is perpendicular to the direction of the
sity distribution of diffracted irradiation with high energy flow at a given point. In a crystal (solid
values of the scattering vector, and the influence of state body) there are as many different EFD vec-
various external actions. tors as there are types of quasi-particles existing
in it. A general expression
for EFD by a j -type
ENERGY GAP in superconductors quasi-particle is Sj = j εj (k)νj (k)nj (k). Here
Basic characteristic of the energy spectrum εj (k), νj (k), and nj (k) are, respectively, the en-
of a superconductor associated with the mini- ergy, group velocity, and distribution function of
mal energy needed for the appearance of quasi- j -type quasi-particles for the same state k. The
particles. The energy of quasi-static excitations EFD associated with lattice vibrations, electro-
εk (in the isotropic model) at T < T c has the magnetic waves, magnetic vibrations, thermal en-
form εk = [ξk2 + Δ2 (T )]1/2 where Δ(T ) is the ergy, etc., can be presented in terms of macro-
energy gap, and ξk is the electron energy rela- scopic quantities: stress tensor σj i , displacement
tive to the Fermi surface. In the Bardeen–Cooper– vector ui , electric and magnetic field strengths E
Schrieffer theory (BSC theory) 2Δ(T ) is the en- and B, respectively, exchange constant A, magne-
ergy needed for breaking up a Cooper pair. The en- tization M, thermal conductivity tensor χik = χki .
ergy gap grows with decreasing temperature from Formulae for these EFD relations are:
410 ENERGY TRANSFER BY EXCITONS
(a) Umov vector of EFD in lattice (1874): form of random jumps or hopping from one lat-
∂uj tice site to another. Such a pattern is characteris-
Si = −σij ; tic of the exciton motion in crystals with narrow
∂t
bands, or at the presence of impurities or defects
(b) Poynting vector for EFD of electromagnetic in large quantities. The value of the diffusion co-
field (J. Poynting, 1884): ef cient D in this case is D ∼ 10−3 cm2 /s. If
Si = [E × H ]i ; excitons are scattered weakly, l
a, then their
motion is related to energy bands, and the dif-
(c) Akhiezer vector for EFD of magnetic waves fusion coefficient is several orders of magnitude
(spin waves, 1958): higher: from 1 cm2 /s (alkali-halide crystals and
cryocrystals) to 103 cm2 /s (semiconductors with
∂ 2M
Si = −A ; “light” excitons). For the diffusion of localized ex-
∂t∂x
citons ln D ∝ T −1 . In the case of free excitons,
(d) thermal energy flux: D ∝ T −1/2 if the transfer is performed by ther-
∂T malized excitons, and D(T ) = const, if the trans-
Si = −χik ; fer occurs via “hot” excitons. If there is a strong
∂xk
exciton–exciton interaction in the crystal, then the
here T is the temperature; and for a crystal probability of energy transfer by many-particle
of cubic symmetry and an isotropic medium bound states: biexcitons, exciton drops, etc. should
χik = χδik . be taken into account.
At diffusion, including thermodiffusion, every
migrating particle carries the transferred heat q, ENHANCED RAMAN SCATTERING OF LIGHT
so the energy flux density Si is given by Si = qJi , A comparatively strong Raman scattering of
where Ji is the particle flux density. The quantity q light by various molecules adsorbed on a metal sur-
may assume both positive and negative values, and face. The scattering cross-section per molecule is
in the context of a microscopic theory q is the 5 to 6 orders of magnitude higher than the scat-
height of a potential barrier which is to be over- tering cross-section of the same molecule in so-
come by the particle during its migration. lution. This is observed on the rough surface of a
metallic electrode in an electrochemical cell or in
ENERGY TRANSFER BY EXCITONS a vacuum, on a metal surface with especially pre-
The transfer of excitation energy from excitons pared relief, in tunnel structures, in metallic granu-
to local impurity and defect centers, as well as lar lms, or on colloidal particles. Despite the great
to the surface states of a crystal. The efficiency diversity of the objects involved, the observed phe-
of such energy transfer depends on the mobility nomena have some common features. First, en-
of the excitation in the crystal. The nature of en- hanced Raman scattering exists for molecules in
ergy transfer by excitons is determined by the re- sharply heterogeneous conditions. The most favor-
lationship among the chief kinetic temporal para- able dimension of a heterogeneity (size of rough-
meters: the lifetime of an exciton in a perfect lat- ness, metallic islands, colloid particles) for en-
tice τ0 , the time of exciton scattering (by phonons, hanced scattering lies in the range 10 to 100 nm.
local centers, etc.) τs , and the lifetime with re- Second, the dependences of the scattering cross-
spect to trapping the excitation on an acceptor τl . section on the exciting-light frequency have the
If τl < τs , τ0 , then the transfer is performed co- form of smooth curves which is evidence of the
herently at a rate equal to the group velocity of an non-resonant nature of the giant enhancement.
exciton in a band. If the scattering time τs turns A possible cause of Raman scattering enhance-
out to be the smallest of these parameters, then ment is the growth of the polarizability of an ad-
the process is diffusive in nature. The mean free sorbed molecule due to its interaction with a metal
path of an exciton in the case of strong scatter- substrate (molecular-chemical amplification). Es-
ing is of the order of a lattice constant (l ∝ a), timates show that in this case the amplification
and energy transfer by excitons takes place in the does not exceed 10–102 . A second possible cause
ENTRAINMENT OF ELECTRONS BY PHOTONS 411
the magnetic field gradient). With the help of sub- at ε = 2. In the theory of systems with continu-
sequent localization of the field E in limited por- ous symmetry groups another version of the ep-
tions of the sample (e.g., using pointed electrodes) silon expansion is often adopted, corresponding to
one can “extract” from an overall EPR signal of an expansion in powers of ε = d − dc , where dc is
the sample the signal from the region of the field the lower critical dimensionality, that is the max-
localization. This problem of obtaining the EPR imum dimensionality at which the thermal fluctu-
signal from a region of field localization can be ations of the order parameter destroy the ordered
also solved by using an alternating electric field as phase, i.e. T c = 0 K. This variant had been pro-
a modulating field (i.e. in the absence of magnetic posed in 1975 by A.M. Polyakov, and it allows one
field modulation). In addition to studying centers to find critical characteristics of three-dimensional
with an ω0 dependence on E, this method also ap- systems as a “rise” in dimensionality from dc = 2
plies for studying reorienting centers for which the (Heisenberg model) or dc = 1 (Ising model). In
field E changes the intensity of various EPR lines. addition to its application in the theory of phase
For a modulating field E, the EPR signal is pro- transitions, the epsilon expansion is used in per-
duced only by portions of the sample where the colation theory, the theory of localization, and in
field E is localized. some other areas of condensed matter theory.
residual austenite. The hardness of the marten- the temperature and pressure dependences of elas-
site ε-phase is higher (to 40%) than the austen- tic moduli, or by other relationships, depending on
ite hardness, the yield limit is 2 to 3 times higher, the process under investigation.
and the plasticity is lower than the correspond-
ing values for austenite. The maximum amount EQUILIBRIUM OF PHASES
of martensite ε-phase in iron–manganese alloys is See Phase equilibrium.
70 to 80%, and at strain it reaches 100%. Upon
plastic deformation or hydroextrusion the ε-phase EQUILIBRIUM SEGREGATION
transforms into alpha martensite. The repeated A metastable state of a solid containing impu-
transition of austenite into the ε-phase and the rities and their traps (dislocations, stacking faults,
inverse transformation lead to phase wear hard- grain boundaries, phase interfaces, etc.). In such a
ening, i.e. a type of hardening of the alloy. Mul- state, the content of impurity atoms differs in the
tiple γ ↔ ε transitions in iron–manganese al- material matrix and at its structural defects. It is
loys do not change the hexagonal symmetry of determined by the conditions of minimum ther-
the ε-phase, and the addition of 0.25 to 0.60% modynamic potential yielded by the impurity con-
of carbon to them leads to the formation of an centrations at traps, and by the free parameters of
18-layered martensite phase. The ε-phase can be structural defects (e.g., by the width of the stack-
detected with the help of X-ray structure analysis ing faults between partial dislocations). Depending
and metallographic analysis. on the value of the bonding enthalpy between the
impurities and the traps one may consider com-
petitive segregation of the various impurities to a
EQUATIONS OF STATE
given trap, and competition between various traps
Equations which describe the relationship be- in the process of pinning a given impurity. In case
tween the variables that characterize the state of a the number of atoms of an impurity is smaller
thermodynamic system, such as a gas, a liquid, or a than or comparable to the number of the positions
crystal. Thermal equations of state relate intensive available for them at traps, such competition may
quantities (independent of system volume) to ex- lead to nonmonotonic temperature dependences of
tensive ones (proportional to the system volume), the content of a given impurity at a given trap.
and to the temperature T , e.g., f (P , V , T ) = 0, Such dependences exhibit a minimum at a temper-
where P is pressure, V is volume. Caloric equa- ature T = Tmin and a maximum at a temperature
tions of state represent thermodynamic potentials T = Tmax > Tmin . Equilibrium segregation lowers
(free energy, enthalpy, etc.) or entropy as functions the impurity content in the matrix and, hence, low-
of pressure and temperature (or V and T , etc.). ers the temperature of dissolution of carbides, ni-
The relation between thermal and caloric equa- trides, and other phases. Equilibrium impurity seg-
tions of state is established by the application of regation affects the strength and plastic properties
fundamental laws of thermodynamics. Equations of materials (see Strength, Plasticity), and the local
of state, which serve as a necessary connecting electrophysical properties of semiconductors.
link that allows applying thermodynamic laws to
particular systems, cannot be derived from funda- EQUILIBRIUM, THERMODYNAMIC
mental thermodynamic laws alone. There are no See Thermodynamic equilibrium.
universal equations of state for solids, rather the
role of these equations is played by certain approx- ERBIUM, Er
imate expressions obtained on the basis of sim- Chemical element of Group III of the peri-
ple theoretical models or experimental data. These odic system with atomic number 68 and atomic
theoretical models are, e.g., the Debye theory of mass 167.26; belongs to lanthanides. There are
speci c heat, and the lattice model of crystal vi- 23 known isotopes with mass numbers from 151
brations developed either with or without regard to 173; naturally occurring isotopes are: 162 Er
for anharmonicity (see Crystal lattice vibrations). (0.14%), 164 Er (1.61%), 166 Er (33.6%), 167 Er
The role of empirical laws may be played by, e.g., (22.95%), 168 Er (26.8%), 170 Er (14.9%). Outer
ETCHING, ION 417
shell electronic configuration is 4f 12 5d 0 6s 2 . (18.0 K < T C < 19.6 K) a first-order phase tran-
Atomic radius is ≈0.1754 nm; radius of Er3+ sition yields (“freezes out”) the sinusoidal mod-
ion ≈0.0886 nm. Oxidation state +3, electro- ulation of the magnetic moments with respect to
negativity ≈1.2. the 6 axis: the magnetic moments are oriented at
Erbium is a silvery-white metal. At low tem- the angle ∼29.6◦ to the 6 axis, and the perpendic-
peratures erbium has a hexagonal close-packed ular components conserve the helicoidal ordering
crystal lattice, space group P 63 /mmc (D6h 4 ), in the basal planes (i.e. cone-like structure of the
with parameters a = 0.35592 nm, c = 0.5585 nm ferromagnetic type of spiral is formed). Magnetic
(at room temperature). At ≈1690 K the poly- susceptibility of erbium is +44500·10−6 CGS
morphous transformation of erbium into a body- units; nuclear magnetic moment of isotope 167 Er
centered cubic modification is possible, space (22.95% abundant) is 0.48 nuclear magnetons.
group P n3n (Oh2 ), with parameter a = 0.3940 nm Pure erbium is used for the purposes of investiga-
(extrapolated value) after hardening. Density is tions, and its alloys (with cobalt, iron and nickel)
9.051 g/cm3 at 293 K, T melting ≈ 1790 K, are used for manufacturing permanent magnets,
T boiling ≈ 2865 K. Heat of melting is 17.2 etc.
kJ/mole, heat of sublimation is 280.5 kJ/mole,
heat of evaporation is 271 kJ/mole; specific heat EROSION, DROP-IMPACT
is 0.168 kJ·kg−1 ·K−1 at room temperature. De- See Drop-impact erosion.
bye temperature is ≈180 K; linear heat expan-
sion coefficient of erbium polycrystal is
EROSION, SURFACE
≈10.8·10−6 K−1 ; thermal conductivity coeffi-
cient is 9.6 W·m−1 ·K−1 . Adiabatic elastic mod- See Surface erosion.
uli rigidity of erbium polycrystal: c11 = 83.88,
c12 = 29.26, c13 = 24.21, c33 = 84.32, c44 = ESAKI DIODE
26.45 (GPa) at 300 K; adiabatic bulk modulus is See Tunnel diode.
44.913 GPa, Poisson ratio is 0.238. Vickers hard-
ness is ≈55 HV; tensile strength is 0.280 GPa. ETCHING
Activation energy of self-diffusion of atoms in A method of chemical or electrochemical treat-
erbium polycrystal at the melting temperature is ment of the surface of solid materials aimed to re-
300 kJ/mole. Effective thermal neutron trapping move dirt and oxides, to reveal the material struc-
cross-section is 162 barn. Resistivity of erbium ture, to produce a desired microstructure of the
monocrystal (at 298 K) is 750 n·m along the surface, to remove the surface layer disturbed by
principal axis 6 and 440 n·m in the perpen- the mechanical processing, and so on. Chemical
dicular direction, and that of erbium polycrystal etching is brought about by the direct interaction
is ≈940 n·m, thermal coefficient of electric re- of the material surface with the etching agent solu-
sistance is 0.00201 K−1 . Work function of er- tion, or with an appropriate reagent in the gaseous
bium polycrystal is 3.12 eV. At the Néel point phase. At electrochemical etching, the treated sur-
T N = 84.4 K erbium transforms from paramag- face is submerged into an electrolyte solution and
netic state to antiferromagnetic state with sine used as an electrode when passing direct or alter-
wave-modulated component of magnetic moment nating current through the solution. See also Se-
along the principal 6 axis and with disordered lective etching, Ion etching.
components of magnetic moment in the basal
plane. Below the critical temperature T c = 52.4 K
ETCHING GROOVE, THERMAL
the modulation of the magnetic moment along the
principal axis 6 is conserved, but there appears See Thermal etching groove.
helicoidal ordering of the perpendicular compo-
nent of the magnetic moment in the basal plane (of ETCHING, ION
the complex helical type). At the Curie point T C See Ion etching.
418 ETCHING, SELECTIVE
EXCHANGE FIELD of a magnetic material part should be antisymmetric and, accordingly, for
The characteristic magnetic eld associated S = 0, its spin part is antisymmetric and its co-
with the exchange interaction. The order of magni- ordinate part is symmetric. The difference in the
tude of the magnetic exchange field Be is J /(2μB ), symmetries of the coordinate parts of the wave
where J is the exchange integral, and μB is the functions results in a difference between the cor-
Bohr magneton. For magnets with several mag- responding particle interaction energies (with each
netic sublattices, Be is a basic material constant other and with other charges, such as atomic nu-
defined as the magnetic field magnitude at which clei in a molecule). The exchange integral may be
the spin-flip magnetic phase transition occurs, either positive (parallel spin orientation, S = 1, is
e.g., where the magnetizations of sublattices be- more favorable in energy) or negative (state S = 0
come aligned. The exchange field for various an- is favored in energy). Here are some of the sim-
tiferromagnets varies from a few dozen tesla (for plest examples. The Coulomb repulsion of sev-
low Néel points TN of several kelvins) to a few eral electrons in a partly filled atomic shell (e.g.,
kilotesla (for TN up to kelvins in the hundreds). a transition atom d- or f -shell) is minimized by
The value of B e is relatively small for metamag- the “most antisymmetric” wave function, so that
netic media (several tesla). J > 0 and the shell has its maximum possible
spin value (Hund rules). For electrons of covalent
EXCHANGE INTEGRAL bonds (e.g., in the molecule H2 ), a configuration
See Exchange interaction. with a symmetric wave function is energetically
favored, so the electron density clusters in the re-
EXCHANGE INTERACTION (W. Heisenberg, gion between two charged nuclei. Because of that,
1926; W. Heitler, F. London, 1927; P.A.M. Dirac, J < 0 and the net spin of a covalent bond is zero.
1929) The exchange interaction of electrons of different
A specific mutual interaction between identi- atoms is appreciable when the atomic wave func-
cal particles whereby the system energy depends tions of these electrons overlap. The exchange in-
on its resultant spin. The exchange interaction is tegral is of the order of ξ(e2 /a B ) where a B is the
a purely quantum effect, involving Fermi–Dirac Bohr radius, and the factor ξ which describes the
statistics or Bose–Einstein statistics (for fermions extent of overlap of the wave functions (usually
and bosons, respectively). It can involve particles ξ 10−2 ) falls off exponentially with increasing
interacting either with each other (e.g., via the interatomic distance. The exchange interaction of
Coulomb interaction) or with external fields. The electrons localized on different atoms is often re-
Hamiltonian of the exchange interaction He for duced to that of atomic spins. Equation (1) also de-
two electrons with spins S 1 and S 2 has the form: scribes the exchange interaction between the total
spins of unfilled atomic shells, S 1 and S 2 , but such
H e = −J S 1 · S 2 , (2)
a treatment is not rigorous in the case S > 1/2.
where J is called the exchange integral. It is To pursue a rigorous approach, additional terms
the difference between the interaction energies must be introduced, e.g., biquadratic in the atomic
of electrons, US , for the two states with differ- spins (so-called biquadratic exchange interac-
ent total spin values S (S = 1 or S = 0), J = tion). The exchange interaction is often related
U0 − U1 . Within the framework of nonrelativistic only to the electrostatic electric interaction of
quantum mechanics (expressing the wave function electrons with each other and with atomic nu-
as a product of spin and space coordinate terms), clei. In that case it includes the scalar products
the origin of the exchange interaction, i.e. why of spins only, and remains invariant with respect
U1 = U0 , is found in the Pauli principle. Accord- to arbitrary spin rotations. When one considers
ing to it, the wave function for fermions (elec- other types of interactions (e.g., spin–orbit inter-
trons in our case) is antisymmetric with respect action) anisotropic terms of the form Jαβ Sα Sβ
to the interchange of particle variables (both coor- may appear in the exchange interaction Hamil-
dinate and spin). Since the spin part of the wave tonian, where α, β = x, y, z (see Spin Hamil-
function is symmetric for S = 1, its coordinate tonian). In the case when Jαβ Sα Sβ = J S · S, one
EXCIMER 421
speaks of the anisotropic Dzialoshinskii–Moriya calculated by taking into account the specific elec-
exchange interaction (see Dzialoshinskii interac- tronic structure of the paramagnetic center.
tion), the so-called antisymmetric exchange inter- Exchange scattering is responsible for such
action. This is described by antisymmetric tensor phenomena as the spin relaxation of paramag-
(a) (a)
components Jαβ , Jαβ = −Jβα ; in order of mag- netic centers, conduction electrons, and excitons
nitude J (a) ∼ (
g/g)J , where
g is the change in semiconductors (see Paramagnetic relaxation),
of the g-factor due to the spin–orbit interaction. electronic polarization, the formation of local-
The symmetric exchange interaction is described ized magnetic moments in metals, and the indi-
(s) (s) rect exchange interaction of atoms through con-
by symmetric components Jαβ = Jβα , and deter-
duction electrons (see Ruderman–Kittel–Kasuya–
mines both the ordinary Heisenberg exchange in-
Yosida interaction), as well as some other phe-
teraction, and the exchange anisotropy. The Ising
nomena which involve participation of electron
model is a limiting case of the anisotropic ex-
spins.
change interaction. Non-Heisenberg exchange in-
teraction terms are particularly large for rare-
earth element and actinide ions. There is also EXCHANGE SPLITTING
an indirect exchange interaction involving either An energy difference between levels differing
an intermediate (nonmagnetic) ion, or conduction in the value of the resultant spin S for a dou-
electrons (see Ruderman–Kittel–Kasuya–Yosida in- blet, triplet, or more complicated complex involv-
teraction). The exchange interaction plays an im- ing exchange-interacting paramagnetic ions. For
portant role in the binding energy of atoms, mole- an isotropic exchange interaction characterized by
cules, and solids (especially metals and covalent the parameter J , the position of the energy level
crystals); it explains trends in the spectroscopy of is found to be ES = J S(S + 1)/2, and this differ-
solids (see Exchange splitting, Exchange-coupled ence is easy to calculate. The quantity J was first
pairs), and underlies the physics of magnetism. measured and calculated for a helium atom (ortho-
and parahelium), and for a hydrogen molecule.
EXCHANGE INTERACTION, INDIRECT
See Indirect exchange interaction. EXCIMER
An excited dimer formed by two atoms or two
EXCHANGE MAGNETOSTRICTION molecules in an excited state. Such an elemen-
See Magnetostriction. tary excitation in an organic compound is char-
acterized by chemical bond formation between
EXCHANGE SCATTERING neighboring molecules as a result of the excita-
Scattering of electrons (see Current carrier tion of one of them, when no bond exists be-
scattering) with a change in the projection of fore the excitation. The crystals where the ex-
their spin angular momentum. The presence of cimers are formed exhibit monomer absorption
exchange scattering involving identical particles (monomer) with broad structureless bands in u -
makes it necessary to antisymmetrize their elec- orescence spectra which are specific for pairs of
tronic wave functions. To describe the scatter- molecules called dimers (see Exciplex). A typical
ing of delocalized band electrons by local para- example of an organic compound where excimers
magnetic centers when the wave vector of the form is pyrene (C16 H10 ). There are four mole-
scattering electron is smaller than the recipro- cules in the unit cell of pyrene with the molecules
cal radius of the local center state, the exchange arranged in pairs, thereby satisfying the conditions
scattering makes use of a zero-radius potential for the formation of an excimer state. Excimers are
= [2π2 /m∗ ]aS · S e δ(r) (where S e is the spin
V observable in gaseous, liquid and solid states of or-
of the scattering electron, m∗ is its effective mass, ganic molecules. Critical parameters for excimer
S is the spin of the paramagnetic center, and a is formation are the distance between molecules, and
the exchange scattering length that is either con- the possibility of their interaction at the optical ex-
sidered as a phenomenological parameter, or is citation of one them. Excimers play a significant
422 EXCIPLEX
role in the optical properties of many organic com- oscillator strength is small then such mixing is in-
pounds, in particular, in processes of photogenera- significant.
tion of current carriers in them. Excimers are used There are two kinds of excitons: the Frenkel
in excimer lasers. exciton (small-radius exciton) and the Wannier–
Mott exciton (large-radius exciton). A Frenkel ex-
EXCIPLEX citon is basically an excited electronic state of an
An excimer (excited dimer) compound formed ion. It can move around the crystal by transferring
as a result of an interaction between two dif- the excitation from ion to ion, so no ions need to
ferent molecules, one of which is in an excited change places. A example of a Wannier–Mott ex-
state (dimer in excited state, see Excimer). Such citon is an electron in the conduction band bound
an excited molecular charge complex is unstable to a hole in the valence band by the Coulomb inter-
in the ground state. It may form from an excited action V (r) = −e2 /(4πεr), where the dielectric
state acceptor molecule interacting with a ground constant of the crystal ε is greater than that ε0 of
state donor, and vice versa. Exciplex uorescence free space, typically by a factor of 10. The energy
arises in lattices constructed from physical dimers, levels are the same as the Rydberg series levels of
donor and acceptor molecules, which do not inter- a hydrogen atom, but with energy values reduced
act in their ground (unexcited) states. by the ratio (ε0 /ε)2 :
me4
EXCITATIONS, COLLECTIVE E = Eg − ,
See Collective excitations. 22 (4πε)2 n2
where E g is the energy gap, and n is the principal
EXCITON quantum number. The exciton Bohr radius aB =
An elementary collective electronic excitation 4πε2 /(me2 ) exceeds that of a hydrogen atom by
in nonmetallic crystals, free of current, and ex- the ratio ε/ε0 . There also exist excitons of an in-
isting in the crystal as a wave of excitation. The termediate type with a binding energy E b ∼ 1 eV
exciton concept was introduced by Ya.I. Frenkel and a small radius (R = 10−7 –10−8 cm) in alkali-
(1931). Exciton energy states involve a band with halide crystals, and in crystals of noble gases. The
a particular dispersion law that depends on the exciton radius is the mean distance between the
crystal properties. A specific feature of the band excited electron and its non-excited state that re-
structure of dipole-active excitons is its splitting mained vacant (hole) and bound in the exciton
into branches of longitudinal and transverse exci- wave.
tons (with regard to the angle between the vectors An exciton existing in the form of a wave (so-
of the wave and the dipole moment). The value called free exciton or coherent exciton) can be
of the transverse–longitudinal splitting is propor- trapped by a lattice structure defect and bond with
tional to the square of the dipole moment of the it, thus forming a bound exciton. The energy levels
exciton transition, and it determines the range of of bound excitons are discrete and situated lower
existence of long-wave excitons (those with small than the band of free excitons for the same value
wave vector magnitude). In optical processes the of binding energy with defects.
main excitons that participate are those with small The exciton absorption spectra in perfect crys-
wave vectors which correspond to the magnitude tals are represented by a series of discrete bands
of the light wave E vector. In most cases excitons whose width and shape are determined by the
are induced by an electromagnetic wave. If the ex- structure of the exciton band, and the character of
citon transition has a large oscillator strength, the the interaction of the excitons with photons and
quanta of the electromagnetic field in the medium phonons. When Wannier–Mott excitons are ex-
(photons) are mixed with the quanta of “me- cited with the help of a low-power light source,
chanical” (electronic) excitations, thereby forming their concentration is small (to 10−14 cm−3 ); this
exciton-induced polaritons (photoexcitons). In this state can be regarded as a gas of non-interacting
case there is no longer a sharp difference between quasi-particles. Under intensive laser excitation
the excitons and the photons in the crystal. If the high-density excitons are produced. In this case the
EXCITON LUMINESCENCE 423
density is so high that the exciton size is compa- many-electron atoms. The shell degeneracy is de-
rable with the inter-exciton distance; their interac- termined by the multiplicity (by principal and or-
tion becomes significant and leads to the appear- bital quantum numbers). The electrons and holes
ance of exciton molecules (biexcitons), which ei- are in a self-consistent field whose symmetry co-
ther condense or decay with the formation of an incides with the symmetry of the impurity center.
electron–hole plasma, or drops of an electron–hole Therefore, the structure and the degeneracy of the
liquid. Excitons play an essential role in processes shells are the same as in simple donors and accep-
of excitation energy transport in crystals (see En- tors.
ergy transfer by excitons), in radiation processes, The description of the energy spectrum and
for laser generation in homogeneous semiconduc- shell filling based on the single-electron approach
tors, and so on. is a crude approximation that fails to explain
See also Biexciton, Dislocation exciton, Frenkel the fine structure of the spectra of the exciton–
exciton, High-density excitons, Photoexciton, Poly- impurity complex caused by the electron–electron
excitons, Self-localized exciton, Surface exciton, (hole) correlation.
Trion, Triplet excitons, Wannier–Mott exciton, X-ray
excitons. EXCITON LUMINESCENCE
Luminescence in which the radiation arises
EXCITON, ENERGY TRANSFER from transitions involving excitonic energy lev-
See Energy transfer by excitons. els. The exciton luminescence is observed in most
crystal phosphors and organic luminophors in the
EXCITON–IMPURITY COMPLEX form of spectrally narrow (in comparison with
Bound state of two or more excitons with a other kinds of luminescence) bands (lines) with
small neutral impurity center (donor or acceptor) line positions and shapes which depend on the
in indirect-gap semiconductors (e.g., Si and Ge). temperature, the level, and the mode of excitation,
The stability of exciton–impurity complexes re- the presence of defects in the material, the action
sults from the degeneracy of the bands. In a direct- of external electric and magnetic fields, deforma-
gap semiconductor with simple bands, at the cen- tions, and so on.
ter point Γ of the Brillouin zone, there is a 1s A radiative quantum transition that produces
orbital state which can only hold two electrons exciton luminescence is exciton recombination oc-
with opposite spins. In indirect gap Si which has curring with the emission of light. The energy
six electronic valleys situated at the points Δ, of the emitted quantum is close to the energy of
twelve electrons with antiparallel spin orientations the corresponding exciton. Many possible exci-
or Bloch functions can coexist in this state. The ton states (exciton bands; bound excitons; exci-
valence band of Si at an extremum point is four- tons due to self-localization) result in the emission
fold degenerate; hence four holes with angular of a few bands (lines) specific for the exciton lu-
momenta projections jz equal to ±3/2, ±1/2 can minescence of each individual material. The exci-
be in the same orbital state. In radiative recom- tons of the exciton band, under action of the ex-
bination spectra at high excitation densities the citation, can spread the energy over a broad con-
exciton–impurity complexes are observed as a se- tinuous range. However, the corresponding exci-
ries of narrow lines. The long-wave lines of the ton luminescence bands are usually narrow. This
series appear at high levels of excitation. An indi- is due, first, to selection rules which provide a suf-
vidual series is associated with each impurity cen- ficient emission probability only for those elec-
ter. trons with a certain energy, and second, to the non-
In the single-electron approximation the elec- uniform distribution of excitons over the energies
tronic structure of an exciton–impurity complex of the band. If the time delay between the excita-
is described in terms of a shell model: the elec- tion and the emission is sufficient for a great num-
trons and the holes of the complex fill the electron ber of collisions of excitons with phonons to take
and hole shells in accordance with the Pauli prin- place, then the excitons become thermalized, i.e.
ciple, similar to the way electron shells are filled in the energy no longer depends on that acquired at
424 EXCITON MECHANISM OF SUPERCONDUCTIVITY
excitation, they will occupy a range of states with of bands, their transformation into other types of
the thermal energy k B T while passing their excess bands, e.g., into an emission band of an electron–
energy to the lattice; consequently, the exciton lu- hole plasma, etc.).
minescence will consist only of quanta with the In most cases the strong temperature depen-
same energy. The dipole approximation selection dence of luminescence quenching is specific for
rules only permit radiative transitions of excitons the exciton luminescence. In this case processes
with the quasi-momenta close to zero. The exci- competing with exciton luminescence may be ei-
ton states related to these quasi-momenta are all ther dissociation of excitons into pairs of non-
within a narrow range of the possible energies that bound current carriers, or recombination with the
will appear in the emission. emission of many phonons.
In the general case, both the kinetics of the
change of the exciton energy after its excita- EXCITON MECHANISM OF
tion (i.e. its relaxation), and the selection rules SUPERCONDUCTIVITY
(taking into account interactions of excitons with Hypothetical non-phonon mechanism of high-
temperature superconductivity which assumes that
phonons, plasmons and other units in the crystal)
Cooper pairs of degenerate free carriers (conduc-
are more complicated. For example, for a strong
tion electrons) form through the exchange of vir-
interaction of excitons with the crystal electro-
tual quanta of collective excitations of bound elec-
magnetic field, the relaxation of exciton polari-
trons in atoms and molecules of solids (i.e. by
tons is characterized by a sharp dependence of the
excitons). The mechanism had been put forward
mean free path on the polariton energy, and the
by W. Little and independently by V.L. Ginzburg
probability of emission is determined, in addition,
(1964). Little considered analytically a quasi-one-
by the conditions of transformation of polaritons
dimensional model of conductive polymer chains
into phonons at the surface. The exciton energy
in organic (metal-organic) compounds, in which,
distribution is determined in this case by a com-
as a result of the Cooper interaction, the conduc-
bination of probabilities from different modes of
tion electrons polarize the electron shells of mole-
relaxation. The dependence of the probability of cules (radicals), thereby reducing their mutual re-
emission on the exciton energy also becomes com- pulsion. The result of this polarization is an effec-
plex and nonmonotonic. All these factors affect tive interaction between the electrons in the chain
the shape of the exciton luminescence bands, thus which takes the form:
making it possible to determine from experimen-
tal data the relaxation mechanisms, their probabil- 4πe2
V (q, ω) = 2 ,
ities, and their selection rules. q ε(ω)
(1)
The application of external fields can change Ωp2 f0
the internal energy of excitons, thereby causing ε(ω) = 1 + 2 ,
ω0 − ω 2
them to shift, split, and broaden the exciton lumi-
nescence bands. For high excitation levels when where ω and q are the transferred energy and
a high-density of excitons and other contributors momentum, respectively; ε(ω) is the dielectric
(current carriers, nonequilibrium phonons) appear, constant of organic molecules surrounding the
their interaction also changes the internal energy conductive chain; ω0 and f0 are respectively the
of the excitons, and correspondingly, the exciton frequency and oscillator strength of the optically
luminescence spectrum. As a result, the changes allowed dipole transition; Ω p is the plasma fre-
involved in combining excitons in “exciton mole- quency of bound electrons. In the frequency range
cules” (biexcitons, drops of electron–hole liquid 1/2
and exciton liquid) can be detected. The change ω0 < ω < ω02 + Ωp2 f0 (2)
of the exciton energy, or the decay of the exciton the interaction V (q, ω) of Eq. (1) is negative cor-
states as a result of the interaction with the current responding to an attractive interaction involving
carriers (electric charge screening), is also accom- the exchange of virtual excitations (excitons) be-
panied by corresponding changes in the exciton lu- tween molecules. It was supposed that this attrac-
minescence spectra (appearance or disappearance tion, owing to the high energy of the excitons
EXCITON–PHONON INTERACTION 425
(ω0 1 eV), could lead to superconductivity an orbital eg -doublet, this ion is a two-level sys-
with a high critical temperature (T c 100 K) if tem with the splitting of the levels due to a low-
one accepts the validity of the BCS formula for the symmetry component of the ligand field. When
transition temperature k B T c ≈ ω0 exp(−1/λex ) traveling through the lattice, the quasi-free hole
(see Bardeen–Cooper–Schrieffer theory) with the described by hybridized p–d-states can excite a
assumption that the dimensionless electron–exci- transition between the two eg -levels of the hole
ton interaction constant is not small, i.e. λex 1/2. localized at the Cu2+ ion. This transition cor-
As the range of attraction (2) is a narrow peak responds to the formation of a quadrupole-type
situated far from the Fermi level, under realistic Frenkel exciton. At the same time d–d transitions
conditions the optimal values are λex 0.1, and can also be excited that are linked with the change
consequently T c cannot exceed 1 K. At present of the spin projection, and which in turn, could be
a series of quasi-one-dimensional (chain) organic one of the causes of breaking the long-range anti-
crystals with metallic conductivity (see Quasi- ferromagnetic order in the presence of charge car-
one-dimensional crystals) that transform into the riers. By treating the d-excitons as carriers of the
superconduction state at T ≈ 10 K have been inter-hole interaction, one might approach a theory
synthesized (see Organic conductors and super- similar to the Bardeen–Cooper–Schrieffer theory
conductors), but there are no convincing exper- with a relatively high superconducting transition
imental proofs of the operation of the exciton temperature due to the high frequency (to 0.5 eV)
mechanism of superconductivity in such systems. of the d–d excitations.
V.L. Ginzburg considered a possibility for this
mechanism in quasi-two-dimensional layered struc- EXCITON MOLECULE
tures of the “sandwich” type (semiconductor– See Biexciton.
metal–semiconductor) when the Cooper pairing of
electrons in ultrathin metal film of d ≈ 1 μm thick EXCITON–PHONON INTERACTION
takes place due to the attraction caused by their A dynamic process perturbing exciton states.
interaction with virtual excitons in the semicon- This is described by a matrix element V e-ph which
ductor plates. There have been reports about the depends on the mechanism involving:
observation of superconductivity in monatomic
films made of silver and gold at the surface of (i) the interaction of excitons with acoustical
phonons by means of the strain potential:
germanium monocrystals, or in multilayer struc-
tures Au–Ge, but it was not been demonstrated
|q|D 2
that the exciton mechanism acts in this case. Ve-ph = ; (1)
In recent years the interest in this mechanism 2Vρu
has grown in connection with the cuprate high- (ii) the interaction of excitons with acoustical
temperature superconductors (T c ≈ 30 to 165 K) phonons resulting in a piezoelectric effect
which have a layered crystal structure (see Quasi- (non-centrally symmetric crystals):
two-dimensional crystals) and a complicated stron-
gly anisotropic band spectrum whose particular 4πeεij
Ve-ph = (qe − qh ); (2)
feature is intensive (corresponding with high os- 2ρu ε0 q 1/2
cillator strength) interband transitions which prob-
ably enhance the Cooper coupling and thereby (iii) the interaction of excitons with phonons of
raise T c . There is one more branch of excitations the Fröhlich kind (see Fröhlich Hamiltonian):
of the boson (exciton) type in copper-containing
layered metal-oxides which is associated with a q ε∞
Ve-ph = π −1
particular electronic structure of bivalent copper 2μ ε0
(configuration 3d 9 ). The carriers are capable of
3
exchanging these excitations. Since in a cubic me − mh 2 aex
crystal e ld the d-state of the Cu2+ ion forms
× ω, (3)
me + mh V
426 EXCITON SPECTROSCOPY
initial stages of the breakdown of metals and non- (1) energy hν, polarization P , and angle of inci-
metals, including polymeric materials and com- dence α of the photon;
posite materials, and for the investigation of ad- (2) kinetic energy E, quasi-momentum k, spin σ ,
sorption processes and chemical reactions at the emission angle, initial state energy Ei and fi-
surface. nal (excited) state energy Ef of the electrons;
(3) crystallographic structure, orientation of the
EXOPHOTON EMISSION surface, electron af nity and electron work
Emission of photons from a solid surface dis- function, surface concentration of adsorbed
turbed by either mechanical processes, radiation, atoms, atomic structure of adsorbed lm s,
or another action such as bombardment by atomic, temperature and other characteristics of the
molecular, or ionic beams, as well as by some material under investigation.
physicochemical process (adsorption, recombina- The diversity of experimental parameters de-
tion of active particles, etc.). Exophoton emission termines the diversity of photoelectron spectro-
takes place as a result of energy released during scopy methods: Ultraviolet Photoemission
the recombination of active particles which are Spectroscopy (UPS), X-ray Photoemission Spec-
formed under the above-indicated processes and troscopy (XPS) and synchrotron photoemission
interactions. Exophoton emission is a surface ef- spectroscopy are named in accordance with the
fect which is strongly influenced by the pressure energies of the exciting radiation. The exper-
and the chemical nature of gases in the ambi- imental parameter of Angle-Integrated Photoe-
ent atmosphere. The photon emission centers are mission Spectroscopy (AIPES) is the angle of
situated either at the surface or in a thin near- emission; whereas Angle-Resolved Photoemis-
surface layer of the solid. This emission can be sion Spectroscopy (ARPES) and normal Photoe-
employed for studying surface defect formation, mission Spectroscopy (PES) take into account
material failure, and physicochemical processes at the angular resolution. The spectroscopic meth-
the surface (including strain and fracture of metals, ods which deal with energies of initial and final
disintegration of materials, sputtering, chemisorp- electron states are Constant-Final-State Spectro-
tion, and surface chemical reactions). Exophoton scopy (CFS), Constant-Initial-State Spectroscopy
emission accompanies exoelectronic emission in (CIS), Normal Constant-Final-State Spectroscopy
many cases, and both processes are described by (NCFS), and Normal Constant-Initial-State Spec-
the same laws. troscopy (NCIS). The values of quantum yield are
measured using Y-spectroscopy techniques: Yield
EXPANSION, HIGH-TEMPERATURE Spectroscopy of total quantum yield (YS) and Par-
See High-temperature expansion. tial Yield Spectroscopy (PYS), in which the photo-
electrons are distinguished with respect to energy
EXPANSION, THERMAL and angle of emission. There is also Electron Spec-
See Thermal expansion. troscopy for Chemical Analysis (ESCA). Some of
these acronyms are, and some are not commonly
EXPERIMENTAL METHODS OF used in the Western literature.
PHOTOELECTRON SPECTROSCOPY The analysis of the dependences of the yield
Experimental techniques for determination of Y (hν, P , σ, α) and number N(E, k, σ, hν, P ) of
the spectral characteristics of photoelectron emis- photoelectrons may provide information on the
sion in photoelectron spectroscopy, including density of empty and occupied (band and inner
quantum yield, energy, degree of spin polarization, core) electron states, dispersion of energy bands,
and angular distribution of photoelectrons N , etc. symmetry of wave functions and spin state of elec-
Varying the test conditions and recording different trons, geometric structure of the surface, orienta-
quantities, one may carry out various methods of tion and symmetry of orbitals of the surface atoms
photoelectron spectroscopy, which are sensitive to and molecules, composition in atomic percent and
diverse physical properties of solids. The experi- chemical composition. It is also used for exam-
mental parameters may be: ining the positions of localized adsorbed atoms
428 EXPLOSION
and molecules and the orientation of their orbitals etc.) are also related to explosions. In scientific in-
in the electron spectrum. The methods of photo- vestigations involving explosions extremely high
electron spectroscopy are used for determination values of pressure, temperature and density are
of the potential barrier characteristics: work func- achieved. It is used for the attainment of super-
tion, electron affinity, location of Fermi level, band strong magnetic elds, for the instigation of phase
edges and bendings at the surface. Information on transitions (see Phase transitions under pressure),
scattering mechanisms can also be obtained. and for the formation or synthesis of new sub-
stances (see High pressures).
EXPLOSION
A process of energy release within a restricted EXPLOSION, COULOMB
volume during a short time interval. See Coulomb explosion.
As a result of an explosion, the material fill-
ing the volume where energy is released is trans- EXPLOSION DEFORMATION, explosion strain
formed to a strongly heated gas under very high Deformation of a material due to the passage of
pressure. During the course of the explosion a shock waves initiated by an explosion. The mag-
shock wave is formed and spreads into the envi-
nitude of strain ε is determined by the expression
ronment. An explosion in a solid medium is ac-
ε = (4/3) ln(V /V0 ), where V and V0 respectively
companied by its collapse and destruction. Differ-
are the specific volumes of the material before and
ent types of explosions can be distinguished by the
after the passage of the shock wave. Explosion
physical nature of the energy source, and by the
deformation is distinguished from quasi-static de-
manner of its release. An explosion can occur due
formation by several special features: pulse wave
to the emission of chemical or nuclear energy, re-
character, shortness of the effect, high effective
leasing the compressed gas energy, or transform-
tangential stresses, lack of significant residual
ing the magnetic, electric, radiation or kinetic en-
macroscopic deformation. These features condi-
ergy into heat energy. A heat explosion can occur
tion multiple slip even at small degrees of defor-
when thermal equilibrium cannot be established
mation, initiation of deformational twinning (see
between the reacting material and its environment.
Twinning of crystals) at room temperature for face-
At sufficiently large values of activation energy
the speed of the chemical reaction ω increases centered cubic metals, twisting deformation mech-
with increasing temperature T in accordance with anisms (see Plastic twisting deformation), and the
Arrhenius law. A chain explosion occurs when formation of a substructure with a high density of
a chemical reaction develops as a branched chain defects. The possibilities of an explosion deforma-
reaction during the course of which large concen- tion are greater than that of a quasi-static one. At
trations of active particles or radicals appear (com- present shock waves are investigated in materials
parable with source material concentrations). The with a pressure within the wave front of about a
thermal and chain modes of propagating the ex- TPa.
plosion can occur during nuclear transformations,
i.e. the reactions of nuclear fission and fusion. An EXPLOSION NOISE
explosion can be caused by sharp external effects, See Telegraph noise.
by shock, by friction, etc. The cause through shock
is probably the result of local heating of the mate- EXPLOSIVE ELECTRON EMISSION
rial. A shock wave causes the specific type of ex- Emission of an intense electron flow, brought
plosive transformation which does not appear si- about by transforming cathode metal material
multaneously along the whole charge, but rather from the condensed phase (see Condensed state
propagates in the material at a constant speed, i.e. of matter) to a dense plasma (see Solid-state
detonation occurs. Processes during which no in- plasma) as a result of local heating. The metal-
ternal energy of the material is released, but rather to-plasma transition can be initiated by the ex-
the energy of an external source (e.g., the collision plosion of the metal due to its heating by a e ld
of bodies moving with large speeds; or the mat- electron emission current of high density (j =
ter where high-power laser radiation is focused, 108 –109 A/cm2 ), due to the impact of particles at
EXTERNAL FRICTION 429
the cathode, due to radiation flux incident on the load F = A + BP χ , where A and B are inde-
cathode, etc. The initial explosion and subsequent pendent of the pressure. Sometimes the two limit-
explosive electron emission are accompanied by ing cases are considered, corresponding to χ = 1,
the near-cathode generation of plasma, which ex- A = 0 (Coulomb’s law) and χ = 1, A = 0. In
pands with the speed v ≈ 106 cm/s, and consti- practice it is convenient to use the friction coef-
tutes a single peak of current ficient f = F /P . The Michin–Kragelsky equation
vt is the best known expression
√ for this coefficient:
I = 3.7·10−5 ·U 3/2 , f = τ0 /P r + β + kα H h/R, where P r is the ac-
(d − vt)
tual load at the contact. The two first terms consti-
where U is the voltage between the cathode and tute the molecular component of the friction coef-
anode during the course of the emission, d is the ficient, and the last represents the mechanical (de-
distance between them, and t is the time. This formation) component; β and k are numerical pa-
electron emission is accompanied by the carrying rameters. The mechanical component is the resis-
off of the material from the cathode. This latter tance arising from pressing together the material,
effect can be reduced by decreasing the electron defined by the coefficient of hysteresis loss α H and
current, but if this current becomes less than some by the intrusion depth h of the unevenness, sim-
critical value the electron emission stops. Explo- ulated by a spherical indenter of radius R. This
sive electronic emission is used in strong-current component is very sensitive to the variation of the
amplifiers of electrons, and in pulse sources of elastic modulus of the softer material of the fric-
high-intensity X-rays. It also plays a fundamental tion pair, and can be 50 to 0.01% or less than the
role in vacuum discharges, in some types of gas coefficient of friction. The first part of the molecu-
discharges, and in the generation of shot noise. lar component τ0 /P r depends on the load P r and
is determined by the value of specific adhesion τ0 .
EXPONENT, CRITICAL For laminated compounds (graphite, NB, MoS2 )
See Critical exponents. which are frequently used as antifriction materi-
als, τ0 is determined by the value of the lamina-
EXTERNAL FRICTION tion energy along the basis plane (E001 ) and by
A dissipative process at the area of contact be- the characteristic dimension of the crystallite (La ):
tween two bodies pressed together as they slide τ0 ≈ E001 ρ001 /(La N a ), where ρ001 is the atomic
past each other, accompanied by heat release, and density of the basis plane. For metallic pairs the
the possibility of electrification (triboelectricity), specific adhesion τ0 and the coefficient of molecu-
erosion (surface wear), etc. lar bond hardening are determined experimentally.
The work of friction (W fr ) is determined by Boundary and hydrodynamic regime of friction are
thermal (∼99.9% of W fr ) and non-thermal losses distinguished. In the first case the structure and
connected with the increase of irregularities of the properties of the surface of the friction pair mate-
layers near the surface, and the possible destruc- rial have an essential influence which determines
tion of surface material. External friction can be the value of actual area of contact, and the value
characterized by the friction force F = dW fr /dl of the adsorption into the contact gap of lubrica-
which appears in the plane of contact of the slid- tion materials which can be in either the gaseous
ing bodies, and is directed opposite to direction of or liquid states (liquid friction) in the contact gap.
the displacement dl. For small relative displace- Upon adsorption the lubrication materials at the
ments
l which are less than a “zero” value
l0 contact form wedging layers which create normal
the friction force grows from zero to a maximum, pressure up to 152·106 Pa (1500 atm) that provides
and beyond this value (
l >
l0 ) it no longer de- the sliding. Here the contact gap can be regarded
pends on
l. Typical zero displacements
l0 lie as a slot micropore with a characteristic dimen-
in the interval 10–0.01 μm. In the general case sion of about 10−9 m. In the boundary lubrication
there are two types of relevant factors: those de- mode the coefficient of friction does not depend
pending on the specific load normal to the fric- on the lubricant viscosity (η) and does not depend
tion surface (P ) and those independent on the (or depends weakly) on the sliding speed (V ). In
430 EXTINCTION
EXTINCTION
Reduction of the intensity of a radiation beam
that propagates in a medium. For weakly diffract-
ing radiation (e.g., light propagating in an almost-
transparent body), a single mechanism of the ex-
tinction is associated with the inelastic processes
of the radiation interaction with the matter, which
Dependence of the coefficient of friction f on the sliding
is accompanied by energy loss. In this case the ra-
speed V , and on the Sommerfeld criterion ηV /(P d).
diation intensity I decays with the depth of pen-
etration l in an exponential manner (Beer’s law):
the second case the coefficient of friction is deter- I = I0 exp(−μl), and the extinction is character-
mined by the hydrodynamic properties of the lu- ized by the absorption coefficient μ which de-
bricating liquid and depends on the dimensionless scribes the inelastic processes. If the medium
Sommerfeld criterion Sm = ηV /(P d), where d is is inhomogeneous (“turbid”) then elastic scatter-
the bearing length. At the transition from bound- ing on the medium inhomogeneities takes place.
ary friction to the hydrodynamic mode (Fig.) the In this case the intensity is expressed as I =
coefficient of friction decreases and then increases I0 e−(μ+ν)l +
I , where the first term describes
in value proportional to the liquid viscosity. The the intensity in the direction of the propagation of
gas phase composition strongly influences the fric- the initial beam,
I (l) is the intensity of the radi-
tion coefficient between the two surfaces. ation scattered in other directions (“diffuse back-
At the working-in of the friction pairs the ground”), and ν is the additional attenuation of the
roughness corresponding to the minimal coeffi- transmitted beam caused by diffusion scattering.
cient of friction is set spontaneously. For metallic The measurement of ν provides information on the
systems fragmentation of the surface layer is ob- characteristics of the medium inhomogeneities. In
served in the friction process, and for laminated the case of interference during the propagation
compounds a film forms with its basis planes ori- of the radiation (X-ray, neutron, electron, or any
ented with respect to the friction surface. When other) in crystals, the extinction acquires a more
the load increases, the depth of the surface layer complicated form. For directions approaching the
changes, it breaks into fragments, mutual disorien- Bragg direction (see Bragg law), at a distance close
tation and rotation of these fragments begins, and in order of magnitude to the extinction length Λ,
with further increases in the load the external fric- the beam intensity transforms partially into the
tion transforms to internal friction, i.e. there is no intensity of the diffracted beam. As a result, an
jump of speed at the transition from one body to interference wave field appears, and in the case
another. of Bragg diffraction, the intensity I (z) decays ex-
Sliding friction and rolling friction can be dis- ponentially with the depth (i.e. I ∝ exp(−z/Λ)).
tinguished, each characterized by its own coeffi- This phenomenon is called dynamic extinction.
cient. For rolling friction f r = M/N , where N is In the case of Laue diffraction, I (z) is described
the normal load, and M is the moment of resis- by a more complicated expression that includes
tance to rolling. Rolling friction is less than sliding the specific effects of dynamic radiation scatter-
friction. It is conditioned by losses due to elastic ing (e.g., anomalous passage of X-rays, and so on)
hysteresis, by the material work of deformation at which cannot be reduced to the ordinary extinc-
the formation of a bead in front of the rolling body, tion of the exponential type. In addition, in real
and overcoming of the forces of molecular inter- crystals with a mosaic structure one can observe
action between the contacting surfaces. At mod- a phenomenon related to the extinction when the
erately high speeds resistance to rolling increases light propagates in a medium with random inho-
EXTRAORDINARY RAY 431
433
434 FAILURE
the figures. The first three mechanisms are based interaction of two different dislocation pile-ups
on the concept of slip bands blocked by obsta- (Fig. (e, f)), or during the interaction of disloca-
cles, which are regarded as flat disclination heaps tion pile-ups with inclusions (Fig. (g)) or pores
that produce high stress concentrations in their (Fig. (h)) in a crystal. Even partial disclinations
head areas. Such a concentration may result in ei- (or their dipoles) are a significant source of inter-
ther a restart of a slip or an opening of a crack. nal stress, and may also result in the generation
Fig. (a) shows a mechanism for the merging of of microcracks. The probability of opening mi-
dislocations in the head of a dislocation heap re- crocracks at various interfaces (grain boundaries,
sulting in the formation of an embryo microcrack. subboundaries) is often greater due to the pres-
The mechanism of hampered shear (Stroh theory) ence of segregated impurities (see Segregation),
is presented in Fig. (b). The crack here devel- and to the availability of energy stored during plas-
ops at an angle of approximately 110◦ relative to tic deformation.
the slip plane, as determined by the distribution Stage III. Crack growth, the final break-up and
of stress at the head of the dislocation heap. The disintegration of the specimen into two parts.
Rozhanski–Gilman mechanism (Fig. (c)) is based The description of processes happening during
on the concept of the bending of the slip plane in this stage generally focuses on the concepts of
which the edge dislocations accumulate. The crack linear mechanisms of failure that entail the con-
develops from the point of braking in the slip plane cept of a critical coefficient of stress intensity
so as to meet that pile-up. The Cottrell mecha- (crack resistance). The micromechanism of failure
nism (Fig. (d)) assumes a crossing of two slip lines is controlled by the character of the interatomic
{110} in body-centered cubic metals accompanied bonds, temperature, type of state of stress, and
by the formation of dislocations in the {100} plane specimen structure. Micromechanisms of failure
with its Burgers vector a[001], as prescribed by may be characterized by the fracture toughness
the reaction (a/2)[1̄1̄1] + (a/2)[111] → a[001]. and the structural indicator of failure. By taking
The merging of those dislocations into bands ini- into account the failure viscosity one may separate
tiates the formation of microcracks. There are the types of failure into brittle failure, quasi-brittle
more advanced comprehensive models available failure, and tough failure. Meanwhile, according
for describing the generation of cracks during the to structural indicators, transcrystalline failure or
FAILURE KINETICS 435
intercrystalline failure may take place. These fac- disrupt that tendency. Studies of failure mecha-
tors of fracture toughness and structural indicators nisms provide data for developing materials with
may be found in any combination, and they pro- improved mechanical and service characteristics.
duce the following six basic types of failure: See also Brittle failure, Intercrystalline failure,
Quasi-brittle failure, Strati cation-type failure,
1. Brittle transcrystalline failure, a catastrophic Tough failure, Transcrystalline failure.
failure of crystals along their cleavage planes
(see Crystal cleavage) with the formation of
FAILURE KINETICS
typical spalling streams without any noticeable
The development and interrelationships be-
plastic deformation in the crack apex.
tween elemental actions that specify mechanical
2. Quasi-brittle transcrystalline failure, a failure
failure and the dependence of strength on tem-
by spalling accompanied by plastic deforma-
perature, and on the time during which the stress
tion.
continues. It has been experimentally established
3. Tough transcrystalline failure, a typical mi-
that the time τ which elapses until failure (see
cromechanism for it being the generation and
Durability), for solids with various types of inter-
merging of microcavities accompanied by the
atomic bonding and lattice structure, obeys the
formation of pitted relief. The major crack then
equation: τ = τ0 exp[(U0 − γ σ )/(k BT )], where
propagates by way of successive lacerations of
σ is the tensile stress, and τ0 , U0 and γ are para-
the neck between the cavities formed in front
meters of a material. The exponential dependence
of the apex of the growing crack. Variations of
of τ on the temperature indicates that thermal ac-
that type of failure include rupture (transverse
tivation underlies the failure. The durability of a
narrowing is then ψ = 100%) and shear failure
body under load depends on thermal fluctuations
of microcrystals.
resulting in the generation of stable material dis-
4. Intercrystalline brittle failure propagates as a
continuities called cracks, and their accumulation
spalling along the grain and subgrain bound-
up to a critical concentration in the failure zone.
aries.
Large-scale density fluctuations serve as a source
5. Quasi-brittle intercrystalline failure involves
for the formation of micro-discontinuities in a con-
spalling along the grain boundaries while a pre-
tinuous body. The initiation of a crack may result
ceding plastic deformation forms characteristic
from a fluctuation of the number of phonon colli-
steps at fracture surfaces (see Fracture). sions over their mean free path, Λ, resulting in a
6. Tough (pit) intercrystalline failure during which kinetic instability with ensuing phonon pumping,
the major crack propagates via successive lac- tensile strain, and break-up of interatomic bonds
erations of necks between the cavities formed inside the fluctuation region. The size of a nucle-
as a result of plastic deformation situated in ating crack, r, will be greater than or equal to Λ
subboundary volumes. In that case the local- according to this model, and its activation energy
ization of plastic deformation in the subbound- U is given by
ary volumes results from the presence of an
intergrain boundary layer that is softer than the a3 Λa 2 σ
internal volumes of the grains. U= 2
− ,
G κ G
In certain cases, particularly in the presence where a is an interatomic distance, G is the
of structural texture of metals, anisotropies are Grüneisen constant (see Grüneisen tensor), and κ
observed in various types of failure. A combina- is the bulk modulus; these relations agree with ex-
tion of several varieties of failure then merge as periment. Cracks are generated randomly over the
the crack develops in various directions. With any material because of the random nature of thermal
increase in temperature one may observe a suc- fluctuations, so their accumulation is accompa-
cession of micromechanisms of failure, extending nied by a spontaneous clustering that results in
from brittle fragmentation to viscous failure. How- the formation of enlarged cracks which then grow
ever, a change in the structural type of failure may to generate the failure itself. The transformation
436 FAILURE OF AMORPHOUS METALLIC ALLOYS
served in the exciton spectra of absorption and odd powers of B, and thus they are unsymmetrical.
luminescence of semiconductors, in electron ab- At high B intensities, especially near resonant fre-
sorption spectra of the impurity centers in crys- quencies, there is a contribution from terms with
talline matrices, in the spectra of Raman scattering higher powers of B than the first one.
of light of photon and electron excitations in crys- It is used for the determination of the band
tals. Investigations of the Fano resonance in solids structure parameters of semiconductors, the prop-
gives information about the energy spectrum of erties of impurity states, the structure of mole-
excitations, about their life times, and their inter- cules, the magnetization of solids. The effective
action mechanisms. masses of the carriers in many semiconductors
have been determined by measuring the Fara-
FARADAY, Faraday number, F day effect of free carriers in the IR range of the
A fundamental physical constant numerically spectrum. The Faraday effect is used in light mod-
equal to the amount of electric charge which ulators.
liberates 1 mole of a monovalent substance at
each electrode after passage through an elec- FATIGUE
trolytic solution. It has the value F = NA e = Progressive accumulation of damage in a ma-
9.6485·104 C/mole, where NA is the Avogadro terial under action of alternating stresses. This
number, and e is the elementary electric charge.
results in a change of properties, generation and
propagation of cracks, and failure of the mater-
FARADAY EFFECT, Faraday rotation
ial. The tendency to counteract the development
(M. Faraday, 1845)
of fatigue is referred to as the fatigue resistance
Rotation of light polarization plane of electro-
magnetic radiation (light) propagating within a of a material. By analogy with the term long-term
strength, this property is sometimes called cyclic
material parallel to the direction of a constant ap-
plied magnetic field (see Magnetooptical effects). strength (see also Endurance limit).
After passage through a layer of nonmagnetic ma- The problem of fatigue is one of the most com-
terial with the thickness d in a magnetic field plicated and important ones for those branches of
with intensity B, the plane of polarization ro- physics and engineering dealing with the strength
tates through an angle θ = V Bd, where V is the of solids. Materials science and the mechanics of
Verdet constant which depends on the properties deformable bodies offer various models for the
of the material, on the wavelength of the light λ, origin and propagation of fatigue damage, but
etc. The Faraday effect is caused by the differ- the quantitative characteristics of fatigue are best
ence in the speeds of the right- and left-circularly obtained experimentally by comparing materials
polarized components of the electromagnetic ra- and structural members under various conditions
diation which propagates longitudinally along an which depend on the task to be achieved. The
applied magnetic field; the superposition of these data provided by relatively simple fatigue tests can
components forms linear-polarized radiation. In be used to calculate the characteristics of fatigue
magnetically ordered materials the Faraday effect for complicated cases of loading, to determine the
is determined not only by B, but also by the spon- durability of structural members at a given load, or
taneous magnetization. conversely, to determine the permissible load for
The sign of the Faraday effect does not de- a given durability. Many of these problems have
pend on the direction of propagation with respect been solved by analyzing the mechanics of cracks
to the B direction, which follows from the pseudo- (see Fatigue crack growth).
vector character of the magnetic field intensity. Numerous methods have been developed for
This distinguishes the Faraday effect from the ro- increasing fatigue resistance by applying spe-
tation of light polarization plane in optically active cial treatments to those parts of machines and
media (helical symmetry). In transparent materi- constructions which are exposed to the greatest
als the Faraday effect is expressed in terms of stresses. From the materials science standpoint,
off-diagonal components of the magnetic perme- the major goal is the preparation of materials
ability tensor μij , which are proportional to the (both homogeneous and composite) which offer
438 FATIGUE CRACK GROWTH
enhanced resistance to the initiation and propa- leads to the appearance of secondary microcracks,
gation of fatigue cracks. The problem is that the located parallel to the fatigue grooves. The third,
attainment of high material static strength is not and concluding, stage of fatigue crack growth pro-
necessarily accompanied by a corresponding in- ceeds with acceleration after reaching the critical
crease of its cyclic strength. length of the fatigue crack. In this section, the
See also Adsorption fatigue, Corrosion fatigue, crack growth proceeds at high speed. Failure in
Thermal fatigue.
this region may have a tough, as well as a brittle
FATIGUE CRACK GROWTH
nature (see Tough fracture, Brittle fracture).
Cyclic nonstationary process of increasing
crack length, which causes failure of a material
FAULTING
under cyclic loading (see Cyclic strength). At low-
cycle fatigue the growth of cracks is of a quasi- A fault is a fracture in which movement of ma-
static nature in every loading cycle, leading to the terial takes place along its plane. Faulting involves
advancement of the crack, and to the formation of the plastic deformation of crystals associated with
a fracture with a neck. After many fatigue cycles the nonuniform distribution of slip (gliding) within
the local dislocation structure changes (see Dislo- the sample. Nonuniformity of sliding upon fault-
cation) in the surface layer at the stage of crack ing during the course of strain (usually upon com-
generation, and near the vertex of the advanc- pression) is caused by the loss by the crystal, at
ing crack at the growth stage prior to the fatigue first of elastic stability, and then of plastic stability.
failure. The most characteristic defects which pre- The sliding develops at regions with orientations
pare for the formation of the fatigue crack are that are more favorable for it. At the moment
intrusions and extrusions, or microcavities and of faulting, temporary unloading of the sample
microbulges at the surface, formation in the sur- takes place, which is manifested through the sharp
face layer of stable slip bands, where irreversible
falling of the level of stresses seen on a deforma-
dehardening of the material occurs. The Cottrell–
tion diagram. Faulting is observed in those crystals
Hull model (A.H. Cottrell, D. Hull, 1957), where
two surface dislocation sources in crossing slip where it is possible to create a state of deforma-
systems are effective, is the most widely used tion which hinders the development of gliding: in
model for explaining the mechanism of formation the case of hexagonal metals, it is necessary to
of extrusions and intrusions. At the first stage of perform compression or stretching approximately
fatigue crack growth, its propagation occurs at the along the basal plane, and in the case of CsCl type
angle of 45◦ to the direction of the load appli- (BCC) crystals the compression or stretching is
cation axis, or in the direction of action of the carried out close to one of the mutually orthog-
maximum tangential stresses. The duration of this onal gliding directions of the type 100. At the
stage of growth is no more than 5% of the total earliest stages of sliding, there appear sections of
time until failure. At the second stage character- stress concentration, most often in the fractures
ized by a stable growth speed the generated crack, at the surface, where upon crank flexure the lat-
having achieved its macroscopic size, changes its tice is subjected to the greatest deviation from
trajectory and spreads perpendicular to the axis of the source orientation. In these areas short slip
application of the stresses. The rate of the crack
bands are generated, spreading rapidly perpendic-
advancement is determined by the structure and
ular to themselves toward the center of the sample.
phase composition of the material, by the magni-
tude of the amplitude of cyclic loading, and also by Several such bands may be generated in different
the testing medium. The sections of flexure being parts of the sample, and then multi-layer faults or
formed at the stage of stationary growth are char- several separate bands of faults are formed. As
acterized by specific grooves, reflecting the cyclic faulting is associated with the behavior of large
interlacing of the sharpening and blunting of the collective bodies of dislocations, it may be de-
advancing crack. Evolution of the material struc- scribed in terms of the disclination model of plastic
ture in the vicinity of the growing fatigue crack deformation.
FERMI–DIRAC STATISTICS 439
the fermion states below some energy εF are occu- The gas of conduction electrons in metals and
pied, all the states lying above εF are empty; and in degenerate semiconductors may be character-
the value of εF coincides to the value of μ: ized by other than a quadratic dispersion law. Then
at T = 0 K the conduction electrons fill momen-
pF2 tum space in regions of a more complex shape (see
εF = ,
2m Fermi surface). The location of the Fermi energy
2 1/3 (2)
6π N in a partly filled electron band is called the Fermi
pF = , level. In the case of semiconductors the position
gV
on the energy scale of the chemical potential at a fi-
where m is the fermion mass, N is number of par- nite temperature is also called the Fermi level. It is
ticles in the volume V , and g is the degeneracy. necessary to remember that this designated Fermi
The quantity εF is called the Fermi energy, level may be located within the gap or band of for-
and pF is the Fermi momentum. In metals and bidden energies, and hence not coincide with any
degenerate semiconductors with a nonquadratic actually existing energy level.
dispersion law the dependence εF (pF ) has a more
complex form than Eq. (2) (see Fermi surface), and FERMI LEVEL
is often anisotropic. For T > 0 a certain number See Fermi energy.
of fermions has an energy greater than εF ; empty
states vacated by these elevated fermions at ε < εF FERMI LEVEL PINNING, Fermi level limiting
are called holes. position
Establishment of a fixed value of the Fermi
FERMI ENERGY level μF of a semiconductor in the presence of
The value of energy, below which all the states electrically active centers or defects, a value which
of a system of particles obeying Fermi–Dirac sta- does not change with further additions of these
tistics (fermions), are occupied at absolute zero centers or defects. Fermi level pinning occurs
temperature, and all higher states are vacant. The when donors and acceptors are introduced into
existence of a Fermi energy is a consequence of the sample in equal concentrations Nd = Na (e.g.,
the Pauli exclusion principle according to which levels of vacancies and interstitial atoms with for-
there can be no more than one fermion particle in mation of Frenkel defects). Upon reaching a suffi-
a state (or two electrons with opposite spins). ciently high concentration, when the numbers of
For an ideal degenerate gas of fermions the charged donors and acceptors Nd+ = Na− domi-
Fermi energy coincides with the value of the nate the electroneutrality condition, μF does not
chemical potential at T = 0 K, and may be ex- depend on Nd , Na , but rather is equal to the limit
pressed as follows in terms of the number N of value μF = (εa + εd )/2, where εa and εd are the
gas particles in a unit volume: positions of donor and acceptor levels (measured
2/3 from a common origin). A Fermi level limiting po-
(2π)2 3N
εF = , sition is also realized when there is a dominant
2m (2S + 1)4π
concentration of defects of one type, and these de-
where m and S are the mass and spin of the parti- fects are amphoteric centers, especially if they are
cle, respectively. The quantity pF = (2mεF )1/2 is centers with a negative correlation energy.
the Fermi momentum. At T = 0 K all the states
with momenta p < pF are occupied, and those FERMI LIQUID
with p > pF they are empty. In other words, Quantum liquid, whose elementary excitations
at T = 0 K fermions in an isotropic momentum (quasi-particles) have half-integer spin (fermions),
space occupy all the states within the sphere p2 = and due to the Pauli exclusion principle (see
2mεF (Fermi sphere), with the radius pF in mo- Fermi–Dirac statistics) they have a characteristic
mentum space (k-space). The value vF = pF /m, Fermi momentum pF determined by the density
called the Fermi velocity, determines the upper of particles of the liquid pF = (3π 2 N/V )1/3
limit of fermion speeds at T = 0 K. (V is the volume of liquid, N is the number
FERMI LIQUID 441
of fermions). The Fermi momentum determines The quantity fˆ called the Landau correlation
the Fermi energy εF , from which the energies function is a 4 × 4 matrix in spin space which pre-
of quasi-particles ξ = |ε(p) − εF | are derived. determines a set of phenomenological parameters
The excitations of a Fermi liquid are generated in that are taken from experiment. In metals, where
pairs: a hole quasi-particle with p < pF and a cor- the correlation length and wavelength of electrons
responding electron quasi-particle with p > pF . are of the order of interatomic distances, it is con-
A strong interaction between the quasi-particles sidered to be independent of the coordinates.
is characteristic of a Fermi liquid, so the single- In the presence of only the exchange interac-
particle state is not the natural state of a many- tion the matrix fˆ is expressed with the help of two
particle system, and it breaks down after a finite scalar functions η(p, p ) and ξ(p, p ):
time. Due to the features of the Fermi spectrum fˆ(p, p ) = η(p, p ) + σ σ ξ(p, p ) (3)
(presence of filled states), regardless of the in-
teraction strength, the decay time τdec is long (σi are Pauli matrices).
for the single-particle states near the Fermi sur- In an isotropic system the functions η(p, p )
face ε(p) = εF (at sufficiently low temperatures): and ξ(p, p ) depend only on the angle θ be-
τdec ∼ εF /ξ 2 (at T = 0, ξ ∼ kB T ). Thus the tween the vectors p and p , and may be expanded
conditions for applying Fermi liquid theory are: in terms of Legendre polynomials. Dimension-
ξ kB T kB Tf = εF . Conduction electrons in less Landau parameters Al and Bl are coeffi-
a metal, the liquid helium isotope 3 He, and nu- cients of the expansion the Legendre polynomi-
clear matter serve as examples of Fermi liquids. als of the functions η(θ) and ξ(θ) multiplied by
For conduction electrons in a metal Tf ∼ 104 to g(εF )(2l + 1)−1 , where g(εF ) is the density of
106 K. states of the quasi-particles at the Fermi surface.
The theory of a Fermi liquid was formulated The function fˆ satisfies the stability requirements
by L.D. Landau on the basis of two assump- for the ground state of a Fermi liquid with re-
tions: (1) the interactions of quasi-particles do not spect to both deformations of the Fermi sphere,
disturb the systematics of levels near the Fermi and small spin vibrations. This imposes specific
surface, and in the presence of interactions the conditions at the parameters Al and Bl (Pomeran-
original fermions are replaced by quasi-particles; chuk criterion, I.Ya. Pomeranchuk, 1958):
(2) interactions of particles with one another may 1 + Al > 0,
(4)
be described with the help of a self-consistent eld 1 + Bl > 0.
by the introduction of a distribution function for the
Inequality (4) for B0 expresses a condition for a
quasi-particles n̂ = nαβ (p, r, t). The function n is
Fermi liquid paramagnetic state with respect to a
a 2 × 2 operator-matrix in spin space. The quasi-
transition to the ferromagnetic one (Stoner crite-
particle energy depending on the spin εαβ (p, r, t)
rion, see Ferromagnetism of metals and alloys).
is determined as the first variational derivative of
The Fermi distribution of quasi-particles follows
the total system energy E[n̂]:
from the requirement that the overall Fermi en-
tropy be maximal (for a fixed number of particles,
δE = Trσ ε̂(p, r, t) δ n̂ dτ d3 r, δN = 0, and a fixed energy, δE = 0):
(1) −1
d3 p ε̂ − εF
dτ = , n̂(ε̂) = exp +1 , (5)
(2π)3 kB T
where p is the momentum, and ε̂ = ε̂0 + δ ε̂, so where the single-particle energy ε̂ is a functional
of n̂. The speci c heat C depends linearly on
the temperature, as also is the case for a Fermi
δ ε̂(p, r, t) = Trσ fˆ(p, r; p , r )
gas, C = (π 2 /3)g(εF )kB T . Within the framework
of the isotropic model the magnetic susceptibility
× δ n̂(p , r , t) dτ d3 r . of a Fermi liquid χ = g(εF )μ2B (1 + B0 )−1 con-
(2) tains the parameter B0 (at B0 < 0 the condition
442 FERMI MOMENTUM
|B0 | → 1, χ → ∞ corresponds to the transition to without any variation of the liquid density. In the
the ferromagnetic state). The sound velocity is ex- simplest case η(θ) = A0 g −1 (εF ) the inequalities
pressed with the help of the parameter A0 : s 2 = u > v, A0 > 0 are the conditions for the existence
[pF2 /(3m0 m∗ )](1 + A0 ). When Galilean invari- of zero sound, and the Fermi surface is axially
ance is satisfied the quasi-particle effective mass symmetric and drawn toward the direction of the
m∗ is related to the source particle mass m0 by the wave propagation. Spin waves (in a neutral Fermi
expression m∗ = m0 (1 + A1 ). In the presence of liquid) are vibrations of the spin characteristic of
an external electromagnetic field with vector and a Fermi liquid associated with the second term in
scalar potentials A(r, t) and ϕ(r, t), respectively, Eq. (3). In the absence of an external magnetic
the quasi-particle energy ε̂(p) is replaced by the field the expression for these vibrations is obtained
expression from Eq. (7) by replacing η by ξ . In the partic-
ular case ξ = B0 g −1 (εF ), for the propagation of
e
ε̂ = ε̂ p − A(r, t)r, t a spin wave B0 > 0 is necessary. For liquid 3 He
c B0 < 0, which supports the impossibility of prop-
− μB σ B + eϕ(r, t), (6)
agating waves of this type.
B = ∇ × A. In the case of a charged Fermi liquid (e.g., elec-
trons in a metal) vibrations of the electron density
In a neutral Fermi liquid two modes of sonic
lead to the origin of the electric fields that al-
vibrations are possible. The low-frequency ones ter the conditions for the existence of zero-sound
(ωτ 1, τ is the mean free path time of the vibrations. This means that vibrations of the elec-
quasi-particles, τ ∼ 1/T 2 ) correspond to ordi- tron density at frequencies ω < ωp in a metal are
nary sound, which is sometimes called first sound. absent (ωp is the plasma frequency). Although
The velocity s of this sound, which involves den- zero-sound vibrations are possible in a Fermi liq-
sity variations, is determined by the compress- uid in which there is no variation of the electron
ibility, and in the isotropic case it is expressed density, zero sound has not yet been detected ex-
through parameters A0 and A1 (see above). At perimentally in an electron liquid (see Spin waves
reduced temperatures (ωτ ∼ 1) ordinary sound in nonmagnetic metals). See also Microscopic the-
strongly attenuates. An undistorted Fermi surface ory of Fermi liquid.
shape is characteristic of ordinary sound: both the
Fermi momentum and the density of the liquid FERMI MOMENTUM
vary, and the Fermi surface vibrates as a whole See Fermi energy and Fermi–Dirac statistics.
with an amplitude whose value is associated with
the speed of the liquid motion within the wave. FERMION, HEAVY
High-frequency vibrations (ωτ
1, collision-free See Heavy fermions.
mode) are called second sound or zero sound. This
sound is sometimes referred to as a temperature FERMIONS IN A SOLITON FIELD
wave which does not involve density variations. A specific state of a fermion, in particular, of
The velocity u = ω/k of zero sound is determined a conduction electron, described by some one-
from the equation dimensional models of interacting fermion and
boson fields with a spontaneous symmetry break-
∂n
(kv − ω)δ n̂− kv 0 η(p, p ) δn δτ = 0, (7) ing due to the double degeneracy of the ground
∂ε state of the free boson field. These two states are
where v = ∂ε/∂p is the quasi-particle velocity. In associated with the spatially nonuniform solution
the general case the presence of several branches of the classic equation for a boson field, called a
of zero sound vibrations is possible, each prop- soliton or kink. In a fermion–soliton system there
agating with different speeds. Unlike ordinary appear two associated states, degenerate in en-
sound, zero sound is associated with periodic dis- ergy, and carrying the charges of ±e/2 (e is unit
tortions of the Fermi surface shape. If in this fermion charge), depending on whether the given
case the Fermi surface deformations are not ax- state is occupied or empty. These states are inter-
ially symmetrical, then the vibrations take place preted as soliton states with a non-integer charge.
FERMI QUASI-LEVELS 443
Fig. 1. Hexagonal close packed structure Fermi surface of zinc showing the symmetry points Γ , A, K, L, and M: the
first Brillouin zone exhibits hole “pyramids” at H points; the second zone comprises a hole “monster”; the third zone
involves electron needles at K points (the electron lens centered at the center Γ point is not shown).
inside the surface. A Fermi surface can have sev- In an applied magnetic field B, under the ac-
eral convolutions (see Figs. 1–4). Among them tion of the Lorentz force ev × B a conduction
closed and open, electron and hole varieties are electron between collisions moves along a trajec-
encountered. The volume of each enclosure of a tory in k-space which is at the intersection of the
Fermi surface Ωi (calculated for one unit cell of Fermi surface with the plane k·B = const. The
k-space) is associated with the density of elec- projection of the electron trajectory in r-space to
trons: ni = 2Ωi /(2π)3 , whereas the plane r·B = const is similar to its trajectory
in k-space. This circumstance served as a basis
ni = n, (2) for devising various methods for investigating the
i conduction electron motion (on Fermi surfaces of
where n is the density of conduction electrons of metals) according to their properties in a magnetic
the given metal (n = z/V0 , z is the number of field (see Galvanomagnetic effect, De Haas–van
valence electrons in the unit cell of volume V0 ). Alphen effect, Shubnikov–de Haas effect, Size ef-
The volume of a hole-type Fermi surfaces is usu- fects, etc.). Fermi surfaces of all the monatomic
(h) metals and of many intermetallic compounds are
ally associated with the density of holes ni , so
Eq. (2) is naturally modified by this. Metals (Be, known with good accuracy. The branch of the elec-
Bi, P, Sb, etc.) where the volume of the hole Fermi tronic theory of metals which focuses attention on
surfaces equals the volume of the electron Fermi geometrical properties of Fermi surfaces (topol-
surfaces are called compensated metals. For them ogy, including presence of singularities), and on
the equation ne = nh (ne,h are the total densities the relationships of these geometrical properties
of electrons and holes, respectively) replaces the with physical properties, is referred to as fermi-
condition (2). ology.
446 FERMI SURFACE
Fig. 4. Fermi surface of divalent metals with a hexagonal close-packed lattice according to the model of almost free
electrons: (a) “monster”, (b) “pyramid”, (c) “needle”, (d) “lens”, (e) “star”, (f) “cigar”.
through oxygen ions and in alloys acts through spectrum. Ferrimagnetic resonance can occur in
conduction electrons, can lead to ferrimagnetic ferrimagnets with noncollinear sublattice magne-
ordering. Ferrimagnets form the most numerous tization vectors, and it can also occur near the
group among magnetically ordered media, with compensation point of the ferrite.
applications in various areas of engineering. See
also Ferrimagnetism. FERRIMAGNETISM
A phenomenon involving antiparallel ordering
FERRIMAGNETIC RESONANCE of the magnetization vectors of inequivalent mag-
Selective absorption by a ferrimagnet of elec- netic sublattices, which leads to the presence of
tromagnetic wave energy within the radio fre- an uncompensated net magnetic moment. A neces-
quency and the infrared ranges; it is a type of sary condition for the appearance of ferrimagnetic
electron magnetic resonance. The presence of ordering is the existence of a negative exchange
two or more magnetic sublattices in a ferrimag- interaction between the magnetic ions (atoms) as-
net leads to the presence of multiple branches of sociated with the different sublattices. Ferrimag-
ferrimagnetic resonance. Each branch is character- netism is observed not only in crystalline magnetic
ized by the dependence of the resonance frequency materials, but also in amorphous ones, e.g., in al-
ω on the value of the external magnetic field B loys of d- and f -metals, in the temperature range
corresponding to the excitation of a specific type from 0 to the Néel point TN . Ferrimagnetic order-
of resonance vibration of magnetization vectors of ing allows the presence of magnetic compensation
the sublattices with respect to one another. The at the temperature Tk < TN , where the resul-
low-frequency branch of ferrimagnetic resonance tant magnetic moment becomes zero. Intrinsic
corresponds to the excitation of the precession of ferrimagnetism and induced ferrimagnetism are
the spontaneous magnetization vector of the fer- distinguished. The existence of intrinsic ferrimag-
rimagnet M s in the effective field B eff which is netism is caused by the crystallographic structure
determined by the external field, by the fields of itself, i.e. by the presence of inequivalent posi-
magnetic anisotropy, and by the demagnetization tions for the magnetic ions of the same or different
e lds, as in the case of ferromagnetic resonance. types; induced ferrimagnetism appears only in me-
The uniform precession takes place in such a dia with magnetic ions of different types due to
manner that the magnetization vectors of the sub- the formation of a superlattice. It is necessary
lattices remain antiparallel, so ω = γeff Beff . This to distinguish ferrimagnetism from weak ferro-
type of ferrimagnetic resonance is analogous to magnetism (see Antiferromagnetism), when the
ferromagnetic resonance, with its one particular resulting magnetic moment is caused by small de-
value of the gyromagnetic ratio γeff . In the sim- viations from antiparallelism of the magnetization
plest case of a ferrimagnet with two sublattices of vectors of two equivalent magnetic sublattices.
magnetizations M1 and M2 we have
FERRITES
M1 − M2
γeff = , 1. Ordinarily the term “ferrite” is used to des-
M1 /γ1 − M2 /γ2
ignate double or more complex crystalline ox-
where γ1 and γ2 are the gyromagnetic ratios of the ide compounds which contain iron sesquioxide
sublattices. Fe2 O3 . In the general case the chemical for-
The high-frequency branches of ferrimagnetic mula of ferrites may be written in the form
resonance correspond to precessions of sublattice n1 Fe2 O3 ·n2 Ma Ob ·n3 Rc Od , where the ni are in-
magnetization vectors that violate their antiparal- tegers, and M and R are metal ions. Ferrite spinels
lelism condition. These branches are sometimes correspond to the case n1 = n2 = 1, n3 = 0,
called exchange resonances. Their frequencies are a = b = 1 (e.g., Fe2 MgO4 ); orthoferrites consti-
proportional to the exchange fields which act be- tute the case n1 = n2 = 1, n3 = 0, a = 2, b = 3;
tween the sublattices: ω = γ αM, where α is an iron garnets comprise the case n1 = 5, n2 = 3,
exchange interaction constant. These frequencies n3 = 0, a = 2, b = 3 (e.g., gadolynium iron gar-
are located in the IR range of the electromagnetic net Gd3 Fe5 O12 ; the names take their origin from
FERROELASTICS 449
the crystalline structures of single crystals formed 2. The term “ferrite” is also a name used for
by the corresponding compounds); for n1 = 2k1 an allotropic modification of iron with a body-
6, n2 0, n3 1 and a = b = c = d = 1 the centered cubic lattice (space group F m3̄m, Oh5 ;
corresponding oxide compounds are called hexa- a = 0.286653 nm at 298 K) which exists at tem-
ferrites; for the orthoferrites and iron garnets M peratures below 1184 K. This ferrite is ferromag-
is an ion of a rare-earth element or yttrium. The netic up to the temperature Tc = 1040 K (alpha
great majority of ferrites are ferrimagnets with phase), and above the Curie point it transforms
two or more magnetic sublattices. Magnetic ions to the paramagnetic state (β-phase). The high-
which enter the tetrahedral or (d)-sublattice are temperature phase (1665 to 1810 K) of iron (delta
surrounded by four oxygen ions, the octahedral or phase or delta ferrite) with a body-centered cubic
[a]-sublattice is formed by the magnetic ions sur- lattice (space group I m3̄m, Oh9 , a = 0.29322 nm
rounded by six O2+ ions, in iron garnets rare-earth at 1667 K, paramagnet) is also called ferrite.
element ions are located in dodecahedral coor- 3. Another use of the term “ferrite” is to des-
dination (12 nearest O2+ ions), thus forming a ignate one of the phases in iron alloys, steels, and
{c}-sublattice. The exchange interaction between cast irons, which is a many-component solid solu-
different magnetically active (metallic) ions acts tion of interstitial (usually C, N) and substitutional
through oxygen ions, i.e. it constitutes an indirect (Si, Al, Mn, Cr, Ni, etc.) elements in α-iron. The
exchange interaction. Both intrasublattice, as well proportion of this ferrite may reach 95% depend-
as intersublattice, exchange interactions in ferrites ing on the alloy composition. The properties of
are negative; the [a]–(d) interaction is the domi- these ferrites depend on the type and concentra-
nant one, which leads to ferrimagnetic ordering. tion of the doping elements, on the dimensions of
The total magnetization of ferrites is the vector the crystalline grains, the density of dislocations,
sum of the sublattice magnetization vectors. the presence of dispersed particles.
Ferrites may form solid solutions, i.e. they al-
low the replacement of the initial metallic ions by FERRITE SPINELS
other dia- or paramagnetic ions with close ionic Oxide monocrystals (ferrites) with the struc-
radii. Undiluted from the magnetic point of view, tural formula MO·Fe2 O3 (M = Mg, Ca, Zn, Cd,
ferrites possess a high Néel point (up to 800 K); Mn, Fe, Co, Ni, Cu, or their combination), iso-
the replacement of iron ions by diamagnetic ones morphous to the spinel mineral MgO·Al2 O3 . They
leads to a reduction of the Néel temperature. Many belong to the cubic system, space group F d 3̄m
ferrites have a magnetic compensation point at (Oh7 ). The lattice constant is 0.833 to 0.851 nm;
which the net magnetization goes to zero. density is 4.75 to 5.42 g/cm3 ; magnetic moment
Ferrites belong to the class of magnetic insu- per formula unit at T = 0 is (1.1–5.0)μB , where
lators, and they have narrow ferromagnetic res- μB is the Bohr magneton; Néel point is from
onance line widths
B. For example, yttrium 9.5 to 670 K. The O2− ions in ferrite form a
iron garnet Y3 Fe5 O12 exhibits record low values face-centered cubic lattice, where metallic mag-
of
B ≈ 0.01 mT. Due to this fact ferrites find netic ions occupy the tetrahedral ((d)-sublattice)
their application in UHF engineering for the de- and octahedral ([a]-sublattice) coordinations. In-
sign of filters, gates, circulators, etc. Ferrites and direct exchange interactions in ferrite spinels are
the more complex compounds based on them are negative; the [a]–(d) interaction is the strongest
also used in radio engineering as cores, magnetic one, which in the great majority of cases leads to
circuit elements, etc.; for creating permanent mag- collinear ferrimagnetism.
nets with a relatively low coercive force. Because
of their extensive technical applications the term FERROELASTICS (term rst introduced by
“ferrite” is often applied to any ferromagnetic in- K. Aizu, 1969)
sulator. Epitaxial iron garnet lms are characterized Insulating monocrystals in which particular re-
by a pronounced magnetic anisotropy, and they are gions (ferroelastic domains) differ in the sponta-
used in magnetic bubble domain devices. neous strain of their crystal lattice relative to some
450 FERROELECTRIC CERAMICS
initial structure. When the sample temperature de- similar to pyrochlore, with the common chemi-
creases, spontaneous deformation can develop as a cal formulae: A2 B2 O7 , A2 (BI BII )O6 , those with
result of a structural phase transition from the ini- a pseudoilmenite structure, ABO3 , etc. Since ce-
tial paraelectric phase to a ferroelastic phase of ramics are agglomerates of randomly arranged
lower symmetry. The decomposition of the crystal crystallites they appear isotropic in their prop-
into separate domains corresponds to minimizing erties. These properties depend quite strongly
the elastic energy of the crystal. In contrast to lin- on the ratios between the different crystalline
early elastic materials (see Hooke’s law, Elasticity), phases, the ratio between the relative distribution
the dependence of the deformation of a ferroelas- of crystalline phases and the intercrystallite lay-
tic on the applied mechanical stress has the shape ers, on the stoichiometric composition of the basic
of a hysteresis loop. Moreover, at a certain value phase, the nature of impurities and doped addi-
of stress, called the coercive stress, the crystal tions that are introduced, the choice of production
undergoes a transition to a single-domain state, ac- and processing technologies and regimes. For ex-
companied by a change of sign of its spontaneous ample, a BaTiO3 fine crystallite ceramic with
deformation. Among known ferroelastics are crys- grain size below 1 μm exhibits a relative dielec-
tals of KH3 (SeO3 )2 (transition temperature to tric constant ε ∼ 3000–5000 which is 2–3 times
the ferroelastic state Tc = −62 ◦ C), KD3 (SeO3 )2 higher than that found in common ceramic sam-
(Tc = 12 ◦ C); Nb3 Sn, V3 Si, DyVO4 , TbVO4 , and ples with grain sizes of 20–100 μm. The dielectric
RbMnCl3 . Some ferroelastics are ferroelectrics at constant of these materials varies only slightly
the same time. Ferroelastics are promising ma- with temperature, which is promising for building
terials for building acousto-electric and acousto- compact ferroelectric ceramic capacitors. Adding
optical devices (see Acousto-electronics, Acousto- oxides of certain elements (e.g., Sn, Zr, Bi, Ca
optics). and other compounds based on them) results in a
significant broadening diffusion of the phase tran-
FERROELECTRIC CERAMICS sition. Compounds, solid solutions, and hetero-
A polycrystalline material produced using ce- geneous mixtures with a diffuse phase transition
ramics technology, with a basic crystalline phase serve as a basis for numerous ferroelectric ce-
exhibiting ferroelectric properties. Ferroelectrics ramic capacitor materials with significantly non-
of the oxygen-octahedron type are the ones most linear properties within broad temperature (−60
easily mastered in state-of-the-art ceramics tech- to 85 ◦ C) and frequency (103 –107 Hz) ranges.
nology. Their chemical composition and crys- The parameters of certain ferroelectric capac-
talline structure vary widely. Both simple and itor materials are summarized in Table 1. Mate-
complex ferroelectric compounds may serve as rials with high nonlinearity produced by orien-
the basis for ferroelectric ceramics, as well as tational polarization (see Polarization of insula-
solid solutions based on them. Such are, first of tor) are widely used to produce various nonlinear
all, materials with an ABO3 chemical formula ferroelectric elements. Solid solutions based on
and a perovskite-type structure. Numerous fer- BaTiO3 , SrTiO3 , PbTiO3 yield a basis for produc-
roelectrics crystallize in that structure, the most ing capacitor ferroelectric elements controlled by
important of them being barium titanate, BaTiO3 . either an alternating or a direct electric field, the
Some ferroelectric ceramics have the tetragonal so-called variconds. Using ferroelectric ceramics
potassium–tungsten bronze type of structure with with intergranular layers, such that the composi-
the common chemical formula: A6 B10 O30 or tion and structure of phases at the grain surface
AB2 O6 . Others have a layered structure and com- and inside it are different, makes it possible to
mon chemical formulae: A2 B2 O7 , A2 Bi2 B3 O12 , reach ultrahigh values of ε (up to 50,000). Fer-
ABi2 B2 O9 , A3 Bi2 B4 O15 , etc. Another group al- roelectric materials reduced in an atmosphere of
ternates perovskite-type layers of oxygen octahe- dry hydrogen or doped with oxides of rare-earth
dra with layers of (Bi2 O2 )2+ , where the number elements feature a low specific resistance (10−3 –
of perovskite layers usually varies from two to 102 ·m) and are semiconductors. Using semi-
five. There are also materials featuring a structure conducting ferroelectric ceramics of BaTiO3 with
FERROELECTRIC DOMAINS 451
Ceramic Dielectric Relative varia- Relative variation Dielectic loss Resistivity Electrical
type constant ε at tion in dielectric in reversible di- tangent at two ρ, G·m at strength E,
a temperature constant
ε, % electric constant temperatures 100 ± 3 ◦ C MV/m
of 20 ± 5 ◦ C
εr , % 25 ◦ C 85 ◦ C
Notes:
1. The relative changes in the dielectric constant
ε are its changes in the temperature interval −40 ◦ C to +85 ◦ C
compared to ε at a temperature of 20 ± 5 ◦ C.
2. The relative change in the reversible dielectric constant
εr is the change in ε with the applied electric field growing
from 0 to 0.5 MV/m.
surface barriers or insulating layers several mi- cal ferroelectric ceramic most thoroughly studied
crometers thick one may manufacture low-voltage and most widely used is zirconium–titanium–lead
capacitors with a maximum specific capacity of (PZT), modified with lanthanum (PLZT). To ob-
about 3 μF/cm3 . Semiconducting ferroelectric ce- tain highly uniform and defect-free electrooptical
ramics display a positive thermal resistance fac- ferroelectric ceramics, one uses the techniques of
tor, and are used to produce a particular type joint chemical precipitation from solutions, fol-
of thermal resistors – posistors. The nonlinear lowed by hot pressing. Such electrooptical ferro-
current–voltage characteristics of semiconducting electric ceramics feature high transparency, and
ferroelectric ceramics make it possible to build exhibit pronounced electrooptical effects.
varistors with relative resistance variation to volt- Using ferroelectric ceramics one may produce
age variation ratios that exceed the value of 3. Fer- polar-anisotropic textures (with a predominant ori-
roelectric materials with a high pyrofactor value entation of the elementary electric moments of
serve as a basis for various pyroelectric transform- crystal grains, or of separate grain regions). Such
ers. The operative spectral range of such devices textures exhibit piezoelectric properties. Ceram-
is quite wide, stretching from the far IR to short- ics with high piezoelectric parameters are usually
wave X-rays and γ -rays. Some of the materials called piezoelectric ceramics. To produce piezo-
found among pyroelectric ferroelectric ceramics electric ceramics, solid PZT solutions as well as
(see Pyroelectric materials) are solid solutions of solid solutions based on BaTiO3 and NaNbO3
PbTiO3 –PbZrO3 (PZT); their composition usu- are used, with compositions close to the ferro-
ally corresponding to the morphotropic region, electric morphotropic region. Certain piezoelectric
in which the tetragonal and rhombohedral phases ceramic materials with properties meeting various
coexist. Their pyrofactor, as well as many other piezoelectric instrumentation requirements have
factors reach their maximum values there. The ex- been perfected. High workability, availability of
treme compositions of either PbTiO3 or PbZrO3 raw materials, possibility of task-specific alter-
differ in their properties: the former is ferroelec- ation of their electrophysical properties, and repro-
tric (tetragonal structure), while the latter is an- ducibility in samples of varying size and shape,
tiferroelectric (rhombohedral structure). The mor- provide ferroelectric materials with a wide range
photropic region in the mixture (PbTi1−x Zrx O3 ) of applications in various branches of technology.
is of the order of
x and reaches 0.15 at around
the 50% composition. One may separate the fer- FERROELECTRIC DOMAINS
roelectric ceramics used to fabricate electroopti- Separate regions differing in their spontaneous
cal devices into specific types. The electroopti- polarization direction, into which an ideal crystal
452 FERROELECTRIC DOMAIN SPLITTING
The ionic radius of the impurities causing the where f (E) is the random field distribution func-
decrease of the lattice constant of the crystal is tion (see Random potential) of E exerted on
smaller than that of matrix atoms. In this situa- the ith dipole by the nearby impurities: f (E) =
tion, the impurity ions, as a rule, become non- δ(E − Ei ), δ(x) is the Dirac delta function, and
central ions and generate impurity electric dipole lE defines the mean value of the electric di-
moments in an ionic crystal. The electric dipole– pole moment of the random field. The molecular
dipole interaction forces result in the ferroelectric field approximation corresponds to the substitu-
ordering of dipoles. The spatial dielectric disper- tion f (E) = −δ(E − Ei ). Under this approach,
sion that is usually significant in media with large ferroelectric order always exists at low tempera-
polarizability enhances the possibility of ferro- tures. However, thermal fluctuations can result in
electric phase transitions in similar compounds. the smearing of the δ-function, so the solution with
The presence of spatial dispersion characterized L = 0 does not always exist. It is precisely the
by the correlation radius of the polarization rc thermal fluctuations that lead to the suppressing
changes the effective potential of the dipolar cou- of the ferroelectric phase transition in alkali-halide
pling, so that interaction between impurities that crystals. Eq. (1) allows one to calculate the critical
are located at distances r < rc from each other impurity concentration ncr above which the fer-
appears to be chiefly ferroelectric, and more long- roelectric phase transition induced by off-center
range than the usual dipole–dipole interaction. ions (ncr rc3 ∼ 10−2 ) is activated. At concentra-
Owing to this long-range action, configuration tions lower than this critical value, the dipole glass
fluctuations of local fields (destroying the fer- state exists.
roelectric phase transition in weakly polarized,
e.g., in alkali-halide crystals) decrease, and long- FERROELECTRICITY, VIBRONIC THEORY
range order can appear in the system under certain See Vibronic theory of ferroelectricity.
conditions (see Long-range and short-range or-
der). Such a phase transition is accompanied by FERROELECTRIC LIQUID CRYSTALS
the emergence of a spontaneous polarization (see Liquid crystals with spontaneous polarization
Ferroelectricity) connected with displacements of (see Ferroelectricity) that varies under the influ-
lattice atoms, since oriented dipole moments po- ence of an external electric field. These crystals
larize the lattice. were predicted by R. Meyer (1975) and detected
The theoretical description of phase transitions by him and his collaborators in 1977. Unlike in
in ferroelectrics with impurities (disordered ferro- solids, spontaneous polarization in liquid crystals
electrics) differs from that of phase transitions in does not arise from dipole–dipole interactions, but
ordered crystals similar to BaTiO3 or KH2 PO4 . rather is usually due to a short-range steric in-
The molecular eld theory is applicable to the lat- teraction. Dipolar ordering can arise in a smectic
ter but not to the former, since the local field C∗ -phase (see Smectic liquid crystal), consisting of
chiral (see Chirality) molecules with long axes in-
fluctuations are large in impurity systems, and this
clined at an angle relative to the smectic layers.
violates the conditions required for applying mole-
In this case, the rotation of molecules around long
cular field theory.
axes is restricted, and the spontaneous polariza-
The thermodynamic mean l of the electric
tion P is directed perpendicular to the plane of
dipole moment directions of impurities, averaged
the molecule tilt. Hence, dipolar ordering arises
over the configurations of impurities (indicated
as a side effect of a steric interaction. The value
by the overhead bar) L = l defines the relative
P ≈ 10−6 –10−4 C/m2 is proportional to the or-
number of coherently oriented electrical dipole
der parameter of the molecular short axis. Possible
moments, and can be chosen as the order para-
uses of ferroelectric liquid crystals are in infor-
meter of the ferroelectric phase transition. The
mation handling devices, imaging, and recording,
quantity L obeys the self-consistent equation
because they ensure liquid crystal device per-
formance times in the range 10−6 –10−3 s (see
L = dE f (E, L)lE , (1) Optical techniques of information recording).
FERROELECTRIC PHASE TRANSITION 455
FERROELECTRICS
Crystalline insulators (or semiconductors)
which exhibit ferroelectricity. Ferroelectrics may
be divided into two types: displacive ferroelectrics
and order–disorder ferroelectrics, depending on
the shape of the single-particle potential of their
ions.
Displacive type ferroelectrics are characterized
by low anharmonicity of their single-particle po-
tential, so that their ferroelectrically active ions Sketch of single-particle potential in ferroelectrics of the
order–disorder type (Ω is the tunnel splitting).
are found in a single-well potential in the para-
electric phase. Within the harmonic approximation,
the dynamics of their lattice may be described positions in the unit cell. Above the ferroelectric
in terms of phonon modes. Their crystal lattice phase transition temperature TC these positions
is stable with respect to minor deformations pro- are statistically uniformly populated, however, be-
vided all the normal phonon modes have real low TC a spontaneous population asymmetry ap-
frequencies. In certain ionic or partially ionic crys-
pears, resulting in polarization of the crystal. Such
tals, the frequency ω0 (q) of one of the polar
cases are encountered frequently, e.g., in crystals
lattice modes may turn purely imaginary in the
with hydrogen bonds such as potassium dihy-
harmonic approximation, thus resulting in a fer-
drogen phosphate (KDP), KH2 PO4 , where the
roelectric phase transition of the displacive type
proton can move between two equilibrium po-
(if q = 0). The imaginary nature of ω0 (q) is
sitions (see Fig.) in a hydrogen-bond potential.
a corollary of the fact that the harmonic force
The proton position on the bond is usually de-
constant in ionic crystals, and therefore ω02 (q),
scribed in terms of a pseudospin operator that is
contains two contributions with opposite signs
the usual S = 1/2 spin operator used for quantum
– short-range repulsive forces between ions and
two-level systems. As in displacive type ferro-
long-range Coulomb forces. If the square of the
harmonic frequency ω02 (q) is negative for a given electrics, the phase transition in systems of the
normal mode, as a result of compensation between order–disorder type can be expressed in terms of
short-range repulsive and long-range attractive a soft mode. It is not phonons, however, but rather
forces, anharmonic interactions will stabilize the unstable pseudospin waves that are represented
system above the Curie point TC . Taking into ac- in this case of weak oscillatory excitations. The
count the oscillation anharmonicity effects in the pseudospin mode is present in order–disorder type
quasi-harmonic approach yields a value of the ob- ferroelectrics, in addition to usual phonon modes
served quasi-harmonic frequency that is real and arising from harmonic lattice oscillations. There is
positive, however, it depends on the temperature, another important distinction between displacive
and the anharmonic contribution drops as the tem- and order–disorder systems in respect to soft mode
perature decreases. The frequency of the so-called dynamics. In displacive type systems, the soft
soft mode decreases down to zero at the approach mode always has a resonant nature, although it can
to the Curie point, with the Cochran relationship be highly damped. In order–disorder type systems
(W. Cochran) observed, ω02 (q) = k(T − TC ). The the imaginary part of the soft mode frequency is
low-temperature phase structure is determined by always nonzero, and the real part is not zero ex-
frozen displacements corresponding to the soft cept for a strongly split ground states determined
mode. by tunneling (as shown in Fig.); otherwise the sys-
Order–disorder type ferroelectrics are charac- tem response has a relaxation but not a resonant
terized by strong anharmonicity of the single- nature. This takes place in ferroelectrics similar to
particle potential (see Fig.). Owing to this effect, NaNO2 or triglycinesulfate that are characterized
ions (or radicals) determining the crystal polariza- by negligible tunneling. The resonant nature of the
tion can be located in two or more equilibrium po- response has been detected in crystals similar to
sitions, displaced relative to centrally symmetric KH2 PO4 characterized by strong tunneling.
FERROMAGNET 457
as a permanent magnet (see also Hard magnetic they are localized within the atom of the crystal in
materials). During a variation of the magnetiza- almost the same way as in an isolated atom.
tion of a ferromagnetic material its dimensions and In the theory of ferromagnetic metals there
size vary slightly (see Magnetostriction), and in the are two main approaches: for d-electron ferro-
adiabatic case the temperature may also change magnetic metals there is the model of delocal-
(see Magnetocaloric effect). In addition anomalies ized electrons or band magnetism (Frenkel, Bloch,
are observed in some nonmagnetic properties of Stoner, Moriya, et al.), for f -electron ferromag-
the material: elastic moduli, specific heat, thermal netic metals there is the model of localized elec-
expansion coefficient, etc.; the maxima of these trons (see Heisenberg magnetism). The fact that
singularities occur in the vicinity of TC or at the in the ferromagnetic metals Fe, Co, Ni the average
positions of other phase transitions. magnetic moment (in Bohr magneton units μB :
Ps = Ps = Ms /(nμB ), where n is the number
FERROMAGNETISM OF METALS AND ALLOYS of atomic magnetic moments per unit volume) is
Phenomenon of ferromagnetism observed in less than that in the isolated atoms, and in addi-
three of the 3d-transition metals (Fe, Co, Ni), in tion they have fractional values such as 2.2, 1.7,
six of the 4f -transition metals (lanthanides, rare- 0.6 instead of the corresponding 4, 3, 2 values
earth elements: Gd, Tb, Dy, Ho, Er, Tm), in a in the isolated atoms, gives support to the model
series of alloys (ordered as well as amorphous), of delocalized electrons. Further support is the
and in intermetallic compounds. They may be clas- high value of the low-temperature electronic spe-
sified according to the electronic structure of the ci c heat of transition metals compared to normal
component atoms: metals. In the case of d-electron metals and al-
loys the magnetic properties are determined only
(1) Alloys of the d- and f -transition metals with
by the spin magnetic moment of the d-level, be-
each other: (a) pure ferromagnetic transition
cause the orbital moments are “quenched” and the
metal alloys (Fe–Ni, Co–Ni, Fe–Co–Ni, Fe–
coupling between spin and orbital components of
Gd, Gd–Dy, etc.); (b) transition metal fer-
the d-electron angular momentum in the level is
romagnetic materials with antiferromagnetic
broken by the strong crystalline electric field (lig-
and paramagnetic metals (Fe–Cr, Ni–Ti, Co–
and field), therefore the exchange interaction acts
Pt, Gd–V, Y–Ni, etc.); (c) transition antiferro-
between the individual electrons with spin J =
magnetic metals with transition paramagnetic
S = 1/2. Fig. 1 shows that for the d-electron fer-
metals (Cr–Pt, Mn–Pt, etc.).
romagnetic metals the theoretical dependence of
(2) Alloys of transition metals with normal (non-
the magnetization on the temperature Ms (T ) cal-
magnetic) elements: (a) transition ferromag-
culated for J = 1/2, coincides most closely with
netic metals with normal elements (Ni–Cu,
experiment.
Co–Ag, Fe–Al, etc.); (b) transition 3d-anti-
The magnetism of the delocalized electrons
ferromagnetic metals (Mn, Cr) with normal
may be described by the single-electron approxi-
metals (Heusler alloy types): Cu2 MnM, where
mation. A free electron in the periodic crystal e ld,
M = Sm, Al, Ge, Zn, As, In, etc., and also
taking into account only its interaction with the
Mn4 N, Mn–Au, Mn–Bi, Mn–Te, etc.; (c) tran-
electric field of the ion cores (neglecting dynamic
sition paramagnetic metals with normal ele-
electron correlation) has a standard band theory
ments (Zr–Zn, Sc–In, Au–V, etc.).
spectrum which consists of bands of finite width,
In the ferromagnetic alloys and intermetallides separated by the forbidden gap
E, with a com-
a transition metal having an unfilled internal d- or plex dispersion law E(p), only coinciding with the
f -shell with an uncompensated spin and orbital quadratic form near “the bottom” and “the top”
magnetic moment (Hund rules) can be one of the of the bands. The external magnetic field H in
components. The d-electron wave functions in the a paramagnetic metal changes the distribution of
crystal overlap considerably, which can lead to the electrons with the spin projection values, equal to
bonding or delocalization of these electrons. For +1/2 or −1/2 (i.e. up “+”, or down “−” spins):
deep lying f -electrons there is no such overlap, so the levels which were degenerate at H = 0 are
460 FERROMAGNETISM OF METALS AND ALLOYS
Fig. 1. Dependence of the normalized magnetization Fig. 2. Shift of the density of states curves for spin up
on the normalized temperature. Theoretical curves are (+ spin) and spin down (− spin) electrons in an applied
shown for spin J = 1/2, 1 and ∞. magnetic field.
Fig. 3. Dependence of the atomic saturation magnetic moment on the number of electrons per atom.
At T < TC the magnetic properties are well ex- the paradox concerning the Curie–Weiss law in
plained by a correlated electron model, although the model of delocalized electrons, and provides
at T > TC the Curie–Weiss law is closely obeyed. a unified picture of ferromagnetic metals. In met-
The same fact is observed for paramagnetic met- als close to ferromagnets (Pd and Pt), due to
als, close to ferromagnets (of Pt type). T. Moriya the exchange amplification in the paramagnetic
developed a general theory of spin fluctuations state, local spin fluctuations (paramagnons) are
which provides a unified description of ferromag- formed which tend to be conserved, i.e. to pass
netic metals, ferromagnetic semiconductors and over to the state of “nuclear paramagnetism” or to
insulators, including in the limit the models of de- that of a paramagnet with exchange amplification.
localized and localized electrons (Moriya theory). Yu.P. Irkhin et al. showed that the strong depen-
For a general approach to spin fluctuations the dence χpm (T ) at T > TC in the ferromagnetic
Rhodes–Wohlfarth curve (P.R. Rhodes, E.P. Wohl- metals may be explained by the sharp dependence
farth, 1963) (see Fig. 4) provides the dependence of the density of states g(E) near EF . This is im-
of Pc /Ps = Mc /Ms on TC for ferromagnetic met- portant in the general theory of spin fluctuations
als, alloys and intermetallides. Here Ps and Pc are because the spin fluctuations themselves depend
the average atomic magnetic moments in the units strongly on the type of g(E). It is also necessary
of μB , calculated respectively from Ms at T = 0, to take into account the hybridization of s- and
and from the constant C in the Curie–Weiss law d-bands (s–d exchange); there are some reasons
under the assumption of localized magnetic mo- to think that the d-band in ferromagnetic metals is
ments, that is C = nμ2B Pc (Pc + 2)/(3kB ). split to two bands.
In the model of localized electrons Pc /Ps = 1, The reduction of
Ms with the increase in
and in the model of delocalized electrons Ps → 0, temperature at low temperatures follows from
which follows from the general theory of spin fluc- the appearance of magnetic excitations, i.e. spin
tuations, and agrees with experiment (see Fig. 4). waves (magnons). In the case of ferromagnetic
The values of Pc /Ps are widely distributed along metals, in addition to the exchange of magnons,
the length of the curve. This points to the fact Stoner type excitations are also possible. Magnons
that general theory of spin fluctuations explains lead to the Bloch law dependence
Ms ∝ T 3/2 ,
462 FERROMAGNETISM OF METALS AND ALLOYS
Fig. 4. Dependence of the ratio Pc /Ps of various materials on the Curie temperature TC .
and Stoner excitations lead to the parabolic depen- it is necessary to consider the many-electron sys-
dence
Ms ∝ T 2 . For weak ferromagnetic metals tem of a 4f -shell. Due to weak overlapping of the
the model of delocalized electrons leads to the wave functions of 4f -electrons of adjacent lay-
dependence
Ms ∝ T 2 at not very low temper- ers of the crystal lattice direct f –f exchange is
atures, T < 0.5TC . vanishingly small. Moreover, the values of TC in
The magnetic properties of ferromagnetic met- ferromagnetic metals of the rare-earth elements
als, alloys and intermetallides are caused by rare- reach 300 K (in Gd), which is brought about
earth atoms, whose partly filled inner 4f -shells by the indirect exchange interaction with adja-
which are shielded by surrounding outer closed
cent 4f -shells through the conduction electrons,
5s 2 5p6 shells determine the magnetic moment.
the so-called s–f exchange (Shubin, Vonsovski,
Unlike the d-metal case, these shells are not sub-
Zener, Ruderman–Kittel–Kasuya–Yosida interac-
jected to the effect of the surrounding ligand fields,
tion). The exchange integral Aind has an alternat-
which quenches the orbital angular momentum
in d-metals. In 4f -metals the magnetic coupling ing sign, is a variable function of the distance r
is small compared to the electrostatic fields, and between electrons with the dependence ∼ r −3 ,
even relative to the spin–orbit interaction within and its maximum value is of the order of A2sf /EF .
the shell, therefore the magnetic moment of a rare- The oscillating characteristic of the Ruderman–
earth ion is determined by the sum of its spin and Kittel–Kasuya–Yosida (RKKY) exchange interac-
orbital momenta. Thus for 4f -shells it is not pos- tion has been experimentally verified by nuclear
sible to use the single-electron approximation, but magnetic resonance. The long-range character of
FERRONS 463
FIBER OPTICS
A branch of optics involving the propagation
of light along fiber light guides, and related phe-
nomena. Fiber optics appeared in the 1950s, and
during the first 20 years of its development single
fiber light guides or bunches of them with regu-
lar and irregular packing and lengths up to several
meters were used. They were made from multi-
component optical glasses, and had a transmission Fig. 1. Optical loss spectrum of a single-mode fiber light
factor in the visible region of only 30–70% at the guide with SiO2 core and SiO2 :F coating, showing the
length of one meter. The numeric aperture of light attenuation, dB/km (vertical scale), plotted versus the
guides can reach 0.5–1. Light guides are mostly wavelength λ, μm (horizontal scale).
used in instrument making, particularly in techni-
cal and medical endoscopy for illuminating not so
easily accessible regions (nasopharynx, stomach,
cavities in the machines, etc.) and for transmission
of images. In the latter case a regular packing of
many fibers is used. The resolution is determined
by the number of fibers, the diameter of the light
guide core and the quality of their fabrication, and
it is usually about 10 to 50 lines per mm.
Fiber light guides of quartz based glass with
small optical losses (∼1 dB/km) have been de-
veloped for the near-IR region. These guides have
minimal optic losses in the range of wavelengths
1.3 to 1.6 μm, as shown in Fig. 1. Their trans-
Fig. 2. Cross-section and profile of the refractive index
mission factor is about 50% at the length of sev- along the cross-section for single-mode (a) and multi-
eral kilometers. As a result, new applications of mode (b) light guide.
fiber optics had appeared, such as fiber telecom-
munication systems, transmission of telemetric
information, and sensors for monitoring differ- the near-IR range, and from several tens to sev-
ent physical values (magnetic field, temperature, eral hundreds micrometers for multimode ones.
acoustic waves, rotation speeds, etc.). An infrared The difference between the refractive indices of
guide in its simplest version is a long filament with the shell and core materials, providing light chan-
a core from highly-transparent glass, surrounded neling by means of total internal re ection at their
by a shell from material with a lower refractive in- boundary, is, as a rule, 1 to 2% for multimode
dex (Fig. 2(a)). Usually glass is used for the shell, light guides, and several tenths of a percent for the
but different polymers can also be employed. Fiber single-mode ones.
light guides are divided into multi- and single- In the mid-1980s, multimode light guides with
mode ones depending on the number of modes a refraction index profile distributed over the
which can be propagated along them. The core di- cross-section (gradient light guides) and single-
ameter is usually 5 to 8 μm for a single mode in mode light guides (Fig. 2(b)) became widely used.
466 FIELD DESORPTION
FIELD DESORPTION
See Field emission of atoms.
FIELD EFFECT
Bringing about a change of the near-surface
electrical conductivity by the application of an elec- (a) Electrical connections of a metal (M), insulator (I),
tric field normal to a solid surface (see Fig.). As semiconductor (S), metal structure showing the applied
a rule, the field effect is observed in semicon- voltage V and the current density J in the space charge
ductors, but its appearance in a metallic phase is region (SCR). (b) Plot of the voltage V as a function
also possible (in particular, in thin lms ). An in- of the distance x from the insulator (I) into the semi-
duced electric charge is distributed between the conductor (S) in the subsurface regions of enhancement,
bulk semiconductor (space charge region) and the depletion, and inversion. The energies of the bottom of
the conduction band E C , the top of valence band E V , the
surface states. The concentration of free electrons
Fermi energy E F , and the surface band bending energy
(holes) in the space charge region is determined
eϕ s are indicated. In Fig. (b) electrons (solid circles) are
by its surface potential ϕ s (surface band bending shown just above the bottom of the conduction band, and
eϕ s ) (see Surface phenomena in semiconductors) holes (open circles) are shown just below the top of the
and the position of the Fermi level E F in the bulk valence band.
semiconductor. An external field establishes three
states of the space charge region (and, correspond-
ingly, three regimes of the field effect): carriers near the surface exceed in concen-
tration the majority carriers in the bulk, and
(i) the enhancement with majority carriers (in an inversion layer emerges. When the surface
comparison with the bulk of the n-type semi- bending becomes so pronounced that the va-
conductor this means a strong bending of the lence band crosses the level E F (not shown),
bands downward, i.e. eϕ s > 0); then a degenerate electron gas forms (in cer-
(ii) the formation near the surface of the deple- tain situations this includes two-dimensional
tion layer where the bands are bent up; in quantization of the carrier band spectrum).
this case the conductivity relative to the field The formation of the inversion layer again
effect decreases with the subsequent upward produces an increase of the conductivity, but
bending; this time due to minority carriers. Thus a de-
(iii) as the concentration of minority carriers (with pendence of the conductivity on the applied
respect to the bulk) increases near the sur- voltage V has the shape of a curve with a
face, a point is reached where the minority minimum. From the analysis of this curve
FIELD ELECTRON EMISSION 467
emission, and then to the thermionic emission am- FIELD EMISSION MICROSCOPY, eld electron
plification by the field due to the Schottky effect microscopy
(see Schottky diode). At j ∼ 1012 A/m2 , a rapid A method for obtaining a factor of 105 –106
destruction (explosion) of the emitter tip occurs, times magnification of the image of the surface of
a cathode plasma torch arises, the emission cur- a solid by using eld electron emission from the
rent rises by about a factor of 100, and field surface.
electron emission transforms to explosive elec- The sample which serves as the cathode in
tron emission which is the initial stage of vacuum the field of an electron microscope is a needle
breakdown. of a conducting material having a tip radius of
For applications see Field-emission cathode. curvature r ∼ 10−7 –10−8 m. The anode is a lumi-
nescent screen placed at a distance of R ∼ 10 cm
from the tip. A high vacuum (residual pressure
FIELD-EMISSION CATHODE 10−7 –10−8 Pa) is maintained inside the micro-
A cold cathode that emits electrons by means scope. Several thousand volts are applied between
of eld electron emission. the screen and tip, which induces an electric field
Field-emission cathode materials are various of strength 109 –1010 V/m near the tip surface, as
metals (clear or coated with activating layers), and needed for the field electron emission. The emit-
metal-like compounds of semiconductors. A typi- ted electrons are accelerated radially and produce
cal field-emission cathode is a needle produced by on the luminiscent screen an image of the needle
chemically or electrochemically etching a wire or tip magnified approximately R/r times. The reso-
a rod. The limiting field electron emission current lution of 2–3 nm is limited by electron diffraction
density of a needle is determined by the heat- of the tangential components of the electron ve-
ing of the needle tip by the current, and it has locity. The device thus cannot quite resolve the
the value ∼1011 A/m2 . With typical needle tip atomic structure of the surface. The contrast of
radii r ∼ 0.1–1 μm, the limiting field electron the field electron image is determined by the lo-
emission current is 0.01–1 A. To obtain higher cal curvature (determining the local electric field)
currents the emitting area can be increased by in- and the work function of different parts of the
creasing r and thereby increasing the operating tip surface. Due to its small size, the tip is typi-
cally a monocrystal. The field electron image of
voltage (e.g., V ≈ 5Er and E ∼ 109 –1010 V/m),
a tip, smoothed and cleansed from impurities by
or by using many-needle systems, blades, etc. The
high-temperature heating, exhibits the lattice sym-
stability of a field-emission cathode is degraded
metry of the tip material. Such a surface is used
by the adsorption of impurities which change its
to study the changes in work function of differ-
work function, and by ionic bombardment which
ent faces under adsorption of different materials,
pits its surface. Ways to increase the stability are
as well as the surface diffusion of adsorbed atoms,
to improve the vacuum; to use materials that ad-
phase transitions in the adsorbed layers, field des-
sorb residual gases; to separate the electron and orption of atoms (see Field emission of atoms),
ion trajectories; and to use field electron emission etc. Processes involving crystal shape changes un-
of microcells where, due to small (1–5 μm) inter- der thermo-field treatments are also investigated
electrode distances and low (20–50 eV) electron by means of field electron emission.
energy, ionic bombardment is strongly suppressed.
Field-emission cathodes are used in e ld emis- FIELD EMISSION OF ATOMS, eld evaporation
sion microscopy, in vacuum microelectronics, and of atoms
in infrared receivers (see Autophotoelectronic Emission or evaporation of surface atoms of a
emission). Electron guns with a field-emission solid or liquid electrical conductor under the ac-
cathode are 103 –104 times brighter than those tion of a strong electric field (of intensity
with heated cathodes, and are used in electron ∼1010 V/m). This emission of adsorbed layers
microscopy and scanning electron microscopy, mi- is also called field desorption. Field evaporation
croanalysis, X-ray microscopy, etc. is the limiting case of positive surface ionization
FIELD ION MICROSCOPY 469
in strong electric fields; its mechanism consists the barrier away from the atom. The probability
in thermal evaporation of ions through the sur- of tunneling depends exponentially on the barrier
face energy barrier, which is lowered by the field. height and thickness, and these depend on the elec-
Emission of multiply charged ions (charge of tric field strength, E, and on the electron energy.
emitted ion may reach 6) is a distinctive property A pronounced field ionization of simple gas mole-
of field evaporation. These ions are believed to be cules occurs at E ≈ (2–6)·1010 V/m, and that
generated by post-ionization, i.e. additional field of alkali atoms and excited molecules occurs at
ionization of outgoing ions at short distances from 108 –109 V/m. The area of the potential barrier
the surface. Field evaporation of refractory materi- near the metal surface becomes reduced due to im-
als is enhanced at adsorption of certain molecular age forces, and the probability of field ionization
gases (H2 , N2 ); as this takes place, emission of with the same E increases noticeably. The field
ions of complexes (hydrides, nitrides) is observed. ionization starts at a critical distance xcr between
Surface evaporation of semiconductors and liquid the atom and surface, at which the energy level of
metals produces a large quantity of polyatomic an atomic electron rises above the Fermi level in
ions. the metal owing to the applied electric field, and
Field evaporation is used in e ld ion microscopy the possibility for an electron tunneling into the
for primary cleaning of the surface of the sample, empty energy levels in the metal is enhanced. If
for step-by-step removal of layers to investigate x > xcr , the tunneling probability decreases due
the internal composition of the sample, at the op- to the widening of the potential barrier, and thus
eration of field ion microscope in the mode of an the ionization zone half-width is a few tenths of
atomic probe, etc. Field evaporation is the basis an angstrom. The field ionization phenomenon is
of operation of a field desorption microscope, and used in e ld emission microscopy, in the construc-
liquid metal sources of ions. tion of ion sources for mass-spectrometers (see
Mass spectrometry), in scanning tunneling micro-
FIELD EMISSION OF ELECTRONS, cold scopes, etc. A considerable increase of the ion
emission of electrons current is achieved in cryogenic ion sources where
Electron emission by conducting bodies at low the field ionization of gas molecules is concen-
temperatures under the influence of external and trated at the surface of the tip.
internal electric fields. Cold emission involves
eld electron emission and emission of nonequi- FIELD ION MICROSCOPY
librium electrons (including hot ones) excited by Method of investigation of atomic structure of
internal electric fields. Generally the term cold pure metals and their alloys. The method is based
emission does not apply to various types of elec- on eld ionization of noble gas atoms near the
tron emission arising from irradiation of a cathode surface of the sample under investigation. The au-
(photoelectron emission, secondary electron emis- toionic image of the surface is produced by an
sion, ion–electron emission, etc.) or from other ion projector for a sample in the form of a needle
types of action upon it (exoelectronic emission, with the radius of the curvature of the point in the
etc.). range 10–100 nm. Because of its high magnifica-
tion (>106 ) an ion projector provides information
FIELD EMISSION, THERMAL on the arrangement of individual atoms, and struc-
See Thermal electron eld emission. tural details of atomic sizes on the surface of the
point. Field ion microscopy is used to study point
FIELD IONIZATION, autoionization defects of a crystal structure (vacancies, intrin-
The ionization of atoms and molecules in a sic and impurity atoms), grain boundaries, stacking
strong electric field. faults, twin boundaries, segregations of impurities
When a bound atomic electron resides in a at defects and various surface phenomena, e.g.,
potential well and an electric field is applied, a po- surface diffusion, adsorption, corrosion, oxidation,
tential barrier is formed at one side of the well. deposition of thin lms . Limitations of field ion mi-
The basic mechanism of field ionization is the tun- croscopy are related to the necessity to use needle-
neling of an electron (see Tunnel effect) through shaped samples, which must be cooled to helium
470 FILAMENTARY CRYSTALS
temperatures, and the fact that the surface of many FILM VIBRATIONS
metals is severely damaged or destroyed by strong See Vibrations of rods and plates.
electric fields. Field ion microscopy does not pro-
vide a direct identification of the chemical nature FILTER, NUCLEAR
of observed atoms. This problem is solved by us- See Nuclear lte rs.
ing a combination of an ion projector with a mass
spectrometer (see Mass spectrometry). FINE STRUCTURE of electron paramagnetic
resonance (EPR) spectra
FILAMENTARY CRYSTALS An observable line splitting for S 1 para-
See Thread-like crystals. magnetic centers (here S is the electron or ef-
fective spin) due to the indirect influence of the
FILM GROWTH, RADIATION-INDUCED crystalline electric field (molecule, ligand) upon
See Radiation-induced lm growth. the position of spin levels, via the interaction of
the electric field with the orbital angular momen-
tum, and of the latter with the spin. Fine structure
FILM IONIC MODIFICATION
energy splittings are often comparable to Zee-
See Ionic modi cation of lms.
man level splittings for paramagnetic materials in
magnetic fields close to one tesla in magnitude.
FILMS
This structure is described by including in the
Layers of a particular material, commonly spin Hamiltonian the irreducible tensor Tlm (S) or
grown on a substrate for different purposes. As a Stevens’ operators Olm (S) (see Irreducible tensor
rule, the properties of a film differ from those of operators) with even l (values lm allowed for par-
the bulk material: there exists a transition region ticular symmetry groups of the spin Hamiltonian
between lm and substrate, the film shows size are listed in Spin Hamiltonian). Spin Hamiltonian
effects (see Thin lms) and exhibits internal me- parameters for operators with l = 2 are sometimes
chanical stresses. The first applications of films represented in the form of a Cartesian D-tensor.
in engineering were for protective and decorative Typical forms of the spectra demonstrating the fine
coatings. Films form the basis of solid-state elec- structure and the dependences of splitting on the
tronics, hence numerous preparation techniques orientation of the applied magnetic field are shown
have been developed, and electrical, optical, me- in the figures in Hyper ne structure and Angular
chanical, and other properties of films have been dependence of spectra.
widely studied. The development of production
methods like epitaxy and photolithography, fol- FINE STRUCTURE of X-ray absorption spectra
lowed by the attainment of microminiaturization, A system of extrema in X-ray absorption spec-
reproducibility of parameters, and high reliability tra. The greatest interest is associated with the
of radio electronic circuits and devices based on structure located in the range of threshold ion-
them, promoted the wide application of films in ization energy maxima of internal atomic shells
electronic engineering. Modern methods of film in solids (so-called absorption edges) which are
technology, including molecular beam epitaxy and related, mainly, to X-ray excitons, and the struc-
electron, ion, and X-ray lithography, continually ture whose energy exceeds the energy of edges
increase the number of active elements that can be by 10–1000 eV. The latter structure is called ex-
deposited per unit surface area of film. tended X-ray absorption fine structure (EXAFS);
See also Thin lm growing, Thin lms, Amor- this results from the interference of the photo-
phous lms, Diamond-like carbon lms, Epitaxial electron wave emerging from the region of the
iron garnet lms, Granular lms, Insular lms, ionized atom, with the wave of the same elec-
Langmuir–Blodgett lms, Magnetic lms. tron reflected from the neighbors of this atom.
Such structure contains information about the spa-
FILMS, GRANULAR tial distribution of atoms in the solid, and this is
See Granular lms, Insular lms. employed for the structural analysis of crystalline
FINITE ELASTIC STRAIN 471
Oscillating part of K-absorption spectrum of Fe in deoxyhaemoglobin, and (insert) its Fourier transform which reflects
the radial distribution of the Fe neighbors (solid lines are experimental data; dashed lines correspond to analytical calcu-
lation).
and noncrystalline solids, as well as polymers and space, the components of curvature tensor calcu-
macromolecules (see the figure). lated with the help of the metric tensor gij or Gij
are equal to zero. This imposes on the deforma-
FINITE ELASTIC STRAIN, nite elastic tion tensor components some constraints that are
deformations called compatibility criteria. The application of
Elastic strains with values comparable to unity. these criteria means that disclinations and dislo-
Rubber is an example of a medium in which large cations do not appear during deformation of the
deformations are possible (see Elasticity). The the- medium. The equations of motion of an elastic
ory of finite elastic deformations is expressed in medium for finite elastic deformations have the
the language of differential geometry. If a medium form Di σ ij + ρF j = ρa j , where σ ij is the stress
before deformation is characterized by the met- tensor, ρ is the medium density after deformation,
ric tensor gij and after deformation by Gij then F j is the bulk force density per unit mass, and a j
the strain tensor is uij = (Gij − gij )2 . Let the is the acceleration of a unit mass of the medium.
The stress tensor σ ij is symmetric if the moments
deformation vector v = vi Gi (see Displacement
of stresses acting on the medium are equal to zero,
vector), where the Gi are basis vectors after defor-
otherwise it can be asymmetric. If the surface
mation, then uij = (Dj vi + Di vj − Di v r Dj vr )2 ,
forces P are defined on the surface of an elastic
where the symbol Dj means covariant differ-
body characterized by a normal vector n then the
entiation: Dj vi = ∂vi /∂q i − Γjsi vs , the q i are boundary conditions look like σij nj = P i . The
Lagrangian coordinates of a point in the medium, deformed medium is called an elastic medium if a
and the Γjsi are Christoffel symbols. Deformation unique dependence exists between stresses and de-
tensor invariants (see Invariants of elasticity the- formations, and it is called a hyperelastic medium
ory) I1 , I2 , I3 are defined as the coefficients in the (see Hyperelastic material) if it can be described by
expansion of the determinant |g rm Gms + δsr | = an elastic energy W which is a unique continuous
λ3 + I1 λ2 + I2 λ + I3 in powers of λ. Thus, I1 = scalar function of the deformation tensor uij .
g rm Grm , I2 = Grm grm , I3 = G/g. Here G = It is possible to consider W as a polynomial
|Gij |, g = |gij |. For an incompressible medium of the components uij without loss of general-
I3 = 1. If a deformable medium before and af- ity, as it is possible to approximate a continuous
ter deformation can be located in a Euclidean function to any precision by a polynomial, and
472 FINITE ZONE POTENTIAL
the definition of W is reduced to the creation to compute any single-particle property of an elec-
of a polynomial basis of deformation tensor in- tron within the finite zone potential, e.g., the
variants. For a known deformation energy the electron–phonon interaction constant. Generally,
deformation tensor is determined by the formula the finite zone potential is a quasi-periodic func-
σ ij = (δW/δuij + δW/δuj i )2 . For a uniform tion and becomes periodic only when its periods
isotropic body W is a function of I1 , I2 , I3 satisfy the mutual commensurability condition. In
only. In practice the following expressions for the the case of a smooth periodic potential, the band
deformation energy of a uniform isotropic body gaps in the spectrum of the respective Schrödinger
are commonly used: harmonic (semilinear) en- operator rapidly shrink with energy, so any peri-
ergy, Murnaghan deformation energy (F.D. Mur- odic potential could be approximated by the finite
naghan, 1967), Rivlin–Saunders deformation en- zone potential within any desired accuracy by ne-
ergy (R.S. Rivlin, D.W. Saunders, 1951), etc. glecting small band gaps. This potential is an
Solving elasticity theory problems for finite elas- exact solution to the Peierls problem of the self-
tic deformations can be carried out by the method consistent state of conduction electrons within the
of successive approximations when the deviations lattice (see Peierls transition), yielding polarons
from linear behavior are small, or by the so-called and charge density waves as particular cases. The
semi-inverse method when the deformed state is finite zone potentials are a periodic general case of
defined beforehand with an accuracy to within non-reflective potentials.
constants and some unknown functions, and it is
necessary to determine these constants and func- FIRST-ORDER PHASE TRANSITION
tions to satisfy the boundary conditions. Finite The phase transition of a material from one
elastic deformation theory can be converted into thermodynamic phase to another, accompanied by
linear elasticity theory when it is possible to ne- a certain quantity of heat (heat of phase transition)
glect higher powers of the deformation vector being either absorbed or evolved. Examples of first
components and their derivatives in comparison order phase transitions are melting and crystalliza-
with their first powers. tion, certain ferroelectric and magnetic transitions,
etc. If the symmetry of the system does not change
FINITE ZONE POTENTIAL at the transition, then the curve of the first-order
A potential u(x) of a one-dimensional Schrö- transition on the phase diagram may end at the
dinger operator H = −∂x2 + u(x) (where ∂x de- critical point. The free energy and the thermody-
notes differentiation with respect to coordinate x) namic potential of the system are continuous at the
with the spectrum as a sum of a finite number first-order transition point, but their first deriva-
of intervals of a double absolute continuum. An tives have discontinuities at this point. The order
arbitrary N -zone potential, its respective eigen- parameter, susceptibility, speci c heat and other
functions, boundaries of spectra, density of states, quantities undergo an abrupt change at a first-order
etc. are expressed via the N -dimensional Riemann transition. An abrupt decrease or increase of the
Θ-function: temperature may result in thermal hysteresis, and
Θ(z) = exp iπ(Bn, n) + i2π(z, n) , the substance may reach a supercooled or super-
heated metastable state (pure water, e.g., may be
n∈CN
supercooled down to −40 ◦ C without crystallizing
where z ∈ CN , B is a symmetrical complex to ice). The presence of thermal hysteresis is often
N × N matrix possessing a positive definite imag- used for identifying first-order phase transitions.
inary part. An example of the N -zone potential is
FISKE STEPS in tunnel junctions (M.D. Fiske,
u(x) = C − 2d2x ln Θ(αx + β),
1964)
where C ∈ R, α, β ∈ CN are constants. With the Stepwise dependence of the current on the
eigenfunctions in hand, making use of the theory voltage at a Josephson tunnel junction or Joseph-
of residues, one is able to determine matrix ele- son junction, caused by the excitation of standing
ments of any single-particle operator, and hence, electromagnetic waves matched to the Josephson
FLASH HARDENING 473
(h is the plate thickness; ν is the Poisson ratio) boundary of flicker noise, where the dependence
can propagate in plates. These waves exhibit dis- Sf ∼ 1/f α no longer holds, has not been reached
persion, and in both cases their phase velocities are experimentally, although measurements have been
√
cph ≈ ω, the group velocities being cgr = 2cph . carried out down to extremely low frequencies
(∼10−6 Hz). Existing theories of flicker noise
FLEXURE do not explain the absence of the lower limit,
A strained state (see Strain) which emerges in a although in some systems they predict the pos-
specimen under action of forces and moments di- sibility of noise with a dependence close to 1/f
rected normally to its axis, and it is accompanied over a broad spectral range. The spectral density
by distortion of the axis. Normal stress σ devel- of electromotive force noise of the 1/f α type is
oping in a cross-section of the sample results in a generally proportional to the square of the current
moment M of force, normal to the axis, referred flow, indicating that it arises from a broad spec-
to as a bending moment. Static bending tests are trum of two-level systems wth extremely low level
commonly employed for assessment of mechani- cutoffs so the 1/f noise can persist to arbitrarily
cal properties of a material. Two loading schemes low temperatures.
can be used in bending tests of a specimen at rest
on stationary supports: FLUCTUATION DIAMAGNETISM
Anomalous diamagnetism of superconductors
(i) The load is applied by a concentrated force at
above the superconducting phase transition tem-
the midpoint between the supports.
perature Tc , arising from the fluctuation forma-
(ii) The load is applied by two equal forces at two
tion of Cooper pairs. The fluctuation diamagnetic
points equally distant from the supports.
susceptibility χfl is much lower in magnitude
The first scheme leads to a distribution of bend- than the susceptibility of a perfect superconductor
ing moments along the sample length in the shape χ = −1 in SI units (−1/4π CGS, see Meissner
of an isosceles triangle, the moment attaining its effect). Fluctuation diamagnetism becomes more
maximum at the midpoint of the span between the pronounced with the reduction of the system di-
supports. Under the second scheme, the bending mensionality (in thin lms , thread-like crystals and
moment distribution follows the trapezium law, small particles), and also upon contamination of
whereas the magnitude of the moment in the inter- the material by impurities (see Fluctuations in
val of “pure flexure” is constant. Bending tests are superconductors). Measurements of χfl in small
useful to determine mechanical properties of brit- superconducting particles with the size d λ
tle materials, to evaluate the temperature of their (λ is penetration depth of magnetic eld) allow
transition from the brittle state to the viscous state one to determine the width of the critical region,
(see Cold brittleness), and to study dislocation mo- and its dependence on d. The value of χfl may be
bility at a given level of stress and temperature. differentiated from the background diamagnetic
susceptibility of a metal by its strong temperature
FLICKER NOISE, excess noise dependence, which follows a power law.
Fluctuations of the dependence of the spec-
tral density Sf on the frequency of the form FLUCTUATION-DISSIPATION THEOREM
1/f α , where the exponent α is close to unity Theorem of statistical physics, establishing the
(usually 0.8 < α < 1.4). This is observed in di- connection between the uctuations of a system
verse types of conductors: in carbon resistors, in in its equilibrium state and its nonequilibrium
semiconductors, in devices based on semiconduc- properties. It was proposed by H.B. Callen and
tor junctions, in continuous metallic insular lms, T.A. Welton (1951). The reaction of a system to
etc. The quantity Sf decreases with an increase some perturbation under the action of a force f
in frequency, and at high frequencies flicker noise (depending on time t as cos ωt ) which enters as the
is hardly noticeable in the background of other additional term −f x̂ in the system Hamiltonian
noise sources. The highest frequencies at which (x̂ is an operator corresponding to the physical
it has been observed are ≈ 106 Hz. The lower value of x), leads to a variation of the average
476 FLUCTUATION LEVELS
value x of the quantity x to δx = α(ω)f , where elementary excitation spectra in disordered solids.
α(ω) is called the generalized susceptibility of the The identification of this region of the spectrum,
system which determines its nonequilibrium prop- and its detailed investigation, was initially carried
erties. According to the fluctuation-dissipation out by I.M. Lifshits. A series of important and ex-
theorem, the Fourier transform of the correlation acting results substantiating the picture of such a
function fluctuation spectrum was obtained by J. Fröhlich
+∞
and his colleagues.
1
ϕ(ω) =
x(t)
x(0)
2
−∞ FLUCTUATION REGION
The same as Critical region.
+
x(0)
x(t) eiωt dt
is associated with α(ω) by the relationship ϕ(ω) = FLUCTUATIONS in solids
coth[ω/(2k B T )] Im α(ω), where Im denotes Random deviations of physical values xi , char-
imaginary part. The fluctuation-dissipation theo- acterizing a solid, from their average values x̄i .
rem is important for the evaluation of noise in The random, undetermined character of the fluctu-
quantum systems. It is applicable to many systems ations means that in the same physical system sub-
in which correlations are studied, and frictional or ject to constant external conditions the value of xi
fluctuating forces are operating. depends on time (xi = xi (t)), but it is impossi-
ble to predict the explicit form of this dependence
FLUCTUATION LEVELS from the experimental conditions. A description
Energy levels of quasi-particles in the fluctu- of the state of a solid based on a restricted set of
ation region of a spectrum (see Elementary exci- values xi corresponding to the temperature, mag-
tation spectra of disordered solids). The quantum
netization, polarization, concentration of current
states corresponding to fluctuation levels are local-
carriers, etc., can only be approximate. The tran-
ized; they appear at low-probability uctuations as-
sition from a fundamental description based on
sociated with the potential wells of quasi-particles,
mechanical equations of motion for the particles
and they are concentrated at these fluctuations.
belonging to the body, to a simpler description
In a macroscopic (infinite) system the spectrum
based on actual measurements, requires that ap-
of fluctuation levels becomes dense, but the dis-
proximations be made, which obscure detailed
tance separating the centers of the localized of
states corresponding to two fluctuation levels (i.e. information about the initial state of the particles
between corresponding potential wells) grows ex- when their motion is averaged over short intervals
ponentially in proportion to the convergence of of space or time. Thus the fluctuation δxi = xi − x̄i
their energies. As a result, the movement in each provides a measure of the incompleteness of the
such well, which “seems” to the quasi-particle to statistical description of the system. Fluctuation
be infinitely deep, is almost independently quan- process may be characterized with the help of a
tized. The density of fluctuation levels at any correlation function S(t, t ) = δxi (t) δxi (t ), where
particular energy is, as a rule, exponentially small, the top bar means averaging over all possible val-
and it is in fact formed by a single optimal fluc- ues of δxi (t) at t < t , or of δxi (t ) at t > t . In
tuation, for which the energy E corresponds to stationary systems S(t, t ) depends only on the
the ground state in the fluctuation potential well. time difference t − t : S(t, t ) = S(t − t ). The
Nevertheless, within the scope of this exponential function S(t, t ) indicates how the presence of a
smallness in different parts of the fluctuation re- fluctuation δxi at a moment t influences its value
gion there may appear energy dependences of the at the later instant t , where t < t . The function
fluctuation level density, which reflect a different S(t, t ) declines with the growth of |t − t |, eventu-
structure of the optimal fluctuations forming the ally vanishing with the characteristic memory time
spectrum at these regions. The presence of fluc- of small fluctuations. Instead of a time correlation
tuation levels is a characteristic singularity of the function it is possible to use the spectral presenta-
FLUCTUATIONS IN SUPERCONDUCTORS 477
tion determined by the expression where d < ξ(T ) is the smallest dimension of
+∞ the material, εF and kF are the Fermi energy
and momentum, respectively. In pure (type I)
Sω = dt eiωt δx(t) δx(0),
three-dimensional superconductors the width of
−∞ the critical region ε3 ≈ 10−15 is not experimen-
where Sω is called the spectral density of fluctua- tally measurable. However, dirty superconductors
tions of the value x. It is often possible to correlate of reduced dimensionality (films, filaments, gran-
fluctuations with values characterizing the linear ules, lamellar and chain crystals) exhibit values
response of the system to an external field. For of εn ∼ 10−1 to 10−3 , and fluctuation effects
example, the spectral densities of fluctuations of are experimentally observed. Their presence is
heat flow, and of electric currents in a state of ther- evident in the experimentally measurable effects
modynamic equilibrium, determine the values of uctuation diamagnetism, paraconductivity, spe-
the corresponding thermal conductivity and elec- ci c heat anomalies. Above the critical temper-
trical conductivity, respectively. It is possible, by ature the fluctuations lead to the possibility of
measurements of Sω , to find the values mentioned combining electrons into nonequilibrium (fluctu-
above. In addition, fluctuations are the cause of ation) Cooper pairs, with life times near Tc of
noise appearing in weak signal amplifiers, and also τ = π/[8kB (T − Tc )]. Due to the relatively small
in various generators. size ξ of a Cooper pair and the large number of
particles with a negligible probability of forming
FLUCTUATIONS, CRITICAL one, the fluctuations of a superconducting region
See Critical uctuations. should be involved in the pair generation process.
Fluctuations influence various physical para-
FLUCTUATIONS IN SUPERCONDUCTORS meters of superconductors: specific heat, conduc-
Random deviations of the order parameter in tivity, magnetic susceptibility, resistance of tun-
superconductors from its average value which nel junctions, sound absorption factor, etc. Sev-
minimizes the free energy F . The probability W eral diverse factors contribute to this. The first
of the appearance of fluctuations is determined
concerns the influence of fluctuation-generated
by Gibbs factor W ≈ exp(−
E/T ); it depends
Cooper pairs themselves on these physical charac-
on the volume of the fluctuations
V , and on
teristics of a normal metal. Thus in superconduc-
the free energy of the condensate in the super-
tors at temperatures somewhat above the critical
conductor
F ≈
V Bc2 (T )/(2μ0 ), where Bc is
temperature the mechanism of paraconductivity,
the thermodynamic critical field. The extent of the
predicted by A.G. Aslamazov and A.I. Larkin
fluctuations is of the order of the coherence length
(1968), results in the additional transport of elec-
ξ(t) = ξ(0)(1 − T /Tc )−1/2 . Here
trical charge by fluctuation-generated Cooper
vF vF 1/2 pairs. The corresponding correction factor to the
ξ(0) = min , l , conductivity appears to be singular near Tc :
kB Tc kB Tc
σ/σn ∝ [Tc /(T − Tc )]2−n/2 , where n is the di-
where vF is the Fermi velocity, l is the electron
mensionality of the superconducting system. The
mean free path, and Tc is the critical temperature.
second contribution to the physical parameters
Fluctuations are strongest in superconductors of
of superconductors above Tc is caused by vari-
reduced dimensionality n < 3 (n is a number of
ations of the single-electron density of states. It
dimensions along which the size of superconduc-
is not so significant very near Tc because it has
tor exceeds ξ(T )), and they show their presence
a much weaker dependence on T − Tc than the
within a narrow temperature range δT near the
first one, but it is comparable to it in respect to the
phase transition point at T = Tc . The width of the
temperature range associated with its contribution
critical region where fluctuations are strong is de-
(
T Tc ). Finally, in sufficiently dirty supercon-
termined by the Ginzburg number
ductors (l < ξ ) above Tc there is a third fluctuation
1/(2−n/2)
δT εF kF−3 contribution, the so-called Maki–Thompson con-
εn ≡ = , tribution (K. Maki, 1968; R.S. Thompson, 1970).
Tc Tc ξ n (0)d 3−n
478 FLUCTUATIONS, MESOSCOPIC
pairs (see Frenkel defect). At sufficiently high an- atmosphere of light gas molecules in thermo-
gles of momentum transfer to the primary atom, a dynamic equilibrium with each other, is de-
defocusing of atomic collisions takes place. scribed by the equation
∂f 3
∂ σ ∂2f
3
FOKKER–PLANCK EQUATION (A.D. Fokker, = −γ (Pi ; f ) + ,
∂t
i=1
∂Pi 2
i=1
∂Pi2
1914; M. Planck, 1917)
Equation for the propagation density of prob- where P1 , P2 , P3 are the Cartesian compo-
abilities f (Q, t), where Q = (q1 , . . . , qn ) is a set nents of the heavy-particle momentum vector,
of dynamic variables. A general representation of f = f (P1 , P2 , P3 , t) are probability distribu-
the Fokker–Planck equation has the form tion densities for the heavy particles, and σ
is the strength of the effect that the light gas
∂f ∂ n
molecules have on the heavy particle.
=− ai (t, Q)f
∂t ∂qi For the applicability of the Fokker–Planck
i=1
equation to the description of the time variation of
1
n
∂2 the distribution density f (Q, t) it is necessary to
+ bik (t, Q)f . fulfill the following conditions (see Kinetic equa-
2 ∂qi ∂qk
i,k=1 tion):
The coefficients ai (t, Q) are called drift coeffi- (1) the variation of f (Q, t) in a short time inter-
cients, and bik (t, Q) are matrix diffusion ratios. val
t is totally determined by the function
f (Q, t) itself, and by the forces applied at the
Many equations of physical kinetics are re-
moment t , i.e. there is no “memory” of the
ducible to the Fokker–Planck equation, such as
character of the system evolution in the past,
(a) self-diffusion and interdiffusion processes (see (2) the complete evolution process of the set of
Interdiffusion) in systems with large numbers physical values (q1 , . . . , qn ) consists in a large
of particles (gases, liquids) are described by number of individual elementary acts, which
the Fick equation occur at random times, and have random du-
rations. The process of variation of the dis-
tribution of values qi is far slower than the
∂f ∂2f
3
=D , elementary processes which generate it, i.e.
∂t ∂x 2
i=1 i the average duration of events comprising the
microscopic mechanism of evolution is much
where x1 , x2 , x3 are Cartesian coordinates of less than the characteristic times of variation
a designated particle, f = f (x1 , x2 , x3 , t) are of the values qi ;
probability density distributions of the particle (3) the average variations of the values q1 , . . . , qn ,
location within the space, and D is the coeffi- on which f (Q, t) depends in every elemen-
cient of diffusion; tary act, are low compared to their individual
characteristic values.
(b) the process of Brownian motion of a foreign
particle in a liquid is described by the Smolu- For all of the above-mentioned physical phe-
chowski equation nomena these conditions are fulfilled because the
described processes proceed much slower than the
3
3 elementary acts (collisions of particles) generat-
∂f Fi ∂f ∂2f
=− +D , ing them, and the average relative change of the
∂t γ ∂xi
i=1
∂x 2 i=1 i dynamic variables (coordinates, momenta) at each
collision is very small. It is necessary to take into
where F1 is the force of gravity which acts on consideration that the actual processes involved in
the particle, γ is a coefficient of friction; the variation of the physical values have a statisti-
(c) a relaxation distribution process involving the cal nature, hence conditions 1 and 2 are only ap-
momentum of a heavy gas impurity in an proximately satisfied, and therefore describing the
FORCE CONSTANTS 481
physical values with the help of the Fokker–Planck states, forbidden via one type of interaction, might
equation is only an approximation, whose accu- be easily observable in other circumstances. For
racy depends on the accuracy of fulfilling these instance, electric dipole-forbidden transitions un-
conditions. This is clearly demonstrated by the ex- der light absorption can be allowed for Raman
ample of diffusion processes of atoms in solids, scattering of light, and vice versa (see Mutual ex-
and especially by the passage of gases through clusion).
pressed powders when a random arrangement of
the medium reinforces the random variation of the FORBIDDEN TRANSITION, DIPOLE
particle motion. Since this state varies so slowly, See Dipole-forbidden transition.
there remains some memory of the past evolution
FORCE CONSTANTS
of the particle, which is expressed in the statistical
Coefficients in the quadratic term for the se-
connection between the separate elementary acts ries expansion of the crystal potential energy U
associated with the particle motion, and this con- for displacements of atoms from equilibrium po-
nection may become pronounced to such an extent sitions us (n) (n labels the unit cells in the lattice,
that condition 1 would not be satisfied. In par- s labels the atoms in the unit cell):
ticular, there is a so-called percolation threshold
1 ik
(see Percolation theory) depending on the para- U = U0 + βss (n, n )uis (n)uks (n ),
meters characterizing the medium, above which 2
diffusion ordinarily cannot occur, and hence the where the summation is carried out over all indices
Fokker–Planck equation only applies in the above (here i, k are the indices of Cartesian coordi-
mentioned physical systems when the parameters nates). The symmetrical and real force constants
ik (n, n ) define the type of crystal lattice vibra-
βss
of the medium are much less than their threshold
values. tions and the spectrum of their frequencies:
ik (n, n ) = β ki (n , n).
βss ss
FORBIDDEN BAND
See Band gap. They obey the following relations:
ik (n, n ) = 0,
βss
FORBIDDEN ENERGY ZONE ns
See Band gap.
ik (n, n )R l (n ) =
βss il (n, n )R k (n ).
βss
s s
FORBIDDEN QUANTUM TRANSITIONS n s n s
Conventional term for transitions with an in- Here Rsk (n) is the kth component of the coordinate
tensity much less than that of allowed transitions. of the sth atom in the nth cell. The correlations
They include the transitions with
M = ±1, between force constants may be obtained from the
n = 1, and, in strong magnetic fields, those tran- symmetry of each specific lattice (note that some
sitions with the simultaneous changing of the pro- terms may vanish). In an infinite (uniform) crystal
jections of two spins
M1 =
M2 = ±1,
n = we have
1, or
M1 = −
M2 = ±1,
n = 1, where
n is ik (n, n ) = α ik (n − n ).
the change in the number of photons (see Fig. in βss ss
Hyper ne structure). Included also in the forbid- In the limiting case of long-wavelength acous-
den category are many-quantum transitions with tic vibrations the equations of crystal lattice vi-
n = 2, 3, . . . . With intersecting energy levels brations are equivalent to the dynamic equations
forbidden quantum transitions become compara- of motion of elasticity theory, where the elastic
ble to the allowed ones in their intensity. The moduli are expressed with the help of the force
position, width and intensity of lines of forbidden constants. For instance, in a monatomic crystal
quantum transitions can provide additional in- 1 ik
formation not available from allowed transitions, ciklm = − α (n)R l (n)R m (n),
2V0 n
e.g., details about low-symmetry spin Hamiltonian
parameters. Quantum transitions between some where V0 is the unit cell volume.
482 FORCE, IMAGE
Infrared spectrum of Ho3+ doped LiYF4 (a), and enlargement of the low-frequency region (b).
A+A→0 nA ∼ t −1 , Ds > 2, –
Annihilation of triplet excitons nA ∼ t −Ds /2 , Ds < 2
crystals, with the concentration of one component model of plastic deformation). Macroscopic ob-
close to the percolation threshold, etc.). The time stacles (boundaries of composite layers, pliable
dependence of concentrations (nA , nB ) of excita- insertions, welded joints, riveted stringers, slots,
tions and reagents for different types of reactions local thermal and elastic fields) as well as mi-
are given in Table 1. croscopic ones (interblock boundaries and grain
In the case of an infinite percolation cluster the boundaries, triple grain junctions, slip bands and
Alexander–Orbach relation, Ds ≈ 4/3, is satisfied faults, second phase inclusions, etc.) may serve
for a space of any dimensionality (S. Alexander, as barriers. At the beginning of the spreading of a
R. Orbach, 1982). See also Percolation theory. crack the external stress should break the bonds at
its opening, with accounts for the work expended
FRACTURE in plastic deformation and overcoming barriers.
The surface damage resulting from the failure As a quantitative measure of fracture toughness
of a mechanical part, a semifinished item, or a three main parameters, fixed at the moment of
specimen. The study of fracture is one of the an- the crack movement under the load, are com-
cient methods of metal quality control performed monly used: (1) G1c , the critical velocity of energy
either visually or with the help of an optical mi- release, J/m2 , (2) K1c , critical stress intensity fac-
croscope; at the present time one makes use of a tor, MPa·m1/2 , (3) δc , critical crack opening, mm.
transmission electron microscope (replica method) For quasi-elastic (low plasticity) materials (see
or a scanning microscope. The type of the fracture Quasi-brittle failure), the values G1c , K1c and δc
depends on the variety of crystal lattice, the metal are interrelated as
structure, the testing (or operating) temperature, 2 (1 − ν 2 )
K1c
the loading rate, and the strain state. Using struc- G1c = σs δc = ,
E
tural indicators, a distinction can be made between
where σs is the yield limit of the material, ν is the
transcrystalline failure and intercrystalline failure;
Poisson ratio, and E is Young’s modulus. The para-
and using a fracture toughness criterion one distin-
meter K1c is determined experimentally by tensile
guishes between brittle failure, quasi-brittle failure
tests of special samples with an applied fatigue
and tough fractures. Some fractures may exhibit
crack (see Fatigue) and a fixed minimum load,
these signs in various combinations. In visual ob-
whereby the crack increases via flat deformation,
servation, the brittle fracture surface obtained by
i.e. by flat break-off under the effect of normal
spalling is shiny and bright; that of a ductile frac-
stresses. The factor K1c is almost independent
ture is dull gray. of the sample dimensions (for sufficiently large
dimensions) and at a given temperature and de-
FRACTURE BRANCHING formation speed it is a constant determined by the
See Branching of fractures. structure of the material. For construction steels
of average strength (σb = 400 to 600 MPa) typ-
FRACTURE, BRITTLE ical values are K1 ≈ 25 to 30 MPa·m1/2 (∼80
See Brittle fracture. to 100 kg·mm−3/2 ). Fracture toughness is not ap-
plicable for describing stress resistance involving
FRACTURE TOUGHNESS, fracture ductility, small cracks and defects of submillimeter dimen-
crack resistance sions (see Defect resistance).
The ability of a material to resist the spreading
of large macroscopic cracks within itself. Fracture FRAGMENTATION
toughness depends on the level of intermolecu- Process of formation of a fragmented struc-
lar bonding forces, on plastic deformation in the ture in crystalline materials, i.e. the generation
crack, and on barriers in the way of its spread- of small structural units which are called frag-
ing. The mechanism of plastic flow in front of the ments. There is no unambiguous definition of the
crack vertex involves dislocations at low levels, concept “fragment”. Originally the term referred
and rotations at high levels of plastic deforma- to a structural unit immediately below polycrystal
tion (see Plastic twisting deformation, Dislocation grain in the size scale. In this context, fragments
FRANCK–CONDON PRINCIPLE 487
of polycrystal grains and monocrystals result from elements occurring in nature. The only natural ra-
various processes: crystallization, recrystallization, dionuclide 223 Fr (produced at one of the stages
phase transition in the solid state, plastic deforma- of 227 Ac decay) has a half-life of approximately
tion under conditions of both creep and resistive 21.8 min. The surface layer of the Earth’s crust
load, including pressure treatment of metals. In all (1.6 km thick) contains only 24.5 g of Fr. The
cases, except for resistive plastic deformation, the electronic configuration of the outer shell is 7s 1 .
disorientation angles between the fragments are Atomic radius 0.280 nm; radius of the ion Fr+
θ ≈ 0.2–2◦ , and a block structure is often observed 0.180 nm. The work function is 1.8 eV, and the
inside the fragments. The sizes of fragments range electronegativity is 0.7.
from decimal fractions of a micrometer to several The following characteristics of francium are
millimeters (in monocrystals). deduced through extrapolation calculations. It is
Substructures with dislocation boundaries, expected that francium exhibits a body-centered
which arise under active plastic deformation, can cubic lattice; Tmelting ≈ 293 K, Tboiling ≈ 913 K,
be called cellular structures. Slightly (θ 2–4◦ ) heat of melting 2.1 kJ/mole, heat of sublimation
and highly (larger θ , up to tens of degrees) dis- 69 kJ/mole; specific heat 32 J·mole−1 ·K−1 ; re-
oriented cellular structures are distinguished. The sistivity 0.45 μ·m (at 290 K). 223 Fr is used in
dimension of cells ranges from 0.2 μm to sev- radiochemical determinations of the radioactive
eral micrometers. Sometimes the cells (particu- element actinium.
larly highly disoriented ones) are assumed to be
types of fragments. Two size ranges of fragments FRANCK–CONDON PRINCIPLE (J. Franck,
are distinguished: (1) dimensions ranging from 0.5 E. Condon, 1925)
to 1 μm, θ = 1–15◦ ; (2) dimensions ranging from A principle which governs the distribution
several micrometers to several tens of microme- of intensity in the optical transition band of an
ters, with disorientation angles of tens of degrees, electronic-vibrational system such as a molecule
and narrow “knife-edge” boundaries. In certain or impurity center of a crystal. The Franck–
cases (particularly at moderate temperatures) the Condon principle is an analogue of selection rules
fragmented structure (of the second size range) for coupled electronic-vibrational (i.e. vibronic)
has a streaky appearance. Fragments of this size transitions. However, this principle does not im-
often serve as boundaries for the texture compo- pose strict exclusions on the possibility of partic-
nent region of a deformation texture. In the context ular transitions, like selection rules do; rather it
of this approach, it is appropriate to use the term indicates which transitions are more likely. The
“cell” when they are slightly disoriented with wide classical Franck–Condon principle states that dur-
dislocation boundaries. These various structural ing the period of an electronic transition in a
scale levels may occur both simultaneously and molecule (or a crystal) the atomic nuclei do not
separately. The tendency toward fragmentation appreciably change their coordinates or momenta.
increases with the elevation of the strain temper- It is also assumed that the transition occurs at
ature, increase of the number of slip systems, the turning points (points of slowest speed) of
increase of the stacking fault energy. Fragmenta- the classical motion of the nuclei. The semiclas-
tion exerts a profound influence on mechanical sical Franck–Condon principle, though accepting
properties, and in a number of cases it results the claim that the coordinates and momenta of nu-
in an increase of the strength of a material with clei remain fixed during the course of the vibronic
a concurrent decrease of the temperature of the transition (which is inconsistent with the uncer-
brittle–ductile transition. tainty relation), allows for transitions at all values
of the vibrational coordinates R of the atomic nu-
FRANCIUM, Fr clei, with probabilities that are determined by the
Radioactive element of the Group I of the pe- quantum-mechanical distribution of R. A rigor-
riodic table with atomic number 87 and atomic ous calculation of the probability of a vibronic
weight 223.0197; belongs to alkali metals. Fran- transition (quantum-mechanical Franck–Condon
cium is the least stable among all the radioactive principle) involves the assumption that the ma-
488 FRANK DISLOCATION
D is the point of the dislocation pinning. Frank– π 4πKii 1/2
Ec = ,
Read sources have been repeatedly observed in d εa
crystals.
where d is the sample thickness, Kii are elastic
constants which assume different values depend-
FRANZ–KELDYSH EFFECT (W. Franz,
ing on the geometry of the experiment, χa and
L.V. Keldysh, 1958)
εa are the respective values of the diamagnetic
The shift of the intrinsic light absorption edge
and dielectric anisotropy of the liquid crystals (see
of a semiconductor toward lower frequencies,
Anisotropy of liquid crystals).
which takes place in the presence of an external
Typical values are Bc ∼ 0.1–0.5 T, Ec d =
applied electrical field. This effect was experimen-
Vc ∼ 1–5 V. In the case of an inclined orientation
tally discovered by V.S. Vavilov and I.K. Britsyn
of the director with respect to the field direction
in silicon. In the absence of an electric field, the
the Frederiks transition is a non-threshold type. An
frequency corresponding to the absorption edge is
analogous effect may be observed in smectic liquid
given by the formula ω = Eg /, where Eg is the
crystal C. In the case of cholesteric liquid crystals,
width of the band gap. In an electric field E the
in which χa > 0 and εa > 0 and the direction of
absorption edge becomes blurred because of elec-
the field is perpendicular to the spiral axis, the
tron tunneling (see Tunnel effect) from the valence
Frederiks transition involves uncoiling the spiral
band to the conduction band, and the absorption of
structure and a cholesteric–nematic transition.
light of frequencies ω < Eg / becomes possible.
The effective shift of the band edge is given by the
FREE ENERGY
formula
E ∼ (e2 E 2 2 /m∗ )1/2 , where m∗ is the
The quantity of energy available to perform
effective mass of the electron. Accompanying a
work, or the capability of a system to do useful
change of the absorption coefficient edge there is a
work. The Helmholtz free energy F for constant
change of the refractive index. The Franz–Keldysh
volume processes, F = U − T S, and the Gibbs
effect is used for the modulation of optical radia-
free energy G for constant pressure processes,
tion (see Modulation spectroscopy).
G = H − T S, are defined thermodynamically in
terms of the energy U , the enthalpy H , the temper-
FREDERIKS TRANSITION (V.K. Frederiks, 1927)
ature T , and the entropy S. See Thermodynamic
The change of orientation of the director n in
potentials, Gibbs free energy, Helmholtz free en-
liquid crystals, as a result of the action of mag-
ergy.
netic or electric fields. The physical cause of the
Frederiks transition is that the liquid crystals tend
FREE RADICALS
to reach the field configuration n of the lowest
Paramagnetic compounds with unpaired elec-
free energy. In the case of a nematic liquid crys-
trons at outer atomic or molecular orbitals, and
tal with finite and infinite values of energies of
which usually display high reactivity. Some free
adhesion of molecules to the supporting surfaces
radicals are stable (have long lifetimes), but most
the Frederiks transition is a threshold process. The
are unstable. Stable free radicals may be isolated
physical cause of the existence of a threshold lies
as chemically pure substances, and unstable free
in the fact that the change of the director orienta-
radicals can be obtained and studied in solid ma-
tion (the director orients along the external field
trices or in solutions. One of the principal methods
direction) is hindered by elastic deformations in
used to detect and study free radicals is elec-
the neighborhood of the supporting surfaces. As
tron paramagnetic resonance (EPR). Free radicals
the field increases, the undistorted texture n turns
having two or more uninteracting unpaired elec-
into a distorted one through a second-order phase
trons are called biradicals or polyradicals, respec-
transition. The critical field values for the Fred-
tively. If they are charged they are called radical
eriks transition for magnetic and electric fields are,
ions. Free radicals form from corresponding mole-
respectively:
cules under the effect of electromagnetic radiation,
π Kii 1/2 high-energy particle fluxes, at catalyst surfaces,
Bc = , during heating, etc. Free radicals are widely found
d χa
490 FREE SPIN COMPLEXES
in nature: at the Sun, in stars and comets, in inter- where u is the thermal motion velocity of atoms.
stellar space, etc. Respiration and photosynthesis From the above formula, there follows an expres-
as well as many individual processes, such as sion for the viscosity coefficient, η = AkB T /(aD),
combustion, polymerization, etc., involve free rad- where A is a constant of the order of unity. In
icals. this form the above formula coincides with the
empirical expression for the viscosity coefficient
FREE SPIN COMPLEXES known as the Vogel–Fulcher–Tamman formula:
The same as High-spin complexes. η = B exp[C/(T − T0 )], where B and C are pa-
rameters. Within the model of free volume, this
FREE VOLUME MODEL formula is satisfied by amorphous metals and
A theoretical model used to describe the phys- metallic alloys around the vitri ca tion temperature,
ical properties of liquids and solids in the amor- so that the model may be applied to metallic
phous state. The model is based on the assumption glasses.
that the amorphous state is completely determined Using the model of free volume, a phenomeno-
by the distribution of volume in it. The free vol- logical description has been suggested for plastic
ume vf per given atom is the difference between deformations, diffusion, and glass–liquid transi-
the volume of the cell vn occupied by the given tions in metallic glasses. The applicability of the
atom and the volume of that atom itself, va ∝ a 3 model to such glasses, however, has not yet been
(here a is the atomic diameter). The basic assump- proven.
tions underlying the model are as follows:
FRENKEL DEFECT
(a) free volume may redistribute itself between Crystal structure defect which arises when an
cells without any changes in either the total atom migrates from its regular lattice site to an in-
volume or the total free energy of the solid; terstitial one. Thus, Frenkel defects consist of two
(b) diffusion of atoms takes place along fluctua- components: the vacancy and the interstitial atom,
tionally developing cavities, their volume ex- therefore this defect is also called a Frenkel pair
ceeding a certain critical value v ∗ that is close (interstice–vacancy pair). In nonmetallic crystals
to va (these form during the course of free vol- the components of a Frenkel defect may exist in
ume redistribution); various charge states, e.g., in alkali-halide crystals
(c) the total free volume changes with changes in the neutral Frenkel pair consists of color centers of
temperature, and under the effect of tensions type F (halogen vacancy with an electron) and H
that produce plastic deformations. (interstitial halogen atom); a charged Frenkel pair
consists of color center α (halogen vacancy with-
For a liquid in thermodynamic equilibrium the
out an electron) and I (negatively charged halogen
distribution function p(vf ) of free volume was
ion at interstice). In crystals of complex com-
found to have the form:
position there can exist several types of Frenkel
γ vf defects which are associated with different sublat-
p(vf ) = exp − ,
v̄f v̄f tices; e.g., the binary crystal MX contains pairs
VM –IM and VX –IX (V are vacancies in sub-
where γ is a geometric factor of the order of unity,
lattices M and X, I are interstitial atoms of M
the average free volume v̄f may be represented in
and X). Frenkel defects can be generated at suf-
the form of a polynomial expansion over temper-
ficiently high temperatures as a result of thermal
ature T as v̄f = ξ(T − T0 ); here ξ is the thermal
uctuations. The probability that an atom escapes
expansion coefficient, T0 is a phenomenological
the potential barrier of height U , which sepa-
parameter. An expansion of the diffusion coeffi-
rates the site and the interstice, is of the order
cient D follows from the formula for p(vf ) and
of ν exp[−U/(kB T )], where ν is the oscillation
assumption (b) above:
frequency of the atom at the lattice site. Frenkel
1 γ v∗ defects are also produced at elastic collisions be-
D ≈ au exp − , tween crystal atoms and high-energy particles at
6 v̄f
FRENKEL–KONTOROVA MODEL 491
exposure of the crystal (e.g., alkali-halide) to nu- stresses (see Stress relaxation). The Frenkel ef-
clear radiations, and they can arise from excitation fect in multicomponent systems is accompanied
and subsequent annihilation of electron–hole pairs by swelling of the diffusion sample by an amount
or excitons. The energy U for the thermal gen- which equals the volume of the pores produced.
eration of a Frenkel pair (of the order of eV) is The swelling manifests itself clearly at sintering
often several times less than the energy released at of the mixture of powdered mutually insoluble
the impact generation of defects. This difference substances. This effect may be prevented by the
is due to different patterns of strain of the atomic application of low (∼ 102 atm) uniform compres-
environment associated with the slow thermal and sion (see Barodiffusional effect). In alkali-halide
rapid impact mechanisms of defect generation. crystals in an external electric field, the Frenkel
Separated and bound Frenkel defects are dis- effect shows itself in the form of vacancion break-
tinguished. A separated Frenkel pair consists of down.
isolated vacancies and interstitial atoms, which do
not interact because they are too far removed from FRENKEL EXCITON (Ya.I. Frenkel, 1931)
each other. The evolution of the bound Frenkel An exciton consists of a single excited atom
pair, the members of which are close to each with a finite lifetime in a lattice (or a single excited
other, may lead either to its annihilation or to its molecule in a molecular crystal). It was intro-
dissociation into isolated defects. The character duced to explain photoelectrically inactive light
of this process has a pronounced effect on the absorption. By virtue of its interaction with the
efficiency of the exposure of the crystal to irra- environment, the excitation may be transferred
diation with various energies of defect-producing (without energy loss) to neighboring (identical)
particles. Frenkel defects and more complicated atoms (or molecules). If the excitation is not ac-
objects, which include Frenkel defects and other companied by a lattice deformation, then it can
types of crystal defects, may have a profound ef- travel throughout the crystal, being characterized
fect on the crystal properties. See also Unstable by a certain wave vector k (or by momentum
Frenkel pair. p = k) and by the dispersion law ε(k) = p2 /2m∗
with the effective mass m∗ = 2 (∂ 2 ε/∂k 2 )−1 . In
FRENKEL EFFECT this case a Frenkel exciton is a free or traveling
Generation of pores (cavities) in the diffusion exciton. When there is an interaction with the en-
zone during the course of diffusion homogeniza- vironment then a deformation of a certain region
tion; the pores are generated because of the coag- of the lattice (which provides obstacles for migra-
ulation of surplus vacancies in systems where dif- tion of the excitation) may take place; in this case a
fusion occurs by a vacancy mechanism. The pres- self-localized exciton is formed, which may travel
ence and oversaturation of surplus vacancies are throughout the crystal carrying along this region
maintained by the inequality of counter diffusion of deformation. Therefore, a self-localized exciton
flows of the system components. In the simplest is characterized by a substantially larger effective
case of diffusion homogenization, the Frenkel ef- mass than the free exciton. There exists a possibil-
fect may be observed also in a single-component ity of deexcitation of the crystal, which involves
system, in which the directional flux of vacancies the transfer of the Frenkel exciton energy to pho-
is maintained by the structural nonuniformity of tons (luminescence) and phonons.
the sample. The strength or effectiveness of the
source of surplus vacancies, which are responsible FRENKEL–KONTOROVA MODEL (Ya.I. Frenkel,
for the generation of pores, is given by the equa- T.A. Kontorova, 1937)
tion gv = div Iv , where Iv is the flux of vacancies, A one-dimensional microscopic model of a
which is equal to the difference between the fluxes crowdion or dislocation in a crystal. An infinite
of the system components. The role of centers of chain of atoms linked together by elastic (spring)
pore nucleation is played by intrinsic defects of forces is considered to be located within the range
the structure (boundaries, cracks) and the defects, of an externally applied periodic potential. In
which arise at relaxation of the diffusion-induced the context of a linear (with respect to relative
492 FRENKEL PAIR
shifts) approximation the equation of atomic mo- mixing is a multiphoton process in the solid state,
tion takes the form with the same starting and ending energy con-
d2 un 2πun ditions. When looked upon as a macroscopic
2
= α(un+1 + un−1 − 2un ) − aω02 sin , phenomenon, the mixing of frequencies is charac-
dt a
terized by a nonlinear susceptibility of the second,
where un is the shift in position of nth atom, third, and higher order (see Nonlinear optics of
a is the equilibrium interatomic distance, ω0 is solids).
the frequency of uniform small-amplitude vibra- Because of the frequency dispersion of the re-
tions of the chain, and α is the elastic interatomic fractive index, the spatial distribution of the initial
interaction constant. The dislocation satisfies the nonlinear wave polarization in the medium does
following boundary conditions: not coincide with that of the electromagnetic wave
un = a, n → −∞, excited at the new “mixed” frequency. There-
fore, phenomena like interference are observed:
un = 0, n → +∞. the characteristic direction of the mixed frequency
radiation has a sharp minimum and maximum, and
Assuming that the length of the region of lo- the dependence of its power on the path length
calization of the inhomogeneous strain is much in the medium (in the direction of maximum in-
greater than a, we take the limit to the equation tensity) has an oscillating nature. The frequency
for the continuous function u(x): mixing processes of greatest practical importance
∂ 2u ∂2u 2πu(x) are the optical generation of sum and difference
2
= c2 2 − aω02 sin . frequencies ω3 = ω1 ± ω2 used for frequency con-
∂t ∂x a
version (e.g., for generating the second harmonic),
This sine-Gordon equation is widely used in one- and four-wave frequency mixing of the types of
dimensional nonlinear dynamics. The solution of ω4 = 2ω1 ± ω2 that is the basis for a nonlin-
this equation with prescribed boundary conditions ear spectroscopy method corresponding to active
corresponds to the solitary wave of the nonlinear spectroscopy of Raman scattering of light. See also
deformation field ∂u/∂x, and it is the simplest Optical parametric oscillator.
type of soliton. The dislocation described in this
continuum approximation may be displaced by an
infinitesimal force. There is an initial or threshold FRESNEL EQUATIONS (A.T. Fresnel, 1823)
stress which must be exceeded to start the disloca- Relations which connect the amplitudes,
tion motion (Peierls stress for dislocation in a one phases and polarizations of reflected and the re-
dimensional crystal; see Peierls–Nabarro model). fracted light waves. The reflected and refracted
waves arise at the passage of light through the
FRENKEL PAIR plane boundary between two transparent media,
See Unstable Frenkel pair. and they are described in terms of corresponding
parameters of the incident wave. Fresnel obtained
FREQUENCY MIXING in a solid these relations using the now discredited concept
A nonlinear process during which a pump of a vibrating elastic ether. The Fresnel equations
wave (electromagnetic) field containing two or follow from the electromagnetic theory of light,
more frequency components (e.g., ω1 and ω2 ) and may be obtained from the solution of Maxwell
is transformed in a nonlinear medium into waves equations with boundary conditions at the inter-
with frequencies that are linear combinations of face between two homogeneous isotropic media
the pump wave frequencies (e.g., ω3 = ω1 ± ω2 , with refractive indices n1 and n2 . For plane waves
ω4 = 2ω1 ± ω2 , . . .). The wave transformation the amplitudes of the electric vectors of the in-
takes place with conservation of the total electro- cident (i), reflected (r) and transmitted (t) waves
magnetic field energy, so the nonlinear medium are respectively resolved into components, which
does not absorb or radiate energy. From a are parallel (subscript p) or perpendicular (sub-
quantum-mechanical point of view the frequency script s) to the plane of incidence, and the Fresnel
FRICTION CONSTANT 493
equations take the form forces (friction forces) which counteract the rela-
tive body motion. In the case of the dependence
2 sin θt cos θi
Tp = Ip , of these forces on generalized velocities q̇, it is
sin(θi + θt ) cos(θi − θt )
convenient to consider a dissipation function Φ =
2 sin θt cos θi (1/2) ij aij q̇i q̇j , where qj are the generalized
Ts = Is ,
sin(θi + θt ) coordinates of the system, aij are the coefficients
tan(θi − θt ) (which, in general, depend on qi ) representing a
Rp = Ip , measure of the scattering of the total mechani-
tan(θi + θt )
cal energy per unit time, and E = −2Φ. In this
sin(θi − θt ) case, when writing the equations of motion of the
Rs = − Is ,
sin(θi + θt ) mechanical system it is necessary to take into ac-
count the terms proportional to q̇ which represent
where θi , θt are respectively the angles of reflec-
tion and refraction of light, which are related to the frictional forces. Friction in momentum space
the refractive indices n1 and n2 by Snell’s law: is the process of a particle decellerating (braking)
n1 sin θi = n2 sin θt . It follows from the above under the action of small forces uncorrelated in re-
equations, that the phase of every component of gard to either direction nor magnitude. Diffusion in
the reflected and transmitted wave is either equal momentum (P ) space (see Fokker–Planck equa-
to the phase of the corresponding component of tion) can be described by a kinetic-type expression
the incident wave or differs from it by π . There-
∂
fore, e.g., if the incident wave is linearly polarized, f = b· gradp P f + gradp (kB Tf ) ,
then the reflected and refracted waves are also lin- ∂t
early polarized. where f is the particle distribution function over
The applicability of the Fresnel equations is momenta, and the first term of the right-hand side
limited by the requirement that the irregularities of determines the force acting upon a particle which
the surface must be small compared to the wave- is directed opposite to its motion, and proportional
length, and refraction indices of the media must to its speed. It can be shown that the time av-
not depend on the amplitude of the E-vector of eraged momentum increment is
P = −bP τ ,
the light wave. In the case of very intense light where τ is the correlation time of the forces acting
sources (laser light) the reflection and refraction upon the particle, and b is the friction constant.
are accompanied by a number of new effects, and Such a consideration is used, e.g., when calculat-
the amplitudes of the waves at the boundary be-
ing the coefficient for the transfer and dissipation
tween the two media no longer satisfy the Fresnel
of the energy of channeled particles (see Friction
equations (see Nonlinear optics).
of channeled particles). By introducing the dissi-
pation term the kinetic equation can be regarded as
FRICTION
including dynamic friction forces. The energy dis-
A dissipative process arising as a result of the sipation produced by dipole moment oscillations
resistance to motion of a body or a system of bod- can be described as the action of a radiation fric-
ies, or a relative displacement of parts of the same
tion force.
system which leads to scattering and the loss of en-
ergy. The sliding motion of bodies in contact leads
to external friction; and this situation in respect to FRICTION CONSTANT
relative displacements of parts of the same system A coefficient in the Fokker–Planck equation
is referred to as internal friction. In a mechan- that determines the force acting on a moving
ical system friction causes the total mechanical particle, proportional to the particle velocity and
energy E to decrease with its partial transforma- directed oppositely to the direction of motion.
tion into other types of energy, including heat By considering the particle motion as a random
energy, energy accumulated in material defects, process, and introducing the total force F (t) act-
and so on. As a rule, one introduces dissipative ing on the particle at every instant of time t , it is
494 FRICTION, EXTERNAL
possible to determine the friction constant b: FRIEDEL OSCILLATIONS (J. Friedel, 1952)
Periodic space oscillations of the electron den-
t
1 sity ne , which occur with the period L = π/kF
b= F (t + t˜)F (t) dt˜,
mkB T (where kF is Fermi wave number), and are due
−τ to excessive electric charge in the bulk of a metal
where m is the particle mass, τ is the corre- or at its surface. The magnitude of Friedel os-
lation time of the forces acting on the particle, cillations has a power-law-type dependence on
and F (t + τ )F (t) is the correlation function of the distance r from the charge. The particular
the effective forces. Here the averaging . . . is form of the decay law depends on the nature (di-
performed over all the particles of the system ex- mensionality) of the electron configuration, and
cluding the probe particle under consideration, and on the shape of the Fermi surface. In the case
in accordance with the assumption that the ran- of a point charge in an isotropic metal with a
dom nature of the process does not depend on spherical Fermi surface, the electron density per-
the time, and determines the subsequent increment turbation decreases at large distances r
kF−1
of momentum
P . During the time τ ,
P = in inverse proportion to the cube of the distance
−bP τ , where P is the momentum of the particle. ∝ cos(2kF r)/r 3 . In anisotropic layered (see
The friction constant is similar to the coefficient Quasi-two-dimensional crystals) or chain (see
of mobility in the diffusion equation, and is asso- Quasi-one-dimensional crystals) metals with
ciated with the diffusion coef cient in momentum nearly cylindrical or slightly corrugated flat Fermi
space Dp = 2bmkB T . The friction constant b is surfaces the asymptotes of Friedel oscillations are
used to obtain expressions for the transport coeffi- of the form ∝ cos(2kF r)/r 2 in the plane of the lay-
cients in liquids and gases, in investigations of the ers or ∝ cos(2kF z)/z in the direction of the chains.
energy dissipation during the channeling of parti- The generation of Friedel oscillations is due to
cles (see Friction of channeled particles), etc. the diffraction of de Broglie waves of electrons
with the wavelength λD = 2π/kF (correspond-
FRICTION, EXTERNAL ing to conduction electrons in the neighborhood
See External friction. of the Fermi surface) by point defects (impu-
rity atoms, vacancies), and by extended defects in
FRICTION, INTERNAL crystals (dislocations, phase boundaries, twinning
See Amplitude-dependent internal friction, In- surfaces). Friedel oscillations play an important
ternal friction. role in the screened Coulomb interaction of atoms,
which are adsorbed by the metal surface, and in
FRICTION OF CHANNELED PARTICLES the indirect interaction of spins (see Ruderman–
A characteristic of the process whereby chan- Kittel–Kasuya–Yosida interaction).
neled particles dissipate energy by scattering off
electrons and thermally displaced atoms (see Ther- FRIEDEL SUM RULE
mal displacements of atoms) of the crystal. When A relation which permits the determination of
describing the energy change of the particle (in a the change in the density of electron states
N
direction normal to the forward motion,
E⊥ ) per in a solid due to impurities introduced into it. It is
unit channel length z, one can introduce an index derived in terms of a wave function which is a sum
of dynamic braking
E⊥ /z where the averaging over products l (r)Pl (cos θ) of radial functions
. . . is carried out over the entire region available and Legendre polynomials. The Friedel sum rule
for a particle with a given transverse energy E⊥ .
given by
Considering this value as the dynamic friction
coefficient, one can write down a kinetic equa- 2
N = (2l + 1)δl
tion (similar to Fokker–Planck equation) where the π
l
above-defined term is related to the change of
transverse energy due to braking (see Channeling, expresses
N in terms of the phase shifts δl of the
Neutral particle channeling). wave functions of electrons at the Fermi surface
FROZEN CONDUCTIVITY 495
which are scattered by impurities. Ordinarily the and spin σ , the phonon of momentum q be-
phase shifts are negligible for l greater than 3 or 4. j
longs to the j th branch of frequency ωq , V is
the normalizing volume, and Δ(q) takes into ac-
FRÖHLICH CONDUCTIVITY count the quasi-momentum conservation law. The
Contribution to the conductivity of a quasi- constant g in covalent crystals (insulators, semi-
one-dimensional semiconductor, which takes conductors) is proportional to the strain potential.
place at temperatures below that of the Peierls The Fröhlich constant in ionic crystals is equal to
transition, and is due to the motion of an elec-
tronic charge density wave (CDW) in the electric 2e 1 1
gF (q) = π − , (2)
field. The Fröhlich conductivity may be observed q ε∞ ε0
when the magnitude of the electric field exceeds
where ε0 and ε∞ are the static and high-frequency
a certain threshold value ET which is determined
dielectric constants, respectively. In the case of
by the pinning of the charge density wave at
metals, one should include in Eq. (1) the ma-
lattice defects, and by commensurability effects.
trix element of the electron–phonon interaction for
Typical values of ET for a number of quasi-one- j j
dimensional inorganic compounds (NbSe3 , TaS3 , gkk (q)[ωq /(2V )]1/2 :
K0.3 MoO3 ), in which the motion of a CDW is ob- 1/2
n
k |∇Vei |keq , (3)
j j
served, lie in the range between 0.01 and 1 V/cm. M (q) = −i
It is assumed that the phenomenon of Fröhlich
kk Mωq2
conductivity is related to certain unusual prop- where Vei is the pseudopotential of the electron–
erties of the semiconductors with charge density j
ion interaction, eq is the polarization vector of
waves: the sharp frequency dependence of the con- the j th phonon branch, q = k − k , M is the
ductivity in a weak alternating field at frequencies ion mass, n is the concentration of free electrons.
∼ 10–100 MHz; the nonlinear nature of static con- The Fröhlich Hamiltonian describes processes of
ductivity in fields with magnitudes that exceed ET single-phonon electron scattering, and the related
at T < TP (TP is the temperature of the Peierls effective attraction between electrons in second-
transition); the generation of narrow-band noise order perturbation theory.
at a frequency which is proportional to that of
the transport current carried by the charge density FROZEN CONDUCTIVITY, residual conductivity,
wave; the generation of Shapiro steps (see Joseph- stimulated conductivity, metastable conductivity
son effects) on the current–voltage characteristic A physical phenomenon in semiconductors,
in the regime of Fröhlich conductivity, etc. consisting in the persistence of an elevated value
of electrical conductivity for a long time after the
FRÖHLICH HAMILTONIAN (H. Fröhlich, 1950) external action (e.g., illumination) is terminated.
Hamiltonian of the electron–phonon interaction Frozen conduction is not a common property of
in the quasi-momentum space in the solid, which is semiconductors; it appears only after special tech-
written in the formalism of second quantization: niques of sample preparation. A fast relaxation
from the frozen conduction state to the equilibrium
gkk (q)ak+ ,σ ak,σ (bqj
+
j
HF = + b−qj ) state can be caused by increasing the temperature,
k,k ,q,j,σ by applying a strong electrical field, or by illumi-
nating with light having quanta of energy less than
j
ωq the width of the band gap (thermal, field, and IR
× Δ(q + k − k), (1) extinction of frozen conduction).
2V
There are several mechanisms of frozen con-
j ductivity. In inhomogeneous semiconductors (in-
where gkk (q) is the vertex part (triple pole) of the cluding, in particular, strongly compensated semi-
+ +
electron-phonon interaction, akσ (akσ ) and bqj conductors), the separation of nonequilibrium car-
(bqj ) are respectively creation and annihilation riers by the fields of inhomogeneities gives rise to
operators of the electron of quasi-momentum k a sharp increase of life-time, which is displayed
496 FRUSTRATION
via the long-term relaxation of conductivity, and FULLERENE, buckyball (named for architect
hence frozen conduction. Frozen conduction in- Buckminster Fuller, inventor of geodesic domes
duced by light can result from photochemical built from straight bars)
reactions in the bulk of the sample, or on its sur- Molecule of general formula Cn (n is an even
face. It can also be caused by optical ionization number) which has the shape of a closed hollow
of deep centers which strongly interact with the (in most cases convex) polyhedron; constitutes the
crystal lattice. The trapping of carriers at such cen- fourth allotropic modification of carbon.
ters requires lattice reconstruction, and proceeds The best known fullerene molecules are C70 ,
extremely slowly. and especially C60 . The molecule of the for-
Frozen conduction can be used in the systems mula C60 exhibits the highest symmetry among all
of information recording. The phenomena involved molecules known to the present day. This mole-
in frozen conduction can also be responsible for cule exists in the shape of a truncated icosahedron
degradation effects in photoelectronic devices. consisting of 20 hexagons and 12 regular pen-
tagons, resembling a soccer ball (see Fig.). Carbon
FRUSTRATION atoms in the fullerene molecule C60 are close to
The situation in a crystal, when conflicting sp2 -hybridized, and each of them forms one dou-
spatial requirements prevent complete atomic or ble bond C=C and two single bonds C–C. The
magnetic ordering from being established. This length of the single bonds (located along pentag-
can occur, e.g., when a regular lattice of antiferro- onal edges) is 0.146 nm, and the length of the
magnetically interacting spins cannot establish an double bonds (at boundary between two hexagons)
ordered arrangement in which each spin has only is 0.140 nm. The average bond length, 0.144 nm,
oppositely directed nearest neighbors. (see Long- is close to that of graphite (0.142 nm). These
range and short-range order). bonds are located at the distance of approximately
0.351 nm from the polyhedron center. Fullerene
FULDE–FERRELL–LARKIN–OVCHINNIKOV molecules are diamagnetic, i.e. their ground elec-
STATE (P. Fulde, R. Ferrell, A.I. Larkin, tron state is a singlet; the fullerene of the formula
Yu.A. Ovchinnikov, 1964) C84 is gyrotropic (see Gyrotropic medium).
Nonuniform superconducting state which The fullerene of formula C60 exhibits in its
arises in a uniform magnetic field B. The su- IR spectrum four bands at 527, 576, 1182 and
perconducting order parameter Φ in the Fulde–
Ferrell–Larkin–Ovchinnikov state varies periodi-
cally in space with a period which is close to
the coherence length ξ0 in the superconductor.
The phase transition in the presence of B from
the normal state to the Fulde–Ferrell–Larkin–
Ovchinnikov state at zero temperature T = 0 is
a second-order phase transition, taking place in
the field B = 0.76Δ0 /μB (μB is Bohr magneton,
Δ0 is the value of the energy gap Δ at T = 0 and
B = 0). When B reaches the Clogston–Chandra-
sekhar paramagnetic limit √ (A.M. Clogston,
B.S. Chandrasekhar) B =
0 /( 2μB ) (see Para-
magnetic limit) the Fulde–Ferrell–Larkin–Ovchin-
nikov state gives way to the uniform supercon-
ducting state through a rst-order phase transition.
This state can only be observed in pure (type I)
superconductors, because in the case of an impure
superconductor the periodic changes of Φ are sup-
pressed by the diffusional motion of electrons of Structure of C60 fullerene molecule. The alternating
mean free path l ξ0 . single–double bonds are not indicated.
FULLERENE 497
1428 cm−1 due to allowed electric dipole intra- metals. The metal atom in this complex is gener-
molecular vibrations. The 13 C NMR spectrum ally outside the fullerene molecule, but it can be
of this fullerene is a singlet (chemical shift inside it, or may be incorporated in the fullerene
δ = 142.68 ppm), demonstrating that all 60 carbon lattice. Chemical reactions, which result in the
atoms are equivalent. generation of fullerides, ordinarilly begin with
Fullerenes of all compositions (including hy- the elimination of one or several pairs of carbon
perfullerenes of formula Cn , where n > 70 and atoms. On the other hand, the generation of, e.g.,
up to ≈ 200) are formed in the gaseous phase at the fulleride of the formula C60 H36 takes place
the evaporation of graphite through laser or elec- through rearrangement of intrinsic π -bonds of the
tric arc heating, and exhibit unambiguous signals C60 molecule. The broad diversity of fullerides,
of highly stable ions C+ −
n or Cn in mass spectra especially metal-containing ones with the formula
(see Mass spectrometry). Fullerenes are soluble in [(C6 H5 )3 P]2 MC60 or {[(C2 H5 )3 P]2 M}6 C60 ,
aromatic, saturated and chlorinated hydrocarbons; where M = Pt, Ir, Pd, etc., has opened up new
they may be separated from solution in the form of trends in the physics and chemistry of organic
a solid phase, which is sometimes called fullerite. compounds.
The study of physical and physicochemical Fullerite is the solid phase of fullerene. The
properties of fullerenes is important for the un- most attention has been given to the study of ful-
derstanding of various processes, in particular lerites based on C60 molecules. Fullerites can be
catalysis, oxidation, combustion. Investigation of precipitated from solution, or generated through
molecules of the formula Cn is of much inter- condensation of graphite vapor with the subse-
est also from the point of view of cosmology and quent extraction with a hydrocarbon, chromato-
astronomy, because certain spectral bands, which graphic treatment, and crystallization. Fullerite is
had not been previously identified and are now a semitransparent crystalline material of dark-
ascribed to fullerene, have been spectroscopically yellow, brown or dark-brown color (color depends
observed in the atmospheres of carbon stars, and on C60 /C70 ratio in the composition). A C60 crys-
the tails of comets. tal has a cubic close-packed (FCC) structure with
Many unique opportunities for the application lattice constant 1.417 nm and density 1.72 g/cm3 .
of fullerenes are in sight; fullerenes may provide This substance is stable in air and sublimes at
the basis of new lubricants, molecular traps for 434 ◦ C.
atoms (including radioactive atoms), new medi- Solid fullerenes are molecular insulators with
cines, etc. the value of the band gap 2 eV; the interaction
The possibility for the existence of a chem- forces between the molecules are mostly van der
ically stable cluster with the formula C60 had Waals in nature (see Van der Waals forces) with a
been predicted by E. Osawa (1970) and, inde- slight intermixture of covalent bonding (see Cova-
pendently, by D.A. Bochwar and E.N. Galperin lent bond), which leads to weak anisotropy of the
(1973). This fullerene was first observed in mass intermolecular forces. Within the dipole approxi-
spectra in 1985 by H.W. Kroto and R.E. Smalley, mation, the lowest optical transitions are forbidden
who subsequently proposed the name “fullerene” in both fullerites and fullerenes.
for this new class of molecular carbon compounds. The fullerites of formula C60 or C70 may be
Fullerene was, for the first time, obtained in a easily turned into semiconductors (of both n- and
pure state through a synthesis carried out by p-types) and metals through doping. Both semi-
W. Krätschmer and D. Huffman (1990). conducting and metallic properties of fullerites are
Fullerides are chemical compounds which in- due to inclusion of atoms of alkali (Na, K, Rb, Cs)
clude the molecule fullerene as one of the con- or alkaline-earth (Sr, Ba) metals into the cavities
stituents. In particular, fullerides are found in the (octahedral, tetrahedral sites) of the face-centered
form of stable complexes of the formula C60 Mn lattice (or lattice of another type); outer shell (va-
(M = Na, K, Cs, Ca, Sr, Ba, La, U, etc.) in the lence) electrons of these atoms are transferred into
mass spectra of vapors obtained through evapo- lowest empty bands of fullerites; and these bands
ration of a mixture of graphite and salts of the become the conduction bands of the material.
498 FUNCTIONAL ELECTRONICS
At compositions close to stoichiometric, ion- “pockets” and packets in charge coupled devices,
radical salts of the general formula M3 C60 double Bloch lines, repolarization domains in fer-
(M = K, Rb, Cs) exhibit not only metallic, but roelectrics, etc. The application of dynamic inho-
also superconducting properties, which are unique mogeneities is possible not only in solids, but also
for three-dimensional organic compounds. In par- in liquid media (e.g., liquid crystals), in gels, and
ticular, the critical temperature of the doped triple in vacuum-gaseous media.
fullerite Rb2 CsC60 is Tc ≈ 31 K. There is a well- The devices of functional electronics may op-
defined correlation between Tc and the atomic erate due to a single particular physical effect and
radius of the doping metal: the critical temperature a single type of dynamical inhomogeneity, or they
increases with the increase of the doping alkali may involve several continuous media with the in-
metal atomic radius; this enhancement of Tc is tegration of different physical effects in a single
caused by the increase of spacing of the fullerite device, i.e. a system of media with various dy-
lattice, and the corresponding increase of the den- namical inhomogeneities which interact with each
sity of states at the Fermi level (which is in full other. Functional electronics, as a circuitry-less
conformity with the Bardeen–Cooper–Schrieffer branch of electronics; it is not in opposition to tra-
theory). However, the attempts to observe the ditional circuit engineering (e.g., IC) since they
phenomenon of superconductivity in doped ful- can exist in parallel and complement each other.
lerites based on C70 and other fullerenes have An advantage of functional electronics is that it
not yet been successful. Solid fullerenes resem- utilizes no intercellular connections, which im-
ble high-temperature superconductors in a number
pose strict limitations on the speed of response and
of superconducting properties (values of correla-
the reliability of integrated and large integrated
tion length, critical fields and currents, etc.), but
circuits with minimal dimensions 1 μm or smaller
are sharply different from them in crystal and
(nanotechnology). The possibility of carrying out
electronic structure. Solid fullerenes and fullerides
not only sequential input (output) and transport of
are considered as promising materials for various
data, but also “instantaneous” transport of large
applications: semiconductor and superconductor
data files, is another advantage of functional elec-
electronics, light-weight rechargeable batteries of
tronics. This may be considered as data processing
high capacity, transistors, heterojunction diodes,
at the level of elementary functions of higher or-
catalysis, hydrogen storage, etc.
der than those of IC and VLIC. This principle
FUNCTIONAL ELECTRONICS is a promising one for processing large data files
Trend in microelectronics, in which the role of in real time, e.g., when solving problems of pat-
the main data carrier is not played by the elec- tern recognition and identification. Examples of
tric state of a certain circuit cell (e.g., logic gate functional electronics devices are magnetic bubble
or memory storage register), as in the cases of domain devices, solid-state photoelectric convert-
integrated circuits (IC), and very large integrated ers with charge couplers (see Charge coupled de-
circuits (VLIC), but rather by a dynamic inho- vices), a variety of acousto-electronic devices (see
mogeneity, i.e. a local nonequilibrium state in a Acousto-electronics) using surface acoustic waves,
continuous homogeneous medium. Examples are electron–optical image converters (electrooptical
magnetic bubble domains, surface acoustic waves, modulators), liquid-crystal diodes (LCD) (see,
magnetostatic waves (see Spin waves), charge e.g., Liquid-crystal displays).
Gg
GADOLINIUM, Gd pressure); Young’s modulus is 56.19 GPa; en-
An element of Group III of the periodic ta- durance limit is 0.1824 GPa; yield limit is
ble belongs to lanthamides, atomic number 64, 0.182 GPa (at 293 K). Brinell hardness is 0.588
atomic mass 157.25; possesses 20 stable iso- GPa. Gadolinium is easily mechanically processi-
topes with mass numbers from 143 to 162. Nat- ble. Activation energy of self-diffusion of atoms
ural gadolinium is mainly a mixture of isotopes (at T boiling ) is 136 kJ/mole (i.e. 1.41 eV/atom).
with mass numbers 154 to 158 and 160, plus Natural gadolinium has very high effective ther-
weakly radioactive 152 Gd which has a half-life of mal neutron capture cross-section 45,000 barn
1.1·1014 years. Electronic configuration of outer (around 160,000 and 70,000 barn for 157 Gd and
electronic shells 4f 7 5d 1 6s 2 , successive ioniza- 155 Gd, respectively). Sommerfeld coefficient (of
tion energies 5.98, 12.1, 20.6 eV. Atomic radius low temperature electronic specific heat)
0.179 nm, Gd3+ ionic radius ≈0.0954 nm. Oxi- 19.5 mJ·mole−1 ·K−2 in the temperature range
dation state is +3, more rarely +2 and +1. Elec- 2.5 to 10 K. At 298 K electric resistivity of
tronegativity is ≈1.20. α-Gd monocrystal along the principal axis 6 is
In free form a silvery-gray metal. It exists in 1220 n·m, and in perpendicular directions it is
α and β modifications. α-Gd has hexagonal close 1390 n·m, temperature coefficient of electrical
packed lattice with parameters a = 0.3634 nm, resistance of gadolinium polycrystal is
c = 0.5781 nm (space group P 63 /mmc, D6h 4 )
0.00176 K−1 . Work function for polycrystal is
under normal conditions. Above 1535 K up to 3.07 eV.
T melting = 1585 K, α-Gd is stable. β-Gd with Gadolinium is the only rare-earth element
the body centered cubic lattice (a = 0.405 af- which transforms from the paramagnetic to the
ter hardening; space group I m3m, Oh9 ). Density ferromagnetic state directly at the Curie point
is 7.886 g/cm3 at 293 K. T boiling = 3553 K. T c = 293.2 K (second-order phase transition).
Binding energy is 4.14 eV at 0 K. The latent Gadolinium is a collinear ferromagnet below T c
heat of α → β transformation is 3.9 kJ/mole, la- and above Tc = 232 K. At low temperatures
tent heat of melting is ≈12 kJ/mole, latent heat the magnetic moments precess around a six-fold
of sublimation is 301.5 kJ/mole, latent heat of axis, forming a cone-like ferromagnetic struc-
evaporation is ≈330 kJ/mole. Specific heat is ture (phase transition is of “order–order” type).
≈0.20 kJ·kg−1 ·K−1 at room temperature. De- Magnetic susceptibility 75,500·10−6 cgs; nuclear
bye temperature is ≈180 K; linear thermal expan- magnetic moment of isotope 157 Gd (15.64% abun-
sion coefficient of monocrystal is 12.5·10−6 K−1 dant) is −0.34 nuclear magnetons. Gadolinium is
along the principal axis 6 and 6.3·10−6 K−1 in a prospective material for control rods (neutron
perpendicular directions (at 600 K); thermal con- absorbers) of nuclear reactors. Gadolinium is used
ductivity is ≈10 W·m−1 ·K−1 at room tempera- for the alloying of steel, titanium, and magnesium;
ture. Adiabatic elastic moduli of α-Gd monocrys- as a component of the alloys (with iron, nickel, and
tal: c11 = 66.7, c12 = 25.0, c13 = 21.3, c33 = cobalt), possessing high magnetic induction and
71.9, c44 = 20.7 GPa (at 298 K); adiabatic com- magnetostriction, for the luminophors and other
pressibility is 37.8 GPa (at 298 K), isothermal materials. Some strongly paramagnetic gadolin-
compressibility is 35.48 GPa (at zero external ium salts are used for cooling to extremely low
499
500 GALLIUM
temperatures, ∼0.001 K (see Adiabatic demagne- detectors. Mirrors made from gallium possess high
tization cooling). reflectivity. Gallium is added to special types of
glasses.
GALLIUM, Ga
A chemical element of Group III of the peri- GALLIUM ARSENIDE
odic table with atomic number 31 and atomic mass A binary semiconducting compound (see Semi-
69.723. The electronic configuration of the outer conductor materials) with the sphalerite (zinc
shells is 3d 10 4s 2 4p1 . Successive ionization ener- blende, ZnS) structure with each As atom centered
gies are 5.998, 20.514, 30.71 eV. Atomic radius in a regular tetrahedron formed by four Ga atoms,
of neutral Gd is 0.139 nm, radius of Ga3+ ion and vice versa.
is 0.062 nm. Work function is 4.19 eV. Oxidation
Thus each Ga atom is in the same position rela-
state is +3, more rarely +2 and +1. Electronega-
tive to its closest neighbors as As. The GaAs crys-
tivity is 1.65.
tal lattice can be thought of as two interpenetrat-
In free form, gallium is a silvery-white metal.
ing face-centered cubic sublattices built of Ga and
Under normal conditions the α-form is stable (α-
As atoms. The two sublattices are shifted along
Ga or GaI) with an orthorhombic crystal lattice
the cube diagonal relative to each other by 1/4 of
with parameters a = 0.4529 nm, b = 0.4519 nm,
its length. The space group is cubic F 43m (Td2 ),
c = 0.7657 nm. The peculiarity of this lattice
which has three 2-fold (π) symmetry axes, four
lies in the fact that the atoms are arranged as
3-fold (2π/3) symmetry axes and six symmetry
pairs, as diatomic molecules Ga2 . Under the pres-
planes, and in addition it lacks a center of symme-
sure 1.2 GPa and temperature 275.5 K GaI trans-
try. The distance between the positions of Ga and
forms to GaII (tetragonal lattice with parame-
As nearest neighbors is 2.44 Å, the sum of their
ters a = 0.396 nm, c = 0.437 nm); at the pres-
atomic radii Ga (1.26 Å) and As (1.18 Å). The lat-
sure 3 GPa and temperature 323 K the form
GaIII has been detected. For α-Ga T melting = tice constant is 5.653 Å, the density 5.32 g/cm3 ,
302.9 K, T boiling ≈ 2600 K; density of solid Mohs hardness is 4.5, specific heat under con-
stant pressure is 2.76 cal·mole−1 ·K−1 at 80 K,
gallium is 5.094 g/cm3 (at 303 K), density
heat conductivity is 0.37 cal·cm−1 ·s−1 ·K−1 . De-
of liquid gallium is 6.095 g/cm3 (at 303 K);
bye temperature is 362 K, static dielectric constant
latent heat of melting is 5.59 kJ/mole, latent
is 12.9, the optical (high-frequency) permittivity
heat of evaporation is 256 kJ/mole; specific heat
11.6, the melting temperature 1510 K, electron
cp = 25.9 J·mole−1 ·K−1 . Debye temperature is
affinity 4.07 V, work function 4.7 eV, tempera-
333 K; electrical resistivity is 0.272 μ·m (at
ture coefficient of linear expansion
L/L
T =
273 K); thermal coefficient of electrical resis-
5.9·10−6 deg−1 . The energy minima E(k) (k is
tivity is 8·10−6 K−1 . In the solid state gal-
quasi-wave vector) in the conduction band are lo-
lium is a diamagnet with magnetic susceptibility
cated along the crystallographic directions (100)
−0.31·10−9 CGS units (at 290 K). Hall constant
is −0.63·10−10 m3 /C. Transition temperature to (X-valleys), (111) (L-valleys) and near the cen-
the superconducting state is 1.085 K. Thermal co- ter point (000) (Γ ). The energy distance between
efficient of linear expansion is 2·10−5 K−1 (at 223 the Γ -minimum and the maximum of the va-
to 293 K), thermal conductivity is 41 W·m−1 ·K−1 lence band equals the width of the gap E g =
(at 273 K). Brinell hardness is 25 MPa. 1.43 eV at T = 300 K. The temperature depen-
Metallic gallium is used for manufacturing dence E g (T ) = E g (0) − αT 2 /(T + β), E g (0) =
high-temperature (873 to 1573 K) thermometers, 1.519 eV; α = 5.405·10−4 eV/K (100 < T <
manometers, thermoregulators, it substitutes for 500 K), β = 204 K. At T = 300 K, E g grows with
toxic mercury in diffusion pumps. Compounds of pressure P , dE/dP = 12.6·10−6 eV·kg−1 ·cm2 .
gallium with elements of Group V of the periodic Energy gaps between the Γ -points and the two
system (GaP, GaAs, GaSb, and some others) are other valleys are E
k = 0.28–0.36 eV, E
L ≈
semiconductors used in the high-temperature rec- 0.45 eV, respectively. The effective mass of elec-
tifiers, transistors, solar batteries, and IR radiation trons near the bottom of the conduction band is
GALVANOMAGNETIC EFFECTS 501
0.068m0 (m0 is the free electron mass). The de- their quasi-momenta P . The observation of Hall
pendence E(k) in the valence band has one dou- fields of opposite signs justified the concept of
bly degenerate maximum. Effective masses of the holes, excitations with a specific charge e/m∗ > 0
light and heavy holes are equal to 0.12m0 and (m∗ is the cyclotron effective mass). Current carri-
0.5m0 , respectively. The effective density of states ers with specific charges of opposite sign moving
at T = 300 K is 4.7·1017 cm−3 in the conduction in directions opposite to each other, deflect in the
band and 7·1018 cm−3 in the valence band; the in- same direction. Hence, in an intrinsic semiconduc-
trinsic concentration of electrons is 1.1·107 cm−3 . tor with an equal mobility of electrons and holes,
The mobility of electrons is 9000 cm2 ·V−1 ·s−1 , and in a compensated metal with the same num-
that of holes is 400 cm2 ·V−1 ·s−1 . The main ac- ber of electrons N1 and holes N2 , the Hall field is
ceptors are atoms of Group II of the periodic sys- very small or does not appear at all. The Hall field
tem that substitute for Ga and As atoms. The main can also be removed by the proper selection of the
donors are atoms of Group VI at As sites. Atoms sample geometry (so-called Corbino disk geome-
of Group IV can be donors, if they occupy the try).
places of Ga, or acceptors if they belong to the In experimental studies of galvanomagnetic ef-
As sublattice. The most important acceptors are fects, a current with a given magnitude and di-
Zn and Cd. At the melting temperature their sol- rection flows along a conductor, and the electric
ubilities are Zn 1020 cm−3 , Cd 1019 cm−3 . field components in three orthogonal directions
At low concentrations, Mg is a donor, and at high are measured. There are a transverse galvanomet-
concentrations (1018 cm−3 ) it is an acceptor. ric effect (j ⊥B) and a longitudinal effect (j B),
Selenium, tellurium and sulfur are shallow donors the latter being quite small. A dimensionless para-
(see Shallow levels). Oxygen creates deep donor meter determining the role of the magnetic field is
levels (see Deep levels in semiconductors). El- the ratio r c / l where r c is the radius of the current
ements of the transition group Cr, Mn, Fe, Co, carrier orbit (electron radius), and l is the mean
Ni are deep acceptors. GaAs is the most widely
free path. The dependence of ρ on r c / l at var-
used of all AIII BV compounds. Manufactured on
ious temperatures approximately scales to a sin-
its basis are transistors, diodes of ultra-high fre-
gle curve, and this is called the Kohler rule. The
quency range, tunnel, switching, emission diodes,
bending of current carrier trajectories in a weak
devices of the ultra-high frequency range with ap-
magnetic field (r c
l) leads, as a rule, to the ap-
plication of bulk effects, the devices with nega-
pearance of a small increase in resistance which
tive differential resistance, parametric amplifiers,
is quadratic in the magnetic field. For r c l the
heterostructures, functional integrated circuits, mi-
behavior of the resistance provides information
crowave varactors, injection lasers, infrared lasers,
on the carrier dispersion law. The main contribu-
solar converters.
tion to the electrical conductivity comes from elec-
GALVANOMAGNETIC EFFECTS trons with energy very close to the Fermi energy.
Manifestations of the action of a magnetic When moving in a magnetic field these electrons
field B on the conduction properties of a solid. travel along curves laying on the Fermi surface
These effects are due to the Lorentz force j × B ε(p) = εF in the plane with p · B = const; the tra-
which deflects the moving charge (current carri- jectories of the electrons in the coordinate space
ers) from their direction of motion. This leads to in the plane perpendicular to B are similar to the
a dependence of the resistivity ρ on B (magne- trajectories on the Fermi surface.
toresistance) and to the Hall effect, that is the ap- In metals with closed Fermi surfaces the mag-
pearance with an electric Hall field E H which is netic field confines the current carrier drift to the
orthogonal to both B and the electric current den- plane perpendicular to B 0 . The drift in this plane
sity J that compensates the action of the Lorentz is due to collisions. However, in non-compensated
force. The experimental study of galvanomagnetic metals (N1 = N2 ), the resistivity ρ(B) is the same
effects stimulated the development of the concept order of magnitude as ρ(0), growing and reaching
of quasi-particles in crystals, i.e. elementary exci- saturation at B → ∞. Even in very clean single
tations with energy ε that are periodic functions of crystals there is no pronounced anisotropy of the
502 GALVANOMAGNETIC EFFECTS
Gamma diffractometer.
504 GAMMA PHASE
Dependence of the absorption cross-section μa on the incident gamma ray energy for a number of target materials. The
right-hand figure is a semi-logarithmic plot covering a wide energy range, and the left-hand figure is a log–log plot for
low energies.
are due to the formation of electron–positron pairs of knocking out electrons from deep lying (inner)
near atomic nuclei, and this is possible beginning atomic electron shells. Nuclear reactions between
from Eγ 2mc2 (where m is the electron mass, high-energy gamma quanta and atoms in solids
c is the speed of light). The scattering of gamma can lead to a nuclear doping of materials.
quanta on atomic nuclei is much weaker than on
GANTMAKHER–KANER EFFECT
electrons. An interesting feature is resonance scat-
A radio-frequency size effect in a magnetic
tering of gamma quanta whose energy coincides
field B normal to a metal surface, caused by a pen-
with a difference between the ground state and any
etrating component of the electromagnetic field,
excited state of a nucleus. A process of resonance
the so-called Gantmakher–Kaner mode. Its exis-
scattering consists of absorption and subsequent
tence is connected with ballistic transport along the
emission of resonance gamma quanta. Due to the
constant magnetic field of the perturbation in the
recoil of the nucleus absorbing and emitting a pho- electron system induced by the applied radiofre-
ton, the latter is shifted in frequency toward lower quency wave. Associated with it is an oscillating
values. In solids, the recoil energy is transferred to dependence of the electric field E in the metal
the photon. However, in a Mössbauer effect, the on the coordinate zB due to the electrons with
recoil is absorbed by the solid as a whole, and the extremal displacement along B during a cy-
the frequency shift when absorbing or emitting a clotron period: u0 = (1/eB)(dS/dpz )ext (where
gamma quantum is insignificant. S is the area of the cross-section of the Fermi
The incoming gamma ray photons possess en- surface normal to B, and pz is the longitudinal
ergies sufficient for knocking out atoms from their (parallel to B) component of the electron quasi-
positions in the lattice, and collisions with them momentum). For z
u0 this dependence has the
can produce radiation defects. In addition, the de- form E(z) ∼ (δ/z)α exp(ik GK z) (where δ is the
fect nucleation events can be stimulated by Auger skin depth, k GK = 2π/u0 , α is an exponent de-
processes (see Auger effect) occurring as a result pending on the shape of the Fermi surface). The
506 GAPLESS SEMICONDUCTORS
presence of the preexponential factor introduces In zero-gap semiconductors of the second kind,
a difference between the coordinate dependence the gap vanishes because of the crystal lattice sym-
of a Gantmakher–Kaner mode and the coordi- metry. The so-called inverse zone scheme in these
nate dependence of the field of weakly-attenuating zero-gap semiconductors is possible if both the
electromagnetic waves (helicon, doppleron). This conduction band and the valence band belong to
preexponential factor is associated with a branch the same irreducible representation of the crystal
point singularity in the non-local conductivity as a lattice symmetry group. In all the zero-gap semi-
function of a complex wave number. Experimen- conductors of the second kind, the wave func-
tally the Gantmakher–Kaner effect is observed tions of the conduction band in the center of the
as an oscillatory dependence of the surface im- Brillouin zone are of p-type symmetry. Due to a
pedance of a metal plate on its thickness, and on strong spin–orbit interaction, the sixfold degener-
the magnitude of the external magnetic field. The acy inherent in a p-symmetry state (taking into ac-
oscillations induced by this effect exist over a wide count the spin) is lifted, and the latter splits into
range of sufficiently strong magnetic fields, with fourfold and doubly degenerate states in accor-
Ω
ν, where Ω is the cyclotron frequency and ν dance with the total angular momentum compo-
is the electron relaxation frequency). Their princi- sition law. The fourfold degenerate state in an or-
pal features are the following: (a) constant period dinary semiconductor gives rise to bands of light
in respect to the magnetic field; (b) small ampli- and heavy holes, while that in a zero-gap semi-
tude and its independence of the magnetic field in conductor of the second kind splits into the con-
the range of the anomalous skin-effect; (c) the ex- duction band and the valence band. The zero-gap
istence of oscillations for both senses of circular semiconductors of the second kind crystallize in a
polarizations of an incident electromagnetic wave sphalerite-type lattice; they include gray tin (α-Sn)
on the metal. The temperature dependence of the and the chalcogenides HgTe, HgSe and β-HgS.
amplitude of the Gantmakher–Kaner oscillations This zero-gap state disappears only under distur-
allows one to determine the temperature depen- bances that lower the crystal lattice symmetry. Hy-
dence of the frequency ν. drostatic pressure does not change the symmetry;
hence, the energy gap width for excitation of free
current carriers remains equal to zero in the zero-
GAPLESS SEMICONDUCTORS
gap semiconductors of the second kind. Both uni-
An intermediate class of substances between axial compression and an applied magnetic field
metals and semiconductors. induce the appearance of an energy gap.
They can be defined as metals with a point- The electron mass is less than the hole mass
wise Fermi surface or as semiconductors with a in all known zero-gap semiconductors of the sec-
zero band gap. There are two kinds of zero-gap ond kind. At finite temperatures, the electrons
semiconductors differing by the mechanism caus- and the holes are created; the Fermi level is sit-
ing the gap to disappear. The first kind includes uated in the conduction band, the electrons are
those in which the gap vanishes due to the ac- degenerate and the holes are nondegenerate. In
cidental coincidence of the energies of the edges this case, the Hall coefficient (see Galvanomag-
of the conduction band and the valence band. In netic effects) is a power function of the tempera-
such zero-gap semiconductors, the zero-gap con- ture. The electric properties of the zero-gap semi-
dition could be destroyed under the action of any conductors are extremely sensitive to the presence
disturbance including those that do not change the of impurities. The donors are ionized at any tem-
crystal symmetry. This kind of zero-gap semicon- perature, so there are free electrons in the zero-
ductors is exemplified by Bix Sb1−x , Pb1−x Snx Te gap semiconductors even at very low tempera-
and Pb1−x Snx Se alloys with certain ratios be- tures. There can also be resonance acceptor lev-
tween components x. The gap in these alloys van- els on the background of continuous spectrum of
ishes at four equivalent points L of the Brillouin the conduction band. The small level width is re-
zone, whereas the energy of electrons and holes lated to the small ratio of the electron and the hole
depends linearly on the wave vector. masses. In contrast to ordinary semiconductors,
GAPLESS SUPERCONDUCTORS 507
GAPLESS SUPERCONDUCTORS
Superconducors with a nonzero density of
Fig. 2. The density of states of a superconductor with
states N(ε) for single particle excitations at and
paramagnetic impurities. (1) n < 0.9nc (energy gap
in the vicinity of the Fermi level, as distinct
εg < Δ); (2) n = 0.9nc ; (3) 0.9nc < n < nc (zero-gap
from ordinary superconductors where the density state, εg = 0).
of states is zero within the the energy gap Δ
in the spectrum of quasi-particles (see Bardeen–
Cooper–Schrieffer theory), and has singularities at would be superconducting despite the fact that
ε = ±Δ (see Fig. 1). Gapless superconductors are density of states of the system as a whole has no
exemplified by superconducting alloys with para- gap (see Proximity effect). Superconducting alloys
magnetic impurities having uncompensated mag- with paramagnetic impurities are characterized by
netic moments. At high enough impurity concen- two relaxation times, one without a change of the
trations, N(ε) does not vanish at any energy (zero- electron spin direction (τ1 ) and the other with a
gap state) although it does dip in the vicinity change of the spin projection (τ2 ). Electron spin-
|ε − εF | ∼ kB Tc of the Fermi energy (Fig. 2). ip scattering occurs with probability of 1/τ s =
At the first glance, the absence of the gap con- 1/(2τ1 ) − 1/(2τ2 ), and corresponds to the decay
tradicts the Landau super uidity criterion for the of a Cooper pair and a lowering of the critical tem-
critical condensate velocity v c = Δ/pF (where perature T c . The value of T c is related to τ s by the
pF is the Fermi momentum). However, this cri- Abrikosov–Gorkov relationship
terion is applicable only to homogeneous sys-
Tc
tems with translational invariance. For example, ln = ψ(1/2) − ψ(1/2 + ρ),
in a bimetallic plate consisting of a normal metal T c0
and a superconductor the current flowing along where T c0 is the critical temperature of the pure
the superconducting layer parallel to the surface metal, ρ = /(2πτ s k B T c ), and ψ(x) is the Euler
508 GARNET FILMS
digamma function. The value of ρ as well as the occurs via a rst-order phase transition referred
scattering frequency τ s −1 are both proportional to to as Frenkel orientational melting. A gaseous
the paramagnetic impurity concentration. Accord- crystalline state manifests itself by singularities of
ing to this equation, T c vanishes at a critical value physical quantities (e.g., dielectric constant) or a
of ρ corresponding to a critical paramagnetic heat of phase transition in the vicinity of the ori-
impurity concentration nc . In the concentration entational melting point. This state is observed in
range 0.9nc < n < nc , the energy gap vanishes some molecular crystals.
(curve 3 in Fig. 2). At n < 0.9nc , the “normal”
gap is restored although N(ε) has no singularity GAS EXPANDER
(i.e. divergence of N(ε)) at the gap edge. In the A piston or turbine device designed to allow
range 0.9nc < n < nc , the dynamic characteris- an adiabatically expanding gas to cool itself and
tics of gapless semiconductors are similar to those perform external work. The action of a gas ex-
of normal metals; in particular, their specific heat pander is based on the adiabatic process. Gas ex-
has a linear temperature dependence. However, the panders are used in cryogenic technology, and are
electrodynamics of zero-gap superconductors re- employed for the liquefaction and separation of
tain the same salient features that are typical of su- gases (nitrogen, oxygen, helium, etc.).
perconductors with a gap (BCS theory). Thus, the
Meissner effect and the finite supercurrent are re- GAS, IDEAL
tained. The Josephson current in alloys with para- See Ideal gas.
magnetic impurities also remains finite in the zero-
gap range. Superconductivity with a zero gap is a GAS PHASE EPITAXY
feature which can occur in high-T c superconduc- An epitaxial deposition (see Epitaxy) on a
tors when the wave function symmetry is d-wave. monocrystalline substrate of layers of the same
(autoepitaxy) or different (heteroepitaxy) mater-
GARNET FILMS ial from a gaseous phase. Gas phase epitaxy rep-
See Epitaxial iron garnet lms. resents one of the most widely used methods of
forming semiconductor epitaxial lm s on conduct-
GARNET, IRON ing and insulating substrates during the manu-
See Iron garnets. facture of semiconductor devices and integrated
circuits. It has the following advantages: simplic-
GARNET, YTTRIUM–ALUMINUM ity and high productivity of the equipment; high
See Yttrium–aluminum garnet. deposition rates in comparison with vacuum va-
porization; relatively low deposition temperatures;
GASEOUS CRYSTALLINE STATE convenience for doping and control of stoichiom-
A state of matter of a solid in which the centers etry (see Stoichiometric composition); a possibil-
of molecules form a regular crystal lattice, while ity of depositing layers with a complicated geom-
the molecular orientational axes are randomly dis- etry; a possibility to clean the surface by chem-
tributed in space with spherical or axial symme- ical etching directly before the layer deposition.
try. In parallel with liquid crystals, a gaseous crys- The main disadvantages of the method are the in-
talline state is an example of a mesomorphous volvement of some toxic, explosive or corrosive
phase (see Mesophases), in which reorientations gaseous products, the difficulty of producing mul-
of molecules can take place without a noticeable tilayer structures with layers of controlled compo-
change in the packing density. A special character- sition, and the uncertainty about precisely preserv-
istic of a gaseous crystalline state is the presence ing the superlattice period.
of scales of order, long-range order (see Long- In the design of apparatus for gas phase epi-
range and short-range order) in the arrangement taxy, one can distinguish a closed process and an
of the centers of the molecules, and the absence of open process. In the closed process a substrate, a
order in their relative orientations. A transition be- source of a stream of the transport material (e.g.,
tween a gaseous and an ordinary crystalline state iodine), and a source of the depositing material
GAUGE INVARIANCE 509
are positioned in different temperature zones of to the charge transformations of the unitary U (1)
a break-sealed or dismountable ampoule filled, as group:
a rule, with hydrogen or inert gas. The tempera-
ϕ → ϕ = eieθ(x) ϕ, (2)
tures of the source and the substrate are chosen
to facilitate the appropriate movement of products where e is the charge, and θ is a parameter. This
of disproportionation or decomposition reactions. transformation is referred to as a global one if
For example, the reaction GeI4 ⇔ Ge + 2I2 near θ = const, and a local one if θ depends on x. If one
the source should take place in such a manner that replaces ordinary derivatives ∂μ with covariant
there is a continuous transfer of germanium from ones Dμ (where Dμ = ∂μ + ieAμ , and Aμ is re-
the source to the substrate. In an open process, the ferred to as the gauge field) in a Lagrangian L that
substrate is washed by the transport gas flux (as a is invariant relative to the global transformation,
rule, this is hydrogen) with additives of halogens, i.e. L(ϕ, ∂μ ϕ) = L(ϕ , ∂μ ϕ ) for θ = const (in
arsenes, or phosphenes of the depositing substance this case the charge conservation law is fulfilled),
which are decomposed or reduced by hydrogen on one obtains a new Lagrangian LG = LG (ϕ, Dμ ϕ)
the substrate surface. Disproportionation reactions which is invariant with respect to the local
are also used in this process in the presence of transformation ϕ → ϕ provided the gauge
a source. Doping of the deposited layers is per- field Aμ transforms according to Eq. (1), i.e.
formed by a doping source which adds gaseous LG (ϕ, (∂μ + ieAμ )ϕ) = LG (ϕ , (∂μ + ieAμ )ϕ ).
compounds of the doping impurity in the transport Transformations (1) and (2) are referred to as local
gas flux (in the open process). An accompanying gauge transformations, and the invariance of LG
(usually undesirable) process is autodoping, i.e. a relative to this transformation is called local gauge
diffusion transfer of impurities from the substrate invariance. The requirement of local gauge invari-
to the growing epitaxial film. ance leads to the appearance of new fields Aμ and
the construction of a new Lagrangian (to be sup-
GAUGE INVARIANCE plemented with a Lagrangian of the gauge fields).
The invariance of field equations (of observ- From the physical point of view, the gauge field
able quantities) with respect to symmetry trans- is the carrier of the interaction of charged fields.
formations unrelated to the properties of space– In the case of electrodynamics the gauge field is a
time (so-called internal symmetries). An exam- field of photons. In the case at hand, the symmetry
ple of gauge invariance is electrodynamics where group U (1) (see Group theory) of the Lagrangian
the electromagnetic field Fμν = ∂μ Aν − ∂ν Aμ re- is Abelian. The scheme of constructing a theory
mains unchanged under the transformation with a non-Abelian symmetry group remains un-
Aμ → Aμ = Aμ − ∂μ θ, (1) changed. The gauge theories of fields with a given
symmetry group (see Continuous symmetry trans-
where θ is an arbitrary function of coordinates and formation groups) are built from a Lagrangian that
time x μ = {x 0 = t, r}, ∂μ ≡ ∂/∂x μ , is invariant relative to the global transformations
of the given group (this yields conservation laws
Aμ = {A0 = Ψ, A},
as required from the physical point of view).
Ψ and A are the scalar and vector potentials (here- Abelian gauge theory is in full correspondence
inafter, the system of units with c = = 1 is used). with the Ginzburg–Landau theory of superconduc-
The above transformation, referred to as a gauge tivity (GL theory) where the order parameter can
transformation, acquires fundamental importance be treated as a complex-valued scalar Higgs field,
for a theory in the presence of charged particles. In and the vector-potential as an Abelian gauge field
the electrodynamics of zero-spin charged particles in the Abelian Higgs model (see Higgs mecha-
(see Spin) described by the complex-valued scalar nism). The GL theory was the first gauge theory
field ϕ(x) (the field ϕ can be considered also as of a field with a spontaneous symmetry breaking.
the order parameter in a superconductor), the La- Abelian and non-Abelian gauge fields appear also
grangian L = L(ϕ, ∂μ ϕ) for the free field ϕ dis- in the theory of condensed matter, such as liquid
plays certain properties of symmetry with respect crystals and spin glasses.
510 GAUSSIAN DISTRIBUTION
GAUSSIAN DISTRIBUTION, normal distribution changes its properties with time; aging and the loss
(K.F. Gauss) of the thixotropic property associated with phase
A Gaussian function layering take place. The removal of the disperse
medium leads to the formation of xerogels, i.e. dis-
1 x2
y(x) = exp − perse solids possessing an extended surface area
σ (2π)1/2 2σ 2 reaching 500 m2 /g and more. A particular fea-
normalized to unity that gives the frequency of oc- ture of xerogels is their porosity, with the pore
currence (probability of appearance) of a parame- size varying over a wide range. The value of the
ter x (σ characterizes the deviation of x from its specific surface and the xerogel porosity are to a
most probable value at x = 0). The Gaussian dis- great extent established at the stage of the sol-to-
tribution is used in solid state physics when ana- gel transition (drying temperature, nature of wash-
lyzing errors of measurements (as a normal law of ing liquid, and so on). Xerogels are employed as
error distribution), to describe the shape of a reso- adsorbents and heterogeneous catalysts.
nance line in spectroscopy (see Gaussian shape),
to account for the distribution of fields of defects GEMINALS (fr. Lat. gemini, twins)
in crystals, and for characterizing uctuations. Wave functions of two interacting electrons
ψ(r1 , r2 ) that cannot be reduced to a simple or an-
GAUSSIAN SHAPE, Gaussian tisymmetrized product of one-electron functions
A spectral line contour conforming to the ex- ϕ1 (r1 ), ϕ2 (r2 ). Geminals are used in the theory
pression I (x) = I m exp[−x 2 ln 2/δ 2 ] where I m of molecules and molecular crystals when greater
and δ are the amplitude maximum (for x = 0) and precision is required for the description of a many-
the line half-width at a half-height (I (x) = I m /2 electron system compared to the Hartree–Fock
method. In this case it is necessary to take into
for x = δ), respectively. When the line is recorded
in the form of the first or higher derivatives of ini- account the correlation motion of at least all the
electron pairs that provide a significant contribu-
tial curves, the Gaussian is described by the cor-
tion to a bonding energy of molecules and crystals.
responding derivative of this expression. In solid
In the simplest cases of a two-electron system (H2
state spectroscopy, the Gaussian shape accompa-
molecule and He atom), a successful selection of
nies line broadening mechanisms of the statis-
geminals allows one to calculate the energy with
tical type (e.g., the hyperfine structure envelope
an accuracy sufficient for comparison with spec-
for lines of electron paramagnetic resonance). As
troscopically determined values. For systems with
a rule, the Gaussian shape characterizes nuclear
the number of electrons 2N larger than 2 it is con-
magnetic resonance lines in solids.
venient to represent a wave function using an anti-
symmetrized product
GEL
A colloidal system consisting of a solid dis-
N
persed phase (see Disperse structure) suspended ψ(r1 , . . . r2N ) = ψ(r2i−1 , . . . , r2i ).
in a liquid (lyogel) or in a gas (aerogel). A gel i=1
emerges as a result of the processes of struc- In this way one can achieve a greater accuracy for
ture formation and sol coagulation. A sol-to-gel ψ than that obtained with a Slater determinant.
transition is not referred to as a phase transi- For even greater accuracy, one can use a linear
tion; under the action of external (mechanical or combination of such ψ(r1 , . . . , r2N ) with differ-
physicochemical) factors a reverse gel-to-sol tran- ent geminals. In this case, the method of geminals
sition (peptization) is possible, where a sol is more competes successfully with a many-con gurational
liquid-like and a gel is more solid-like. Gels pos- approximation. From a technical viewpoint, a cal-
sess mechanical properties specific for solids such culation with geminals is more complicated than
as plasticity and elasticity; and in addition gels one with orbitals, but it can be simplified by im-
exhibit a property of thixotropy, i.e. the ability posing the requirement of strong orthogonality on
to restore its spacial structure after a mechani- different geminals. This, however, results in sys-
cal fracture. Being a nonequilibrium system, a gel tematically overestimating the calculated energy.
GEMS 511
terminal members of this group are pyrope blue light, in UV rays by red, yellow, pale green
Mg3 Al2 Si3 O12 , almandine Fe3 Al2 Si3 O12 , spes- light. It is easily electrified by friction, compres-
sartine Mn3 Al2 Si3 O12 , grossular Ca3 Al2 Si3 O12 , sion, heating. Dielectric constant is 6 to 7; diamag-
andradite Ca3 Fe2 Si3 O12 and uvarovite Ca3 Cr2 - netic susceptibility is 0.42·10−6 cm3 /g. Mohs
Si3 O12 . Crystallographic system (syngony) is cu- hardness is 8, cleavage is perfect in one direction.
bic, hexoctahedral type of symmetry. Crystal lat- Zircon is a mineral ZrSiO4 . Composition: Zr
tice parameters: a = 1.1459 nm (pyrope), 49.5% (ZrO2 67%, SiO2 33%). The additions
1.1526 nm (almandine), 1.1621 nm (spessartine), are possible: Hf, Th, U, Ca, Al, Fe. Crystal-
1.1851 nm (grossular), 1.2048 nm (andradite), lographic system (syngony) is tetragonal, struc-
1.205 nm (uvarovite). The structure is formed by ture is insular with radical-ionic lattice of an-
isolated groups [SiO4 ] located along the helical ion groups SiO4 and cations Zr4+ surrounded
axis of the fourth order, which explains the rhom- by 8 oxygen ions. Parameters of the unit cell a =
bic dodecahedral and tetragonal-trioctahedral habit 0.6586 to 0.6622 nm, c = 0.594 to 0.6025 nm.
of crystals. Density and refractive indices are the Density is 4.7 g/cm3 , density of metamict zir-
following: pyrope, 3.71 g/cm3 , n = 1.705; alman- cons is 3.9 to 4.2 g/cm3 , compressibility coeffi-
dine, 4.32, n = 1.83; spessartine, 4.18, n = 1.80; cient is (47.6–48.3)·10−13 cm2 /dyn. Zircon may
grossular, 3.53, n = 1.735; andradite, 3.83, n = be transparent, translucent and nontransparent.
1.895; uvarovite, 3.78, n = 1.86. Upon melting to- Color: grayish-brown, yellow, barely red, blue,
gether the density drops and the refraction index black, pink, lilac, green, powder may be white
also becomes lower. Color: dark-red, red, pink, and yellow. The coloring varies depending on the
orange-yellow, black, green; there are no blue gar- effect of light, heating, X-rays. Dielectric con-
nets. Magnetic susceptibility of garnets is pro- stant is 3.6 to 5.2, that of the transparent ones
portional to the content of Fe, specific magnetic 8 to 12. Magnetic susceptibility is (−0.19 to
susceptibility is (60–160)·10−6 cm3 /g. Dielectric 0.90)·10−6 cm3 /g. Refraction index ng = 1.968
constant 4.25 to 4.51 (almandine), 4.70 to 4.87 to 2.015, nm = 1.923 to 1.960; birefringence
(pyrope), 5.22 (grossular), 6.40 (andradite). Mohs ng − nm = 0.045 to 0.058. Mohs hardness is 7
hardness is 6.5 to 7.5, it is the highest for alman- to 7.5, cleavage is imperfect. Transparent zircons
dine, pyrope and spessartine (7 to 7.5). Cleavage have a brilliant glow, the translucent and nontrans-
is imperfect. Only andradite is soluble in HCl, the parent ones have glass glow. Zircon luminesces
others only after melting. The transparent varieties under the UV radiation with yellow, orange, green
of garnets are jewels; garnets may be used as abra- light, in cathodoluminescence with green light; it
sive materials, artificial garnets are ferrimagnets. may exhibit thermoluminescence.
Topaz is a mineral Al2 [SiO4 ](F, OH)2 . Fe, Olivine is a mineral. For the continuous iso-
Cr, Mg, Ti, V, Ge and other elements occur as morphous series the end pont formulae are
additions. Crystallographic system (syngony) is (Mg, Fe)2 [SiO4 ], the end terms of the series are
rhombic, rhombic dipyramidal type of symme- forsterite Mg2 SiO4 and fayalite Fe2 SiO4 . The
try; crystal lattice structure is insular, it con- composition: Mg up to 34.3%, Fe up to 54.8%
sists of the columns of Al-octahedra connected (MgO up to 57.3%, FeO up to 70.5%, SiO2 up
by Si-tetrahedra; crystal lattice parameters: a = to 42.7%), additions of Mn, Ni, Co, Ca, Ti. Crys-
0.465 nm, b = 0.880 nm, c = 0.840 nm. Density tallographic system (syngony) is rhombic, rhom-
is 3.4 to 3.6 g/cm3 . Color: colorless, yellow, wine- bic pyramidal type of symmetry, parameters of the
yellow, green, bluish-green, blue, violet-blue, pink, unit cell: a = 0.4700 to 0.4815 nm, b = 1.0197
red-grayish brown. Chromophores (Cr, V, Fe, Ti) to 1.0431 nm, c = 0.5937 to 0.6068 nm, cell
and crystal lattice defects are responsible for col- parameters lengthen with increasing Fe content.
oring. Refraction indices: ng = 1.638 to 1.616; The structure is insular, with (Mg, Fe)O6 octa-
birefringence is 0.009 to 0.011. Colored topaz hedra forming zigzag-like chains along the c axis,
exhibits pleochroism: along ng it is violet-pink, which are held together by SiO4 tetrahedra. Den-
along nm yellowish-pink, along np brownish- sity is 3.2 to 4.4 g/cm3 . Color: olive- or bottle-
yellow. It luminesces in cathode radiation by red, green, yellow, grayish brown, more rarely violet;
GEMS 513
the powder is white. It is transparent to translu- UV rays it sometimes luminesces with weak or-
cent. Refractive indices: ng = 1.670 (forsterite), ange color. Upon heating the intensity of blue col-
nm = 1.651 to 1.869, np = 1.635 to 1.827. Dielec- oring increases. Mohs hardness is 5.5 to 6, cleav-
tric constant is 3.3 to 9.1 for different terms of the age of the crystals is perfect. Under the effect
isomorphous series and 6.8 to 7.1 for forsterite; of acids lazurite decomposes, releasing sulfurous
it is a paramagnetic material, with the magnetic gas.
susceptibility up to 18·10−6 cm3 /g (for the sam- Malachite is a mineral Cu2 [CO3 ](OH)2 . The
ple with contents of Mn 0.07% and of Fe 8.0%) composition Cu 57.4% (CuO 71.9%, CO2 19.9%,
depending linearly on the total contents of these H2 O 8.2%), additions of Fe, Ca. Crystallographic
elements. Mohs hardness is 6.5 to 7, cleavage is system (syngony) is monoclinic, prismatic sym-
average. Decomposes in HNO3 , forsterite gleams metry type, unit cell parameters: a = 0.948 nm,
in UV beams with crimson-colored light. Trans- b = 1.203 nm, c = 0.321 nm. Density is 3.9 to
parent, beautifully colored olivine (chrysolite) is a 4.1 g/cm3 . Color: bright-green, blackish-green;
jewel. streak is pale green; luster is from glass to di-
Quartz (silica) and its varieties may be sub- amond. Refraction indices: ng = 1.909, nm =
divided into the crystalline forms (rock crystal, 1.875, np = 1.655; birefringence is 0.254. Mag-
smoky quartz (morion), pink quartz, amethyst etc.) netic susceptibility (10–25)·10−6 cm3 /g. Upon
and also the hidden-crystalline chalcedonies (sard heating there appears endothermic effect of dehy-
(or carnelian), chrysoprase, agate, onyx etc.). dration and dissociation within the range 533 to
Amethyst is the violet variety of quartz. Color 773 K. Mohs hardness is 3.5 to 4, cleavage is av-
from red-violet to blue-lilac. The type of coloring erage. It decomposes in HCl with release of CO2 .
is caused by the structural addition of Fe: Si4+ It is a copper ore, and is used as a manufacturing
may be substituted by Fe3+ with charge compen- and decorative stone.
sation by Na ions. Radiation also influences the Nephrite is hidden-crystalline matted-fibrous
amethyst coloring. Coloring disappears upon heat- aggregate of amphibole of the tremolite-actinolite
ing up to 543 to 773 K, and is restored under the series, related to the prismatic type of symmetry,
effect of X-rays and gamma radiation. Upon heat- with general formula Ca2 (Mg, Fe)5 Si8 O22 (OH)2 .
ing above 793 K the coloring irreversibly disap- Color: white (tremolite-nephrite), green of dif-
pears. Intensely colored stones possess weak para- ferent tints – apple-green, grayish, light bluish,
magnetism unlike diamagnetic colorless quartz. grass-green, dark green; intensive green coloring
Chrysoprase is the most valuable variety of is caused by the presence of chromium (actinolite-
chalcedonies. The color is apple green. The col- nephrite). It is translucent in plates of thickness up
oring depends on the addition of nickel, which to 2 cm. Density is 2.9 to 3 (tremolite-nephrite)
contents varies from 0.3 to 3.3%. The color is, as and 3.1 to 3.3 g/cm3 (actinolite-nephrite), Mohs
a rule, nonuniform – from white to deep green. hardness is 5.5 to 6.5. It is distinguished by high
Green coloring may be explained by the color of strength and toughness.
gel. Rhodonite, or manganese spar is a mineral,
[There follow stones known as jewelery, artifi- silicate of Mn, (Mn, Ca)·SiO3 , additions of Fe,
cial or natural.] Mg, Al, Zn. Crystallographic system (syngony)
Lazurite is a mineral, aluminosilicate of Na is triclinic, pinacoidal symmetry type. Unit cell
and Ca, (Na, Ca)8 [AlSiO4 ]6 (SO4 , S)2 , additions parameters: a = 0.779 nm, b = 1.247 nm, c =
of Se, Cl, CO3 . The crystals belong to the cu- 0.675 nm. The crystal lattice structure is repre-
bic system (syngony), they are usually met in sented by simple chains of silicon–oxygen tetra-
the form of dense aggregates. Lattice parameter hedra with 5-fold period of identity. Density is
a = 0.905 nm. Density is 2.8 to 2.42 g/cm3 . 3.4 to 3.7 g/cm3 . Color: pink, crimson, scarlet;
Color: blue, violet-blue, light blue, greenish-blue, in the thin layer sometimes translucent. Refrac-
the powder is blue; the coloring depends on the tion indices: ng = 1.730 to 1.744; nm = 1.726
presence of sulfur, which in the brightly blue lazu- to 1.735; np = 1.721 to 1.728; birefringence is
rite is 0.5 to 0.7%. Refraction index n = 1.50. In 0.011 to 0.013. Distinct yellow- and pink-red
514 GEMS
pleochroism. Glows in cathode radiation with red Pearl is formed by the internal layer of shells
color; in UV radiation with red, pink, orange light. of bivalve and gastropod mollusks (mother of
Dielectric constant is 4.68 to 7.56, it increases pearls). It consists of very thin crystals of arag-
with increasing Ca contents up to 11.2; magnetic onite (orthorhombic form of calcium carbonate)
susceptibility is 53.4·10−6 (at 24% of Mn) and 82 to 96% and of an organic substance (protein)
87.7·10−6 cm3 /g (at 36% of Mn and 1% of Fe). conchiolin 10 to 14%. It has an iridescent luster.
Mohs hardness is 5.5 to 6.5, cleavage is perfect. Density is 2.6 to 2.8 g/cm3 , Mohs hardness is 3.5
Charoite is a mineral (Ca, Na, K, Sr, Ba)3 - to 4.5. Small crystals of aragonite are oriented per-
[Si4 O10 ](OH, F)·H2 O (it is around 56.5% of SiO2 , pendicular to the surface of the internal layer of the
1.01 to 1.85% of Al2 O3 , 20.5% of CaO, 8.2 to shell. The luster of pearl is explained by refraction
10.5% of K2 O, and also there are oxides of Ba, Sr, and reflection of beams in the prismatic layers of
Na (0.9 to 3.5%), fluorine and water). The isomor- aragonite. Color of the sea pearls is white, creamy,
phous additions: Pb, Mn, La, Zr. Unit cell parame- pink, blue, red, black.
ters: a = 3.182 nm, b = 0.713 nm, c = 2.210 nm, Corals are the external skeletons of marine in-
β = 94.15◦ . Crystallographic system (syngony) vertebrates (polyps) forming tree-like branching
is monoclinic, cleavage is perfect in three direc- constructions. In the composition of corals cal-
tions. Angle between the cleavage planes is 124◦ . cite with the addition of Mg carbonate prevails,
The color is lilac, of other tints up to violet; lus- sometimes there are oxides of Fe, Mn, etc. There
ter is glass-like. The spectrum of optical adsorp- are small quantities of organic material. The color
tion shows that the coloring of charoite is deter- of coral depends on the additions captured during
mined by the presence of Mn ions. Mohs hardness
the course of their growth, it may be white, pink,
is 5 to 5.5. Density is 2.54 to 2.68 g/cm3 . Mi-
flesh-pink, red, dark red, sometimes light blue and
croscopically in the passing light charoite is col-
black (containing almost 100% organic material).
orless along ng and pink along np ; refraction in-
Density from 2.7 to 1.32 g/cm3 (decreases with
dices: ng = 1.559, nm = 1.553, np = 1.550; bire-
content of organic material); Mohs hardness is 3
fringence is 0.009. It is not soluble in acids.
to 3.5. Refractive index is 1.65 to 1.49 for pink
Cordierite or dichroite, iolite, or “lynx sap-
coral, 1.56 for the black type. In UV rays coral lu-
phire” is a mineral, Mg alumosilicate; (Mg,Fe)2 -
minesces with violet or red light.
Al3 [AlSi5 O18 ]·nH2 O, additions of Mn, Ca, Ti,
Na. Crystals of rhombic crystal system (syngony), Amber is a mineral of the organic substance
rhombic dipyramidal type of symmetry; crystal class, fossil resin of coniferous trees, mainly of
lattice consists of the rings of silicon–oxygen the Paleogene period. Chemical composition: C
tetrahedra located in parallel to {0001}, thus form- 76 to 81%, H 10.0 to 10.5%; O 7.5 to 13.0%,
ing channels along the c axis where alkali cations N and S tenths of a percent. Amber is amor-
or H2 O molecules may be located. Crystal lat- phous, it is a volume polymer. Color: watery-
tice parameters: a = 1.708 nm, b = 0.710 nm, transparent (barely), milky-white, reddish-brown
c = 0.933 nm. Density 2.54 to 2.75 g/cm3 . Color: (oxidized amber), usually it is yellow. Depend-
grayish-blue, blue up to bluish-violet, greenish- ing on its turbidity the following types of amber
blue, colorless, yellow, grayish-brown. Refraction are differentiated: cloudy (semitransparent), hy-
index: ng = 1.538 to 1.568, nm = 1.532 to 1.560, brid (translucent in thin chips), bony and foamy
np = 1.527 to 1.558; birefringence is 0.009 to (nontransparent). Amber exhibits a specific IR
0.010, sometimes 0.003, rarely 0.012. Magnetic spectrum in the range of 700 to 1900 cm−1 , which
susceptibility of cordierite is proportional to its facilitates reliable identification. Mohs hardness is
content of Fe2+ : 6·10−6 cm3 /g at 2% of Fe2+ , 2.0 to 2.5. Density is 1.00 to 1.10 g/cm3 . It is eas-
∼10·10−6 at 4% of Fe2+ , and 2.4·10−6 at 8.5% ily machinable (with exception of the foamy type).
Fe2+ . Mohs hardness is 7 to 7.5, cleavage is per- It melts with decomposition at 573 to 613 K; with-
fect. out contact with air it softenes at 413 K, small
[We follow by listing stones of organic origin: pieces may then be pressed into large blocs of so-
amber, pearl, coral, agate, which are related to jew- called pressed amber; turbid varieties transform to
els.] the transparent state. Amber is a good insulator. It
GENERATION–RECOMBINATION NOISE 515
is formed upon petrifaction of resin as a result of duce new carriers. At ω > 2εg it is possible to
oxidative polycondensation of the saline acids and have η > 1 since a single photon can excite more
terpenes. It is used as a manufacturing stone, and than one electron.
also in the production of paint and varnish mate- For ω
εg (X-ray or gamma radiation), the
rial. Pressed amber is used for the fabrication of generation of current carriers consists of primary
electrical insulators. ionization events at which high energy (∼ω)
carriers appear, and multiple impact ionization
GENERATION OF CURRENT CARRIERS
processes which produce electron–hole pairs. In
this case 1 η < ω/εg . The latter inequality
Excitation of electrons to the conduction band
is connected with the requirement of momentum
and holes to the valence band where they func-
conservation in the elementary events of electron–
tion as current carriers. This generation of cur-
hole pair creation and interaction with lattice vi-
rent carriers is stimulated by the thermal motion of
brations. For ω
εg , an approximate formula
the crystal lattice atoms (thermal generation) and
η ≈ ω/(3εg ) is often used. The carrier gener-
also by some external factors: illumination (opti-
ation proceeds in a similar way when, instead
cal generation), particle beam irradiation, action
of photons, high-energy charged particles are in-
of strong fields, etc. A measure of this process is volved (electrons, protons, α-particles, etc.). At
the rate of generation, i.e. the number of carriers high light intensity (laser radiation) when multi-
produced per unit volume per unit time. Thermal quantum light absorption processes are operative,
carrier generation in an equilibrium semiconduc- the dependence of the rate of carrier generation on
tor is balanced by their recombination (see Re- the intensity becomes nonlinear (see Multiphoton
combination of current carriers), hence, the heat processes, Semiconductor lasers).
generation rate G is equal to the recombination Current carrier generation also occurs in the
rate: G = n0 /τ , where n0 is the equilibrium con- presence of a strong electric field as a result of im-
centration of carriers, and τ is the lifetime of the pact ionization and tunneling transitions of elec-
nonequilibrium current carriers. trons from impurities, and from the valence band
In the case of optical generation the nonequi- to the conduction band (so-called Zener break-
librium carrier concentration can exceed the equi- down).
librium value by many orders of magnitude. The
interband light absorption that occurs when the GENERATION–RECOMBINATION NOISE
quantum of light energy ω exceeds the band gap A type of noise in semiconductors caused by a
width εg results in the generation of electron–hole stochastic aspect of the generation and the recom-
pairs (Ge = Gh ), while impurity absorption leads bination of current carriers. If a variation of the
to the generation of electrons (Ge = 0, Gh = 0) number of carriers δN decays in time following
or holes (Ge = 0, Gh = 0). The rate of the optical an exponential law, the spectral density of fluctua-
generation of current carriers at ω > εg depends tions (see Fluctuations in solids) of the number of
on the light intensity. At low intensities, this de- carriers is
pendence is usually linear, described by the for- 2 τ
mula Sω = 2δN ,
1 + ω2 τ 2
2
G = ηαI0 exp(−αx), (1) where δN is the self-correlation function, and
τ is the carrier recombination time. In the low-
where I0 is the beam density of light quanta (num- frequency limit (ωτ 1), Sω grows with increas-
ber of quanta arriving per unit area unit per unit ing τ . This fact plays an important role in the
time), α is the light absorption coefficient, x is design of photosensitive detectors based on the
the depth of the penetration, η is the quantum photoconductivity effect. Since the low-frequency
yield (frequency dependent efficiency coefficient sensitivity of such devices increases with the
for producing charge carriers). At ω εg , η 1 growth of τ , it is necessary in this case to take into
because intraband light absorption does not pro- account the increase of the noise.
516 GENERATION–RECOMBINATION PARTICLE COAGULATION
positions of a strong interaction of the electrons with their symmetry axes along the 111 direc-
with the sound (k · V = 0, V is the electron tions. Effective mass of electron: longitudinal one
speed) over the distance of a sound wavelength. is 1.64m, transverse 0.082m (m is mass of free
An oscillation period of a geometrical resonance electron). Valence band energy maximum is in the
is
(1/B) = 2πe/(kP extr ); which depends on the center of the Brillouin zone at k = 0, the band is
extremal magnitude P extr between the points of threefold degenerate, effective mass of the heavy
noncollision absorption at a Fermi surface mea- holes is 0.28m, of the light holes is 0.044m, due
sured in a direction normal to the vectors k and B. to spin–orbit interaction the third band is split off
The measurement of
(1/B) at various orienta- to 0.28 eV, effective mass of holes in the split-off
tions of the vector B relative to crystallographic band is 0.077m.
directions allows one to determine the Fermi sur- The impurities of Groups III and V in ger-
face topology. manium are hydrogen-like acceptors and donors,
respectively. Shallow donor states are introduced
GERMANIUM, Ge also by lithium. The deep acceptors are zinc,
A chemical element of Group IV of the peri- beryllium, cadmium, mercury, copper, silver, gold.
odic system with atomic number 32 and atomic The deep donors are oxygen, sulfur, selenium, tel-
mass 72.59. Its contents in the Earth’s crust is lurium. Germanium is an amphoteric impurity in
7·10−4 % in the mixture of 5 stable isotopes: 70 Ge semiconductors of the A3 B5 type. Heat treatment
(20.55%), 72 Ge (27.35%), 73 Ge (7.78%), 74 Ge and the effect of nuclear radiations bring about
(36.5%), 76 Ge (7.86%). Outer electron shell con- the so-called electron–hole (n–p) conversion of
figuration is 4s 2 4p2 . Successive ionization ener- germanium, or changing the type of conductivity,
gies are 7.899, 15.934, 34.2, 45.1 eV. Atomic ra- caused by the binding of atoms of the Group V
dius is 0.139 nm; radius of ions Ge2+ is 0.065 nm, into the electrically inactive complex with non-
of Ge4+ is 0.044 nm. Oxidation state is generally equilibrium vacancies. By irradiation with elec-
+4, occasionally +2. trons with an energy of about 1 MeV at T < 65 K
In a free form, germanium is a solid with a metastable pairs may be created: vacancy + inter-
metallic shine. It exists in one amorphous and sev- stitial atom (unstable Frenkel pair). Upon irradia-
eral crystalline modifications. The crystal modifi- tion by the slow neutrons the most important re-
cation that is stable under normal conditions has action of transmutation occurs: 74 Ge(n, γ )–75 Ge
a cubic structure of the diamond type with the which disintegrates with a half-life 82 min to form
lattice parameter a = 0.56575 nm, space group 74 As. This process is the basis of neutron dop-
F m3m (Oh5 ). The density of solid germanium is ing of germanium. Monocrystals are obtained by
5.327 g/cm3 (at 298 K), of liquid germanium is the Czochralski method or by zone melting (see
5.557 g/cm3 (at 1273 K); T melting = 1211 K, Monocrystal growth) to produce germanium with
T boiling = 2973 K. Melting heat is 443 kJ/kg; the highest purity of all the elements with only
heat of evaporation 4700 kJ/kg; coefficient of 109 cm−3 of electrically active impurity atoms,
thermal conductivity is 60.7 W·m−1 ·K−1 ; coeffi- so germanium can be used for manufacturing nu-
cient of linear expansion is 4.5·10−6 K−1 (in the clear radiation detectors and also of photodiodes,
range 73 to 273 K). Mohs hardness is 6.25. IR radiation receivers (transparency range 1.5 to
Germanium is a typical semiconductor with the 20 μm), and avalanche diodes.
band gap 0.78 eV (at 273 K); temperature coef-
ficient of the band 3.9·10−4 eV/K; electrical re- GERMANIUM–SILICON ALLOYS
sistivity of extremely pure germanium 0.60 ·m; A continuous series of substitutional diamond
mobility of electrons 3800 cm2 · W−1 ·s−1 , structure solid solutions with unlimited mutual
of holes 1900 cm2 ·W−1 ·s−1 (at 298 K). Con- solubility of the components. Due to liquation
duction band has four equivalent energy minima and weak interdiffusion, the production of uniform
(valleys) located near the boundary of the Brillouin monocrystals in the range of concentration ratio
zone, the surfaces of constant energy as a func- 0.08 < Si/Ge < 0.95 is technologically compli-
tion of quasi-wave vector are ellipsoids of rotation cated. Monocrystals of germanium–silicon alloys
518 GERMAN SILVER
are characterized by a high density of structural active material (getter), a passive part (drain), or a
defects (dislocations) and non-uniformity of com- vacuum. The term gettering is taken from a vac-
position. The lattice parameter and the density of uum technology process which involves cleansing
the alloys depend linearly on the composition over a gaseous medium from undesirable components
its entire range. The band structure of these al- during either adsorption, absorption, or bonding
loys has been calculated theoretically, and also re- chemical reactions at the surface or in the bulk
constructed from experimental measurements of of disperse layers of chemically active substances.
optical adsorption and reflection spectra, as well The gettering process in solids includes three main
as from thermal conductivity, electrical, and other stages: (a) activation of particles involved in get-
properties. The dependence of the band gap width tering; (b) mass transport; (c) reaction or interac-
on the alloy composition is described by two linear tion with a getter accompanied by fixation or re-
regions with a point of inflection, corresponding to moving the particles from the sample. Depending
Si/Ge = 0.15. Variations in the alloy composition on the type of activation, gettering is divided into
bring about the restructuring of the energy spec- the thermal-, chemical-, and radiation-activated
trum of the conduction band and the valence band. types. The highest temperature process is the ther-
The functional dependences of the kinetic coef - mally activated one.
cients point to the non-uniformity of the alloys, A radiation-activating process is possible even
which is determined by the characteristics of their at cryogenic temperatures. With respect to mass
structure: by the difference of ionic radii and by transport gettering is divided into diffusion (in-
fluctuations of component distributions. cluding radiation-induced diffusion), drift (via elec-
Germanium–silicon alloys are used for man- tric field or mechanical force) and dynamic (with
ufacturing thermoresistors, thermooscillators and mass transport, relay-race method, acoustic wave
memory chips; they are favorable for the design of including impact wave, and dilation collective
receivers of IR radiation with the long-wave limit transport, see Dilatation). With respect to posi-
of sensitivity 14 μm, of tensometers, and of ther- tion we can differentiate: (a) bulk gettering (in-
mal neutron monochrometers. clusions of gettering phase embedded in the bulk,
e.g., in silicon these are SiO2 , SiC, Si3 N4 , rare-
GERMAN SILVER earth elements, metals, phosphorosilicates, and
Copper alloy with nickel (13.5 to 16.5%) and so on); (b) surface gettering (gettering layers
zinc (18 to 30%), sometimes with additions of lead Si3 C4 , SiO2 , SiO2−x Px , SiO2−x Pbx and oth-
(1.6 to 2.0%), and other metals. The density of ers deposited on the surface, mass transport is in
German silver is 8.7 g/cm3 , electrical resistivity the bulk); (c) planar gettering (mass transport at
is 26·10−8 ·m; thermal linear expansion coef- the surface or over the interface separating the
ficient at temperatures 293 to 373 K is equal to body and the contacting phase) (see the figure). In
16·10−6 K−1 ; coefficient of thermal conductivity view of facilitated migration at the surface, this is
at 293 K is 89.5 W·m−1 ·K−1 ; normal elastic mod-
ulus is 115 GPa; shear modulus is ≈45 GPa; limit
of linearity is 120 to 140 MPa in the annealed
state and 480 to 590 MPa in the strained state;
elasticity limit upon e xure σ0.005 = 560 MPa.
German silver is used in instrument making for
components of timepieces and fine mechanical de-
vices, the contact plates of automobile mechanical
relays, and the spring parts of electromechanical
television devices and computers. Model of a getter, where
σ is the gradient of the me-
chanical pressure, Jn is the current of the getter particles
GETTERING in solids in the region of flow, μG is the chemical potential of the
A process of extraction of impurities or struc- particles in the getter region, and μC is the chemical po-
tural defects from a solid through contact with an tential of the particles in the cleansing region.
GIANT QUANTUM OSCILLATIONS 519
characterized by a strong acceleration of the get- (Tb3 Ga5 O12 , TbLiF4 and Dy2 Ti2 O7 ) exhibit a
tering process (by several orders of magnitude) giant magnetostriction much greater than that of
and, therefore, will function at low temperature. nickel in the ferromagnetic state, which exem-
In terms of the bonding mechanism we can distin- plifies the phenomenon in paramagnets (
l/ l ∼
guish phase gettering (absorption due to increased 10−3 –10−4 ). Microscopic mechanisms underly-
solubility of an impurity in a getter) and struc- ing giant magnetostriction are the interaction of an
tural gettering (absorption due to defect structure anisotropic f -orbital electron cloud with the crys-
and the capability of absorbing point defects which tal eld, and the sharp dependence of the exchange
bond impurities). Gettering is widely used in mi- energy (in some crystallographic directions) on
croelectronics; making it possible to increase the interatomic distances. Another occasion for giant
lifetimes of minority current carriers and prevent magnetostriction is the presence of a phase transi-
the formation of defects under various kinds of tion in the lattice of a magnetic material, such as a
treatment. martensitic transition.
tion condition l Pl = 1:
El F
Z= Wl exp − = exp − ,
kBT kBT
l
a magnetic field B in the vicinity of critical tem- of the upper critical eld B c2 (T ) and the thermo-
perature T c takes the form: dynamic critical magnetic eld B c (T ), the pene-
β tration depth of magnetic eld λ(T ), the coher-
Fs = Fn + α(T )|ψ|2 + |ψ|4 ence length ξ(T ) = [2m∗ α(T c − T )]−1/2 , and
2
the Ginzburg–Landau parameter κ = λ(T )/ξ(T ).
1 2
∗ A)ψ 2 + B , This theory also accounts for the structure of
+ (−i∇ − e (1)
2m∗ 2μ0 Abrikosov vortices in the mixed (vortex) state, crit-
where Fn is the Helmholtz free energy in the ical currents and critical magnetic elds in thin
normal state, ψ is the complex coherent wave superconductor films and wires, the Little–Parks
function (order parameter) of the superconducting effect and the quantization of ux in multiply-
charge carriers called Cooper pairs with effective connected superconductors, the nucleation of in-
mass m∗ = 2m and charge e∗ = 2e (m and e are cipient superconductivity at the sample surface,
free electron values), and A is the vector-potential and so on.
of the magnetic field (B = ∇ × A). The coeffi- The behavior of a superconductor in an exter-
cients α and β obtained by Gorkov (1959) from nal magnetic field B depends essentially on the
the Bardeen–Cooper–Schrieffer theory are given value of Ginzburg–Landau parameter. In pure el-
by emental superconductors κ < 1 (with the excep-
tion of niobium), but in dirty superconductors (al-
2 T
α(T ) = −1 ; loys) with an electron free path length l ξ(0)
m∗ ξ02 Tc κ can be much more than
(2) √ unity. In type I super-
2 conductors with κ < 1/ 2, the surface energy
0.098 2 α s between the boundary of the superconduct-
β= ,
N(0) m∗ ξ 2 kB Tc ing and normal phase is positive. They undergo
0
a rst-order phase transition to the normal state
where ξ0 = vF /(πΔ0 ) is the coherence length or
when the applied magnetic field reaches the value
the size of a Cooper pair, Δ0 is the energy gap
of the thermodynamic critical magnetic field √ Bc.
in the spectrum at temperature T = 0, vF is the
Fermi velocity, N(0) is the density of states at the
In type II superconductors with κ > 1/ 2, the
Fermi surface, and the superelectron density nS = value α s < 0, and in the range of applied fields
2|ψ|2 . The condition of the free energy minimum B c1 < B app < B c2 the material is in the mixed
(1) provides the first Ginzburg–Landau equation: state. Upon reaching√ the lower critical field B c1
(B c1 ≈ B c ln κ/( 2κ) at κ
1), it becomes ener-
1
(−i∇ − e∗ A)2 ψ + α(T )ψ getically favorable to form filamentary inclusions
2m∗ of the normal phase called vortices (see Abrikosov
+ β|ψ|2 ψ = 0, (3) vortices) each of which carries a ux quantum
Φ0 = h/(2e). The state of minimal energy is a reg-
which is supplemented by the quantum-mechanical ular triangular lattice of vortices. Their number per
expression for the supercurrent density, called the unit area of cross-section ν = B/Φ0 increases as
second Ginzburg–Landau equation: the field√increases, and when its value approaches
ie∗ ∗ (e∗ )2 B c2 = 2κB c , the distance between adjacent vor-
JS = − ∗ (ψ ∇ψ − ψ∇ψ ∗ ) − ∗ |ψ|2 A tices becomes close to ξ(T ). In fields exceeding
2m m c (4) the upper critical field B c2 bulk superconductivity
and by Maxwell’s equations. disappears by way of a second-order phase transi-
The Ginzburg–Landau–Abrikosov–Gorkov tion, but it persists in a surface layer of thickness ξ
(GLAG) theory provides solutions of the non- up to the third critical eld B c3 (T ) = 1.69B c2 (T ).
linear equations (3) and (4) in an applied magnetic In the presence of an electric current density J ,
field which describe the macroscopic electromag- the Lorentz force J × B acts on the vortex fila-
netic and thermodynamic properties of supercon- ments. In an ideal homogeneous material, the vor-
ductors. In particular, this theory provides values tex motion under action of this force is hindered
GLASS CERAMICS 523
by viscous friction only, and a finite resistance ap- value of the energy gap, and r0 is close in or-
pears in the superconductor. In practice many vor- der of magnitude to the mean interatomic distance
tices are held or pinned at defects and structural (d ∼ 10−8 cm), so that Gi 10−12 , thus provid-
heterogeneities (see Vortex pinning), and this pin- ing a measure of the precision of the Ginzburg–
ning has a dominant influence on the critical cur- Landau theory of superconductivity. In contrast, in
rent of a type II superconductor. See also Meissner liquid helium (see Super uidity) in the vicinity of
effect and Gauge invariance. the phase transition point (lambda point) to the su-
perfluid state, Tλ = 2.17 K, Gi ∼ 1, and the self-
GINZBURG NUMBER, Ginzburg–Levanyuk consistent field theory is not applicable.
parameter (V.L. Ginzburg, A.P. Levanyuk, 1960)
A dimensionless parameter designated by Gi GLAG THEORY
(or τ G ) which characterizes the width τ = See Ginzburg–Landau theory of superconduc-
|T /T c − 1| of a critical region near the critical tivity.
temperature T c of a second-order phase transition
outside which the Landau theory of second-order GLASS
phase transitions is applicable. The Ginzburg See Vitreous state of matter, Dipole glass, Fiber
number Gi is connected with the coefficients of glass, High molecular weight glasses, Ideal glass,
the Ginzburg–Landau theory expansion, in the ab- Metallic glasses, Optical glass, Photochromic glass,
sence of a magnetic field (i.e. the vector potential Spin glass, Structural glasses.
A = 0), of the Helmholtz free energy F (h, T ) in
terms of the order parameter ψ GLASS CERAMICS, pyroceramics
Polycrystalline bodies of fine-grained struc-
1 ∂ψ 2
F = Fn + α(T )|ψ|2 + β|ψ|4 + g , ture, which are obtained by controlled directional
2 ∂h
crystallization of glasses of special composition.
where h = x/ξ0 , and ξ0 is the coherence length. The composition of the glass (see Vitreous state of
For the usual approximation, α = α0 (T − T c ), the matter) for the production of glass ceramics is cho-
Ginzburg number has the value sen in accordance with the phase diagram in such
a way that the end product of the crystallization
T c α0 β 2
Gi = τ G = , exhibits certain predetermined properties. In order
4g 3 to carry out uniform bulk crystallization and ob-
and Landau theory is applicable at τ > τ G . tain a microcrystallite structure (dimensions range
In the region τ < τ G the thermodynamics (and from fractions of a μm to 1 μm) the glass is often
dynamics) of the system is significantly affected doped with a crystallization catalyst (metals Au,
by critical uctuations of the order parameter. The Ag, Pt, Cu, oxides TiO2 , Cr2 O3 , P2 O5 , SnO2 ,
number Gi is determined by the nature of the in- fluorides CaF2 , Na2 SiF6 , etc.). The transforma-
teraction forces between particles (atoms, mole- tion of glasses into glass ceramics is performed
cules, electrons, their spins, and so on) causing the through a two-stage heat treatment. The tempera-
phase transition. For a particular material Gi is the ture of the first stage fits the optimal conditions for
order of magnitude of (r0 /r c )6 , where r0 is the the generation of crystal nuclei and their growth
particle interaction radius, and r c is the correla- up to critical sizes, whereas the temperature of
tion length of uctuations at a temperature far from the second stage is adequate for a maximal crys-
the phase transition point, when |T − T c | ∼ T c , tal growth rate (curing time determined by re-
i.e. when τ ∼ 1. A criterion for the applicabil- quirements imposed on the phase composition of
ity of the thermodynamic theory of phase transi- glass ceramics). Owing to their fine-grained struc-
tions (self-consistent eld theory) up to T ≈ T c ture, glass ceramics feature high strength char-
is the inequality Gi 1. For example, for su- acteristics, high thermal shock resistance, and
perconductors r c coincides with the coherence low thermal expansion, whereas their other spe-
length of the Bardeen–Cooper–Schrieffer theory cific properties depend mainly on the quantity
ξ0 = v F /(πΔ0 ) 10−6 cm, where Δ0 is the (60–90%) and type of the resulting crystalline
524 GLASS’ CONSTANT
phases, as well as on the composition of the resid- binders cement together separate filaments, and
ual glass phase. Glass-ceramic photomaterials, contribute to a uniform load distribution. Glass-
obtained by S.D. Stookey (1957), form a spe- reinforced plastics are divided into the three fol-
cial class of glass-ceramics produced from light- lowing groups: glass textolites (laminated sheet
sensitive glasses containing Au, Ag, Cu by ex- materials) obtained by hot pressing previously im-
posing the latter to the action of UV or ionizing pregnated glass fabric; oriented glass-reinforced
radiation followed by heat treatment. This causes plastics obtained by winding, drawing, or packing
the segregation of colloid particles of the metal, ber glasses and multiple glass fibers, combined
which catalyze the further crystallization. The se- with simultaneous impregnation of these materi-
lective irradiation of such glasses can provide pho- als with resin; fiberglasses obtained, as a rule, by
tographic images. Of high practical value are ash continuous impregnation followed by refinement
glass-ceramic materials, petrol glass-ceramic ma- of filaments and multiple fibers; glass-reinforced
terials, and slag glass-ceramic materials; these plastics based on linens and mats available from
types being obtained from industry wastes and pressing under low pressures; and soft rolled
rocks. glass-reinforced plastics produced by continu-
ous impregnation and heat treatment of various
GLASS’ CONSTANT (A.M. Glass, 1974) roll glass-fiber materials under standard pressure.
A parameter characterizing the magnitude of Glass-reinforced plastics find wide application in
a bulk photovoltaic effect. The photocurrent den- the building trade (roofing and facing materials,
sity J is proportional to the incident light inten- damp-proofing, sound insulators), in the chemical
sity I , and the proportionality coefficient for het-
industry (materials for pipes, containers, etc.), in
erogeneous absorption has the form kα:
electrical engineering (insulators, collectors, etc).
J = kαI, (1)
GLIDE-REFLECTION PLANE, mirror glide plane
where α is the light absorption coefficient inde-
Crystallographic symmetry operation combin-
pendent of I ; and the parameter k is called the
ing translation along a plane for a distance equal
Glass’ constant. Eq. (1) is for the isotropic case.
to the half of the lattice translation period t , and
Customary units are [J ] = A/cm2 , [I ] = W/cm2 ,
reflection across this plane (see Symmetry plane).
[α] = cm−1 , and [k] = A·cm·W−1 . The value of
The glide vector t/2 is sometimes called the glide
the Glass’ constant depends upon the mechanism
component. The designation of the glide symme-
of the bulk photovoltaic effect, and often involves
try plane depends on the glide direction: if the
both intrinsic and extrinsic defects. Therefore,
this value can depend strongly on the individual glide vector is parallel to the a, b, or c axis, then
properties of the sample, and change under var- the plane is designated respectively a, b, or c; if
ious kinds of influences, e.g., under the action the glide vector is directed along the body diag-
of high-energy irradiation, annealing, and so on. onal or the base diagonal and equals half of the
In addition, Glass’ constant depends on the fre- corresponding diagonal, then the glide-reflection
quency and polarization of the exciting light. In plane is designated n and referred to as the diag-
the crystal LiNbO3 , the value k in the vicinity onal glide plane or clino-plane; if the glide vec-
of the maximum of the absorption spectrum is tor is aligned with the base diagonal of the face-
close to 10−9 A·cm·W−1 in order of magnitude, centered lattice, or with the body diagonal of the
while in crystals of sillenite Bi12 SiO20 it reaches body-centered lattice and equals 1/4 of the corre-
∼10−8 A·cm·W−1 . sponding diagonal, then the plane is designated d
and named diamond plane.
GLASS-REINFORCED PLASTICS
Composite materials consisting of fiberglass GLOBAR
fillers (fibers, multiple fibers, tissues, etc.) and A radiator consisting of a rod made of com-
polymeric binders (polyester, epoxy, organic-sil- pressed refractory material (silicon carbide) which
icon, or other structure). Under operating condi- glows during the passage of an electric current,
tions the fillers carry the main load, while the and can be used as a source of infrared radiation.
GONIOMETER, X-RAY 525
A typical working temperature of a globar is close GOLDSTONE THEOREM (J. Goldstone, 1961)
to 1500 K; with the maximum of the radiation in- A statement concerning the excitations of a
tensity close to the wavelength of about 2 μm. system with spontaneous symmetry breaking. Ac-
cording to the Goldstone theorem, a system with
GOLD (Lat. Aurum), Au such a broken symmetry (see Continuous symme-
A chemical element of Group I of the peri- try transformation groups) possesses excitations
odic table with atomic number 79 and atomic mass whose frequency ω goes to zero as the wave vec-
196.967. Natural gold consists of one stable iso- tor k is reduced to zero. These zero-gap (massless)
tope 197 Au. Outer shell electronic configuration is excitations are called Goldstone modes (Gold-
4f 14 5d 10 6s 1 . Successive ionization energies are stone bosons). In the case of a Heisenberg mag-
9.266, 20.5, 30.5 eV. Electron affinity energy is net, the symmetry group of the Hamiltonian is
2.31 eV. Atomic radius of Au is 0.144 nm, radius the three-dimensional rotation group; the symme-
of Au+ ion 0.137 nm, of Au3+ ion 0.085 nm. Ox-
try group of the ground state (there is a contin-
idation state is +3, more rarely, +1 and +2. Elec-
uum of states with the same energy) is the group
tronegativity is ≈1.5.
of two-dimensional rotations about a certain ar-
Gold is a soft yellow metal. It has a face-
bitrary direction of the spontaneous magnetiza-
centered cubic crystal lattice with parameter
tion M. The symmetries relative to rotations about
a = 0.40783 nm and density 19.32 g/cm3
the other two axes (orthogonal to each other and
(at 293 K). T melting = 1340 K, T boiling =
2950 K. Latent heat of melting 12.5 kJ/mole, the axis M/|M|) are broken. The appearance of
heat of vaporization 349 J/mole, specific heat Goldstone excitations, i.e. small oscillations about
25.4 J·mole−1 ·K−1 . Coefficient of linear expan- the indicated axes, can be associated with the ten-
sion is 14.2·10−6 K−1 (at 273 to 373 K), thermal dency of the system to restore its symmetry by
conductivity is 311 W·m−1 ·K−1 . Electrical resis- undergoing transitions from one degenerate state
tivity is 2.25 μ·cm, thermal coefficient of resis- to another, which requires an infinitesimal energy
tance is 3.96·10−3 K−1 (at 273 to 373 K). Gold (hence, ω → 0 when k → 0). The dispersion law
is diamagnetic. Elastic modulus is 77 GN/m2 ; the of the Goldstone modes and their number are de-
tensile strength for annealed Au is 100–140 MPa. termined by the equations of motion in each par-
Brinell hardness is 177 MPa (for gold annealed ticular case. The Goldstone excitations in a ferro-
at 673 K), Mohs hardness is 2.5. Adiabatic elas- magnet are spin waves or magnons with a square
tic moduli of gold monocrystals are c11 = 192.9, dispersion law (ω ∝ k 2 ). In many non-relativistic
c12 = 163.8, c44 = 41.5 (GPa) at room tempera- systems the Goldstone theorem fails to hold. In
ture. particular, the presence of long-range forces (e.g.,
Gold is used in engineering in the form of al- the Coulomb forces in superconductors) results
loys with other metals, which improves its hard- in the appearance of a gap in the spectrum of
ness and strength. It is used in production of chem- Goldstone modes (e.g., a finite plasma frequency
ically stable equipment, in electrical engineering, ωp when k → 0; see Plasma oscillations). A gap
etc. Coating of surfaces with thin layers of gold also arises in the presence of interactions affect-
provides high corrosion resistance and high reflec- ing the degeneracy of the ground state. For exam-
tivity. ple, both a magnetic field and magnetic crystalline
anisotropy (see Magnetic anisotropy) lead to the
GOLDSTONE EXCITATIONS appearance of a gap in the magnon spectrum. A vi-
See Goldstone theorem and Spontaneous sym- olation of the Goldstone theorem is also possible
metry breaking. due to the Higgs mechanism.
GOLDSTONE MODES
See Goldstone theorem and Spontaneous sym- GONIOMETER, X-RAY
metry breaking. See X-ray goniometer.
526 GONIOMETRY
GRAIN-BOUNDARY EFFECTS
Properties of monocrystals and phenomena in-
volving them, which are due to grain bound-
aries. To some extent all properties specific to
polycrystals are associated with the presence of
grain boundaries. The recrystallization rate is con-
trolled by the mobility of boundaries μ = vm /F =
μ0 exp[−U/(kB T )], where vm is the migration
rate, and F is the motive force. Plots of the de-
pendences of μ and μ0 on the disorientation an-
gle θ exhibit maxima, and the plot of the activa-
Fig. 3. Grain-boundary defects. Twinning dislocation. tion energy U dependence on θ exhibits minima.
An elementary act of migration involves many
atoms, and one possible mechanism is the step-
easily). The shift by the vector b of the shifted
wise activated motion of a grain-boundary disloca-
grain boundary lattice is not a vector of the com-
tion (see Grain-boundary defects). Self-diffusion
plete superposition lattice; it disturbs the superpo-
and impurity diffusion of substitutional atoms (see
sition pattern and the atomic neighborhoods at the Diffusion) along grain boundaries occurs, as a
boundary. As a result, a grain-boundary stacking rule, according to a vacancy (see Vacancy) mech-
fault appears at the boundary. If the shift by b anism. The boundary diffusion coefficient Db =
covers only a part of the boundary area rather Db0 exp[−Ub /(kB T )] is greater than Dv , its vol-
than all of it, the boundary of the shift is a partial ume diffusion counterpart. Db for special bound-
grain-boundary dislocation. It separates the grain- aries is less than it is for ordinary ones. Since
boundary stacking fault and the defect-free bound- the potential relief of crystals in the boundary
ary, or it separates two such defects of different plane and along the normal to it is nonuniform,
packing types. Another linear grain-boundary de- the measured activation energy Ub is an effective
fect is the shelf type sketched in Fig. 2. The value and relates to the effective width δ of the
boundary passes along the shelf from one plane grain boundary, which equals several interatomic
of the grain-boundary shift lattice to a parallel spacings. Typically, what is measured experimen-
one. The shelf may be related to a grain-boundary tally is the product D b δ(U b ≈ (1/2)Uv ), where
dislocation, and the twinning dislocation sketched Uv is the activation energy of volume diffusion.
in Fig. 3 is an example (see Twinning of crys- Some types of impurity atoms are attracted to grain
tals). Since the defect-free lattice is a sublattice boundaries (horophilic impurities), while others
of the complete superposition lattice, the Burg- are repelled by them (horophobic impurities). The
ers vector of the lattice dislocation is a vector of incorporation of impurities in grain boundaries
the fully aligned lattice, and the lattice dislocation (intercrystallite internal adsorption) may cause
may be built into the boundary. Typically, such embrittlement of a polycrystal, and accelerate in-
a boundary-trapped lattice dislocation restructures tergrain corrosion. Facilitated sliding along grain
GRANATO–LÜCKE MODEL 529
boundaries with participation of grain-boundary constitute the complete superposition lattice are
dislocations increases the creep rate and plasticity shown within the boundary region. In actuality the
(see Superplasticity). boundary will tend to have the same density of
The restructuring of atomic configurations of atoms as the two individual crystal lattices, so it
grain boundaries by an alternating voltage re- will have a grain boundary lattice with many less
sults in internal friction Q−1 , characterized by atoms than those corresponding to the sites of the
a grain-boundary peak on the temperature func- complete superposition lattice. The atoms in the
tion Q−1 (T ). In semiconductor polycrystals (e.g., boundary region will arrange themselves in a con-
polysilicon), many grain boundaries contain dan- figuration of lowest energy which involves a ten-
gling bonds that are centers for recombination of dency to align themselves with the atoms outside
current carriers. the boundary.
Another view of a grain boundary arrangement
GRAIN BOUNDARY LATTICE with the same angle θ is shown in Fig. 1, see Grain
An auxiliary lattice for describing the location boundary.
of atoms in a bicrystal with a specific boundary
oriented in such a way that the surface density σ −1 GRANATO–LÜCKE MODEL (A. Granato,
of coincident sites (see Coincident-site lattice) in H. Lücke, 1956)
its plane (or parallel to it) is smaller than its max- A model of the dissipation of energy of elas-
imum possible value (σ −1 )max . Coincident sites tic vibrations (see Internal friction) which arises
are sites of the complete superposition lattice that from dislocation motion. The Granato–Lücke (GL)
are occupied by atoms of both individual bicrystal model is based on an analogy between dislocation
lattices. At σ −1 = (σ −1 )max the grain boundary motion and a string vibration, including the ten-
lattice coincides with the complete superposition sion, viscous and inertial forces. It assumes that
lattice so that all of the sites are coincident. a monocrystal contains a dislocation grid consist-
Fig. 4 shows a boundary between two simple ing of screw dislocations and edge dislocations.
cubic lattices Σ = 5 with the rotation axis [001], A dislocation line possesses two kinds of pinning
tan(θ/2) = 0.5, rotation angle θ = 53.13◦ . The centers, sturdy ones (nodes of the dislocation net-
sites of the atoms of the lattice of crystal No. 1 at work) and weak ones (impurity atoms and other
the bottom are designated by (◦), those of the other point defects). Correspondingly, two kinds of dis-
crystal No. 2 at the top are denoted by (×). The location segments, LS and LW , are considered;
boundary region is delimited by coincident sites, LS
LW . The GL model takes into account dif-
indicated by darkened circles, ABC at the bottom ferent length distributions of line segments LW for
and GF ED at the top. Sites of both lattices, which zero temperature. When the magnitudes of the ap-
plied stress σ are small (magnitude independent
losses), line segments LW vibrate between pin-
ning points (like in an elastic string). As long as
the motion is hindered by friction there is a phase
lag between the strain and the stress; hence energy
loss Δ appears as well as variations in the elas-
tic modulus. It follows from the GL model that for
amplitude-independent internal friction at low fre-
quencies (ω ∼ 1 kHz) we have Δ ∼ ΛL4W , where
Λ is the dislocation density (see Dislocation den-
sity tensor); while at high ω there is a peak in the
losses Δ(ω) of the resonance or relaxation type
which depends on the magnitude of the damping
constant. For further increases in σ the disloca-
tions tear off from the weak pinning centers, so
Fig. 4. Grain boundary lattice. their motion is limited by the strong pinning at
530 GRANITE
Table 1. Some properties of different classes of materials obtained on the basis of graphite
Density, g/cm3 1.6 to 1.7 1.72 to 1.98 2.15 to 2.23 1.45 to 1.52 1.4 to 1.45
Compression strength, MPa 33.0/33.0 98 295 to 345 225 410
Elastic modulus E, GPa 6.4/4.9 10.3 26.5 26.5 to 34.0 132
Coefficient of thermal expansion
α · 10−6 , K−1 (at 20 ◦ C) 3.7/4.1 6.6/6.0 10/0.55 – –
Thermal conductivity, W·m−1 ·K−1 (at 20 ◦ C) 120/103 94 70/227 – –
Typical crystallite dimension in
plane of basis La , nm 40 60 100 5 to 10 –
Note: The numerators provide values of properties in the direction parallel to the axis of compression, and the denomi-
nators denote those in the perpendicular (transverse) direction.
Natural graphite is in the form of sheets with- production, etc. For polycrystalline graphite the
out distinct crystalline shape, and their aggregates. anisotropy is insignificant.
It is transparent in the IR range. It is stable towards Besides conventional graphites, natural and ar-
heating in the absence of air, and it burns in air. tificial ones obtained from roasted and unroasted
It has 10 to 20% impurities (hydrogen, nitrogen, coke, there is also a whole series of carbon materi-
oxides and dioxides of carbon, methane, ammo- als based on graphite which differ from each other
nia, hydrogen sulfide, water). It is used as a lubri- in many of their properties. Quasi-monocrystals of
cant after chlorine purification at about 2000 ◦ C. graphite used as monochromators, possess sharply
Graphite polycrystalline materials are related to expressed anisotropy (disorientation of crystallite
semimetals with zero (or close to zero, depending packing does not exceed 1 to 5 ) and the char-
on impurities and submolecular structure) overlap- acteristic crystallite dimension in the basal plane
ping of the valence band and conduction band. is about 1 μm. At the same time on the basis of
The electric resistivity depends on porosity, grain the so-called nongraphitizing resins glass carbon
dimensions, crystallite dimensions and the ther- is obtained, which is a completely anisotropic gas-
mal treatment temperature, thus varying for differ- impermeable material, stable in corrosive media.
ent brands of graphites by factors of 10 or even Glass carbon has a globular structure with closed
100. In monocrystals it is highly anisotropic; in porosity ∼30%; and a characteristic crystallite di-
the plane of the layers it has a metallic character mension of about 5 to 10 nm.
(0.385 μ·m), and in the perpendicular direction Materials based on graphite are differentiated
(52.0 μ·m) graphite is a semiconductor. The re- by their supramolecular structure, porosity, crys-
sistivity of polycrystalline materials parallel to the tallite size and manner of packing. The wide spec-
basal plane (perpendicular to the axis of pressing) trum of properties of these materials is shown in
is several dozen times lower than along the axis. Table 1 for the cases of: material obtained on the
The strength parameters of graphite are also basis of roasted coke; material on the basis of un-
anisotropic. Thus, e.g., for pyrographite obtained roasted coke; pyrographite – obtained by deposi-
by deposition from the gaseous phase, the aver- tion from the gaseous phase; glass carbon – on the
age limiting compression strength σ comp parallel basis of phenol-formaldehyde resin; composition
to the deposition plane has a value 312 MPa, and materials of the carbon–carbon type, obtained on
perpendicular to this plane the value is 61 MPa. the basis of carbon fibers.
The strength parameters of anisotropic graphites The various properties of graphite determine
depend on the submolecular structure associated their areas of application. High electrical con-
with the treatment temperature, the method of ductivity, thermal stability in an oxygen-free at-
532 GRAPHITIZATION
mosphere and inertness with respect to the whole modifications. The hexagonal structure of graphite
series of corrosive media allow the use of graphite is related to the space group C6/mmc (D6h 4 ),
for manufacturing heaters, electrodes, crucibles with four atoms per unit cell, which is a prism
and cuvettes. The main consumers are the met- with the height 0.671 nm and a rhombus in the
allurgy, chemistry and semiconductor industries. base. The sides of the rhombus are 0.246 nm
Due to its low cross-section for thermal neutron with the angle 60◦ . The displacement of layers
capture (∼0.0045 barn) and its ease of machin- in relation to each other is 0.1418 nm, and every
ing, graphite is used as moderator of neutrons third layer repeats the first one. The rhombohedral
and a construction material for thermonuclear re- modification corresponds to the space group R3m
actors. Together with conventional graphite ma- (D3d5 ). The packing of layers is such that every
terials, carbon becomes more and more widely fourth layer repeats the first one. The rhombo-
used: carbon composites, carbon plastics, struc- hedral structure is usually found in natural graphite,
tural materials in rocket and aircraft industry, where its content may reach 30%; it almost never
sporting goods, in medicine for implantable non- appears in artificial graphite.
detachable prostheses, etc. Graphite high-modulus The forces of interaction of the basal planes are
fibers and graphite “whiskers” have values of the weak van der Waals forces. The binding energy
elasticity modulus ∼500 GPa. between the planes is 4.2 to 18.2 kJ/mole and the
The structure of graphite is a typical lamel- distance between them is 0.3354 nm. Due to the
lar type (see Quasi-two-dimensional crystals). The
low values of the bonding forces between the basal
layers form a series of parallel flat planes, which
planes the latter may be the main planes of dis-
consist of regular hexagons with carbon atoms
placement and crystallite stratification, and cleav-
at their vertices (see the figure). In the plane
ages are possible along them. The ease of displace-
of the hexagonal lattice (basal plane) the atoms
ment along the basal planes has for a long time
are bound together by strong covalent bonds
been incorrectly associated with the good antifric-
formed by overlapping of sp2 -hybrid orbitals
(see Hybridization of orbitals). The shortest dis- tion parameters of the material (L. Bragg hypoth-
tance between the atoms in the plane is a = esis) similar to those of a deck of playing cards.
0.1415 nm, and the energy of the covalent bonds is The lubricating ability of graphite is explained by
710 kJ/mole. The order of the location of planes the formation in the contact gap of cleaving layers,
along the c axis and the symmetry in this direc- adsorbed at the surface of molecules (see External
tion determine the space group. There are two friction).
During the production of artificial graphite, at
the stage of annealing, a material is obtained with
the turbostrate structure, characterized by the ab-
sence of ordering in the direction of the c axis. The
characteristic interplane distance for turbostrate
carbon is more than that for the ordered modifi-
cation, and has a value 0.34 nm. The amount of
turbostrate carbon that is produced decreases in
proportion to the increase of the treatment temper-
ature.
GRAPHITIZATION
The formation of graphite like material in
carbide-containing metal alloys (mainly based on
iron) whose carbides are unstable at atmospheric
pressure. Graphitization occurs at increasing tem-
perature, and is represented by nucleation and
growth of graphite nuclei in a metallic alloy host
Graphite. (due to the diffusion influx of carbon atoms of
GREEN’S FUNCTIONS 533
the dissolving carbide, and self-diffusion of metal particular physical properties assigned to the ver-
atoms from the graphite surface). The higher the tices and edges. The lattice gas model widely used
temperature, the higher the rate of graphite forma- in solid state physics assumes that the atoms and
tion. Graphitization is stimulated by preliminary molecules of the gas occupy positions that are not
quenching of alloys, deformation, irradiation by random, but rather are situated at crystal lattice
high-energy particles, and introduction of silicon, sites, each site having at most a single particle.
or more rarely, aluminum. The process is hindered This model is convenient for studying the ran-
by the elements which increase the sturdiness of domness of substitutions in disordered systems.
carbides (e.g., chrome and manganese). Some- Configuration integrals over particle coordinates
times the term graphitization is used for the for- transform into lattice sums which, in some sim-
mation of graphite in alloys which contain no car- ple cases, can be calculated exactly. Diagram rep-
bides. In such alloys graphite separates out from resentations of expansions of interaction integrals
the carbon-supersaturated solution during solidifi- and Green’s functions as sums lead to the neces-
cation and further cooling. In addition, variations sity of summing the corresponded series only over
in the structure of carbon-based materials are also connected diagrams, based on a fundamental theo-
called graphitization. rem of graph theory (diagrams). The resistance of
a random network of resistors can be treated from
GRAPHS in solid state physics the viewpoint of the theory of graphs (percolation
Geometrical figures consisting of a number of theory, seepage theory), and by analogy with the
points (vertices) and segments (edges) connect- Kirchhoff problem one can study the passage of
ing these points. Properties of graphs and vari- electric current over a tree with random links.
ous problems related to graphs are studied in the
theory of graphs, which is a branch of discrete GRAVIMETRIC ANALYSIS, weight analysis
mathematics. Their first applications in physics A quantitative analysis of the chemical compo-
occurred in the 19th century when G.R. Kirch- sition of a material through measurements of its
hoff employed them for the calculation of the elec- mass by weighting. Gravimetric analysis is based
tric circuits and A. Cayley used them for cal- on the mass conservation law during chemical re-
culating the number of isomers of the various actions, and on the law of the constancy of com-
saturated hydrocarbons, so-called “trees” or con- position of chemical compounds. As a rule, the
nected, non-oriented graphs without cycles. At analysis includes weighting the sample (or mea-
present graphs are commonly used in various ar- suring its volume), dissolving it, quantitatively
eas of the solid state physics for designing diverse isolating it from the solution in its elemental con-
structures, processes, interactions, and so on. dition or as a barely soluble compound, transfor-
Concrete applications of graph theory involve mation (by drying or heating) of the product into a
the topology of disordered structures, the calcu- stable compound with an exactly known composi-
lation of electron and phonon properties of disor- tion, and weighting the final product. The amounts
dered solids using model lattices (see Bethe lattice, of the desired components in the analyzed product
Cayley tree), the development of the Ising model, are calculated.
and the problem of dimers in a theory of adsorp-
tion (number of nearest neighbor compounds is GREEN’S FUNCTIONS in solid state physics
determined for a distribution of diatomic mole- (H. Green)
cules on a doubly-periodic lattice-surface). A the- Green’s functions are constructed from two-
ory of graphs is used for the graphic representation time correlation functions. It is in terms of two-
of a perturbation theory series, for example, for time Green’s functions (or correlation functions)
Green’s functions (see Feynman diagrams). In a that one can express observed macroscopic char-
Feynman diagram the segments represent “quasi- acteristics of solids (see Macroscopic descrip-
particles” in motion, and the vertices represent tion in solid state physics), microscopic proper-
events; every segment has its own “momentum”. ties (e.g., positions and widths of energy levels,
More generally, diagrams are planar graphs with see Microscopic description in solid state physics),
534 GREEN’S FUNCTIONS
and kinetic coefficients (e.g., electrical conductiv- tem as a whole in state α. The diagram technique
ity, susceptibility). For example, an electrical con- is built for the system of functions
ductivity tensor has the form +
G> αβ (t1 , t2 ) = aα (t1 )aβ (t2 ) ,
σxx (ω) +
G< αβ (t1 , t2 ) = aβ (t2 )aα (t1 ) ,
ie2 n 1 ω
=− δxx + tanh
mω ω 2kB T G+
αβ (t1 , t2 ) = −iΘ(t1 − t2 )Aαβ (t1 , t2 ),
∞ G−
αβ (t1 , t2 ) = iΘ(t2 − t1 )Aαβ (t1 , t2 ).
× jx (0)jx (t) + jx (t)jx (0) e−iωt dt,
−∞ Here A = G> + ηG< , η = 1 for fermions, −1 for
bosons, and 0 for the subsystem; . . . means aver-
where n is the electron concentration, j (t) is the aging over the non-equilibrium state. In a spectral
Heisenberg operator of the current density; and av- representation, i.e. a Fourier representation rela-
eraging is performed over an ensemble in the equi- tive to the time difference t1 − t2 , the following
librium state at temperature T . relations (in a matrix form) are valid:
In solid state physics there are Green’s func-
tions built on creation and annihilation opera- >
− >
−
tors (see Second quantization) of various particles: G < (ω) = G+ (ω)Σ < (ω)G− (ω),
electrons, phonons, magnons, photons (partial 1
G± (ω) = ,
Green’s functions), states of a quantum (sub)sys- ω − Ω − Γ ± (ω)
tem (e.g., an impurity center), and the system as
a whole (global Green’s functions). The technique Σ > + ηΣ < = i Γ + − Γ − ≡ Γ,
of unlinking and the diagram technique are em-
where Ω is determined by a zero-order Hamil-
ployed. The technique of unlinking is based on
tonian, Γ and Σ are eigenvalue energy functions,
obtaining (with the help of time differentiation) a
Γ ± is the mass operator, Σ > and Σ < correspond
chain of equations containing more and more com-
to leaving and entering the state, respectively. The
plicated Green’s functions, and the subsequent
expression Γ = Σ > + ηΣ < has a simple physi-
closing of the chain by unlinking, i.e. by the ap-
cal sense: the width of a level for the system as a
propriate substitution of a mean value of a prod-
whole is equal to the rate of particle interchange;
uct of a certain number of operators by a mean
in the case of fermions (bosons) this is a sum (dif-
value of a product of a lesser number of operators.
ference) of the rates of leaving and entering. The
The diagram technique is based on the expansion
result for the case of bosons means a narrowing of
of a Green’s function in a power series of interac-
the line with the increase of input; and this effect
tions of the particles or parts of the system, bring-
is implemented in a laser.
ing the individual members of the expansion into
The possibility of introducing exact eigen-
correspondence with certain Feynman diagrams,
value energy functions is important for calculating
and the summation of an infinite (this is the heart
Green’s functions. The diagram technique was de-
of the method) number of the most essential dia-
veloped for these functions, and the following two
grams (so-called partial or selective summation).
expressions are valid:
In those cases when one succeeds in applying a
diagram technique the method becomes the most dω
Aαα (ω) = 1,
effective. 2π
Quite general results can be obtained with the
dω <
help of non-equilibrium Green’s functions built on G (ω) = nα ,
2π αα
creation and annihilation operators. Let aα+ and aα
be the operators of creation and annihilation of ei- where nα is the population number operator of
ther particle (fermion or boson, see Fermi–Dirac state α, Aαα (ω) describes the shape of the level
statistics and Bose–Einstein statistics) or a subsys- responding to state α, and G<
αα (ω) is the spectral
GREEN’S FUNCTIONS AND MAGNETISM 535
distribution of the level population. For equilib- imaginary part. Then use is made of a known re-
rium states, the usual single-particle Green’s func- lation between the density of states and the imagi-
tions are commonly used: nary part of the Green’s function, plus a relation
between the Green’s function of an ideal lattice
G0αβ (t1 , t2 ) = Θ(t1 − t2 )G>
αβ (t1 , t2 ) and a perturbed Green’s function of a lattice with
+ ηΘ(t2 − t1 )G< a defect (Dyson equation).
αβ (t1 , t2 ), The method of Green’s functions was initially
where Θ is the step function, and the temperature developed by I.M. Lifshitz in 1940s for vibrations
Green’s functions are obtained from the usual ones of a lattice with defects, and then reformulated
by the exchange t → −iτ , 0 τ (k B T )−1 , by Slater and Koster (1954) for electronic struc-
where −iτ is the imaginary “time” parameter. ture (see Koster–Slater model). At present, vari-
ous versions of the method as well as the molecu-
GREEN’S FUNCTIONS AND DEEP ENERGY lar cluster method are widely used in calculations
LEVELS of point defects (vacancies, substitutional impuri-
A method for calculating one-electron energy ties, interstitial atoms) and complicated defects in
levels and wave-functions of non-metallic crystals insulators and especially in semiconductors, and
with defects that introduce energy levels into the also in calculations of electronic states in crys-
band gap of the main crystal. In this method a one- tals with a surface which is considered as a two-
electron Green’s function is used, i.e. the opera- dimensional defect. The method of Green’s func-
0 (E) = (E − H
tor G 0 )−1 , where E is the energy, tions differs from the cluster method in correctly
and H0 is the effective one-electron Hamiltonian taking into account the “infinite” crystal lattice,
of the crystal. In practical applications the oper- but this becomes labor-consuming when calculat-
ator G0 is replaced by the matrix G0 . The latter ing the complete energy and studying multielec-
is calculated in a basis of localized orbitals, as a tron effects.
rule atomic orbitals based on the band structure
GREEN’S FUNCTIONS AND MAGNETISM
and Bloch functions of an ideal lattice (see Bloch
Methods of quantum field theory applied to
theorem). The enormous dimensionality (order of
magnetism. A principal element is a spin Green’s
the Avogadro number) of matrix G0 does not cause function in terms of which one can describe in a
any complications since, in fact, it is only neces- self-consistent manner the main characteristics of
sary to know a comparatively small submatrix G 0
a quantum magnetic system: the susceptibility, the
of matrix G0 for the portion of the crystal lattice magnetization, the thermodynamic potential, the
around the defect. The defect is considered as a kinetic coef cients, the spectrum of intrinsic ex-
perturbation of an ideal lattice, and is described citations, etc. (see Green’s functions). There are
by the perturbation operator V with the matrix V
retarded, advanced, causal, and spin temperature
expressed in a basis of localized orbitals. In the Green’s functions. Their analytical properties are
latter only a submatrix V is non-zero for the per-
determined both by general requirements of statis-
turbed region near the defect. The unknown levels tical physics, and by symmetry properties of a sys-
in the gap are then determined from the equation tem Hamiltonian. Like the case of Bose and Fermi
Det 1 − G = 0.
0 (E)V operator Green’s functions (Fermi–Dirac statistics
In addition to these equations corresponding and Bose–Einstein statistics), the poles of retarded
to localized electronic states (see Local electronic spin Green’s functions determine the energy and
levels), the defect-induced resonance states can attenuation of the intrinsic vibrations of the spin
be calculated with the help of Green’s functions. system. One of the calculation modes consists in
These states appear as peaks and valleys on the breaking the infinite chain of the equations for
curve
n(E) = n(E) − n0 (E), where n0 (E) and Green’s functions so that in a particular approxi-
n(E) are, respectively, the densities of states in an mation the chain is reduced to a closed one. A cal-
ideal crystal and in a crystal with a defect. To find culation of the spin Green’s function can also be
these, the energy E in the Green’s function is con- carried out by using a representation of spin oper-
sidered as a complex number with an infinitesimal ators in terms of Bose or Fermi operators (see Spin
536 GREEN’S FUNCTIONS AND SUPERCONDUCTIVITY
operator representations). The Wick theorem was where Γ (p1 , p2 , p3 , p4 ) is the vertex part (four
formulated and the diagram technique was devel- pole) of the effective non-local (retarded) elec-
oped for use with spin operators. However, since tron–electron interaction, p{k, ω} is the energy–
a spin operator commutator is also an operator, momentum four-vector (here and below = 1).
the algebraic grounds for such a mode of Green’s The four-pole Γ contains the electron–phonon in-
function calculation consists in a subsequence teraction due to the exchange of virtual phonons
of more complicated relations (in comparison (see Virtual transitions) and the Coulomb repul-
with Bose and Fermi operators). The perturbation sion shielding due to the exchange of virtual
approach to spin type Green’s functions is suffi- plasmons (see Electric charge screening). The
ciently straight forward. There exist several vari- electron–phonon interaction is described by the
ants of the perturbation technique (expansion over phonon Green’s function which is determined by
reciprocal spin, magnon density, reciprocal radius the Dyson equation:
of interaction) with each suitable for a particular
(p)D(p),
D(p) = D0 (p) + D0 (p)Π (3)
application, and they provide descriptions of high-
frequency, thermodynamic and kinetic properties where D0 is the bare (un-renormalized Green’s
of magnetic systems. is the self-energy part of the phonons:
function), Π
d4 p
GREEN’S FUNCTIONS AND
Π(p) = −2ig G(p )G(p − p)
SUPERCONDUCTIVITY (2π)4
A field theory treatment of the superconducting g (p , p − p; p),
× (4)
properties of conduction electrons in metals with
the help of the Dyson–Gorkov equations for nor- where g and g are the bare and renormalized ver-
mal and anomalous Green’s functions (see Green’s tex parts (three pole) of the electron–phonon in-
functions, Gorkov–Nambu method). In the case of teraction. The shielded Coulomb interaction in the
singlet Cooper pairing (see Cooper pairs), these random phase approximation takes the form:
equations in momentum–energy space take the
form: (k, ω) = V (k) 4πe2
V ≡ 2 , (5)
1 − V (k)Π(k, ω) k ε(k, ω)
ω − ξ(k) − Σ(k, ω) G(k, ω)
where
− Δ(k, ω)F + (k, ω) = 1,
4πe2
(1) ε(k, ω) ≡ 1 − Π(k, ω)
ω + ξ(k) + Σ(−k, −ω) F + (k, ω) k2
is the dielectric constant, and
−
+ (k, ω)G(k, ω) = 0,
4
d p
where ξ(k) is the energy of electrons with quasi- Π(p) = −2i G(p )G(p − p)
(2π)4
momentum k in the conduction band measured
from the Fermi level, Σ and Δ are, respec- is the electron polarization operator. Eqs. (1)–(5)
tively, the normal and anomalous eigenenergy together with the expression
parts (mass operators of quantum mechanics) of
electrons which are determined by the following
g (k, ω) = g(k) + V (k)Π(k, ω)
g (k, ω)
integral equations: g(k)
≡ (6)
ε(k, ω)
d4 p
Σ(p) = i Γ (p, p ; p, p )G(p ), form a closed set of nonlinear equations that is
(2π)4
(2) used for the description of physical properties of
d4 p superconductors with strong coupling (see Eliash-
Δ(p) = i Γ (p, p ; p, p )F (p ), berg equations).
(2π)4
GROUP CHARACTER 537
GREEN’S TENSOR for atomic bond breakage, so that the crack can
A tensor of rank two, Gik , whose elements are propagate under the constant external stress with-
the proportionality coefficients between the com- out external energy input. Since the energy gain
ponents of the displacement vector ui and the increases as the square of the crack length, the
components of the force Fk applied to a certain crack spontaneously accelerates either up to the
point of an infinite elastic medium: ui = Gik Fk . total fracture of the body, until it enters a region
The components of Green’s tensor are homoge- without tensile stress, or until it enters a region
neous functions of the coordinates to first order. where there is compression stress in a plane nor-
For an isotropic medium the tensor is given by mal to the crack. The Griffith critical stress needed
1+ν 1 for crack development under load in the case of a
Gik = (3 − 4ν)δik + ni nk , planar stressed state is the following:
8πE(1 − ν) r
where ν is Poisson ratio, E is Young’s modulus,
2γ E
δik is the unit tensor (Kronecker delta), ni (i = σG = ,
1, 2, 3) are components of a unit vector, and r is πL
the distance from the force application point.
where E is Young’s modulus, and γ is the surface
energy. The thermodynamic criteria for the crack
GRIFFITH THEORY (A.A. Grif th, 1920)
growth are necessary conditions only. To carry out
A theory that describes possible reasons for
the process it is necessary to provide at the crack
the lack of conformity between the actual strength
tip a local stress that is within an order of magni-
and theoretical strength of crystalline and vitre-
ous bodies (see Vitreous state of matter). Griffith tude of the theoretical tensile strength of the metal.
suggested that when a body is fracturing, the stress The Griffith criterion could be both a necessary
need not be equal to the theoretical strength over and a sufficient condition for fracture of a crys-
the whole body bulk but only at the tip of a sharp talline solid if the crack tip radius is comparable
and narrow crack. Such a crack may become un- in size to an interatomic distance.
stable under relatively low nominal stress. His rea-
soning was based on a thermodynamic approach GRINDING
that correlates an increment of crack length with Treatment of sample surfaces by abrasive de-
a change of the potential energy of the stressed vices (see Abrasive materials). The grinding of
system. If there is an external elongation stress it
metallic wares is accomplished on grinding ma-
produces a penetrating crack of length L which
chines by rotating abrasive wheels or whetstones.
is small compared to the dimensions of the body,
In the case of surfaces that are difficult to treat,
and a zone free of elastic stress arises in the region
electrochemical (electrolytic) grinding wheels are
adjacent to the crack. Increasing the crack size in-
provided with a constant electric current, and an
creases the volume of the relaxation zone and, con-
electrolyte is used for the cutting section. Grinding
sequently, it increases the relaxed elastic energy.
The concomitant formation of crack voids requires may be carried out by a soft abrasive (see Polish-
an expenditure of energy for breaking chemical ing), by abrasive powder suspended in a liquid, and
bonds. As the crack size increases, the energy bal- also by means of a vibrational treatment. For the
ance of the system changes, and the energy needed grinding of stones one may employ carborundum
for forming new interfaces increases linearly with plates, whetstones, etc., of various grain sizes.
the crack length L, while the energy gain increases
in proportion to L2 . As a result at a certain critical GRINDING, DISPERSION
value Lcr the energy expenditure reaches a max- See Dispersion grinding.
imum, beyond which the total potential energy
of the system begins to decrease continuously. It
means that the gain of elastic energy in the bulk in GROUP CHARACTER
the vicinity of the crack exceeds the expenditure See Character of group representation matrix.
538 GROUP
corresponds to normal dispersion, u < v to anom- growth) off an individual face. If the crystal has
alous dispersion, and u = v indicates the absence the shape of a polyhedron (see Crystal polyhe-
of dispersion. The direction of the group velocity dron), its body consists of growth pyramids cen-
for an electromagnetic wave coincides with the di- tered at the nucleus or seed, from which the crys-
rection of the Poynting vector S = E × H whose tal has grown. If a certain face is generated at
magnitude equals the the energy S passing in a a later stage of crystal growth, then the vertex
unit time through a unit area normal to the energy of the corresponding growth pyramid is at the
flux direction. In anisotropic media the group ve- point, at which the face had originated. A particu-
locity direction, in general, does not coincide with lar case of a growth pyramid is a tubular domain,
the phase velocity direction, and we have the re- which is left in a crystal pulled from the melt by
lation u cos α = v, where α is the angle between an atomically smooth (singular) face, and differs
the directions of the group velocity and the phase from other tubular (cylinder-shaped) domains that
velocity. frame it, and have formed on atomically rough sur-
rounding regions of the crystallization front. The
GROWTH ANISOTROPY growth pyramid is often bounded by nonplanar
Differences in properties of solids along differ- surfaces, which are constituted by traces of the
ent directions, induced during the growth process. motion of edges that frame the face, which forms
It is most prominently represented by the pres- this pyramid. The crystal consisting of growth
ence of magnetic anisotropy in magnetically or- pyramids exhibits the so-called sector structure
dered media, i.e. in the surface regions of large (the existence of these sectors is the consequence
monocrystals, and in thin monocrystalline or amor- of the lack of thermodynamic equilibrium in the
phous lms. In amorphous films growth anisotropy processes of crystal growth; it breaks the ideal mi-
often plays a dominant role (see Epitaxial iron gar- croscopic crystal symmetry to a variable degree
net lms, Metallic glasses). The existence of a spe- with respect to different properties). The growth
cial direction, such as the normal to the surface of pyramids differ from one another in properties be-
a growing crystal or film, is a physical character- cause of the difference in concentrations of impu-
istic of magnetic growth anisotropy. The contribu- rities (sometimes by several fold), inclusions, dis-
tion of growth anisotropy to the free energy f of locations, etc. (see Defects in crystals). Hence they
a medium depends on the orientation of the mag- deviate in values of crystal lattice constants (up to
netization vector with respect to the growth direc- 10−3 %). Therefore, there occur internal stresses
tion n, i.e. at the interfaces between the growth pyramids in
a crystal. Even the growth pyramids of crystallo-
fg.a. = kij βi βj + kij kl βi βj αk αl + · · · ,
graphically equivalent faces can differ in proper-
where kij , kij kl , . . . are the components of the ten- ties, for the defects “memorize” not only the index
sors of the growth magnetic anisotropy constants, of a face, but also the direction of its growth. For
βi and αk are the respective direction cosines of instance, the so-called anomalous pleochroism in
the magnetization vector M with respect to n and the growth pyramids of the minor rhombohedron
to the edges of the crystal unit cell (in amorphous of quartz is due to the fact that the substitution
films there is no contribution of crystallographic of Si4+ for Al3+ + Na+ proceeds with different
anisotropy). Growth anisotropy is metastable; and probabilities in three silicon–oxygen tetrahedra
it disappears during long-term, high-temperature which are inequivalent with respect to a particu-
annealing. lar face.
which is a measure of the frequency shift induced Frederiks transition, the dye molecules reorient in
by the change of volume V for an arbitrary strain. common with the host molecules. If the dye has a
The value of γS can be estimated from experimen- dichroic absorption band in the visible range, then
tal data as a mean value this reorientation leads to a change of the color
intensity of the system. In the case of two mutu-
1
3N
V ally perpendicular polarized absorption bands in
γ= γS = βK,
3N CV the dye spectrum, a switching of color is observed.
S=1
When using luminescent dichroic dyes, the lumi-
where β is the thermal expansion coefficient (see
nosity intensity of the system changes. The guest–
Thermal expansion), K is the bulk modulus, CV is
host effect is employed in colored liquid-crystal
the speci c heat at constant volume, and N is the
displays.
total number of particles. The absolute values of
the dimensionless parameter γ , which are some-
GUINIER–PRESTON ZONES, G-P zones
times called Grüneisen constants, vary from 0.5
(A. Guinier, G.D. Preston, 1938)
for covalent crystals to 3.0 for crystals with weak
Regions or clusters without a definite crystal
van der Waals forces, thus reflecting to some ex-
structure that emerge from a supersaturated solid
tent the anharmonicity of vibrations (see Anhar-
solution during the initial stages of precipitation.
monic vibrations). Generalizing γS for any kind of
They involve a process of solid solution decompo-
deformation leads to the relation
sition (see Alloy decomposition) that is accompa-
(S) d ln ωS V nied by concentration variations in small regions
γij = − and γij = cij kl αkl ,
d ln uij CV of the crystalline host. In a general case, the for-
where uij , cij kl , and αkl are the tensors of defor- mation of these zones induces atom displacements
mation, elasticity and thermal expansion, respec- relative to the sites of the initial lattice (see Al-
tively. Thus, γij is a symmetric tensor of rank loy aging). At early stages of aging the X-rays
two whose components within the accuracy of scattered from Guinier–Preston zones of Al–Cu
the known scalar V /CV are equal to the com- monocrystals indicate the presence of layer-like
ponents of the thermal stress tensor cij kl αkl in a units 0.5 nm thick and 5–10 nm wide oriented par-
“squeezed” crystal. The Grüneisen tensor is ap- allel to {100} planes of the host cubic lattice. The
plied in studies of thermal expansion, the depen- formation of Guinier–Preston zones can result in
dencies of cij kl on temperature and pressure, the the strengthening of alloys.
behavior of solids in the state of hydrostatic and
impact compression, the lattice thermal conduc- GUNN DIODE
tivity of semiconductors and insulators, the scat- A semiconductor diode with a negative differ-
tering of various ionizing radiations and particles ential resistance arising in an active layer of a
on crystal lattice vibrations, the absorption and at- semiconductor as a result of the Gunn effect. The
tenuation of ultrasonic waves as a result of their transformation of energy from a source to the mi-
interaction with phonons, quantum zero point mo- crowave oscillations in a diode is carried out in
tion, and so on. various ways depending on the level and the pro-
file of the impurity distribution in a sample, the
GUEST–HOST EFFECT active layer dimensions, the cathode contact prop-
An electrooptical effect involving the change of erties, the offset value of the diode, and the type of
the intensity of coloration, color, or luminosity of external circuit. A Gunn diode is used for the gen-
a system consisting of a “dichroic dye in a liq- eration, amplification, and transformation of mi-
uid crystal” (guest–host) at a Frederiks transition. crowaves in the frequency range 4–200 GHz; it has
Due to the geometric anisotropy, the dye mole- a low noise level (see Gunn effect).
cules are oriented in the liquid crystal host in ac-
cordance with its director orientation. Therefore, GUNN EFFECT (J.B. Gunn, 1963)
when an electric field is applied to the system with The appearance of a portion of a semiconduc-
a strength higher than the threshold value of the tor current–voltage characteristic with a negative
GYRATION 541
can be distinguished on the basis of Onsager re- a molecular ordering in cholesteric liquid crystals,
lations (see Onsager theory) according to which, and spins in magnets with modulated magnetic
e.g., structures. It can manifest itself in both amorphous
and liquid phases of a matter if the lack of bilat-
χik (k, ω, B 0 ) = χki (−k, ω, B 0 ), (1)
eral symmetry is inherent in particles composing
where B 0 is the magnetic ux density. On the the material. Weak but observable gyrotropic ef-
strength of this relation (or an identical one for fects appear due to the violation of space parity in
εik ), the gyrotropy of the medium can be caused weak interactions.
by either a spatial dielectric dispersion (depen- In the presence of a magnetic induction B 0 =
dence on k), or by the presence of a magnetic μ0 (H 0 + M 0 ), the medium can be gyrotropic
field B 0 . even without taking into account space disper-
If the gyrotropy is due to spatial dispersion, sion. In these cases, magnetic gyrotropy is said to
the corresponding gyrotropic medium is called a be present, and the gyrotropic medium is called
medium with natural optical activity (see Gyra- magnetogyrotropic. In magnets, the magnetogy-
tion). Based on Eq. (1), a tensor form typical for rotropy is present even without the external mag-
this is the following: netic field B 0 due to the magnetization M 0 (in
(0) magnets with a complicated type of magnetic or-
εik = εik + iεikj
gk j , (2) dering, e.g., in antiferromagnets, the contribution
where εik (0) = εki (0), εikj is the totally antisym- of the antiferromagnetism vector L can be signifi-
metric third-rank tensor, g is the gyration ten- cant). A magnetogyrotropic medium can be both
sor, and an identical relation is valid also for χik . gyroelectric (e.g., a plasma in a magnetic field,
Apparently, on the basis of Eq. (2), a gyrotropic including a solid-state plasma) and gyromagnetic
medium must be invariant with respect to some (magnets: paramagnets and materials with mag-
mirror reflections. This property can be caused by netic ordering) at the same time. A gyrotropic
the symmetry of a gyrotropic medium: by the pres- medium is characterized by a variety of particular
ence of a screw axes in the crystal (see Crystal features in regard to electromagnetic wave propa-
symmetry) as well as by a screw-like character of gation (see Rotation of light polarization plane).
This Page Intentionally Left Blank
Hh
HADFIELD STEEL (after English metallurgist a = 0.351 nm, space group I m3m (Oh9 ). Bind-
R.A. Had eld) ing energy of pure hafnium is 6.35 eV/atom
High-manganese austenite steel which con- (at 0 K). Density is 13.3 g/cm3 , T boiling =
tains 11 to 15% Mn and 1 to 1.5% C. It can 4870 K; latent heat of melting is 21.8 kJ/mole;
be strongly hardened under dynamic loads. Af- latent heat of sublimation is 712 kJ/mole; latent
ter slow cooling the structure consists of austen- heat of evaporation is 650 kJ/mole; specific heat
ite and double carbides of the cementite type 143 J·mole−1 ·K−1 (at 298 K); Debye temperature
(Fe, Mn)3 C. To obtain a single-phase austenite is 252 K for hafnium with purity 99.96%; linear
structure it is heated up to 1050–1100 ◦ C for thermal expansion coefficient is 5.9·10−6 K−1 (in
several hours with subsequent cooling in water.
the temperature range 273 to 1273 K); thermal
Hadfield steel is used as a resistant to wear mate-
conductivity coefficient with 2 at% Zr decreases
rial under conditions of moderate wear, but it has
from 22.32 to 20.52 W·m−1 ·K−1 (in the range of
low resistance to abrasive wear. It has good yield
temperatures 323 to 723 K); adiabatic coefficients
(in the liquid state) which allows its use as a cast-
ing material (see Casting). A strong tendency to of elastic rigidity of α-Hf monocrystal: c11 =
strain hardening complicates rolling and forging. 181.1, c12 = 77.2, c13 = 66.1, c33 = 196.9, c44 =
Poor thermal conductivity is a cause of forming hot 55.7 GPa (at 298 K); bulk modulus is 108.9 GPa,
cracks in casting and cracks upon the accelerated Young’s modulus is 83.4 GPa; rigidity is 30.4 GPa,
heating for quenching. Poisson ratio is 0.37 (at 298 K); Vickers hardness
is 152 in HV units for hafnium iodide (0.72% of
HAFNIUM, Hf Zr); thermal neutron capture cross-section is 105
A chemical element of Group IV of the pe- barn; ion-plasma frequency is 13.3 THz. Sommer-
riodic table with atomic number 72 and atomic feld linear temperature term in low-temperature
mass 178.49. Natural hafnium consists of 6 sta- specific heat is 2.15 mJ·mole−1 ·K−2 ; electrical
ble isotopes with the mass numbers 174, 176 resistivity is 306 n·m (at 295 K), temperature
to 180. Electronic configuration of outer shells coefficient of electrical resistivity is 0.00351 K−1
is 4f 14 5d 2 6s 2 . Successive ionization energies (273 to 1073 K); work function of polycrystal is
are 7.5, 15.0, 23.3, 33.3 eV. Atomic radius is 3.53 eV; superconducting transition temperature
0.159 nm; radius of Hf 4+ ion is 0.078 nm. Ox- is 0.13 K; hafnium is paramagnetic with mo-
idation is state is +4 (less common +3, +1).
lar magnetic susceptibility +70·10−6 cgs; nuclear
Electronegativity is 1.23.
magnetic moment of 177 Hf isotope (18.39% abun-
In the free form hafnium is a silvery-white
dant) is 0.61 nuclear magnetons.
metal; it is pliable, easily subjected to cold and
hot processing (rolling, forging, stamping, milling, Hafnium is used in nuclear power engineering
drilling). It exists in α- and β-modifications. (control rods of reactors, screens protecting from
α-Hf has hexagonal close-packed lattice with neutron radiation) and in electronic engineering
parameters a = 0.3188 nm, c = 0.5042 nm at (cathodes, getters, electrical contacts), manufac-
273 K; space group P 63 /mmc (D6h 4 ). Above ture of high-temperature alloys. A solid solution
2013 K up to T melting ≈ 2500 K β-Hf is sta- of hafnium and tantalum carbides is the most re-
ble, body-centered cubic with lattice parameter fractory ceramic material (T melting = 4270 K).
545
546 HAGEN–RUBENS APPROXIMATION
HAGEN–RUBENS APPROXIMATION (E. Hagen, described by the Petch relation, where the value of
H. Rubens, 1903) the cell size is substituted for the grain size. Strain
Theory of optical absorption in metals and hardening is carried out during the process of plas-
narrow-band semiconductors at low frequencies tic working (pressing, forging, rolling, drawing,
ω ωτ (ωτ = 1/τ is the collision frequency, etc.). The nature of the dislocation structure is de-
τ is the electron mean free path time); accord- termined by the degree and the temperature of the
ing to this theory, optical properties of materials strain.
depend only on their static conductivity σ0 . For The hardening due to the formation of dis-
this case the refractive index ñ = n + iχ , where persed particles of the second phase (see Precipit-
n ≈ χ = [σ0 /(2ε0 ω)]1/2 , ε0 = 8.85·10−12 F/m ation-hardened materials, Dispersion hardening)
is the vacuum dielectric constant. In this case, may be either direct or indirect. Direct hardening
the re e ctance R ≈ 1 − 2/n. The deviation from is related to the direct interaction between dislo-
total reflection is given by the Hagen–Rubens for- cations and dispersed particles, the latter acting
mula 1 − R ≈ (8ε0 ω/σ0 )1/2 , i.e. the lower the as barriers for dislocations, which slip during the
conductivity of metals and semiconductors, the course of plastic deformation. Indirect hardening
lower is their reflecting power. In the Hagen– is related to the fact that the presence of dispersed
Rubens approximation, the absorption coef cient particles of the second phase brings about an
α ≈ (2μ0 σ0 ω)1/2 ∝ ω1/2 (ω ωτ ), where increase of stability of the nonequilibrium struc-
μ0 = 4π·10−7 H/m is the magnetic permeabil- tural state, and an increase of the recrystallization
ity of vacuum. At high frequencies, ω
ωτ , temperature. Precipitation-hardened systems are
the Drude formula is valid, according to which obtained by means of alloy aging (age hardening),
α ∝ ω−2 . by application of powder metallurgy methods, etc.
Hardening through alloying is widely used in
HALL EFFECT engineering; the method being based on the in-
See Galvanomagnetic effects. teraction between dislocations and the atoms of
dissolved elements (see Solid-solution hardening).
HARDENING of materials In the case of alloys subject to phase tran-
A process for increasing the strength of a mate- sitions, the hardening is achieved through phase
rial. Strain hardening is widely used for increasing recrystallization during the course of a heat treat-
the strength of metals. The strain may give rise ment (quenching, tempering, etc.). The hardening
to a “forest of dislocations” (uniformly distrib- in this case is due to a directional change of the
uted dislocation structure), or to a cellular structure phase composition, grain refining, phase peening
(see Fragmentation). In the former case, the hard- (see Strain hardening, Work hardening).
√ See also Dispersion hardening, Flash hard-
ening achieved is higher (
τ ≈ αGb ρ, where
α = 0.1–1.5, G is the shear modulus, b is Burg- ening, Irradiation hardening, Linear hardening of
ers vector, ρ is the dislocation density), but it crystals, Solid-solution hardening, Bulk sound
is accompanied by a decrease of plasticity. The hardening, Shock hardening, Surface hardening,
hardening achieved in the latter case is somewhat Surface sound hardening.
lower, but the structure obtained is more stable,
the hardening process being accompanied by a HARD MAGNETIC MATERIALS, high-coercive
decrease of the cold brittleness temperature, and materials, hard magnetics
an increase of plasticity at low temperatures. At Materials characterized by a large coercive
high enough degrees of reduction, when the dis- force (B c = 10−2 –1 T). The origin of high B c
orientation of individual dislocation cells relative in hard magnetic materials lies in the lag of
to one another exceeds a certain critical value magnetic reversal in ferromagnets, arising from
(∼2–3◦ of arc), the cell boundaries start to behave three main mechanisms: (i) irreversible magneti-
as grain boundaries, acting as barriers to disloca- zation rotation processes; (ii) lag in the forma-
tion motion. In this case, the hardening process is tion and growth of remagnetization nuclei; and,
HARD MAGNETIC MATERIALS 547
Table 1. Principal characteristics of the most important hard magnetic materials (1 kA/m corresponds to 1.26 mT and
12.6 Oe). Note that Bc = μ0 Hc
Magnetic properties
Material Composition Hc Br (BH )max Remarks
(iii) pinning of domain walls at various crystal ir- HARD MAGNETIZATION AXIS
regularities and defects. Hard magnetic materials The direction of the magnetization vector of a
find their use in permanent magnets, hysteresis ferromagnet, M, which corresponds to the maxi-
motors, and magnetic recording. Employed in the mum of the magnetic anisotropy energy. To mag-
last two cases are magnetic materials with mod- netize an infinite magnetic substance (with no
erate B c = 10−3 –10−2 T, while hard magnetic elastic strains present) along the difficult magne-
materials with B c = 10−2 –1 T are used for per- tization axis, it is necessary to subject the crystal
manent magnets. The appraisal of a hard magnetic to the action of an external magnetic field with
material is mainly based on its demagnetization a strength greater than that needed to magne-
curve, i.e. on the portion of the major hysteresis tize it along any other direction. If the magnetic
loop lying in the 2nd quadrant. anisotropy is not unidirectional, the difficult mag-
The most important parameters of a hard mag- netization axes are axial directions; so their num-
netic material, along with B c , are the remanence, ber is always even. The direction corresponding to
B r , and the maximum value of the product of the absolute maximum of the magnetic anisotropy
fields on the demagnetization curve, (BH )max , energy is sometimes referred to as the most diffi-
the so-called energy product. cult magnetization axis. See also Easy magnetiza-
The high-coercive state in hard magnetic ma- tion axis.
terials is formed following a certain production
cycle (sometimes, a rather complex one). In the
context of present day technology hard mag- HARDNESS
netic materials can be classified into the following The ability of materials to resist plastic de-
groups (see Table 1): formation or brittle failure under local loading.
Its magnitude and dimensionality depend on the
(1) steels hardened into martensite, displaying method and conditions of measurement. The most
relatively low B c , and rarely used; widespread method is the static indention of an
(2) rigid cast alloys like alni, alnico and ticonal, especially loaded solid indenter into the material
displaying a broad spectrum of magnetic prop- to be studied. In this case, the hardness is deter-
erties; these alloys are the most common ma- mined by a ratio between the load and the imprint
terials for permanent magnets; area. Applied also are tests with dynamic indent-
(3) ductile alloys like vicalloy, cunife, cunico, the ing, pendular methods, methods of elastic recoil,
alloys Fe–Co–Cr, Mn–Al–C, and also noble sclerometric methods (scratching tests), and so
metal alloys: Pt–Co, Pd–Fe and Pt–Fe; on. There are established correlations of hardness
(4) powdered hard magnetic materials produced with various mechanical characteristics of mate-
by powder pressing followed by heat treat- rials, such as elastic moduli, yield limit, ultimate
ment. Metalloceramic, and metalloplastic hard
strength. See also Microhardness and Macrohard-
magnetics are distinguished here, along with
ness.
oxide magnets and powder magnets.
HARTREE–FOCK METHOD (D. Hartree, 1928; and β are spin wave functions corresponding to
V.A. Fock, 1930) the two spin directions. In the restricted Hartree–
Quantum-mechanical method of description Fock method, the functions of the space coordi-
and calculation of systems of interacting fermi- nates (orbitals) ϕm and ϕ̄m are assumed to have
ons, which is widely applied to many-electron an identical form (spin limitation) and to trans-
systems: atoms, molecules, crystals. Sometimes form with respect to the irreducible representation
the term “Hartree–Fock method” is used in refer- of the symmetry group of the system under con-
ence to a purely multiplicative approximation for sideration (symmetry limitation). For crystals ϕm
the wave function, a set of coupled equations to can be expressed in Bloch form. The nonlinear-
solve,
one for each electron, containing
Coulomb ity of the Hartree–Fock equations is responsible
( ψk∗ V ψk dτ ) and exchange ( ψj∗ V ψk dτ ) in- for their plurality of solutions. The solutions of
tegrals, and occasionally exchange terms are ne- the restricted Hartree–Fock method may be un-
glected. In the usual Hartree–Fock method the stable, i.e. may correspond to a saddle point of
wave function of a system of N electrons has the energy surface. Then the decrease in energy
the form of a Slater determinant, the product of at the removal of spin or symmetry restrictions
one-electron wave functions (spin-orbitals) ψ1 , (unrestricted Hartree–Fock method) can be ac-
ψ2 , . . . , ψN . As a rule, the spin-orbitals are as- companied by the appearance of components of
sumed to be orthonormal, and are determined different wave function multiplicities of the sys-
through the variational principle. The conditions tem, or by a violation of its symmetry (solutions
for a stationary energy functional lead to the of the charge density wave or spin density wave
problem of determining pseudo-eigenvalues of the type in crystals).
Fock operator (Fockian): F (P )ψk = εk ψk , where The Hartree–Fock equations are integrodiffer-
F (P ) = h + J (P ) − K(P ), h includes the oper- ential because of the nonlocality of K(P ). Even
ator of the kinetic energy of the electrons and the in the simplest cases they cannot be solved analyt-
potential energy of attraction of the electrons to ically. The spin-orbitals in crystals are constructed
the nuclei. The Coulomb term J (P ) and exchange through the linear combination of atomic orbitals
operator K(P ) are defined on spin-orbitals ψk (LCAO) approximation, or through expansion in a
through the Fock–Dirac density matrix: plane wave basis set, with coefficients determined
by the variational method. Atomic orbitals are
P= |ψk ψk |. constructed from Slater (exponential) or Gaussian
1kN functions, or through the use of solutions of the
The term J (P ) includes the diagonal elements atomic problem with the so-called muf n-tin po-
of P , which represent the density of particles. In tential. The necessary matrix elements are found,
practice, the solutions are obtained through iter- as a rule, by numerical integration, but sometimes
ative procedures and self-consistency. In deriving (e.g., with the aid of Slater’s basis set) they may
the variational equations, the self-action terms are be found analytically.
formally included in operators F and K; these Besides this nonempirical (ab initio) approach,
terms compensate each other, and this justifies based on an algebraic modification of the Hartree–
the formation of the Hermitian Fockian, which is Fock method, various semiempirical and simu-
common for all spin-orbitals. This is significant lation methods have been developed for calcu-
for fulfilling the conditions of spin-orbital ortho- lating the electronic structure of molecules and
normality. The initial Hartree method used a wave crystals. In these methods certain parts of mole-
function product of spin-orbitals without antisym- cular integrals are neglected, so the remainder
metrization. In this case exchange terms do not are calculated using simplified equations, or are
appear, and each spin-orbital ψk is related to its considered as parameters (see Hoffmann method,
operator Fk , which does not contain the self-action Mulliken–Wolfsberg–Helmholtz method). A draw-
term of the kth electron in J (P ). back of the Hartree–Fock method is that it neglects
In the Hartree–Fock approach the spin-orbitals electron correlations, which are due to Coulomb
are specified in the form ϕm α and ϕ̄m β, where α interactions between electrons; these correlations
HEAT CONTENT 551
are especially important for electrons of oppo- of solids most affected are the decrease of the yield
site spin. The error of the Hartree–Fock method limit and strength (see Dehardening), and the in-
with respect to the determination of the binding crease of the plasticity. The healing of defects is
energy (correlation energy) of atoms and mole- accompanied by the release of stored elastic en-
cules reaches 1–2 eV per electron pair. To take ergy, and by an increase in density. The change in
into account electron correlations, which are not electrical conductivity during healing of defects de-
always included in perturbation theory, it is suffi- pends on the electronic structure. The annihilation
cient to employ self-consistent approaches, which of defects in metals increases this conductivity,
make use of the basic set of several Slater deter- whereas the inverse phenomenon is observed in
minants (many-configuration approach; see Many- semiconductors and insulators. Type II supercon-
con guration approximation). ductors change their parameters upon the healing
of defects due to the decrease in the vortex pin-
HASIGUTI PEAK (R.R. Hasiguti, 1960) ning. Healing of defects affects ferromagnets also:
Maximum point of the temperature depen- the coercive force decreases, but the magnetic per-
dence of the internal friction of crystals with a face- meability increases. The healing of defects can
centered cubic structure, which contain point de- influence magnetic properties indirectly through
fects: impurity atoms, vacancies, interstitial atoms. the change in texture.
It is observed in samples which are deformed at
some temperature and quickly cooled (quenched). HEALING OF SOLIDS with cracks, restoration of
The Hasiguti peak is caused by thermally acti- solids with cracks
vated disconnections of dislocations from point Recovery of continuity and strength of a dam-
defects under the action of oscillating stresses aged solid, i.e. healing of an already existing crack
which accompany low-frequency vibrations of the (see Healing of defects). The latter may go either
sample. There are several variations of Hasiguti via diffusion (see Diffusion) or via direct recovery
peak theory (Hasiguti himself proposed five mod- of molecular bonds between the lips of the opened
els), which specifically take into account the crack (in the absence of mass transport). An ex-
motion of dislocations, and interactions between ample of this restoration is the collapse of the
dislocations and point defects (see Point defect– Rosette channels when twinning interlayers leave
dislocation interaction). the volume of a monocrystal.
The strength of a restored crystal typically
HEALING OF DEFECTS reaches only 70–80% of its initial strength. This is
Processes involving redistribution and annihi- due to the bonding being only partially recovered
lation of defects (see Annihilation of dislocations) through the body cross-section. The healed crack
in crystals. Healing of defects includes restoration contains various discontinuities, e.g., hollow chan-
of crystals, elastic recovery, recovery of proper- nels along the spalling steps and fronts formed at
ties, polygonization, and recrystallization. Practi- crack stops.
cally complete healing of atomic scale defects
(nonequilibrium point defects, dislocations) occurs
HEAT CAPACITY
upon heating (recrystallization annealing), while
See Speci c heat, Surface speci c heat.
partial healing of defects is also possible dur-
ing the course of plastic deformation (dynamic
recovery, dynamic recrystallization). To heal de- HEAT CONDUCTION
fects of a macroscopic scale (macrodefects, pores See Thermal conductivity.
in solids, cracks), the heating is combined with
high pressure processing. The process of hydroex- HEAT CONDUCTIVITY FACTOR
trusion can bring about the dynamic healing of See Thermal conductivity coef cient.
defects of a microcrack type. Through this heal-
ing, the totality of structurally sensitive properties HEAT CONTENT
of the solid is changed. The mechanical properties See Enthalpy.
552 HEAT EMISSION
not always, present in a heat sink. A phonon sink Heat transfer), and by the interaction of its surface
is a particular case of a heat sink when the phonon with the environment (see Heat loss). The main
energy (of lattice vibrations) is the only energy be- kinds of heat transport are thermal conductivity,
ing utilized. convection, thermal radiation, and heat transfer at
phase transitions (heat of phase transition). The
HEAT SWITCH heat transport depends on the geometric dimen-
A device which provides a sharp change of sions and the shape of the body, its temperature
heat transfer between objects (e.g., between the and thermal conductivity, the degree of brightness
consecutive circuits of a cooling system). The of its surface, and also on the temperature and
most widespread types of heat switches are the fol- the physical properties of the external medium. Its
lowing: mechanical heat switches where the heat intensity serves as a quantitative measure of the
linkage is through mechanical contact (e.g., by the transport of heat, and this is expressed through a
compression of elastic plates which detach at the thermal transport coefficient with the dimension-
end of the compression); gaseous heat switches ality W·m−2 ·K−1 . The heat transport at phase
where the heat transfer takes place through the transformations in solids is due to the change in
gas thermal conductivity (the gas is evacuated to enthalpy (see Thermodynamic potentials); this is
break the contact); superconducting heat switches determined by both the thermal conductivity and
which utilize the large difference in the thermal the ability of a body to accumulate heat, i.e. by its
conductivity of metals in normal and supercon- diffusivity.
ducting states (see Superconductivity). The latter is
widely used in obtaining temperatures below 1 K HEAT TREATMENT
owing to the simplicity of controlling such a heat Modification of the structure, phase composi-
switch: its transition from the superconducting to tion, and related physical and mechanical prop-
the normal state is easily brought about by apply- erties of metals and alloys by subjecting them to
ing a magnetic field that exceeds a critical value. heating and cooling. The temperature and its rate
of change can vary over wide ranges, determined
HEAT TRANSFER by the combination of a number of material char-
Heat transport between two carriers of heat acteristics.
through a solid wall or interface separating them. There is a hardening and dehardening treat-
A heat transfer process includes a heat loss from ment, and also treatments for imparting some tech-
a hot liquid or gas to a wall, thermal conductiv- nological characteristics to a metal. The hardening
ity in the wall, and the heat loss from the wall to effect is achieved by raising the degree of pre-
a colder movable medium outside. The magnitude cipitation of hardening phases (see Precipitation-
of the heat transfer is characterized by a heat trans- hardened materials) which are associated with
fer coefficient k = δQ/(
T dS) where δQ is the the host, or with the formation of other concen-
heat flux through a surface element of area dS, tration and structural inhomogeneities enabled to
and
T is the temperature difference (sometimes play a role as obstacles for dislocation motion,
called temperature pressure) between the medium e.g.: grain boundaries, segregations, twins (see
and the surface. The value R = 1/k is called the Twinning structure), and so on. In the case of a
overall thermal resistance. In most cases k is de- dehardening heat treatment, the role of the above-
termined by testing. mentioned factors is less. The principal kind of
hardening heat treatment of steels is their quench-
HEAT TRANSPORT, thermal transport, heat ing with subsequent tempering, and quenching
exchange for the next aging. Dehardening can involve var-
A process of transport of energy in the form ious kinds of annealing applied to remove the
of heat through a medium with a nonuniform tem- consequences of cold plastic deformation (strain
perature distribution, or other nonuniformities of hardening).
physical fields. Heat transport in solids is carried To improve the technological properties of
out by the heat propagation inside a body (see metals, special kinds of heat treatment are used.
HEAVY FERMIONS 555
Classified among them are spheroidizing, graphi- CeCu6 , etc.) the peak in g(E) is the many-particle
tizing, and homogenizing annealings, patenting, Abrikosov–Suhl resonance generated at EF due
normalizing, and full graphitization (ferritization) to the antiferromagnetic exchange interaction be-
of cast iron. Particular kinds of heat treatment in- tween the conduction electrons and the localized
clude decarburizing annealing carried out either magnetic moments of f -ions that make up the reg-
in air or in a hydrogen atmosphere, and oxida- ular lattice. By analogy with a separate magnetic
tion in order to form an oxide film at the surface. impurity in a metal (Kondo impurity), the latter
The oxidizing of iron-based alloys combined with lattice consisting of localized spins that interact
tempering and burnishing, improves the estheti- with conduction electrons is given the name Kondo
cal appearance of the metal. Combinations of heat lattice. The properties exhibited by heavy-fermion
treatment with other methods of material modifi- systems are unusual for metals (see Table 1): the
cation are widely used: combined heat treatment, enormous value of the electronic speci c heat co-
thermomechanical treatment, chemical heat treat- efficient, the enhanced Pauli paramagnetism, high
ment, and so on. The term heat treatment is applied value of the thermoelectric coefficient, the appre-
to practically all metal alloys when they undergo ciable contribution of electron–electron scattering
phase and structural transformations during heat- to the resistivity, extremely low value of the Fermi
ing and cooling. Related research approaches and velocity (∼105 –106 cm/s).
investigations in the field of practical applications Of chief interest are superconducting heavy
are the subject of such sciences as metal science, fermion systems represented by CeCu2 Si2 , UBe13 ,
the physics of metals, the heat treatment and tech-
UPt3 , URu2 Si2 . Although exhibiting low val-
nology of metals.
ues of the critical temperature (Tc = 0.5–0.9 K),
See also Mechanical heat treatment of metals,
these superconductors feature extremely high up-
Chemical heat treatment, Combined heat treat-
per critical elds Bc2 , extremely high values of the
ment, Cyclic heat treatment, Rapid heat treatment,
dBc2 /dT derivative (at T → Tc ) (see Table 1),
Rapid electric heat treatment.
and are also highly sensitive to the presence of
HEAVY FERMIONS in metals both magnetic and nonmagnetic impurities. The
Charge carriers of half-integer spin in electronic heat capacity, sound absorption, spin–
solids of abnormally large effective mass lattice relaxation time, and other characteristics of
m∗ = (102 –103 )m0 . Heavy fermions exhibit a superconducting heavy-fermion systems behave at
narrow, high amplitude peak in the density of elec- T = Tc as power law functions of temperature
tron states g(E) in the neighborhood of the Fermi rather than as exponential functions. This sug-
level EF in a number of metallic compounds based gests that the superconducting gap Δ(k) goes to
on rare-earth elements (Ce, Eu, Sm, Tm, Yb) and zero for certain directions of the wave vector k.
the actinides (U). Compounds of this type are re- These anomalous properties of superconducting
ferred to as heavy fermion systems. The nature heavy-fermion systems are indicative of the pecu-
of heavy fermions is still not completely under- liar nature of their superconductivity. In particular,
stood. It is generally assumed that for the major- ordinary BCS-pairing of two electrons with oppo-
ity of heavy-fermion systems (CeAl3 , CeCu2 Si2 , site spins (see Bardeen–Cooper–Schrieffer theory)
Table 1. Characteristics of heavy fermion superconducting systems, and their comparison with super-
conducting tin
dBc2
Compound γ (T → 0), mJ·mole−1 ·K−1 Tc , K Bc2 (0), kG − (T = Tc ), kG/K
dT
CeCu2 Si2 1050 0.4–0.6 10–18 100–250
UBe13 1100 0.86 150 250–400
UPt3 450 0.54 16 63
Sn 1.78 3.733 0.3 0.14
556 HECTO. . .
in heavy-fermion systems may turn out to be less nonpolar nature of the nematic; e.g., two hedge-
efficient than p-state or d-state pairing, which hogs with equal charges |N| = 1 may merge either
gives rise to a Cooper pair with a nonzero orbital to annihilate (|N| = 0) or to form a hedgehog
moment l = 0 exhibiting singlet (s = 0) or triplet with |N| = 2. This depends on the merging path
(s = 1) types of superconductivity (see Triplet pair- and on the presence of linear disclinations in the
ing). Table 1 compares the superconducting char- sample: in tracing around them, the director orien-
acteristics of three heavy-fermion systems with tation changes by 180◦ . In nematics and smectics
those of the “normal” superconductor Sn. A hedgehogs similar to those sketched in the fig-
ure have been observed experimentally.
HECTO. . .
A decimal multiple prefix, now rather anti- HEISENBERG HAMILTONIAN
quated, signifying a 102 times increase of the A model spin Hamiltonian describing the ex-
initial unit of a physical value. For example, 1 hW change interaction in insulators containing atoms
(hectowatt) = 102 W. or ions with magnetic moments
HEDGEHOG H = −J
Sm Sn ,
A point defect in a vector-type field describing (m,n)
the long-range orientational order of a condensed
medium. Hedgehogs are generated in isotropic fer- where α (α = x, y, z) is the operator of αth pro-
Sm
romagnets in the magnetization vector field, as jection of the spin at the mth site of a crystal
well as in nematic liquid crystals, and in type A lattice; J is the exchange integral for the atoms
smectic liquid crystals in the director field. Elastic at sites m and n. The Heisenberg Hamiltonian
distortions due to a hedgehog can spread over the is a many-particle generalization of the Heitler–
entire bulk of the medium. The simplest hedge- London approximation for a hydrogen molecule.
hogs are a radial hedgehog and a hyperbolic The standard approximation is to take into account
hedgehog where every possible spatial orientation only the interactions of nearest neighbors, and to
of the vector occurs only once (see the figure); this assume that they have a constant exchange inte-
allows a topological invariant (charge) N equal to gral Jmn = J , so the summation is taken over pairs
unity to be correlated with each of them. When of nearest neighbors. For J > 0 this Hamiltonian
the direction of the arrows is reversed, N takes on describes a ferromagnetic material (see Ferromag-
the value of −1. In ferromagnets the coalescence netism) whose total spin in the ground quantum
and decay of hedgehogs are describable as sim- state has the maximum value NS (N is the number
ple addition of the charges N . For nematics the of atoms in lattice, S is the spin of an individ-
result of hedgehog merging and decay is ambigu- ual atom), i.e. classically all the spins are parallel.
ous, since the sign of N is indeterminate due to the For J < 0 the Heisenberg Hamiltonian describes
an antiferromagnetic material (see Antiferromag-
netism) with minimal total spin in the ground state
(0 or S, depending whether N is even or odd), for
which the wave function and the ground state en-
ergy are known only approximately (exactly in the
one-dimensional case for spin S = 1/2). Since the
exchange interaction involved in the Heisenberg
Hamiltonian is isotropic, the total spin of the sys-
tem is conserved, and the stationary states can be
classified in accordance with the projection of the
total spin along a coordinate axis. A state with a
single spin deviation from its ground state consti-
tutes an elementary excitation of a ferromagnet,
i.e. a magnon, which can propagate along the spin
Radial (a) and hyperbolic (b) hedgehogs. system, and states with several spin deviations are
HELICAL DISLOCATION 557
spin complexes which describe excited states of in solid state physics to study impurities and ex-
magnons. citons in molecular crystals. In a wider sense
this approximation provides wave functions of the
HEISENBERG MAGNET crystal in the form of a product of the wave func-
A crystal with magnetic ordering described tions of all the molecules in the system, and in a
by the Heisenberg Hamiltonian. In the case of a limited sense it is associated with neglecting the
positive exchange integral J > 1 (see Exchange contributions to the energy of crystal states involv-
interaction), a Heisenberg magnet is a ferromag- ing two or three excited molecules. This means
net, otherwise (negative exchange integral, J < 1) that in the following Hamiltonian for low-density
it is an antiferromagnet. If the signs of the ex- Frenkel excitons (<1017 cm−3 ):
change integrals for the nearest and next to nearest
= + f +
neighbors differ, the magnetic structure is more H Ef Bnf Bnf + Mnm Bmf Bnf
complicated, for example, helical (see Modulated n,f nf,mf
magnetic structures). The isotropic character of 1 f + +
the exchange interaction affects the energy spec- + Mnm Bnf Bmf + Bnf Bmf ,
2
trum and macroscopic properties. The elementary nf,mf
excitations, i.e. magnons, in a Heisenberg magnet +
the third term is neglected. Here Bnf and Bnf are
are Goldstone modes (zero-gap). Hence, the low-
the creation and annihilation operators of the ex-
temperature additives to thermodynamic values
citations of an f -state molecule at lattice site n;
are temperature-dependent in the form of a power
Ef is the energy of the excited molecule with-
(but not exponential) law. There is no long range
out accounting for the transfer of the excitation to
ordering at finite temperatures in low-dimensional f
(one- and two-dimensional) Heisenberg magnets other sites; and Mnm is the matrix element of the
due to the formation of topological defects. In resonance excitation transfer. The Heitler–London
three dimensions there is a finite critical temper- approximation is applicable when the width Δ
ature (Curie point or Néel point) with a magnitude of an exciton band is small in comparison with
determined by the exchange integral. energy Ef . For some molecular crystals this con-
dition is satisfied (e.g., for the lowest singlet ex-
HEITLER–LONDON APPROXIMATION
citations of anthracene and naphthalene Δ ≈ 500
(W. Heitler, F. London, 1927)
and 150 cm−1 , respectively, and Ef ≈ 25,100
A method of constructing the wave functions and 31,500 cm−1 ).
of the electrons of a molecule from products of
wave functions of electrons of individual atoms, HELICAL DISLOCATION
taking into account the identity of the electrons. It Dislocation lines of a specific configuration in
was first applied for the calculation of the energy the form of a coil arising from the creeping of
spectrum of the hydrogen molecule H2 . According a dislocation segment (see Dislocations) of the
to the principle of identity, the coordinate part of Burgers kind, or of a mixed kind due to the absorp-
the wave functions of H2 in the zeroth-order ap- tion or emission of vacancies. Helical dislocations
proximation of perturbation theory can be either were often observed in experiments involving
a symmetric (singlet) state, or an antisymmetric the high oversaturation of intrinsic point defects
(triplet) state, with respect to the permutation of emerging at high temperatures during quenching,
electrons: or when subject to reactor irradiation (see Radi-
ation physics). The figure shows an example of
Ψs = αs ψA (r 1 )ψB (r 2 ) + ψA (r 2 )ψB (r 1 ) , the formation of a helical dislocation. There exist
right- and left-screw types differing in the direc-
Ψt = αt ψA (r 1 )ψB (r 2 ) − ψA (r 2 )ψB (r 1 ) ,
tion of the winding of the coil. In a crystal over-
where A(B) (r j ) is the wave function of elec- saturated by intrinsic point defects, the process
tron r j ; and αs,t is the normalization coefficient. of formation of the helicoids can closely corre-
The Heitler–London approximation is employed late with the nucleation and the decay of prismatic
558 HELICAL MAGNETIC STRUCTURE
In typical metals this condition can be satisfied (see Polymorphism). Solid helium has a low den-
at a frequency of tens of megahertz, and in fields sity (up to 0.19 g/cm3 ), high plasticity (up to
up to 10 T. The wave arises due to the electron– 3.5·10−8 Pa−1 ); yield limit at shear deformations
phonon interaction. Owing to a helicon–phonon of about 103 Pa; the Debye temperature is very
resonance, the spectrum of collective vibrations low (up to θ D = 25 K). It is the best known repre-
becomes reconstructed: with increasing magni- sentative of a quantum crystal.
tude of the wave vector a phonon branch of the Helium is used for reaching low and superlow
spectrum transforms into a helicon branch and a temperatures (4.2 to 10−3 K), for thermometry in
helicon branch transforms into a phonon one. the range of temperatures 1 to 80 K, for providing
a protective inert medium at melting, cutting and
HELIUM, He welding of metals and alloys, growing of semicon-
A chemical element of Group VIII of the pe- ductor monocrystals, etc.
riodic table with atomic number 2 and atomic
mass 4.002602. Electron configuration is 1s 2 . Ion- HELIUM-3, SUPERFLUID
ization energy is 24.59 eV (the highest of all See Super uid phases of 3 He.
elements). Radius of helium atom 0.122 nm. The
element is chemically inert. Natural helium has HELMHOLTZ FREE ENERGY
stable isotopes 4 He (99.999862%) and 3 He. See Free energy.
Helium is a light colorless monatomic gas; its
density is 0.1785 kg/m3 (at 273 K and 1.013·105 HERZBERG–TELLER INTERACTION
Pa); it is barely soluble in water. Heat conductiv- (G. Herzberg, E. Teller, 1933)
ity is 0.1438 W·m−1 ·K−1 (at 273 K), viscosity The part of the interaction between the elec-
is 18.60 μPa·s (at 273 K). Dielectric constant is tron motion and the nuclear motion in polyatomic
1.000074 (at 273 K and 1 atm pressure). 4 He is systems, which is treated by the separation of
weakly diamagnetic, χ = −0.78·10−13 m3 ·kg−1 . the electron and the nuclear variables in an adia-
Index of refraction for the yellow Na line nD = batic wave function (see Adiabatic approximation).
1.000034; T boiling = 4.22 K, the lowest for all The presence of a Herzberg–Teller interaction ex-
liquids. Liquid helium is a transparent colorless plains (i) a possibility of electronic-vibrational
liquid. Helium isotopes remain liquid at the pres- transitions involving odd vibration sublevels of an
sure of saturated vapors up to arbitrarily low tem- excited electronic state in the case of non-totally-
peratures (see Low-temperature physics), which is symmetric vibrations; (ii) the lack of reflection
associated with the weakness of the interaction symmetry in the intensity distribution of conju-
between helium atoms and the high amplitude of gated optical bands of the light absorption and light
their zero-point vibrations. uorescence of impurity molecules; (iii) tempera-
Since at very low temperature the de Broglie ture and other dependences of the optical spectra
wave length of the helium atoms is comparable of crystals (see Optical spectroscopy).
with their interatomic distance, helium is a quan-
tum liquid: more specifically, 3 He forms a Fermi HETERODESMIC STRUCTURES
liquid, and 4 He forms a Bose u id. Liquid helium (fr. Gr. ετ ερoς , other and δεσ μoς , bond)
undergoes a second-order phase transition to the Crystalline structures which contain two or
superfluid state (see Super uidity). The superfluid more types of chemical bonds. Included among
phase of 4 He (called He II) exists at temperatures heterodesmic structures are both monatomic crys-
below the lambda (λ) point 2.172 K, and the su- tals (e.g., graphite) and crystalline compounds,
perfluid phases of 3 He (A–, B–, A3 –3 He) exist for example, various molecular crystals, coordi-
below 2.7·10−3 K. Solutions of 3 He–4 He form nation compounds, polymers. It is specific for
quantum liquids. heterodesmic structures to include portions (lay-
Solid helium is a colorless transparent crys- ers, chains, molecules) in which the atoms are
tal, which exists only at high pressures (P > linked by strong bonding such as a covalent bond,
2.5 MPa). It experiences polymorphic transitions ionic bond, metallic bond; as well as having parts
560 HETERODIFFUSION
connected with each other by weak bonds (van der tice constants and their elastic moduli. Heteroepi-
Waals bond, hydrogen bond). Crystals with a sin- taxy plays the following roles in solid-state-device
gle kind of bonding are referred to as homodesmic technology: design of semiconductor integrated
structures (diamond, silicon, germanium, alkali- circuits on dielectric substrates (heteroepitaxy oc-
halide crystals, metals, etc.). curs for semiconductors on metals and metals on
semiconductors); design of devices on heterojunc-
HETERODIFFUSION, chemical diffusion tions (e.g., lasers) with a double injection of current
A flux of foreign atoms diffusing into a ho- carriers; design of new solids with alternating ul-
mogeneous medium. In this case the diffusion trathin layers made of semiconductor or other
mobility (see Diffusion) of the foreign atoms is materials with various band structures (artificial
determined not only by the parameters of the superlattices). The synthesis of a superlattice can
medium, but also by the interactions of the dif- be a chain of sequentially alternating processes of
fusing atoms which depend on their concentration. a heteroepitaxy.
The heterodiffusion coefficient is a function of
concentration. It is defined in the same way as
HETEROGENEOUS STRUCTURE
the partial diffusion coef cients which arise in
the theory of interdiffusion formulated within the A structure of a heterogeneous material; el-
framework of the thermodynamics of irreversible ements of a heterogeneous structure are distrib-
processes. The driving force of heterodiffusion is uted irregularly (distinguished from homogeneous
the gradient of the chemical potential. Introducing structure). A heterogeneous structure can be ex-
the chemical potential μ as μ = k B T ln(cγ ) (c is emplified by the structure of a system consisting of
the atomic concentration of a component, and γ is parts differing in physical properties or chemical
its thermodynamic activity coefficient) the expres- compositions (phases), a so-called heterophase
sion for the heterodiffusion coefficient takes the structure. The parts are separated by interfaces in-
form: volving a sharp change in material concentration,
crystal lattice parameters, or other characteristics.
∂ ln γ
Dh = D ∗ 1 + , We can distinguish macro- and microheterogene-
∂ ln c ity of the electron distribution in the system. Such
where D ∗ is the self-diffusion coefficient. a structure can be formed through thermal, radi-
ation and other kinds of material treatments. For
example, a single-phase alloy which is thermody-
HETEROEPITAXY
namically unstable at a certain temperature can
Oriented crystallization of a material on the
decay during annealing to phases with different
crystal surface of another material. This phenom-
chemical compositions. In the process of hetero-
enon is widely found in both natural and artificial
crystallization processes. The degree of ordering geneous structure formation, the grain boundaries
of a solid formed as a result of heteroepitaxy can of a host (main) phase as well as the shear lines
vary from a predominant orientation of polycrystal and twinning boundaries (see Twinning of crystals)
grains to a structurally perfect monocrystal. The play a significant role which is connected with
latter case is possible only for crystallogeometric the decrease of the energy of the interfaces due
compatability of a substrate and the growing ma- to the accumulation of impurity atoms. A macro-
terial, at least in the plane of growth. As no two scopic heterogeneity manifests itself, for exam-
materials have precisely the same crystal lattices ple, through various shrinkage cavities, cracks,
the inevitable differences between their parame- martensitic phases, and so on.
ters leads to the appearance of dislocations (see Producing a heterogeneous distribution can
Epitaxial dislocations) which can hinder a lm from provide materials with desired combinations of
attaining a particular thickness. A critical thick- properties (see Alloy heterogenization), such as
ness at which dislocation failure occurs depends composite materials and semiconductor hetero-
on the relative misalignment of the crystal lat- structures.
HETEROJUNCTION LASER 561
HETEROGENEOUS SYSTEM
A system consisting of physically or chemi-
cally heterogeneous (differing) regions, separated
spatially but in contact. In many cases heteroge-
neous systems are of natural origin (for example,
minerals), and synthetic counterparts are widely
used in science and technology. Such a system
may consist of phases differing in their states of
matter, for example, gas–liquid, gas–solid, col-
loidal particles in solution, and so on. Included Sketches of potential energy changes (solid lines) and
among heterogeneous systems are solids having a Fermi energy level position (εF , dashed lines) near the
interface for two types of heterojunctions, namely, a
heterophase structure, for example, ferroelectrics
junction involving a difference in work functions (a) and
and magnetic substances with a magnetic do- a junction involving surface states (b). The energies of
main structure; metal–semiconductor systems (see the top of the valence band εv , the bottom of the con-
Metal–semiconductor junction), metal–insulator– duction band εc , and the band bendings ϕ1 and ϕ2 at the
semiconductor, etc. A system consisting of two interfaces are indicated.
or more contacting semiconductors of different
chemical compositions corresponds to a solid het-
erogeneous structure. Due to a difference between current passes through a heterojunction as a re-
polarity and between energy band widths, het- sult of the difference of the potential barriers for
erojunctions can form in such structures (e.g., electrons and holes, producing a current–voltage
n–p and multicomponent heterojunctions); quan- characteristic which is sharply asymmetric. Dif-
tum wells and more complicated multiwell struc- ferent heights of barriers for holes and electrons
tures based on them; superlattices consisting of allow one to obtain either hole or electron cur-
alternating layers of different chemical composi- rents. Under application of an alternating electric
tion. Heterostructures and heterojunctions based field the heterojunction is characterized by a paral-
on them are particularly important components of lel combination of resistive and reactive (as a rule,
diverse devices of microelectronics. capacitive) impedances whose magnitude depends
on the applied constant electric field. The compli-
HETEROJUNCTION cated energy structure of a heterojunction provides
A contact of two semiconductors differing a wide spectrum of applications in semiconductor
in chemical composition and physical proper- electronics for lasers, light emitting diodes, pho-
ties. Depending on the type of conduction of the todetectors, etc.
semiconductors in contact, there are distinguished
anisotropic heterojunctions (different types of con- HETEROJUNCTION, ISOPERIODIC
duction) and isotropic heterojunctions (similar See Isoperiodic heterogeneous systems.
type of conduction) (see Inversion layer, Quan-
tum well). Combinations of various heterojunc- HETEROJUNCTION LASER
tions form a heterostructure. At the formation of A semiconductor laser (the most common in-
a heterojunction, due to a difference between work jection laser) based on a heterostructure that in-
functions of the semiconductors in contact (see corporates one or more heterojunctions adjacent
Fig. (a)) or due to the presence of a charge in the to an active region. Potential barriers limit the
surface states (see Fig. (b)), band bending ϕ1 and region of recombination of excess current car-
ϕ2 as well as potential barriers for electrons and riers (active region), preventing their migration
holes arise near the interfaces. to neighboring regions (electronic confinement).
Due to differing electron affinity, jumps in On the other hand, optimization of optical prop-
energy of conduction band and valence band erties can be achieved by preparing a dielectric
of semiconductors,
εc and
εv , respectively, waveguide (optical confinement). A comparatively
occur. Under constant applied voltage, a direct low generation threshold (below 100–200 A/cm2 )
562 HETEROJUNCTION LASER
and high efficiency (up to 50%) at room tempera- sorption (out of the contact area). Heterostruc-
ture are achievable. A basic model of a heterolaser tures have been designed with a refractive index
is a planar heterostructure with an active (narrow- gradient both in side directions and in vertical
band) layer sandwiched between two passive emit- cross-section, corresponding to a two-dimensional
ter (wider-band) layers. Optimized multilayer het- dielectric waveguide (refractive confinement). The
erostructures with gradient or stepwise waveguide figure provides a microphotograph of an over-
which includes an active layer are also employed. grown heterostructure based on InGaAsP/InP for
Strip-geometry (V-groove) heterolasers with a 1.3 μm wavelength heterolaser. The threshold
a contact in the form of a strip at the planar current in such heterolasers can be reduced to 5–
heterostructure serves as a waveguide configu- 10 mA at room temperature.
ration which provides the proper distribution of An important class is represented by the quan-
optical amplification (under the contact) and ab- tum-well heterolasers in which the active layer (or
HETEROSTRUCTURE 563
active few layers) thickness is comparable with or Characteristics of the motion of charge carri-
less than the de Broglie wavelength of the current ers in heterostructures are determined by relative
carriers. In such a quantum well, the quantization energy positions of band extrema (i.e. by the type
of levels takes place with a substantial modifica- of heterojunction) and by the geometry of these
tion of the energy spectrum of the active region, structures. In structures based on GaAs–AlAs,
and with the appearance of new selection rules GaAs–GaP and some others the gap of a narrow-
for the optical transitions. In addition, in such het- gap semiconductor is situated on the background
erolasers the active region volume is quite small, of a wide-band material. This band diagram (i.e.
close to 10−12 cm3 , which allows one to signif- a dependence of band extremum positions on a
icantly reduce the threshold current. On the basis coordinate) corresponds to a heterostructure of
of these structures not only single-strip lasers, but the first kind. Heterostructures of the second kind
also multiple-element lasers with more than 1 W (e.g., InP–Al0.48 In0.52 As) correspond to a band
continuous radiation have been designed. diagram where the breaks of c- and v-bands at
The initial developments involving physics and the heterojunction are of the same sign. The case
engineering of heterolasers were concerned exclu- of non-overlapping band gaps (e.g., in the sys-
sively with the heterosystem GaAlAs/GaAs. To tem InAs–GaAs) is referred to as a heterostructure
expand the wavelength range improved heterosys- of the third kind. More complicated variants of
tems became necessary, and these are mainly qua- band diagrams occur for heterostructures formed
ternary ones such as InGaAsP/GaAs, InGaAsP/InP, by materials of different symmetry (e.g., hetero-
and GaAlSbAs/GaSb. Some characteristics of het- junctions with band inversion, or those between a
erolasers based on various materials are given in gapless semiconductor and an ordinary semicon-
Table 1. In their manufacturing various methods ductor), or in the case of the structures formed by
of epitaxy are used. semimagnetic semiconductors (spin heterostruc-
tures) or materials in which the band extrema
HETEROSTRUCTURE
are situated at different points of a Brillouin zone
A semiconductor structure formed by a few
(different-valley heterostructures); various combi-
heterojunctions. The geometry of these junctions
is established during the growth of a heterostruc- nations of these above variations also occur.
ture in contrast to heterogeneous systems for To account for the change in the geometry
which the shape of heteroboundaries is deter- of structures various cases of low-dimensional
mined by the conditions of phase equilibrium, electron states as well as tunnel processes in het-
or by the random conditions existing during pro- erostructures (see Tunneling phenomena) became
duction. Several parameters of the component available. In Fig. 1 the band diagrams, positions
materials (e.g., positions of the band extrema, ef- of the energy levels, and the wave functions are
fective masses, dielectric constants, etc.) provide shown for various heterostructure types: (a) sin-
possibilities for controlling the properties of the gle quantum well, (b) multiple quantum wells,
charge carriers and the propagation of the elec- (c) superlattice, and (d) selective-doped single
tromagnetic waves. The reflection and refraction heterojunction. Two-dimensional electron states
of electromagnetic waves propagating in a het- are formed in narrow-gap materials; and in the
erostructure is accompanied by the appearance case of a superlattice a mini-band energy spec-
of waveguide modes localized in layers with a trum appears due to a tunneling between these
large refractive index. Due to a difference between states. Structures with tunneling-transparent bar-
the crystal lattices forming the heterojunctions the riers (Fig. 2) are used for transverse charge carrier
phonon spectrum of a heterostructure changes, and transfer. At these two-barrier structures separated
local phonon modes can arise which can influence by a narrow quantum well, an increase of the trans-
the nonequilibrium charge carriers. The scattering parency for the energies close to the resonance
of electrons and holes (as well as their distribution level is observed (resonance tunneling effect).
in the structure) can also be controlled by selective In addition to the above-described plane het-
doping when donors and acceptors are introduced erostructures, quantum wires and quantum dots
only in selected regions of the structure. have been intensively investigated (see Fig. 3). In
564 HEUSLER ALLOYS
Fig. 1. Diagrams of (a) single quantum well, (b) multiple quantum wells, (c) superlattice, and (d) doped single hetero-
junction. The first three figures show the potentials (solid lines), energy levels (a, b) and miniband (c), and wave functions
(dotted lines) plotted versus position. The fourth sketch is a plot versus the Fermi energy εF .
remagnetization brought about by uniform rota- drift space in the device, this leads to a very large
tion of the spontaneous magnetization (see Mag- drift velocity, up to 4·107 cm/s, a shortening of the
netization rotation process). The ordered alloys delay time to 1 ps, and a raising of the frequency
Pt–Co, Pt–Fe, Mn–Al–C are the other class of limit to hundreds of GHz. This two-dimensional
high-coercivity alloys, in which the magnetic re- electron gas exhibits the quantum Hall effect and
versal processes are difficult, mainly due to pin- the Shubnikov–de Haas effect. In metrology this
ning of the magnetic domain walls at the anti-phase makes it possible to measure the fine structure
boundaries of the ordering domains. The proper- constant, and to define the international ohm stan-
ties of many alloys are given in Table 1 in Hard dard.
magnetic materials.
HIGH-ENERGY CHEMISTRY
HIGH-DENSITY EXCITONS The branch of physical chemistry dealing with
Situation arising at high levels of optical radia- high temperature activation processes caused by
tion in crystals when interactions between excitons electronic, atomic, molecular, ionic, neutron, pho-
become appreciable. This high exciton density ton impacts, absorption of photons and phonons,
favors the formation of exciton molecules (biex- application of electric and magnetic fields, me-
citons), and the appearance of coherent nonlinear chanical stress, shear, shock. A common fea-
phenomena and new luminescence bands. At den- ture of these processes is that the energy which
3 1 (here a is the exciton radius)
sities nex aex is absorbed from the external source creates in
ex
Wannier–Mott excitons metallize and transform the medium nonequilibrium concentrations of in-
into an electron–hole plasma. The presence of termediate highly reactive particles and quasi-
this plasma at low temperatures in crystals like particles: holes and conduction electrons, local-
Ge and Si with a complex electron band struc- ized holes and electrons, ions, ionic clusters,
ture leads to the formation of an electron–hole excitons and other excited states, free radicals,
liquid and electron–hole drops. It was found exper- etc. At the instant of formation these particles
imentally that crystalline Si exhibits the presence are very far from thermal equilibrium with the
and inter-transformation of excitons, biexcitons medium, and in a number of cases they take part
and electron–hole drops. The quantum statistics in chemical reactions until they slow down, cool
of excitons and properties close to Bose–Einstein down, and approach thermal equilibrium. Thus
condensation are found in the crystal Cu2 O. high-energy chemistry involves the study of ther-
mal transformations in chemical systems of all
HIGH ELECTRON MOBILITY TRANSISTOR types in the absence of thermodynamic equilib-
(HEMT) rium. Within this scope the following branches
A eld-effect transistor whose channel con- are included: photochemistry, laser chemistry, ra-
tains a heterojunction interface between wide- and diation chemistry, plasma chemistry, chemistry
narrow-gap semiconductors. In a narrow potential of shock waves, as well as chemistry of sound
well (∼10 nm) formed at the heterojunction inter- (phonons), chemistry of infrared and extra-high
face, both the quasi-momentum and the electron frequency radiations (at high densities of corre-
energy are quantized, and there appears a finite sponding actions), chemistry resulting from nu-
density of states, independent of the energy, which clear transformations, chemistry of hot atoms,
in combination with selective (modulated) doping chemistry of “exotic” atoms (positronium, muo-
leads to the suppression of scattering mechanisms, nium, etc.), chemistry of processes of dissolving
and a sharp rise of the mobility of current carri- defect-containing substances, chemistry of mole-
ers. The mobility growth is promoted by a thin cular beams, photoelectrochemistry, emission of
undoped layer formed at the heterojunction inter- ions and electrons, electric discharge in condensed
face. The current carriers in the channel form a medium, etc. The main tasks of high-energy chem-
two-dimensional electron gas with the carrier mo- istry are: development of methods of production of
bility reaching 2·106 cm2 ·V−1 ·s−1 at 4 K, and new substances and materials including those re-
5·105 cm2 ·V−1 ·s−1 at 77 K. In the case of a small sistant to superthermal activation of various kinds,
HIGH-FREQUENCY SIZE EFFECTS 567
determination of the resistance of substances and AC current density j (ω) and the strength of the
materials towards various types of corrosive ac- applied AC electric field E(ω):
tions, the study of the mechanisms of chemical
j (ω) = σ (ω)E(ω). (1)
processes, of the composition and physicochem-
ical properties of intermediate particles and quasi- Isotropic conductors (metals, semiconductors)
particles, etc. usually have their high frequency conductivity
quite well described by the classical Drude for-
HIGH-ENERGY ELECTRONICS of the solid state mula (P. Drude, 1900):
A discipline at the border of solid-state physics 1 + iωτ
and electronics, studying the properties of solids σ (ω) = σ0 , (2)
1 + ω2 τ 2
upon the passage of high density electron beams
through them. Such investigations underwent in- where σ0 = e2 nτ/m∗ is the static electrical con-
tensive development during the 1970s after the ductivity, n and m∗ are the concentration and the
introduction of high-current pulsed electron beam effective mass of free current carriers (conduction
accelerators. These were based on explosive elec- electrons, holes), and τ is the characteristic time
tron emission, with a pulse period up to 1 ns, of scattering of carriers by each other, acoustic
and with electron energies from 0.3 to 0.5 MeV, phonons, crystal lattice defects, impurity atoms,
regulated by a current I up to 5,000 A with a den- etc.
sity up to 2,000 A/cm2 . These accelerators had Eq. (1) holds in a local approximation, pro-
been developed in Russia and the USA. Upon ir- vided the characteristic distance over which the
radiation by strong-current beams and high-energy amplitude of the external electric field E changes
hole plasmas with densities of about 1013 to (e.g., the depth of the field penetration into the
metal through the skin-effect) is much larger than
1014 cm−3 some specific new effects appeared.
the effective mean free path, leff = |l/(1 − iωt)|,
One such effect is a new type of luminescence,
where l = vτ and v is the average carrier velocity.
caused by band-to-band transitions of electrons
and holes. This emission has a wide spectral band
HIGH-FREQUENCY SIZE EFFECTS
and a short relaxation time, much less than a
Dependence of the impedance and/or the trans-
picosecond. The electrical conductivity of the ir-
parency of an electrically conducting medium
radiated body even in a weak external electric
(metal or semiconductor) on its dimensions (see
field is determined by the strongly excited con-
Size effects). During bilateral antisymmetric ex-
duction electrons and holes with energies of about
citation of electromagnetic waves under the con-
an electron volt beyond the edges of the cor-
ditions of the normal skin-effect (the skin depth
responding bands. The specific resistance of an
δ exceeds the electron mean free path, l), the
insulator penetrated by the beam drops by 1011 –
impedance is at its maximum when δ equals a
1015 orders of magnitude, and can reach values
half of the plate thickness, d, i.e. the entire sam-
of about 10 ·cm, and a current density of thou-
ple volume placed in the resonant cavity takes
sands of A/cm2 . An electron–hole plasma, which part in lowering the cavity quality factor Q. For
is formed by the pulsed strong-current irradia- δ > d/2, the intensities of the waves reflected from
tion, greatly influences the brittle failure of crystals. and passing through the plate fall off with de-
Pulsed accelerators can be used to investigate the creasing frequency ω due to the superposition
fine details of the process of Frenkel defect forma- of waves inside the plate. The Fischer–Kao size
tion via the excitation of the electronic subsystem effect (H. Fisher, Y.-H. Kao, 1969), which pro-
in non-metallic crystals (see Pulse spectroscopy). vides a direct determination of the skin depth
by a radio-frequency measurement, has been ob-
HIGH-FREQUENCY CONDUCTANCE served in semiconductors. In metals this effect
A complex kinetic coef cient σ (ω) in the gen- takes place even at room temperatures up to fre-
eralized Ohm’s law that is frequency (ω) depen- quencies of ω 1014 Hz. It may be employed
dent and defines the linear relation between the to control the state of the specimen surface, and
568 HIGH-FREQUENCY SIZE EFFECTS
eventually collide with the layer surface. Mean- conductors electrons form drops isolated from one
while, resonance effects are only possible when another. Within each drop the conductivity has a
there is almost specular reflection from that sur- metallic nature, but the drops occupy only a small
face. Experimental studies of these effects make it fraction of the volume. They are separated from
possible to directly observe the Andreev reflection one another by potential barriers with height γ and
of electrons. thickness R, and at not very low temperatures con-
duction takes place by the relegation of electrons
HIGHLY-DOPED SEMICONDUCTORS to the processes of percolation (see Percolation
Semiconductors whose concentration of impu- theory), i.e. it has an activated nature. The tran-
rities N is so high that weak compensation causes sition from metallic conductivity to the activated
electronic states at the Fermi level to become de- variety with the increasing extent of compensation
localized, and the low-temperature electrical con- (Anderson transition) in highly-doped semicon-
ductivity assumes a metallic, non-activated nature. ductors has been experimentally observed in n-Ge,
Experiments show that the transition from acti- GaAs, CdTe and n-InSb. Such a transition may
vated conductivity to the metallic type with the occur under the effect of a magnetic field, be-
growth of N (Mott metal–insulator transition) takes cause the magnetic field, by restricting the motion
place at the concentrations that fulfill the condi- of electrons in directions transverse to the field
tion NaB 3 ≈ 0.02, where a is the Bohr radius direction, helps to bring about their localization
B
of the impurity atom. The theory of highly-doped (see Anderson localization). The limiting case of
semiconductors usually rests on the assumption strongly-compensated highly-doped semiconduc-
that the inequality NaB 3
1 holds. In this case, tors is sometimes investigated theoretically for
in the weakly-compensated highly-doped semi- the purpose of modeling properties of amorphous
conductors the electrons are degenerate, and the semiconductors.
structure of the density of states is described by
the linear theory of screening (see Electric charge HIGH MELTING TEMPERATURE MATERIALS
screening). For this situation the frequency depen- See Refractory materials.
dence of the band-to-band light adsorption coeffi-
cient at frequencies below the threshold provides
HIGH MOLECULAR WEIGHT COMPOUNDS
the density of states of minority carriers. With the
growth of compensation (or temperature), when Compounds containing in a single molecule
the Fermi level lies deep in the band gap, the di- a large number (from several thousands to many
rect association of the frequency dependence of millions) of chemically bound atoms, e.g., pro-
the absorption coefficient with the density of states teins, starch, polyethylene, and other polymers. In
vanishes. the case of flexible-chain polymers the location of
A special class arises with strongly-compens- the boundary between the low molecular weight
ated highly-doped semiconductors, where the elec- and the high molecular weight categories depends
tron density at sufficiently low temperatures ap- on the exibility of the polymer chain. In the case of
pears to be highly nonuniform in space. The rigid-chain molecules forming three-dimensional
screening of the potential associated with fluctu- networks the transition from the low to the high
ations of impurity concentrations in such a system molecular weight categories is blurred.
is nonlinear. The greatest spatial R and energy γ
scales of fluctuations, which are not screened by HIGH MOLECULAR WEIGHT GLASSES
the electrons, are equal in order of magnitude to Amorphous polymers in the solid state. The
1/3 2/3
the values R = Nt /n2/3 , γ = e2 Nt /(εn1/3 ), properties of high molecular weight glasses (e.g.,
respectively. Here Nt is the overall concentration speci c heat) depend on the temperature of the
of impurities, n is the difference in concentra- transition to the vitreous state of matter (the tem-
tion of basic and compensating impurities, ε is perature of vitri ca tion), and on the chacteristics
the dielectric constant of the system. In such semi- of the vitrification process. In particular they de-
HIGH-PRESSURE CELLS 571
pend on the rate of cooling during vitrification, the in them, which provides high-permeability mag-
type of solvent, and the velocity of its evaporation netic materials with μa and μmax within the limits
upon the formation of the polymer glass from the 2,000 to 5,000, and 20,000 to 300,000, respec-
solution. Concerning mechanical effects, the tran- tively. All the Fe–Ni alloys are pliable, they can be
sition to the glass-like state depends on the loading subjected to cold rolling up to micron thicknesses,
speed and on the frequency of vibrations (mechan- to cutting and welding; after thermal treatment in
ical vitrification). High molecular weight glasses the state with maximum permeability they are very
are capable of extensive strain when loaded in the sensitive to mechanical stresses and deformations,
thermal range below vitrification (induced high under the influence of which their magnetic prop-
elasticity). erties sharply deteriorate.
2. Alloys containing 33 to 40% Ni, 25 to 30%
HIGH-PERMEABILITY MAGNETIC MATERIALS Co, additions of: Cr, Si, Mo, Cu, and residual Fe.
Materials which have high values of magnetic With the help of a thermomagnetic treatment they
saturation, and which can be easily remagnetized exhibit a rectangular hysteresis loop with Br /Bs =
(see Magnetic reversal) and demagnetized. This 0.85 to 0.98 and μmax = 4·104 to 8·105 .
is achieved by forming alloy compositions with 3. Alloys based on the Fe–Al system contain-
a highly uniform atomic magnetic structure, and ing about 12% Al have a high hardness; they
close to zero constants of magnetic anisotropy are used for manufacturing magnetic circuits and
(K1 ) and of magnetostriction (λs ). products which are routinely subjected to mechan-
There are several systems of high permeability ical loads.
magnetic materials: 4. Some metallic glasses, obtained by quench-
1. In the Fe–Ni alloy system there are three ing from the liquid state, have high values of
types of compositions, used for production of μa and μmax . For example, Fe–Ni glasses with
high-permeability magnetic materials, called 45% Ni after annealing in a longitudinal mag-
permalloys. At 70–85% of Ni with the help of netic field have μmax = 310,000. Some Co-based
annealing in the range of temperatures 1100–
glasses after tempering in vacuo at 300 ◦ C have
1300 ◦ C in vacuo (or hydrogen) and cooling at
μa = 30,000 and μmax = 370,000.
a regulated rate, a state with K1 and λs , close
to zero, is obtained, in this case with values
of magnetic permeability (initial and maximum) HIGH-PRESSURE CELLS
μa = 2·104 to 2·105 , μmax = 105 to 106 , but Devices for the creation of static pressure in
these alloys have low saturation induction Bs = excess of 100 MPa.
0.5 to 0.8 T. To facilitate the process of thermal High-pressure cells with a gaseous or liquid
treatment of such alloys and obtain certain com- pressure-transmitting medium are capable of pro-
binations of magnetic, electrical and mechanical ducing a hydrostatic pressure of up to 3 GPa.
properties, they are alloyed with the following el- These cells are made in the shape of hollow
ements which influence the ordering processes: thick-walled (often multilayer) cylinders, closed
Mo, Cr, Cu, V and Si. The best properties in this (if the pressure is created by means of a pump)
group are exhibited by supermalloy containing or with a movable piston (if they are driven by
80% Ni, 5% Mo, 15% Fe and with μmax > 105 . a press). A higher (but quasi-hydrostatic) pres-
The alloys with 55 to 68% Ni after thermomag- sure (up to 170 GPa) is produced in solid-state
netic treatment and thermomechanical treatment high-pressure cells, i.e. with a solid pressure-
acquire a magnetic texture with an easy magne- transmitting medium. These cells are press-driven
tization axis, and for magnetizing along this direc- and consist of a container (material: pyrophyl-
tion they have a rectangular magnetic hysteresis lite, talc, lithographic stone) with the material to
loop and μmax > 800,000. The compositions with be compressed inserted into it, and force bearing
45–50% Ni have an increased Bs = 1.3 to 1.5 T. hardened instrumental steel, hard alloy or diamond
During the course of secondary recrystallization at placed around the container. The shape of the
1150 to 1200 ◦ C a crystallographic texture appears force applicators are chosen so that a cavity ca-
572 HIGH PRESSURE MEASUREMENTS
pable of reducing its volume is created around (1012 Pa) (see Units of physical quantities). Their
the container. This cavity may be either initially interrelationships are given below:
closed (high-pressure cells with gliding punch), or
kbar Mbar MPa GPa TPa
with clearances left between power details (high-
kbar 1 10−3 102 10−1 10−4
pressure cells with compressible burr). When the
force applicators come closer together, a part of Mbar 103 1 105 102 10−1
the container’s material is forced out into these MPa 10−2 10−5 1 10−3 10−6
clearances to form the burr locking the cavity, and GPa 10 10−2 103 1 10−3
clamp together the surfaces of power details ad- TPa 104 10 106 103 1
jacent to the cavity. The lateral surfaces of the
HIGH PRESSURES in solid state physics
force applicators are clamped together by press-
Conventional name for a range of pressures at
ing them into steel cartridges. The steel solid-state
which the distance between the atoms in solids
high-pressure cells create a pressure up to 5 GPa,
decreases more through compression than it does
and the diamond ones up to 170 GPa. For synthe-
during cooling of the same body at a normal pres-
sis of diamond, cubic boron nitride and sintering sure. Pressures above 100 MPa are classed as high
of composite materials on their base, the industry pressures.
employs solid-state high-pressure units for pres- In solid state physics the high pressures that
sures 5–8 GPa in the shape of two anvils with are generally employed are of the static type. In
depressions, together with a cylinder with flared this case the value of the attainable pressure is
holes at both ends and two punches (see also High determined by the design characteristics of the
pressures, High pressure measurements). high-pressure cell and chamber. Depending on the
application, the specifications required for partic-
HIGH PRESSURE MEASUREMENTS ular investigations can put limits on the range of
Measurements of pressure carried out with working pressure. For example, a high pressure
the help of various manometers whose operation chamber can be weakened by windows incorpo-
is based on the pressure dependence of many rated for the admission of light, by conductoring
diverse characteristics of matter, viz., thermody- wires inserted for resistance measurements, or by
namic, electric, magnetic, spectral, and so on. the incidence of ionization radiation. For magnetic
Manometers must be calibrated by comparison ei- investigations the chambers should be manufac-
ther with results of absolute measurements using tured from non-magnetic materials; for chambers
piston gauges, or with secondary data recom- working at high or low temperatures appropriate
materials must be selected. The pressure, together
mended by the special group of the International
with the temperature, both enter the main thermo-
Association on Development of Research in the
dynamic relations, and can appear as differentials
field of high pressures (1986). Below 1.4 GPa, the
of energy, entropy and volume. The dependence
pressure measurements can be based on mercury
of the pressure on the volume and temperature
melting temperature data. Manometer calibration
is described by an equation of state. Several ap-
can also be based on characteristics of phase
proximate and empirical equations of state have
transitions at room temperature, calculated data
been proposed which give satisfactory agreement
relating the pressure (in kbar) to the compress- with experimental data, at least for pressures up to
ibility of NaCl, or the spectral shift of the R1 20 GPa.
line in the spectrum of ruby at room temperature: Under compression changes can occur in vari-
2.746
R1 (Å) = P (kbar) (this equation is valid ous interactions, and in particular, interactions be-
up to 200 kbar). The Raman spectrum of ruby is tween electrons of the same configuration, interac-
currently the most popular calibration method. tions with nuclei, as well as inter-configurational
Measurements of high pressures are often re- interactions. The values of these interactions
ported in non-standard units, such as: kilobar and sharply differ for “inner” and “outer” electron
megabar, as well as derived SI units: megapas- shells. For instance, the variation of electrosta-
cal (106 Pa), gigapascal (109 Pa), and terapascal tic interactions of f -electrons can bring about
HIGH-TEMPERATURE ALLOYS 573
a shift of energy levels up to approximately tially full d-shell of the central transition atom
10 cm−1 /GPa, and the shift of outer configuration exceeds that of the crystal field of the ligands.
levels can be an order of magnitude higher. High-spin complexes are formed preferentially
Using such relations and the fact that high pres- with strongly electronegative ligand atoms and
sures can provide a large variation of inter-atomic groups utilizing strongly ionic type bonds, without
distances, it is possible to single out these interac- essential participation of d-orbitals. In high-spin
tions and to acquire some information about them complexes the structure of the d-shell is not dis-
which cannot be obtained by the other methods. turbed by the influence of weak ligand fields, and
Quantitative investigations of this type have been the number of unpaired d-electrons, which has a
carried out using magnetoacoustic resonance, tun- maximal value according to the Hund rule for the
neling spectroscopy, electron paramagnetic reso- main term of the isolated central atom, remains the
nance, nuclear magnetic resonance, etc. It often same in complexes with ligands. Low-spin (spin-
happens that at the attainment of a particular de- paired) complexes are also distinguished; they are
gree of compression qualitative changes in prop- complexes of transition metals, where the max-
erties occur, the characters of interactions change, imum possible spin of unpaired electrons in the
and phase transitions are observed. For exam- unfilled d-shell of the central atom is not realized
ple, metallic cesium undergoes an electronic phase due to the strong interaction with the ligands. This
transition (see Phase transitions under pressure) interaction brings about the splitting of the d-level
in the vicinity of 4.5 GPa; its volume significantly of the central atom into sublevels, the multiplicity
decreases without a change of the close packing of of which is determined by the crystal field sym-
the atoms or the lattice symmetry. Such transitions metry (see Crystal e ld). Differences in the total
are studied by measuring the electric resistance, spin of the ground state, depending on the ligand
because they often involve a metal-to-insulator type, are possible for the electronic configuration
transformation (and also superconductor). Mod- d n of the central atom with n = 4, 5, 6, 7 in
ern experimental installations permit one to make the case of octahedral complexes, and with n =
simultaneous measurements of electric resistance 3, 4, 5, 6 in the case of tetrahedral complexes.
under high pressures (up to 50 GPa) and low tem- In low-spin complexes the energy gain from plac-
peratures (below 1 K). Different spectral investi- ing electrons at the lower energy sublevels exceeds
gations (absorption, luminescence, Raman scat- the loss in exchange energy. These complexes are
tering of light, etc.) performed at low temperature characteristic of the second, and especially of the
(4.2 K using “piston–cylinder”, and 25 K using third series of transition elements; they are formed
“diamond anvil” type chambers) and high pressure preferably with weakly-negative ligands, bound to
(up to 6 GPa in two-step “piston–cylinder”, and up metals by covalent bonds.
to 16 GPa in “diamond anvil” chambers) are de-
voted to the study of structural phase transitions HIGH-TEMPERATURE ALLOYS
and electronic phase transitions. Alloys based on nickel, iron–chromium–nickel,
During magnetic investigations at high pres- cobalt, or a mixed base, known for their high-
sures (up to 1 GPa) and low temperatures (below temperature strength. The main features character-
0.05 K) some rare-earth substances characterized izing high-temperature alloys are high resistance
by magnetic ordering loose this ordering down to creep and fatigue, and long-term strength.
to the lowest temperatures. Investigations of the Deformable and cast high-temperature alloys
Mössbauer effect have shown, in particular, that are distinguished. The heat resistance of the de-
these materials do not change their valence up to formable high-temperature alloys is due to the
very high degrees of compression. At present the structure, where the particles of intermetallic com-
Mössbauer effect can be studied (at room temper- pounds, borides and carbides are evenly distrib-
ature) at pressures up to 70 GPa. uted. Such a structure appears as a result of heat
treatment (homogenizing and aging, see Homoge-
HIGH-SPIN COMPLEXES nization) bringing about the alloy heterogenization
Complexes of transition metals, in which the of the microstructure. Besides, this appears dur-
electrostatic interaction of electrons within a par- ing the course of alloying with refractory chemical
574 HIGH-TEMPERATURE EXPANSION
In 1946 R.A. Ogg Jr. observed a sharp drop mechanism of electron superconductivity involv-
in the resistance of a metal–ammonium solution ing the screened Coulomb interaction in transition
Na:NH3 frozen in liquid nitrogen (T = 77 K). metals and alloys. No account, however, was taken
Based on his experiments, Ogg suggested that of the effects of retardation that develop during the
bound electron pairs with zero spin form as a exchange of acoustic plasmons.
result of the polarization of dipole molecules After the discovery of high temperature super-
of NH3 (prototype of singlet Cooper pairs or conductivity in 1986, numerous new theoretical
bipolarons), and he attributed zero resistance (i.e. concepts appeared to explain this unique phenom-
superconductivity) to the super uidity of such enon. Such concepts may be separated condition-
charged bosons. Ogg’s experimental results, how- ally into the traditional ones that treat this or
ever, were later found to be due to the precipitation that mechanism of Cooper pairing (see Bardeen–
of sodium fibers from the solution. Cooper–Schrieffer theory), and the nontraditional
Almost twenty years later the idea of high- varieties that offer entirely new approaches to
temperature superconductivity appeared in studies the solution of the problem of high-temperature
by W. Little (1964) who suggested its possibility superconductivity. Among the first are the mod-
in organic polymers at Tc ∼ 102 –103 K. Simi- els with strong electron–phonon interaction that
larly, V.L. Ginzburg pointed to a possibility of are based on the Eliashberg equations for the
high-temperature superconductivity at a metal– energy gap in superconductors and the critical
insulator interface or in “sandwich” systems of temperature. Various non-phonon mechanisms of
the insulator–metal–insulator or semiconductor– superconductivity are considered, in particular
those stemming from the exchange of dipole and
metal–semiconductor type. These studies consid-
quadrupole permitted excitons (see Exciton mech-
ered a non-phonon mechanism of Cooper pairing
anism of superconductivity) or of collective exci-
of conduction electrons (see Cooper effect), which
tations of charge density (e.g., of plasmons, see
might occur due to exchange of high-energy exci-
Plasmon mechanism of superconductivity). Such
tations of the Bose type, i.e. excitons (see Exciton
excitations may take place in layered (quasi-two-
mechanism of superconductivity). V.L. Ginzburg
dimensional) metals and in narrow-band metals
and D.A. Kirzhnits wrote a book entitled High
with charge carriers almost localized at lattice
Temperature Superconductivity in 1977. The above
points. In 1962 I.Ya. Privorotsky suggested an-
ideas stimulated experimental research in systems other mechanism of Cooper pairing for the case
of low dimensionality, such as the quasi-one- of triplet p-wave pairing of electrons that could
dimensional crystals (both organic and inorganic) go via the exchange of virtual antiferromagnetic
of the TTF-TCNQ, KPC type, the (SN)x polymer, magnons (see Magnon mechanism of supercon-
Nb(Se, S)3 , or the quasi-two-dimensional crystals ductivity). D. Pines (1990) applied this mechanism
of the type of Nb(Se, S)2 and Ta(Se, S)2 . However, to a modified case of singlet d-wave coupling
the maximum values of Tc ≈ 10 K could only be of holes in CuO2 cuprate layers to explain the
reached in the (ET)2 Cu(NCS)2 compound. nature of high temperature superconductivity. Ac-
In 1967–1968 J. Garland, H. Fröhlich, and cording to a concept suggested by A.A. Abrikosov
E.A. Pashitsky independently suggested a mech- (1993), an important role in Cooper pairing that
anism of high-temperature superconductivity in takes place in high-temperature superconductors
the transition metals and in degenerate semicon- based on layered copper metal-oxide compounds
ductors, semimetals and layered semiconducting is played by the quasi-one-dimensional saddle
structures where there seemed to be a possibility anomalies (lines of saddle points) found in the
of another type of collective boson excitation ex- quasi-two-dimensional dispersion law (band spec-
change of an electronic nature, namely acoustic trum). These are experimentally observable using
plasmons. Such excitations had been predicted angle resolved photoelectron spectroscopy. Such
earlier by D. Pines (1956) for a two-component anomalies are characterized by Van Hove singular-
Fermi liquid with “light” and “heavy” charge car- ities in their density of electron states.
riers (see Plasmon mechanism of superconduc- Other nontraditional theories of high tempera-
tivity). B.T. Geilikman (1966) treated a similar ture superconductivity which have been proposed
HIGH-TEMPERATURE SUPERCONDUCTORS 577
HIGH-TEMPERATURE SUPERCONDUCTORS
Solid-state materials (chemical compounds
Fig. 1.
such as alloys, solid solutions, especially ceram-
ics) with rather high critical temperatures (Tc )
of their superconducting phase transition (as-
sumed Tc > 30 K). The first high-temperature
superconductor was discovered by J.G. Bednorz
and K.-A. Müller (1986) in the form of ceramic
samples of a LaBaCuO metal-oxide compound,
Tc ≈ 30 K. At present, several types of copper-
oxygen containing (cuprate) high-temperature
superconductors are known: La2−x Mx CuO4
(M = Ba, Sr, Ca), the maximum value of the crit-
ical temperature, Tc max ≈ 40 K; YBa2 Cu3 O6+x
(123) and YBa2 Cu4 O7+x (124), Tc max ≈ 90 K;
Bi2 Sr2 Can−1 Cun Ox , Tc max ≈ 110 K (for n = 3);
Tlm Ba2 Can−1 Cun Ox , Tc max ≈ 115 K (for m = 1
and n = 5), or Tc max ≈ 125 K (for m = 2,
n = 3); HgBa2 Can−1 Cun Ox , Tc max ≈ 135 K
(for n = 3) and Tc max ≈ 160 K under pres-
sure (30 GPa). All such compounds feature a
perovskite-type structure (see Perovskites) and
contain from 1 to 5 of the so-called cuprate lay-
ers of CuO2 in their unit cell, the latter serv-
ing as the basic conducting and superconduct-
ing channel in a high-temperature superconductor.
Layers of CuO2 alternate with layers of metal Fig. 2.
ions (Ca2+ , Sr2+ , Y3+ ) and oxide layers (LaO,
BaO, SrO, BiO, TlO, HgO) in their crystal lat-
tice. Figs. 1–3 demonstrate the crystalline struc- of Bi2 Sr2 CuO6 and Tl2 Ba2 CuO6 with a single
tures of certain high-temperature superconductors. cuprate layer, n = 1. Fig. 2(b) shows the struc-
Fig. 1 shows the structure of LaCuO4 (a) and ture of Bi2 Sr2 CaCu2 O8 and Tl2 Ba2 CaCu2 O8
YBa2 Cu3 O7 (b); Fig. 2(a) shows the structure with two cuprate layers, n = 2. Fig. 3(a) shows
578 HIGH-TEMPERATURE SUPERCONDUCTORS
Fig. 3.
Fig. 5.
newly synthesized, are ongoing, and have flour- the single-electron Hamiltonian H from the eigen-
ished during their active development phase. One value equations
should not completely rule out the possibility of
discovering “room temperature” high-temperature H Ci = εi Si Ci , Ci+ SCj = δij ,
superconductors (Tc > 300 K). It is hard even to
grasp the perspectives of practical applications of where S is the matrix of orbital overlap, Ci+ is
such an eventuality. the transposed column-vector Ci ; εi are the en-
ergies of the one-electron levels; above the lowest
HIGH-TEMPERATURE X-RADIOGRAPHY levels of the range the electrons are distributed in
Use of X-ray diffraction to investigate solid- accordance with the Pauli principle. The matrix el-
state physics and material science problems at ements of H are determined by the formulae
high temperatures (500 to 3000 K). One main ap-
plication is the investigation of rst-order phase Hμμ = −Iμ ,
transitions and second-order phase transitions at Hμμ + Hνν
high temperatures, including polymorphism, the Hμν = kSμν , μ = ν,
2
crystal structure of the high-temperature modi-
fications, determination of the phase transition where Iμ is the ionization potential which re-
temperature, finding volume and orientation rela- sponds to removal of an electron from a valence
tions at the transformations, and elucidating the atomic orbital μ, and k is a numerical coefficient
kinetics of different transformations in the solids. (as a rule, k = 1.75). A principal drawback of the
The X-ray diffraction data can provide high- Hoffmann method is its neglect of the Coulomb
temperature phase diagrams of multicomponent interaction between electrons; however, owing to
system equilibria. This technique is also used for its simplicity, it has been used to assess the elec-
studying heat expansion of crystals at high temper- tron density distributions and the spin density in
atures (X-ray dilatometry). Although the accuracy the ground state of molecules and radicals, and to
of X-ray dilatometry is not very high, it has the estimate the energy of conformations, internal ro-
advantage of allowing one to measure the ther- tation barriers, and characteristics of deformation
mal expansion tensor of anisotropic crystals. The vibrations. The method is not satisfactory for de-
X-ray diffraction data can provide estimates of the termining bond lengths.
equilibrium concentration of point defects. Mea-
surements of the temperature dependence of the HOLE
intensity of X-ray diffraction lines, together with A quasi-particle introduced to account for the
heat diffusion dissipation at high temperatures,
cooperative motion of electrons in energy bands
can provide the characteristic Debye temperature.
of solids. Holes are found in the valence band
X-ray diffraction intensity measurements are also
of semiconductors, and in the conduction band
used to study alloy ordering processes.
of hole-type metals and semimetals. The intro-
HILBERT RELAXATION TERM
duction of holes replaces taking into account the
See Landau–Lifshits equation. actual motion of all the band electrons in external
electric (E) and magnetic (B) fields by considera-
HOFFMANN METHOD, extended Hückel method tion of the motion of fictitious particles, i.e. holes,
(R. Hoffmann, 1963; E. Hückel, 1931) each of which has a positive charge e equal in
A semiempirical method, no longer much in magnitude to electron charge; each hole occupies
use, for the quantum mechanical calculation of one of the free (unoccupied) states of the band and
the electron structure of multiatomic molecules. In has the quasi-wave vector k of this state. As this
the Hoffmann method the single particle states of takes place, the following circumstances should
a molecule (molecular orbitals) are determined in be noted: (1) the concentration of holes is equal
the form of linear combinations of valence orbitals to the concentration of free-electron states in the
of component atoms, whose coefficients are com- band, and can be calculated with the help of the
ponents of particular vectors Ci of the matrix of statistical equations which describe these states;
HOLMIUM 581
(i.e. the external field, the “internal” photovoltaic Lorentzian shape and Gaussian shape. The func-
field, or the diffusion field). The lack of a center tion I appears in the first-order statistical theory of
of symmetry in photorefractive crystals determines line shape of magnetic resonance when the field
the possibility of a special nonlocal nonlinear re- components fk , which cause a frequency shift,
sponse providing a stationary amplification of the fall off with distance from the resonance center
weak light wave on account of the diffraction of as r −2 . An example is the electric fields of charged
the strong beam at the emerging hologram (see point defects in crystals, which lead to inhomo-
Real-time holography). geneous broadening of an electron paramagnetic
resonance line. This line shape is rarely used.
HOLOGRAPHY, ACOUSTICAL
See Acoustical holography.
HOMEOTROPY
HOLOGRAPHY, POLARIZATION A homogeneous state of a liquid crystal layer
See Polarization holography. with a director normal to principal surfaces. In this
case the direction of the director is called the opti-
HOLOHEDRIES cal axis of the crystal.
Centrosymmetric lattice point groups of sym-
metry of crystal systems; highest point symme-
HOMOEPITAXY
try group of a crystal family. There is a single
holohedry in one dimension. There are four holo- See Autoepitaxy.
hedries in two dimensions, three of which are
symmetry groups of parallelograms (oblique, rec- HOMOGENEOUS BROADENING
tangular, square), and the fourth of a hexagon This is the broadening of spectral lines result-
(hexagonal). In three dimensions there are seven ing from irreversible phase relaxation, i.e. related
holohedries, six of which are the allowed sym- to the finite lifetimes of particles in various en-
metry groups of parallelepipeds (triclinic, mono- ergy levels. Its effect on general broadening may
clinic, orthorhombic, tetragonal, trigonal, cubic),
be determined in the most direct manner by the
and the seventh is the symmetry group of a hexag-
spin echo technique (in the case of magnetic res-
onal prism (hexagonal). A holohedry can also be
onance), or by using a light echo, since only irre-
defined as a symmetry group of the Wigner–Seitz
versible phase relaxation contributes to decaying
cell. The holohedries of direct and reciprocal lat-
echo signals. While the resonance is saturated by
tices are the same.
a monochromatic variable field, all the portions of
the uniformly broadened line behave identically:
HOLSTEIN–PRIMAKOV REPRESENTATIONS
See Spin operator representations. in contrast to the case of inhomogeneous broaden-
ing, their amplitudes monotonically decrease. The
HOLTSMARK SHAPE, Holtsmarkian
corresponding mechanisms include processes of
(T. Holtsmark, 1919) spin–lattice relaxation and ion–lattice relaxation;
A line shape described by the function relaxation processes due to internal molecular
motion, by atomic diffusion, etc. The inhomo-
∞ geneous contribution is always present during
ρ
IΩ (ω − ω0 ) = (πΩ)−1 cos(ω − ω0 ) dipole–dipole broadening due to the presence of
Ω
0 static local fields that significantly influence the
× exp −ρ 3/2 dρ, situation, particularly in magnetically diluted sys-
tems. See also Inhomogeneous broadening.
where ω0 is the frequency of the resonance line
in the absence of broadening mechanisms, char-
acterized by the parameter Ω. The Holtsmark HOMOGENEOUS STRUCTURE
line shape is an intermediate one between the See Heterogeneous structure.
HOMOLOGY 585
and hence Young’s modulus can be written E = hot atoms. At any temperature hot atoms are used
3K(1 − 2ρ). in the synthesis of labeled compounds, the separa-
tion and enrichment of isotopes, etc.
HOPPING CONDUCTIVITY
A mechanism of conductivity working through HOT ELECTRON EMISSION
localized states. Such a mechanism operates Emission of very fast, hot electrons from a solid
through activated tunneling of charge carriers to its surroundings. The majority of the free elec-
(electrons, holes, polarons) from an occupied state trons in a solid have an energy close to the mean
with energy εi to a vacant state with energy energy. Hot electrons have energies much higher
εj . This mechanism bypasses activation of cur- than the mean; for a conductor this is an energy far
rent carriers into the band of delocalized states. above the Fermi energy, and for a semiconductor
The energy deficit εj − εi is compensated by the it is an energy far in excess of the conduction band
absorption of phonons. It entails a temperature de- energy. The number of electrons whose energy ε
pendence for the activation of jump conduction. is significantly higher than the mean energy at
At moderately low temperatures when “hops” be- thermodynamic equilibrium decays exponentially
tween adjacent levels dominate, the activation en- with their increase in energy, i.e. in proportion to
ergy is constant, in other words, ln ρ ∝ T −1 (ρ is the Boltzmann factor exp[−ε/(k B T )], where T
the resistivity). With decreasing temperature hops is the temperature of the solid (lattice). Among
only take place between levels that are close to these fast electrons there exist some whose energy
each other in energy (although spatially they may exceeds the work function ϕ. A portion of these
be quite far apart). Such a regime is called the electrons (whose velocities are perpendicular to
variable range. Then ln ρ ∝ T −n , 0 < n < 1 (see the surface) contribute to the thermionic emis-
Mott law). In strong electric fields, when the en- sion current, with a magnitude determined by the
ergy accumulated during average hops becomes Richardson equation I = B(T ) exp[−ϕ/(k B T )].
equal to the activation energy in zero field, an The dependence of the preexponential factor
activation-free conductivity develops with a char- B(T ) on the temperature is described by a power
acteristic dependence on the field strength, with law. The presence of hot electrons causes the ther-
ln ρ ∝ E −n . Hopping conductivity is observed in mal emission current to sharply increase. In the
disordered solids, in which the quantum states simplest situation when (owing to the dominant
localized at various points in the system have role of interelectron collisions) the electron gas
different energies. In doped or compensated semi- can be characterized by a certain electron temper-
conductors the value of the hopping conductivity ature for the emission current, we again have the
displays a strong exponential dependence on the Richardson formula, but with the electron tem-
impurity concentration. Calculating the concentra- perature replacing the lattice temperature. In this
tion, temperature, and field strength dependences case the electron temperature, with its dependence
of hopping conductivity may in certain cases be on the heating electric field strength, can be sev-
reduced to calculating the Miller–Abrahams resis- eral times or even an order of magnitude greater
tance network (A. Miller, E. Abrahams) plotted than the lattice temperature, and the current of
over random points (see Percolation theory). hot electron emission will exceed by many orders
of magnitude the thermal emission value. How-
HOT ATOMS ever, in most cases the distribution of fast electrons
Recoil atoms in nuclear transmutations that ac- in a semiconductor differs from the exponential
quire an extra kinetic energy ∼10–100 eV due to (Maxwell) law, and depends on the magnitude of
the emission of neutrons, protons, α or β particles, heating field, the scattering mechanisms, and the
or gamma rays by a radioactive nucleus. These band structure of the given material. Therefore, it
atoms can effectively participate in chemical re- is not possible to write for the hot electron emis-
actions in solids (see Radiation chemistry). The sion current a general formula (like the Richardson
rate of such reactions is independent of tempera- formula) with a single parameter that would be the
ture due to nonequilibrium energy distribution of work function of a given material.
588 HOT ELECTRON GENERATION
Hot electrons can be obtained from semicon- InP) as samples. The spectrum of hot electron
ductor–metal insular lms on an insulator sub- photoluminescence provides a type of time evo-
strate, as well as from “sandwiches” of, e.g., the lution of energy relaxation processes. Excitation
metal–insulator–metal type. of electron–hole pairs by linearly polarized light
The phenomenon of hot electron emission has gives rise to the optical alignment of electrons and
prospects for the production of cold cathodes. holes with respect to velocities, which manifests
The advantage of these in comparison with heated itself in the appearance of polarization in hot elec-
cathodes is obvious. The evaporation of cathode tron photoluminescence. The study of depolariza-
material does not take place at a cold lattice with tion of hot electron photoluminescence in a mag-
hot electrons. In addition, such a cathode is al- netic field allows one to determine characteristic
most noninertial (no time is needed to preheat electron relaxation times (e.g., for the processes
it). Insular metal films are, in addition, promising of emission of longitudinal optical phonons, in-
for electronics from the viewpoint of the con- tervalley transitions). Spectra of hot electron pho-
tinuous trend toward the microminiaturization of toluminescence provide a number of important
electronic devices. parameters of band structure (the width of the va-
lence band). The final stage of energy relaxation
HOT ELECTRON GENERATION involves thermalization, which takes place due to
Almost unbounded increase of the energy of electron–electron collisions: a Fermi–Dirac dis-
conduction electrons in a constant electric field. tribution of electrons of temperature Te becomes
The hot electron generation occurs when the rate established. At low temperatures, Te often no-
at which the energy increases exceeds the rate of ticeably exceeds TL even at moderate levels of
energy dissipation through interaction with the lat- excitation. The magnitude Te − TL is determined
tice. A similar result is predicted by a theory which by the ratio between the pumping power and the
assumes a fixed scattering mechanism whereby, rate of transmission of energy from the electron
e.g., energy and momentum are dispersed through subsystem to the lattice.
the agency of optical phonons or ionized impu-
rities (see Current carrier scattering). In actual HOT ELECTRONS
situations the increase of energy is limited by the Conduction electrons in a solid with an in-
increasing role played by other scattering mecha- creased (owing to excitation from an external
nisms, e.g., intervalley scattering of electrons in- source) mean energy. Any input of energy into the
volving transitions to valleys of higher energy, or electron subsystem (due to electric current, laser
impact ionization. In a single-valley semiconductor irradiation) disturbs the thermodynamic equilib-
(see Many-valley semiconductors), the hot elec- rium between the electron and lattice subsystems.
trons may cause electron gas overheating, with the The electrons become “hot”, i.e. their mean en-
result that the differential conductivity may take ergy exceeds the equilibrium value corresponding
on a negative value (and the current–voltage char- to the lattice temperature. The extent of electron
acteristic may become S- or N-shaped); the latter heating is greater, the higher the input power and
corresponding to current instability due to over- the lower the magnitude of the electron–phonon
heating. interaction. The heating of electrons is easily ob-
served in semiconductors with weak electron–
HOT ELECTRON PHOTOLUMINESCENCE, hot phonon coupling (Ge, Si, InSb, GaAs, and so on).
photoluminescence The materials with p-conductivity allow the for-
Radiation, which is produced at recombina- mation of hot holes.
tion of photoexcited electrons, the kinetic energy The energy distribution of hot electrons can
of which far exceeds kB TL (hence the name hot differ considerably from the equilibrium one:
electrons), where TL is crystal lattice tempera- when heated, the electrons “populate” new parts
ture. Thorough investigations of hot electron pho- of the energy band (see Intervalley redistribution).
toluminescence are performed using direct band Therefore, there appears a number of new phe-
gap semiconductors of the AIII BV group (GaAs, nomena dependent on the band structure, and
HOT MAGNONS 589
crystal lattice. In ferromagnetic semiconductors is more convenient to use the range of the inter-
(e.g., EuO) a situation could occur in which the action region W = 2zt , where z is the number of
conduction electrons scatter strongly off magnons nearest neighbors.
(see Magnetic semiconductors) which are only For U W the Hamiltonian (1) corresponds
weakly coupled to phonons. In this case, the elec- to a Fermi liquid, and in the opposite “tight bind-
tron heating by the external electric field leads ing” limit (U
W ) the Hamiltonian (1) is consis-
to a heating of the magnons due to the trans- tent with a strongly correlated system. In the latter
fer of energy from electrons to magnons. The case it is convenient to operate on the premise of
phonons serve as a heat bath because the lattice the atomic limit, which allows us to discard the ki-
is maintained at a constant temperature. Due to netic (hopping) term t of the Hamiltonian. In this
the weak coupling, the magnon temperature can limit every site has two electron levels, separated
substantially exceed the phonon (lattice) tempera- by U . When the kinetic term is taken into account
ture. Heating the magnons causes a dependence of these levels become smeared and turn into bands,
magnetic characteristics (e.g., magnetization) on so at high values of U the correlation splitting of
the electric field. Heating both the electrons and the interaction zone into two Hubbard subbands of
the magnons can be accomplished with the help one- and two-particle states is to be expected. In
of laser irradiation (see Laser). When several co- the initial approach, the Green’s function splitting
herent light beams are incident on a ferromagnetic (“Hubbard 1 state” approximation) leads to band
semiconductor, superlattices can arise both from formation, which, however, in the limit of low
hot electrons and from the hot magnons (superlat- values of U lacks physical sense. The more so-
tices in concentration, effective temperature, and phisticated (and more realistic) “Hubbard 3 state”
other parameters) due to interference effects. approximation gave correct results: the splitting is
observed when U exceeds a certain critical value
of the order of W . Thus, in the case of half-filling
HOT PHOTOLUMINESCENCE
(number of electrons per site n = 1), the system
See Hot electron photoluminescence.
is a metal for low values of U and an insulator
for high values of U . The metal–insulator phase
HUBBARD MODEL (J. Hubbard, 1963–1965) transition takes place at intermediate values of the
One of the fundamental models which treats a parameter U/W , so there is no small parameter
system of strongly interacting electrons in a crys- in the theory. The region of intermediate values
tal. The one-state Hubbard model is the simplest of this parameter may be handled in two ways:
model of correlated electrons. This model de- the nearly free electron limit (U W ) and the
scribes the motion of electrons on a lattice, taking tight binding limit (U
W ). The crossover of
into account only the local Coulomb interactions these modes can be studied using various methods:
between the electrons at a particular lattice site. analytical (variational method, mean-field theory,
The Hubbard Hamiltonian contains only two etc.) and numerical (exact diagonalization of small
parameters: the hopping amplitude matrix ele- clusters, quantum Monte Carlo method).
ment t , which corresponds to jumping from one Different methods of investigating the inter-
lattice site to a neighboring one, and the Coulomb mediate region sometimes yield erratic results,
repulsion term U at a site: therefore, of prime importance are exact solutions,
which are known for a number of particular cases
+
H = −t Ciσ Cj σ + U ni↑ ni↓ . (1) of the the Hubbard model. Thus, the exact solu-
ij σ i tion of one-dimensional Hubbard model (d = 1)
for all values of U was found by E. Lieb and
+
Here Ciσ (Ciσ ) are Fermi annihilation (creation) F.Y. Wu (1968), and investigations in the limit
operators of an electron at the site i with projec- of infinite spatial dimensionality were begun by
+
tion σ , and niσ = Ciσ Ciσ is the electron number W. Metzner and D. Vollhardt (1989). In the context
operator at the site with a given spin projection of many-particle systems theory, the limit d = ∞
(up ↑ or down ↓). Instead of the parameter t , it corresponds to the mean eld approximation. It has
HUME-ROTHERY PHASES 591
been shown that in the limit of d = ∞ the statisti- from occupying a site which is already occupied
cal mechanics of the Hubbard model is equivalent by another electron of spin σ . The Hamiltonian (2)
to that of a certain auxiliary problem involving describes the electron motion over the lattice for
an impurity center in the Anderson model, which the above condition, and the exchange interaction
has been more widely studied than the Hubbard at the neighboring sites of the antiferromagnetic
model. The evolution of the quasi-particle spec- integral J = 2t 2 /U . When the band becomes half-
trum in the Hubbard model during a variation of filled, the first term in Eq. (2) vanishes, and we
the parameter U is more complicated than that arrive at the Heisenberg model (see Heisenberg
predicted by an approximate approach like the Hamiltonian) of spin S = 1/2 and an antiferro-
“Hubbard 3 state” approximation. Depending on magnetic interaction. If the number of electrons
the U/W ratio, there may occur a one-band and a per lattice site n = 1, then the ground state of the
two-band spectrum, as well as a spectrum with a system is that of an antiferromagnetic insulator.
gap, in which there is a sharp Kondo resonance Deviations of n from unity may result in the gen-
at the Fermi surface (see Kondo effect). Within eration of charge carriers in the system; and these
the context of this approach it is possible to de- charge carriers undergo a strong interaction with
scribe a whole set of interrelated phenomena: the the magnetic order. The carrier (electron or hole)
metal–insulator transition, the generation of local- becomes surrounded by a cloud of spin deviations,
ized magnetic moments, and the disturbance of the and is a quasi-particle called a magnetic polaron.
Fermi liquid pattern by an increase of the parame- The Hubbard model is used as a test model for
ter U . describing band magnetism in metals, a metal–
The exact solution of the one-dimensional insulator phase transition, and various interrela-
Hubbard model, which is obtained through the tionships between magnetism and the electrical
application of the Bethe ansatz, shows that the properties of materials.
ground state is insulating for all values of U . The The Hubbard model is often used to describe
third exact result in the Hubbard model, known as narrow-band metals with d- and f -electrons, and
the Nagaoka theorem (Y. Nagaoka, 1966), states compounds based on these metals. In recent years,
that as U tends to infinity, a single hole in the the Hubbard model and its limiting case, the
half-filled system makes the ground state ferro- t –j model (d = 2, electron concentration near
magnetic. This allows one to expect that for a n = 1), have been used to describe cuprate high-
large and finite U value the ground state will be temperature superconductors.
ferromagnetic over a wide range of electronic con-
centrations. HUME-ROTHERY PHASES
In the limit of U
W , one may change from Phases formed in alloys of Cu, Ag, Au and
Hamiltonian (1) to the effective Hamiltonian of the some transition metals with metals of Groups IIB,
t –j model which replaces the Coulomb
repulsion IIIB and IVB of the periodic system of elements.
term U by the Heisenberg term J Si Sj : The crystal structure of a Hume-Rothery phase is
determined by the electron concentration, that is
+
H = −t (1 − niσ )Ciσ Cj σ (1 − nj σ )
the number of valence electrons per atom of the
ij σ
crystal lattice. An alpha phase (α-phase) with a
+J Si Sj , (2) face-centered cubic lattice covers solid solutions
ij with the electron concentrations up to 1.4. Beta
phases (β-phases) are disordered solid solutions
where Ciσ (Cj+σ ) are operators of annihilation with a body-centered lattice (e.g., CuZn, Cu3 Al,
(creation) of an electron on the lattice site pro- Cu5 Sn, etc.) that are characterized by an elec-
vided that there is no other electron at
that site, tron concentration equal to 1.5. Compounds CoAl,
S is the spin operator at the site, ni = σ niσ is NiAl, and FeAl with a body-centered cubic lattice
the number of electrons at the site i. The factors can be referred to as β-phase under the supposi-
1 − nj σ arise from the Pauli exclusion princi- tion of a zero valence for the transition metals. As
ple, which forbids electrons with spin projection σ the temperature decreases, alloy ordering (CuZn)
592 HUME-ROTHERY RULES
or eutectoid alloy decomposition (see Eutectoid al- (2) a total orbital quantum number L with the
loys) (Cu3 Al, Cu5 Zn) take place. Gamma phases highest value permitted after the choice of (1);
(γ -phases) with a complicated ordered cubic lat- (3) a total quantum number J of the resultant an-
tice with 52 atoms per unit cell (Cu5 Zn8 , Cu9 Al4 , gular momentum J = L + S with the value
Cu31 Si8 , and some others) are characterized by (magnitude) J = |L − S| for a less than half-
an electron concentration equal to 1.62. Epsilon full shell, and J = L + S, if the shell is more
phases (ε-phases) with a hexagonal close-packed than half-filled.
structure (CuZn3 , Au5 Al3 , Cu3 Sn, and so on) are
characterized by the electron concentration equal In accordance with the Hund rules, the atomic
to 1.75. The β-, γ -, and ε-phases are called elec- magnetic moment of an atom with a partly occu-
tronic compounds. The phase boundaries between pied internal electronic shell (3d, 4d and 5d; 4f
α- and β-phases in brass were calculated from and 5f ) differs from zero. This fact has a dom-
the analysis of a difference of vibrational compo- inant role in the phenomenon of magnetism in
nents of the entropy in the face-centered cubic and solids.
body-centered cubic lattices. As the temperature
decreases the contribution of the entropy to the to- HUSIMI CACTUS (K. Husimi, 1950)
tal free energy decreases, and the β-phase lattice Graph of a special type. Each block of this
becomes unstable. graph represents a unique ordinary cycle, and none
of its edges may belong to more than one ordi-
HUME-ROTHERY RULES (W. Hume-Rothery) nary cycle. Initially the Husimi cactus was studied
Semiempirical rules which determine the con- in connection with the computation of cluster in-
ditions for two elements to form solid solutions teraction integrals in the condensation theory of
over a wide range of concentrations (on phase di- J.E. Mayer (1938). The Husimi cactus is used in
agrams). These conditions can be reduced to the solid state theory as a pseudo-lattice in various
following: if the difference in atomic size (scale problems involving elementary excitations. All
factor) of the components which form an alloy sites of such a lattice (pseudo-lattice) are equiv-
exceeds 10–15%, the solubility in the solid state alent, and a transition from one site to another one
becomes limited. The formation of stable inter- is determined by simple recurrence correlations
mediate compounds would narrow the ranges of (see Bethe lattice). A Husimi cactus is constructed
restricted solid solutions, so solubility is greater using regular triangles, quadrangles, pentagons,
when the two elements have similar electronega- etc. with their apices in contact. These appear as
tivities (electronegative valence effect). Solubility graphical representations of ordinary cycles.
is highest if the elements have the same valence;
and if the valences differ then solubility in the ele- HYBRIDIZATION OF ORBITALS
ment with smaller valence always exceeds that in Formation of mixed (hybrid) orbitals at chem-
the element with larger valence (relative valence ical bonding. Hybridization utilizes the Pauling’s
effect). For complete solid solution solubility (over maximum overlap principle. The most common oc-
the entire range) all these rules should be satisfied. currence involves the mixing of s and p (px , py ,
pz ) orbitals; and when transition ions are involved
HUND RULES (F. Hund, 1927)
in the bonding, d orbitals also participate. For ex-
Empirical rules determining the relative ori- ample, a better approximation to the energy of the
entations of spin and orbital angular momenta of hydrogen molecule than the Heitler–London ap-
electrons in an atom. According to these rules a
proximation is obtained if the two-electron wave
partially full electron shell with a given principal
function is built up from 1s orbitals with a small
quantum number n and orbital quantum number l
admixture of 2p orbitals oriented towards each
is characterized by:
other to achieve a larger overlap. Hybrids of
(1) a total spin quantum number S with the high- 2s and 2p orbitals of carbon atoms provide the
est value permitted by the Pauli exclusion strong in-plane chemical bonds in crystals of
principle; graphite and the tetrahedral bonding in diamond.
HYDRIDES 593
nuclear magnetic resonance (NMR) or nuclear distribution of magnetization and the density of
gamma ray resonance (Mössbauer effect). The the electron wave function within the volume of
fields B n only weakly affect the NMR frequencies the nucleus. As a result, the isotropic hyperfine
(Knight shift, chemical shift) of paramagnetic cen- interaction aI · S and a I · S for two isotopes
ters in which the values of M are usually low due with nuclear magnetic moments μ and μ yields
to thermal motions. In magnetically ordered mate- an anomaly in the hyperfine structure character-
rials with small thermal oscillation amplitudes M, ized by the parameter Δ in the ratio a/a =
the average values of B n may reach 10–100 T, (1 + Δ)μ/μ . The value of Δ may exceed 10−2
so the hyperfine interaction is the main factor con- for heavy nuclei, with Δ < 10−5 for the light ones.
trolling the behavior of nuclear spins. There result Depending on the mutual orientation of the
certain features in the NMR, such as enhancement crystal axes, and on the total electron J and nu-
effects, nuclear spin waves, etc. Measurements of clear I spins, the anisotropic part of the hyperfine
B n provide information on the electric prop- . It is characterized
interaction has the form J AI
erties of solids, and the distribution of magnetic by the parameters Bpq = Apq −Tr A, and is deter-
atoms in them. mined by averaging the dipole–dipole interaction
The isotropic part of the hyperfine interac- between the electron moment, distributed with the
tion (contact interaction) is described by the spin probability |ψ|2 over space, and the point nu-
Hamiltonian, aJ · I , where I is the nuclear spin, clear moment. For distances between J and I
and J is its electronic counterpart. The isotropic that exceed the radius of the electron cloud, the
hyperfine interaction parameter (Fermi constant) anisotropic hyperfine interaction is close to the
is a = 16πgβgn βn (|ψ↑ |2 −|ψ↓ |2 )/3 where |ψ↑ |2 dipole–dipole interaction of point magnetic dipole
is the sum of the electron densities at the atomic moments βqJ and βn q n I . The parameters Bpq
(ion) shells, with spins aligned upward (↑) at the are found from the angular dependences of spec-
nucleus, and analogously for |ψ↓ |2 . Note that the tra of EPR and ENDOR.
nucleus is considered to be a point charge. Next
come g(gn ) and β(βn ). These are the electron HYPERFINE STRUCTURE
(nuclear) g-factor and the magneton, respectively. The splitting of spectral lines due to intrinsic
Depending on the filling of the electron shell and nuclear spin interactions with a paramagnetic ion.
the value of the nuclear moment, the value of The hyper ne interaction of the magnetic moment
an atomic moment a may vary from several kHz (spin) of the nucleus I with the magnetic mo-
to tens of GHz. It is measured using the tech- ment of the net electronic angular momentum J ,
niques of electron paramagnetic resonance (EPR), as well as the Zeeman interaction of the nuclear
NMR, nuclear γ -resonance, electron–nuclear dou- spin with an external magnetic field B, and the in-
ble resonance (ENDOR), selective saturation of teraction of the quadrupole moment of the nucleus
EPR, and spin echo. These measurements yield with the electric eld gradient, can all be described
information on the spins and magnetic moments by combinations of terms in the generalized spin
of the nuclei, on the isotropic hyperfine interac- Hamiltonian of the form J k I l B n (here k, l, n =
tion of the paramagnetic center with surrounding 0, 1, 2, . . . , k + l + n is an even number), with
nuclei, and also on the spatial distribution of its their symmetry corresponding to the point symme-
electron cloud. The isotropic hyperfine interac- try of the paramagnetic center. It is a “fingerprint”
tion serves as a measure of the covalence of of an ion, since it contains the information on its
complexes. Different isotopes and different charge nuclear spin and magnetic moment, its hyperfine
states of ions also have differing values of a. and quadrupole interactions, and the relative abun-
The isotropic hyperfine interaction is a factor en- dance of the given isotope (see Fig. 1).
hancing the effect of radio frequency fields upon The superhyperfine structure (SHFS) of an
the nucleus (sometimes that enhancement reaches electron paramagnetic resonance (sometimes
several thousand). Disregarding the approxima- called “ligand hyperfine structure”) is the splitting
tion of a point nucleus results in a weak depen- that results from the interaction of the net mag-
dence of the isotropic hyperfine interaction on the netic moment of the electronic angular momentum
598 HYPER-RAMAN SCATTERING
Fig. 1. A typical spectrum of Mn2+ : (1, 2, 3) are the sextets of hyperfine components (M = 5/2 → 3/2, 3/2 → 1/2,
1/2 → −1/2, respectively) of allowed transitions that comply with the following selection rules for changes in the
electronic, M, and nuclear, m, spins:
M = ±1,
m = 0. The two higher field sextets (M = −3/2 → −5/2 and
−1/2 → −3/2) symmetric with sextets (1) and (2) are not shown. Panels (4, 5) identify the forbidden transitions with
M = ±1,
m = ±1 which appear between the stronger allowed lines of sextets (2) and (3), respectively. The terms δ1
and δ2 identify the splittings that characterize the fine structure.
tons accompanied by the generation of a scattered the initial stress. This idea is a generalization of an
photon, and the creation (Stokes process) or anni- intuitive concept of an elastic material for a case of
hilation (anti-Stokes process) of a quantum of an finite deformations.
elementary excitation (phonon) of matter.
The magnitude of this scattering is propor- HYSTERESIS
tional to the square of the intensity of the incident Dependence of a physical state on the his-
light and, as a rule, is extremely small. There- tory of controlling parameters, rather than just
fore, it is necessary to use pulse-periodic lasers for on their values. Such a multivalued correspon-
excitation, and photon counting systems with gat- dence of the state with the parameter values is
ing for its detection. The efficiency of the process observed in any process which exhibits relaxation,
becomes significantly enhanced when the frequen- whose time constant (relaxation time) determines
cies ωl and/or 2ωl approach the frequency of an the time for reaching the equilibrium state under
intrinsic transition in the medium where a reso- the variation of external conditions. The ambigu-
nance hyper-Raman scattering takes place. ity in the value of the state of the system parameter
Important advantages of hyper-Raman scatter- depends on how the external condition parame-
ing result from selection rules which permit this ter relates to the system relaxation time. In the
process to occur for excitations that are inac- case of a rst-order phase transition one can ob-
tive in both Raman and infrared spectra (silent serve a difference between the temperatures of
lattice vibrations). In this scattering IR-active ex- the phase transition occurring during cooling and
citations are permitted; consequently, polaritons heating. This phenomenon is called temperature
(quanta of interaction between phonon and pho- hysteresis, and it is a consequence of the possi-
ton) in media of any symmetry are also permitted, ble existence of metastable states within a certain
and there are favorable conditions for studying temperature range where the thermodynamic po-
both phonon and exciton polaritons. When the tentials of the system are double-valued. For a
polariton frequency is far removed from an in- given material this temperature range depends on
trinsic resonance frequency of the medium (po-
the rate of variation of the temperature. For com-
lariton energy corresponds to “photon” portion
plete thermodynamic equilibrium in the system
of dispersion curve) the hyper-Raman scattering
(which is never reached due to the lag of the nucle-
on polaritons is also called hyperparametric light
ation processes and other causes), the temperature
scattering, four-photon light scattering, or light on
range of the hysteresis must be zero, and the
light scattering in matter. See Raman scattering of
phase transition should occur at the temperature
light.
corresponding equilibrium of thermodynamic po-
HYPER-RAYLEIGH SCATTERING tentials. Another kind of hysteresis phenomenon
A process of nonlinear interaction of a suffi- is observed in many physical systems with mag-
ciently intense light wave of frequency ω0 with netic or electric polarizations, when a transition
matter, which results in the appearance of scat- from one phase to another includes a state with do-
tering at frequencies 2ω0 , 3ω0 , and so on, with mains (an intermediate state), and the ambivalence
the magnitude proportional to, respectively, I02 , I03 of the dependence of values does not disappear
and so on (here I0 is the intensity of the incident under infinitesimally slow variations of the exter-
light wave). See also Rayleigh scattering of light, nal conditions. Examples of this are ferroelectric
Light scattering. hysteresis and magnetic hysteresis observable in
ferroelectric and magnetically ordered materials,
HYPOELASTIC MATERIAL (introduced by respectively, and also elastic hysteresis, i.e. a lag
C. Trusdell, 1955) in the phase of a strain relative to that of a stress
A material in which the components of stress (see, e.g., Shape memory effect, Martensitic trans-
variation rates under a strain are homogeneous formation). The phenomenon of elastic hysteresis
functions of the deformation rate components. is inherent in all solids. In experiments, these
A hypoelastic material is non-viscous and its in- hysteresis phenomena appear as loop-like depen-
finitesimal deformations are reversible relative to dences (hysteresis loops) of values characterizing
600 HYSTERESIS IN PHASE TRANSITIONS
the system state (polarization, magnetization, in- metastable states at rst-order phase transitions.
ternal stress, etc.) under a cyclic variation of It is observable under sufficiently fast changes of
corresponding values specifying the external con- the external conditions owing to the slow response
ditions of a hysteresis loop (electric field, magnetic of the system, and its persistence in a state with an
field, applied stress). Hysteresis is also typical of order parameter value corresponding to a relative
certain states of condensed media with a noner- minimum rather than the absolute minimum of the
godic behavior (spin glasses and other amorphous thermodynamic potential (see Hysteresis, Phase
magnetic substances, dipole glasses, and so on). transitions).
601
602 ICE BOMB
tetrahedra with a common edge there arises a√de- IDEAL SOLID SOLUTION
ficiency of space angle β = 4π − 20 arcsin(1/ 3) A homogeneous, multicomponent (i = 1, 2,
(Fig. (a)). Elimination of the deficiency is pos- . . . , n) solid phase of varying composition, its for-
sible either with the help of elastic closure of mation being accompanied by the same changes
the slit, which is equivalent to the introduction of thermodynamic potentials as in mixing ideal
of a positive wedge disclination along the com- gases. In contrast to ideal gases, interactions be-
mon edge of tetrahedra with the magnitude ω = tween particles in an ideal solid solution do exist
β/2 ∼= 7◦ 20 (Fig. (b)), or by adding an extra reg-
but undergo no changes with variations of com-
position, while the partial enthalpies Hi and en-
ular tetrahedron into the existing slit (Fig. (c)). The
tropies Si (both vibrational and rotational) are
latter operation leads to the appearance of a nega-
conserved. Only the configuration component of
tive disclination with ω ∼= 2π/5 = 72◦ . Therefore, the entropy S changes. Hence, the changes of
tetrahedra with local tetrahedral packing turn out thermodynamic potentials are
H = 0,
S =
to be distorted, and the packing geometry can be
−R ni=1 xi ln xi and
G = −T
S (R is the
described by the number q of tetrahedra adjoined gas constant, xi is the relative concentration of
along the linking edge. In the examples given, the ith component, and G is the Gibbs free en-
q = 5 and q = 6. ergy). These solutions are idealized states; close
An ideal glass with tetrahedral packing is an to them are isotopic solid solutions. More com-
atomic configuration in three-dimensional space, mon are regular
solid solutions with
H = 0,
which is transformed with the help of a relax-
S = −R ni=1 xi ln xi . Real solid solutions de-
ation process to a condition involving more or less viate from both ideal and regular solutions in that
differing bond lengths. For an ideal glass, mean the entropy increases, S > S id , with the formation
values of q and Z, related by the expression q = of defects (e.g., Schottky defect), and the entropy
6 − 12/Z, are equal to q i = 2π/ arccos(1/3) ∼ = decreases S < S id with the reduction of the parti-
5.104 and Z i ∼ = 13.397. Values close to Z i are
cle oscillation frequencies under ordering.
specific for coordination numbers of the so-called IMAGE FORCES in electrostatics
Frank–Kasper phases (F.C. Frank, J.S. Kasper) in Electrostatic attractive or repulsive forces af-
complex alloys with a large number of atoms in fecting an electric charge near a solid surface or
a unit cell (e.g., Z = 13.358 for Mg32 Zn49 ). In near the interface between different media. Forces
the framework of the present approach, Frank– of the image type arise because of the polarization
Kasper phases can be treated as near-ideal glasses (redistribution of natural charges) in a medium
that degenerate to crystals due to their chemi- acting on a test charge e at rather long (macro-
cal composition. One can define various defects scopic) distances x from the interface. These
with respect to ideal glasses. Disclinations that are image forces are determined from the classical po-
the lines of distorted atomic coordination in local tential:
tetrahedral packing play a principal role in glasses e2 ε1 − ε2
with local tetrahedral packing. W (x) = ,
4x ε1 + ε2
604 IMAGING, EPR
where ε1 is the relative dielectric constant of the of the image converge to zero, while in a spatially-
medium with the charge e, and ε2 is the dielectric restricted body of arbitrary shape they depend on
constant of the medium bordering it. In particu- this shape, and on the arrangement of real defects
lar, if the charge is in a vacuum (ε1 = 1) close relative to its boundary. Despite the compara-
to a metal surface (ε2 → ∞), then image forces tively small value of corresponding deformation
are a maximum and correspond to reflecting im- “images”, bound by surface image forces (and
age forces with the potential W (x) = −e2 /(4x), inversely proportional to the cube of the charac-
which arises because of the redistribution of free teristic crystal size L), they are evenly (if defects
conduction electron density in the metal. The di- do not concentrate near the surface layer) spread
vergence of W (x) at the point x = 0 is eliminated throughout the crystal volume ∼L3 . Therefore,
by systematically taking into account the effects of image forces do make a finite contribution to the
spatial dielectic dispersion at distances compara- change of some characteristics (e.g., volume) upon
ble to the screening radius, and also by accounting the introduction of defects, which is comparable
for quantum effects. For rapidly moving charges to changes arising from the emergence of static
the image forces are attenuated due to retardation deformation fields that decrease at long distances
effects arising from the inertia of natural charges from defects.
in a medium. For metals this happens at times The energy contribution to the total strain inter-
of flight comparable to the plasma oscillation pe- action of defects related to image forces is gener-
riod of about 10−15 s. Under the action of image ally an “attractive” one and varies very smoothly
forces surface electron states can arise, which are even at distances r between defects comparable
localized near the interface, e.g., on the surface with L, i.e. it is almost independent of r for
of liquid helium (see Levitating electrons). Image a macroscopically uniform average distribution
forces play an important role in the transverse in- of defects. It is particularly insensitive to fluc-
teraction of adsorbed atoms or molecules at a solid tuational redistributions of their concentration at
state surface (see Adsorption). In particular the in- distances r L.
teraction between adsorbed atoms with charge Q,
located a distance x from the metal surface and IMAGING, EPR
separated from each other by the distance R > x, See EPR imaging.
has the following form for R
x as a result of
IMAGING, NMR
reflecting image forces:
See Magnetic resonance imaging.
Q2 Q2 P2
W (R) = − # + ≈ , IMPACT IONIZATION, collision ionization
R 2 + 4x 2 R 2R 3
Ionization, which occurs when an ionizable ion
where P = 2xQ is the effective dipole moment, acquires energy as a result of a collision with an-
i.e. W (R) exhibits a dipole–dipole type repulsion other particle (electron, ion, atom). In the context
(see Dipole–dipole interaction). of solid state physics, the term impact ionization
Image forces in imperfect crystals are those is used in reference to ionization brought about
taking into account boundary conditions in a crys- by applying a strong electric field to the sample.
tal with defects. In an infinite crystal, displace- Conduction electrons are accelerated by the field
ment vectors u due to defects decrease, according and give up part of their kinetic energy at colli-
to the continuum theory of elasticity, as the dis- sions with atoms of the solid. In a strong enough
tance from the defect grows. However, in a crystal field, the energy acquired by electrons may ex-
of finite dimensions such a solution does not obey ceed the band gap width E g , in which case an
the surface boundary conditions (e.g., condition of electron can jump from the valence band to the
absence of stress on a free surface). To meet such conduction band as a result of collisions between
conditions image forces are introduced, which are conduction electrons and vibrating atoms. The ap-
virtual forces acting from the sites of imaginary proximate criterion for impact ionization is the
external defects inducing so-called displacements equality of the energy E g gained during the pas-
of “images”. In an infinite medium, displacements sage along a mean free path length l, i.e. during
IMPEDANCE, SURFACE 605
the average interval between two successive colli- ultimate strength, see Shock hardening). As a re-
sions of an electron with phonons or lattice defects sult of a nonuniform distribution of stresses and
(see Defects in crystals), i.e. eE = E g /(αl), where stress concentrations in certain regions, the plas-
E is electric field intensity, and α is a numerical ticity of a structural member under impact load is
factor specifying the amount of energy lost by a lower than that under static load (see Construction
high-speed electron in a collision. The process of material strength). The impact strength of a struc-
impact ionization results in an avalanche-like in- tural member is improved by a design plan that
crease of the concentration of free current carriers. allows for increasing the uniformity of stress dis-
For field intensities from 105 to 106 V/cm, this tribution throughout the bulk of material.
increase is limited by the recombination of elec-
trons and holes produced by impact ionization. At
IMPATT DIODE
higher fields (of the order of 107 V/cm), however,
impact ionization can bring about the electrical The same as Impact ionization avalanche transit
breakdown of a semiconductor (see Breakdown of time diode.
solids).
IMPEDANCE (fr. Lat. impedio, am hindering)
IMPACT IONIZATION AVALANCHE TRANSIT A complex-valued quantity that is an analogue
TIME DIODE, IMPATT diode of the electrical resistance for harmonic (AC)
A semiconductor diode with a dynamic nega- processes. The impedance has the complex form
tive resistance in the UHF range that forms due Z = R + iX, where the real part R is the re-
to an avalanche breakdown (see Semiconductors) sistive or lossy term, and the imaginary part X
precipitated by a phase shift between the voltage is the reactive term. A distinction is drawn be-
and the current at the device output. This effect tween the impedance of an AC circuit element
arises from the finite inertia of the avalanche mul- (two-terminal impedance) and the impedance of
tiplication of carriers and the finite time, τ , that a (planar) surface in a monochromatic electro-
it takes for these carriers to traverse the carrier- magnetic field (surface impedance). The concept
depleted layer (transit time). The efficient transfer of impedance was introduced into electrodynam-
of IMPATT energy input into UHF oscillations of ics by O. Heaviside and O. Lodge; that of sur-
frequency f occurs when f · τ ≈ 0.5. IMPATT de- face impedance belongs to S. Schelkunoff (1938).
vices are used to generate, amplify, and transform Impedance characteristics are used not only in
oscillations in the range from units to hundreds of electrodynamics, but also in descriptions of trans-
GHz. They remain the most powerful solid-state mission lines for wave disturbances of any nature
sources of UHF oscillations in the 20–400 GHz (e.g., Acoustic impedance). In AC electrical cir-
band. cuit theory the impedance Z of a resistance R,
inductance L and capacitor C in series has the
IMPACT MALLEABILITY
magnitude Z = [R 2 + (XL − XC )2 ]1/2 , where
See Toughness.
XL = ωL and XC = 1/(ωC) are the inductive and
capacitive reactances, respectively, ω = 2πν is the
IMPACT STRENGTH
frequency, and ω0 = 1/(LC)1/2 is the resonant
The strength of materials under short-time im-
frequency. The voltage V and current I are com-
pact loading, one of the forms of dynamic strength.
plex quantities related by Ohm’s law V = I Z, and
Impact strength determines the resistance to failure
the phase angle φ between the current and voltage
under short-time (of the order of 10−3 s and less)
is given by the expression tan φ = (XL − XC )/R.
loads of high severity. The action of these loads is
The losses in the circuit arise from the resis-
characterized by wavelike propagation, and hence
tance R.
by transient stress distribution over the bulk of
material under load, as well as by the high-rate
deformation of the material which, in a number of IMPEDANCE, SURFACE
cases, enhances its strength properties (yield limit, See Surface impedance.
606 IMPREGNATION
When a sample possesses trapping levels, con- of the projection of the magnetic field on the metal
ditions appropriate for impurity breakdown may surface; the lattice has the structure of triangles
be produced by filling the levels with current carri- elongated in this direction. Inclined vortices deter-
ers. For example, one may irradiate a sample with mine the current-carrying capacity of the surface
photons of energy exceeding the band gap width layer, which is due to the pinning of inclined vor-
(so-called induced impurity breakdown). tices by defects of the surface structure (see Vortex
pinning).
IMPURITY CONDUCTIVITY
Low-temperature hopping conductivity involv- INCLUSION, MACROSCOPIC
ing impurities in amorphous or doped (impurity) See Macroscopic inclusions.
semiconductors. The hops (jumps) may be accom-
panied by the absorption or emission of phonons. INCOMMENSURATE–COMMENSURATE
At moderately low temperatures the hop occurs PHASE TRANSITION
by thermal activation through a potential barrier A phase transition in a crystal, which displays
that separates the impurities. At low temperatures, a superlattice; the transition is accompanied by
quantum tunneling appears to be essential. See a reduction of the incommensurability parame-
also Electrical conductivity, Semiconductors. ter to zero (see Incommensurate structures) and
Sometimes the term impurity conductivity is the resulting restoration of translational symmetry.
used in another sense. When the doping of a An incommensurate–commensurate phase transi-
semiconductor is sufficiently high and the current tion is, as a rule, a rst-order phase transition.
carriers that determine conductivity are formed The transition may be detected through diffrac-
by ionization of the impurities, these carriers can tion methods; changes of certain physical prop-
move in an impurity band. erties (electric resistance, magnetic susceptibility,
etc.) are also indicative of an incommensurate–
IMPURITY DISTRIBUTION
commensurate phase transition. The transition is
See Interphase impurity distribution coef cient.
observed in dichalcogenides of transition met-
als with the formula MX2 (M=Ta, Nb; X=Se,
IMPURITY-INDUCED FERROELECTRICITY
Te, S), over the premartensitic temperature range
See Ferroelectricity induced by impurities.
of compounds based on TiNi; in ferroelectrics
IMPURITY INTERACTION
K2 SeO4 , (ND4 )2 BeF4 , etc.
See Magnetic impurity interaction.
INCOMMENSURATE MAGNETIC STRUCTURE
IMPURITY STATE, SURFACE See Modulated magnetic structures.
See Surface impurity states.
INCOMMENSURATE STRUCTURE
INCLINED VORTICES in superconductors, Kulik’s A superstructure in a crystal, with a period in-
vortices commensurate with the lattice constant (may be
A specific kind of vortex lattice formed in pla- found in more than one crystallographic direc-
nar superconducting layers with a thickness of the tion). It originates when atoms are displaced off
order of the coherence length ξ ∼ 10−5 –10−6 cm. their positions in the initial structure in such a
The layer is located in the neighborhood of the sur- way that it becomes impossible to find a transla-
face and retains superconducting properties under tion of the lattice that would superpose the crystal
the action of an applied field exceeding the field with itself. An incommensurate structure fails to
needed to destroy bulk superconductivity. The vor- fall into any of 230 crystallographic space groups.
tex state is realized at all angles of magnetic field The transition to this structure occurs as a phase
inclination, except for θ = 0 (i.e. strictly paral- transition when the temperature is lowered. The
lel to surface) in the field interval from Bc2 (θ) to modulation responsible for this structure may be
Bc3 (θ) (see Surface superconductivity). The vor- produced not only by displacement waves, but also
tex state is a lattice that is oriented in the direction by charge density waves or spin density waves.
608 INCOMPATIBILITY TENSOR
Wave vectors of the incommensurate and the ini- the plate where the plate magnetization M is at
tial structures, Q and q, are related to each other its maximum, or where the surface energy density
as Q = n(1 + δ)q where δ is the nonintegral in- of the domain walls σ is at its minimum. When
commensurability parameter, and n is an integer. the distribution profiles of either M or σ exhibit
Incommensurate structures are identified either by several respective minima or maxima through the
using diffraction techniques (e.g., by appearance plate, there can be a variety of subsurface and
of anomalous satellites during electron (neutron) internal volume magnetic domain structures of
diffraction analysis), or by tracking the changes in various types.
certain macroscopic properties (e.g., electric resis-
tance, etc.). Such structures have been identified INCONGRUENT MELTING
in certain laminated dichalcogenides of the MX2 Melting when the liquid phase (the melt) dif-
type (M = Ta, Nb; X = Se, Te, S), in compounds fers in chemical composition from the solid phase.
based on TiNi, Cr, in quasi-one-dimensional crys- Incongruent melting is observed in solid solu-
tals, etc. tions and in phases of some compounds of vary-
ing composition. This melting can complicate
INCOMPATIBILITY TENSOR the processes of growing monocrystals with a
A second-rank tensor T = Ink A is the curl of given stoichiometric composition (see Monocrystal
the transposed curl of a given second rank ten- growth). The composition stability of a growing
sor A. In coordinate notation, crystal is ensured by the change of melt com-
position calculated from the compositional phase
∂ 2 Aqn
Tik = eimn ekpq , diagram.
∂xm ∂xp
where eij kl is the alternating sign tensor. If A is INDENTER
symmetric, then Ink A is also symmetric. Hence, A hard tip of a device serving for local loading
under the condition Ink A = 0, the tensor A can of materials in testing their hardness. The most
be written in the form of the deformation of some common indenter shapes are as follows. A hard
vector ui : ball is used as the indenter in testing the Brinell
hardness, with the relation d/D ≈ 0.375 main-
1 ∂ui ∂u
Aik = + k . tained (where d is the indentation diameter, D the
2 ∂xk ∂xi ball diameter). A cone with an apex angle of 120◦
This property of an incompatibility tensor is used and a rounded tip is used in the Rockwell hardness
for an invariant formulation of the conditions for test. A regular tetrahedral pyramid with an angle
the strain compatibility. of 136◦ between opposite faces is used in testing
Vickers hardness. A regular trihedral pyramid with
INCOMPLETE MAGNETIC DOMAIN, an angle of ∼77◦ between a face and the axis is
non-see-through domain used for Berkovich hardness. A tetrahedral pyra-
One of several possible types of magnetic do- mid with a rhombic base is used in testing Knopp
mains in highly anisotropic plates with a single hardness with the ratio between the long and short
magnetic axis (magnetic lms with quality factor diagonals of the imprint equal to seven. Bicylindri-
exceeding unity), such that their easy magnetiza- cal Yegorov’s indenter and a flat die type are also
tion axis remains normal to their developed sur- known. Equivalence of the indenters of various
faces. Characteristic of such domains is the lack shapes (i.e. a possibility to obtain the same hard-
of contact of a magnetic domain wall with either ness values) is due to the same value of the elastic
one or both surfaces of the plate. In the first case strain and plastic deformation under the inden-
one speaks of subsurface incomplete domains, and ter. Indenter material must possess high values of
in the second case one speaks of internal volume hardness and an elastic modulus exceeding those
domains. Incomplete magnetic domains are sta- of the material tested. Local loading with the rigid
ble only in plates that are inhomogeneous along indenter allows one to determine the hardness as
their thickness, and are localized in that region of well as the fracture toughness (crack resistance),
INDIRECT EXCHANGE INTERACTION 609
Young’s modulus, breaking stress, cold brittleness INDIRECT ELECTRON SPIN–SPIN COUPLING
temperature, and other mechanical properties of a The coupling of electron shells of paramag-
material. netic centers caused by various fields of elemen-
tary excitations in crystals: phonons, plasma os-
cillations, conduction electrons, spin waves, and so
INDENTER–HARDNESS DIAGRAM
on. This coupling of paramagnetic ions through
A graphic presentation of the dependence of
a phonon field was first considered by K. Sugi-
the loading of the indenter P and the depth of
hara (1959). He made use of an electron–phonon
its penetration (indentation) h during hardness interaction mechanism which changes the orbital
checking. The indenter–hardness diagram exhibits motion of valence electrons, taking into account
the following stages: (1) active loading at a con- spin–orbit coupling which brings about the cor-
trolled rate (load increased from zero to maxi- relation of the spin variables of paramagnetic
mum), (2) imprint hold time under pressure, and centers with the phonon fields. In the second or-
(3) active unloading and possible repeated load- der of perturbation theory the spin variables of
ing with passage through a hysteresis loop. The the paramagnetic centers are “mixed”, the sys-
appearance of the indenter–hardness diagram de- tem energy involves two paramagnetic centers,
pends on the indenter shape, e.g., the pressing and phonon field terms appear which depend on
of the prism with a flat end produces a diagram the spin variables of the paramagnetic centers, and
which resembles a single-axis compression. The their reciprocal distances. These terms correspond
hardness Hh , calculated from the depth depen- to the energy of the indirect electron spin–spin
dence P (h) of the applied load, is a measure of coupling, and written in operator form they rep-
the resistance to elastico-plastic deformations that resent the effective Hamiltonian of the indirect
come into play, unlike the Vickers or Brinell hard- electron spin–spin coupling of the paramagnetic
ness H (see Macrohardness) which to a greater centers. Couplings of this type are anisotropic in
extent reflects the resistance to plastic deformation. the spin variables, have a long-range nature, and
The correlation between Hh and H is an important are comparable in magnitude with the magnetic
structure-sensitive parameter of the measurement, dipole–dipole interaction. For S-state ions the anal-
depending on the interatomic bond type, on the ogous mechanism is less important. This indirect
plasticity, and on the degree of material loos- coupling can be initiated by any elementary exci-
ening. The information content of the indenter– tation field which can influence the orbital electron
hardness diagram is especially high when testing motion in any manner. In addition to the spin–orbit
surface layers of coatings and powder compacts indirect coupling mechanism described above, an
(see Powder metallurgy). The resolution reaches alternative mechanism is possible that involves ex-
10 nm, with the loading resistance limited by el- change coupling caused by fields of elementary
ementary mechanisms of mass transport. When P excitations.
is reduced the quantities H , Hh and their non-
uniformity grow due to the statistical nature of INDIRECT EXCHANGE INTERACTION
strength. Modern theories permit a separation by Exchange interaction of spin degrees of free-
this method of the elastic and plastic deformations dom of localized electrons (or atomic nuclei)
of the material under the indenter, and also permit through the perturbation of another subsystem
a determination of Young’s modulus. The indenter such as interstitial (diamagnetic) ions located
among magnetic ions in magnetic insulators, or
hardness test, its nonuniformity, the ratio H /Hh ,
conduction electrons in metals and semiconduc-
and the roles of elastic (reversible) deformation
tors.
all come under the scope of “kinetic hardness” or
In magnetic insulators containing ions of non-
“microhardness”.
transition series metals, where the direct overlap of
transition metal ion d-orbitals is small, the main
INDEX OF REFRACTION mechanism of exchange coupling is Kramers–
See Refractive index. Anderson indirect exchange (superexchange). The
610 INDIRECT EXCITATION
specificity of the superexchange arises from the increase being caused by an increase of cova-
nature of the localized wave functions of the mag- lent bonding.
netic electrons, represented by the superposition of
atomic d-orbitals (f -orbitals for rare-earth com- The application of these rules provides the
pounds) with the s- and p-wave functions of the following relations involving the exchange para-
interstitial anions (F− , O2− , S2− , Se2− ligands), meters Iab between ions with various numbers of
and the influence of the overlap between them. d-electrons on the centers:
For d-orbital transition ions with octahedral or 180◦ exchange:
tetrahedral coordination the parameter Iab of the
isotropic exchange coupling Iab (Sa Sb ) is deter- • I (d 8 −d 8 ) > I (d 5 −d 5 ) > I (d 3 −d 3 ) > 0 (an-
mined by the superposition of the contributions of tiferromagnetic bonds),
unpaired magnetic electrons in d-states (eg or t2g ) • I (d 3 − d 8 ) < I (d 3 − d 5 ) < 0 (ferromagnetic
split by the crystal field: bonds);
Iab = (na nb )−1 Iia ,jb nia njb , 90◦ exchange:
ia ,jb
• I (d 3 − d 5 ), I (d 5 − d 5 ) > 0 (antiferromagnetic
where na and nia are the numbers of unpaired bonds),
electrons on the center, and on the ith orbital • I (d 8 − d 8 ) < 0; I (d 3 − d 3 ) can have different
of this center. In the Anderson theory (P.W. An- signs.
derson) two basic superexchange mechanisms are
considered: (1) potential ferromagnetic exchange In substances with metallic conductivity the in-
(direct exchange coupling of orthogonal molecu- direct exchange coupling of electronic or nuclear
lar orbitals of magnetic ions), and (2) second-order spins occurs through conduction electrons (see Ru-
antiferromagnetic kinetic exchange generated by derman–Kittel–Kasuya–Yosida interaction (RKKY
the virtual transport of unpaired electrons from interaction), Indirect nuclear spin–spin coupling).
one center to another. The dependence of the mag-
nitude and sign of the exchange parameters Ii,j INDIRECT EXCITATION
(and hence the dependence of Iab ) on the filling A many-wave effect in dynamic scattering of
of d-states, and the angle of the superexchange X-rays by monocrystals. It occurs in cases when
metal–ligand–metal bond are explained empir- three or more points of the reciprocal lattice all lie
ically by the Goodenough–Kanamori–Anderson on the Ewald sphere (see X-ray structure analysis),
rules (J.B. Goodenough, J. Kanamori, P.W. Ander- e.g., points (000), (hkl), and (h k l ), with (hkl)
son), which we summarize as follows: corresponding to a forbidden reflection. If the re-
flection (h − h , k − k , l − l ) is not forbidden,
1. If half filled orbitals on adjacent ions overlap
a wave scattered in the direction [hkl] appears,
then the exchange coupling parameter Ii,j is
since in this case the reflection (h k l ) plays the
antiferromagnetic and relatively strong. In the
role of a new zero point, and the wave field at first
case of a 180◦ bond the eg -state coupling is
scatters in the direction [h k l ], and then in the di-
the strongest. For a 90◦ bond relatively strong
rection [hkl]. See also Many-wave approximation,
exchange coupling takes place in octahedral
Dynamic radiation scattering.
coordination for the dz -state electrons of one
center, and for dxy on the other.
2. If there is no overlap between half filled or- INDIRECT NUCLEAR SPIN–SPIN COUPLING
bitals then the exchange is ferromagnetic and, Coupling of nuclear spins through orbital and
as a rule, weaker. spin magnetic moments of nearby electrons. In-
3. For each magnetic ion the magnitude of the ex- direct nuclear spin–spin coupling takes place be-
change parameters Iab and angle of the bond cause of the polarization of electronic shells by
increase with the replacement of one ligand by magnetic fields arising from nuclear moments.
another in the series F− , O2− , S2− , Se2− , this Nuclear spins, because of their coupling with or-
INDIUM 611
Doping of Si and Ge with microscopic quantities axis, as a rule, vanishes (because of relaxation
of indium provides hole conduction and p–n junc- processes). The following fields may be used for
tions. Besides, indium is used as sealing, soldering inducing an optic axis: electric field E (see Kerr
and corrosion-resistant materials in the electronics effect), magnetic field B (see Voigt effect, Fara-
industry. Indium coatings possess high reflectivity day effect), field of elastic forces, electromagnetic
and may be used for manufacturing mirrors and field (see Self-induced light polarization change).
reflectors. In the context of phenomenological crystal optics,
induced optical anisotropy may be treated as a
INDUCED DIPOLES relationship between the dielectric constant ten-
Dipoles produced at locations of atoms and sor εij (ω, k) and external fields E, B, and also
molecules by an external electric field. The phys- mechanical stresses σij . If the external fields are
ical cause for the dipole moments to appear is rather weak, then the tensor εij may be expanded
the displacement of ions from their equilibrium in a power series in E, B, σij , and taking into
positions, and the shifted polarizations of atomic account crystal symmetry considerably decreases
electron shells. Induced dipoles are responsible for the number of independent tensor components
the polarization of a dielectric material. In electric that appear in the expansion. Starting from this
fields of complicated configurations, e.g., in crys- expansion the polarization of normal electromag-
tal elds, the induction of higher order multipoles netic waves is determined. Molecules in solids
such as quadrupole moments (see Quadrupole) is are anisotropic and are characterized by a polariz-
possible. ability tensor. When the molecules are arbitrarily
spatially oriented, then the medium as a whole
INDUCED LIGHT SCATTERING
is macroscopically uniform. On the application of
Light scattering of high intensity off a solid,
an external electric field, the molecules can orient
with the simultaneous creation in the solid, under
themselves with the axis of highest polarizabil-
the influence of the incident radiation, of co-
ity along the field, a process responsible for the
herent oscillations or wave-like excitations. The
initiation of anisotropy. Three polarizability con-
scattered radiation contains Stokes’ components
tributions are usually singled out: an electronic
with lower frequencies, and anti-Stokes’ compo-
one due to displacement of electronic shells from
nents with higher frequencies than the incident
nuclei in the external field; an ionic one asso-
beam. From the classical point of view the Stokes’
components of the induced light scattering can be ciated with shifts of ions relative to each other;
interpreted as arising from elementary excitations and an orientational (dipole) one due to change of
in the medium which subtract energy from the orientation of elementary dipoles in the external
incoming Raman wave and produce a lower fre- field. Each of these polarizability components has
quency scattered wave. Induced light scattering is its characteristic time of dielectric relaxation; and
exemplified by Raman scattering of light, as well these times differ sharply. Thus, e.g., orientation
as by Brillouin scattering of light. This scattering relaxation in solids is characterized by relaxation
is used for transformations of laser radiation, for times in the range from 10−10 s to several hours,
generating ultra-sound and other types of excita- depending on the physical nature of the induced
tions in solids, and for investigation of the internal optical anisotropy and the frequency of the ini-
structure of matter (nonlinear spectroscopy). tiating radiation. If molecules possess permanent
magnetic moments they may be oriented by a con-
INDUCED OPTICAL ANISOTROPY stant applied magnetic field.
Optical anisotropy arises in optically isotropic
media in the presence of external fields which in- INDUCED RADIATION
duce a preferred direction in these media. The Emission of electromagnetic waves by an ex-
decrease of symmetry causes a difference of op- cited quantum system under the influence of ex-
tical properties of the media along the preferred ternal radiation, such as the stimulated emission
direction (optical axis) and in the plane perpen- of a laser. The probability of induced radiation
dicular to it. On removal of the field, the optic is proportional to the external radiation intensity.
INELASTIC NEUTRON SCATTERING 613
The frequency, propagation direction, phase and INELASTIC NEUTRON SCATTERING in solids
polarization of the emitted and inducing radia- Scattering accompanied by energy and mo-
tions coincide with each other. If external photons mentum exchange between the neutron and the
of frequency ν = (E2 − E1 )/ h enter a medium scattering system during the course of their in-
where the upper energy level E2 is more populated teraction. Thermal neutrons are commonly used
than the lower level E1 then the process of induced because their energy and momentum is compara-
radiation can develop into an avalanche. The phe- ble to those of elementary (vibrational) excitations
nomenon of induced radiation is used in lasers in the condensed state of matter. The presence
for the generation and amplification of electro- of the neutron magnetic moment makes it pos-
magnetic waves (see Quantum ampli er, Quantum sible to study the magnetic properties of solids.
radio-frequency generator). The types of inelastic neutron scattering include
coherent and incoherent, nuclear and magnetic
INDUCTANCE (see Neutrons). Coherent inelastic neutron scat-
A physical quantity, L, that characterizes mag- tering is governed by the collective dynamics of
netic properties of an electrical circuit. The induc- all the particles, corresponding to the interac-
tance is numerically equal to the ratio of magnetic tion between a neutron and the collective excita-
u x Φ threading a contour (e.g., loop or coil of tions of a lattice (phonons, magnons) when both
wire) to the strength of the current I in the contour the energy and the momentum of the interact-
producing this flux, L = Φ/I . The inductance de- ing particles are conserved. Experiments on this
pends on the size and shape of the contour, on the scattering by monocrystals yield complete infor-
magnetic permeability μ of the conductors com- mation on phonons (nuclear neutron scattering)
posing it, and on the magnetic susceptibility χ of and magnons (magnetic neutron scattering) in a
the environment. The magnetic energy storage in crystal, including phonon and magnon dispersion
an inductive circuit is LI 2 /2. A changing current curves, largely unattainable by other techniques.
through an inductive circuit produces the voltage During incoherent inelastic neutron scattering, the
V = −L dI /dt . particles interact with individual nuclei. However,
Inductance coils with cores of soft magnetic due to the strong bonds in a lattice, other nuclei
materials are used to increase the inductance. Ow- also affect the process of such scattering, and a
ing to the dependence of the magnetic permeabil- large group of particles takes part in it. Therefore,
ity of ferromagnets on the magnetic field strength, incoherent scattering may also be treated as an in-
the inductance of such coils depends on I . The in- teraction of the neutron with quanta of elementary
ductance of a solenoid situated in a medium with excitations when only the energy of the colliding
relative permeability μ is L = μN 2 S/ l, where μ0 particles is conserved. Incoherent inelastic neu-
is the permeability of free space, N is the number tron scattering enables retrieval of data on phonon
of turns of the coil, S and l are the coil cross- spectra of crystals. Compared to other (e.g., op-
section and length, respectively. The SI unit of tical) techniques, using neutron scattering allows
inductance is a henry, H, where 1 H = 1 T·m2 /A. one to carry out experiments over a wide range
of wave vectors at very low vibration frequencies.
Besides, this technique is not limited by selection
INDUCTION
rules since all the vibrations are active. Because of
This term involves electric ux density D, mag-
the large element-to-element difference in the neu-
netic ux density B (electromagnetic field char-
tron scattering cross-sections, this inelastic scat-
acteristics); and also electromagnetic induction,
tering permits one to study the dynamics of atoms
electrostatic induction.
of a specific type in complex compounds. With the
appearance of powerful neutron sources (such as
INDUCTION, NUCLEAR high flux reactors, accelerators) the capabilities of
See Nuclear induction. this technique are greatly expanding. For exam-
614 INFILTRATION
ple, its sensitivity (energy resolution of 1 μeV ∼ in devices with charge transfer. There, the action
0.8·10−2 cm−1 ) is comparable to that provided of light on a semiconductor releases charge carri-
by optical techniques, and its high intensity makes ers which, after amplification, are collected at the
it possible to conduct studies with small samples metallic layer of the MOS structure. The visual-
(down to 10−2 cm3 ), including surfaces. Inelas- ization of images in MOS arrays is possible using
tic neutron scattering data demonstrate its unique an electroluminophor. Devices with charge trans-
capabilities for studying phonons, structural phase fer are used in TV engineering, for visualization
transitions, magnons, collective excitations in liq- and amplification of IR images, etc. To process
uids, diffusion of hydrogen in metals, rotational information, the phenomenon of low-temperature
excitations and tunneling in molecular crystals, superconductivity can also be employed.
crystal e ld effects in compounds of rare-earth ele- Magnetic information recording via ferrites is
ments, dynamics of disordered solids, etc. See also most widely used in random access and read
Neutron spectroscopy. only memories (RAM, ROM) of large comput-
ers. The principle of recording in binary is based
INFILTRATION by reversing the magnetization direction; and to
Process of penetration or permeation. Pene- accomplish this each magnetic memory element
tration of liquids or solids through microscopic is supplied with two current circuits. Early com-
cavities and/or pores under action of capillary puters used ferrite storage systems in the form
forces. The phenomenon of infiltration underlies of plates with holes (0.5 mm) with magnetizing
the technological processes of impregnation and wires passing through them. In order to design
soaking.
more compact storage systems, ferrite cores (di-
mensions 0.5 × 0.1 mm2 ) or thin lms (6 μm;
INFORMATION RECORDING in solids
core diameter 2 mm) can be used. Periods of in-
Processes based on using external effects (de-
formation retrieval 100 ns long and total memory
formation, electromagnetic fields, chemical reac-
volumes of 1011 bits are achieved with ferrite
tions, etc.) for storing information via residual
memories. Information recording in magnetic ma-
changes of physicochemical properties of solids.
terials requires rather high values of magnetizing
The information recording technique and the in-
currents (power ∼10 W). Thin magnetic lms on
formation carrier determine, to a large extent, the
an aluminum substrate are used in ROM. Good
design of a computer, or of an information sys-
prospects for magnetic memory systems are ther-
tem in general (data input and output, storage
momagnetic information recording systems, and
system, processor, etc.). At present, the following
systems using magnetic bubble domains.
techniques of information recording in solids are
extensively employed: electrical, magnetic, opti- Optical information recording is widely used
cal, and molecular. in devices for the reproduction of sound and video
Electrical information recording is widely used images (see Optical storage disks, Optical record-
in processors and in storage systems in the form ing media).
of semiconductor transistors and large-scale inte- Molecular information recording is now at the
grated circuits (LSIC). The LSICs are the main research stage. The huge information capacity
processors in microcomputers. A representative of protein molecules (organic polymers), genetic
LSIC contains about 106 elements per cm2 and molecules (DNA, RNA), etc., can be used for
has a response time of 1–10 ns. The operat- it. When they are utilized in the form of thin
ing principle of semiconductor transistors involves films on a solid substrate, one can class them
switching between two states (1; 0). In recent as solid systems. Attempts have been made to
years metal–insulator–semiconductor structures use various tautomeric reactions, polymerization,
(or metal–oxide–semiconductor structures, MOS reactions of enzymes with a substrate, etc., for in-
structures) have found wide use for processing in- formation recording. To indicate a record, one can
formation, and especially images. MOS structures use changes in electrical, optical and other proper-
in the form of multicomponent arrays are used ties.
INFRARED SPECTROSCOPY 615
help of spectrophotometers, or tunable sources of the case of spectral diffusion. The second mech-
monochromatic infrared radiation such as lasers. anism makes it equally probable for an excited
spin to enter any packet. The “burned-out hole”
INHOMOGENEOUS BROADENING does not broaden with time; but as the microwave
Spectral line broadening resulting from the power grows, the whole line saturates evenly.
spread of resonance frequencies from various The spread of resonance spin frequencies that
atoms. When resonance frequencies lie close to produces inhomogeneous broadening may be re-
each other there appears an unresolved spectrum, lated to microscopic inhomogeneities of the crys-
with an envelope having the shape of an in- tal, to the presence of defects and impurities (all
homogeneously broadened line. This shape is a result in random changes of the internal field), or
convolution of a uniformly broadened line, i.e. a to a hyper ne interaction of electrons with spins
shape obtained by integrating the spectral line pro- of the surrounding nuclei. Nonuniformity of the
duced by an isolated atom or a group of atoms external magnetic field, the polycrystalline nature
with the same resonance frequency over a spec- of the sample, the presence of slightly disoriented
tral distribution of resonance frequencies. In the macroscopic blocks in an imperfect monocrys-
case of magnetic resonance, individual compo- tal (so-called mosaic structure), and the dipolar
nents of the line contour are called spin packets, interaction between spins with differing Larmor
and a set of spins in identical local fields con- frequencies (see Larmor precession) all result in
tributes to each such packet. The model of spin inhomogeneous line broadening. To calculate the
packets is particularly useful when describing in- shape of a spectral distribution of resonance fre-
teractions of a spin system with a microwave field, quencies, the statistical theory of line shape is
its frequency ωp fitting into the contour of an sometimes used. See also Homogeneous broaden-
inhomogeneous line of electron paramagnetic res- ing.
onance. According to such a model, only spins
belonging to the packet of frequency ωi = ωp INJECTION of current carriers
are directly affected by the field. If one scans Introduction of nonequilibrium (excessive) cur-
a strongly saturated electron paramagnetic reso- rent carriers into a semiconductor (insulator) from
nance line using an additional weak microwave an external source that could be either a contact
frequency there appears a characteristic dip or (semiconductor or metal contact) or an incident
trough in that line, a “burned-out hole” centered at electron beam. The remaining semiconductor bulk
the frequency ωp (see Hole burning). The width of may be activated by a limited region (usually a
the hole depends on the extent of the overlap of the subsurface one) of enhanced generation of carri-
neighboring packets, and on processes spreading ers by light, fields, etc.
the saturation along the overall inhomogeneous Injection from a contact arises when an exter-
line contour. Two mechanisms for spreading sat- nal electric field reduces the contact potential bar-
uration are cross-relaxation or spectral diffusion, rier compared to its equilibrium value that is the
and packet-to-packet spin “jumps” due to fluctua- contact potential difference. When two solids with
tions of local fields. In the first case, the saturation different work functions Φ1 and Φ2 (Φ1 > Φ2 )
is transferred from the “burned-out hole” first to are in contact, there is an exchange of electrons.
packets closest in frequency, and then it spreads The subsurface layers of both bodies become
further by the process of diffusion. The role of the charged, the solid with Φ1 being charged neg-
characteristic length in frequency space is played atively and the solid with Φ2 positively. These
by the quantity ΔD = (T1 Dω )1/2 , where T1 is the charges produce a potential barrier whose height
spin–lattice relaxation time, and Dω is the spec- at thermodynamic equilibrium is (Φ1 − Φ2 )/|e|.
tral diffusion coef cient. The diffusion spin length The equilibrium barrier field produces drift fluxes
ΔD defines the frequency range over which the of electrons and holes that compensate exactly
spin excitation spreads within its lifetime (until for their diffusion fluxes. If the external field is
absorbed by the lattice). This value also deter- directed against the barrier field, it reduces the bar-
mines the half-width of the “burned-out hole” in rier height so the drift fluxes become less than the
INJECTION LASER 617
diffusion ones, thus inducing the flux of excessive carrier concentration. Therefore, in these semicon-
electrons directed into the solid with Φ1 , while ductors efficient nonstationary injection is pos-
those of holes are directed into the solid with Φ2 . sible even for quite slow processes; such injec-
Injection efficiency of the contact with a semicon- tion noticeably increases the material conductivity
ductor is described by the injection factor. This is compared to its stationary value. The excessive
the ratio between the current of carriers entering carriers are able to penetrate into the semiconduc-
the semiconductor from the contacting conduc- tor depth up to the length determined by current
tor through the near-contact potential barrier and carrier drift during trapping.
the total current through the contact. The injec- If one of the contacts supplies the semicon-
tion factor of a perfect injecting contact is close ductor with excessive electrons and the other one
to unity. with holes, then bipolar (double) injection occurs.
Under monopolar injection from the contact In this case, the excessive minority carriers intro-
an excess of charge carriers in the semiconduc- duced into the semiconductor stimulate, by their
tor bulk grows with time. It produces a field field, the influx of excessive majority carriers up
that precludes further penetration of carriers into to almost complete neutralization of their charge.
the semiconductor, and it also makes the poten- Due to the lack of charge hindering the injec-
tial barrier higher and reduces the injection. The tion, efficient stationary double injection is easily
charge stops growing when the injection ceases achieved even at high semiconductor conductivity.
completely. If there is a second contact through The excessive carriers under stationary conditions
which the carriers can drain from the semicon- can also penetrate into the semiconductor depth
ductor then stationary monopolar injection is pos- a long distance from the contacts, their path be-
sible. Its level is controlled by the competition ing determined by their lifetime. The penetration
between the injection-inducing external field due depth in weak fields is equal to the ambipolar
to the voltage applied to the semiconductor, and diffusion length; this depth in strong fields is de-
the anti-field produced by the excessive carrier termined by the ambipolar drift.
charge. Hence, in a semiconductor with a high One should differentiate injection from a re-
free-carrier concentration one can obtain a high- lated phenomenon called accumulation. The oper-
level injection (significant concentration growth ation of most semiconductor devices is based on
in the whole semiconductor) only for strong ex- injection.
ternal fields in samples with very small distances
between the contacts. If the equilibrium concen- INJECTION LASER
tration is small due to low doping, the conditions The commonest kind of semiconductor laser
needed to produce stationary monopolar injection (diode laser, laser diode). Its particular feature is
are made easier. current carrier injection through a nonlinear elec-
This conclusion is, however, not valid for semi- trical contact (e.g., the semiconductor junction)
conductors with a high charged impurity concen- used as the pumping mechanism. The advantages
tration (i.e. donors and acceptors in equal numbers of this type of laser are the small size of its ac-
compensating each other); here low free-carrier tive element, simplicity of its design, economy of
concentration is due to weak generation of cur- operation (efficiency up to 50%), possibility of di-
rent carriers from impurity levels to allowed bands rect modulation of radiation over a wide frequency
(at low temperature and/or due to high impurity range (∼1010 Hz), simplicity of frequency tun-
depth). For semiconductors with low generating ing and selection of the active medium to cover a
rates the appearance of even small (comparable to wide wavelength range (0.6–45 μm), durability (to
equilibrium) concentrations of excessive carriers 106 h), low-voltage power supply (1–2 V), com-
leads to the accumulation of a large charge of ex- patibility with other semiconductor devices, and
cess carriers at impurities due to the increase of possibility of its incorporation in integrated cir-
the trapping rate; and this charge inhibits injec- cuits. A simple version of an injection laser is
tion. The accumulation of a large impurity charge a crystal diode (Fig. 1) with two plane parallel
is very prolonged, running slower for a lower free faces forming a plane laser resonator. So-called
618 INJECTION LASER
rigid body, and usually aligned along a principal high-voltage generators, transformers, cable man-
axis of the body. The totality of instantaneous axes ufacturing, etc. Examples of good electrical insu-
forms cones in both spaces; the motion of the solid lating materials are resin, wax, varnish, various
being a result of the rolling of the body cone about polymeric materials, ceramics, mica, and so on.
the space one, the two cones being in contact along The presence of defects in a real dielectric together
instantaneous axes at every instant of time. with thermal excitations determine the final value
of resistance, and hence there can exist a small cur-
INSULAR FILMS, granular lms rent leakage through an insulating material. The
Systems of non-contacting small particles on resistivity of electrical insulating materials reaches
a substrate, obtained by spraying or by disper- 1018 –1019 ·cm. These materials are also speci-
sion within thin continuous films as a result of fied by the value of their dielectric strength, that is
heating. Of interest are electrophysical, emission, the critical electric field Eth at which breakdown
optical, and magnetic properties of granular or in- occurs. The value of Eth in solid electrical insula-
sular films. The electrical conductivity of these tors is 104 to 105 V/cm. Special requirements are
films, which is of a tunneling nature, is sensitive imposed on the insulating materials employed in
to minor changes in the distances, and the potential microelectronics (sapphire, quartz, etc.) in connec-
barrier height between the islands or grains. This tion with very thin insulating layers (≈10−6 cm).
property is employed to produce strain gauges, va-
por pressure pickups for various substances, and INSULATOR POLARIZATION
temperature-sensitive elements. By changing the
See Polarization of insulator.
substrate and island materials, and the morphology
of the islands, one can obtain positive, negative or
INSULATORS, dielectrics (fr. Gr. δια , through,
even zero temperature coefficients of resistance.
and ηλεκτρoν , amber)
The system of islands covered by the adsorbate
Term introduced by M. Faraday for substances
can provide an N-shaped current–voltage charac-
into which an external electric field penetrates. If
teristic, and thus switching and storage elements.
the material contains free current carriers the ap-
The films where the island size varies monotoni-
plied electric field E causes them to move, and
cally along a coordinate axis exhibit conductivity
they establish a spatial charge distribution that
asymmetry. Small metal particles in close thermal
contact with an insulator substrate are capable of compensates for the field within the conductor.
withstanding high power flux without destruction. In the absence of the free carriers only partial
This property, in combination with the reduction compensation of the field is possible, caused by
of the magnitude of the electron–phonon interac- small displacements of the bound charges (see
tion as the particle size decreases, allows heating Polarization of insulator). Thus the other defini-
the electron gas of small metal particles to in- tion of insulators as materials with low electrical
cite hot electrons. The appearance of hot electrons conductivity is closely connected with the first
(during passage of a current or under laser irradi- one. Materials with resistivity in the range 108 to
ation) causes electron emission (see Hot electron 1017 ·cm (i.e. insulators) are referred to as di-
emission) and glow of such films. During the tran- electrics in the Russian literature. Many solids as
sition from continuous films to granular or insular well as liquids and gases possess the properties
ones, optical constants change, the dynamic po- of insulators. In real insulators there are always
larizability increases, and the magnetooptical effect some free charge carriers, but at their low concen-
becomes stronger. tration their screening of the electric field appears
to be weak. Low concentrations of free electrons
INSULATING MATERIALS, electrical insulating in crystals with large band gaps E g 3 eV (i.e.
materials E g
k B T ) produce a low value of the con-
These are insulators, or dielectric materials ductivity. The efficiency of compensation of the
with very high resistance, which are used to isolate external field E in insulators is determined by
conductors electrically, as well as capacitor ele- the dielectric constant (also called permittiv-
ments in electrical circuits. They are employed in ity) ε of the substance; thus in Coulomb’s law
INTEGRATED OPTICS 621
F = e1 e2 /(4πεr 2 ) the charges e, placed into the substrate using layer-by-layer planar technology
dielectric medium, could be replaced by effective (see surface smoothing) to form all the elements
√
charges e∗ = e/ ε/ε0 . into a single technological unit. Layers with a
Based upon the polarization mechanisms insu- given topology of circuit components (transistors,
lators are subdivided into three classes: condensers, diodes, resistors, etc.) are sequen-
tially formed on the surface of crystals made of
(1) nonpolar insulators, where only polarization
materials with various electrical properties, and
of the electronic shells is present (crystals of
are arranged under one another. Combinations of
O2 , inert gases, diamond, etc.). At frequencies
such layers produce various circuit elements. Af-
far from those of resonance absorption ε does
ter forming the multilayer structure the circuit
not depend on the frequency and ε/ε0 = n2 (n
elements are interconnected with a surface con-
is the refractive index);
ductive layer of the required configuration. The
(2) ionic insulators, which consist of oppositely
basic material for chip production is silicon single
charged ions capable under the effect of the
crystals as well as gallium arsenide single crys-
electric field to undergo only small displace-
tals featuring high carrier mobility. Materials used
ments from their equilibrium positions (alkali-
as insulator layers include silicon dioxide SiO2 ,
halide crystals, oxides, sulfates, metal phos-
sapphire Al2 O3 , spinels AB2 O4 , iron garnets, and
phates etc.). For them there is a significant dif-
others. According to the application, the IC can be
ference between the static and high-frequency
divided into digital and analogue types. Depend-
dielectric constants;
ing on the packing density (number of devices
(3) polar insulators or dipole dielectrics consist-
per unit area on a crystal), the integrated circuits
ing of polar molecules (having intrinsic dipole
can be classified into small-scale IC containing
moments). Another class of crystals has two
10–12 devices on a crystal with the surface area
or more different equilibrium positions for
within a mm2 or cm2 ; mid-scale IC containing
ions in the unit cell. This group contains wa-
12 to 100 devices; large-scale IC containing more
ter (in the solid and liquid state), solutions
than 100 devices; very-large-scale IC that con-
of polar molecules, electrolytes; materials of
tains 10 000 and more digital devices. The ICs
the second group with lattice defects, etc.
are used for microminiaturization of radioelec-
For these dielectrics the value of ε is associ-
tronic many-purpose systems like microcomputers
ated with the dipole moment by the Onsager
for automatic control (in space complexes, in ro-
relations (see Onsager theory). At low tem-
botics) as well as for data transmission and data
peratures reorientation of the dipole moments
processing. See also Microelectronics.
is “frozen” and the behavior of the dielectric
constant becomes similar to the behavior of ε
INTEGRATED OPTICS
for ionic insulators.
A branch of optoelectronics oriented to de-
Dielectrics or insulators of the types of ferro- signing solid-state devices for optical information
electrics, piezoelectrics, pyroelectrics (see Pyro- processing that are manufactured with the help of
electricity) possess their own specific properties. planar technology (see Surface smoothing) devel-
Insulators are used in quantum electronics, in par- oped for integrated circuits. The application areas
ticular, in optoelectronics for the creation of the of these integrated optics are optical selectors,
active laser elements, in holographic devices, in retransmitters, commutators, spectral multiplexes
infrared techniques, in capacitors; as insulating of optical channels in fiber-optic communication
materials etc. (see also Dielectronics). lines, spectral analyzers in radio-beam systems,
optoelectronic and fiber-optic sensor devices, in-
INTEGRATED CIRCUIT tegrated optoelectronic circuits, planar optoelec-
A tiny solid-state electronic device produced tronic instruments and systems.
by incorporating a given arrangement of circuit The constructive base of integrated optics (see
elements in a single microcrystal. An integrated Fig. 1) is the dielectric waveguide (DW) that can
circuit (IC) is manufactured on a semiconductor be planar (a), three-dimensional (b) or stripe (c).
622 INTEGRATED OPTICS
Three-dimensional dielectric waveguides provide (0.1–0.08 dB/rad for glasses), which restricts
additional limitation of the optical radiation in the the permissible relative bending radii R/λ that
traverse direction. This increases the integration are the greater, the smaller
n/n2 . Level fre-
level and efficiency of integrated optics circuits quency and phase characteristics of a dielectric
and simplifies the interfacing of the dielectric waveguide are easiest to achieve when using a
waveguides to other integrated optics components single waveguide type of transmission. Single-
and the fiber-optical cable (see Fiber optics). The and few-mode regimes limit the transverse di-
dielectric waveguide is an optical line with low mensions of a dielectric waveguide (its format)
loss (1 dB/cm) at the operating radiation wave- to the order of the of transmitted radiation wave-
lengths (from 0.62 μm to 1.6–2 μm and more). The length (0.3–3 μm in height, 1–5 μm in width),
waveguide modes, e.g., the fundamental modes of and they should be maintained within the error of
integrated optics devices, are solutions to the wave a few hundredths of μm. A particular waveguide
equation mode is ensured by preset refractive index profiles.
There are dielectric waveguides with stepwise and
∇ 2 E(r) + k02 n22 (r)E(r) = 0, k0 = 2π/λ0 , continuous (graded) refractive index variations.
in the form In a gradual microwaveguide it is easier to pro-
vide single-mode operation that results in much
E(r, t) = E(x, y) exp i(ωt − βz) , less loss. At a given frequency the number of
where β = ω/v ph is the propagation constant; ω waveguide modes is limited by the relation
and v ph are the angular frequency and the phase 1/2
m 1/a + ah/λ0 2n3 (n2 − n1 ) ,
velocity of the electromagnetic wave; n2 is the re-
fractive index of the dielectric waveguide carrier where a = 2 and 4 for the stepwise and gradient
layer; and λ0 = 2πc/ω is the wavelength in free dielectric waveguides, respectively.
space. A waveguide mode with the field varying For passive applications there are dielectric
harmonically inside the dielectric waveguide and waveguides on glass substrates where a given pro-
decaying exponentially in a transverse direction file n2 (x, y) is produced with the help of the
outside of the waveguide satisfies the condition ion exchange method, ion implantation, hydro-
n2 < n1 , n3 (see Fig. 1). The absence of the ra- lysis, or the deposition of thin lms (including
diation in the substrate (as a rule, n3 n1 ) for a chalcogenide and polymeric ones). To control the
given mode configuration is controlled by the fol- light beam, the waveguide material must have
lowing inequality: certain electrooptic, acousto-optic, and other ac-
tive properties determining the dependence of its
(2m − 1)2 λ0 refractive index on the controlling action. Ac-
n = n2 − n3 > ,
32n2 n21 tive substrates in dielectric waveguides are ex-
emplified by a crystal of lithium niobate with
whereby we have
n ≈ 10−2 –10−3 for the ba- the required refractive index profile in the crys-
sic modes and λ0 = 1 μm. The bending of a tal produced by the diffusion of Ti (transmis-
dielectric waveguide introduces some extra loss sion range is λ 3 μm, with SiO2 antireflection
INTERATOMIC INTERACTION POTENTIALS 623
For covalent and molecular crystals interatomic empty bands becomes possible. The probability
3/2
potentials are also obtained on the basis of pseudo- of tunneling is proportional to exp[−πm1/2 Eg /
potential theory. However now it is necessary to (2eE )], where m is the reduced effective mass of
introduce additional assumptions (e.g., the model the electron and hole, Eg is the width of the band
of charges on bonds) in order to take into ac- gap, and e is the electron charge. In parallel with
count the directionality of the interatomic bonds tunneling, electron oscillations also occur inside
in these crystals. From an empirical approach, in the band, caused by reflection of electrons from
addition to the pairwise potentials three-particle band boundaries (see Quantizing electric eld).
terms are introduced, depending on the angles be- In strong fields, interband tunneling may cause
tween the bonds. By combining different angular breakdown (Zener breakdown). Band-to-band tun-
functions with the previously considered pairwise neling provides the basis for tunnel diodes, where
potentials it is possible to obtain (usually in the an electric field is induced in the region of the
harmonic approximation) the potential energy of semiconductor junction.
a crystal lattice: Keating potential, Vukcevich po-
tential (M.A. Vukcevich). This facilitates the cal- INTERCALATED STRUCTURES
culation of properties of crystals that depend on Crystals of a laminated type with intercalants,
small displacements of atoms from their equilib- i.e. extra atoms, atomic groups, or molecules,
rium positions (elastic constants, phonon spectra). introduced between the layers. The most com-
The many-parameter potential functions for de- monly studied intercalated structures are com-
scribing the transition to the liquid state can be pounds based on laminated dichalcogenides, tran-
found by the same empirical approach. sition metals compounds, and graphite-based com-
pounds. Typical examples of the first type are the
INTERBAND TUNNELING, band-to-band metals K0.4 MoS2 , Fe0.05 TaS2 , Cax (NH3 )y MoS2 ,
tunneling, Zener tunneling (C. Zener, 1934) and TaS2 (Py)1/2 where Py is the pyridine
Tunneling of electrons from the valence band (C5 H5 N) molecule. The source crystals are semi-
to the conduction band of a semiconductor, which conducting MoS2 and metallic TaS2 or NbSe2 .
takes place in a strong electric field. The energy Graphite is also easily intercalated by metals and
diagram of a semiconductor, shown in Fig. (a), molecules. Intercalated compounds possess phys-
changes in an electric field: the electron energy ical properties that often differ greatly from those
is augmented with the electrostatic energy of the of source crystals and intercalants due to elec-
applied field E . As a result, the plot of the elec- tron transfer from the intercalated atoms to the
tron energy E versus the coordinate x becomes layers. For example, MoS2 is a semiconductor,
inclined as shown in Fig. (b), and electron tunnel- but upon intercalation with alkali metals it be-
ing (see Tunneling phenomena in solids) between comes a metal, and superconductivity has been
states of equal energy in nearby occupied and observed below 6 K (see Superconductivity). In
a similar way, neither graphite nor alkali met-
als exhibit superconductivity, but the intercalated
compounds superconduct below 5.5 K, due to the
appearance of new electron states occupied by
electrons that had left the alkalis. At the same
time, transition metal dichalcogenides intercalated
by molecules exhibit very little change in their
properties if these properties arise from electron
motion inside the layers. However, as a rule, in-
tercalated compounds differ from source crystals
by a stronger anisotropy of the electronic and me-
chanical properties. For example, in TaS2 (Py)1/2
the electrical conductivity across the layers at 20 K
Interband tunneling. drops by 5 orders of magnitude compared to the
626 INTERCALATION
source crystal TaS2 , while this conductivity along for graphite, corresponding to different stages of
the layers is only two times lower due to re- intercalation.
duction of the carrier concentration at the layer
spacing. Therefore, intercalated compounds pro- INTERCRYSTALLINE CORROSION
vide so-called quasi-two-dimensional systems (see Selective corrosion of boundary layers of grains
Quasi-two-dimensional crystals) where the elec- (crystallites) of metallic materials, which spreads
tron motion is close to being two-dimensional. inside the metal along intercrystalline bound-
The intercalated structures are also interesting aries; a kind of local corrosion. The tendency to-
from the physical viewpoint since they provide the ward intercrystalline corrosion often results from
potentiality to obtain new superconductors with a heating which causes microstructural transforma-
possible hypothetical exciton mechanism of super- tions along grain boundaries. Extra phases (often
conductivity. These structures admit the existence chromium-based carbides) are generated, enriched
of combinations of layers with different electronic with certain alloy components, and next to them
properties in the same crystal, e.g., metallic layers bands of solid solution are formed, which are de-
associated with semiconducting ones, or super- pleted of the above-mentioned components (often
conducting layers with magnetic ones (Fe0.05 TaS2 Cr). The most typical intercrystalline corrosion
is such a system). In this sense these structures are occurs near welding seams (so-called knife-line
similar to metallic superlattices. corrosion). Depending on the properties of the
hostile medium and the chemical composition of
INTERCALATION structural components along the grain boundaries,
Introduction of extra atoms, or atomic groups, intercrystalline corrosion may develop in the form
or molecules (intercalants) into the space be- of selective corrosion of either depleted bands, ex-
tween crystal layers. Usually intercalation is found cessive phases, or both. The mechanical properties
in crystals with laminated structures and weak of a metal that shows a tendency toward inter-
chemical or van der Waals binding of layers crystalline corrosion, or has been exposed to it,
(see Van der Waals forces). Easily intercalated become degraded.
structures are exemplified by dichalcogenides of
transition metals of the TaS2 type with mainly INTERCRYSTALLINE FAILURE
van der Waals interactions between the layers, Nucleation and propagation of the main crack
as well as by graphite. The intercalation process (Griffith crack, see Grif th theory) along bound-
could be carried out by immersion of the source aries of adjacent crystallites. Grain boundaries or
crystal into the gas or liquid phase of inter- cell boundaries (see Cellular structure), and those
calants. During the intercalation procedure, the of twins (see Twinning of crystals) and marten-
layers of the original crystal move apart to ac- sitic crystals, could serve as interfaces between
commodate the intercalants, and the expansion crystallites. The best studied is the failure along
can be appreciable. For example, in the com- grain boundaries in polycrystalline alloys. In lim-
pound TaS2 (octadecylamine)1/3 the TaS2 layers iting cases, intercrystalline breakdown may occur
are 0.3 nm thick, are spaced 0.3 nm apart, and in either a brittle manner (see Brittle failure), i.e.
separate to a distance of 5 nm to accommodate without formation of a strain relief on the surface,
two layers of decylamine linear molecules 2.5 nm or in a tough or viscous manner (see Tough fail-
long. The intercalation process is energetically fa- ure) according to the mechanism of nucleation and
vorable because the bond between the intercalants merging of pits over a sufficiently wide range of
and the layers proves to be stronger than that of temperatures and strain rates. The physical cause
the layers between each other. Ordinarily this is of intercrystalline breakdown is the decrease of
due to partial or complete transfer of electrons metal–vacuum surface energy γ0 during crack for-
from the intercalants to the layer. The intercalants mation at an interface in a polycrystal by the value
form regular structures between the layers, with γb of the energy of existing boundaries. In the
the structures depending on the extent of the inter- presence of surface-active impurities (see Surface-
calation. For example, many structures are known active agents) or doping agents (see Doping) that
INTERDIFFUSION 627
cause weakening of the binding forces at bound- increase of the interstitial impurity content (or
aries, the quantity γ0 changes to γc . Here, γc is change of limit of possible oversaturation at heat
the surface energy of the crack that has crossed the treatment or thermomechanical treatment) causes
interface which existed in the crystal, and where the formation of disperse structures. The latter
segregations of dissolved atoms have significantly hinder the nucleation of intercrystalline cracks
changed γ0 and, accordingly, the surface energy which results in a decrease of Tc.g. (from A
b = γ − γ /2.
of the intercrystalline failure γeff c b to 2). Along line segment 2 to 3 brittle fail-
The impurity doping atoms nitrogen and oxygen in ure either reverts to transcrystalline (dashed curve
body-centered cubic transition metals noticeably with crossover in Fig.) or remains intercrystalline
reduce the value γ0 . In a similar way, phospho- (upper solid line curve with no intersection),
rus and sulfur affect iron, sulfur affects nickel but the cold brittleness temperature unambigu-
and copper, etc. In deformed metals the inter- ously decreases. A pronounced rise in temperature
crystalline breakdown along cell boundaries (see takes place at concentrations above point 3 when
Cellular structure) can result from the increase
structure-free layers of excessive brittle phase
of γb due to the increased density of unrelaxed
b de- (interstitial phase, intermetallic compounds, etc.)
dislocations at the subboundary. The value γeff
appear at the grain boundaries, these layers hav-
creases sharply in this case. The strength σcr under
ing very high cold brittleness temperatures. The
intercrystalline breakdown is described by the de-
b /d, where β is a constant; formation of continuous interlaying of the second
pendence σcr βεγeff
ε is the strain related to the amount of mobile dis- phase along grain boundaries embrittles the al-
locations; and d is the structural element size. loy in a deleterious manner. At the decline of γc
b → 0), failure occurs below the yield limit
(γeff
The relationship between the cold brittleness
temperatures under transcrystalline failure and of the alloy. This phenomenon is called inter-
intercrystalline failure can be illustrated by a di- crystalline brittleness (such as oxygen-induced
agram (see Fig.). Alloying the solid solution causes brittleness of polycrystalline molybdenum alloys,
a monotonic increase of the cold brittleness tem- steel temper brittleness, antimony-induced brittle-
perature Tc.g. of the host grain. At small contents ness of copper, etc.). Doping and heat treatment
of surface-active agents (especially interstitial im- can stimulate impurity redistribution between the
purities of body-centered transition metals), the grain bulk and the grain boundary, thus chang-
sharp increase of the cold brittleness temperature ing both the effective surface energy of formation
Tc.b. occurs near boundary regions due to the ap- of an intercrystallite crack, and the conditions for
pearance of segregation (Fig. from 1 to A). The crack relaxation. This, in turn, causes a change in
the cold brittleness temperature at the boundaries
and in the bulk of the grains.
INTERDIFFUSION
Mutual penetration of atoms and molecules
across an interface of contact between two mate-
rials (solid, liquid, gas) of different compositions,
caused by thermal motion of the atoms or mole-
cules.
Interdiffusion is established with respect to
the displacement of the front of constant con-
centration. Because of counter-diffusion flow the
mechanical equilibrium in the system is disturbed,
but then it is immediately restored by the mechan-
Sketch of variation of temperatures of intercrystalline ical shift of layers, and this reestablishes the front
(Tc.b. ) and transcrystalline (Tc.g. ) embrittlement under of constant concentration (see Kirkendall effect).
doping with carbon. The theory of interdiffusion for two-component
628 INTERFACE
in a superconductor was clarified by L.D. Landau place only partially throughout the material. Then,
(1937), who demonstrated that it represents a ther- as the external field B 0 gradually increases, the
modynamically stable domain structure consisting value of magnetic induction, B, inside the sample
of alternating patches of metal in the normal and linearly grows from zero at B0 = Bc∗ (at the transi-
the superconducting states. tion from the superconducting to the intermediate
The intermediate state has been found and state) to B c when B0 = Bc (at the transition
studied in the region of first order phase transitions from the intermediate to the normal state). Mean-
in antiferromagnets, as well as in ferrite spinels while, accounting for demagnetization eld, we
and orthoferrites, where it develops during various have Bc∗ < B0 < Bc for the whole intermediate
magnetic phase transitions. state range, with the Maxwell field inside the spec-
imen H = Bc /μ0 . The value of Bc∗ depends on the
INTERMEDIATE STATE OF value of Bc and the shape of the material.
SUPERCONDUCTORS For an arbitrary sample shape the normal and
Thermodynamically stable state of a type I the superconducting phases may coexist because
superconductor of finite dimensions in a certain of the inhomogeneity of the demagnetizing fields.
range of external magnetic fields. Such fields have However, they are always separated from each
values close to those that bring about field induced other by an intermediate state region. In the case
phase transitions back to the normal state. It is a of an ellipsoid placed in a uniform external field,
structure consisting of alternating macroscopic do- the induction and the field inside it are always uni-
form. Therefore, the entire ellipsoidal specimen
mains of the normal and superconducting phases
transforms into the intermediate state as a whole,
in the shape of plates parallel to the external field,
with a single value of H = Bc∗ /μ0 through-
with a characteristic transverse size much smaller
out it. When the direction of B 0 coincides with
than the specimen size. Such a structure forms due
the ith principal axis of the ellipsoid, the direc-
to the positive free surface energy.
tions of B and H also coincide with it, so that
When affected by the magnetic field produced
Bc∗ = Bc (1 − Ni ), where Ni are the dimensionless
by current flowing through the specimen, a dy-
demagnetizing factors (i = 1, 2, 3) (see Demag-
namic intermediate state develops in which the
netization factor tensor). Considering the extreme
domains are normal to the direction of the current cases of the field parallel to the axis of a long
flow, and move continuously through the speci- cylinder (N = 0) we find the intermediate state
men. They are generated at one of its surfaces and virtually absent, while in the case of the field nor-
end at the opposite one. mal to a flat plate (N = 1) the intermediate state
When a superconducting sample is placed in develops starting from a close to zero value of B0 .
an external homogeneous field B 0 , that field is The above description explains the basic exper-
pushed out of the sample due to the Meissner ef- imental results, but says nothing about the inter-
fect. The field force lines then concentrate at the mediate state structure, and it fails to clarify the
surface meaning that the field in that range exceeds physical content of H and B inside the specimen.
B 0 and may reach the value of the thermody- The Landau (1937) theory of the intermediate
namic critical magnetic eld, B c , where B0 < Bc . state represented the macroscopic field H inside
In such a case the superconducting state should the specimen as the result of averaging micro-
start to disintegrate in the neighborhood of the scopic fields over the normal phase domains, while
maximum concentration of flux lines, with the the induction, B, was arrived at by similar aver-
concomitant penetration of the magnetic field into aging over the entire specimen. The fraction of
the specimen. However, the normal phase ap- the normal phase in the intermediate state equals
pears to remain in the range of magnetic field (B0 − Bc∗ )/(Bc − Bc∗ ). The spatial period d of the
strengths that are lower than the critical field, domain structure is determined by minimizing the
so there is an apparent contradiction. To over- free energy of the magnetic field forced outside the
come it R. Peierls (1936) and F. London (1936) specimen by the Meissner effect, plus the surface
both independently suggested that the intermedi- energy of the domain walls, so there is a com-
ate state is one in which the Meissner effect takes petition between the bulk term [Bc2 /(2μ0 )]
Vn
INTERMEDIATE VALENCE 631
equalizes the valences. The magnetic properties other special materials. They are involved in com-
are defined by the relation between the Heisen- positions of the most important alloys such as
berg exchange interaction (that results in either structural, refractory, hard, high-temperature, and
ferromagnetism or antiferromagnetism) and double corrosion-resistant alloys. See also Superconduct-
exchange. For strong double exchange the cluster ing intermetallic compounds.
has a maximum spin, as shown in Fig. 2 for a two-
ion cluster with total spin S = S 1 + S 2 obtained INTERMETALLIC SUPERCONDUCTOR
by vector addition. The interaction among clus- See Superconducting intermetallic compounds.
ters may stabilize the charge-ordered phase, and
two phase transitions become possible under cer-
tain conditions, the ordered state forming within a INTERNAL ENERGY
finite temperature range T > 0. See Thermodynamic potentials.
Stress tensor). At the free surface of a solid the Particularly useful is Vol. A edited by T. Hahn
internal stresses satisfy the boundary conditions which tabulates atom positions for the 230 space
σj k nk = 0 (i, k = 1, 2, 3), where nk is the pro- groups, and in addition has a great deal of back-
jection of the external normal to the surface on ground material on crystallography.
the coordinate axis xk . If within the volume of
the solid there are stresses caused by static ex- INTERPHASE IMPURITY DISTRIBUTION
ternal forces, they add to the internal stresses. COEFFICIENT
They can appear in solids under the action of The ratio of the concentration of a substance
mechanical and thermal effects due to the non- dissolved in a solid phase to that in a liquid phase.
uniformity of plastic deformations, and due to the The equilibrium interphase impurity distribution
possible occurrence of phase transitions. Internal coefficient is defined using a phase diagram as
stresses which appear after such processes are fre- the ratio of abscissae of the solidus and liquidus
quently called residual stresses. These stresses points for a given impurity concentration. If this is
can have different degrees of localization, i.e. they smaller than 1, then the solid phase that forms as a
can be counterbalanced within different regions result of crystallization contains fewer impurities,
of a solid. In polycrystals the following types of and vice versa. In actual crystallization schemes,
internal stresses are distinguished: (1) first-order the ratio of impurity content in the crystal to that in
residual stresses which are balanced within the the melt is determined by the interphase impurity
overall volume of a solid, and produce a shift of distribution coefficient that depends on the ratio of
an X-ray diffraction line, (2) second-order resid- the rate of the crystallization front motion to the
ual stresses (microstresses) which are balanced rate of equalization of impurity concentration in
within a volume about the size of a crystallite or the melt. The latter process occurs by the diffuse
a mosaic block and broaden an X-ray diffraction transition of impurities in a thin stationary layer
line, and (3) third-order residual stresses which are of melt that adjoins the crystallization front, and
balanced within a volume the size of a crystalline by forced or natural convective agitation of the re-
lattice unit cell and weaken X-ray diffraction lines. maining melt. Allowing for the above factors, the
X-ray diffraction methods (see X-radiography, distribution coefficient can be calculated for actual
Nondestructive testing techniques), and also (for crystallization schemes.
first-order stresses) various mechanical methods
based on the measurement of deformations, with INTERSTITIAL ALLOYS
subsequent recalculations that convert them into Alloys, one component of which consists of
stresses with the aid of Hooke’s law, are used interstitial atoms. These alloys form for the ra-
for the determination of residual stresses. These tio r x /r m 0.59 of atomic radii of interstitial
(especially the first-order ones) can strongly influ- element r x to metal r m . Interstitial phases are di-
ence the strength of solids. vided to two classes: interstitial solid solutions
of metals, and interstitial compounds or hybrids:
INTERNATIONAL TABLES FOR carbides, nitrides, and some borides, oxides and
CRYSTALLOGRAPHY silicides.
Reference guides for crystallography and struc-
ture analysis. “International Tables for Determina- INTERSTITIAL ATOM
tion of Crystal Structures” were first published in An atom that is localized in a crystal between
1935 (Berlin). The second generation comprises crystal lattice sites. It is common for an interstitial
four volumes published between 1952 and 1974. atom to be situated at an interstice, i.e. in a cavity
Since 1983, the third generation “International Ta- bounded by neighboring atoms at regular crystal
bles for Crystallography” has appeared as volumes lattice sites. However, in a number of cases more
A, B and C (Space Group Symmetry; Reciprocal complicated configurations of an “extra” atom in
Space; Mathematical, Physical and Chemical Ta- a crystal can occur, such as when the atom in-
bles, respectively) which provide comprehensive trudes into an atomic chain so that there are n + 1
coverage of the status of modern crystallography. atoms for n sites. These are dumb-bell interstitial
INTERVALLEY REDISTRIBUTION 635
defect (n = 1) and crowdion (n ≈ 5–10) configura- anisotropic. The partial currents of these groups
tions. Intrinsic interstitial atoms appear in crystals may then be nonparallel, and intervalley noise in
as non-stoichiometric components (see Stoichio- such crystals arises in the direction transverse to
metric composition). This may happen, e.g., when the direct current. The dependence of intervalley
a crystal is in thermodynamic equilibrium with noise on the orientation of the electric field rel-
the high-pressure vapor (see High pressures) of ative to the crystallographic axes allows varying
one of its components, or through the generation the magnitude of intervalley noise, and facilitates
of equilibrium or nonequilibrium (during nuclear its identification in experimental studies.
radiations, elastic strain) Frenkel defects. Impurity
atoms of certain chemical elements can be lo- INTERVALLEY PHONONS
calized in the interstices of various crystals. The Phonons involved in interaction electron tran-
following processes can take place: transitions of sitions from one valley to another (see Many-valley
interstitial atoms between different types of in- semiconductors). By virtue of quasi-momentum
terstices (e.g., between hexagonal and tetrahedral conservation (see Conservation laws) within an
ones in a diamond-type lattice); trapping of an in- accuracy of a reciprocal lattice vector, a phonon
terstitial atom by a vacancy; dislodgment of an quasi-wave vector which participates in an el-
impurity atom from a site by an intrinsic atom; ementary transition between valleys is equal to
coagulation of intrinsic interstitial atoms with the k phonon = ±(k 1 − k 2 ), where, k 1 and k 2 are the
formation of dislocation loops; and other reac- quasi-wave vectors of the initial and final electron
tions that involve intrinsic and impurity interstitial states, the “+” sign refers to emission of a phonon,
atoms, as well as defects. The activation energy of and the “−” sign refers to absorption of a phonon.
intrinsic interstitial atom diffusion is usually much Since valleys are often situated near the Bril-
lower than that of vacancies. The non-activation louin zone boundary (e.g., in germanium, silicon),
nature of the motion of interstitial Si atoms gen- the values of |k 1 | and |k 2 | and, accordingly, of
erated via nuclear radiation on a silicon crystal
|k phonon | are close to the maximum value of a
has been experimentally confirmed at low temper-
quasi-wave vector in the Brillouin zone. Such a
atures; the same is true of interstitial Ge atoms
value of |k phonon| corresponds to the energy of
in germanium crystals. The presence of interstitial
intervalley phonons, which is close to the peak
atoms can have a significant effect on the proper-
phonon energy in the crystal. See Intervalley re-
ties of crystals.
distribution, Intervalley scattering.
INTERVALLEY NOISE
Fluctuations of current flowing through a many- INTERVALLEY REDISTRIBUTION
valley semiconductor, or of voltage across the Population change or readjustment of equiv-
latter. These result from population fluctuations alent (or inequivalent) valleys of a many-valley
of individual valleys. In the general case, the mo- semiconductor energy band. Typically, it occurs
bilities of current carriers in inequivalent valleys when electrons of different equivalent valleys ob-
along the external electric field direction differ tain unequal energies from the electric field by
from each other. Therefore, random transitions of virtue of the difference in their effective masses
carriers from one valley to another cause corre- along the field direction due to intervalley scatter-
sponding changes of current (or of voltage), i.e. ing. Since the intervalley transition is related to the
the appearance of intervalley noise. If the external change of the electron quasi-wave vector (of the
field does not warm up the carriers, the spec- order of a reciprocal lattice constant), satisfying
tral density of such noise is proportional to the the conservation law of quasi-momentum requires
square of the current flowing through the crystal. participation of a phonon or an impurity in the
Its frequency characteristic is of the same form as intervalley scattering process. Intervalley redistri-
generation–recombination noise, i.e. proportional bution has a significant effect on semiconductor
to (1 + ω2 τ 2 )−2 , where τ is the lifetime of an properties: a high anisotropy of the conductiv-
electron in a valley in intervalley scattering. Mo- ity arises from the misalignment of the directions
bilities of carriers in isolated valleys are usually of the current flow and the electric field strength
636 INTERVALLEY SCATTERING
(Sasaki effect); when the intervalley distribution with phonon emission, the lower refers to one with
depends strongly on the electric field, then neg- phonon absorption. The quantity C is defined with
ative diffusive conduction can be observed. The the help of the explicit form of the Bloch wave
Hall effect and magnetoresistance (see Galvano- functions |k 1,2 (see Bloch theorem).
magnetic effects) exhibit some characteristic fea- The transition probabilities in scattering by im-
tures in their dependences on the electric and purities may be obtained from the above formula
magnetic field intensities; absorption of light by by setting ω12 = 0, N + 1/2 ± 1/2 = 1 in it. Then
free carriers decreases, and birefringence arises. the quantity C is defined by the short-range part
Intervalley redistribution may also occur as a of the electron–impurity interaction potential. The
result of a spontaneous symmetry breaking, which change in electron momentum, required for an in-
leads to a multivalued electron distribution (see tervalley transition, may occur at its collision with
Sasaki effect). This redistribution can also take the crystal surface. The possibility of such a tran-
place via the Gunn effect between valleys of dif- sition is determined by geometric considerations,
ferent energy bands. and depends on the orientation of the k 2 − k 1 vec-
tor relative to the surface. In this case the laws of
INTERVALLEY SCATTERING the conservation of energy and of the projection of
Quantum transitions of electrons or holes be- quasi-momentum on the surface (see Conservation
tween valleys of an energy band (see Many-valley laws) must be taken into account. See also Current
semiconductors). The term is used mostly for de- carrier scattering.
scribing scattering processes in semiconductors
and semimetals. Intervalley scattering may take INTRAMOLECULAR VIBRATIONS
place in the interaction of carriers with phonons,
Oscillations associated with relative displace-
impurities, or crystal surfaces. Since the distance
ments of atoms within a molecule or a molecular
between valleys in momentum space is compa-
ion in a molecular crystal. The frequency range of
rable to the Brillouin zone size, this scattering
the vibrations is 400–3000 cm−1 . For an impu-
involves transfer of a large quasi-momentum at
rity molecule in a crystal lattice the intramolecular
each individual event. The probability of interval-
vibrations are often local vibrations, significantly
ley scattering is usually considerably lower than
higher in frequency than the maximum frequency
that of intravalley scattering, in which both ini-
tial and final states of the carrier are in the same of the matrix crystal (e.g., NO− − 2−
2 , NO3 , SO4 , etc.
valley. Yet, intervalley scattering plays an impor- in an alkali-halide crystal lattice).
tant role, since it controls relaxation processes of
level populations of different valleys, and deter- INTRINSIC CONDUCTIVITY
mines steady-state populations and the magnitude Electrical conductivity of a semiconductor due
of the energy that current carriers absorb under the to the excitation of electrons from the valence
action of external fields (see Intervalley redistrib- band to the conduction band, and the generation
ution). Intervalley scattering is described by the of holes in the valence band, with the concentra-
probability of an intervalley transition, which (for tions of electrons and holes equal to each other.
an electron–phonon interaction) is Intrinsic conductivity predominates in intrinsic
semiconductors. For more details, see Electrical
1 1
Wk± →k = C N + ± δ(εk 1 − εk 2 ∓ ω12 ) conductivity.
1 2 2 2
where quasi-momenta k 1 and k 2 determine the INTRINSIC DEFECTS in crystals
initial and final states of electrons in valleys 1 Crystal lattice defects that are deviations from
and 2, respectively; εk1 and εk2 are the energies the regular arrangement of the atoms A, B, . . . of
of these states; ω12 is the frequency of a phonon a formula unit Am Bn . . . at preassigned sites. The
with a wave vector that is close to the distance be- simplest intrinsic defects are point vacancies V
tween maxima 1 and 2 to within a reciprocal lattice and interstitial atoms I of crystal components. One
vector; and N is the number of phonons with fre- of the mechanisms for the production of both
quency ω12 . The upper sign refers to a transition V and I is the thermal motion of the atoms.
INVARIANCE 637
in the spinor representation, we obtain to within to describe the dynamics of elementary excita-
a similarity transformation that Li = σi , the Pauli tions: the energy of a band state and its natural
matrices. Invariance (or symmetry) with respect to frequencies of vibrations, with invariants includ-
time inversion means that together with a state Ψ ing quantities that determine the dynamics of a
of an object, which is not in a magnetic field, a quasi-particle (spin, quasi-momentum).
state ΘΨ is possible that is characterized by op- With the help of group theory methods one can
posite directions of spins, and motions comparable write down the invariants for a physical system.
to those in the state Ψ . Applying the Pauli prin- This permits the analysis of material characteris-
ciple, one can establish the antiunitarity property tics: electrical conductivity, dielectric constant, and
of Θ, its analytic forms, and the matrix forms of magnetic susceptibility, including those with spa-
operators of the group G + ΘG (the so-called co- tial dielectric dispersion taken into account. For
representations). example, the conductivity tensor σij enters the
The crystal e ld is regarded as external with formula for Joule heat Q = σij Ei Ej ; therefore
respect to the quasi-particles in a crystal, hence one can analyze the tensor σij from the proper-
there arises an invariance of the potential energy ties of the quantity Q represented as a sum of
U (x) = Tg U (x), where x is the set of coor- invariants. All variables and their combinations,
dinates, g is an element from Ga = G + a0 G, including spin operators, can be classified in terms
G is the space group or Shubnikov group (mag- of their irreducible representations of a group,
netic space group), and a0 is a certain antiuni- and depending on the required approximation one
tary operator. Physical quantities, depending on can consider not only the first but higher pow-
their nature, implement representations or co- ers of parameters (e.g., Ei , Ej2 , etc.) as well.
representations of the group Ga (see also Space If the sets of functions fiα and ϕiα simultane-
groups). Most often, the term invariance is used ously transform by the same (αth) representation,
with respect to the group H or the Euclidean group the invariant is constructed as Iα = i (fiα )∗ ϕiα .
Ge = {g + (α/ h)}, where α is a translation vector, This invariant enters the Hamiltonian, the thermo-
and Ψ (x) is a scalar invariant if Tg Ψ (x) = Ψ (x); dynamic potential, etc., with a certain phenom-
the space {Ψi (x)} is invariant if hik Ak (g −1 x) = enological parameter determined for particular
Ai (x); and S(x) is a scalar invariant operator if systems from comparison between theory and ex-
Tg S(x) = S(x)Tg . Application of Tg to an ex- periment, or using calculations based on basic
pression containing xi means substituting hik xk − principles. This scheme has been very success-
hki αk for xi . ful for the construction of spin Hamiltonians, in
providing a phenomenological description of mag-
netically ordered substances, and in band theory
INVARIANCE, SCALING THEORY
for the case of complicated forbidden gaps (e.g.,
See Scaling invariance hypothesis.
Luttinger Hamiltonian).
determined by the nature of the material. In addi- by the magnetic-field-induced shift of the val-
tion to the above-mentioned mechanisms, radiative leys of a semiconductor close to a high symme-
quantum transitions stimulated by incident elec- try point of the Brillouin zone (see Many-valley
trons can also occur with the particles adsorbed at semiconductors). Phenomena caused by inversion
the surface. asymmetry are observable in AIII BV type semi-
The principal physical characteristics of in- conductors, mercury selenide, and tellurium.
verse photoelectron emission are the following:
the intensity of the radiation (photon yield Y ), its INVERSION AXIS
spectrum, polarization, and direction. The value An element of symmetry, which includes ro-
of Y , which is defined as the ratio of the number tation about a rotation symmetry axis (n, Cn )
of photons emitted in a certain wavelength range through a certain angle (2π/n) with a simulta-
to the number of incident electrons, depends both neous reflection (m) in a perpendicular plane at
on the nature and the structure of the material, the center of symmetry situated on this axis. In
the condition of its surface, and the kinetic en- crystals, there can be inversion axes of the same
ergy E and angle of incidence of the electrons. orders as the ordinary rotation axes: of the first
In metals the photon yield is small (Y 10−4 (1, Ci = S2 ), second (2 = m, C1h = Cs ), third (3,
photon/electron at E = 103 eV in the wavelength C3i = S6 ), fourth (4, S4 ), and sixth (6, C3h ) or-
range 200–600 nm). The study of these physical ders, where each inversion axis is listed in both
characteristics provides information about prop- the Hermann–Mauguin and the Schönflies nota-
erties of the surface (see Electron–photon spec- tion. Only axes 3, 4 and 6 have an independent
troscopy). Inverse photoelectron emission of met- significance, since axis 1 is equivalent to a center
als can serve as a secondary standard source of symmetry, and axis 2 corresponds to reflection
of optical radiation. The luminescence produced (m) in a symmetry plane.
by luminophors is widely used in practice (see
Cathodoluminescence). INVERSION-INEQUIVALENT POSITIONS
Two locations of an atom (an ion) that differ
INVERSION ASYMMETRY one from another by the inversion of the coordi-
Lowering of symmetry of a crystal structure by nates of all atoms of a crystal relative to a par-
removal of a center of symmetry with respect to a ticular lattice site (see the figure). The inversion-
related centro-symmetric structure (e.g., transition inequivalent positions exist, for example, in tung-
from the diamond to the sphalerite structure). The sten, corundum, and Si. The coordinate systems
term inversion asymmetry is used to characterize for the inversion-inequivalent positions that are
features arising in the electron energy spectra, e.g., related to the crystal environment in the same
term splitting due to the inversion asymmetry (see
Inversion splitting) at a lattice site. Of particular
importance is the lifting of the spin degeneracy
of spin states at points of the Brillouin zone: in a
crystal structure without an inversion center, only
those states with oppositely directed wave vec-
tors k and spins σ are found to be degenerate
(owing to time inversion): E(k, σ ) = E(−k, −σ ).
There are features of electron energy spectra in
Two inversion-inequivalent positions (1 and 2) in
semiconductors lacking inversion symmetry that
cube centers of a diamond-type lattice: • inversion-
produce physical effects that cannot be observed inequivalent positions (sites or interstitial sites), ◦ lattice
in crystals with an inversion center. Examples of atoms (only first coordination sphere is shown). The
these are: one type of combined resonance, the crystal field environment of position 1 will be the same
precession mechanism of electron spin relaxation, as for its inequivalent counterpart 2, if the crystal is in-
circular photogalvanic effects (see Photoelectric verted in the coordinate system with origin centered at
phenomena), and the linear Zeeman effect caused position 1.
640 INVERSION LAYER
way are mutually inverted. The crystal character- the free carrier bulk charge is applied to a metal
istics (in particular, the tensors corresponding to electrode (gate or field electrode). The inversion
external fields) in these inversion-different coordi- layer can also arise near a semiconductor–metal
nate systems are the same, whereas the external interface. However, if the surface charge sign is
fields are inverted in the systems. The magnetic opposite to the majority charge carrier sign, the
field B is, by virtue of its pseudovector nature, enriched layer appears in the subsurface layer.
invariant under space inversion, while the elec- The potential distribution in an electrical dou-
tric field strength E changes its sign at the in- ble layer is determined by the Poisson equation
version operation. The inversion nonequivalence (characteristic scale of its change is the screen-
has no effect in many experiments. For example, ing radius). For a low density of surface charge
the electron paramagnetic resonance spectra of the inversion layer does not appear. These charges
inversion-inequivalent paramagnetic centers coin- are neutralized by a subsurface depletion layer
cide. One can distinguish between these centers in which the majority carrier concentration is re-
only upon application of the electric field along duced, and an uncompensated layer of acceptor
with the magnetic one (see Electric eld effects). (donor) impurities arises. As the surface charge
A similar situation exists in optics. (or the voltage across the MOS-structure gate)
grows, this depletion layer shifts into the bulk
INVERSION LAYER semiconductor, and an inversion layer appears
A charged subsurface region formed by mi- near the surface. Such charge redistribution leads
nority current carriers in semiconductors. The to the change of the MOS-capacitor capacitance
inversion layer is a part of an electric double as a function of the electrical voltage Vg across
layer (see Fig.), which appears near the surface of the gate. This capacitance–voltage characteristic
the p-type (n-type) semiconductor where a pos- method allows one to study the charge distribu-
itive (negative) charge is localized. This charge tion over the interface of a semiconductor and an
is due to the trapping of majority carriers at the insulator.
Tamm states (see Tamm levels), i.e. levels of The inversion layer is separated from the semi-
surface defects and adsorbed atoms. There is a conductor bulk by a nonconductive depletion
widely used method for producing an inversion layer, so the MOS-capacitor can transform into a
layer in a metal–insulator–semiconductor structure transistor structure by establishing the source and
(or metal–oxide–semiconductor structure, MOS), sink contacts on the surface close to the field elec-
where an electrical voltage of the same polarity as trode. In this case, the conduction would not be
Energy diagrams of the subsurface region of p-type (left), and n-type (right) semiconductors.
ION BEAM CHANNELING 641
shunted by the bulk. The resistance of the source- high crystallographic index; the method of control-
sink channel is modulated by the voltage Vg ( e ld ling the doping of the interface Si–SiO2 with Na+
effect) due to the change of carrier concentra- ions; and studying the two-dimensional plasma
tion, and in the inversion layer the carrier mobility electron vibrations in the inversion layer.
varies, e.g., with changing thickness (size effect).
Studies of galvanomagnetic effects in the inversion INVERSION OF POPULATION
layers not only yield information on the electronic See Level population.
properties of a semiconductor–insulator interface,
but also enable optimization of the eld-effect tran- INVERSION SPLITTING, tunnel splitting
sistor parameters on the Si–SiO2 structure that is (I.B. Bersuker, 1960)
a basic element of modern electronics.
Splitting of energy levels in a multiatomic
Charge carriers in an inversion layer are local- system in the presence of electron degeneracy
ized in a potential well formed by a barrier on or pseudodegeneracy where the Jahn–Teller ef-
the semiconductor surface on one side, and by
fect produces two or more equivalent, quite deep
the bulk electrostatic charge on the other side. For
minima of an adiabatic potential with tunneling
strong inversion this potential well is rather narrow
between them (see Tunnel effect). This results in
(comparable to the quasi-particle de Broglie wave-
a lowering of the symmetry and a raising of some,
length), and the electron (hole) motion is quan-
or all, of the degeneracy. The value of the inversion
tized in the direction normal to the surface. How-
splitting depends strongly on the ratio between the
ever, in the direction along the inversion layer, the
stabilization energy of the Jahn–Teller effect and
motion remains free, corresponding to a quasi-
the vibration quantum in a minimum energy state;
two-dimensional electron gas. For dimensionally
it varies over a wide range from fractions to hun-
quantized states to appear the scattering needs to
dreds of cm−1 .
be weak (attained at a high-quality interface at low
temperatures when the thermal vibrations are in-
ION
effective). In the latter case the spacing of levels
should exceed their impact broadening. It is also An electrically charged atomic particle, or an
necessary to fill the lower subbands to provide low atom that accepted (negative ion, called an an-
carrier concentrations in the inversion layer. Di- ion) or gave up (positive ion, called a cation)
rect observation of the of two-dimensional nature one or several electrons (singly charged ion or
of the energy spectrum is possible using the oscil- multiply charged ion, respectively); also a neg-
lations of the magnetic conduction of the inversion atively or positively charged molecule (molec-
layer of Si–SiO2 by the variation of Vg (due to ular ion). The charge of an isolated ion such
the Shubnikov–de Haas effect), with the period of as Na+ and F− is a multiple of the electron
the oscillations depending on the magnetic field charge e. In crystals with ionic bonds the elec-
component normal to the surface. In strong mag- tron density redistribution between the positively
netic fields the motion of two-dimensional carriers and negatively charged ions leads to fractional,
in the inversion layer is quantized (see Quantiz- i.e. non-integral values of the charge e, called ef-
ing magnetic eld), and their spectrum becomes fective ionic charges, which are related to their
discrete. In this case, the character of galvanomag- ionic radii (see Atomic radii). For instance, in an
netic phenomena changes qualitatively and the NaCl crystal, the number of electrons is approx-
quantum Hall effect appears. Using this effect, one imately 10.2 (11 in the atom) in Na+ and 17.8
can measure the value of the fine structure constant (17 in the atom) in Cl− , for an effective ionic
on a silicon field transistor, and design a resis- charge of 0.8e. In a LiF crystal the charge of Li+
tance standard with an accuracy noticeably higher is 2.1e (3 in atom) and of F− is 9.9e (9 in atom)
than that achieved with other methods. There are so the effective ionic charge is 0.9e.
other specific electronic features of the inversion
layer: the appearance of an oriented superlattice ION BEAM CHANNELING
in the case of a surface parallel to a plane with a See Channeling of ion beams.
642 ION BOMBARDMENT
ION BOMBARDMENT of solids work function of the target material. The energy
Bombardment, or irradiation of solid targets spectrum of the emitted electrons has a maximum
by a beam of accelerated ions. This results in at the energy 1–3 eV, and the Auger emission spec-
backscattering of ions, which is nearly total for trum can exhibit fine structure. The principles of
the ion energy range 102 –103 eV (for normal inci- ion–electron emission underlie methods of analy-
dence). At higher energies the ions penetrate into sis of the composition and electronic structure of
the bulk, decelerate due to collisions with target matter, viz. ion-neutralization spectroscopy and
atoms, and stop, thus accumulating in the mate- Auger ion spectroscopy.
rial (Ion implantation, see also Ion-mixing, Ionic
synthesis). Sputtering of target atoms, appearance ION EMISSION
of radiation-induced defects, and phase transitions Emission of negative and positive ions, espe-
take place in the irradiated matter at the ion bom- cially by solids or liquids to a vacuum or gaseous
bardment. This is widely used in technology for matter, but ionized gases (plasma) can also serve
treating solid surfaces of various types (see Ion- as ion emitters. Thermal-equilibrium ion emis-
stimulated surface modi cation, Ionic modi cation sion arises at the thermal excitation of particles.
of lms, Ion-plasma treatment, Ion lithography), This is the surface ionization of vaporizing atoms
and for analysis of the substance composition of the solid itself, of the atoms of impurity el-
(see Secondary ion mass spectrometry). The in- ements diffusing from the bulk, or of particles
cidence of collimated ion beams at small angles adsorbed from the gaseous medium, as well as
to low-index directions in monocrystals results in the reaction products of the latter at the surface.
channeling of the beams. Thermally nonequilibrium ion emission is stim-
ulated by radiation of various kinds, these are
ION–ELECTRON EMISSION ions (ion–ion emission), electrons (ion–electron
Emission of electrons by a solid surface into emission), neutral particles (in ground or excited
a vacuum under ion bombardment. The ion– states), the photons of various energy ranges (ion–
electron emission coefficient γ = ne /ni is the photon emission). Ion emission is used to study the
ratio of the number of emitted electrons ne to composition and properties of solids.
that of ions ni incident on the surface. Ordinar-
ily there is no energy threshold. For low-energy ION EMISSION, SECONDARY
ions γ is almost independent of their energy and See Secondary ion emission.
mass, but depends on their charge (γ ∼ 0.2–0.3
for single-charged ions, and may exceed unity for ION ETCHING
multicharge ones). A removal of material from a solid surface un-
Depending on the mechanism of energy trans- der action of ion bombardment. Ion etching results
fer to an electron, potential ion–electron emission in cleaning the surface, exposing its structure, and
and kinetic ion–electron emission are known. The forming the relief. The bombarding low-energy
potential type takes place at the approach of a low- ions originate from various ion sources, glow dis-
energy ion to the surface when an electron from charge or radio-frequency plasma formed in a
the solid can transfer to the ion and neutralize it. closed cavity containing the object with the sur-
The energy released at such a transition can be face to be etched (see Ion-plasma treatment). The
taken up by other electrons in the solid, which main process involved in etching is cathode sput-
are then emitted into the vacuum. The kinetic type tering as well as ion-stimulated chemical reactions
emission occurs under high-energy ion bombard- which produce volatile substances at the surface,
ment due to intensive electron exchange with the as well as chemically active plasma components.
excitation of valence and inner level shells of the The etching rate is determined by the ion flux den-
target atoms, and involves electrons with a wide sity incident on the surface (up to a few mA/cm2 ),
energy range as well as Auger electrons. and by the material properties. The bombardment
The ion–electron emission depends on the ion- of crystals reveals grain boundaries, dislocation
ization and excitation energies, as well as on the outcrops, and so-called etch figures on the target
IONIC CONDUCTIVITY 643
surface. This allows one, in particular, to detect proposed to consider a deviation from this mean
the regions with different states of stress (selective value as a measure of the contribution of the ionic
etching). Ion etching is used to cleanse oxidized component of the bond. According to Pauling the
or soiled details and blanks during the technologi- dependence of the parameter β on the difference
cal procedures of manufacturing integrated circuits between the electronegativities of atoms A and B,
in microelectronics, for revealing microstructure XA − XB , has the form
of metallography samples, for preparing objects
β = 1 − e−0.25 (XA − XB )2 .
for electron microscopy studies, and for obtaining
atomically clean surfaces in ultrahigh vacuum. Even for typical alkali-halides β fails to reach
unity; for example, for LiF β ∼ 0.92, for KCl
ION EXCHANGER β ∼ 0.82, and for NaCl β ∼ 0.75.
See Ionite.
IONIC CONDUCTIVITY
IONIC BOND, electrovalent bond Electrical conductivity by ions as the charge
An interatomic chemical bond with an asym- carriers. Unlike electron conductivity, it involves
metric charge distribution between the adjacent the transfer of mass to electrodes. The substance
atoms. The ionic bond exhibits a molecular elec- deposition is described by the Faraday law: at
tric dipole moment; and its value can be regarded the transfer of the electrical charge of 96,500 C,
as a measure of the chemical bond ionicity. An one gram-equivalent of ion components is car-
ionic bond arises between atoms with a large dif- ried to the electrodes. The substance decomposi-
ference in electronegativity. tion under the action of electric current is called
The concept of ionic bond was introduced by electrolysis, so ionic conductivity is frequently re-
W. Kossel (1916) who noted that it is energet- ferred to as electrolytic conductivity. This is the
ically favorable to transfer an electron from a prevailing type of conduction in many dielectrics,
metal (M) to a halide (X) with the formation in particular, in ionic crystals, e.g., in alkali-halide
of a molecule M+ X− consisting of ions with crystals. Interstitial (intrinsic or impurity) ions
the electron shell configurations corresponding to can be current carriers such as Ag+ in crystals
those of noble gases. The resulting ionic bond is AgCl, H+ and OH− in ice crystals, or charged
caused by the Coulomb attraction between op- vacancies (e.g., anion vacancies in alkali halide
positely charged ions. Typical compounds having crystals). At low temperatures the ionic conductiv-
ionic bonds are alkali-halide crystals. However, a ity values are determined by the concentration of
perfect ionic bond is an idealization. The wave impurity or nonstoichiometric current carriers, and
function is actually a superposition of those of can differ from one sample to another made of the
covalent and ionic bonds, ψ co and ψ ion , respec- same material. The temperature dependence of the
tively: conductivity σ has an Arrhenius character, σ =
ψ = C1 ψ co + C2 ψ ion . σ0 exp[−U g /(k B T )], where U g is the activation
energy of the current carrier jump process. At high
In this case the bond ionicity (β) is given by the temperatures, the ionic conductivity is determined
expression by the concentration of stoichiometric current car-
riers that are the components of Frenkel defects
|C2 |2
β= . or Schottky defects, and it is the same for every
|C1 |2 + |C2 |2 sample of a given substance. The activation energy
There are several methods for empirically estimat- of ionic conduction at high temperature is deter-
ing the value β. L. Pauling established that in the mined in this case by a sum U = U0 + U g where
case of a predominantly covalent bond (C1
C2 ) U0 is the energy of thermally-stimulated nucle-
in a molecule A–B, one can calculate the bond ation of a stoichiometric defect. The maximum
energy as an arithmetic (or geometric) mean of conduction attained at a temperature close to that
the experimentally measured (covalent) bond en- of melting is 10−3 –10−4 ·cm. Solid electrolytes
ergies of the molecules A–A and B–B. Pauling exhibit abnormally high ionic conductivity.
644 IONIC CONDUCTIVITY COMPENSATION
IONIC CONDUCTIVITY COMPENSATION atomic charges that deviate greatly from those as-
Sublinear dependence of the ionic conductiv- cribed to their valence.
ity on the composition of multicomponent ma-
terials in the condensed state of matter for an IONIC CRYSTALS
equimolar interchange of any two of their compo- Crystals where the ionic type of chemical
nents. Ionic conductivity compensation is called bond prevails. Ionic crystals consist of positively
the polyalkaline effect when it involves relative re- (cations) and negatively (anions) charged ions that
placement of alkali atoms in oxide-coated glasses. can be monatomic or multiatomic. For example,
This compensation is an analogue of the sublin- ionic crystals of the first type are alkali halides,
ear dependence of the electric conductivity on alkaline-earth metal halides and oxides where the
the composition in some alloys for equimolar re- cations are metal ions, and the anions are halogen
placements of the components in them, and in or oxygen ions (NaCl, CaF2 , Al2 O3 , etc.). The
semiconductors for the equimolar replacement of ionic crystals of the second type are nitrates, sul-
donor impurities by acceptor types. All these cases fates, phosphates and other salts of metals where
can be explained by a general model for the si- the anions are oxygen radical groups consisting of
multaneous transport of any two types of current
a few atoms (NO− 2− 3−
3 , SO4 , PO4 , etc.). A certain
carriers which makes use of a matrix defining their
(small) ionicity is inherent in the chemical bonds
possible states with different migration abilities.
of covalent crystals.
There is a so-called bound case in which the mo-
The lattice energy of ionic crystals consists
bility is very small or equal to zero, and the free
mainly of an electrostatic part (Madelung energy)
case in which the mobility is large enough for elec-
and a positive energy of interaction of the ion elec-
tromigration observation.
tron shells (Pauli energy), and it amounts to 6–10,
20–30, and 30–150 eV per pair of ions for alkali
IONIC-COVALENT BOND
halides, alkaline-earth halides, and oxide crystals,
A type of a chemical bond intermediate be-
respectively. As to the electrical properties, most
tween the purely ionic bond and purely cova-
ionic crystals at normal conditions are insulators,
lent bond. Crystals built of atoms with differ-
and some of them are superionic conductors (see
ent degrees of electronegativity possess more or
Solid electrolytes). In view of a wide band gap
less differing average atomic charges when form-
(up to 13.6 eV in the case of LiF), ionic crys-
ing valence bonds. Therefore, in addition to a
purely covalent bond, there appears also an admix- tals are transparent to electromagnetic radiation
ture of ionic bond due to the attraction between in the visible and ultraviolet spectral ranges. The
charges of opposite sign. An example is the com- high-energy (up to 0.2 eV) optical phonons are
pounds AIII BV and AII BVI where the elements of well suited for determining the spectrum of atomic
Groups V and VI acquire a small (∼0.2–0.3) elec- vibrations of ionic crystals. Physical properties
tron charge taken from the elements of Groups III of ionic crystals are strongly dependent on their
and II, respectively. On the other hand, many crys- defect content (anion and cation vacancies and
tals considered as ionic crystals have a partially interstitial atoms and ions, dislocations, impurity
covalent bond due to overlapping of the near- atoms). The latter result, e.g., in the appearance
est neighbor wave functions. For such univalent of selective absorption bands in the transparency
crystals as alkali-halide crystals (and also silver- range of pure crystals (see Color centers). Typical
and thallium-halide ionic crystals) the charges fail for many ionic crystals, e.g., halides, is the elec-
to reach the limiting values ±1, and amount to tron excitation decay (mostly excitons) with the
±0.96. This effect is even more noticeable for ox- formation of lattice defects.
ides, sulfides, and selenides (e.g., MgO, SiO2 , PbS
and CdSe), since the ion O2− (to say nothing of IONICITY
the ions S2− and Se2− ) is unstable in the iso- Semi-quantitative characteristic of a chemical
lated state. Intermediate cases are possible when bond, which is intermediate between an ionic bond
it is difficult to unambiguously establish average (with ionicity = 1) and a covalent bond (with
IONIC SYNTHESIS 645
zero ionicity). There is no unified and univer- of inelastic processes is apparent from the oscil-
sally recognized definition of ionicity. The most lating dependence of the ionic effect on the ion
commonly used definitions are those proposed by energy in the low energy range (<200 eV), which
L. Pauling, C.A. Coulson, and J.C. Phillips. The reflects the character of the energy dependence
first of these is based on the chemical bond en- of the cross-section of ion resonance recharge at
ergy, the second proceeds from a consideration of small graphite clusters.
the coefficients of the expansion of single parti- In irradiation of already formed carbon films
cle orbitals in terms of atomic states, and the third by high and medium energy ions, the graphitiza-
is based on the static dielectric constant. The pa- tion process dominates, as is the case for heat
rameters of the ground state of crystals correlate treatment. Here, the film conductivity increases up
with ionicity, which allows its use for rough in- to values more than 103 (·cm)−1 .
terpolations (see Ionic bond, Ionic crystals). More
generally, the term ionicity can refer to the ionic IONIC RADIUS
properties of a solid. A size characteristic of ions in ionic com-
pounds. The ionic radius is defined so that the
IONIC MODIFICATION OF FILMS
distance between the nearest neighboring ions in
An alteration of the structure and properties of a crystal lattice of NaCl type would equal the sum
of the radii of the corresponding ions. Historically
lm s through ion bombardment. Irradiation is most
there have been several internally consistent sys-
often performed by ions of noble gases, hydrogen,
tems of ionic radii. For example, the radius of
and atoms involved in the film composition with
the ion O2− is 0.132 nm in the Goldschmidt sys-
the energy of about 102 –103 eV. Subjected to the
tem; 0.136 in the Belov and Bokij system; 0.140
irradiation are either the substrate before the film
in the Pauling system; 0.146 in the Zachariasen
deposition, or the entire system during the film
system. Nowadays researchers use standard, ac-
synthesis. The ionic modification of carbon films
cepted values tabulated in handbooks, and for O2−
has been investigated in detail, with irradiation by
this is 0.132 nm. When using ionic radii to de-
low-energy ions (1000 eV), the most efficient for
termine the interionic distances in crystals with
carbon film synthesis. The term “i-carbon” is used
structures different from that of NaCl, one may
to denote the carbon film formed during the irradi-
have to introduce corrections close to 0.01 in or-
ation. It is distinct from the graphite-like carbon
der of magnitude (see also Atomic radius).
of films condensed without the irradiation. The
“i-carbon” exhibits diamond-like properties: en- IONIC SYNTHESIS
hanced hardness, density and transparency, high A formation of new phases consisting of the
electric resistivity, chemical resistance, the band atoms of a target and atoms introduced during ion
gap width specific for diamond (see Diamond-like implantation in the solid. As a rule, the energy
carbon lm s). Varying the ion energy one can ob- of the implanted ions is between tens and hun-
tain carbonic films with a wide spectrum of prop- dreds of keV, the dose is more than 1017 ion/cm2 ,
erties: density d = 1.9–4 g/cm3 ; Mohs hardness the ion flux density is 1013 –1016 ion·cm−2 ·s−1 ;
1–10; electric resistivity ρ = 103 –1012 ·cm; the target during the irradiation can be either at
band gap width 0.5–2 eV. The action of the ions an elevated temperature (to ∼1000 ◦ C) or room
causes the selective suppression of the graphite- temperature (or lower). The phases under synthe-
like phase stable under normal conditions, and sis are formed in a thin surface layer (1 μm)
stimulates the formation of metastable phases, the in either an amorphous (see Amorphization) or
foremost being diamond. In addition, formerly a crystalline state directly during the implanta-
unknown crystal modifications of carbon were ob- tion process. To stimulate the crystallization and to
tained: a face-centered cubic phase with lattice improve the synthesized compound structure, sub-
constant a = 0.357 nm (γ -phase) and a body- sequent thermal, light pulse, electronic and other
centered cubic phase with a = 0.428 nm (phase kinds of annealing are often used.
C8 ). The ion-stimulation effect is caused by both Phase transformations (see Phase transitions)
elastic and inelastic processes. The important role that take place during the synthesis process under
646 ION IMPLANTATION
the ion implantation have some particular fea- The ionic synthesis method has prospects for
tures: solving problems of microelectronics (such as
producing silicides, insulator and semiconductor
(i) Because of essentially nonequilibrium condi-
layers), of metal science by modifiying surface
tions due to high-rate (t ∼ 10−10 –10−11 s)
properties (improvement of wear resistance, hard-
nonthermal excitation of the lattice and the
ness, corrosion resistance, thermal resistance, etc.
electron subsystem in local regions along
via thin films of carbides, borides, silicides, and
the ion tracks, where very high pressures
other compounds on the surface), and of many
and temperatures arise, both equilibrium and
other areas of science and engineering.
metastable phases are synthesized (see Meta-
stable state), as well as some compounds
ION IMPLANTATION
unobtainable by ordinary methods.
Introduction of atoms into the surface layer
(ii) The phases under synthesis are formed by
of a solid by bombarding its surface with ions
way of nucleation of elementary molecular
of energy from a few kiloelectronvolts to a few
complexes at the crystal nuclei (most often,
megaelectronvolts (see Ion bombardment). The
various radiation-induced defects and defect
penetration of the impurity atoms into the solid
clusters); such phases are usually finely dis-
at implantation is due to the ion kinetic energy
persed.
far exceeding the binding energy of the atoms
(iii) The influence of the base material structure
in the crystal. The process does not need high
and orientation on the crystallization and type
temperatures as is the case with diffusion doping
of synthesized compounds is predominant
or alloying. Ion implantation is a nonequilibrium
due to adaptation of the newly formed struc-
process so there are no thermodynamic restrictions
tural elements to details of the source phase.
on the concentration of implanted impurities, and
(iv) The course of solid state chemical reactions
this concentration can far exceed values limited by
during ionic synthesis is greatly affected by
equilibrium solubility. This method allows one to
the high concentration of radiation-induced
introduce any element into any material at room
defects, structural disordering, and high local
temperature, and in a quantity strictly controlled
pressure.
by the ion current density and the exposure time.
(v) Important parameters of the ionic synthesis
The penetration depth of ions of a given chemi-
that influence the phase composition and the
cal type into the selected material is determined
crystal structure of forming lm s are the ion
by their initial energy, and by the stopping ability
flux density and the sample temperature dur-
of a target (see Charged particle paths). One can
ing the irradiation. The ion energy determines
freely change the concentration profile by vary-
only the embedding depth and the effective
ing the ion doses and their energies in a controlled
thickness of the synthesized layer.
manner, and produce latent (“buried”) doped lay-
The ionic synthesis method for thin lms pos- ers by implanting ions of sufficiently high energy
sesses some essential advantages compared to (hundreds of electronvolts). The use of electrical
other traditional methods: purity of the process; or mechanical scanning of the ion beam across
opportunity to obtain a homogeneous gas mix- the surface can assure high homogeneity doping.
ture of strictly controlled composition; high repro- The ion implantation method allows one to carry
ducibility of results; opportunity to obtain deeply out local doping by using a sharply focused ion
lying layers and multilayer systems, i.e. to gov- beam or protective masks. Since the ions experi-
ern the chemical and mechanical properties of ence numerous collisions with the lattice atoms, a
the subsurface layers; opportunity to carry out great number of radiation-induced defects appears
the synthesis at low temperatures; opportunity to in the implanted region. As a rule, the number of
form metastable phases; high adhesion of the syn- these defects is more than 100–1000 times that
thesized layers to the substrate; opportunity to of the implanted ions. In semiconductor materi-
synthesize the phases on local regions of the sub- als, the radiation-induced defects can mask the
strate. effect of chemical doping; so annealing is called
ION IMPLANTATION DOPING 647
for. Pulsed annealing of semiconductors is effi- data. In parallel with these applications, ion im-
cient after ion implantation (in particular, laser plantation is used to investigate metastable phases
annealing). The dose range for ion implantation and obtain metallic amorphous alloys on the sur-
doping of semiconductors varies within the lim- face of metals and alloys. It also finds wider use
its 1012 to 1015 ions/cm2 . Ion implantation into for controlling the properties of insulators: ce-
metals and alloys requires doses as large as 1015 – ramics, glasses (see Vitreous state of matter), and
1018 ions/cm2 . This method allows one to control polymers. In particular, ion implantation into poly-
the solid surface properties over a broad range mers allows transforming various polymers into
independently of the bulk properties, thus provid- conductors through increasing their electrical con-
ing unique possibilities in the fields of solid-state ductivity by many orders of magnitude.
electronics and metallurgy. Initially the ion im-
plantation method was applied to semiconductors
as a means for introducing electrically active im- ION IMPLANTATION DOPING of semiconductors
purities for producing semiconductor junctions. Method for the formation of controlled distrib-
Owing to the universality of the method, its utions of electrically active impurities in semicon-
complete compatibility with planar-epitaxial tech- ductors. The method is based on ion implantation
nology, and also its consistency with diffusion, ion of atoms of chemical elements, which are capable
implantation is used in the production processes of serving as donors or acceptors in the treated
of most semiconductor devices and integrated cir- semiconductor, e.g., atoms of Groups V and III in
cuits. The guiding principles behind the develop- silicon and germanium, respectively. Selection of
ment of solid-state electronics, viz., the increase the ion bombardment energy serves to control the
of component packing density and speed and the doped layer depth (see Charged particle paths),
improvement of quality, are well implemented by while lithography helps to establish the layer distri-
combining ion implantation with ion lithography. bution relief over the irradiated surface. However,
Ion implantation is also employed for introducing many of the implanted atoms fail to occupy electri-
radiation-induced defects into the solid surface to cally active states during the implantation. For ex-
modify its mechanical, electrical, electrophysical ample, the atoms of Groups V and III in Si and Ge
and optical properties. For example, the implanta- can be at interstitial positions or enter compounds
tion of the ions H+ and N+ in gallium arsenide with radiation-induced defects that appear at im-
provides the disordered layers that are the basis of plantation, i.e. be in electrically inactive states, or
switching devices. The introduction of implanta- in states with deep levels. This reduces the mobility
tion disturbances into devices based on magnetic of current carriers. At high doses the amorphiza-
bubble domains allows one to control their static tion of semiconductors is also possible. Therefore,
and dynamic characteristics. The ion implanta- the technology of ion doping includes annealing.
tion method is applied for governing the surface Thermal annealing can cause the diffusion spread-
properties of metals and alloys. By appropriate ing of the impurity distribution profile established
selection of ions, it is possible to improve wear at implantation. More efficient is laser annealing
resistance, corrosion resistance, radiation resis- that does not lead to thermally active diffusion.
tance, fatigue failure resistance, and resistance to Semiconductor ion doping imparts diamond lay-
the surface oxidation of various materials, as well ers with semiconductor properties and produces
as to reduce their friction coefficients. The micro- semiconductor junctions there, and doping gallium
scopic aspects of metallurgy (e.g., to find impurity arsenide crystals and other semiconductor com-
diffusion rates in metals, phase diagrams, en- pounds provides them with donors and acceptors,
thalpy and entropy of reactions, trapping of atoms etc. Due to the possibility of producing sharp-
of dissolved matter), as well as the mechanisms relief microprofiles of electric resistance in various
of corrosion and surface strengthening, can be un- materials, ion semiconductor doping is highly ef-
derstood better with the help of ion implantation ficient in integrated circuit manufacture.
648 IONITE
shake-off, and other many-particle effects. In a two relaxation times describe the ion–lattice re-
number of cases, e.g., for the L23 -level of sili- laxation, namely: the longitudinal relaxation time
con, phosphorus, the K-level of carbon, etc., the (τ1 ) and transverse relaxation time (τ2 ). The time
ground state level broadening does not exceed τ1 is measured by electrocalorimetric means using
1 eV, and the ionization line fine structure does pulse or continuous saturation, or it is estimated
reflect the distribution of the density of unoccu- from the temperature broadening of the paraelec-
pied electron states. The ionization spectroscopy tric resonance lines. The time τ2 can be estimated
of outer shells reflects the loss for transitions from
from the uniform temperature-independent part of
the valence band to higher levels, and is sensitive
these lines. Both relaxation times depend on the
to the condition of the solid surface. If indepen-
concentration of the noncentral ions; and τ1 also
dent photoelectron spectroscopy data on the va-
lence band structure are available, then ionization depends on the temperature (e.g., for Li+ in KCl,
spectroscopy provides information on unoccupied τ1−1 ∼ T at T < 4 K, with the value τ1 ≈ 10−8 –
electron states. 10−9 s at T = 1.4 K). See also Relaxation time.
Ionization spectroscopy is used for qualitative
and quantitative analysis of the composition and
the electronic structure of the crystal subsurface, ION LITHOGRAPHY
and the adsorbed layers of solids. An advantage is A method of formation of microstructures in
its insensitivity with respect to the surface charg- the technology of large and very large integrated
ing, its ease of implementation, the possibility of circuits. It includes ion beam irradiation (expo-
obtaining information on the density of free states, sure) of a polymer resist (radiation sensitive coat-
and convenient control of the probe depth by vary- ing) deposited on a substrate; obtaining a protec-
ing the primary electron energy E P . This provides tive resist mask by a development process; etching
for the nondestructive elucidation of the composi-
the material through the mask, then removal of
tion and electronic structure of subsurface layers
the mask (see Lithography). When using polymer
in the depth of a solid.
resists, ion doses of 1011 to 1014 cm−2 are em-
IONIZATION, SURFACE
ployed. The negative resists become polymerized,
See Surface ionization. the molecular mass of the polymer increases and
it becomes insoluble during development. The ir-
ION–LATTICE INTERACTION, ion–photon radiation of positive resists results in the polymer
interaction degradation, which becomes soluble. The resist
A dynamic characteristic of noncentral ions exposure is performed by a spatially modulated or
which arises from the dependence of the potential computer-controlled sharply-focused ion beam.
energy of these ions on lattice atom displacements Ion lithography is superior to other methods
from their equilibrium positions. The ion–lattice of microlithography (photolithography, electron
interaction is responsible for both the influence of and X-ray lithography) in submicron resolution
the elastic stress in a crystal on the spectrum of for short exposure times, and is insensitive to
tunneling states of noncentral ions, and the ion–
proximity effects. This is due to the small lateral
lattice relaxation. Ion–lattice interaction data are
path of secondary electrons (∼10 nm) during the
obtained from experiments on paraelectric reso-
nance in the presence of an applied axial pressure.
exposure and to the efficient absorption of the ra-
diation (∼1 keV/nm) in the resist material. More
ION–LATTICE RELAXATION descriptively, ion lithography is a method of di-
A process of restoration of thermodynamic rect microstructure formation in the technological
equilibrium in a system of noncentral ions after and topological layers of integrated circuits. Layer
the action of some external disturbance (e.g., pulse formation results from the ion, chemical or ion-
saturation of the paraelectric resonance spec- stimulated chemical etching of insulator, metal, or
trum). By analogy with paramagnetic relaxation, semiconductor layers by a controlled ion beam.
650 ION METALLURGY
ION METALLURGY, implantation metallurgy interface “film–substrate” during the film conden-
Production of solid solutions by the ion implan- sation (renewal), and this allows one to reduce
tation method. A particular feature of the method sharply the ion energy (E ∼ 10 keV). These meth-
is the possibility of creating materials that cannot ods of ion mixing are used to produce very thin
be obtained under equilibrium conditions. Such coatings and surface layers with some special
metastable “alloys” can persist up to rather high properties such as improved wear and corrosion
temperatures. An example is the AgCu substi- resistance, hardness, high adhesion, surface elec-
tutional solid solution obtained by implantation trical conductivity, etc.
of silver in copper (this decays at T 310 ◦ C),
the metastable substitutional solutions W–Cu or ION-MOLECULAR CRYSTALS
Ta–Cu appearing at the implantation of tungsten Crystals formed by complex ions, and belong-
or tantalum, respectively, in copper, etc. See also ing to heterodesmic structures, i.e. those with
Ionic synthesis. more than one type of chemical bonding. One can
group ion-molecular crystals into three types:
ION-MIXING during irradiation
(i) those with a monatomic cation (positive ion)
Redistribution of atoms in a solid during irra-
and a complex anion (negative ion; salts of
diation by an ion beam, caused by the appearance
oxyacids): perhalogenates, halogenates, ni-
of recoil atoms and collision cascades (recoil atom
trates, phosphates, sulfates, carbonates, nio-
implantation). This can involve distortion of the
bates, tungstates and similar compounds of
impurity concentration profiles at high-dose ion
alkali and alkaline-earth metals;
implantation. The most interesting processes for
(ii) those with a monatomic anion and a complex
studies and practical applications are:
cation such as ammonium halides;
(1) Ion-mixing with cross-penetration of dissim- (iii) those with a complex cation and anion, e.g.,
ilar atoms at the interface of the two-layer NH4 NO3 , NH4 ClO4 , NO2 ClO4 , etc.
structure “thin film–substrate” aimed to pro-
Ion-molecular crystals are generally low in
duce thin (∼10 nm) boundary layers with spe-
symmetry and polymorphous. For example, five
cial properties (film thickness d ∼ 30–50 nm,
lattice modifications for NH4 NO3 are known. The
ion energy such that the maximum of energy
presence of low-symmetry complex ions can in-
loss is at the interface). At elevated irradiation
volve the orientational ordering or disordering of
temperatures (mostly >300 K), an increase of
ion-molecular crystals, which is related to the
the atom penetration depth is observed due to
possibility or impossibility of a free rotation of
radiation-induced diffusion.
the complex ions at their lattice sites. It is for
(2) Ion-mixing via irradiation of multilayer struc-
this reason that order–disorder phase transitions
tures formed of very thin (d ∼ 3–5 nm) al-
are common for ion-molecular crystals, as well
ternating layers aimed to produce phases (in-
as anisotropy of the crystal properties in most
cluding metastable ones) or compounds with
cases. In view of the purely ionic bond between
a certain structure and stoichiometry.
the anion and the cation they have similarities
Ion-mixing is often studied via Rutherford with ionic crystals like alkali-halide crystals, while
backscattering, Auger microanalysis, and X-ray their status of being covalently bonded complex
methods. Used for this purpose are universal ion ions provides them with properties similar to those
sources of inert gases (less often P+ , N+ , etc.) of of molecular crystals. These features are graphi-
energy E ∼ 50–600 keV with radiation doses Φ = cally revealed in optical spectra where, along with
1015 –1016 ions/cm2 . Using such sources is an ad- the optical phonon bands of ion-molecular crys-
vantage compared to direct ion implantation where tals, there are infrared bands typical of internal
a special ion source is needed for each implanted vibrations in complex ions (intramolecular vibra-
element, and high implantation doses are required tions). The ultraviolet spectra of ion-molecular
(e.g., for metals Φ = 1017 –1018 ions/cm2 ). Dy- crystals are complicated by intramolecular transi-
namic ion-mixing occurs at the irradiation of the tions of electrons from occupied orbitals to free
ION–PHOTON SPECTROSCOPY 651
ones. Therefore, a double-band model involving photon emission factor (photon yield) Y , which is
a valence band and a conduction band is not ap- the ratio between the number of electrons emitted
plicable to ion-molecular crystals, although there in a given frequency range, and the number of ions
are intrinsic allowed electron energy levels (or incident at the surface. The value Y depends on
quasi-bands) between these bands. the nature of the material, the condition of its sur-
face, and the parameters of the bombarding ions
IONOLUMINESCENCE (energy, angle of incidence, and ion type). In the
Luminescence appearing during high-energy case of metals, Y 10−4 photon/ion at the res-
(0.001–10 MeV) ion bombardment of solids. Its onance transition frequency under bombardment
salient features are a high density of excitation en- of the target by ions of intermediate mass and
ergy and the surface nature of the excitation owing energy. This parameter may increase noticeably
to the small ion penetration depth in solids (from (10–100 times) when the surface is coated with
several to hundreds of angstroms depending on atoms of chemically active gases. The principles
the beam energy). Hence, the ionoluminescence of ion–photon emission underly the ion–photon
method yields information that pertains mostly spectroscopy method of elemental and layer-by-
to the subsurface layer; it underpins ion–photon layer analysis of the surface.
spectroscopy. The ionoluminescence spectrum is
ION–PHOTON SPECTROSCOPY
determined by the composition of the target mate-
A physical method based on the ion–photon
rial and by the radiation-induced defects appearing
emission phenomenon for the study of a solid
under the ion bombardment. In addition, the inten-
surface. Ion–photon spectroscopy employs mea-
sity depends on the ion energy and density, and on
surements of the intensity and frequency range of
the crystal orientation relative to the beam direc-
optical emission from particles sputtered from the
tion.
surface, the dependence of these characteristics on
the incident ion position, the sputtered layer depth,
ION–PHOTON EMISSION
the nature, and the surface condition of the tar-
Emission of optical radiation by a solid, or by get. This spectroscopy allows one to determine
excited particles leaving its surface under bom- the elemental composition of the surface, and the
bardment by accelerated ions (or atomic parti- distribution of elements over the depth and the
cles). The radiation emitted by the surface itself surface. One can also study spectra of atoms and
is typical of insulators and semiconductors, and molecules, processes of oxidation, diffusion, sur-
is widely investigated (see Ionoluminescence). face particle adsorption, and mechanisms of the
There is another component of ion–photon emis- interaction of atomic particles with solids.
sion which is specific for each substance, and re- Quantitative elemental analysis is based on
lated to photon emission by excited particles (sput- the correspondence between photon yield, Y , at
tered and scattered atoms, molecules, clusters and the frequency of selected transitions of impurity
their ions) noninteracting with the surface. The atoms with concentration, c, and the atomization
spectra of this component contain, in a general factor, s M of the host atoms: c = Y/(Aτ ns M )
case, atomic and ionic lines as well as molecular (A is the probability of the transition; n and τ
bands, and in some cases continuous radiation is are respectively the probability of excitation and
observable. The principal mechanisms of the ex- the atom lifetime in this state). The determina-
citation of sputtered atoms and ions are, first, the tion of c is facilitated if parameters n and s M are
detachment mechanism when the electronic states known or can be calculated, and we note that the
of the particles are formed only during the process relation between c and Y is linear. In other cases
of their escape from the surface, and second, one should use standards. The layer-by-layer analy-
the kinetic mechanism when the particle excita- sis is performed by continuously measuring the
tion is caused by atomic collisions. Important in dependence of the intensity of a spectral line of
both cases are electron exchange processes in the any impurity atom on the duration of the surface
solid, particle-emitting system. Quantitatively, the etching. The potentiality for quantitative analy-
excited particle emission is characterized by the sis using ion–photon spectroscopy is almost the
652 ION-PLASMA TREATMENT
same as that of the related method of secondary the gas phase by sputtering. Protective coat-
ion mass spectrometry. Advantages of ion–photon ings obtained by this treatment have high ad-
spectroscopy are high sensitivity and accuracy of hesion; the process of their formation does not
elemental identification, remote mode of data ac- need high temperatures; application of several
quisition, and no need to apply an electric field to sputtering targets of various chemical compo-
a target. sition allows one to govern the composition of
The ion–photon spectroscopy method uses ions the forming films;
(typically, Ar+ ) with tens of keV in energy to (3) ion-plasma cleaning, polishing, producing the
probe the surface under conditions of high vac- relief of surface roughness.
uum. The emitted particle radiation is analyzed by
These treatments are efficient in combination with
optical spectroscopy methods.
ion bombardment (so-called reactive ion etching,
reactive ion-plasma deposition of materials). In
ION-PLASMA TREATMENT
these cases, an additional ion beam is used to ac-
Treatment of materials that consists in modify-
celerate the processes of forming volatile products
ing the material surface and subsurface layers by
of chemical reactions by incoming plasma parti-
gas plasma components, i.e. by high-speed elec-
cles on the irradiated surface, and the removal of
trons and ions. To this end, the material is placed
the reaction products.
in the immediate vicinity of or inside the region
where a gas discharge occurs with its strong elec-
ION PROJECTOR, eld-ion microscope
tric (constant or high frequency) and magnetic
A lensless vacuum ion-optical apparatus for
fields. A pressure of 10−2 –102 Pa is usually em-
obtaining an image of a conducting material sur-
ployed with a particle density of 1012 –1016 cm−3 ,
face with a magnification of several million. A tip
and a particle energy of 10–104 eV. Collisions of
of a thin needle cooled by liquefied gas acts as a
electrons and atoms with atoms of the irradiated
positive electrode, with its surface imaged on a lu-
material lead to the following effects: (i) surface
minescent screen. The noble gas atoms that fill the
sputtering (knocked out atoms and ions become
inner volume of the ion projector are ionized in
plasma components and participate in the treat-
the strong electric field (∼5·108 V/cm) at a dis-
ment); (ii) ion penetration into the material (ion
tance 0.5 to 1 nm from the needle tip that accepts
implantation); (iii) deposition of particles from
their electrons. Under the action of the electric
the plasma on to the substrate; (iv) activation of
field, positive ions of the imaging gas are acceler-
various heterogeneous chemical reactions at the
ated and bombard the screen which has a negative
treated surface. Using this ion-plasma treatment
potential. The luminescence distribution of each
one can control material parameters over a wide
screen element is proportional to the density of
range by changing the plasma chemical compo-
the incoming ionic flow, and thus reproduces the
sition, the electric discharge parameters (energy
ion density distribution near the needle tip on an
of excited particles, ratios of excited and ionized
expanded scale. The magnification equals the ra-
plasma particles, multiplicity of ionization), and
tio between the screen radius and the tip curvature
the magnitudes and directions of electric and mag-
radius. The contrast image of the needle tip ap-
netic fields. Sometimes the ion plasma component
pearing at the screen is determined by the presence
only is isolated for irradiation (see Ion bombard-
of the local microscopic relief, and the imaging
ment, Ion-stimulated surface modi cation). Widely
gas pressure that usually does not exceed 0.01 Pa.
used ion-plasma treatments are:
An ion projector is used in e ld ion microscopy.
(1) plasma-chemical etching and ion-chemical
etching, including the chemical reactions of ION-STIMULATED SURFACE MODIFICATION
high-energy (“hot”) plasma particles with the Alteration of the structure and properties of
material; the surface and subsurface layers of solids by
(2) ion-plasma deposition and ion-beam deposi- ion bombardment. The bombardment of solids by
tion for producing films on various substrates. ions with energy E ∼ 1–10 keV leads mainly
The atoms to be deposited are introduced into to sputtering and cleaning the solid surface, to
IRIDIUM 653
modification of the crystal morphology, i.e. the for- bombardment are cathode sputtering and ion im-
mation of “cones”, “craters”, etc.; or, conversely, plantation. This treatment is applied to metals,
to ion polishing of the surface for special condi- semiconductors, and insulators, and is one of the
tions. At high radiation doses Φ with inert gas basic technological processes in the manufacture
ions (Φ ∼ 1017 –1020 cm−2 ), gas blisters may of semiconductor devices and integrated circuits.
form at the surface (see Blistering), and surface The ion treatment of a surface typically requires
swelling and scabbing occurs, which imitates the an ion source, ionic optics, a mass separator, and a
processes connected with problems at the first source of accelerating voltage up to 200 kV.
wall of a thermonuclear reactor. The high-energy
ions implant themselves into the irradiated mate- IRIDIUM, Ir
rial (see Ion implantation). Initial studies involved Chemical element of Group VIII of the peri-
the ion implantation doping of semiconductors for odic system with atomic number 77 and atomic
the modification of their electrophysical charac- mass 192.22. Natural iridium is a mixture of two
teristics, for the formation of p–n junctions, and stable isotopes 191 Ir (38.5%) and 193 Ir (61.5%).
for the design of devices. Implantation metallurgy Outer shell electron configuration is 4f 14 5d 7 6s 2 .
(see Ion metallurgy) involves the modification of Successive ionization energies are 9.1, 17.0, ≈27,
surface properties of metals and alloys to increase ≈39, ≈57 eV. Atomic radius is 0.136 nm; radius
microhardness, wear resistance, ultimate fatigue of Ir3+ ion is 0.073 nm, of Ir+4 ion is 0.068 nm.
strength, corrosion resistance and cavitation re- Electron affinity is 1.97 eV. It is chemically inert;
sistance, as well as modification of optical and oxidation state is +3, +4, +6, less often, +1, +2.
other properties by high-dose implantation of var- Electronegativity is 1.55.
ious elements using beams of ions with energies Iridium in a free form is a silvery-white metal.
of hundreds of keV, and Φ ∼ 1016 –1018 cm−2 . It has face-centered cubic lattice, space group
Mechanisms for the improvement of properties are F m3m (Oh5 ); a = 0.3839 nm at 299 K. Density
diverse and not always clear. Depending on the
is 22.65 g/cm3 at 299 K (one of highest val-
ion–target combination and the regimes and condi-
ues for a simple substance); T melting = 2716 K;
tions of the irradiation which take place simultane-
T boiling ∼
= 4740 K. Bonding energy is −6.93 eV
ously with the impurity implantation to the depth
at 0 K. Latent heat of melting is 26.38 kJ/mole,
of a few thousand angstroms, there is observed
the formation of equilibrium or metastable surface latent heat of evaporation is 629 kJ/mole, spe-
alloys, phase and/or structural transformations in cific heat is 131 J·kg−1 ·K−1 (at 298 K). De-
thin subsurface layers (up to amorphization), the bye temperature is 420 K; the linear thermal
introduction of mechanical stresses, etc. When expansion coefficient is 6.63·10−6 K−1 (293 to
carried out in a technical vacuum, ion-induced ox- 373 K); coefficient of thermal conductivity is
idation and carbonization of the surface are also 148 W·m−1 ·K−1 (273 to 373 K); adiabatic co-
possible. A particular case is ion-nitriding, which efficients of elastic rigidity of iridium are the
improves some mechanical properties of the sur- highest of all cubic group crystals (except dia-
face of these materials. See also Ion-mixing, Ionic mond): c11 = 600, c12 = 260, c44 = 270 (GPa)
synthesis, Ionic modi cation of lms. at room temperature. Isothermal bulk modulus of
iridium is 355.3 GPa (at 298 K); Young’s modulus
ION TREATMENT of surfaces is ∼
=514 GPa; Poisson ratio is 0.26; tensile strength
An action upon a solid surface by a flux of is 0.23 GPa, relative elongation at the moment of
accelerated ions to change its profile, chemical rupture is 2%. Brinell hardness is 1.64 GPa, Vick-
composition, physicochemical and electrophysical ers hardness is ∼ =220 in HV units for annealed
properties, as well as to produce pits or to de- iridium (at room temperature). Thermal neutron
posite foreign substances. The ion treatment of trapping cross-section is 440 barn; ion-plasma
a surface is carried out with the aid of ion etch- frequency is 16.02 THz. Linear low-temperature
ing, ionic milling, ion implantation, ionic surface electronic specific heat term (Sommerfeld coeffi-
passivation, ion lithography, and ion-stimulated de- cient) is 3.2 mJ·mole−1 ·K−2 . Electric resistivity
position. The main processes resulting from ion is 51 n·m (at 295 K); temperature coefficient
654 IRON
of irradiation annealing (see Radiation physics of γ -quanta, etc.) within the temperature range be-
solids). When radiation doses are large, anneal- low (0.4–0.5)Tmelting , where Tmelting is the ma-
ing lowers the rate of accumulation of the defects, terial melting temperature. Since the mechanical
particularly if the intensities themselves are high properties of metals and alloys are highly sensi-
enough. Irradiation annealing mechanisms are re- tive to structural and phase changes, irradiation
lated to the processes of radiation-induced diffusion hardening was found in them earlier than other
of defects, to atomic restructuring triggered by radiation effects, and has been studied in more
collisions of external particles with crystal atoms, detail. The level of irradiation hardening depends
and also to the small-dose effect. on the nature of the irradiated material, the condi-
tions of irradiation, and the dose. For pure crystals
IRRADIATION EMBRITTLEMENT (aluminum, copper) the increase of the amount of
Depletion or total loss of plasticity of materials stress required to induce plastic flow becomes no-
resulting from their irradiation with high-energy ticeable even after irradiating with neutron doses
neutrons, ions, electrons, γ -quanta and other types up to 1012 –1014 n/cm2 . Moreover, when irradiat-
of emissions. Processes of low-temperature irradi- ing these materials with the much higher doses of
ation embrittlement and high-temperature irradia- 1021 –1022 n/cm2 the resulting changes may be
tion embrittlement can thereby develop. greater by a factor of 5 to 10. As a rule, irradiation
Low-temperature irradiation embrittlement hardening in metals and alloys is accompanied
consists in raising the temperature of the brittle– by irradiation embrittlement. Irradiation hardening
viscous transition and lowering the relative elon- in crystals results from the development of new
gation, transverse compression, and impact centers of dislocation pinning, and of barriers for
strength of irradiated metals and alloys within the mobile dislocations. Beside the ordinary interac-
temperature range of tests, that is below approxi- tion forces with the lattice and other initial lattice
mately 0.5Tmelting , where Tmelting is the melting defects, dislocations in an irradiated crystal, as the
temperature. Low-temperature irradiation embrit- natural “carriers” of plastic deformation, have to
tlement is mainly caused by the same processes penetrate through a whole spectrum of other bar-
that are responsible for irradiation hardening. The riers of radiation origin (irradiation defects).
annealing of irradiated metals at temperatures
above 0.5Tmelting lessens the effects of this em- IRRATIONAL TWINNING (a term introduced by
brittlement, and can lead to a complete recovery N.A. Brilliantov, I.V. Obreimov, 1935)
of plasticity back to its initial level. Appearance of strip-like regions with a turned
High-temperature irradiation embrittlement lattice structure in a monocrystal as a result of
constitutes a practically irreversible lowering of plastic deformation. This phenomenon is observ-
the plasticity parameters of irradiated metals and able in alkali-halide crystals of the NaCl type.
alloys at temperatures above 0.5Tmelting . The Unlike ordinary twinning of crystals, irrational
cause of this embrittlement involves a change of twinning is characterized by an irregular angle of
the ratio between the strength of the matrix to that the lattice turn, blurred strip boundaries, and mis-
of grain boundaries. The accumulation of helium alignment of the boundaries with crystallographic
generated during the course of nuclear reactions, planes. The strip boundaries are close to planes of
and the segregation of adverse impurities at grain the {110} type. Irrational twinning takes place un-
boundaries, both stimulated by irradiation, signifi- der compression of planar prisms spalled off along
cantly lower the tear strength at grain boundaries. planes of crystal cleavage. As a result of irrational
twinning the lateral surface of prisms acquires an
IRRADIATION HARDENING accordion-like profile with bevels at a normal to
Technique of increasing strength and elasticity the loaded faces. X-ray, optical polarization, and
of crystalline bodies (elastic modulus, proportion- selective etching methods have been used to estab-
ality limit, yield limit, ultimate strength, hardness) lish that a strain emerges along various crystallo-
by irradiating crystalline targets with high-energy graphically equivalent slip systems in neighboring
particles and photons (neutrons, electrons, ions, disoriented layers. The lattice disorientation angle
IRREDUCIBLE TENSOR OPERATORS 657
in neighboring layers ordinarily does not exceed electric field vectors. Introduction of irreducible
2–3◦ , its maximum value of 6◦ being observed as tensor operators simplifies calculations of matrix
a result of strain at elevated strain temperatures. Ir- elements, which are then reduced to computing
rational twinning also occurs at stretching of NaCl the simplest single reduced matrix element. Thus,
type alkali halide crystals. the set of irreducible tensor operators Tlm (S),
where l = 0, 1, . . . , 2S and m = −l, . . . , l, forms
IRREDUCIBLE POLYNOMIAL BASIS a complete orthonormal system of linearly inde-
Finite set of polynomials P (ci ) invariant un- pendent matrices of rank 2S + 1. In magnetic
der the symmetry group G of a physical system resonance spectroscopy of solids, Stevens oper-
or problem. Any other invariant polynomials of
ators (K.N. Stevens, 1953) also find wide use.
the system can be expressed in terms of these
They represent a set of linearly independent op-
polynomials. In the theory of phase transitions
erators Gm l (J ) composed of polynomials of mu-
G = Φ is the crystallographic space group of the
tually noncommuting components of the angular
high-symmetry phase, and ci are parameters of
momentum (effective spin) vector J . Such oper-
the transition, belonging to only one irreducible
representation. In an expansion of the thermody- ators were initially derived as a generalization of
namic potential the coefficients of the invariants spherical harmonics. Matrix elements of Stevens
of different orders are considered to be compara- operators are easily calculated; and have been tab-
ble in their magnitudes. In this simplified model ulated for l = 1, . . . , 7, m = −l, . . . , l. They are
an irreducible polynomial basis P (ci ) is used to used as equivalent operators to calculate matrix
connect coefficients of the invariants of various or- elements of the operators of the interactions of
ders with possible low-symmetry phases. In this electrons in the unfilled shells of transition ions
case the symmetry groups of the latter (subgroup in the presence of a crystalline electric field and
of the Φ-group) can often be identified with- an external magnetic field. The same matrix ele-
out computation by simple fitting, using a group ments also serve as basis operators to construct the
representation (character) table or the Frobenius generalized spin Hamiltonian using the method of
method. In other nonlinear theories of solid state invariants; they are unnormalized linear combina-
physics the irreducible polynomial basis is con- tions of irreducible tensor operators.
structed from the components of vectors, tensors For radiospectroscopic spectra in solids, equiv-
(e.g., strain tensor), etc. Associated with a particu- alent operators are used that display identical
lar group G there can be a point group, a magnetic transformation properties during rotation of the
group, a space group, etc. The construction of coordinate system. According to the Wigner–
an irreducible polynomial basis is founded on the Eckart theorem (E.P. Wigner, C. Eckart) for a set
Noether theorem (the degree of the polynomials of states with a definite value of angular momen-
from the basis does not exceed the order of the tum, matrix elements of such operators possess a
group).
common factor, the so-called reduced matrix ele-
ment, that is independent of the azimuthal quan-
IRREDUCIBLE TENSOR OPERATORS
Operators Tlm that satisfy certain commutation tum number. Thus, for electrons in an unfilled
relationships, and experience the same transfor- paramagnetic ion shell, one may skip the cumber-
mations during rotation of a coordinate system as some calculation of the matrix elements of oper-
the lmth eigenfunction of the angular momentum ators for the interaction of the electrons with the
operator. Spherical harmonics Ylm (θ, ϕ) represent crystalline electric field or the Zeeman operator
a particular case of such operators, where an ir- (L + 2S)·B. Instead, one simply multiplies tab-
reducible tensor is formed from the components ulated matrix elements of much simpler operators
of a radius vector. Atomic and radio spectroscopy (polynomials of angular momentum components
use irreducible tensor operators composed of the constructed from either Stevens operators or irre-
operators of the total angular momentum, J , the ducible tensor operators) by the reduced matrix
orbital angular momentum, L, the electron and nu- elements which also have been precalculated for
clear spins, S and I , as well as the magnetic and all the transition ions.
658 IRREVERSIBLE PROCESSES
The above operators underlie the technique applied forces in reverse order, or in reverse di-
of irreducible tensor operators used to describe rections.
both optical and radio-frequency spectra of para-
magnetic ions by representing their Hamiltonians ISING MAGNET
as expansions over the full system of irreducible A magnetic material with properties describ-
tensor operators. Expressing a generalized spin able by the Ising model within an acceptable mar-
Hamiltonian as a sum of these operators for the gin of error. Ising magnets include some rare-earth
components of electron and nuclear spins, and compounds, e.g., dysprosium–aluminum garnet
electric and magnetic field vectors, is a powerful (see Iron garnets) and orthoferrites at low tem-
tool. It provides matrix elements of a generalized peratures. Typical of an Ising magnet is a pro-
spin Hamiltonian in closed form without having nounced anisotropy of its magnetic properties
to determine their wave functions (which is par- (see Magnetic anisotropy), which is related to the
ticularly important for multispin systems), thereby anisotropic character of the Ising model Hamil-
simplifying the calculation of the transformations tonian. The discreteness of the energy spectrum in
of the generalized spin Hamiltonian during rota- the Ising model results in an exponential temper-
tions of the coordinate system, and simplifying the ature dependence of the additional magnetization
procedure for excluding extra parameters from the along the selected axis, and of the magnetic part
Hamiltonian through an unitary gauge transfor- of the speci c heat at low temperatures. Unlike
mation. When applying this technique, one often the Heisenberg magnet, the order parameter in the
resorts to the concept of effective spin. The latter is Ising magnet differs from zero not only in a three-
a fictitious angular momentum S that is introduced dimensional crystal, but also in a two-dimensional
to calculate the energies EM of a group (mul- one.
tiplet) of electron energy levels separated from
other levels by a gap appreciably exceeding the
ISING MODEL (E. Ising)
level-to-level spliting within the group itself. The
A spin model describing both magnetic (see
quantity S is chosen in such a way that the number
Ising magnet) and nonmagnetic (ferroelectrics, bi-
of values M = − S, . . . ,
S, which equals 2 S + 1,
nary alloys and others) macroscopic systems. The
coincides with the multiplicity of the multiplet.
Ising model Hamiltonian has the form:
However, it appears to be more convenient to
use integer values of S to describe the doublet H I = −J σm σn − B σm ,
states of paramagnetic ions with even numbers (m,n) m
of electrons, N . One then avoids misunderstand-
ings that stem from the transformation properties where J is the exchange interaction constant, each
of a half-integer effective spins. According to the variable σn = 2Snz assuming the values ±1 (the
Wigner–Eckart theorem, the elements of matrices spin S = 1/2), m and n are the vector indices of
describing interactions of electrons either with the crystal lattice sites, and B is the external magnetic
electric fields of a crystal (or molecule) or with an field. The first summation extends over nearest
external magnetic field may be expressed via the neighbor pairs. The Ising model is much sim-
matrix elements of irreducible tensor operators or pler than other quantum spin models since the
Stevens operators, constructed from components Hamiltonian has no noncommuting operators. All
of the effective spin for different states of a multi- quantum states of this model are determined by
plet. See also Spin Hamiltonian. sets of values σn at the lattice sites.
In the case of one-dimensional and two-dimen-
IRREVERSIBLE PROCESSES sional lattices exact results have been obtained
Processes in a thermodynamic system accom- for some thermodynamic functions. For the one-
panied by an increase of entropy. In the case dimensional Ising model there is no phase tran-
of irreversible processes, the system cannot be sition, and the magnetization vanishes as B → 0.
brought back to its prior state, even if it is sub- The Ising model free energy F for the two-
jected to the action of the same factors such as dimensional square lattice at B = 0 and J > 0 is
ISOELECTRONIC 659
obtained using the Onsager method: temperature, the dependence V (T ) in the single-
% phase region is close to linear. An important para-
1 meter of an isobaric process is the thermal volume
F = −k B T ln(2 sinh 2K)
2 expansion coefficient α = (1/V )(∂V /∂T )P . For
& an ideal gas α = 1/T . For a solid at tempera-
π tures higher than the Debye temperature, a good
dq
+ sinh−1 (cosh 2K coth 2K − cos q) , approximation is considered to be α = const, un-
4π
−π less the system undergoes a phase transformation.
In an isobaric process, the system, owing to the
where K = J /(k B T ). At the Curie temperature thermal flow toward it, not only heats up but also
T C a second-order phase transition takes place, produces work. Therefore, CP > CV where CP
and the speci c heat cp has a logarithmic sin- and CV are the speci c heats at constant pressure
gularity cp = A + B ln |(T − T C )/T C |. In the and constant volume, respectively.
vicinity of the Curie point, the magnetization M
tends to zero, M ∝ (T − T C )1/8 ; the magnetic ISOCHOR (fr. Gr. ισ oς , equal, and χωρα , region)
susceptibility χ diverges as (T C − T )−7/4 ; and the A line on a thermodynamic compositional
correlation length r c grows according to the law phase diagram for a constant volume (isochoric)
r c ∝ |T − T C |−1 . Numerical calculations show process; plots of the functions P (T ) and V =
that the three-dimensional Ising model also ex- const (P , T , and V are pressure, temperature and
periences a second-order transition, and has the volume, respectively).
critical indices α = 0.08, β = 0.33, and ν = 0.64.
ISOCHORIC PROCESS
ISOBAR (fr. Gr. ισ oς , equal, and βαρoς , weight, A process performed in a physical system at
heaviness) constant volume. Thermodynamic compositional
A line on a compositional phase diagram phase diagrams represent an isochoric process by
showing an equilibrium isobaric process taking an isochor. In an equilibrium thermodynamic sys-
place at constant pressure. The isobar of an equi- tem without external fields, an isochoric process is
librium process in a single component system is described by the dependence P (T ) at V = const
the plot of the function V (T ) at P = const (where (where P , T , and V are the system pressure, tem-
V , T and P are the volume, temperature, and pres- perature and volume, respectively). For an ideal
gas, P = RT /V (Charles’ law) where R is the
sure, respectively).
gas constant. The function P (T ) is monotonic in
the single phase region (solid, liquid, gaseous state
ISOBARIC PROCESS
of matter), and has discontinuities at the tempera-
A process occurring in a physical system at tures of rst-order phase transitions.
constant pressure, or, in a narrower sense, an
equilibrium thermodynamic process involving the ISOELECTRONIC
variation of system parameters with the pres- A situation when two different objects possess
sure held constant. The dependence V (T ) at an equal number of electrons, e.g., the 48 electrons
P = const, where V , T and P are the volume, in the unit cell of diamond (8 carbon atoms) and
temperature and pressure, respectively, is called an boron azide BN (4 molecules), and the 20 elec-
isobaric process equation, and its plot is called an trons of the molecules NaF and MgO. For these
isobar. The isobaric process equation can be of an cases the same or similar crystal structures are of-
involved form; but for a given state of matter, V is ten observed: in the first case, the diamond and
a continuous function of T and changes abruptly the sphalerite lattices; in the second case, the rock
at the temperature of a rst-order phase transition. salt (NaCl, simple cubic) lattice. Sometimes the
The simplest form of the dependence V (T ) is that term isoelectronic is interpreted as an equal num-
of an ideal gas, V = RT /P , where R is the gas ber of valence electrons, e.g., when substituting a
constant (Gay-Lussac law, J. Gay-Lussac, 1802). Na atom for a K atom, or Br for Cl in an NaCl
For a solid at temperatures higher than the Debye crystal, or C for Si in diamond, and so on.
660 ISOELECTRONIC SEQUENCE
induced by them in the band gap, and also by the of the matrix atoms and the impurities inside a unit
elastic stress fields arising due to a mismatch be- cell, as well as to the electron polarization and lo-
tween the covalent radii of the impurity and matrix cal lattice distortion. If the solubility of impurity
atoms. Isovalent impurities can give arise to long- atoms is sufficiently high then their concentration
range strain fields. could increase to the point where isovalent impu-
The short-range potential of an isovalent im- rity atom pairs or even larger clusters form.
purity is due to a difference between the potentials
This Page Intentionally Left Blank
Jj
JACCARINO–PETER EFFECT (V. Jaccarino, figuration differing from a linear configuration, in
M. Peter, 1962) the state of electronic degeneracy due to vibronic
A reduction of the paramagnetic limit in su- instability (see Jahn–Teller theorem). Similar ef-
perconductors with magnetic ions due to the mu- fects are observable in the state of pseudodege-
tual compensation of the external magnetic field neracy (Jahn–Teller pseudoeffect) and in linear
and the internal exchange field of magnetic cen- systems (Renner effect). A direct consequence of
ters. The Jaccarino–Peter effect occurs in com- the Jahn–Teller theorem is the formation of a com-
pounds with an antiferromagnetic electron–ion plex adiabatic potential, and the impossibility of
exchange interaction, results in an increase of separating the electronic and nuclear motions in
the upper critical eld, and can cause (e.g., in an adiabatic approximation. In contrast to the lat-
Sm1−x Eux Mo6 S8 ) stimulation of superconduc- ter, for the determination of the energy spectrum
tivity by magnetic fields. and the wave functions (vibronic states) of a poly-
atomic system in an f -fold degenerate electronic
JAHN–TELLER DISTORTION
state, it is necessary to solve a system of f coupled
Spontaneous distortion of the initial (highly equations.
symmetric) configuration of nuclei Q0 of a poly-
Doubly-degenerate E-term. For a doubly-dege-
atomic system in the presence of an electronic
nerate orbital E-term, in a linear approximation
degeneracy to a lower symmetry arrangement due
with respect to the vibronic interaction (linear E–e
to the absence of a minimum of the adiabatic
problem), the adiabatic potential takes the form
potential in this configuration (see Jahn–Teller the-
of a “sombrero” (see Fig. 1). Under free rotation
orem). The term Jahn–Teller distortion is used
along its groove, the system changes its configu-
in a conventional sense since at points of elec-
ration continuously in the space of symmetrized
tronic degeneracy, the adiabatic potential looses its
E-displacements of Qθ and Qε which, for an
meaning as the potential energy of nuclei in the
field of electrons. The lack of a minimum of the octahedral configuration, are related to a tetrago-
adiabatic potential at the point Q0 complicates the nal Jahn–Teller distortion. The groove depth (at
dynamics of the nuclear motion, and in the general the minimum), counted out from the degeneracy
case, may not lead to a static distortion of the con- point, is called the stabilization energy of the
Jahn–Teller effect; EJT E = V 2 /2k where V is the
figuration Q0 , but rather to a more complex set E
of properties. A static Jahn–Teller distortion can vibronic constant, and kE is the force constant. By
be observed either directly, in the presence of a taking into account the quadratic terms of the vi-
low-symmetry perturbation provided by a vibronic bronic interaction (quadratic E–e problem) along
interaction (so-called vibronic amplification), or in the groove of the adiabatic potential, there appear
a cooperative Jahn–Teller effect (see Jahn–Teller three minima Mi at points Φ = 0, 2π/3, 4π/3
effect). which alternate with three peaks Ci together with
the Jahn–Teller deformations of octahedral and
JAHN–TELLER EFFECT (H. Jahn, E. Teller, tetrahedral systems. In the case of low barriers
1937) between the minima, the free “rotation” becomes
A set of phenomena, effects, and trends which hindered, otherwise for a high barrier the nuclear
appear in polyatomic systems which have a con- motion acquires a tunneling character which leads
665
666 JAHN–TELLER EFFECT
effect is characterized by the presence of low-lying degeneracy for an odd number of electrons (to-
and temperature-dependent interacting electronic tal spin S is half-integer). A more limited theorem
and phonon (acoustic) branches. With increasing states that if a magnetic ion is at a crystal site of
temperature the correlations of local distortions such a high symmetry that its ground state is not
weaken, and at the critical temperature T c a phase the Kramers minimum, then it will be favorable
transition to the distortion-disordered phase takes energetically for the crystal to distort and lower the
place. Since the phase transition is structural it symmetry to a sufficient extent to remove the de-
induces anomalies of the elastic crystalline prop- generacy. There are some widespread simple, and
erties. At the same time, as the static lattice dis- in principle incorrect, formulations of the Jahn–
tortions below T c induce substantial changes in Teller theorem which claim that in the presence
the electronic structure which split the degener- of an electronic degeneracy a polyatomic system
ate energy levels (bands), in the vicinity of T c must spontaneously distort to remove the degen-
all the crystalline electronic parameters (mag- eracy. However, in the Jahn–Teller effect which
netic, electric, etc.) also exhibit appreciable anom- follows from the theorem the degeneracy is not
alies. The relationship between the electronic and necessarily removed, but rather it can transform
phonon subsystems, which plays a determining from an electronic degeneracy to a vibronic one
role in crystals with a Jahn–Teller cooperative ef- (see Vibronic states), which lacks a Jahn–Teller
fect, leads to strong interrelationships between the distortion in the strict sense, since it is of a dy-
elastic, electric, and magnetic properties of these namic character.
crystals. This is the cause of some original effects
in these crystals: appreciable influence of a mag- JELLIUM MODEL
netic field on a structural phase transition, giant The simplest model of an interacting electron
magnetostriction and electrostriction, and so on.
gas; it assumes that the positive charge com-
The Jahn–Teller cooperative effect has been found
pensating for the electronic charge is uniformly
and studied in many compounds. The most widely
distributed in space. The only parameter that de-
studied are crystals containing rare-earth ions (like
termines system properties is the electron density,
TmVO4 ), spinel (NiCr2 O4 ), perovskites (KCuF3 )
or the average distance between electrons r s . For
with Jahn–Teller ions of transition metals, and in-
a dense gas r s 1 (in Bohr radii) and in cal-
termetallic compounds (TmCd).
culations of the total energy, the application of
Examples of the Jahn–Teller cooperative and
perturbation theory to the electron–electron in-
pseudoeffects are found in all major areas of the
teraction is justified. The zero-order term in the
physics and chemistry of molecules and crystals,
interaction yields the kinetic energy; the first-order
including spectroscopy over the entire electro-
magnetic spectrum, crystal physics and crystal term is the exchange energy, while higher-order
chemistry, structural phase transitions, phenomena
terms yield the correlation energy. Calculations
of photon transport, physics of impurity crystals, of the latter involve divergences, but their sum
mechanisms of chemical reactions, chemical ac- yields the correct final result: an expansion of the
tivation, and electron-conformational interactions energy in powers of r s , with a nonanalytic con-
in biology (see Vibronic effects in optics, Vibronic tribution proportional to ln r s . For a rarefied gas,
effects in radio-frequency spectroscopy, Inversion r s
1, the uniform distribution of electrons fails
splitting). to represent the ground state, and the so-called
Wigner crystallization (see Wigner crystal) takes
JAHN–TELLER THEOREM place. In this limiting case a strict calculation of
A theorem concerning the properties of a de- energy is also possible; and r s −1 acts as the ex-
generate polyatomic system. It can be formulated pansion parameter. In real crystals (2 r s 6),
as follows: if the adiabatic potential of a nonlin- calculations performed with the help of variational
ear polyatomic system has several branches f that calculus or by application of various interpolation
coincide at the same point (f -fold degeneracy), formulae often give very close results. The jellium
then at least one of these branches lacks a mini- model is often used for investigations of the prop-
mum at this point. An exception is Kramers spin erties and spectra of the ground state in solids,
JOSEPHSON EFFECTS 669
JEWEL
See Gems.
JOSEPHSON JUNCTION j
2e
A point contact or tunnel junction between Aij = A dr,
two superconductors which displays Josephson c
i
effects. For point contacts with direct conductivity,
or for superconducting microbridges, the charac- where the amplitude of the order parameter is as-
teristic dimension of the constriction should not sumed to be constant within a grain or granule.
exceed the coherence length ξ . On the other hand, Here ϕi is the phase of the order parameter in the
to prevent the destruction of the Josephson effect ith granule; Jij is the value of the weak bond be-
by uctuations the contact resistance should not be tween the granules i, j ; A is the vector-potential
too high (usually less than 100 ). According to of the magnetic field B = ∇ × A.
their design, superconducting film microbridges,
superconductor–insulator–superconductor tunnel JOSEPHSON RADIATION
junctions, superconductor–semiconductor–super- Electromagnetic radiation from a Josephson
conductor junctions, and others are singled out. junction due to the passage of an alternating super-
Josephson junctions are used in superconducting conducting current j s = j c sin ωt upon applying a
quantum interference devices (SQUIDs), which direct voltage to the electrodes. Josephson radia-
measure magnetic flux and its space derivatives, tion had been predicted by B. Josephson (1962),
and also any other electrical parameters (e.g., and was detected by I.K. Janson, V.M. Svistunov,
voltage, current), which can be transformed into I.M. Dmitrenko (1965). The radiation frequency ω
a magnetic flux coupled to the superconducting is proportional to the voltage V through the
circuit containing the Josephson junction. These Josephson relation ω = 2eV , which expresses
junctions are used in low-temperature thermo- the energy conservation law at the passage of a
metry, as basic elements of superconductor digital Cooper pair through the junction, with the simul-
electronics, and as nonlinear elements for detect- taneous emission of a photon with energy ω. For
ing and mixing electromagnetic radiation over a 1 μV the frequency ω/(2π) is 483.6 MHz. Joseph-
wide range of frequencies. son radiation may be to a large extent coherent and
672 JOSEPHSON VORTEX
monochromatic (relative width of the frequency The reflections of Josephson vortices from the
band 10−7 ). The power generated by a single boundaries of the tunnel junction causes stand-
Josephson junction is usually small (10−10 W). ing modes of oscillation. The motion of a one-
In addition to the fundamental frequency, har- dimensional chain of vortices may be regarded
monics and subharmonics are often observed, as as a non-stationary Josephson effect, because the
a result of the nonlinear properties of Josephson frequency of the field oscillations caused by this
junctions. motion, and the respective average voltage across
The reciprocal effect, that is the excitation by the barrier V , are related by the Josephson fre-
the external radiation of steps on the current– quency expression 2eV = ω.
voltage characteristic of the junction located at the
voltages Vn = nω/(2e), n = 1, 2, . . . , was ob- JOULE HEAT (J.P. Joule)
served by S. Shapiro (1963) (see Shapiro steps). The heat emitted by an electrical conductor
At a finite temperature noise due to fluctuations of upon the passage of an electric current through it.
the voltage on the junction [Nyquist noise and shot If J is the current density and σ is the electrical
noise (H. Nyquist, 1927)], and also fluctuations of conductivity then the Joule heat emitted per unit
the order parameter, provide a finite width to the time per unit volume is equal to J 2 /σ (Joule–Lenz
Josephson radiation. In the case of Nyquist noise law; established in 1841 by J.P. Joule and con-
the width
f = (2e/)2 RT , where R is the junc- firmed in 1842 by the experiments of E.H. Lenz).
tion resistance in the normal state, corresponds to
f ∼ 40 kHz at T = 4 K and R = 10−3 . JOULE–THOMSON EFFECT (J. Joule,
W. Thomson, 1853)
JOSEPHSON VORTEX, uxon Change of temperature of a gas at a change of
A localized distribution of current density in pressure under adiabatic conditions. The Joule–
a superconducting tunnel junction with the char- Thomson effect is the basis of the most widely
acteristic dimension λJ = [c2 /(8πeΛJ c )]1/2 ∼ used methods of cryogenic cooling (see Cryogenic
10−3 to 10−1 cm, such that the total magnetic temperatures).
flux Φ created by this current is equal to the
ux quantum Φ0 = h/(2e), where λJ is called the JUMP CONDUCTIVITY
Josephson penetration depth, J c is the Josephson See Hopping conductivity.
critical current density, and Λ is the total depth of
JUNCTION
the magnetic field penetration. A Josephson vortex
See Heterojunction, Homojunction, Josephson
is a topological soliton solution of the sine-Gordon
junction, Metal–semiconductor junction, Semicon-
equation,
ductor junction, Tunnel junction.
∂2ϕ 1 ∂2ϕ 1
2
− 2 2 = 2 sin ϕ, JUNCTION CAPACITANCE
∂x c0 ∂t λJ
See Barrier capacitance.
describing the distribution of phase ϕ in the
Josephson junction. Josephson vortices, unlike JUNCTION PHOTOEFFECT, gate photoeffect
Abrikosov vortices in type II superconductors, An internal photoeffect which appears at the
do not have a normal state core, so they only illumination of a junction or gate, i.e. a rectify-
weakly dissipate their energy when moving along ing contact. It depends on the separation of the
the junction. Therefore, in addition to station- electron–hole pairs generated by the light (see
ary (immobile) Josephson vortices there are vari- Semiconductor junction) by means of the electric
ous related dynamic soliton structures. The tran- field at the junction. The cut-off layer (see Metal–
sient formation of “soliton–antisoliton” (vortex– semiconductor junction) at the semiconductor–
antivortex) pairs with a frequency ω in the energy metal boundary involves a p–n junction which
gap for small oscillations of the phase, ω < ω0 = possesses rectifying properties. As a result, the
c0 /λJ (ω0 is Josephson plasma frequency), is one illumination of the rectifying contact causes a cur-
of them. rent to flow in the external circuit, and for the case
JUVENILE SURFACE 673
675
676 KAUZMAN PARADOX
temperature jump is a universal physical phe- the Kellerman model for NaCl have indicated the
nomenon which arises at low temperatures at the existence of two types of phonon waves: acoustic
interface between any heterogeneous media. The and optical. The frequency distribution function is
theory of this phenomenon is based on the as- non-monotonic, in contrast to the Debye case. An
sumption that the heat transport between the con- important shortcoming of the Kellerman model
tacting bodies is carried by thermal phonons. For is that it fails to take into account the polar-
example, the main contribution to Rk is from the ization of ions displaced from their equilibrium
difference between the phonon spectra of He II positions. This shortcoming was overcome by the
and a solid body (I.M. Khalatnikov, 1952). construction of the model of a lattice of polarized
As the heat loading and the corresponding tem- ions (K.B. Tolpygo), and the shell model that is
perature gradient
T increase, processes in the equivalent to it (W. Cochran). See also Phonon
liquid begin to play a more significant role. The spectra.
value of Rk for large heat loading decreases as Q
increases. KELVIN EQUATION (W. Thompson, Lord Kelvin,
The Kapitza temperature jump hampers the 1871)
cooling of bodies to extremely low temperatures, Thermodynamic equation describing the de-
and must be allowed for in designing deep freezing pendence of the vapor pressure Pr (or of the
systems for achieving cryogenic temperatures. solubility Cr ) on the radius of curvature r of the
surface separating the phases:
KAUZMAN PARADOX (W. Kauzman, 1948)
The property of a number of vitrifying systems Pr Cr 2σ v
= = exp ,
(organic, oxide, polymerizing liquids) whereby P0 C0 rRT
the extrapolation of the physical characteristics of
where σ is the surface tension at the interface be-
their liquid phase below the vitri ca tion tempera-
tween the phases, v is the molar volume of the
ture TV leads to nonphysical results. In particular,
liquid or solid, R is the molar gas constant, and
the extrapolated entropy of an overcooled liquid
P0 and C0 are the values for a flat surface.
becomes smaller than that of the crystalline phase.
The Kelvin equation was obtained from the
KELLERMAN MODEL (E.W. Kellerman, 1940) condition that the chemical potentials of the corre-
This first serious model for computing phonon sponding phases be equal to each other. It follows
optical spectra takes into account inter-ionic in- from this equation that the saturation vapor pres-
teraction forces, and it served as the prototype sure (or solubility) on the boundary with drops of
for a number of subsequent models suited for liquid or with crystals increases with the decrease
simple ionic crystals. In addition to the Coulomb of their dimensions, whereas the pressure inside
interaction between the ions regarded as point a bubble or by the surface of a concave menis-
charges, the Kellerman model accounts for the ef- cus (i.e. when the radius of curvature is negative)
fects of charge cloud overlapping between nearest is lower than that over a flat surface. Changes in
neighbors. The contribution of the latter to the in- the saturation vapor pressure (or solubility) only
teraction energy is most often approximated as become significant at very small radii of curva-
b/r n or Ae−αr , where r is the distance between ture (usually, less than 10−5 cm). Thus, the Kelvin
the atoms. The parameters of the overlap force, equation is usually applied when studying systems
b and n, or A and α, are calculated by proceeding containing small objects (nuclei of a new phase,
from known values of the bulk modulus (recipro- colloidal systems, etc.), as well as when dealing
cal of compressibility) and lattice constants. The with capillary phenomena, and with processes as-
potential energy of the crystal, U , is defined as the sociated with the formation of a new phase. For
sum of pair interaction energies of ions in the lat- instance, according to the Kelvin equation, the
tice. The vibrational frequencies are solutions of condensation of a vapor in capillaries and micro-
the secular equation, elements of the determinant cracks occurs at vapor pressures lower than P0 .
of the latter being expressed in terms of the sec- For the establishment of thermodynamic equilib-
ond derivative d2 U/dr 2 . Calculations following rium, the Kelvin equation points out the direction
KIKOIN–NOSKOV EFFECT 677
of transport of the substance from smaller drops or of a magnetized ferromagnet, and the reflected
crystals to larger ones. The nuclei of a new phase light is elliptically polarized, with the major axis
cannot start growing before the supersaturation de- of the polarization ellipse rotated through a cer-
fined by the Kelvin equation has been reached. tain angle relative to the plane of polarization of
the incident light.
KERR EFFECT (J. Kerr, 1875)
A quadratic electrooptical effect involving the
emergence of, or change in, birefringence resulting KHAIKIN OSCILLATIONS (M.S. Khaikin, 1960)
from the action of an electric field on a medium. Sharp nonmonotonic dependence of the sur-
It was first found in isotropic media (glass, po- face impedance of metals on a weak magnetic
lar liquids) by the enhanced transmittance in a field B ∼ 0.01 to 10 mT. Khaikin oscillations are
transverse electric field of a polarization cell con- observed at electromagnetic radiation frequencies
sisting of a crossed polarizer and analyzer with a ω ∼ 1010 –1011 s−1 in the course of experiments
glass plate between them. The Kerr effect occurs at liquid helium temperatures (T < 10 K), when
due to anisotropy of the refractive index exhibited the external magnetic field is applied parallel to
by optically uniaxial crystals. The optical axis is the surface of a metal sample. Initially discovered
aligned along the electric field, and the magnitude by Khaikin in monocrystals of tin, cadmium and
of the double refraction,
n, is proportional to the indium; later these oscillations were found and
square of the electric field intensity,
n = BE 2 , studied in a variety of metals and semimetals (bis-
where B is the Kerr constant which depends muth, aluminum, copper, gallium, etc.). Khaikin
on the temperature, light wavelength, molecular oscillations are due to the resonance absorption
structure, and the phase state of matter.
of an electromagnetic wave at transitions of elec-
The Kerr effect occurs both in static or low-
trons passing from one magnetic surface level to
frequency fields and in the high-frequency (op-
another. Since these surface levels are not equidis-
tical) range; it may be treated in the terms of
tant, Khaikin oscillations are not periodic with
nonlinear optics as a manifestation of the nonlin-
ear polarization of a medium in external electric respect to reversals of the magnetic field, and are
fields. The Kerr effect occurs in crystals of all the result of superpositions of different resonance
symmetry classes (see Crystal symmetry). In crys- series. Resonance values of the magnetic field are
tals that lack a center of symmetry and have a proportional to ω3/2 . The shape of a Khaikin os-
particular shape it is overshadowed by the much cillation resonance line is related to the dispersion
stronger linear electrooptic effect (the Pockels ef- law of conduction electrons, and to their scattering
fect) when the applied electric field is oriented in the bulk and at the surface of the metal.
along the direction of the light propagation. The
phenomenological description of this effect re- KIKOIN–NOSKOV EFFECT (I.K. Kikoin,
quires a polar fourth-rank tensor. M.M. Noskov, 1933)
The Kerr effect in solids is usually associated
A photomagnetoelectric effect involving the
with both lattice strains and a rearrangement of
generation of an electric field under the action
electronic states. The effect in a low-frequency
of light in a semiconductor placed in a mag-
electric field is usually accompanied by elec-
trostriction, thus the Kerr effect for allowed and
netic field. The electric field is perpendicular to
forbidden deformations should be distinguished. both the magnetic field and the direction of flow
Because the Kerr effect “switches” almost instan- of the current carriers, the latter diffusing in the
taneously it is exploited for the high-frequency semiconductor from the illuminated side, where
amplitude and phase modulation of light (see Mod- the absorbed photons produce electron–hole pairs,
ulation spectroscopy). then drifting toward the non-illuminated side. The
In addition to the above described electrooptic Kikoin–Noskov effect manifests itself through a
Kerr effect, there is also a magnetooptical Kerr ef- pronounced non-uniform concentration of minor-
fect (J. Kerr, 1876) which arises when linearly po- ity carriers brought about by the strong absorption
larized light is normally incident upon the surface of light.
678 KIKUCHI LINES
KIKUCHI LINES (S. Kikuchi, 1928) the equations: ∂S(γ )/∂γi = −Yi , where S(γ ) is
Lines found in the electron diffraction pattern a nonequilibrium entropy. The kinetic coefficients
(see Electron diffraction analysis) of thick crystals. obey the symmetry principle Kij = Kj i , i.e. the
The usual Laue–Bragg spots that appear on the Onsager relations. (Note that Kij (B) = Kj i (−B)
photographic plate of an electron diffraction pat- when the kinetic coefficients depend on a mag-
tern from thin crystals vanish in the case of thick netic e ld B.) Consider, for example, the case
crystals, with dark and light lines (the Kikuchi of thermodynamic forces arising from an elec-
lines) and bands (Kikuchi bands) observed in their tric field intensity E, in the presence of gradients
stead. In the general case there is a light Kikuchi of the chemical potential μ, and the tempera-
line parallel to each dark Kikuchi line. If the in- ture T . To describe transport phenomena in a
tensity between the pairs of Kikuchi lines differs metal Eqs. (1) assume the forms
from that of the background, then it is termed a
1 ∂μ 1 ∂T
Kikuchi band (it is a “surplus” band, if its in- Ik = σki Ei − + Kki ,
tensity exceeds the background, and a “deficit” e ∂xi T ∂xi
band in the opposite case). The Kikuchi lines are μ
Qk − Ik (2)
produced by electrons that underwent an inelastic e
scattering (on phonons, plasmons, or on valence
= −Kki E − 1 ∂μ + κ 1 ∂T ,
or inner atomic electrons) by way of their diffrac- i
e ∂xi
ki
T ∂xi
tion at the reflecting planes of a crystal lattice.
The structure of Kikuchi patterns depends upon where Ik is an electric current, Qk is an energy
the diffraction conditions and the crystal thick- flux (thermal current), xi are space coordinates,
ness. Almost all Kikuchi lines get split on crossing and e is the electric charge. The kinetic coeffi-
the Kikuchi bands. With increasing crystal thick- cients here are the conductivity tensor σki = σik ,
ness the dark Kikuchi lines turn into light ones. the thermoelectric coefficient tensor, the Peltier
A rigorous quantitative description of the contrast thermal e.m.f. coefficient (see Thermoelectric phe-
nomena) Kki = Kik , and the tensor κ = κ
of Kikuchi lines and bands is given by the theory ki ik
of dynamic radiation scattering. One can utilize which is related to the thermal conductivity tensor,
the Kikuchi lines to determine the crystal orien- χ̂ , by the expression χik = T −1 (κ̂ − K̂ σ̂ −1 K̂)ik .
tation and to obtain information on crystal lattice When B = 0 the energy flow is expressed in terms
distortions. Kikuchi lines are an electron-optical of the thermal conductivity tensor, χ̂ , under the
analogue of Kossel lines from X-ray diffraction. conditions Ik = 0 and Qk = χki (∂T /∂xi ). The
fact that the electric field E and the gradient of
KILO. . . (fr. Gr. κιλιας , thousand) the chemical potential μ appear in the combina-
Prefix to denote a decimal multiple equal to tion E − (1/e)(∂μ/∂x) indicates the validity of
1,000 times an original unit, abbreviated as k. An the Einstein relations linking the conductivity with
example: 1 km = 1,000 m. the diffusion coef cient.
When analyzing the equations of fluid dy-
KINETIC COEFFICIENTS namics applied to a non-ideal liquid, dissipative
The quantities Kij that define the linear depen- fluxes similar to the fluxes Ii in Eq. (1) appear.
dence of thermodynamic flows or fluxes, Ii , on the These dissipative fluxes are determined by such
thermodynamic forces, Yj , that give rise to these kinetic parameters as the first and second viscosity
fluxes in accordance with the expression coefficients, the diffusion coefficient, the thermod-
iffusion coefficient, and the thermal conductivity.
dγi n
≡ Ii = Kij Yj (1)
dt KINETIC DECOUPLING OF COOPER PAIRS
j =1
The phenomenon of the decrease of the energy
for a physical system close to equilibrium. The pa- gap and of the order parameter in a supercon-
rameters γi describing the nonequilibrium state, ductor brought about by the flow of supercurrent
and the thermodynamic forces, Yi , are related by (kinetic decoupling effect). After carrying out a
KINETIC INSTABILITY OF SPIN WAVES 679
transformation to a moving reference frame, the great number of particles, such as fluids, possess
energy of a Cooper pair traveling with a velocity such internal conditions that the kinetic stage of
v s is found to be equal to Δ − pv s , where Δ is evolution is completely missing.
the value of the gap at v s = 0. Since the electron Kinetic equations are widely used in statis-
momentum assumes values along the direction of tical physics applied to nonequilibrium systems:
motion varying between −pF and pF , where pF in the theory of gases, theory of solids, theory
is the Fermi momentum, the minimum energy gap of turbulence, plasma physics, theory of particle
is given by Δmin = Δ − pF vs . The maximum vs transit through matter, theory of radiation transfer,
value for a gap different from zero is Δ/pF cor- etc. A kinetic equation is ordinarily constructed,
responding to a maximum Cooper pair velocity following L. Boltzmann, based on the phenom-
vc ≈ 104 cm/s. This kinetic decoupling of Cooper enological principle that relates a change in the
pairs indicates the existence of a maximum dissi- function f during a short period of time to the bal-
pationless current density j c in a superconductor ance of incoming and outgoing probability fluxes
(depairing current density) with a magnitude of for an element in the single-particle phase space.
the order of j c ≈ ns ev c ≈ 109 A/cm2 (see Cur- Methods have already been developed for the
rent states). self-consistent construction of a kinetic equation
within the formalism of perturbation theory, with
KINETIC EQUATION reference to some small parameter, proceeding
The basis for the statistical description of non- from the macroscopic dynamic equations which
equilibrium systems comprising a great number describe the motion of particles in the system. For
of similar objects (particles, quasi-particles, etc.); instance, one can derive the Boltzmann equation
it specifies the change with time, t , of a distrib- from the Liouville equation of classical mechan-
ution function, f (t), defined in the phase space ics written for the distribution of probabilities in
of an individual object. The term was coined by the phase space of all the particles in a gas to the
L. Boltzmann in the latter part of the 19th century, lowest approximation in the density. In quantum
and he also proposed the Boltzmann equation, a statistical physics an equation similar to the Boltz-
kinetic equation for the single-particle distribution mann equation (see Quantum kinetic equation),
function of a classical monatomic gas. A kinetic which takes into account quantum-mechanical ef-
equation should satisfy the following condition: fective cross-sections and symmetry requirements,
a change in the distribution function f during a is obtained from the first non-vanishing terms in
short time interval,
t , is completely determined the approximation of the density, proceeding from
by the function f itself, and by the external forces the quantum Liouville equation for the system
acting at that instant of time t , i.e. the history of density matrix.
the past evolution of a system in no way affects
its current change. When this condition is satisfied KINETIC EQUATION, QUANTUM
with a sufficient precision then the system is re- See Quantum kinetic equation.
ferred to as being in the kinetic stage of evolution,
which means being confined within a finite time KINETIC INSTABILITY OF SPIN WAVES
period which depends both upon the system’s indi- The exponential growth of the amplitudes of
vidual characteristics (e.g., mean particle density, some groups of spin waves occurring in the case
nature of their interactions), and its initial distribu- of highly nonequilibrium distributions in the spin
tion. In the limit of short times the kinetic stage is wave spectrum. Such nonequilibrium distributions
restricted by the condition of a sufficiently small emerge, e.g., at the parametric excitation of spin
deviation from equilibrium when the magnitude waves, with their population growing by many or-
of the particle correlation length is of the order of ders of magnitude within a very narrow spin wave
the radius of their interaction. On the other hand, spectral region. When the parametrically excited
the times must not be so long that weak, although waves reach a certain critical amplitude they give
long-lasting, time correlations can influence the rise to kinetic instability at the lower end of their
system’s evolution. Some systems consisting of a spectrum. The corresponding critical amplitude of
680 KINETIC ISOTOPE EFFECT
parametrically excited spin waves is reached when described by frequency-dependent kinetic coef -
the pumping amplitude exceeds its threshold value cients, as distinct from the slow processes which
by 10–20 dB. involve constant or static kinetic coefficients. The
distinction between fast and slow kinetic processes
KINETIC ISOTOPE EFFECT is largely a matter of convention, and depends on
Influence of the isotopic composition of a re- the choice of a characteristic microscopic time
acting substance on the rate of a reaction, where scale. If one takes for this scale the mean free
the reaction rate constants of the molecules de- time of particles or quasi-particles, using the ki-
pend on the isotopes present. The kinetic isotope netic equations with a self-consistent eld, such as
effect arises from the effects of atomic masses on the Vlasov equation, it will be sufficient to deal
the zero point vibrational levels of reacting mole- with fast processes; with slow processes being de-
cules. The primary kinetic isotope effect depends scribed by equations of the fluid dynamics type in
upon the masses of the nuclei directly involved this case. Slow processes correspond to those as-
in the reaction, and the secondary effect involves sociated with transport phenomena.
masses of other nuclei. An abnormally large ef-
fect may be associated with tunneling reactions KINK
in solids, where the mass affects the probability A particle-like excitation (soliton) in one-di-
of tunneling (see Tunneling phenomena in solids). mensional systems with a discrete degeneracy of
A magnetic kinetic isotope effect has been ob- its ground state energy. An example of a kink
served, viz., the dependence of the rates of some is a domain wall in a crystal with quasi-one-
chemical reactions on the nuclear magnetic mo- dimensional ordering (magnetic ordering, ferro-
ments associated with reacting free radicals. The electric ordering, etc.). The simplest soliton so-
kinetic isotope effect is of use in studies of the
lution of the sine-Gordon equation (see Soliton)
mechanisms of chemical transformations, and for
describes a kink. Kinks play an important role
the separation of isotopes.
in the thermodynamics of quasi-one-dimensional
crystals; in particular, they determine the distur-
KINETIC MOMENT
bance of long-range order (see Long-range and
The same as Angular momentum.
short-range order) known as the central peak of
KINETIC PHENOMENA the correlation functions.
Physical phenomena occurring in non-equilib-
rium macroscopic systems under the action of KINKS ON DISLOCATIONS
external factors, or resulting from the collisions Elements of a dislocation line, which connect
of particles or quasi-particles. These processes two of its segments, which lie in adjacent val-
are described by the equations of macroscopic leys of the potential Peierls relief. Depending on
physics, including the kinetic equations, equa- the kink width ω = a[U0 /(2UP )]1/2 , which is
tions of fluid dynamics, Maxwell equations in a determined by the balance between the Peierls
medium, etc. Kinetic phenomena can be classi- energy (UP ) and the linear dislocation tension
fied into slow and fast varieties. The first class energy (U0 ), sharp and smooth kinks are distin-
of slow processes involves those those for which guished. The kink width ranges from several lat-
the characteristic times of change of either the tice constants (a) in the case of covalent crystals to
macroscopic parameters describing a given non- tens of lattice constants in the case of close-packed
equilibrium system, or of the externally acting metals. The kink energy is given by the equation
factors, are long compared to the characteristic Uk = (2a/π)(2U0 UP )1/2 . At zero temperature,
microscopic times in this system. Fast kinetic the dislocation line may have only geometrical
phenomena are those for which the characteris- kinks if its end points are fixed in different val-
tic times of change of the external factors and leys of the potential relief. At T = 0 the thermal
of the macroscopic parameters are comparable fluctuations give rise to pairs of kinks; these pairs
to, or smaller than, the microscopic characteris- consist of kinks of opposite sign. The equilib-
tic times. In the latter case kinetic phenomena are rium concentration of paired kinks is defined as
KIRKWOOD APPROXIMATION 681
ck+ ck− = (1/a 2 ) exp[−(2Uk )/(k B T )]. Kinks on a two-component substitutional solid solutions that
dislocation line are of crucial importance for de- the rate of marker displacement along the x axis,
scribing dislocation mobility in the Peierls relief. v, equals: v = (DA − DB )∂CA /∂x. The mark-
Due to the translational symmetry along the dis- ers move toward the component possessing the
location line, the kink also exhibits a potential larger diffusion coefficient. The Kirkendall effect
relief, which is called the Peierls relief of the per se appears when the flow of vacancies involves
second kind. The magnitude of this secondary re- only dislocations. Other fluxes (nucleating pores,
lief is small in the cases of face-centered cubic cracks, etc.) can be removed, e.g., by applying a
and hexagonal close-packed metals, which exhibit small uniform compression to the specimen (see
smooth kinks, and is comparable to aUP in the Barodiffusional effects).
case of sharp kinks in silicon and germanium.
KIRKWOOD APPROXIMATION (J.G. Kirkwood,
KIRKENDALL EFFECT (E.O. Kirkendall, 1947) 1946)
Transport of material in the diffusion zone dur- One of the so-called superposition approxima-
ing a homogenization process involving unequal tions that relate multiparticle correlation functions
partial diffusion coef cients DA and DB of two with each other. An example of the Kirkwood ap-
interdiffusing components A and B. The process proximation is the following decomposition of a
is sustained by the self-maintaining action of a (3)
three-particle correlation function fAAB in terms
“source” of vacancies whose current has a mag- (2) (2)
of pairwise correlation functions fAB and fAA :
nitude g v proportional to [(DA − DB )/ω]∇C v
fAAB (r 1 , r 2 , r 1 )
(C v is the vacancy concentration, and ω is an (3)
atomic volume), and by the vacancy flow itself.
(2) (2)
The flux which determines the effect involves dis- ≈ fAB |r 1 − r 1 |, t fAB |r 2 − r 1 |, t
locations whose Burgers vectors include an edge
(2)
component. In phenomenological terms the Kirk- × fAA |r 1 − r 2 |, t ,
endall effect is a creep under an effective stress in
with A and B designating the type of parti-
the diffusion zone maintained by supersaturation
cles, r 1 , r 2 and r 1 , are the coordinates of the
with vacancies, Δ/Cv0 (Δ = C v − Cv0 , where
A, A and B particles, respectively; and t is time.
Cv0 is an equilibrium vacancy concentration). The
The Kirkwood approximation can be employed to
effect was experimentally discovered on brass–
break up an infinite chain of coupled correlation
copper specimens (see Fig.) when inert markers
functions that describe the spatial distribution of
that were placed on the plane of initial contact be-
classical particles in statistical physics, and in the
tween the specimen components drifted due to the
kinetics of dense gases and condensed media. Its
transport of matter in the diffusion zone. It fol-
use involves a transformation to a reduced descrip-
lows from the simple Kirkendall effect theory for
tion of the fluctuation spectrum of a multi-particle
system on the level of two-particle correlation
functions. This approximation facilitates the com-
putation of a number of thermodynamic parame-
ters. The validity and the regions of applicability
of the Kirkwood approximation, which does not
involve an expansion in powers of the density,
depend upon the particular problem under consid-
eration. They may be assessed by comparison with
a computer simulation, or with an exact solution
of the problem when one is available. For in-
stance, the Kirkwood approximation applied to the
diffusion-controlled transport of energy coincides
Displacement,
x, of inert markers with time in the with the exact solution for the case of immobile
brass–copper system at T = 785 ◦ C. energy fluxes, it becomes an approximate solution
682 KITTEL’S DOMAIN STRUCTURE
equation for the diagonal elements of the den- ing temperatures comes to a stop since the system
sity matrix in first order perturbation theory. The of impurities transforms through the Ruderman–
off-diagonal matrix elements are then expressed Kittel–Kasuya–Yosida interaction into the “spin
through the diagonal ones. Corrections have been glass” state in which the impurity spin orien-
calculated through terms of order λ4 , where λ is tations remain fixed, and the electron scattering
the dimensionless parameter characterizing the in- channel through spin flips is suppressed. An ap-
teraction force of impurities with electrons. For plied magnetic field also fixes the spin orientation,
higher orders it was found that the distribution and for μB B
kB T (μB is the Bohr magneton)
function equation does not reduce to the usual the minimum in R as a function of T disappears.
Boltzmann equation, but all correction terms are Depending on the particular characteristics of the
smaller than their “Boltzmann” counterparts. sample, switching on a rather strong magnetic
field can either increase or decrease the electrical
resistance.
KONDO EFFECT (J. Kondo, 1964)
Observation of a minimum electrical resis-
KONDO LATTICE (J. Kondo)
tance R in experiments with nonmagnetic metals
A lattice of localized spins of paramagnetic
(Au, Ag, Cu, Al, Zn, etc.) as a function of tem-
ions (f -state ions) experiencing an antiferromag-
perature (T ) caused by the conduction electron
netic exchange interaction with conduction elec-
scattering off impurity atoms with unfilled in-
trons. In such a lattice the collective Abrikosov–
ner electron shells and spin different from zero
Suhl resonance appears that causes a sharp in-
(paramagnetic impurities). For each material the
crease of the effective mass of electrons transform-
effect is observed below a particular temperature
ing them into heavy electrons (heavy fermions).
TK called the Kondo temperature. Such impurity
atoms can be transition series (e.g., Cr, Mn, Fe,
KOOPMANS THEOREM (T.A. Koopmans, 1933)
Co) and rare-earth elements (e.g., Ce, Tm, Yb). This theorem provides a physical interpretation
The Kondo effect is a collective one involving of energies of one-electron orbitals of a mul-
an indirect exchange interaction of the conduc- tielectron system in terms of the Hartree and
tion electrons with a paramagnetic impurity. As Hartree–Fock approximations (see Hartree–Fock
was shown by Kondo, corrections to the Born method). By Koopmans theorem the energy of
approximation for the amplitude of the electron every ith orbital taken with opposite sign is equal
spin- ip scattering from paramagnetic impurities to the ionization energy of the system from the ith
depend on the Fermi–Dirac distribution func- one-electron level. Koopmans theorem holds un-
tion. In cases where the spin exchange interac- der the approximate assumption that removing an
tion constant of the conduction electrons with electron from the ith level does not change the
the paramagnetic impurities is antiferromagnetic, orbitals of the remaining electrons. Some cases
the scattering amplitude increases logarithmically are known where Koopmans theorem is not valid,
with the approach of the scattered electron en- and the arrangement of one-electron levels by
ergy to the Fermi energy of the metal. As a result, their energies is incompatible with the results of
the spin component of resistance increases log- a Hartree–Fock calculation of the ionization en-
arithmically with decreasing temperature, and a ergy as the difference between the ground state
minimum in R appears through the competitive energies of (N − 1)-electron and N -electron sys-
contribution of other scattering mechanisms (in tems. In models different from those of Hartree or
particular, electron–phonon) resulting in the elec- Hartree–Fock, Koopmans theorem may not hold.
trical resistance decreasing with the temperature.
In the neighborhood of the Kondo temperature TK KORRINGA–KOHN–ROSTOKER METHOD
(see Abrikosov–Suhl resonance) the temperature (J. Korringa, W. Kohn, N. Rostoker, 1947), KKR
dependence of the electrical resistance changes method
sharply and tends to a finite limit at T → 0. The A non-linearized method, one of the band
growth of the electrical resistance with decreas- structure computation methods for crystals. The
684 KORTEWEG–DE VRIES EQUATION
widely used KKR method has two equivalent for- is similar to the arrangement of interference lines
mulations: one expressed in terms of multiple on a diffractogram from a monocrystal exposed
scattering theory is widely used in the theory to a divergent monochromatic X-ray beam (see
of disordered systems, and the other led to the Diffraction methods of investigation). Rays which
widespread linear muffin tin orbital approach (see obey the Bragg law are diffracted on correspond-
Linear methods of band structure computation). ing reflecting surfaces and give rise to Kossel
The solution of the Schrödinger equation in lines on a photoplate. The Kossel line positions
a crystal muf n-tin potential by the KKR method are determined by crossing the photoplate plane
is carried out with the aid of a linear combina- by Kossel cones with angular openings equal to
tion of trial basis functions. The solutions of the π − 2θ (θ is the Bragg angle), and axes normal
Schrödinger equation for a given energy inside the to the reflecting planes. The quantitative defini-
muffin tin sphere are used as the trial functions. tion of Kossel line contrast is given by the theory
They are matched continuously and smoothly to of dynamic radiation scattering. Kossel lines are
spherical waves outside the sphere which have the used to determine the lattice orientation relative to
same angular momentum. The condition of can- the crystal surface, to measure the lattice constant
celing the “tails” of all “foreign” basis functions with a relative error ∼10−5 , and to obtain infor-
inside a given muffin tin sphere provides the sec- mation about crystal distortions, and degradations
ular equation of the KKR method; an equation of crystal lattice symmetry (see Crystal symme-
which is non-linear in the energy, and has the form try).
1 1
det1Blmt,l m t (E, k) + δll δmm δt t Plt (E)1 = 0.
KOSTERLITZ–THOULESS TRANSITION, called
The so-called structure constants B depend only Berezinski–Kosterlitz–Thouless transition in the
on the mutual arrangement of the atoms, and not Russian literature
on the lattice parameter, or on the muffin tin poten- A phase transition occurring in planar (two-
tial. The potential parameters P , on the contrary, dimensional) degenerate systems with a two-com-
do not depend on the structure. The order of the ponent order parameter Ψ eiφ(x) such as two-
equation is determined by the number of spheri- dimensional magnets, superconductors and
cal harmonics Ylm in the wave function expansion superfluid helium. Upon tracing around a closed
(usually 9), and the number of atoms labeled by contour, the angular phase ϕ(x) can change only
the index t in the cell. by an integer multiple of 2π , i.e.
ϕ = 2πQ.
An advantage of the KKR method, in compari- The quantity Q determines a resultant topologi-
son with the augmented plane waves method, is cal charge of the region confined by this contour,
the small dimensionality of the matrix, and the and equals the sum of topological charges of indi-
possibility of calculating the structure constants vidual excitations (vortices) situated in this region
for a given structure once and for all; a deficiency (see Topological inhomogeneity). The contribution
is the extremely inconvenient representation of the of each of the vortices to the system energy grows
wave function in the regions between muffin tin logarithmically with the increase of size R of the
spheres. region: δE = πJ ln(R/a) where a is the inter-
atomic distance, and J is the interaction energy
KORTEWEG–DE VRIES EQUATION of nearest neighbor atoms. For several vortices
See Boussinesq equation and Soliton. with zero resultant topological charge, a simi-
lar contribution vanishes; therefore, the pairing of
KOSSEL LINE (W. Kossel, I935) two vortices of opposite topological charge (i.e.
Diffraction pattern by X-rays generated within opposite orientation) in a neutral vortex mole-
a single crystal. The interference lines (dark, light, cule produces a gain in energy (V.L. Berezinski,
mixed light-dark) are captured on a photoplate af- 1971). As the temperature increases, the number
ter irradiation by a beam of high-energy electrons. of “vortex molecules” and the distance between
The arrangement of Kossel lines on a kosselogram vortices in a newly formed molecule increase as
KRAMERS–KRONIG RELATIONS 685
The wave vectors k in crystals of finite dimen- KUBO–TOMITA METHOD (R. Kubo, K. Tomita,
sion take on discrete values. Using the Born–von 1954)
Kármán boundary conditions, we obtain A general formalism used to describe the linear
response of physical variables to a weak exter-
nal perturbation, AF (t). The change in a physical
k = k1 b 1 + k2 b 2 + k3 b 3
variable B(t) produced by an external force F (t)
m1 m m is expressed as
= 2π b1 + 2 b2 + 3 b3 ,
L1 L2 L3 t
B(t) = ϕBA (t − t )F (t ) dt .
where b1 , b2 , b3 are the basis vectors of the recip-
∞
rocal lattice; L1 , L2 , L3 are the number of direct
lattice unit cells in three coordinate directions; The function
m1 , m2 , m3 are integers taking on values, e.g., 1
ϕBA (t) = Tr ρ A, B(t)
0 m1 < L1 , 0 m2 < L2 , 0 m3 < L3 . Thus, i
the total number of discrete values of k is equal is called the system response, and ρ is the density
to the number of crystal unit cells N = L1 L2 L3 . matrix. The system susceptibility or admittance
As the quantity N is very large, the distribution of is a kinetic coef cient calculated as the limiting
wave vectors can be regarded as quasi-continuous, Fourier transform of the response function:
and summations over k can be replaced by inte- ∞
grations carried out inside a Brillouin zone χBA (ω) = lim ϕBA (t) exp[−iωt − εt] dt.
ε→+∞
0
V
··· = . . . dk As long as the reaction of the system to external
(2π)3
k perturbations remains linear, then calculating the
susceptibility is reduced to calculating the time
V
= dk1 dk2 dk3 . . . , correlation functions for the uctuations A − A,
(2π)3
B − B in the equilibrium state. In particular,
this formalism provided the basis for developing
where V is the crystal volume. By correlating ele- the quantum theory of magnetic resonance, which
mentary or collective excitations of an ideal crystal does not make use of a kinetic equation for the
with particular the wave functions ψk (r) it is pos- density matrix.
sible to prove the Bloch theorem, which states that
KULIK FACTOR (I.O. Kulik)
ψk (r) = eikr uk (r), A kinematic factor entering the microcontact
function of the electron–phonon interaction that
determines its dependence on the shape of the mi-
where uk (r + l) = uk (r). Therefore, both the crocontact, on the orientation of crystallographic
wave function ψk (r) and any measured physical axes of the metal in the near-contact region with
attribute of a crystal in k-space are periodic func- respect to the contact axis, and also on the average
tions of k with the period of the reciprocal lattice. path length of elastic scattering of electrons. For a
For particular conditions (e.g., for small perturba- pure contact shaped like a circular hole in an infi-
tions, for long wavelengths, and in certain other nitely thin screen that is opaque for electrons, the
cases) the density of states, the energy spectrum, Kulik factor has the form
and other factors are very little altered for imper- 4|vz vz |
fect crystals, fluids and glassy-type materials. K(v, v ) = θ(−vz vz ),
|vvz − v vz |
where v and v are the velocities of the electron
KUBO THEORY before and after scattering, the z axis is par-
See Linear response. allel to the contact axis, and θ(x) is the theta
688 KURCHATOVIUM
689
690 LAMBERT LAW
Fig. 1. Electron velocity distribution function (a) in an equilibrium Maxwell plasma, and (b) in a plasma pierced by an
electron beam. The region v1 < v < v0 corresponds to the kinetic beam instability.
at v = ω/k and is positive for ∂f/∂v < 0. The Im ε(q, ω) is determined by the condition (J. Lind-
latter condition is satisfied for every v in an equi- hard, 1954; F. Stern, 1967):
librium plasma. However, we can have ∂f/∂v > 0
q q
in a certain velocity range in a highly nonequi- max 0, qvF − 1 < ω < qvF +1 ,
librium plasma (e.g., when an electric current 2kF 2kF
or charged particle beam propagates through the where vF is the Fermi velocity, kF is the electron
electron-ion plasma (A. Akhiezer, Ya. Fainberg, Fermi momentum, and ω and q are the frequency
D. Bohm, E. Gross, 1949; see Fig. 1(b)). In and momentum, respectively (see Fig. 2(a) with
that case γL < 0, and an exponential build-up of the damping region hatched, where ωp is the
plasma waves takes place, resulting in the so- plasma frequency).
called kinetic beam (or current) instability. The In a one-dimensional (1D) degenerate electron
detection of this instability in a solid-state plasma gas found in thread-like crystals, the range of Lan-
is hampered by strong relaxation (scattering) of dau quantum damping is defined by the condition
free charge carriers (conduction electrons, holes) (P. Williams, A. Bloch, 1974):
or vibrations (phonons) and point defects (impu-
q q
rities, vacancies) of a crystal lattice. However, q vF − 1 < ω < q vF +1 ,
various dissipation instabilities become possible 2kF 2kF
in a nonequilibrium solid-state plasma. where q is the longitudinal component of mo-
A quantum analog of the Landau damping mentum in the direction of the chains. Due to
takes place in the degenerate electron (hole) plas- shrinking of the phase volume available for the
ma of metals (semimetals). It results from the de- process of plasmon decay, there exists a “trans-
cay of collective (plasma) excitation quanta of the parency window” (see Fig. 2(b)) with Im ε(q, ω) =
boson type, i.e. plasmons. These decay into one- 0 in the range 0 < ω < q vF (1 − q /2kF ) for
particle electron excitations of the fermion type – q < 2kF .
the electron–hole pairs that follow the conserva- Nonlinear Landau damping (R.K. Mazitov,
tion laws of energy and momentum. For the cases 1965; O’Neil, 1965) takes place when the damp-
of an isotropic three-dimensional (3D) degenerate ing time, γ −1 , is longer than the oscillation period
electron plasma, and of a two-dimensional electron of the resonance particles in the traveling wave
gas (2D) generated in inversion layers and strati- field, −ϕ0 cos(kx − ωt). Phase trajectories of such
fied crystals (see Quasi-two-dimensional crystals), particles are shown in Fig. 3. The region inside
the range of Landau quantum damping with a non- the separatrix corresponds to trapped particles, and
vanishing imaginary part of the dielectric constant that outside it to transient ones. The characteristic
692 LANDAU DAMPING
Fig. 2. Plasmon spectrum, ω(k), and Landau quantum damping region (hatched) of a degenerate electron plasma (a) in
an isotropic 3D-metal, and (b) in a quasi-one-dimensional (chain-like) metal.
where γ = gμB /, g is the gyromagnetic ra- The dissipation term R is often written in the
tio (g = 2 for spin magnetism), μB is the Bohr Landau–Lifshits or the Gilbert (T.I. Gilbert, 1955)
magneton, H eff is the effective field equal to form:
– δW/δM, and W (M) is the magnetic energy ex-
pressed as a functional of the magnetization. For RLL = λγ M0 Heff − m(mHeff ) ,
an n-sublattice magnetic material the equation for α ∂M
the magnetization of the ath sublattice, M a , is RG = M, ,
M0 ∂t
obtained by the substitution M → M a , H eff =
(a) where λ and α are dimensionless relaxation con-
H eff = −δW (M 1 , . . . , M n )/δM a . The term R
stants. For R = R LL and R G , the Landau–Lifshits
provides for relaxation of the magnetization field
equations become equivalent under the substi-
energy. Within the nondissipative approximation
tution λ → α and g → g = g(1 + α 2 ). The
(R = 0), the Landau–Lifshits equation has the
Landau–Lifshits equations containing R LL or R G
following integrals of motion: energy W and mo-
exactly conserve the length of the magnetiza-
mentum P ,
tion vector |M| = M 0 = const, and provide a
dr(My ∇Mx − Mx ∇My ) way to write these equations in angular coordi-
P= ,
2μB M0 + Mz nates. In that case, the right-hand sides of the
equations for ∂θ/∂t and ∂ϕ/∂t involve the terms
of the magnetization field. The length of the vec- α sin2 θ(∂ϕ/∂t) and ∂θ/∂t , respectively. How-
tor M is conserved, and has the Cartesian compo- ever, the variable R in those forms describes
nents the phenomenon of relaxation only qualitatively,
coming into conflict with both microscopic calcu-
Mx = M0 sin θ cos ϕ,
lations based on the method of Green’s functions,
My = M0 sin θ sin ϕ, and experimental data. Since this variable R fails
to account for the dynamic symmetries of interac-
Mz = M0 cos θ.
tions of the Heisenberg magnet, V.G. Bar’yakhtar
The Landau–Lifshits equations have the following (1984, 1986) proposed a relaxation term in the
forms in terms of the angular variables for M: form R B = R e + R r , where R e and R r origi-
nate from exchange and relativistic interactions,
∂θ γ ∂W respectively.
sin θ =− ,
∂t M0 ∂ϕ The term R e = −λe a 2 M 0 ∇ 2 H eff describes
∂ϕ γ ∂W relaxation when the resultant magnetic moment
sin θ = , of a crystal, μ = M dr is conserved, where λe
∂t M0 ∂θ
is the exchange relaxation constant, and a is the
where W = W (θ, ϕ) is the functional of mag- lattice constant. The structure of R r is essentially
netic energy expressed in spherical coordinates θ defined by the magnetic symmetry.
and ϕ. In that form, the Landau–Lifshits Hamil- The Landau–Lifshits equation is widely used
tonian equations are: ∂q/∂t = δH /δp, ∂p/∂t = to describe both linear and nonlinear dynamic
−δH /δq for q = cos θ , p = ϕ and H = W . In phenomena when a microscopic quantum analy-
a one-dimensional case, when M = M(x, t), the sis is either difficult or impossible, such as for
Landau–Lifshits equation for a ferromagnet, with spin waves in inhomogeneous or bounded magnet-
its energy quadratic in powers of the magneti- ics (see Walker modes, Magnetostatic oscillations),
zation, can be integrated explicitly if dissipation and nonlinear magnetization waves (see Magnetic
is ignored. It is equivalent to the method of the domain wall dynamics, Magnetic soliton).
inverse problem of scattering theory, has an in-
finite series of conservation laws, and admits the LANDAU SUPERFLUIDITY CRITERION
existence of many-soliton solutions satisfying the The most general condition imposed upon the
asymptotic principle of superposition (see Soli- dispersion law for single-particle and collective
ton). elementary excitations in a superfluid quantum
LANDAU THEORY OF SECOND-ORDER PHASE TRANSITIONS 695
particle system on the basis of the conservation Fig. 2. Spectrum of quasi-particles in superconductors
laws of energy, E(p), and momentum (quasi- (2Δ is the energy gap).
momentum), p. The Landau superfluidity criterion
has the form vc = min{E(p)/p} > 0 and consti- LANDAU THEORY OF SECOND-ORDER
tutes a sufficient condition for the E(p) spectrum PHASE TRANSITIONS (L.D. Landau, 1937)
to satisfy. Provided the flow rates of the quantum A phenomenological theory of phase transi-
system as a whole fulfill the condition v < vc , tions based on the hypothesis that the nonequilib-
then all its intrinsic excitations resulting from in- rium thermodynamic potential Φ may be expanded
teractions with other external systems (such as in a power series of a variable η called the order
capillary walls, the crystal lattice) remain forbid- parameter :
den. In other words, no exchange of energy or
momentum as well as no transitions of the sys- Φ = Φ0 + Aη2 + Bη4 . (1)
tem into excited states are possible. The Landau The value η = 0 is attained in the phase of the
superfluidity criterion characterizes the degree of highest symmetry, while for any η = 0 the symme-
stability of the ground state of a superfluid compo- try of the state of matter is lower than that of the
nent in a quantum system (the Bose condensate) symmetric phase, and corresponds to an unsym-
with respect to external excitations. metric phase. The expansion coefficients in Eq. (1)
In a neutral Bose-liquid (4 He), the superfluid-
are temperature dependent with A = a(T − Tc ),
ity criterion is also met because, on the one hand,
where Tc is the temperature of a second-order
the acoustic dispersion law is valid for both single-
phase transition. The equilibrium value of η is
particle oscillations, and many-particle excitations
obtained by minimizing Eq. (1), from which it
hybridized with the former at p → 0, while on
follows that the symmetric phase (η = 0) is sta-
the other hand, a minimum in the E(p) plot is
ble at T > Tc , while for the unsymmetrical one
present at p = 0 (see Fig. 1(a)). In a system of
with η = 0 we have η2 = −A/(2B) ∝ (Tc − T ) at
charged bosons (e.g., bipolarons), The Landau cri-
T < Tc .
terion is satisfied for a finite gap, equal to the
In a general case, the order parameter is a mul-
plasma frequency ωp , in the spectrum of elemen-
ticomponent variable (see Structural phase tran-
tary excitations at p = 0 (see Fig. 1(b)). Such
sition). When formulating the Landau theory, the
a gap develops due to the Coulomb interaction.
function of the change of crystal density δρ (or
For a Fermi system (e.g., conduction electrons),
the Landau superfluidity criterion is met when some other physical variable) in an unsymmetric
an energy gap of 2Δ appears in the spectrum of phase is assumed to be expandable in the func-
(m)
single-particle excitations (electron–hole pairs) at tions ϕα that transform according to the αth
the Fermi level (see Fig. 2). The critical velocity row of the mth irreducible representation of the
then equals vc = Δ/pF , where pF is the Fermi symmetric phase spatial symmetry group: δρ =
(m) (m) (m)
momentum, defining the maximum critical current α,m ηα ϕα . All coefficients ηα are zero for
in superconductors. the symmetric phase, while some of them may be
696 LANDÉ g-FACTOR
nonzero for an unsymmetric phase. Therefore, the motion of magnetic ions such as transition ions
(m) and cause the g-factor to deviate from the Landé
ηα coefficients play the roles of order parame-
ters. Since the potential should be invariant with value, sometimes to a considerable extent.
respect to the symmetric phase symmetry group,
the expansion in the powers of the order parame- LANGEVIN–DEBYE FORMULA (P. Langevin,
ter assumes the form P. Debye, 1912)
(m) 2 Formula describing the nonlinear orientation
Φ = Φ0 + Am ηα + Φ3 + Φ4 + · · · .
polarization of noninteracting electric dipoles μ0
m α under the effect of an applied electric field E. For
(2)
low fields a linear dependence on the field is ex-
Expansion (1) is a particular case of Eq. (2) for
pected for the average moment μ = α D E, where
a second-order transition under a one-dimensional
α D is the dipole polarizability given by the Debye
irreducible representation. Two limitations exist
formula, α D = μ20 /(3k B T ). In relatively strong
for a second-order transition under a given irre-
electric fields the polarization saturates, becoming
ducible presentation. The continuity of the tran-
nonlinear and described by the Langevin–Debye
sition requires that Eq. (2) contain no invariants
(m) function:
cubic in ηα that would correspond to one and
μ μ E
the same irreducible representation. The stability =L 0 ,
of the emerging unsymmetric phase with respect μ0 kBT
to the presence of nonuniform strains requires that where L(x) is the Langevin function. Meanwhile,
no invariants linear in the first spatial derivatives the dielectric constant ε of a polar insulator in a
(m) (m) (m) (m)
of the form ηα (∂ηβ /∂xi ) − ηβ (∂ηα /∂xi ) strong field becomes weaker. For small x (in weak
be present. The Landau theory provides a qualita- fields), we have L(x) ≈ x/3, and the Debye for-
tively correct description of a second-order tran- mula follows. The Langevin–Debye formula is a
sition. However, in the close vicinity of Tc (the generalization of the Clausius–Mossotti equation
critical region), nonuniform fluctuations of the or- for polar insulators, and it accounts for the tem-
der parameter (critical uctuations) start to play a perature dependence of ε.
significant role, and may qualitatively alter the re-
sults (see Ginzburg number). LANGEVIN FUNCTION (P. Langevin, 1906)
A function defining the value of the magneti-
LANDÉ g -FACTOR, magnetic splitting factor zation of paramagnets in the classical limit, i.e.
(A. Landé, 1921) of paramagnets with large values of magnetic mo-
The dimensionless gyromagnetic ratio factor ments in their atoms (molecules)
g L which is a measure of the ratio of the magnetic
1 j μB gH
moment to the angular momentum of an atom or L(x) = coth x − , x= ,
subatomic particle in the presence of spin–orbit x kB T
coupling. It is given by the Landé formula where μB is the Bohr magneton, g is the gyro-
magnetic ratio, and H is the magnetic field. The
3 S(S + 1) − L(L + 1)
gL = + Langevin function may be obtained as the classical
2 2J (J + 1) limit of the Brillouin function Bj (x) for j → ∞.
which provides the Zeeman splitting of atomic en-
ergy levels in a magnetic field (see Zeeman effect). LANGMUIR–BLODGETT FILMS (I. Langmuir,
The values of the Landé g-factor depend on the K. Blodgett, 1937)
electron configuration, and are valid for Russell– Films of organic materials of a specific type
Saunders coupling. For a purely orbital moment formed by successive deposition of monolayers
(S = 0, J = L), the Landé factor equals unity, and on a substrate. Usually, the molecules of such
for a purely spin moment (L = 0, J = S), it is 2. materials are elongated, and their opposing ends
A free electron has the value g = 2.0023. Crys- have different chemical compositions, and differ
talline electric fields in solids act on the orbital in the nature of their interaction with the substrate.
LANTHANUM 697
For example, a number of fatty acids and their the lanthanides are called f -elements). In chem-
derivatives have molecules in the shape of long ical compounds, the atoms of lanthanides usually
hydrocarbon chains built of (CH2 )n groups that loose three electrons for the oxidation state +3,
exhibit hydrophobic properties (see Lyophilic and and some also form oxidation states +2 and +4.
lyophobic behavior), one end of such a molecule With increasing atomic number there is a gradual
having a hydrophilic group of atoms (amphiphilic reduction of the radius of neutral atoms and ions
molecules). When placed on a water surface, such from 0.102 nm (Ce3+ ) to 0.080 nm (Lu3+ ), the
a substance spreads into a monomolecular layer. so-called lanthanide contraction.
This layer may be in various states of matter In free form the lanthanides are silvery metals,
depending on the ambient conditions (see Mono- T melting decreases from 1073 (Ce) to 936 K (Lu).
layer). Molecules in a solid monolayer sit close The metals possess comparably low electrical con-
to each other in a parallel arrangement so that ductivity, and most of them are paramagnetic.
their hydrophilic ends face the water surface and Some lanthanides (Gd, Dy, Er) at low tempera-
their hydrophobic ends look outside. A mono- tures possess ferromagnetic properties. Upon fu-
layer structured this way may be transferred to sion, lanthanides mix with each other yielding
a solid substrate. By repeating the operation one solid solutions, the so-called mischmetal. Many
may form Langmuir–Blodgett films of prescribed lanthanide compounds are used in the manufac-
thickness that consist of monolayers alternating ture of special optical glasses, luminophors, crys-
in a prescribed order. The structure, and the op- tallophosphors, laser materials; some lanthanides
tical, magnetic, electric and other properties of are components in high-temperature superconduc-
Langmuir–Blodgett films have been widely stud- tors.
ied, and their applications to microelectronics are
well developed, especially those associated with LANTHANUM, La
the task of producing molecular level microelec-
A chemical element of Group III of the peri-
tronics.
odic system with atomic number 57 and atomic
LANTHANIDES (fr. lanthanum and Gr. ειδoς ,
mass 138.91, belongs to rare-earth elements.
type), general symbol Ln
A natural mixture consists of a stable isotope
139 La (99.911%) and a radioactive isotope 138 La
Family of 14 chemical elements with atomic
numbers from 58 to 71, coming after La in the (0.089%), which decays by K-capture with half-
periodic system; belonging, as does La, to the life T1/2 = 1.1·1011 years. Isotope 139 La is
rare-earth elements. Lanthanides include the ele- formed by fission of uranium (6.3% of mass of
ments: cerium Ce, praseodymium Pr, neodymium all fission fragments) and is “the reactor poison”.
Nd, artificially produced radioactive promethium Configuration of outer electronic shells is 5d 1 6s 2 .
Pm, samarium Sm, europium Eu, gadolinium Gd, Successive ionization energies are 5.577, 11.06,
terbium Tb, dysprosium Dy, holmium Ho, erbium 19.176 eV. Atomic radius is 0.187 nm; radius of
Er, thulium Tm, ytterbium Yb, lutetium Lu. These La3+ ion is 0.114 nm. Oxidation state is +3, and
elements occur rarely and widely scattered in na- electronegativity is ≈1.1.
ture. Based on their properties, lanthanides are In a free form, lanthanum is a silvery-gray
subdivided into the light cerium (from Ce to Gd) metal. It has three allotropic modifications:
and the heavy (Tb to Lu) subgroups. Electronega- α-La, β-La, γ -La. The α-La form with a dou-
tivity ranges from 1.2 to 1.3. ble hexagonal close-packed lattice and parame-
Physical and chemical properties of all lan- ters a = 0.3770 nm, c = 1.2159 nm is stable at
thanides are similar to each other as a result of the temperatures below 533 K. In the range 533 to
similarities of their electronic shell structures. The 1150 K β-La forms a face-centered cubic lattice
configurations of the outer shells of the majority with parameter a = 0.5304 nm. At temperatures
of lanthanides are 5s 2 p6 6s 2 (except Ce, Gd and above 1150 K β-La transforms to body cen-
Lu: each of which has one 5d electron). Electrons tered γ -La with a = 0.426 nm. Density of α-La
added to atoms of the lanthanides with increas- is 6.162, of β-La 6.190, of γ -La 5.97 g/cm3 ;
ing atomic number enter the 4f -sublevel, hence T melting = 1193 K; T boiling = 3743 K. Heat
698 LARMOR PRECESSION
of melting is 6.19 kJ/mole; heat of evapora- uniform gravitational field. The Larmor theorem
tion is 412.15 kJ/mole; and specific heat cp = holds if ωL remains small compared with the nor-
27.8 J·mole−1 ·K−1 . Coefficient of thermal ex- mal frequencies of orbiting particles in the absence
pansion is 4.9·10−6 K−1 (at 298 K). Thermal of a magnetic field. We have ωL ∝ 3·1010 s−1 for
conductivity is 13.8 W·m−1 ·K−1 (at 300 K). electrons in a very strong magnetic field (1 T),
Debye temperature of β-La is 146 K. Electric re- while the electron orbiting frequency in an atom
sistivity of α-La is 0.568 μ·m (at 298 K), of is of the order of 1015 s−1 . Thus, the Larmor
β-La 0.98 μ·m (at 830 K), of γ -La 1.26 μ·m theorem has an extremely wide application range.
(at 1160 K). Temperature coefficient of electric As the result of this precession of charged parti-
resistance of α-La is 2.18·10−3 (at 273 K). Su- cles, a macroscopic system acquires a magnetic
perconducting transition temperature of α-La is moment. Therefore, the Larmor theorem helps ex-
4.90 K, of β-La 5.85 K. Work function is 3.3 eV. plain the phenomena of diamagnetism, the normal
Lanthanum is a paramagnetic substance. Ther- Zeeman effect, and the Faraday effect (magnetic
mal neutron cross-section of 139 La is 9 barn. rotation of light polarization plane).
At room temperature, normal elastic modulus is
≈6·104 MPa; Poisson ratio is 0.288. Yield limit of LASER (acronym for Light Ampli cation by
a cast sample at the temperature 293 K is 125 MPa, Stimulated Emission of Radiation)
ultimate strength is 130 MPa, relative elongation A source of coherent radiation generated dur-
is 8%. Vickers hardness (at 293 K): of a cast sam- ing stimulated emission or light scattering by an
ple 490, of an annealed sample 363, of a forged active medium located in a resonant cavity. Every
sample 1180 to 1740 MPa. At room temperature, laser contains three principal elements: a laser
sufficiently pure lanthanum is forgeable (manu- active medium, a system for pumping the active
facturing sheets is possible) and pressable, but it medium (pumping system), and an optical reso-
does not possess sufficient toughness. Ion-plasma nant cavity (see Resonators). The active medium
frequency is 8.70 THz. Linear low-temperature can be either a gas, a liquid or a solid. To pump
electronic specific heat (Sommerfeld coefficient) this medium light from nonlaser sources can be
is 10 mJ·mole−1 ·K−2 . Lanthanum is used in op- used, as well as an electric discharge, thermal
tical glasses, as a getter in alloys with Ni, as a heating, etc. The simplest optical resonant cav-
component of mischmetal. ity consists of two flat mirrors placed parallel to
each other. To extract the radiation, one of the
LARMOR PRECESSION (I. Larmor, 1895) mirrors is made semitransparent or has a special
Precessional rotation of a stable system of opening. Typical lasers generate radiation within a
identical (usually charged) particles with magnetic range from the UV to submillimeter frequencies,
moments located in a uniform constant magnetic and operate either by continuously emitting power,
e ld, the field direction serving as the rotation axis. or in a pulse mode. Some lasers operate at a fixed
The Larmor theorem states that when a uniform frequency, and others are tunable in their radia-
magnetic field B is impressed upon a system of tion frequency (dye lasers, etc.). Some continuous
such particles (e.g., electrons or atomic nuclei), power lasers can generate in excess of 1 MW, and
the equations of motion retain their form when some pulse lasers can produce output powers ex-
transformed to a system of coordinates in uniform ceeding 10 TW, providing energy flux densities at
rotation about the direction of that field at the Lar- targets that exceed 1015 W/cm2 , with their pulse
mor frequency ωL = eB/(2m) (SI units). Here e energy reaching 10 kJ.
and m are the charge and mass of an electron, Lasers are used in research (spectroscopy,
respectively. The magnetic field thus produces a holography, thermonuclear fusion, etc.), informa-
uniform precession of the orbit of each charged tion technology (communications, data storage,
particle around the field direction. The Larmor etc.), and technological applications (laser chem-
precession is caused by the magnetic component istry, isotope separation, semiconductor produc-
of the Lorentz force upon charged particles, and is tion, cutting, welding, smelting, etc.). Three direc-
similar to the precession of a gyroscope axis in a tions of laser applications in the physics of solids
LASERS WITH DISTRIBUTED FEEDBACK 699
are: (a) studying interactions of laser radiation At T < Tmelt.a , the process of laser epitaxy takes
with solids; (b) using the effects of such interac- place in the solid phase, and when T = Tmelt.a ,
tions to study the composition and properties of it proceeds through the liquid phase (see Liquid-
solids (spectral analysis, mass spectrometry); and phase epitaxy). Solid-phase laser epitaxy is trig-
(c) using laser radiation to process various ma- gered when amorphous layers are pulse-heated
terials and alter or improve their properties (sur- within an interval of several microseconds, and the
face cleansing, surface hardening, surface doping liquid phase type within nanoseconds. Heating the
and cladding, deposition of thin lms , etc.). Laser semiconductor by 6–8 ms long ruby laser pulses at
types include solid-state lasers, gas lasers, ex- power densities of 40–60 J/cm2 brings the layer
cimer, plasma, and free-electron ones. up to the temperature 900–1100 K, and restores
See also Electron beam pumped laser, Hetero- the monocrystalline structure. Heating with 50 ns
junction laser, Holographic laser, Injection laser, long pulses at power densities over 3 J/cm2 melts
Lasers with distributed feedback, Semiconductor the amorphous layer all the way down to the sub-
lasers, Solid-state laser, Streamer laser, Tunable strate.
lasers, X-ray laser. The mechanism of laser epitaxy is described
satisfactorily by the thermal model of pulse heat-
LASER ACTIVE MEDIUM ing which assumes that the mean free path of
A medium capable of amplifying electromag- the charge carriers produced in the semiconduc-
netic radiation by the use of a laser energy transi- tor by the light pulse is shorter than the thermal
tion. Active laser media are known in every state diffusion length, and the carrier lifetime is shorter
of matter, including plasma. Among the most im- than the pulse duration. These conditions provide
portant such media are various solids including for the transformation into heat of energy of the
crystals and glasses (see Active solid-state laser electron–hole plasma produced by the radiation.
materials) activated with ions of the transition Plasma phenomena appear during picosecond du-
and rare-earth elements (ruby (Cr+3 ), yttrium– ration pulses. High crystallization rates attained
aluminum garnet, Nd+3 , neodymium glass, etc.); during laser epitaxy with melting impart semi-
alkali-halide crystals with color centers, e.g., conductor layers with concentrations of impurities
LiF: F+ 2 ; artificially layered structures (hetero- 10–100 times higher than those acquired by crys-
structures) for charge recombination, inhomoge- tallization under equilibrium conditions.
neously doped semiconductor crystals, and het-
erogeneous structures. They are used in lasers LASER MATERIAL, ACTIVE
operating in every mode of generation from con- See Active solid-state laser materials.
tinuous emission to supershort pulses.
LASER RESONATOR, DISPERSION
LASER ELEMENT, ACTIVE See Dispersion laser resonator.
See Active laser element.
LASERS WITH DISTRIBUTED FEEDBACK
LASER EPITAXY A type of laser with enhanced spectral selec-
A technique used to produce monocrystalline tivity of the resonator cavity attained either by
semiconductor layers on oriented substrates using periodic modulation of the laser active medium,
laser pulse or scanning heating. A laser beam inci- or from the optical waveguide along the radiation
dent on a target evaporates atoms which fall onto a direction of propagation. Such a modulation by
substrate to form a thin lm . Laser epitaxy was first diffraction provides feedback resulting in a distrib-
observed during pulsed annealing of semiconduc- uted radiation output from the cavity. The period-
tor surface layers (Si, Ge, GaAs) transformed into icity of the medium may be caused by changing its
the amorphous state by ion implantation. Without volume or by modulating the transverse surfaces
affecting the lower layers of a crystalline mater- of the waveguides, which are then corrugated by
ial, pulse heating can bring the temperature up to etching. First used in dye lasers (see Solid-state
Tmelt.a at which the amorphous layer starts to melt. quantum electronics), the principle of distributed
700 LASER TECHNOLOGY
feedback has found its widest application in injec- possible to weld metals with high thermal con-
tion lasers, in those with optical pumping, and in ductivities (aluminum, silver) as well as materials
those excited by an electron beam. Periodic modu- of sharply differing properties, such as nonmetals
lation of the waveguide thickness is effected either which resist welding by other techniques. At pulse
inside the active zone with gain g (distributed energies between 0.1 and 30 J, typical for solid-
feedback lasers) or outside it (distributed Bragg state lasers, the depth of the material melted is
reflection lasers). Distributed reflection may be 0.05–1 mm. The efficiency of point welding is sev-
combined with a cavity having concentrated end- eral dozen operations per minute, and a continuous
face mirrors. Distributed output makes it possible seam output is 1.5 m/min (for a depth of 0.5 mm).
to obtain a collimated beam with low divergence Pulse laser welding applications include the man-
determined by the length of the periodic output ufacture of vacuum electronic and semiconductor
structure. The direction of the scattered wave de- devices, sealing cases for integrated circuits and
pends on the order m of the dispersing grating: quartz resonators, producing high precision me-
sin θi = 2i/m − 1, where θi is the angle of the chanical devices, etc. Using 400 W/cm2 garnet
scattered wave front relative to the direction of laser output power, continuous seam welding of
its propagation; and i is an integer. The separa- 3 m/min is achievable at melt depths of 0.3–
tion of longitudinal normal frequencies (modes) of 0.4 mm. Carbon dioxide (CO2 ) laser output beams
the resonant cavity of a distributed feedback laser with power levels of 10–20 kW can weld steel up
depends on the length of the cavity L, i.e. the dis- to 50 mm thick.
tance between the boundaries of the homogeneous At power densities of 107 W/cm2 substances
and the periodic media. begin to evaporate, making possible high-precision
laser processing. Laser drilling and cutting have
LASER TECHNOLOGY undergone rapid development. One can produce
Processing and welding technology applica- holes from 0.005 mm in diameter in any material,
tions of laser radiation in various materials. including those extremely resistive to processing
The high directionality of laser radiation al- by other techniques (ceramics, diamond, corun-
lows focusing a ray on a spot with a diameter dum). Using laser drilling to produce diamond
comparable to the wavelength of the radiation. The dies, ruby watch stones, and components for elec-
energy density attainable is sufficient to heat, melt tronic devices increased productivity by more than
and evaporate any natural or artificial material, tenfold. Laser cutting is widely used with ceramic
so welding, boring, cutting and thermal process- components, in particular for inscribing substrates
ing are practable. Solid-state lasers and gas lasers for semiconductor and hybrid integrated circuits.
operating in both the continuous wave (CW) and Gas-laser cutters are actively employed in the ma-
pulse modes are used. chine building industry; the substance melted by
The advantages of laser technology are man- laser radiation being removed by a gas jet (usu-
ifold. The effects inflicted on materials are local ally air or oxygen). Garnet lasers with 100–500 W
in nature; the heating remains very localized, and output power, and CO2 lasers with 300–2000 W
there is no physical contact during processing. output, function as power tools in laser cutters.
It is also possible to process materials in any The material cut in this manner may be up to
transparent medium, or to work through trans- 10 mm thick, with cutting rates 0.5–10 m/min, de-
parent shields. These applications are widespread pending on the thickness.
in the electronic and radio industries, and lasers The manufacture of integrated circuitry and
with output powers of several dozen kilowatts are other electronic devices involves the formation of
routinely used in the machine building industry thin film patterns, and the control of various film
(automobile manufacture and tool making). parameters through selective ablation of film ma-
At power densities of 105 –106 W/cm2 met- terial.
als melt, which is the basis for laser welding, Lasers are used for the laser heat treatment
and provides high-quality joints between stain- of surfaces, predominantly for thermal strengthen-
less steel, nickel, molybdenum, Kovar, etc. It is ing of steels and cast irons by quenching, as well
LATTICE 701
as for annealing and thermal-chemical process- the unique combination of two physical features:
ing. Radiation output power densities for heat high monopolar dark conductivity due to intersite
treatments reach 103 –104 W/cm2 . During laser cations Ag+ (see Frenkel defect) and significantly
hardening it is possible to start cooling at ex- smaller mobility of photo-generated holes relative
tremely high temperatures, including the melting to photoelectrons from the Ag+ ions. In such a
temperature at the surface. Note that a high tem- compound a sequence of elementary reactions is
perature gradient in a thin surface layer provides initiated, such as X− + hν → X0 + e− ; e− + T →
extremely fast cooling rates (up to 106 deg/s) that T e− (T is a deep trap); T− + Ag+ i → TAg ;
0
result in a fine crystalline structure with a heavy − 0 − −
X + hν → X + e ; e + TAg → TAg , 0 −
concentration of martensite and a high disloca- TAg− + Ag+ 0
i → TAg2 , etc. This sequence leads
tion density. Cutting tools thus processed increase via quantum absorption to the formation of Ag0n
their strength twofold, and stamping equipment in- clusters of sizes sufficient for catalyzing the devel-
creases in strength by a factor of three or more. opment through the capture of electrons from the
A CO2 laser with 1–10 kW output provides high developer without further participation of light.
rate continuous hardening of large-area surfaces to See also Nonsilver photography.
a depth of 0.7–1.0 mm. It also finds use in ma-
chine construction to attain higher wear resistance LATERAL INTERACTION of adsorbed particles
of rubbing surfaces of machine parts. Interaction of adsorbed atoms and molecules
Applications of laser heating in technology on a solid surface, either immediate (direct for
include doping, producing amorphous metal lay- high adsorbate density) or indirect (proxy, via
ers, local activation of etching processes, thermo- adsorbent or substrate) (see Adsorption). Several
chemical modification of surface properties, evap- types of lateral interaction are identified: van der
orating substances in vacuo to produce thin lms , Waals (dispersion), deformation (distortions in
communication and manufacturing plates in print- crystal structure of solid surface via chemisorption
ing, eye surgery, cosmetic surgery, other medical forces), electrostatic (charged particles on insula-
applications, etc. tor surfaces), direct exchange (overlap of particle
wave functions) and indirect exchange through
LATENT PHOTOGRAPHIC IMAGE the electron gas of a metal or a semiconductor
The immediate result of light exposure on substrate (type of oscillating Ruderman–Kittel–
photographic material; modification of the pho- Kasuya–Yosida interaction, RKKY interaction),
tographic material with no visible change. For dipole–dipole interaction (image forces on metal
visualization the latent photographic image must surface), and screened Coulomb interaction (see
be converted to a visible one with the help of Friedel oscillations).
some process for which the latent image plays
the role of the initiator, catalytic agent, etc. This LATTICE
process cannot occur in the absence of this hidden A totality of points (sites) arranged regularly in
image. In the AgX (X = F, C1, Br, I) emulsion space with integer coordinates relative to a fixed
layers of the silver halide process the latent im- coordinate system with its origin (main reference
age involves small atomic clusters generated by point) at one of the lattice sites. Each of the three
the light exposure (most sensitive emulsions have vectors forming a primitive unit cell is called a
only 3 or 4 Ag atoms per microcrystal AgX). Dur- principal vector of the lattice. A vector connecting
ing the development the complete reduction of any two sites of the lattice is called a lattice vector.
Ag in the microcrystals takes place. The physi- A parallel displacement through any lattice vector
cal mechanism of latent image formation in AgX converts the lattice into itself, i.e. it is a symmetry
microcrystals offered by R. Gurney and N. Mott transformation of this lattice. Each reference point
(1938) is still accepted with some modifications unambiguously defines the lattice, i.e. it may be re-
and additions (Gurney–Mott model). According to garded as the main reference point of the specific
this model, the high AgX light sensitivity (abil- lattice. The main reference point in a given lat-
ity to form latent photographic image) is based on tice may be selected in an infinite number of ways
702 LATTICE CONSTANT
LATTICE DESTABILIZATION
LATTICE, SURFACE
Approach to zero of the shear resistance for
See Surface lattices.
a special plane and direction (correspondingly
the combination of elastic moduli (C11 − C12 ) LATTICE VIBRATIONS, SILENT
tends to zero) during the variation of thermo- See Silent lattice vibrations.
dynamic parameters (temperature, pressure, mag-
netic or electrical fields, etc.). The lattice becomes LAUE CLASSES
mechanically unstable, and its rearrangement to Eleven classes of point crystal symmetry with
another structure takes place. Almost complete lat- a center of symmetry. Classified by crystal sys-
tice destabilization towards shear {110}11̄0 was tem they are: triclinic (1), monoclinic (2/m), or-
discovered in the compounds V3 Si and Nb3 Sn in thorhombic (mmm), tetragonal (4/m, 4/mmm),
the neighborhood of the temperatures of structural trigonal (3, 3/m), hexagonal (6/m, 6/mmm), and
phase transitions, and in the alloys InTl (28 to 39 cubic (m3, m3m). Laue classes are classes of dif-
at.% Tl) at the transition from a face-centered cu- fraction symmetry (deduced from diffraction pat-
bic to a tetragonal lattice. In other alloys (with the terns) higher than its point symmetry, as follows
body-centered cubic lattice) during a temperature from Friedel’s law. Each of the 32 point groups
change only a partial reduction of the shear mod- falls into one of the Laue classes that results when
ulus and an increase of the degree of anisotropy of a center of symmetry and certain derivative ele-
crystals take place. ments of symmetry are added. The Laue classes
and diffraction symmetry are directly manifested
LATTICE GAS in Laue patterns (see Laue method) and in electron
A mathematical model that deals with the be- diffraction patterns (see Electron diffraction analy-
havior of a system of N particles distributed sis) of single crystals.
among N0 (where N < N0 ) sites of an undistorted
lattice. Only one particle may be located at an in- LAUE CONDITIONS, Laue equations
dividual site, every particle is at some site, but not The relations between the directions along the
all sites are occupied. The state of the system is de- interference maxima for the X-ray diffraction pat-
termined by the distribution of the particles among tern of a monocrystal:
the sites. It is impossible to determine the kinetic
a(cos α − cos α0 ) = H λ,
energy in such a system; yet, for a noninteracting
lattice gas there is an ideal gas equation of state b(cos β − cos β0 ) = Kλ,
P V = RNT . If the energy of the particle inter-
c(cos γ − cos γ0 ) = Lλ,
action is known, one can simulate some structural
properties of alloys, and of substitutional and in- where a, b, c are the periods of the crystal lattice
terstitial solid solutions. of a monocrystal along the X, Y, Z axes; α0 , β0 ,
LAW OF CONSTANCY OF ANGLES 703
γ0 are the angles of the X, Y, Z axes with the The Laue method is used to determine the crystal
direction of the incident (primary) X-ray beam of symmetry (i.e. its Laue class), the level of single
wavelength λ; α, β, γ are the angles of X, Y, Z crystal perfection, its orientation, etc. The method
to the direction of the diffracted beam; H, K, L is used for the routine orientation of single crys-
are integers. tals.
LAUE DIFFRACTION, Laue geometry (M. Laue, LAVES PHASES, lambda phases (λ-phases)
1912) (after Swiss scientist F. Laves)
A scheme of X-ray diffraction of crystals (also A large group of intermetallic compounds with
electrons, etc.) with the diffracted beam exiting the formula AB2 that form between elements
on the side opposite to its entry face (see X-ray (metals) with the ratio of atomic diameters d AA
diffraction). Two cases of Laue diffraction are iden- and d BB within the range 1.1–1.6. Arguing from
tified; symmetric and asymmetric ones, when the geometric requirements, an ideal ratio is
reflecting planes (hkl) are and are not perpendic- d AA /d BB = 1.225. Lambda phases are crystals
ular to the entry face of the crystal, respectively. with close-packed lattices. They fall into one of
Several important effects have been observed dur- three structure types: λ1 has structure type MgZn2
ing Laue diffraction. One is the periodic pumping with a hexagonal lattice, λ2 has structure type
of energy of the passing wave into the diffracted
MgCu2 with a face-centered cubic lattice, and λ3
wave and back across the crystal depth with a
has structure type MgNi2 with a hexagonal lat-
period equal to the extinction length (this occurs
tice. Considering the positions of the large atoms,
for thicknesses, t , less than the absorption path
all three structures involve an alteration of dou-
length, t abs , the so-called Ewald pendulum solu-
bled layers with vertical pairs of A atoms forming
tion). Another effect is the anomalous passage of
a hexagonal lattice. The difference between these
X-rays (the Borrman effect) for thicknesses much
structures consists in the angle of the shift between
greater than t abs . Laue diffraction is used in solid
the alternating doubled layers relative each other.
state physics to determine the diffraction symme-
The phases exhibit metallic conduction and high
try of a material, and is often employed to orient
hardness, but are brittle (see Brittle failure). Al-
single crystals. Using fast electrons in the Laue
though the main factor responsible for the lambda
diffraction regime makes it possible to monitor the
phase formation is geometric, the stability of one
structural state of extremely thin layers or lm s
or another modification is attributed in part to the
(t = 1 μm), while X-rays probe less thin layers
interaction between the Fermi surface and Brillouin
(t = 0.1–1 mm). Thick samples (t 1–10 mm)
zone boundary. See also Brass.
may be studied using γ -quanta (see Gamma dif-
fractometry) and neutrons.
LAW OF CONSTANCY OF ANGLES,
LAUE METHOD (suggested by M. Laue and Steno–Romé de L’Isle law
implemented by W. Friedrich and P. Knipping, An empirical law of crystallography stating
1912) that, in one crystal or in different crystals of the
A technique of retrieving a diffraction pat- same polymorphic modification of a material, the
tern formed by spectrally continuous radiation area and shape of facets, their mutual arrangement
diffracted by a monocrystal. Laue conditions are and number can change, but the angles between
met in an experiment with polychromatic X-rays corresponding facets (i.e. those which are paral-
that vary continuously in wavelength from λmax lel in compared crystals) remain constant. The law
to λmin for any given set of single crystal lat- holds with a great accuracy, up to fractions of
tice constants a, b, c with orientations α, β, γ , minutes. The law was discovered in quartz (SiO2 )
α0 , β0 , γ0 (see Laue conditions). The X-ray pat- and haematite (Fe2 O3 ) crystals. In 1772, J. Romé
tern is recorded on a planar X-ray sensitive film de L’Isle established the law’s validity for crystals
positioned normally to the incident beam behind of all materials after carrying out a very large num-
a single crystal (Laue pattern) or in front of it. ber of measurements.
704 LAW OF RATIONAL INDICES
LAW OF RATIONAL INDICES, Haüy law, law of made at an angle from several minutes to several
integers (R. Haüy, 1783) degrees relative to the surface (destructive meth-
An empirical law of geometrical crystallo- ods of layer-by-layer analysis), or by “looking
graphy stating that, if any triad of non-coplanar through” with the help of physical measuring tools
crystal edges is assumed as coordinate axes a, at the material state at fixed depths without any
b, c, the ratios of the intercepts on the axes by any destruction of the sample. Layer-by-layer analy-
two crystallographic (cleavage) planes are equal sis is used in combination with the methods of
to the ratio of small (usually not exceeding five) radioactive isotopes, Mössbauer spectroscopy, mi-
prime integers: (a1 /a 2 ) : (b1 /b2 ) : (c1 /c2 ) = h : croprobe X-ray spectrum analysis, X-ray analysis,
k : l. There are rules of “alignment” adopted for etc. The removal of layers is performed mechani-
each symmetry type (crystal system) for the se- cally, chemically, electrolytically, by ion and elec-
lection of coordinate axes and individual planes tron etching, by vacuum evaporation, etc. The ac-
with parameters a1 , b1 , c1 . The validity of the curacy of determining the thickness of a removed
law of rational indices follows from the crystal layer in case of destructive layer-by-layer analy-
lattice structure and from the parallelism of edges sis does not exceed 0.1 μm, while nondestructive
and planes to rows and networks of the lattice. The testing techniques allow one to investigate layers
smallness of the h, k, l integers is explainable by from 1 nm thickness and higher.
physical factors: planes with high reticular (atom)
density, form a body with minimal surface free en-
LAYERED STRUCTURE, ORDER–DISORDER
ergy (see Surface energy). The law holds exactly
for microscopic portions of a surface; macroscopic See Order–disorder layered structure.
faces of a real crystal may fail to obey it due to an
atomic step-shaped relief or to structural defects LEAD (Lat. plumbum), Pb
(see Atomic steps on a surface). A chemical element of Group IV of peri-
odic system with atomic number 82 and atomic
LAW OF ZONES, zone law, Weiss’ zone law mass 207.19. Natural lead consists of four stable
An empirical law of crystallography stating that isotopes 204 Pb, 206 Pb, 207 Pb, 208 Pb. Electronic
each plane, parallel to two actual or possible edges configuration of outer shell is 6s 2 6p2 . Successive
or lines of a crystal, is an actual or possible crystal ionization energies are 7.37, 14.91, 31.97, 42.32,
face, while each direction parallel to the zone axis 68.8 eV. Atomic radius is 0.173 nm; radius of
is an actual or possible crystal edge. The zone law Pb4+ ion 0.084 nm. Oxidation states are +2, +4.
is expressible via crystallographic indices of planes Electronegativity is 1.55.
and directions as hu + kv + lw = 0, where h, k, l In a free form, lead is a bluish-gray metal. It
are Miller indices of a plane lying in a crystal zone, has face-centered cubic lattice with parameter a =
and u, v, w are zone indices of directions (edges) 0.49495 nm, space group F m3̄m (Oh5 ). Density
of the zone.
is 11.344 g/cm3 ; Tmelting = 600.4 K, Tboiling =
LAWRENCIUM, Lr
2013 K. Heat of melting is 4.81 kJ/mole, heat
See Transuranium elements. of evaporation is 178 kJ/mole; specific heat is
129 kJ·kg−1 ·K−1 (at 273 to 373 K). Thermal con-
LAYER-BY-LAYER ANALYSIS ductivity factor is 34.8 W·m−1 ·K−1 ; coefficient
Method of investigating the structural state, of linear thermal expansion is 27.56·10−6 K−1 (at
phase, elemental composition, and also concen- 293 K). Coefficient of self-diffusion is 7.32·10−16
tration distribution of atoms of different elements m2 /s (at 500 K). Debye temperature is 95 K.
according to the depth of material. It is used in Adiabatic elastic moduli: c11 ≈ 48.1, c12 ≈ 40.7,
the physics of metals, metal science, and another c44 ≈ 14.7 (GPa) at 298 K. Young’s modulus is
areas of solid state physics. Layer-by-layer analy- 28.2 GPa, Brinell hardness is ≈40 MPa. Tensile
sis is performed by the successive removal of strength is ≈12.5 MPa, relative elongation is 55%.
layers of the required thickness from the investi- Coefficient of linear term in low temperature mo-
gated object, by the investigation of an oblique cut lar electronic specific heat is 2.98 mJ·mole−1 ·K−1.
LEBEDEV CRYSTALLITE HYPOTHESIS 705
Electrical resistivity is 20.68·10−8 ·m. Super- the wetting of the leaded surface the melted lead
conducting transition temperature is 7.17 K. Nu- should contain antimony, tin, copper, or zinc.
clear magnetic moment of 207 Pb is 0.584 nuclear Electrochemical lead plating provides high pu-
magnetons. Lead is used in the chemical industry, rity coatings with better physical and mechanical
in electronics, in semiconductor engineering, for properties. A reliable protection of ferrous metals
protection against radiation. Cases of wires and in corrosion-active media is provided by pore-free
cables are manufactured from lead. See also Lead lead plating at least 200–300 μm thick. Elec-
alloys. trolytes that contain lead salts, and are usable
for electrochemical lead plating, include fluorine
LEAD ALLOYS borite, phenolsulfone, and other electrolytes con-
Alloys based on lead. They are characterized by taining lead salts.
high density, by low strength, hardness, melting
temperature, by good antifriction properties, by LEBEDEV CRYSTALLITE HYPOTHESIS,
relatively high acid strength towards some dilute Lebedev glass structure hypothesis
acids (sulfuric, sulfurous, chromic, etc.). Dilute (A.A. Lebedev, 1921)
and concentrated lead alloys (see Alloying) are A hypothesis on the structure of glasses. Ini-
known. The first group comprises lead alloys with tially, having measured the temperature dependen-
small additions of Fe, Cu, Sb, Cd or Ca, which cies that the coefficients of refraction of light and
do not reduce, and in some cases increase the cor- thermal expansion of dense silicate glasses fea-
rosion resistance of lead, and therefore increase ture within narrow regions of temperatures near
its creep limit and long-term strength. The second structural transformations in crystals of quartz,
group comprises lead alloys which contain con- tridymite and cristobalite, Lebedev supposed that
siderable quantities of elements which increase glasses had a finely dispersed crystal structure.
their strength, hardness and antifriction properties, His later analysis (1933, 1948) included numerous
and lower the melting temperature and shrink- experimental data on temperature dependences of
age upon casting. Lead alloys are obtained by their internal friction and data from X-ray structure
the melting of primary or secondary metals, and analysis. Summing it up, he concluded that glasses
also of their alloying elements. Those with sodium, contain regions of perfect short-range ordering,
potassium and calcium are formed by an electro- 0.7–1.5 nm in size, surrounded by molecular as-
chemical method, i.e. by electrolysis of the melt sociations with less perfect ordering. That is, the
salts together with a liquid lead cathode. There are glass is structurally a random continuous network
antifriction alloys (see Antifriction materials), typo-
graphic alloys and solders. The majority of alloys
are soft, easily subjected to mechanical process-
ing: to rolling, stamping, forging, etc. Lead alloys
are used as bearing materials, typographic and
other low melting temperature alloys.
LEAD PLATING
Coating of various pieces and units with lay-
ers of lead. It is used to manufacture lead storage
(automobile) batteries, to screen X-rays, and to
protect ferrous and nonferrous metals from cor-
rosion that would otherwise result from the effect
of solutions of sulfuric acid or chromic acid and
their salts in chemical installations. To produce
lead plating the pieces are immersed into molten
lead (hot technique), or liquid lead is sputtered Crystallite model of disordered solid (ordered phase ar-
onto their surfaces (metallization). To improve eas are hatched).
706 LED
featuring nonuniform shifting intermediate order where N is the total number of particles, and Ei
(see Fig.), in agreement with diverse experimen- is the energy of the ith level. Thus, according to
tal data on the structure of glass. See also Vitreous Eq. (1), the level population decreases exponen-
state of matter, Polycluster amorphous solids. tially as its energy increases. An opposite situation
is found in the case of population inversion where
LED higher levels are more populated than lower ones.
Acronym for Light emitting diode. This inversion is a necessary precondition for the
functioning of lasers and masers. The most typi-
LEDEBURITE (after German metallurgist cal way to produce inverted populations consists
A. Ledebur) in equalizing the populations of a pair of levels
Structural compound of iron–carbon alloys i–k via external actions such as high power elec-
(mostly, cast irons) that is a eutectic mixture (see tromagnetic wave pumping (see Fig. in Quantum
Eutectics) of austenite and cementite. It forms
ampli er). It may then happen that for a differ-
at a temperature below 1145 ◦ C (carbon content ent pair of levels, i–j , such that Ei > Ej > Ek ,
is 4.3%). Ledeburite at room temperature con-
one obtains ni > nj . Inverted populations during
sists of cementite and pearlite. In steels subjected
magnetic resonance can be produced by causing
to alloying with chromium, tungsten, and other
the system to undergo an adiabatic fast passage
carbide-forming elements, ledeburite appears at
through a resonance transition. The extent of in-
low carbon content (0.7 to 1%) and is a eutectic
version, viz., the number of particles at inverted
mixture of alloyed austenite and carbides of alloy-
higher levels, depends on various external factors:
ing elements.
energy gaps, probabilities of relaxation transitions,
LENNARD-JONES POTENTIAL etc.
(J.E. Lennard-Jones, 1924) When the particles possess a magnetic moment
A pairwise interaction potential between atoms (i.e. spin), their distribution over magnetic sub-
often used to describe the structure of the con- levels is characterized by a spin temperature, Ts ,
densed state of matter and its dynamics: which is determined via the Boltzmann distribu-
1/2 6 tion:
σ σ
U (r) = 4ε − , ni Ei − Ej
r r = exp − . (2)
nj kB Ts
where r is the interatomic distance, the parameter
σ characterizes the range of the interaction, and ε Provided such particles (the spin system) are
is its magnitude (see also Van der Waals forces). in equilibrium with the lattice, one finds from
The attractive r −6 term arises from a fluctuating Eqs. (1) and (2) that Ts = T . Applying an exter-
induced dipole–dipole interaction, and the r −12 nal action, one may either cool the spin system
term represents the ion core repulsion. The po- (Ts < T ) or heat it (Ts > T ). A spin system is
tential works well for inert gas (Ne, Ar, Kr, Xe) cooled when its energy gap Ei − Ej is adia-
crystals. batically narrowed (see Adiabatic demagnetization
cooling) through application of alternating mag-
LEVEL POPULATION netic (electric) fields to a system of magnetic
The number of particles (atoms, molecules, (electric) dipoles (see Magnetoelectric effect), etc.
nuclei, paramagnetic impurities, etc.) occupying A characteristic example of a heated spin sys-
energy levels of a quantum system characterizing tem involves populations ni and nj equalized
its various states. Provided there is no degeneracy, through saturation effects. Indeed, if ni = nj
the number of particles satisfies the Boltzmann (while Ei = Ej ), then, according to Eq. (2) we
distribution for thermodynamic equilibrium at the have Ts = ∞. In the case of an inverted popula-
ith level, given by tion, the distribution of particles (not necessarily
−1 having spin) over the energy levels is character-
E Ej
ni = N exp − i exp − (1) ized by a negative temperature. If ni > nj (while
kB T kB T Ei > Ej ), then we have Ts < 0 in accordance
j
LEVITATING ELECTRONS 707
LEVELS, SHOCKLEY
See Shockley levels.
LEVITATING ELECTRONS
Electrons localized at surface levels formed
near a vacuum-dielectric interface with negative
electron af nity. They are produced under the ef-
fect of electrostatic image forces with an attractive
potential V (z) = −Λ/z, where
e2 (ε − 1)
Λ= , (1)
4(ε + 1)
ε is the relative dielectric constant, and z is the Fig. 1. First derivative dA/dE of absorbed radiofre-
distance measured from the insulator surface into quency power, A, with respect to electric field E versus
vacuum. The surface state spectrum has the form: the potential V of the external clamping field. Res-
onances correspond to electron transitions from the
Δ p2 ground energy level (l = 1) to excited levels (l 2) as
El (p ) = − 2 + , (2) indicated by arrows.
l 2m
mΛ2
Δ= , l = 1, 2, 3, . . . , (3) (ripplons). They dissipate energy in the process to
22
produce a finite resonance width.
where p = (px2 + py2 )1/2 is the longitudinal two- The two-dimensional electron system of fi-
dimensional momentum in the surface plane. Lev- nite surface density, ns ∝ 104 –109 cm−2 , above
itating electrons with their spectrum (2) may exist the liquid helium is an ideal target for studying
on the condition that the width of their localization various collective effects related to the Coulomb
region, λl = 2 l 2 /(Λm), considerably exceeds interaction. It was in the system of levitating
that of the interface, the latter being of the order electrons that two-dimensional plasma oscillations
of an interatomic distance (a ∝ 10−8 cm). Such were first observed, and instabilities of the charged
a condition is met, e.g., when electrons are local- surface of liquid helium were studied, includ-
ized above the surface of liquid helium, for which ing the formation of a two-dimensional honey-
Λ ∼ 0.01. In that case, the electrons would “soar” comb crystal (Fig. 2), and of charged bubbles
above the liquid surface in the gas phase (vapor) (bublons) in strong electric fields, and where the
at the quasi-microscopic height λl 10−6 cm, two-dimensional electronic Wigner crystal (crystal
justifying the term “levitating”. The lifetime of of levitating electrons) was first discovered, with
levitating electrons increases when an additional its existence region corresponding to particular
electric field is switched on to press them to the values of the dimensionless parameter Γ satisfy-
liquid helium surface. Fig. 1 shows the resonance ing the condition Γ ≡ e2 (πns )1/2 /(k B T ) > 137
absorption of radiofrequency energy at a fixed (Fig. 3). At Γ = 137 during the liquid–crystal
frequency. This occurs by varying the clamping phase transition a two-dimensional triangular lat-
electric field strength, E, forcing the electrons to tice of identically charged electrons forms, the
jump between discrete surface levels, the distance first known example of Coulomb crystallization.
between the latter depending on E. Electrons scat- The formation of the Wigner (electron) crystal is
ter off the helium atoms (in the vapor) and off detected by the appearance of specific energy ab-
thermal excitations at the liquid helium surface sorption resonances in the electromagnetic field,
708 LEVITATION CRYSTALLIZATION
Fig. 2. A hexagonal honeycombed crystal on the surface of liquid helium in the clamping field E (under lateral illumina-
tion with a laser).
LEVITATION CRYSTALLIZATION
Technology of growing crystals in the regime
of microgravity (10−3 g) is achievable on board
long-time orbital space stations. Levitation crys-
tallization practically eliminates mixing of liquid
layers by natural convection, and makes it easier
to obtain perfect crystals. It features stronger lam-
Fig. 3. Phase diagram of a two-dimensional system of inar mass transport to the crystallization front, and
levitating electrons above the surface of liquid helium
1/2 prevents contamination of the growing crystals
in the ns versus T plane. The straight line indicates
by container wall material. Levitation suppresses
the boundary of the Wigner crystal–liquid phase transi-
tion from the liquid phase below to the crystalline state the effect of segregation when alloys with differ-
above. ing component densities crystallize, and broadens
the possibilities of capillary formation of vari-
and by other techniques. Specific electron–ripplon ous complex prescribed profiles (see Monocrystal
interactions define the structure of resonances in growth), even hollow shapes. See also Levitation
the crystalline state. There are grounds to believe melting.
LIFETIME 709
its initial value
p0 is determined by the relation was carried out for the density of states, and there-
p =
p0 exp(−t/τp ). This expression becomes fore, for all thermodynamic quantities. All of the
more complicated in the presence of traps. One results mentioned were based on the idea of the
should not confuse the lifetime of an electron existence over a definite region of the spectrum
(hole) in a band with their relaxation times within of representative uctuations, which provide the
the band. The latter are determined by processes main contribution to the density of states. This
of energy and momentum relaxation. In metals idea proved to be especially fruitful for the con-
the lifetime of an excited electron (time of ther- struction of uctuation levels, where it developed
malization) is determined by the electron–electron into an effective computational method for the op-
scattering time (generation of electron–hole pairs), timal fluctuation, which found wide application in
and by the emission of phonons. The rate of the theory of highly-doped semiconductors.
both processes strongly depends on the tempera-
ture, and is also sensitive to the amount of static LIFSHITS–ONSAGER HAMILTONIAN
disorder (impurity scattering). In a similar way (I.M. Lifshits, L. Onsager, 1952)
the lifetime for other non-equilibrium particles A Hamiltonian describing the dynamics of con-
and quasi-particles in solids can be introduced, duction electrons in an external magnetic field.
e.g., for phonons, magnons, vacancies, interstitial The Lifshits–Onsager Hamiltonian (H ) is con-
atoms, etc. For high concentrations of photons or structed on the basis of the dispersion law for free
phonons, induced transitions (see Induced radia- electrons, ε(p) (p is the quasi-momentum), ac-
tion) play the main role in determining the value cording to the following correspondence principle:
of the lifetime. ε → (H ) ≡ ε(p ),
p=P − eA( r ). Here P is the
generalized momentum operator, and A = ∇ × B
LIFSHITS MODEL (I.M. Lifshits, 1963) is the vector potential. The order of operation
A model of structural disorder (in amorphous of the mutually noncommutative components of
materials or liquids) which describes the motion of the kinematic momentum operator, p , is defined
electrons in a field produced by randomly arranged by the symmetrization rule: the periodic function
point scatterers (impurity centers). The Lifshits ε(p) is represented as a Fourier series, and the
model is employed to study the impurity band in substitution p → P is made in the exponent of
the energy spectra of electrons in disordered ma- each term of that series. The Lifshits–Onsager
terials, a band arising due to the broadening of Hamiltonian is approximate, with terms of the or-
a local electronic level. In the leading approxima- der of χ = eBa 2 / (a is the reciprocal lattice
tion (in the concentration of impurities) the energy parameter) dropped from it. In fields of 1–10 T we
levels and quantum states of the impurity band have χ ∝ 10−5 –10−4 . A limit to the applicability
acquire in this model a descriptive geometrical for- of this Hamiltonian is related to the emergence of
mulation, in which wave functions are localized at band-to-band transitions in strong magnetic fields
one or two impurity centers (although the energy (see Magnetic breakdown).
levels also depend on the positions of other cen-
ters), while the density of states and the spatial LIFSHITS–ONSAGER QUANTIZATION RULES
correlations are represented on a certain scale by Rules of quasi-classical quantization of en-
universal functions. In the immediate vicinity of ergy levels of conduction electrons in a uni-
the localized level, where within the framework of form magnetic field B. Lifshits–Onsager quanti-
the geometrical formalism the density of states has zation rules appear when the quasi-classical Bohr–
a dip, other states become important, collectiviz- Sommerfeld quantization rules are applied to the
ing a large number of centers, as a result of which Lifshits–Onsager Hamiltonian, and provide the re-
the dip fills up to a significant extent. The structure lation S(E, pz ) = 2πeB(n + 1/2). Here S(E, pz )
of the spectrum and of the states in the vicinity is the area bounded by a closed curve in mo-
of the edge of the initial spectrum was studied mentum space that a classical electron traces in
on the basis of this model, and an expansion in a magnetic field B at its energy E, and the pro-
terms of the powers of the impurity concentration jection pz of its momentum on the magnetic
LIGANDS 711
case of two Cl− ions in the molecule the number of the applied cascades of polariza-
tion switching, each consisting of a separating
element, a crystal with high birefringence (e.g.,
calcite CaCO3 , calomel Hg2 Cl2 ) and an elec-
The nature of the chemical bond between lig- trooptical polarization switch, see Electrooptical
ands and the central atom and its effects upon the effect). Acoustooptical deflectors produce the light
magnetic, spectral, and other properties of com- deviation at a dynamic diffraction grating, formed
plex compounds are the subject of ligand field the- within the acoustooptical element by a connected
ory and crystal eld theory. The reaction capability piezoelectric vibrator operating at a tunable ultra-
of complex compounds depends significantly on sonic radiation frequency (see Acousto-optics).
the mutual effects of ligands within their inner Traditional light deflectors meet most modern
sphere; in particular those produced by ligands in a requirements except for fast response. The latter
straight line, with the central atom on its opposite has been considerably improved (up to the order of
sides. nanoseconds) by the transition to integral-optical
versions of light deflector design, which are char-
LIGHT ABSORPTION by solids acterized by the simultaneous reduction of energy
Decrease in intensity of radiation due to trans- consumption by three or more orders of magni-
fer of radiation energy to the medium, through tude.
which this radiation propagates. Quantum-me-
chanical processes of light absorption by solids LIGHT DISPERSION
involve the of annihilation of light quanta, with See Dispersion of light.
the energy transferred to electrons, phonons, ex-
citons and other quasi-particles, and to defects. LIGHT EMITTING DIODE (LED)
Therefore, the investigation of light absorption by An optical (visible wavelength) transformer of
solids can provide information on energy bands electrical energy into incoherent light emission. It
of crystals, the density of electron states, the en- consists of a semiconductor junction to which a
ergy spectrum of phonon and exciton excitations voltage (Vg = 1.2–2.5 V) is applied in the con-
(see Band theory of solids), on electronic and ducting direction. To operate the unit one has to in-
vibrational states of impurity centers and other ject secondary current carriers (e.g., electrons) into
defects, and on the interaction between quasi- the contact via a homogeneous or heterogeneous
particles. At high radiation intensities, processes p–n junction, so that the injected carriers recom-
of many-photon absorption are possible (see Op- bine with the principal carriers at the base (e.g.,
tical spectroscopy, Nonlinear optics, Two-photon holes). The energy thus released, eVg , is partially
light absorption, Infrared spectroscopy). dissipated by the diode body as heat (emission-free
recombination, see Nonradiative quantum transi-
LIGHT DEFLECTORS tion), and most of the remainder transforms into
Devices for the smooth or discrete deviation of light (emission recombination). This emitted ra-
a light beam direction to a given angle. Many light diation Phν has an ampere–watt (or ampere–
deflectors use the natural or induced (electrical brightness) profile Phν (j ), that depends on the
or mechanical) field variation of birefringence of injection current density. Typical values of j
electro-acousto-optical crystals, or of transparent and Phν lie within the limits 5–500 A/cm2 and
ceramics. They are employed in projection devices 0.2–0.5 mW, respectively. Light emitting diodes
for the optical processing of information, for the are characterized by their internal, ηi , and ex-
synthesis of electronic integrated circuit topology, ternal, ηe , efficiencies. The internal efficiency ηi
in computer optical storage devices, etc. gives the relative share of emission recombina-
There exist electrooptical deflectors of the tion, while ηe = ηi F . Here F = 0.04 ÷ 0.06 is
analogue type (deviation changes smoothly with the aperture coefficient accounting for radiation
variation of control voltage) and discrete deflec- losses during emission from the LED. Depend-
tors (deviation changes in steps determined by ing on the design and the material efficiency,
LIGHT SCATTERING 713
the ratio Phν /(Ig Vg ) (here Ig is the LED cur- LIGHT SCATTERING
rent) may reach 0.5–2, the corresponding values Phenomenon taking place during the propaga-
of ηe being 10–12% for 50% < ηi < 100%. The tion of light through various optically inhomoge-
probability of emission recombination is higher neous media. The light wave experiences diffrac-
in semiconductors with direct band gaps, com- tion of light by inhomogeneities, thereby deviating
pared to that for semiconductors with an indirect from its initial direction of propagation. The prin-
band gap (see Indirect transitions). Therefore, such cipal quantitative characteristics of light scattering
LEDs are normally more efficient. The princi- are the differential scattering cross-section dss (ra-
pal materials used for LEDs are of the AIII AV tio of the radiation flux dIs , scattered into an
type. These are binary (GaAs, GaP, GaN), ternary elementary solid angle dΩ, to the incident flux I0 :
(Alx Ga1−x As, GaAs1−x Px , Gax In1−x As), and dss = dIs /I0 ); the scattering phase function (de-
quaternary (Gax In1−x Asy P1−y ) compounds, as pendence of scattered intensity on scattering an-
well as AIV BIV compounds (e.g., SiC). They are gle), the frequency spectrum, and polarization of
doped with corresponding impurities to achieve n- light.
and p-type conductivity and the desired radiation Physical causes producing optical inhomo-
spectra from blue (470–483 nm, GaN, SiC) to red geneities are quite diverse. The medium may be-
(690–780 nm, GaP, ZnO, AlGaAs) and IR (from come inhomogeneous, e.g., when one material,
820–910 to 1300–1700 nm, AlGaAs, GaInAs, differing in its refractive index, is included in an-
GaInAsP). To obtain the required emission color other. The scattering phase function depends on
a double transformation may also be used, when the ratio between the inhomogeneity size and the
one combines an IR LED and an anti-Stokes lu- light wavelength λ0 . Light scattering may also be
minophor. To increase ηe and color contrast, and observed in macroscopically homogeneous media
to form the desired phase function, both poly- (i.e. containing no impurities) due, e.g., to ran-
mer and glass lenses are used. The LED current dom thermal motion of the material’s particles. In
Ig ∝ exp[eVg /(kB T )] is controlled by an external that case, the reason for the optical inhomogeneity
limiting resistance and remains within 10–100 mA is fluctuations of the dielectric constant produced
for crystals 400 × 400 μm2 in size. Light emitting by different excitations of the medium. When the
diodes on the market are offered both with and light wavelength λ0 differs considerably from the
without special casing. They serve as indicators, wavelengths of the normal mode frequencies of
small-scale data presentation devices, are used in the medium, then we have σs ∝ λ−4 0 (Rayleigh
optoisolators, function in read-out devices that use law). However, as λ0 approaches the wavelengths
magnetic tapes for data storage, in short-range of the normal modes of the medium, the Rayleigh
fiber optic communication lines, sensor elements, law is no longer satisfied, and the intensity of the
etc. scattered light may be strongly enhanced (reso-
nance light scattering).
LIGHT GENERATION From the quantum point of view light scat-
See Parametric light generation. tering is treated as the absorption of the incident
photon, its energy ω0 , momentum k0 and polar-
LIGHT PIPE ization e0 , by the medium, a process accompanied
Dielectric waveguide (see Integrated optics) in by the simultaneous emission of another photon
the optical range. See also Fiber optics. with energy ωs , momentum ks and polariza-
tion es . Provided the emitted photon frequency is
LIGHT POLARIZATION equal to that of the absorbed one (ω0 = ωs ) the
See Polarization of light, Self-induced light po- light scattering is called elastic or Rayleigh scat-
larization change, Rotation of light polarization tering of light, and when ω0 = ωs it is inelastic
plane. scattering. Meanwhile, the corresponding energy
difference |ω0 − ωs | = ω is transferred to the
LIGHT REFRACTION medium, so the medium transforms to a differ-
See Refraction of light. ent quantum state. Light scattering may take place
714 LIGHT SCATTERING IN LIQUID CRYSTALS
Scattering processes: (1) Rayleigh; (2) nonlinear Rayleigh (hyper-Rayleigh); (3) inelastic Stokes; (4) inelastic anti-
Stokes; (5) nonlinear inelastic (hyper-Raman) Stokes; and (6) nonlinear inelastic (hyper-Raman) anti-Stokes.
when the scattered photon energy is either lower The direction of the optical axis of the liquid
than the energy of the incident photons, ωs < crystal, as defined by the predominant orienta-
ω0 (Stokes scattering), or greater than that en- tion of its molecules, then changes. Scattering
ergy, ωs > ω0 (anti-Stokes scattering). Inelastic in smectic liquid crystals is usually weaker than
light scattering in a solid may be accompanied in nematic types, since the stratified structure of
by either the generation or annihilation of vari- the smectics hinders the development of fluctua-
ous types and numbers of quasi-particles. Typical tions of the optic axis direction. At the same time
examples of such processes are light scattering inhomogeneities at the surface of liquid crystals
by acoustic phonons (Brillouin scattering), optical and defects (e.g., disclinations, confocal domains)
phonons, magnons, surface and bulk polaritons, introduce perturbations into their structure, and
etc. (see Raman scattering of light). result in light scattering that exceeds the scatter-
With the development of powerful sources of ing from thermal fluctuations. Besides, smectics
coherent radiation (lasers), it also became possi- of the C-type, in which the molecules are tilted
ble to study nonlinear light scattering processes with respect to the layer planes, are as opaque as
during which a single quantum of scattered light nematics, because of fluctuations of the azimuth of
is produced by two or more light quanta incident the molecular inclination. Studies of the character-
upon the medium (see Fig.). Lasers make it pos- istics of light scattering provide information on the
sible to use the techniques of active spectroscopy most important parameters of liquid crystals, their
as well. In addition, as is also the case with light elastic moduli, viscosity factors, energy of interac-
emission, induced light scattering becomes possi- tion with the substrate, etc. These parameters are
ble along with spontaneous light scattering during important for the design of liquid crystal displays
intense laser excitation involving already scattered (LCDs).
photons.
See also Induced light scattering, Photoinduced LIGHT WAVEGUIDE
light scattering. See Optical waveguide.
broadening) contributions to the level width are (3) If V (t) strongly disturbs the equilibrium of the
determined, as a rule, by different interactions, quantum system, then the spectral line width
but they could involve the same interaction (e.g., may reflect changes in the spectral line due
electron–phonon interaction for electron impurity to the new equilibrium. For instance, optical
states). transitions associated with the recharging of
The intrinsic level widths are related to the impurity centers may affect the quantity of
experimentally measured line width
Ej k for the recharged ions, and hence result in an ad-
transitions between energy levels Ej and Ek in ditional broadening of the spectral line. The
accordance with the expression
Ej k
Ej + inverse situation is possible (e.g., in high-
Ek . The spectral line recording always involves resolution NMR), when the disturbance cools
an additional (applied perturbation) interaction the quantum subsystem, which is responsible
V (t) giving rise to the transitions. For example, for the spectral line, leading to the suppres-
for electric dipole transitions V (t) is the dipole sion of fluctuations which cause broadening.
moment operator er interacting with the oscillat- The equilibrium can also be disturbed when
ing electric field vector of an incoming electro- the exciting signal is a high-power pulse.
magnetic wave. There are also magnetic dipole,
Experimentally the spectral line width is de-
electric quadrupole, and higher-order multipole
termined by the energy spacing (or that of an-
transitions. The width depends on the method of
other parameter such as wavelength or magnetic
recording the response to V (t), such as the scan-
field B which is scanned) between the points,
ning frequency or magnetic field, and it can also
where the line intensity has dropped to one-half
depend on the strength or intensity of the exciting
of its maximum value (Δ1/2 ). When the spectrum
radiation. Some typical situations are:
is recorded as a first derivative, the spectral line
(1) The spectral line width reflects the level width width is characterized, as usual, by its peak-to-
only partly when V (t) excites transitions in peak value (ΔPP ) which is the spacing between the
such a manner that transition frequencies ωj k extrema of the derivative recording. For most line-
do not depend on the parameters broadening shapes the peak-to-peak width is less than the half
the energy levels. Electron paramagnetic reso- amplitude width (Δ1/2 /ΔPP = 31/2 = 1.7321
nance (EPR) of a shallow donor, whose levels for Lorentzian shaped lines, and Δ1/2 /ΔPP =
are broadened by the fluctuating electric fields (2 ln 2)1/2 = 1.177 for Gaussian lines).
of charged impurities, may be an example of The proper theoretical calculation of the spec-
this type. Such a broadening may mount up to tral line width requires a knowledge of its line
several meV and be observed in optical spec- shape. However, in the case of homogeneous
tra, while the EPR spectral lines of shallow broadening, its width is calculated from the knowl-
donors are broadened by only several μeV, edge of its corresponding relaxation times. In the
through much weaker magnetic interactions case of inhomogeneous broadening the spectral
with nuclear and electronic spins. lines are sometimes characterized by moments of
(2) If a frequency of the quantum transition re- spectral lines.
sponsible for the spectral line is exposed to
the influence of time varying (fluctuating) pa- LINE SHAPE
rameters (fields), then the spectral line width Distribution of spectral line intensity as a func-
will reflect their distribution function gj (E) tion of the variable according to which the spec-
only when the transition frequency ωj k = trum is recorded (wavelength λ, frequency ν, mag-
|Ej − Ek |/ exceeds the reciprocal of the cor- netic field B, etc.); it corresponds to the form factor
relation time τc of the fluctuating fields. For of the line. Experimental line shapes are affected
the case ωj k < τc−1 an averaging of the fluctu- by instrumental settings as well as by physical fac-
ating action takes place, the so-called dynamic tors, so line shape studies can provide information
line narrowing effect. This case, which is not about the nature of the broadening mechanisms.
unusual, occurs in various types of magnetic The line shape may be characterized by a mathe-
resonance. matical expression, or by a set of moments. The
LIQUID-CRYSTAL DISPLAYS, LCD 719
most common line shapes F (x) are a Gaussian resulting line shape function is a convolution of I1
shape, a Lorentzian shape, and also their convo- and I2 , so that
lution (see Line shape convolution):
I (ω − ω0 ) = I1 × I2
FG (x) = exp −x 2 ln 2 (Gaussian),
∞
1 = I1 (ω − ω)I2 (ω − ω0 ) dω .
FL (x) = (Lorentzian),
1 + x2 −∞
where x = (ω0 − ω)/Δ, the center frequency of In the most common and simplest case I (ω − ω0 )
the resonance is ω0 , and Δ is the half-width at is a convolution of a Gaussian and a Lorentzian
half-amplitude. shape
Specific line shapes can be due to orientational
line broadening, and also due to the action of spe- I (ω − ω0 ) = G × L
cific physical mechanisms (Dyson shape, Holts- ∞
mark shape). A line shape can arise from several
= G(ω − ω)L(ω − ω0 ) dω .
sources of inhomogeneous broadening (hyper ne
interaction, superhyperfine interaction, unresolved −∞
fine structure, etc.), or it can result from the su-
perposition of overlapping lines. For the best res- LINE SHAPE, HOLTSMARK
olution and analysis the first, second and third See Holtsmark shape.
derivatives of the line are recorded. Also math-
ematical analogues of complex lines, calculated LINE SHAPE, STATISTICAL THEORY
with the help of a computer, can be fit to exper- See Statistical theory of line shape.
imental spectra with the proper choice of one or
more variable parameters. LIQUATION
Separation (segregation) of a homogeneous
LINE SHAPE CONVOLUTION (single-phase) multicomponent liquid (solution,
A composite line shape that results from the melt) into two or more phases differing in com-
simultaneous effect of two or more spectral dis- position. This can be done, e.g., by heating an
tributions. The line shape function of a resonance alloy to melt the more fusible matter. New phases
absorption line in a paramagnetic system is de- that form during liquation may be either liquid
fined by the spectral distribution of absorption or solid (vitreous or crystalline). Liquation plays
probability for each separate paramagnetic center, an important role during vitri ca tion of multicom-
I1 (ω − ω), and the spectral distribution of the ponent supercooled (metastable) highly viscous
number of particles with a given energy, liquids. Sometimes, though rarely, the term “liqua-
I2 (ω − ω0 ). Depending on the nature of the inter- tion” is used to denote the process of separating
action of a paramagnetic particle with its environ- new phases from the initial one that differ from it
ment, the function I1 (ω − ω) may be described only in structure.
by various distribution laws. The ones most often
encountered in practice are the Lorentzian shape LIQUID CRYSTAL, CHIRALITY
(ω − ω)2 −1 See Chirality in liquid crystals.
L(ω − ω) = π −1 1 +
Γ2
LIQUID-CRYSTAL DISPLAYS, LCD
and the Gaussian shape Devices for visualizing information in the form
of letters, numbers, symbols, lines, etc. based on
(ω − ω)2
G(ω − ω) = π −1/2 exp − , a liquid crystal as an operating agent. The action
Γ2 of a liquid crystal display is based on the ability
where Γ is the linewidth. The function of a liquid crystal to change its optical proper-
I2 (ω − ω0 ) is usually Gaussian, G(ω − ω0 ). The ties under the action of an external electric field
720 LIQUID CRYSTAL, LIGHT SCATTERING
(or temperature). The field causes either a coor- polymers, into thermotropic (formed by change in
dinated rotation of liquid crystal molecules (see temperature) and lyotropic (formed by change in
Electrooptical effects in liquid crystals) or a strong concentration) types, and they form smectic, ne-
liquid crystal flow (see Electrohydrodynamic insta- matic and cholesteric mesophases. Liquid-crystal
bilities). In the first case, the optical axis changes polymers might be used for film dichroic po-
its direction; in the second case, a strong spatial laroids, and color indicators based on them.
nonuniformity of the effective refractive index de-
velops. Both effects are reversible; they can be LIQUID CRYSTALS
used for controllable modulation of the intensity of Thermodynamically stable states of matter with
light passing through the liquid crystal layer. Dyes physical properties intermediate between those of
are introduced into a liquid crystal to bring about a liquid and a solid crystal. Liquid crystals are
a color change. composed of molecules or groups of molecules
Two types of liquid-crystal media suitable for
(e.g., micelles) often unsymmetrical in shape.
displays are ferroelectric liquid crystals (providing
Their most general features are the presence of
a high switching rate) and liquid crystals dis-
long range orientational order in the arrangement
persed into polymer arrays. These media have
of the asymmetric molecules, and the complete or
good prospects for the creation of TV screens, and
partial absence of translational order of the loca-
of flexible wide-angle displays. There are also dis-
tions of the centers of mass of the molecules.
plays based on a thermal method of information
recording. A laser beam that scans across the layer
Liquid crystals can be thermotropic and ly-
of a smectic liquid crystal introduces local changes otropic. Thermotropic liquid crystals form dur-
due to melting, whereby its refractive index is ing a temperature change: during melting a solid
changed. The stored information is erased by an crystal may, before entering the isotropic liquid
electric field which restores the initial structure state, pass through one or more mesomorphic
of the liquid crystal. Due to their simple man- (intermediate) phases or mesophases. Lyotropic
ufacturing technology, high economic efficiency liquid crystals form when dissolving asymmet-
and functional variability, liquid crystal displays ric objects in suitable solvents. Proceeding from
are widely used as alphanumeric indicators, TV rod-like to disc-like shaped component molecules,
screens, as well as in electronic watches, calcula- liquid crystals are subdivided into calamitics and
tors, etc. discotics; and those formed from polymer macro-
molecules are called liquid-crystal polymers. Ac-
LIQUID CRYSTAL, LIGHT SCATTERING cording to the manner of ordering, mesophases
See Light scattering in liquid crystals. are divided into nematic, cholesteric, smectic and
columnar (these do not exhaust the classification,
LIQUID CRYSTAL MODULATION see also Blue phases). The most widely stud-
See Wave-like modulation. ied types are thermotropic calamitics of the ne-
matic liquid crystals, cholesteric liquid crystals and
LIQUID CRYSTAL ORDER smectic liquid crystals consisting of low-molecular
See Degree of order, Order parameters in liquid weight organic materials with the common mole-
crystals. cular formula
LIQUID-CRYSTAL POLYMERS ,
Macromolecular analogs of common liquid
crystals: typical of molecules or their fragments i.e. molecules with two benzene rings connected
(e.g., lateral groups of flexible-chained comb-like by a double or triple A–B bond, terminated by
molecules) in liquid-crystal polymers is the pres- short chains R and R at the ends. Such mole-
ence of orientational order. Liquid-crystal poly- cules are about 2 nm long and about 0.5 nm thick.
mers occupy an intermediate position between Nematics are characterized by the total absence
amorphous and crystalline polymers. They are di- of translational order in the arrangement of the
vided, like low molecular weight liquid-crystal molecular mass centers, and the presence of a
LIQUID–GLASS TRANSITION 721
preferable orientation of the longer axis of rod- between liquid crystals and biological substances.
like molecules (or of normals to the planes of Thus, the term “liquid crystal” encompasses a
disc-like molecules) along some direction denoted wide class of systems – from single-component,
by the unit vector n, which is called the director. low molecular weight organic substances to com-
A cholesteric can be pictured as a twisted nematic: plex biological structures. The term is attributed
when moving along an axis perpendicular to n to the German physicist O. Lehmann (1890) who,
the director n traces a spiral. Cholesterics consist at the request of the Austrian botanist F. Reinitzer
of chiral molecules, i.e. those without planes of (1888), the discoverer of liquid crystals, carried
symmetry. Smectics have a unidimensional trans- out the first systematic studies of liquid crystals.
lationally ordered laminated structure. The best The orientational order in liquid crystals un-
known are smectics A, B and C. The longer axes of derlies the anisotropy of their physical proper-
the molecules are perpendicular to the plane of the ties: optical, electrical, magnetic, elastic, etc. (see
layer in smectics A and inclined to it in smectics C. Anisotropy of liquid crystals). The theoretical de-
The mass centers of the molecules within the layer scription of liquid crystals employs the methods
in both cases are positioned randomly. The lay- of statistical physics and hydrodynamics as well
ers can slip relative to one another, so smectics A as the language of the order parameter, accepted
and C resemble crystals in one direction, and liq- in the theory of ordered systems (see Order para-
uids in the other two directions. The orientation meters in liquid crystals and Elasticity of liquid crys-
of the molecules in smectics B is the same as that tals). Liquid crystals are systems with interesting
in the A-phase, but there is some translational and topological properties giving rise to various de-
orientation ordering (see Smectic liquid crystals). fects and textures that exist in the field of the order
Columnar phases are two-dimensionally ordered parameter (director), and can be easily observed
columns of disc-like molecules. There is no trans-
experimentally (see Defects in liquid crystals, Do-
lational order in the arrangement of molecular
main wall in liquid crystals). Thus, for example, the
mass centers within the columns; i.e. there is two-
term “nematic” (fr. Gr. νημα, a thread) originated
dimensional crystalline order in combination with
from the microscopic observation of disclinations
one dimensional liquid order (see Discotics).
looking like dark threads floating in the sample.
During an increase in temperature the best
Numerous electrooptic, hydrodynamic, and other
known mesophases replace each other in the fol-
effects are observed in liquid crystals (see Wave-
lowing sequence: Crystal ↔ Smectic B ↔ Smec-
like modulation in liquid crystals, Twist effect, Fred-
tic C ↔ Smectic A ↔ Nematic (or Cholesteric)
eriks transition, Electrohydrodynamic instabilities
↔ Isotropic liquid (some mesophases can be ab-
in liquid crystals, Guest–host effect). These ef-
sent for a particular substance). The regions of
the existence of mesophases (from fractions to fects underlie different practical applications (see
some tens of degrees) may fall into the tempera- Defectoscopy, Liquid-crystal displays, Thermogra-
ture range 60–400 ◦ C. Phase transitions between phy, Thermal indicators). Liquid-crystal polymers
the mesophases can be either of first or second serve as the source material for the manufactur-
order. Between cholesteric and isotropic liquids ing of super-durable fibers. Some lyotropic liquid
(in a narrow temperature range) sometimes blue crystals are suitable for the simulation of biologi-
phases occur with exotic and seldom-investigated cal objects such as membranes.
properties. The types of ordering of lyotropic liq- See also Cholesteric liquid crystal, Lyotropic liq-
uid crystals are mostly the same as in the case of uid crystals, Smectic liquid crystal, Ferroelectric
thermotropic liquid crystals. However, the macro- liquid crystal, Nematic liquid crystal.
molecules or micelles can act as structural units
of lyotropic liquid crystals, which renders some LIQUID CRYSTALS, MAGNETOOPTICAL
specifics to their ordering. For instance, an ana- EFFECTS
log of smectic phases is lamellar phases, where See Magnetooptical effects in liquid crystals.
the lipid structural units form, within the solvent
(water), double layers similar to biological mem- LIQUID–GLASS TRANSITION
branes. This similarity establishes the connection See Vitri ca tion.
722 LIQUID METALS
second approach deals with statistical relation- In a free form, lithium is a plastic, very soft
ships for the model of a dense gas, and lacks much silvery-white metal. It has a body-centered cu-
substantiation. bic lattice with parameter a = 0.35023 nm (at
Problems of liquid state models that remain to 20 ◦ C), at 78 K it transforms to a hexagonal lat-
be solved concern the structure of microscopic re- tice. Density is 0.539 g/cm3 (the lowest among all
gions, their lifetime, nature of boundaries, and the metals) at 273 K. T melting = 453.7 K, T boiling =
formation mechanism for the regions (melting). 1600 K. Heat of melting is 3.0 kJ/mole; heat
There are two points of view regarding the struc- of evaporation is 133.7 kJ/mole; specific heat
ture of microscopic regions. One hypothesis as- cp = 24.85 J·mole−1 ·K−1 , coefficient of ther-
serts that the liquid consists of crystal “fragments” mal conductivity is 71 W·m−1 ·K−1 (from 273 to
(microcrystallites) that retain a crystal structure. 373 K). Debye temperature is 370 K. Resistivity
According to the second concept, a liquid is com- is 0.0812 μ·m (at 273 K), the average coefficient
posed of cybotaxes (three-dimensional arrange-
of resistance is 4.5·10−3 K−1 . Mohs hardness is
ments of molecules) that are intermediate between
0.6, Brinell hardness is 5 MPa; tensile strength
crystalline and gaseous states in their structural as-
is 115 MPa; relative elongation is 50 to 70%.
pect.
Adiabatic elastic moduli of lithium: c11 = 13.42,
There are no boundaries between microscopic
c12 = 11.30; c44 = 8.89 GPa at 298 K.
regions in the cybotaxic model. In the micro-
crystallite model, crystal-like regions must have Metallic lithium is paramagnetic and its com-
boundaries, but there is scarcely any evidence for pounds are diamagnetic; magnetic susceptibility
their presence. It is clear that much of the heat is +2.04·10−9 (at 293 K). It is used as an alloy-
of melting in this case goes into the generation ing addition to various alloys. The most impor-
of these boundaries. Frenkel thoroughly examined tant area of lithium application is nuclear power.
plausible reasons for the generation of boundaries Lithium compounds are used in manufacturing
and found them reducible to density uctuations, special glasses, thermostable porcelain and ceram-
generation of vacancies, and systems of micro- ics, etc.
cavities and microcracks (see Cracks). Yet, the
monotonic increase of the number of crystal de- LITHOGRAPHY (fr. Gr. λιθoς , stone, and
fects with increasing temperature is inconsistent γραϕω, am writing)
with the abrupt change of viscosity and density Historically, a printing technique that uses pat-
at melting. Microcrystallite boundaries are likely terns etched in stone. Modern lithography tech-
to arise as a result of the anomalous thermal ex- niques are used to etch systems of conducting
pansion of a crystal surface (and at locations of and insulating regions on the surfaces of semi-
defects). There, the lattice parameters increase conductor plates to form integrated circuits. The
more rapidly than they do inside the bulk of the surface of the plate is coated with a uniform layer
crystal, and shear resistance vanishes at the melt- (or several layers) of a resist sensitive to vari-
ing temperature. The crystal turns into a liquid ous types of radiation (accordingly, one speaks of
owing to the small value of the shear modulus X-ray photolithography, electron lithography and
along boundaries. ion lithography). The intensity and spacial distribu-
tion of the irradiation controls the configuration of
LITHIUM, Li
the protective mask remaining on the surface after
Element of Group I of the periodic system with
atomic number 3 and atomic mass 6.941; it is an processing the plate. It also controls the distribu-
alkali metal. Natural lithium is a mixture of stable tion of impurities in subsequent doping of the plate
isotopes 6 Li (7.42%) and 7 Li (92.58%). Electronic through the processed surface or the surface relief
configuration is 1s 2 2s 1 . Successive ionization en- resulting from etching. The mask then has to be
ergies are 5.39, 75.64, 122.42 eV. Electron affinity removed.
energy is 0.59 eV. Atomic radius is 0.157 nm;
ionic radius of Li+ is 0.068 nm. Oxidation state LITHOGRAPHY, ION
is +1, electronegativity is 0.95. See Ion lithography.
724 LITTLE–PARKS EFFECT
LITTLE–PARKS EFFECT (W. Little, R. Parks, They are employed after thin layers of the sample
1962) are subjected to chemical or electrochemical dis-
A periodic variation
T c in the critical tem- solution, and the components to be identified are
perature, T c , of a thin walled superconducting concentrated.
cylinder or thin film wire loop that appears during
variations of a magnetic field B applied parallel LOCAL BARRIERS for dislocations
to the cylinder axis. The conditions on the wall Obstacles (stoppers) to interactions which re-
thickness d for observing the Little–Parks effect sult in a haphazard quality of displacements, and
are d R, ξ(T ), λ(T ), where R is the cylinder slow down the velocity of dislocation motion.
radius; ξ(T ) is the coherence length, and λ(T ) is Local barriers are typical of materials with low
the penetration depth of magnetic eld into the su- Peierls barriers (see Peierls relief), especially for
perconductor. The Little–Parks effect results from alkali-halide crystals, and they also have an ef-
the quantization of ux. The value
T c (B)/T c fect upon the mobility of dislocations in crys-
is proportional to (n − Φ/Φ0 )2 , where Φ is the tals with high Peierls barriers. The local barri-
magnetic ux through the cylinder cross-section, ers usually include vacancy–impurity complexes
Φ0 = h/2e is the fluxoid or ux quantum, and n is (see Vacancy, Impurity atoms), small-size impurity
an integer. For a given B, there is a value of n that and vacancy clusters, peaks of internal uncom-
minimizes the super uid velocity vs . According to pensated stress fields (see Internal stresses) of
dislocations and sometimes of vacancies, intersti-
estimates, max{
T c (B)} ≈ 10−3 K.
tial atoms, impurity atoms, precipitates of foreign
phases, and microscopic dislocation loops. The
LOADING, MICROSHOCK
nature of local barriers has been studied most thor-
See Microshock loading.
oughly in silicon. The so-called A and B swirl
defects were found there. According to electron
LOCAL ANALYSIS
microscopy data, B-defects form during crystal
Procedures for determining the chemical com- growth, and are agglomerates of implanted sili-
position of microvolumes or thin layers of a solid. con atoms with an admixture of other elements,
They are usually applied to retrieve the elemental e.g., carbon, which regulate the growth kinetics
composition of microphases, of individual inclu- and stability. A-defects form by way of restruc-
sions, of micropatches or of thin layers of the solid turing the B-defects, and are dislocation loops of
surface. Both the basic components of the sam- an intrusion type. One may consider interactions
ple and impurities (10−2 –10−4 %) are identified. of dislocations with both mobile and immobile lo-
Various instrumental techniques are used that fea- cal barriers. The immobility of local barriers stems
ture low detection thresholds (10−11 –10−14 g) from their large size and the low temperature of the
and allow carrying out a multicomponent analysis. experiment, when the diffusional motion of atoms
These techniques are characterized by a certain lo- is hindered (see Diffusion). Local barriers related
cality L (i.e. size of area or volume within which to peaks of internal dislocation stresses are immo-
the analysis may be conducted with a given ac- bile in principle. Immobile local barriers usually
curacy). One may distinguish longitudinal locality have random positions throughout the crystal bulk,
L (depth of analyzed region), and transverse lo- although some ordering can occur. Dislocations
cality L⊥ (measured along analyzed surface). penetrate local barriers in either an activated or
Highly sensitive local analysis techniques nonactivated manner during a finite time. The ac-
(L < 1 μm3 ) are the X-ray spectrum, laser, cath- tual physical mechanisms of penetration depend
ode luminescence, ionic probing microanalysis, on the barrier nature, and in many cases have
Auger spectroscopy, characteristic energy loss not yet been studied. Spatially large local bar-
spectroscopy, etc. Layer-by-layer local analysis is riers (e.g., dislocation fields of internal stresses)
conducted at various depths with a layer step of are often avoided by being by-passed (Orowan
about 0.1 μm. Beside the above techniques, vari- mechanism). According to electron microscopy
ous others are also applicable, such as atomic ab- data, small local coagulation barriers may disin-
sorption, luminescence, radiation activation, etc. tegrate under the forceful action of dislocations.
LOCAL ELECTRIC FIELD 725
The mathematical formalism of local barrier pen- is far broader. A drawback of this approximation
etration is based on geometrical statistics, i.e. on is the irregular (nonzero) limit for the electron–
statistical characteristics of a dislocation line con- electron interaction energy when the number of
figuration. They include probability distributions electrons converges to one, and also the exponen-
of the lengths of dislocation segments, and of the tial, rather than power law, fall off of the effective
angles of coverage of individual local barriers by potential at large distances from the system (the
a dislocation. Simultaneous interactions of a dislo- so-called problem of image forces). This approx-
cation with a large number of randomly positioned imation often gives good results for ground state
local barriers introduce statistical features to its properties like cohesive energies and charge densi-
motion. ties of valence electrons, but the results are poorer
for excitation energies.
LOCAL CORROSION
The corrosion of spatially localized regions of LOCAL ELECTRIC FIELD
a material. Corrosive failure may, depending on the The internal field produced by a sample polar-
properties of the metallic material and the level of ization that affects an atom (or ion). Calculating
hostility of the environment, localize at separate the local field (E loc ) is an important problem
in the theory of insulators, in particular of ferro-
regions of the surface, with other areas remaining
electrics, and also the theory of magnetism. This
practically intact. The characteristic types of local
electric field arises from fixed charges outside the
corrosion are contact, crevice, pitting and inter-
body (external electric field, E0 ) and the sum of all
crystalline. Contact corrosion appears in a device
fields of charges within the body itself, taking into
constructed from alloys differing in their electro-
account the sample polarization (see Polarization
chemical properties. It may also be observed in a
of insulator). The local field acting upon an atom
single homogeneous material containing welds or
at a point r0 differs significantly from the macro-
areas of different structural states (stressed, tem-
scopic electric field, E, the latter being derived
pered, etc.). Crevice corrosion involves an intense
from the values of electric fields averaged over
selective destruction in a crevice (slit) at the junc- the volume of a unit cell. For instance, consider
ture of different parts of a construction, between the center of symmetry point in a cubic environ-
gaskets and metal, etc. During pitting (point) cor- ment assuming that the crystal itself is spherical
rosion, some metal spots dissolve at high rates, in shape. We then have E = E0 − P /(3ε0 ) (the
deeply enough to form penetrating sores. Inter- negative term represents the depolarization arising
crystalline corrosion (the most dangerous of all) from the crystal shape). Meanwhile, E loc = E0
develops selectively along the grain boundaries owing to the mutual compensation of dipole fields
and results in the loss of strength and plasticity of in the sample, so that E loc = E + P /(3ε0 ). Since
the construction material. P = (ε − 1)ε0 E, we have E loc
E for certain
crystals with large relative dielectric constants ε
LOCAL DENSITY APPROXIMATION (e.g., ferroelectrics). For a lattice point with an
An approximate method for calculating the environment other than cubic, the local field is
exchange-correlation energy Exc of an inhomoge- E loc = E + γ P /(3ε0 ), where γ is the Lorentz fac-
neous electron gas, assuming that it depends on the tor (see Lorentz eld). Such a field may strongly
local charge density ρ(r). More specifically, the differ both in value and sign from site to site in
value of Exc is expressed through the exchange- a crystal. The Lorentz model used to derive the
correlation energy density εxc of a homogeneous above expressions for E loc fails to account for the
gas at the given point: Exc = drρ(r)εxc (ρ(r)). field produced by the molecule under considera-
The local density approximation is at present the tion, i.e. for the self-action of the molecule on its
main one employed in the practical application environment (reactive field in the Onsager theory).
of density functional theory. Formally, the local Taking into account the reactive field causes the
density approximation is of zero order in the ex- formula for E loc to differ from the above. Both
pansion of Exc in the gradients of the electron models yield consistent results only when the elec-
density, but the overall area of its applicability tric dipole moments are induced by an external
726 LOCAL ELECTRONIC LEVELS
electric field. Note that the field acting upon a Electron microscopy techniques are commonly
free carrier always agrees with the macroscopic used to observe directly the localization of plastic
field, whereas E loc = E in the general case for strain at the first three levels. Using transmission
localized electrons (e.g., electrons associated with electron microscopy, nonuniformities in the distri-
impurities). That field depends on the radius of the bution of defects are commonly observed, while
localized electronic state. replicas and scanning electron microscopes make
it possible to analyze microrelief that develops
LOCAL ELECTRONIC LEVELS at the free surface of a target under plastic de-
Localized states in the energy spectrum of formation. The localization of the fourth class,
a semiconductor, which correspond with certain observable by an ordinary optical microscope,
discrete levels in the band gap and resonance usually appears as bands that vanish during polish-
(quasi-discrete) levels in allowed bands. As to ing and further etching, the so-called shear bands
their capability for either accepting an electron or persistent slip bands. The latter type is typ-
from the valence band or releasing it into the ical of nonstationary regimes of loaded targets.
conduction band, the local electronic levels are The fifth category of localization, called geomet-
divided into acceptors and donors. Certain im- ric localization of plastic deformation, appears as
purities are amphoteric (see Amphoteric centers) patches of more pronounced changes of shape ob-
and may behave either as donors and acceptors. servable by the naked eye. Such patches in the case
The local electronic levels in crystalline semicon- of stretching are usually called necks. If these de-
ductors are generated by atoms at substitutional formations develop, substances may deform over
or interstitial positions in the lattice as well as great distances without significant strain local-
by complexes comprising several impurity atoms, ization. Superplasticity is often observed in such
and by formations involving other lattice defects substances. When plastic strain localizes as a re-
(vacancies, stacking faults) beside impurities. Im- sult of changes in material structure or loading
purity levels are divided into shallow levels and regime, or when a cellular dislocation structure
deep levels depending on their positions relative appears (see Fragmentation), there is usually an
to the edges of the band gap, and of the properties accompanying loss of superplasticity. The most
of the impurity potential. dangerous case is the fifth degree of localization
since the appearance of a neck sharply lowers the
LOCALIZATION, ANDERSON
level of deformation which leads to failure. The
cause of the localization lies in the spatial hetero-
See Anderson localization.
geneity of crystalline materials, which results not
only from the presence of defects and impurities,
LOCALIZATION OF PLASTIC DEFORMATION
but is also closely connected with the crystalline
Result of a nonuniform distribution of plas- structure of a perfect crystal.
tic shape changes within the bulk of a deformed
material. Localization of plastic deformation ac-
LOCALIZATION, SCALING THEORY
companies plastic deformation of most crystalline
bodies, and reflects the deformation behavior of See Scaling theory of localization.
the regions with highest plasticity. Different scales
of localization of plastic deformation are found, LOCALIZATION, SELF-
characterized by the corresponding linear dimen- See Self-localization.
sions:
(1) submicroscopic (10−9 –5·10−9 m); LOCALIZATION THRESHOLD
(2) microscopic (5·10−9 –5·10−8 m); The same as Mobility edge.
(3) mesoscopic (5·10−8 –3·10−7 m);
(4) quasi-macroscopic (optical) (3·10−7 –10−5 m); LOCALIZATION, WEAK
(5) macroscopic (above 10−5 m). See Weak localization.
LOMER–COTTRELL BARRIER 727
Lomer–Cottrell barrier Aδ + δα + αD, the respec- functions. The coefficients ak are called long-
tive stair-rod dislocation δα having its Burgers range order parameters, or simply order parame-
vector b = (a/6)[101]. Since the Lomer–Cottrell ters. The structural order is usually determined
barrier cannot slip in either of the planes P1 or P2 , by the function RΦk a (r) of the lowest symme-
it blocks both planes. The formation of this barrier try. The values of the order parameters ak are
is an efficient means of strain hardening of crystals measures of ordering. Among the structures with
for the case of multiple slips. the same symmetry properties the one with the
larger order parameter is more ordered. In the
LONDON PENETRATION DEPTH case of the same composition but different sym-
See Penetration depth of magnetic eld. metries the structural states of the solid are called
phases. In addition to long-range order parame-
LONG-RANGE AND SHORT-RANGE ORDER ters, sometimes short-range order parameters are
Order and symmetries (or correlations) in the introduced which reflect the measure of correla-
spatial location of the elements of a solid struc- tions over short distances (usually within the first
ture (atoms, molecules, atomic groups of particu- coordination sphere). The order parameters are
lar configurations, etc.) or the space correlations usually determined by atom-to-atom interactions,
of values of physical parameters (e.g., magne- and depend on temperature (T ), pressure (P ),
tization M, polarization P ). The pair correla- specific volume (V ) and external forces. A jump-
tion function in coordinate space RΦ (r) = Φ(x), wise changing of the order parameters (or their
Φ(x + r), where Φ(x) is a physical quantity derivatives) with variations of P , T , V or exter-
(atomic equilibrium position, local value of mag- nal forces is accompanied by jumpwise variations
netic moment, etc.) and the angular brackets mean of different physical values, and this is called a
averaged over the coordinates x, is often used for phase transition. The values of P , T , V or those of
describing the order. The degree of ordering is applied external forces are called critical parame-
characterized by the correlation radius r c , by the ters. At a rst-order phase transition a jumpwise
symmetry properties of RΦ (r), and by the values variation of the order parameter occurs, and at a
of the order parameters. The correlation radius de- second-order phase transition some order parame-
termines the average size of the area where RΦ (r) ter equals zero at one side of the critical point,
is not negligibly small. If limr→∞ RΦ (r) = 0 differs from zero to the other side of it, and tends
then rc → ∞, and the structure has long-range toward zero at the critical point itself.
order with respect to the property Φ. If rc ∼ a Topological order is described by the correla-
(a is the average interatomic distance) then only tions in the mutual locations of atoms. If Φ(x)
short-range (or local) ordering exists. If rc has a is a function of the distribution of atoms, then
finite value, considerably exceeding a, then the RΦ (r) = ρ 2 + fρ (r), where ρ = Φ(x) is the
structure has intermediate order. If Φ(x) has some average density of the atoms, and the function
symmetry properties, then RΦ (r) inherits these fρ (r) = Φ(x) − ρ, Φ(x + r) − ρ describes
properties. The structure with lower symmetry the correlations of deviations from the average
is considered to be more ordered. The structure value. In liquid and amorphous solid phases the
of RΦ (r) with a single-axis of symmetry, e.g., second term tends toward zero as r → ∞, and
where there is invariance with respect to rotations its behavior at low r determines the short-range
around some space axis, is more ordered, than an and intermediate order. In the ordered crystalline
isotropic structure (RΦ (r) which depends only on phase fρ (r) = 0 at r → ∞, and its symmetry is
|r|). The part of RΦ (r) which does not disappear determined by the lattice symmetry. The varia-
at r → ∞ (let us designate it as RΦ a (r)) may be tion of this symmetry is associated with a phase
the superposition of functions which differ by their transition. For the analysis of order in solids the
symmetry types, or belong to the same symmetry isotropic part of RΦ (r), or the radial atomic distri-
types but with different parameters of the sym- bution function, is often used.
metry group RΦ a (r) = a R a (r) + a R a (r) + Compositional order is determined by the cor-
1 Φ1 2 Φ2
· · · , where RΦk (k = 1, 2, . . .) are normalized
a relations of the mutual ordering of atoms of dif-
LONG-TERM HARDNESS 729
ferent types in polycomponent solids. The varia- there exists long-range magnetic order. For uni-
tions in compositional order may occur without form magnetization RM a = M(x)2 = S 2 in pe-
variations or with negligibly small variations of riodic modulated magnetic structures RM a (r) =
topological order, e.g., in ordered alloys (see Alloy S 2 + ak2 fk (kr), where fk is a periodic function
ordering). Let us consider the alloy Cu–Zn with with period 2π . The quantities S and ak are order
equal concentrations of copper and zinc atoms, parameters, and k is a parameter of the translation
with a simple cubic lattice and therefore topolog- group. At S = ak = 0 the body is magnetically
ically ordered at temperatures below the melting disordered. The temperature at which S becomes
temperature. At a temperature close to zero the different from zero is called the Curie point. Mag-
alloy is compositionally ordered: each atom of netic ordering can occur in topologically ordered
copper is surrounded by six atoms of zinc and each crystalline phases, as well as in amorphous mate-
atom of zinc is surrounded by six atoms of cop- rials which lack topological order.
per, so there is short-range order. Besides this, The ordering of the polarization vector P (x) in
the atoms of copper and zinc individually form ferroelectrics is similar to the ordering of M(x) in
cubic sublattices of a doubled (compared to the magnetic substances. Orientational order is dis-
source lattice) period, i.e. there is also long-range played in the correlations of vectors which deter-
compositional order. It is possible to introduce pa- mine the orientations of structural elements (e.g.,
rameters to specify the long-range and short-range molecules) which form a solid. Orientational order
orderings. Let us designate by indices α and β the may be present together with topological disorder,
values related to the sublattices of copper and zinc as occurs in quasi-crystals. Besides this, topolog-
(respectively). Let us assume that wαCu , wαZn ical ordering of molecules may coexist with their
are the probabilities of locations of copper and orientational disordering (e.g., the orientationally
zinc atoms, respectively, on the α-sublattice, with disordered molecular crystal N2 ).
wβCu , wβZn those values for the β-sublattice. The
parameter for the long-range order η is determined LONG-TERM HARDNESS, long-term
in the following way: high-temperature hardness
An auxiliary method for estimating the po-
1 1 tential high-temperature strength and other high-
η = 2 wαCu − = 2 wβZn − .
2 2 temperature mechanical properties of a material
from the time dependence of the imprint size
It can be seen that for wαCu = 1 and wβZn = 1 during the measurement of high temperature hard-
we have η = 1 so all the points of the α-sublattice ness under permanent load at constant temper-
are occupied by the copper atoms, and the points ature. This depends on the surface state of the
of the β-sublattice are occupied by the zinc atoms. material. For measurements below 2300 K, the
This might be the state of maximum long range or- method of static indention is applied. An indenter
der at absolute zero temperature. With increasing in the form of the regular four-face pyramid with
of temperature there will be interchanges of atoms the angle 136◦ between opposite faces is made
of both types from one sublattice to the other, of sapphire (Al2 O3 ) for use up to 2000 K, and
and at temperatures above some value T c (the of boron carbide (B4 C) for higher temperatures,
critical point) wαCu = wαZn = wβCu = wβZn = where for the latter HV = 1.854P /b2 , where P is
1/2, so η becomes zero and the alloy becomes a the load at the indenter, and b is the arithmetic
compositionally disordered solid solution with the mean of the diagonals of imprint after release.
conservation of topological order. Above 2300 K the hot strength is determined by
Magnetic ordering. The magnetic moment of the method of one-sided flattening of a conical
the atom M(x) located at point x interacts with sample (cone angle 120◦ ) on the plane surface of
the magnetic moments of the other atoms, due the punch. Hc = 1.273P /d 2 , where d is the mean
to which magnetic ordering can occur. When the diameter of the imprint. The long-term hardness H
correlation radius of the magnetic moment tends was found to vary linearly with the long-term
to infinity and RM (r) = M(x), M(x + r) = 0 strength for a series of alloys. Three subsequent
730 LONG-TERM STRENGTH
tests of high temperature hardness, performed with results are approximated graphically by curvilin-
a factor of ten increase in exposure time at each ear or rectilinear (with turning points) diagrams
sequential test (0.5, 5 and 50 min), provided in- in the coordinates “stress versus logarithm of time
dention curves or curves of hardness versus time. prior to fracture”, and they are then extrapolated
If H = ar n , where a and n are the material con- using temperature–time techniques from the the-
stants dependent on its nature, n being the rate ory of plasticity to the longevity corresponding to
index, then one can compare the characteristics of the working conditions (105 h and more). During
long-term strength using the data on the long-term the process of creep that precedes fracture one can
hardness. observe a significant change in the structure, the
phase composition, and the dislocation substruc-
LONG-TERM STRENGTH
ture of a metal. This makes it difficult to construct
The material strength sustained under applica-
physically justifiable equations for the conditions
tion of permanent, long-term stress. As a charac-
of long-term loading and elevated temperatures.
teristic of long-term working ability of, mainly,
Therefore, the predictions of long-term strength
metallic materials, this is particularly important in
made on the basis of shorter duration tests remain
thermal power engineering, gas-turbine construc-
very approximate, and can be misleading.
tion, materials for aircraft and spacecraft engi-
neering, and construction (concrete). A principal
working characteristic of materials for long-term LONSDALITE (after English crystal-chemist
service life is their ultimate long-term strength σlT Dame Kathleen Lonsdale)
(here T means the temperature, l is longevity) An allotropic form of carbon which was syn-
which provides a numeric value of stress that pro- thesized in 1965 and found in an iron meteorite
duces a fracture at working temperature within a in 1967. Density of lonsdalite is 3.51 g/cm3 .
given time. There is no theoretical limit of long- A framework of tetrahedrically coordinated car-
term strength. Its characteristics are determined bon atoms bound to each other by sp3 -hybrid
from laboratory tests of standard samples under orbitals may be obtained in two ways. One of
the conditions of single-axis tension by special test these yields a diamond-like cubic lattice where
machines which maintain constant load and tem- the packing order of atomic layers in the direction
perature during the test (104 h and more). The test 111 follows the sequence ABCABC (Fig. 1(a)).
Fig. 1. Projection of atomic grids of the cubic diamond (a) and lonsdalite (b) lattices.
LORENTZ FORCE 731
Fig. 2. Unit cells of the cubic diamond (a) and lonsdalite (b) lattices.
By changing the sequence of atomic layer pack- constants is an artificial mathematical tool which
ing to ABAB (Fig. 1(b)), the cubic structure is enables one to calculate the fields from the dipoles
transformed to a “hexagonal” wurtzite-like mod- present inside the cavity, and the fields that result
ification. Concerning its interatomic bonds and from the remainder of the sample by integrat-
their spatial arrangements lonsdalite is close to the ing over the effective surface charge density. The
cubic modification of diamond (Fig. 2(a)), but in Lorentz field E L turns out to equal to P /(3ε0 )
the cubic modification all layers built of tetrahedra for the simple case when the neighbors around
are oriented in the same direction, while in lons- a given point in the lattice have cubic symmetry
dalite each successive tetrahedric layer is turned (see Cubic system). Here P is the polarization
by 60◦ relative to the previous one. The lattice of or dipole moment per unit volume of the crys-
lonsdalite may be thought of as a combination of tal. Then the local electric eld acting upon the
two hexagonal close-packed sublattices displaced atom is E loc = E + P /(3ε0 ) (Lorentz formula,
relative to one another by 3/8 of the lattice para- where E is the average macroscopic field). In view
meter c (Fig. 2(b)). of the relation P = (ε − 1)ε0 E between P and
Lonsdalite, with symmetry space group E, we find that E L = (ε − 1)E/3, and E loc =
4 ), crystal lattice constants a =
P 63 /mmc (D6h (ε + 2)E/3. For highly polarizable crystals
2.51 Å, c = 4.12 Å, is unobtainable in pure form. (ε
1), E L
E and E loc ≈ E L . For crystal
Its content in natural impact diamonds is as high field symmetries lower than cubic the Lorentz for-
as 50%, in compression-synthesized ones it is up mula is more complicated: E loc = E + γ P /(3ε0 ),
to 50%, and in those formed by static the com- where γ is the Lorentz factor, with sign and mag-
pression of graphite it is up to 90%. Lonsdalite nitude depending on the lattice angle. The Lorentz
transforms to graphite at far lower temperatures field plays an important role in explaining the na-
than diamond; hence, to obtain lonsdalite through ture of ferroelectric phase transitions. Thus, the
synthesis under dynamic conditions, effective heat interactions of ions from different cells that seek to
withdrawal is necessary. displace the atoms in a single direction reduce to
the difference between the total field acting upon
LORENTZ FIELD (H.A. Lorentz, 1934) an ion and the macroscopic field averaged over the
An electric field that acts upon an atom of an cell in the Lorentz field potential.
insulator, and is derived from polarization charges
at the surface of a fictitious spherical cavity cen- LORENTZ FORCE
tered at the atom under consideration. Introducing The force F acting upon a particle of charge e
such a cavity with a radius of about 10 lattice that moves at the velocity v in an electromagnetic
732 LORENTZIAN SHAPE
sis of the geometry of LEED pictures allows a prevailing type of radiation-induced damage be-
simple determination of the symmetry and the pe- low a certain value of the energy transferred at
riods of the surface lattice. To obtain the chemical collisions of ions with target atoms. In the case
composition of the unit cell and of the coordi- of crystalline targets, the decrease of the bom-
nates of its atoms in the general case requires the barding ion energy lowers the probability of the
analysis not only of the positions but also of the ions to enter lattice channels, and increases the
intensities of the diffraction beams. This may be critical angles for channeling. There is interest in
performed only on the basis of the dynamic the- low-energy ion doping because of the need to pro-
ory of LEED because, due to the high probability duce submicron elements with gaps between them
of elastic scattering of low-energy electrons and in modern chips, and also in connection with the
the effects of multiple scattering, the usual kine- development of molecular beam epitaxy. In some
matic theory of diffraction which neglects these cases (e.g., manufacture of low-cost solar photo-
effects is practically inapplicable. The measured converters, or doping of metals and amorphous
dependences of the diffracted beams on the elec- semiconductors), low-energy ion implantation can
tron energy is compared with those calculated for be carried out with a gas discharge, or with the
various hypothetical positions of atoms in the sur- help of the simplest high-current ion guns without
face lattice, while the generation of the theoretical mass separation, which makes the technology eas-
curves requires time-consuming calculations that ier and cheaper. Low-energy ion bombardment of
can be carried out only for the simplest cases. heated targets causes an accelerated impurity diffu-
In spite of these difficulties, the LEED method sion which can be used as an independent method
finds its wide application in studies of the physics of ion doping (ion injection).
and chemistry of surfaces. With its help the phe-
nomenon of surface reconstruction has been de- LOWER CRITICAL FIELD, rst critical eld
veloped, namely the formation at the surface of a The temperature dependent maximum value
lattice, that differs from the one in the bulk. Also of the external magnetic field strength, at which
surface phase transitions are successfully investi- the full Meissner effect may still be observed in
gated (at clean surfaces, and in adsorbed layers), type II superconductors. The value of the lower
as well as the dynamics of thermal vibrations of critical field (Bc1 ) at absolute zero (T = 0 K)
surface atoms, the processes of formation of thin is given in the BCS theory by the expression
lm s, and other surface phenomena. The discov- Bc1 = [Φ0 ln(λ/ξ )]/(4πλ2 ), where Φ0 is the u x
ery of LEED by C.J. Davisson and L.H. Germer quantum, λ is the penetration depth of magnetic
(1927) was the first experimental proof of the ex- e ld into the superconductor, and ξ is the coher-
istence of the wave properties of microparticles. ence length. Since λ > ξ in type II supercon-
ductors, Bc1 is smaller than the √ thermodynamic
LOW-ENERGY IONS critical magnetic eld Bc = Φ0 /(2 2πλξ ), which
Ions with energies 104 eV used in the ion in turn is smaller than the upper critical eld Bc2 =
implantation doping method for modifying the Φ0 /(2πξ 2 ).
properties of thin subsurface layers to produce
contacts, superfine p–n junctions (in radiation de- LOW-FREQUENCY ENHANCED MAGNETIC
tectors, solar photoconverters, etc.), elements of RESONANCE ABSORPTION
chips, conducting and insulating coatings, and A combined technique of resonance and non-
control for the height of Schottky barriers. The resonance absorption. When a paramagnetic res-
low-energy ions possess quite short mean free onance is saturated with a high-frequency mag-
paths (100–10 nm or less) and, accordingly, small netic field B1 (while energy is simultaneously
statistical dispersion of the paths. This is of impor- absorbed from a low-frequency magnetic field par-
tance when producing miniature integrated circuits allel to the external static magnetic eld B0 ), there
with high packing density by ion doping. In view can develop, due to a change in the spin–spin
of the reduced probability of formation of a so- reservoir temperature of the paramagnetic sys-
called disordered region, point defects are the tem (see Dipole–dipole reservoir), quite a strong
734 LOW-SPIN COMPLEXES
ing with TLS. The features of the atomic structure LOW-TEMPERATURE LUBRICANT
responsible for various low temperature anom- A layer of material spread over the friction
alies in amorphous materials remain mostly ob- surfaces of moving joints in machines and mecha-
scure. The free volume model relates the presence nisms operating at low temperatures (below
of TLS to the appearance of regions containing 120 K). It has low shear resistance at low tempera-
free volume, while polycluster amorphous solids tures, decreases the friction force and rate of wear
may have cluster boundaries with unmatched cross of joined surfaces within the full operating tem-
links partially occupied by atoms. perature range (from room to low), and withstands
multiple cooling and heating. Widely used are:
films of soft metals, mostly those having a face-
LOW-TEMPERATURE DEFORMATION centered cubic lattice (gold, silver, nickel, etc.);
Deformation at temperatures below the Debye solid lubricating composite coatings with a hexag-
temperature (usually at 77, 20.4, 4.2 K), carried onal close-packed lattice (graphite, disulfides and
out under conditions of suppression of dynamic diselenides of Mo, W, Nb. etc., and others), in
recovery, involving an appreciable accumulation combination with polymer binders; coatings pro-
of crystal lattice defects, and a build-up of the duced by chemical heat treatment of the sur-
latent energy of strain. Low-temperature deforma- face (sulfurizing, selenizing, etc.); metal–polymer
tion is a source of information on the mechani- coatings (porous metal layer saturated with flu-
cal properties of solids at low temperatures (see oroplastic or other polymers); magnetic powder
Low-temperature physics), and a powerful tool materials (mixtures of antifriction and magnetic
used to influence the defect structure and physical powders), etc.
properties of samples (including high-temperature
strength). With temperatures decreasing and ap- LOW-TEMPERATURE LUBRICATION
proaching absolute zero, the mechanical proper- Techniques for transferring lubricants to low-
ties of certain metals (Al, Cu, etc.) improve, while temperature friction joints in order to prevent grip-
other metals (Fe, Mo, etc.) become brittle (see Brit- ping and to ensure reliable operation of machines
tle failure). Besides, such features as mechanical and mechanisms at low temperatures. Widely used
twinning, stepwise flow, interaction of disloca- are the preliminary application of a lubricant to a
tions with the electron system, all become more friction surface; offset-duplicator feeding of lubri-
prominent. Finally, the speed of dislocation mo- cant during the operation proper; manufacturing
tion decreases, plastic deformation becomes more elements of friction joints from composite mate-
and more controlled by quantum mechanical tun- rials containing a lubricant; bringing the lubricant
neling (see Tunneling phenomena in solids) and to and applying it at the friction surface by mag-
zero-point vibrations of dislocations, rather than netic forces.
by thermally activated processes. Related to low-
temperature deformation are the conditions for the LOW-TEMPERATURE PHYSICS
development of a high-strength crystalline state A branch of condensed matter physics involv-
with a strength 1/3 of its theoretical limit, attained ing the investigation of phenomena and properties
with the ability to withstand significant elastic of materials at low (cryogenic) temperatures. (The
strains (up to 1%). Certain phenomena are typical 13th Congress of the International Low Temper-
for low-temperature deformations: polymorphism, ature Institute (1971) recommended calling tem-
low-temperature recrystallization, depletion of dis- peratures below 120 K cryogenic.) Investigations
locations during creep, anomalies in the yield limit, of liquid and solid helium, of superconductivity,
dehardening during transition to the supercon- of cryocrystals and cryogenic liquids, of elec-
ducting state (see Superconductivity), slip along tronic properties of metals and semiconductors,
the boundaries of twins (see Twinning of crystals), of low-temperature magnetism etc. are consid-
etc. ered as low-temperature physics. The boundaries
736 LOW-TEMPERATURE PUMPING METHODS
anomalies were experimentally discovered in the relaxation of the center from its excited state,
late 1950s, and have been observed in numerous and emission of a quantum of light. As to the
metals, alloys, and ionic crystals at temperatures excitation type, one may identify photolumines-
below that of liquid nitrogen (T < 77 K). Detailed cence (excited by light); radioluminescence (ex-
studies of such anomalies in face-centered cubic cited by penetrating radiation, e.g., by an elec-
(FCC) and hexagonal close-packed (HCP) metals tron beam, see cathodoluminescence); electrolu-
and alloys have demonstrated that their appear- minescence (excited by electric field); tribolu-
ance and properties depend significantly on the minescence (excited under mechanical action);
concentration of impurities (see Impurity atoms) chemiluminescence (generated during chemical
and on the value of the initial strain. Also of reactions); acoustoluminescence (excited by an
importance is the state of the sample surface. acoustic wave).
Theoretical explanations of those anomalies dif- Of importance in solids is the second stage
fer significantly for various specific mechanisms of luminescence, when the excitation energy ab-
of plastic deformation. When the decisive factor sorbed by a luminophor is transferred (migrates to)
controlling the motion of dislocations is local bar- a luminescent center. Mechanisms of energy mi-
riers (FCC and HCP crystals, ionic crystals with a gration in which free energy carriers (electrons or
CsCl lattice), the appearance of the anomaly is ex- holes) take an active part are as follows: exciton
plainable by the role of thermal inertial properties mechanism, Auger mechanism involving bound
of dislocations (see Dislocation model of plastic current carriers (see Auger effect); intercenter tun-
deformation). This means that the importance of neling of immobilized carriers. Energy migration
thermal uctuations in those crystals decreases results in a time lag between the processes of ex-
with a drop in temperature, while the influence of citation and emission, which may be studied by
dynamic effects increases, and the sensitivity of measuring the kinetics using pulse excitation, and
deformations to structural inhomogeneities grows. monitoring the thermal and spectral emissions. In-
When the role of Peierls barriers is decisive (see cluded in the second stage of luminescence are
Peierls relief) (body-centered cubic metals and al- relaxation processes during emission from the ex-
loys, and ionic crystals with the NaCl lattice), the cited state. This involves the partial transformation
anomaly is explainable by quantum-mechanical of the excitation energy into thermal energy of
tunneling of dislocations and other quantum ef-
the crystal, i.e. emission in the form of phonons.
fects of dislocation motion. From a practical point
The relaxation process may be concurrent with the
of view, these anomalies are quite detrimental to
emission of an optical quantum of luminescent ra-
cryogenic technologies.
diation, so that hot luminescence is observed. In
many cases, the migration of energy to the lu-
LS -COUPLING
minescent center may only occur at high enough
See Russell–Saunders coupling.
temperatures. That is linked with the trapping of
LUBRICANT, LOW-TEMPERATURE free electrons (holes) by attachment centers, the
See Low-temperature lubricant and Low-tem- probability of their ionization being proportional
perature lubrication. to exp[−E a /(k B T )], where E a is the activation
energy. When the ionization temperature of attach-
LUMINESCENCE ment centers increases, it results in thermolumi-
Radiation that exceeds the thermal radiation of nescence. The phenomenon of energy migration
a medium and lasts significantly longer than the makes luminescence qualitatively distinct from
light frequency period. The phenomenon of lu- various kinds of light scattering and reflection,
minescence is typical of optical processes taking or the parametric transformation of the radiation
place in various states of matter. Luminescence in wavelength that is encountered in nonlinear optics
solids is observed in luminophors, materials which (see Parametric light generation). The third stage
contain luminescent centers. of luminescence consists in the act itself of light
The process of luminescence proceeds in three emission. Depending on the mechanism of the el-
successive stages: excitation, energy migration or ementary emission we can identify the following
LUMINESCENCE 739
flaws. Luminescent analysis combined with scan- 164 to 180, the most abundant in nature (97.40%)
ning electron microscopy offers wide capabilities being the stable isotope 175 Lu. Electronic con-
and diverse approaches to solving such problems. figuration of filling outer shells is 4f 14 5d 1 6s 2 .
Successive ionization energies are 5.426, 13.9,
LUMINESCENT CENTERS 20.96, 45.19 eV. Atomic radius is 0.1737 nm; ra-
Local structural elements in solids (impurity dius of Lu3+ ion is 0.0848 nm. Oxidation state
atoms, dislocations, etc.) which cause radiative is +3, and electronegativity is ≈1.28.
quantum transitions between their states to be ob- In the free form lutetium is a silvery-white
served as luminescence. When producing lumino- metal. It has a hexagonal close-packed crystal
phors, luminescent centers are intentionally gen- lattice with parameters a = 0.35052 nm, c =
erated in them using various techniques, e.g., by 0.55494 nm at room temperature, space group
P 63 /mmc (D6h 4 ). There is a polymorphic trans-
introducing particular selected impurities (lumi-
nescence activators). formation at ≈1820 K to a body-centered cubic
modification with parameter a = 0.390 nm upon
LUMINESCENT INDICATOR hardening, space group I m3m (Oh9 ). Density is
A material with luminescence sensitivity to 9.84 g/cm3 , T melting = 1935 K, T boiling is about
some physical or chemical factor, so that one may 3685 K. Bonding energy is −4.4 eV/atom at
detect that factor and assess its characteristics (see 298 K. Heat of melting is 18.85 kJ/mole, heat
Luminescent analysis). The term is also applied to of boiling is 356.6 kJ/mole, heat of sublima-
a device for data presentation based on lumines- tion is 377 kJ/mole; specific heat cp = 26.56
cence (usually electroluminescence). J·mole−1 ·K−1 . Debye temperature is ≈185 K;
linear thermal expansion coefficient of lutetium
LUMINOPHORS (fr. Lat. lumen, light, and monocrystal is 19.0·10−6 K−1 along the 6 prin-
Gr. ϕoρoς , carrying) cipal axis and 5.2·10−6 K−1 perpendicular to this
Materials which produce luminescent radia- axis at 273 to 301 K. Adiabatic elastic moduli of
tion (see Luminescence), and are used in various lutetium monocrystal: c11 = 86.23, c12 = 32.0,
lighting, visualization, and X-ray recording de- c13 = 28.0, c33 = 80.86, c44 = 26.79 (in GPa)
vices, in indicators, in systems for transmission at 300 K; isothermal bulk modulus is 41.1 GPa,
and transformation of video information (oscillo- Young’s modulus is 84.3 GPa, shear modulus is
scopes, television, computer displays, etc.). Many 33.8 GPa, Poisson ratio is 0.233. Vickers hard-
luminophors are synthetic solid-state materials ness of annealed lutetium is 77 HV. Lutetium
which can be either inorganic or organic in com- is easily amenable to mechanical treatment. Ef-
position. In widest use are inorganic crystalline fective thermal neutron capture cross-section is
luminophors (phosphors), that are found in fluo- 108 barn. Electrical resistivity of lutetium poly-
rescent lamps, TV and oscilloscope cathode-ray crystal is ≈73 n·m at 298 K, thermal coefficient
tubes, electric luminescent panels, etc. Depend- of the electrical resistance is 0.0014 K−1 . Electron
ing on their specific purpose, luminophors have to work function of polycrystal is ≈3.0 eV. Lutetium
meet widely differing requirements in their radia- is a paramagnet within the overall temperature
tion color, afterglow time, sensitivity to this or that range because its atomic 4f -shell is completely
excitation method, operating temperature range, full; the magnetic susceptibility is 0.102·10−9 ;
etc. These requirements are met by appropriate se- nuclear magnetic moment of 175 Lu is 2.9 nuclear
lection of luminophor composition and methods of magnetons. Lutetium is a prospective material as
activation for particular luminescent centers. a getter (see Gettering).
valence band dispersion in diamond and in zinc- LYOPHILIC AND LYOPHOBIC BEHAVIOR (fr. Gr.
blende (ZnS) semiconductors. The band energy λυω, am dissolving, ϕιλoς , love, ϕoβoς , fear)
extremum in cubic semiconductors is attained A characteristic of the interaction of molecules
in their valence band at the center of the Bril- of a solid substance (or particles of a colloidal
louin zone. Bloch functions (see Bloch theorem) system) with a contacting liquid. Depending on
at that point transform as p-functions, their an- the sign of the interaction one may identify sub-
gular momentum being j = 1, or j = 3/2 (taking stances which are lyophilic (strongly attracting),
into account the spin–orbit interaction in the latter and lyophobic (repelling) suspensions in a liquid.
case). A matrix Hamiltonian describes the motion Materials displaying lyophilic properties with wa-
of holes in the vicinity of the zone center: ter are called hydrophilic, and those displaying
lyophobic behavior are called hydrophobic. Ma-
5
H(k) = − γ1 + γ2 k 2 I terials exhibiting such behavior in oil are called
2 oleophilic (lipophilic) and oleophobic (lipopho-
2 2 bic), respectively. Substances lyophilic with re-
+ 2γ2 Jx kx + Jy2 ky2 + Jz2 kz2
spect to one liquid may be lyophobic with respect
+ 4γ3 [Jx Jy ]kx ky + [Jy Jz ]ky kz to another. One may quantitatively characterize
the degree of lyophilic versus lyophobic behav-
+ [Jz Jx ]kz kx ior by the edge angle, or by the heat of swelling
(or heat of solution). Changes in lyophilic ver-
+ k(J · B) + q Jx3 Bx + Jy3 By + Jz3 Bz ,
sus lyophobic behavior (lyophilization or lyopho-
where the J are matrices that satisfy the angu- bization of substances, bodies, or surfaces) may
lar momentum commutation relation J × J = iJ , occur during chemical transformations or physico-
k is the operator of quasi-momentum of a hole in chemical interactions. For example, modi cation
the magnetic field B, while the constants define by adsorption, a process involving surface-active
the effective masses and the g-factor. The Lut- agents, makes it possible to intentionally switch
tinger Hamiltonian gives a good description of cy- between the lyophilic and the lyophobic states
clotron resonance, acceptor states (see Acceptor) during flotation (separation of fine particles of dif-
in p-Ge, p-Si, and intermetallides of the AIII BV ferent substances) and during the manufacture of
type. This is the first example of a Hamiltonian composite materials.
applicable to the method of effective masses and
written in the form of a sum of invariants. This LYOTROPIC LIQUID CRYSTALS
approach has proved fruitful for constructing the Crystals formed during the course of dissolv-
equations of motion in other semiconductors (see ing certain organic compounds in water or in non-
Group theory). polar liquids. Molecules in low molecular weight
lyotropic liquid crystals are amphiphilic, i.e. they
LYDDANE–SACHS–TELLER RELATION have a polar and a nonpolar part. Salts of long
(R. Lyddane, R. Sachs, E. Teller) chain acids, such as potassium laurate, exemplify
The relationship between the frequencies of the such molecules. A lyotropic liquid crystal struc-
long-wave (k → 0) transverse, ωTO , and longitu- tural unit is a micelle, a conglomerate of 102 –103
dinal, ωLO , optical phonons in cubic crystals: molecules. Micelles may be normal, with polar
parts of their molecules at the outside surface, or
2
ωLO ε0 inverted with the polar parts inside. High molec-
i
= ,
2
ωTO ε∞ ular weight compounds have structural units that
i i
may be relatively short rigid fragments of a chain,
where ε0 is the low-frequency (static) and ε∞ is e.g., a section of a DNA molecule about 300 nm
the high-frequency (optical) relative dielectric con- long. The rigidity of such a fragment results from
stant (permittivity) of a crystal. The product in the its interaction with the solvent.
above relation is taken over all the optically active In terms of structural organization, lyotropic
dipole oscillations. (as well as thermotropic) liquid crystals are divided
LYOTROPIC LIQUID CRYSTALS 743
into nematics, cholesterics, and smectics. The lat- properties of lyotropic liquid crystals are similar to
ter are lamellar phases. Low molecular weight ne- those found in thermotropic ones. However, long
matics are built of anisometric micelles (cylinders relaxation times (minutes to hours) are typical for
or disks). Cholesterics have twisted cylindrical or lyotropic liquid crystals owing to their consider-
discotic micelles due to the presence of chiral able viscosity, 1–103 Pa·s. The lyotropic liquid
molecules (see Chirality). Lamellar phases often crystal type structure is an attribute of many sys-
feature double layers, with molecules of the sol- tems of biological origin.
vent located between them. Textures and physical
This Page Intentionally Left Blank
Mm
MACHINING, ELECTRIC PULSE ways to determine macrohardness are as follows:
See Electric pulse machining. Meyer hardness (E. Meyer), or contact pressure
H = P /A, where P is the load on the indenter,
MACROCYCLIC COMPOUNDS A is the projected area of the indentation; Brinell
Compounds that contain cyclic arrangements hardness (J.A. Brinell) is determined when pen-
of nine or more atoms in their molecules. Of most etration is carried out with a steel ball; Vickers
importance are macroheterocyclic compounds hardness is measured using a regular tetrahedral
whose cycles includes heteroatoms (atoms of ni- diamond pyramid with an angle of 136◦ between
trogen, oxygen, sulfur, etc.) besides carbon opposite faces. In both latter cases, H = P /A.
atoms. Typical of macrocyclic compounds is the Rockwell hardness (S.P. Rockwell) is determined
inclusion of neutral or charged particles into the using a diamond cone by the depth of plastic
macro-ring cavity, with the resulting formation impression at a given indenter load. Hardness
of complexes, compounds of the “host–guest” of minerals is often determined with the help
type (see Guest–host effect), clathrate compounds. of Mohs’ scratch hardness test (F. Mohs). Mohs
According to the type of donor atoms incor- hardness scale contains ten standard hardness val-
porated into the macro-ring, macrocyclic com- ues, with diamond the hardest (10) and talc the
pounds may be classified under oxamacrocycles lowest value (1).
(crown ethers, O), azamacrocycles (N), thiamacro-
cycles (S), etc. Inclusion of donor atoms of more
MACROLATTICE, imperfect superlattice
than one type is possible (e.g., azathiamacro-
A regular arrangement of the particles of co-
cycles). Crown ethers form strong complexes
herent precipitate phases in the bulk of an aged
with alkali and alkaline-earth ions, as well as
alloy crystal, which forms a modulated structure
compounds of the “host–guest” type. Aza- and
thiamacrocyclic compounds bind ions of transi- resulting from a diffusive phase transition (alloy
aging, alloy ordering, an eutectoid alloy decom-
tion metals, i.e. Pb, Hg, Zn. Macrocyclic com-
plexes of transition elements can form sandwich position). Dimensions of the macrolattice cells
compounds exhibiting metallic conduction. Since amount to 0.5–5 nm, while the orientation of their
many macrocyclic compounds favor the solid– axes depends upon the matrix of anisotropy of
liquid phase transition of ionic compounds, they elasticity; for cubic crystals, it may be either 100
serve as phase-transfer catalysts. Macroscopic com- or 111. Unlike a superlattice, the macrolattice is
pounds include some naturally occurring sub- imperfect: there is no strict periodicity in the posi-
stances, e.g., porphyrins, certain antibiotics, and tions of particles along fixed directions.
cyclodextrins. Depending on the difference in the crystal
structure matrix and the phase separation, differ-
MACROHARDNESS ent types of macrolattices may develop. Under
Hardness measured when an indenter carries an isomorphic (see Isomorphism) or an isostruc-
an appreciable load (10 N or more). Macrohard- tural phase transformation in a high-temperature
ness H characterizes the resistance of a mater- superalloy (see High-temperature alloys) or a hard
ial to plastic elastic strain (see Elasticity, Plas- magnetic alloy of the cunife, alnico, etc., type,
tic deformation) under local loading. The main it is generally a primary primitive cubic lattice
745
746 MACROSCOPIC DEFECTS
that later turns into a tetragonal one. If the trans- proper selection of physical quantities (scalar, vec-
formation is not isomorphic (or isostructural) (in tor, tensor) to describe the phenomena, and from
high-carbon and high-nitrogen steels, in beryllium expressions of mathematical relations between the
bronze, in hard magnetic alloys Co–Pt, etc.), it said quantities. A macroscopic description is ex-
is a tetragonal or rhombic lattice. In the latter emplified by Ohm’s law (G. Ohm, 1826) which
case, a many-step twinning (see Twinning struc- provides a linear relationship between the current
ture) of the macrolattice fragments of different density j in a conductor and the electric field in-
orientations springs up, which leads to relaxation tensity E. Both quantities are vectors. This linear
of elastic distortions. relationship is formulated through the application
of the electric conductivity tensor σ (see Electrical
MACROSCOPIC DEFECTS ρ =
conductivity) or resistivity tensor ( σ −1 ):
Defects in crystals that have a characteristic
ji = σik Ek , Ei = ρik jk (i, k = x, y, z). (1)
linear size greater than the interatomic distance
in all dimensions. Macroscopic defects have ex- These second-rank tensors σik and ρik provide the
tra energy both in the form of surface energy, macroscopic description (phenomenological de-
and from the field of the stresses that they induce. scription) of conduction phenomena.
Macroscopic surface defects involve details of sur- Invoking geometrical considerations and gen-
face relief, including the shape of scratches, and eral laws of nature often allows one to reach
elements of statistical surface roughness. These important conclusions about the components of a
are smoothed by the molecular mechanism of macroscopic description. Thus, for a body in the
mass transport. Three-dimensional macroscopic state of equilibrium, Onsager–Casimir relations
defects include pores and cracks. Pores may arise based on time reversal symmetry considerations
in a crystalline body during its growth from the require the conductivity and resistivity tensors to
melt as a result of negative jumps of volume at be symmetric (σik = σki , ρik = ρki ), whereas
crystallization, due to the interdiffusion that oc- crystal symmetry limits the number of indepen-
curs through differing partial diffusion coef cients dent components of the tensors σik and ρik . In a
of component atoms (see Diffusion porosity), dur- cubic crystal, e.g., the tensors σik and ρik reduce
ing sintering of powders, and under the effect of to scalars (σik = σ δik , ρik = ρδik , ρ = σ −1 ). In a
nuclear radiation. Cracks are formed during brit- hexagonal crystal, σik and ρik each have two inde-
tle failure of a solid. An estimate of the critical pendent principal values (one along the symmetry
size l ∗ of a crack capable of further growth is axis and the other in the plane, perpendicular to
obtained by equating the elastic energy released this axis). In accordance with the second law of
during crack formation (∼σ 2 l 3 /2E, where E is thermodynamics the principal values of the con-
the elastic modulus) to the crack surface energy ductivity and resistivity tensors (see Principal axes
(∼l 2 α, where α is the surface energy density): of a tensor) must be positive. The analytical re-
l ∗ ≈ 2αE/σ 2 . Macroscopic defects have a con- sults obtained from the application of general laws
siderable effect on many physical features and of nature to constituent quantities are inter-related
application properties of solids. through Kramers–Kronig relations which link to-
gether real and imaginary parts of generalized
MACROSCOPIC DESCRIPTION in solid state susceptibilities through relationships involving in-
physics tegrals over the frequency.
Summarizing statements of solid state exper- Included in a macroscopic description is the
imental results, often involving averaging proce- classification of solids by their characteristic fea-
dures, and sometimes including a phenomenolog- tures. For example, metals (normal metals and
ical description. Thermodynamic relationships, superconductors), insulators and semiconductors
which establish a correspondence between vari- are classified according to their electrical conduc-
ous thermodynamic quantities, may serve as the tivities; paramagnets (ferromagnets or antiferro-
conceptually simplest example of a macroscopic magnets at low temperatures) and diamagnets are
description. Such a description proceeds from the classified according to magnetic properties; brittle
MACROSCOPIC QUANTUM PHENOMENA 747
and plastic bodies (see Brittle failure) are classified MACROSCOPIC QUANTUM COHERENCE
according to their strength and elasticity; crystals in superconductors
and amorphous bodies (see Amorphous state) are A collective (cooperative) quantum phenom-
classified according to structure, etc. An impor- enon, whereby a superfluid or superconducting
tant component of a macroscopic description is electron-pair (Cooper pair) Bose condensate can
based on equilibrium thermodynamics, as well as be described by a unified complex function ψ(r, t)
the thermodynamics of nonequilibrium processes. called the order parameter of the superconduct-
The limits within which the description holds true ing state. The existence of such a function for
are often clear from the description itself, although superfluid electrons in the bulk of a superconduc-
the precise formulation of the limitations of a tor implies that their motion is highly correlated
macroscopic description necessitates, as a rule, on a macroscopic scale. That is, the whole ensem-
a microscopic treatment (see Macroscopic defects, ble of electrons forming the superfluid condensate
Macroscopic inclusions). exhibits behaviors similar to a single quantum-
mechanical particle. Related to this macroscopic
MACROSCOPIC INCLUSIONS quantum coherence are, e.g., the quantization of
Particles of a foreign phase in the bulk of a u x and Josephson effects.
solid phase matrix that have a characteristic linear
size exceeding the interatomic spacing in all three
MACROSCOPIC QUANTUM PHENOMENA
dimensions. They can either form because of the
in superconductors
prior history of the solid phase, or be purposely
Phenomena related to the wave function phase
introduced into this phase. Solid, liquid, gas, and
χ(r, t) of Cooper pairs in superconductors:
combined inclusions can be distinguished. Macro-
Josephson effects, quantization of ux (fluxoid),
scopic inclusions include hollow pores, gas-filled
macroscopic quantum tunneling, and macroscopic
cavities, and solid phase particles insoluble in the
quantum coherence. According to the BCS mi-
matrix. The latter are generated through internal
croscopic theory of superconductors, the super-
oxidation, through decomposition of oversaturated
conducting phase transition is described by a com-
solutions (see Alloy decomposition), or via liquid
plex order parameter ψ = Φeiχ , which acts as
inclusions captured from the liquid phase during
crystallization. In many actual situations macro- the wave function of a Cooper pair. By virtue of
scopic inclusions give rise to fields of elastic the Bose–Einstein condensation of pairs, the wave
tension, whose existence is responsible for some function of the whole system is ψT ∼ ψ N , where
interactions of these intrusions. In anisotropic me- N is the number of Cooper pairs. Therefore, ψ can
dia such interactions may occur in both attractive be considered as a macroscopic characteristic to
and repulsive modes (see Anisotropy of elastic- describe not only a single particle, but the super-
ity). The presence of macroscopic inclusions may conductor as a whole.
be responsible for consolidation of a crystalline The phase χ(r, t) is coherent, i.e. it retains
matrix through mobile dislocations decelerated by fixed values at different points r 1 , r 2 of a sample,
inclusions. In various force fields, external or in- including those spaced at macroscopic distances,
ternal with respect to the matrix, macroscopic as well as at instants of time t1 and t2 separated by
intrusions may move as an integral whole due to macroscopic time intervals. Observable physical
the existence of directed mass transport main- quantities, such as charge density and current den-
tained by a corresponding force field. The pres- sity, are expressed in terms of derivatives ∂χ/∂r
ence of macroscopic inclusions in a solid affects and ∂χ/∂t in gauge-invariant combinations with
the kinetics of both recrystallization and diffusive vector (A) and scalar (ϕ) potentials:
homogenization, decomposition of oversaturated
ns e
solutions, long-term strength, and propagation of j= (∇χ − 2eA),
elastic vibrations. All the above features of macro- 2m
scopic inclusions find wide use in technological ∂χ
practice. ρ = Ns e + 2eϕ .
∂t
748 MACROSCOPIC QUANTUM TUNNELING
The uniqueness condition of a wave function in perform zero-point vibrations ω0 with frequen-
a closed superconducting ring leads to a phase cies of 1011 –1012 Hz at metastable minima of
change by a multiple of 2π when tracing round the the potential energy. This allows the inequality
ring. This is in accordance with the quantization of ω
kB T to be realized, and the observation of
magnetic flux in ux quantum units Φ0 = h/(2e) a transition from the thermally activated penetra-
(2e results from the electron pairing). tion of the barrier to a temperature-independent
An outcome of this for a bulk hollow cylin- macroscopic quantum tunneling mechanism at
der is the quantization of the flux, frozen in a temperatures of about 1 K. Macroscopic quan-
hole, in units of Φ0 . In thin-walled shells (with tum tunneling disappears not only with an increase
wall thickness less than London penetration depth in T , but also with increased energy dissipation
of magnetic eld), the flux is not quantized. Yet, in the system. Yet more sensitive to dissipation
a circulating current arises with a magnitude and is the macroscopic quantum interference of more
direction periodically changing in flux with incre- than one quantum state of a macroscopic system.
ments of Φ0 (Aharonov–Bohm effect). The flux, A quantum-mechanical description of the phe-
which penetrates into a superconductor in the form nomenon takes dissipation into account. Also clas-
of Abrikosov vortices, is also quantized in fluxoid sified as macroscopic quantum tunneling phenom-
units Φ0 . In the case of superconductors sepa- ena are quantum features of the low-temperature
rated by a thin insulating layer (tunnel junction plasticity of solids (of dislocation motion).
or weak link), the phase is coherent on either
side of the insulating barrier. The phase differ- MACROSTRESSES
ence χ determines the superconducting current First-order residual stresses in solids. Macro-
I that runs through the junction in accordance stresses appear as a result of nonuniform heating
with the Josephson relation I = Ic sin χ . Given the and cooling (welding, flame cutting, casting com-
voltage difference across the junction, the edges plicated shapes), pressure shaping and dressing
remain at equilibrium, i.e. ρ = 0. This leads to the of finished products, heat treatment (in particu-
relationship ∂χ/∂t = 2eV / for the phase differ- lar, at the surface) and sintering, the application
ence, with consequent oscillations of the Joseph- of different coatings, chemical or mechanical sur-
son current in the course of time at a frequency face processing (turning, grinding, polishing, shot-
(the Josephson frequency) determined by the re- blasting), exposure of the surface to energy fluxes
lationship ω = 2eV . A tunnel junction of small (laser or electron irradiation), ion treatment, diffu-
area may undergo transitions between states with sive surface saturation, etc. Macrostresses can be
different coherent phases by means of quantum tensile or compressive. They generally concentrate
tunneling through a Josephson barrier Ic /(2e), in the subsurface layer, a few microns to 2 mm
which separates these states (macroscopic quan- thick. These stresses may originate from such
tum tunneling). In the latter case, the phase χ physical causes as: pile-up of dislocations of a sim-
ceases to be a classical variable, and the system is ilar sign; changes in crystal lattice constants due to
described by a “second quantization” wave func- a nonuniform distribution of impurity atoms; elas-
tion ϕ(χ). tic strain of the lattice at its coherent bonding with
a coating or at a coherent phase boundary interac-
MACROSCOPIC QUANTUM TUNNELING tion, etc. The magnitude and sign of macrostresses
Exhibition of quantum behavior by a macro- depend upon the material of a mechanical part,
scopic system, with a collective dynamic variable upon the manner of its treatment, upon the man-
for a degree of freedom (see Collective variables ner of fixing, way of cooling, etc. If a macrostress
method). The phenomenon of macroscopic quan- exceeds the yield limit in a plastic material, then
tum tunneling occurs with Josephson junctions relaxation occurs by way of plastic deformation.
and in superconducting quantum interference de- In low plasticity materials macrostress may ex-
vices (SQUIDs), where the collective variable is ceed the ultimate strength, leading to the formation
the phase difference across the junction, or the to- of cracks, spallings, and other surface defects.
tal magnetic flux in the squid loop. These systems Macrostresses and their changes during the course
MAGNESIUM 749
of operation strongly affect the performance char- heat of sublimation is 152.6 kJ/mole. Specific
acteristics of mechanical parts. heat is cp = 1.02 kJ·kg−1 ·K−1 (298 K). De-
bye temperature is 405 K; linear thermal ex-
MADELUNG CONSTANT (E. Madelung, 1918) pansion coefficient of magnesium monocrystal is
A dimensionless constant, α, dependent solely 26.8·10−6 K−1 along the main sixfold axis 6,
on the structure of a binary ionic crystal, which al- and 24.7·10−6 K−1 perpendicular to this axis
lows evaluating the potential ϕ arising from the at 280 K; coefficient of the thermal conductiv-
Coulomb interaction at a lattice site produced by ity is 138 W·m−1 ·K−1 at room temperature and
all the other ions from the equation: ϕ = αe/r0 , normal pressure. Adiabatic elastic moduli of mag-
where ±e are ionic charges, the sign of the poten- nesium monocrystal: c11 = 59.28, c12 = 25.90,
tial ϕ is opposite to the sign of the charge at the
c13 = 21.57, c33 = 61.35, c44 = 16.32 GPa at
site considered, and the characteristic length r0 is
298 K; bulk modulus is ≈35 GPa, Young’s modu-
the distance between nearest neighbor ions. The
lus is ≈45.5, shear modulus is ≈18 GPa; Poisson
value of α is calculated by summing over all the
ratio is ≈0.28 (at 298 K); at 293 K the properties
atoms in the lattice, and it is a universal constant
of cast and deformed magnesium are characterized
for a given lattice type. In crystals with several
ions carrying different charges in their unit cell, by the following parameters: ultimate strength is
the Madelung potential at the sth site is given by 0.113 and 0.196 GPa, relative elongation is 8.0 and
the equation: 12.0%, yield limit is 0.025 and 0.088 GPa, respec-
tively. Brinell hardness is 0.294 and 0.353 GPa;
p
e fatigue limit of annealed magnesium with ba-
ϕ(rs ) = αss s ,
r0 sis 5·108 is 0.062 GPa; recrystallization tem-
s =1
perature is 423 K. The strength of magnesium
where the ionic charges are designated by es , is increased by cold hardening. Self-diffusion
s = 1, 2, . . . , p (p is the number of atoms in the coefficient of magnesium is 3.04·10−14 m2 /s
unit cell), and the p2 quantities αss form a sym- along the axis 6 and 3.75·10−14 m2 /s perpen-
metric square matrix. dicular to it at 738 K. Ion-plasma frequency
of magnesium is 17.78 THz; Sommerfeld coef-
MAGGI–RIGHI–LEDUC EFFECT
ficient of linear low-temperature electronic spe-
See Thermomagnetic phenomena.
cific heat is 1.32 mJ·mole−1 ·K−2 . Resistivity of
monocrystal is 3.48·10−8 ·m along the princi-
MAGNESIUM, Mg
pal axis 6 and 4.18·10−8 ·m perpendicular to
A chemical element of Group II of the peri-
it, temperature coefficient of electrical resistivity
odic system with atomic number 12 and atomic
is 0.00408 K−1 along the principal axis 6 and
mass 24.305. Natural magnesium consists of the
stable isotopes 24 Mg (78.60%), 25 Mg (10.11%) 0.00390 K−1 perpendicularly to it, Hall constant
and 26 Mg (11.29%). Outer shell electronic con- is −8.42·10−11 m3 /C (room temperature, normal
figuration is 3s 2 . Successive ionization energies pressure); coefficient of the absolute thermal elec-
are 7.645, 15.035, 80.144 eV. Atomic radius is tromotive force is +3.3 μV/C along the principal
0.1602 nm, radius of Mg2+ ion is 0.066 nm. Oxi- axis 6 and +3.5 μV/C perpendicular to it. Work
dation state is +2. Electronegativity is ≈1.29. function of electron from polycrystal is 3.64 eV.
Magnesium in a free form is a silvery-white Magnetic susceptibility of paramagnetic magne-
metal. It has a hexagonal close-packed lattice with sium is +13.25·10−6 cm3 /mole in CGS units
the parameters a = 0.32088 nm, c = 0.52095 nm (300 K). Nuclear magnetic moment of 25 Mg iso-
(at 298 K), space group P 63 /mmc (D6h 4 ). Den- tope is +0.855 nuclear magnetons. Magnesium is
sity is 1.74 g/cm3 . T melting = 922 K; T boiling = mainly used in production of superlight alloys (see
1376 K. Bonding energy of pure magnesium Magnesium alloys). Metallic magnesium is used
is 1.53 eV/atom at 0 K. Heat of melting is for reduction of Zn, Th, U and other metals and
8.5 kJ/mole, heat of evaporation is 130 kJ/mole, their compounds.
750 MAGNESIUM ALLOYS
MAGNETIC AXIS
The common principal axis of the g-tensor
and of the crystal field tensor of the quadrupole
and hyper ne interactions (see spin Hamiltonian
parameters). The axis is parallel to one of the crys-
tal symmetry directions, or perpendicular to one
of the symmetry planes. The magnetic axis di-
rection coincides with the applied magnetic field
orientation corresponding to extrema in angular
dependences of spectra. If low-symmetry effects
appear in the spectra, viz., if the principal tensor
axes are dissimilar, pseudoaxes averaged over all
the resonance transitions will sometimes be con-
ventionally selected. Such axes may deviate from
Fig. 1. Classical orbits (1 and 2) of different energy
the symmetry axes and from normals to the sym- bands approaching one another, having the same mo-
metry planes by as much as tens of degrees, and, mentum pz of an electron of energy E in the form of
with their orientation depending on the observa- two hyperbolae whose asymptotes intersect at the center
tion frequency (field), they have little physical point 0. The region of magnetic breakdown is indicated
significance. by hatching. 1− , 2− and 1+ , 2+ are sections of classical
A coordinate system that includes a magnetic orbits respectively entering and exiting the breakdown
axis is called a magnetic coordinate system. For region. Arrows indicate directions of classical motion.
high-symmetry spectra its axes coincide with spe-
cific crystallographic directions like 100, 110,
at its minimum in the Bragg reflection planes
111, 112, etc. This approach is useful if the (see Bragg diffraction). Therefore, the band-to-
crystal has several equivalent paramagnetic cen- band tunneling probability (magnetic breakdown
ters, since describing the spectrum of each such probability) depends only on the orientation of the
center in its own magnetic coordinate system is magnetic field, B = (0, 0, B), relative to the crys-
much simpler than presenting it in a general crys- tallographic axes, and on the value of the projec-
tallographic system of coordinates. tion of the electron quasi-momentum (pz ) along
the B direction. Since the magnetic breakdown re-
MAGNETIC BREAKDOWN gions are rather narrow they can be considered as
Quantum tunneling of conduction electrons in point centers of two-channel quantum scattering:
metals in a magnetic field, B, between classi- with or without a change of energy band num-
cal electron orbits of different energy bands. The ber (see Fig. 1). Corresponding to the former is a
term “magnetic breakdown” reflects the classi- magnetic breakdown scattering through the chan-
cally forbidden aspect of the interband tunneling nels 1− → 2+ and 2− → 1+ , and correspond-
in a magnetic field. Because of the small magni- ing to the latter, through the channels 1− → 1+
tude of the dimensionless quasi-classical approx- and 2− → 2+ . A closed configuration of quasi-
imation parameter, κ = eB/pF2 1 (pF is the classical trajectories corresponds to an arbitrarily
electron Fermi momentum and ωc = eB/m is the aligned B. In a field B parallel to the recipro-
cyclotron frequency), the band-to-band transitions cal lattice vector b, an open periodic configuration
only occur within narrow classically forbidden re- with period b often results (Fig. 2). Generally the
gions of momentum space (magnetic breakdown electron dynamics at magnetic breakdown is of a
regions). There the gap Δ between the bands is quantum nature, determined by the coherent in-
so small that the classical electron orbits from terference of quasi-classical waves arising from
different bands approach one another to within multiple electron scattering at breakdown cen-
distances comparable to those estimated from the ters. The energy spectrum of electrons (magnetic
quantum uncertainty of the transverse momentum: breakdown spectrum) is defined in this quantum
(eB)1/2 ≈ κ 1/2 p0 (Fig. 1). The gap is usually interference situation by the quasi-classical phase
754 MAGNETIC BREAKDOWN
of the thermodynamic parameters coupled to them the magnetic bubble domain converts to a band-
(de Haas–van Alphen effect), contain only har- type domain or, more often, to the labyrinthine
monics with S cl / phases, where S cl is the action magnetic domain structure. In samples where the
increment for all the possible closed paths of the magnetic characteristics are substantially nonuni-
breakdown. form in thickness over their surface area, internal
Also typical for a coherent magnetic break- domains in the magnetic bubble domain may ex-
down is a variety of nonlinear effects in any ist, which are partly or entirely located inside of
external field capable of changing the electron the film rather than on its surface, as in case of
momentum by a rather small value, of the order ordinary domains. Under the influence of an inho-
of ∼κpF , during a time ∼τ B (or τ Q ). Coherent mogeneous magnetic field, temperature gradients,
magnetic breakdown exhibits an abnormal sensi- and other factors, a magnetic bubble domain may
tivity of kinetic coefficients to small inclinations move over the film plane. When present with a
of B to the symmetry axes (planes), or to the fairly high density, due to magnetic dipole repul-
planes perpendicular to reciprocal lattice vectors. sive forces, they form a magnetic bubble domain
A very unusual situation shows up in the latter lattice. The possibility for controlled magnetic
case: a pure metal behaves like an incommensu- bubble domain creation and annihilation at arbi-
rate structure with the quantum localization of trary sample points, as well as their motion by
conduction electrons. The above characteristics means of a magnetic field, permits their use in
of the magnetic breakdown kinetics open a way various devices (see Magnetic bubble domain de-
to design fundamentally innovative electronic de- vices).
vices.
MAGNETIC BUBBLE DOMAIN DEVICES
Solid-state computer engineering devices
MAGNETIC BUBBLE DOMAIN, cylindrical
which utilize a magnetic film with magnetic bub-
magnetic domain
ble domains. The magnetic bubble domain devices
A small solitary magnetic domain shaped like are generally used for primary and auxiliary mem-
an upright (perpendicular to the surface) circular ory functions, but they are also capable of per-
cylinder. Magnetic bubble domains are observed forming logic operations. These devices are con-
in single-crystal films 0.1–100 μm thick with the structed using a set of integrated micro-assemblies
easy magnetization axis normal to the sample sur- of different capacities (up to several megabits)
face, as well as in the presence of a magnetic controlled by an external electronic system. The
phase transition. Examples of such films are or- micro-assembly contains: films of magnetic bub-
thoferrites, iron garnets of rare-earth elements, ble domain materials with functional modules ap-
amorphous magnetic substances, alloys of rare- plied on the surface; a set of permanent magnets
earth and transition metals, and other magnetic for establishing the bias field; a system for estab-
substances. The diameter of a magnetic bubble lishing the control magnetic field (control current).
domain cross-section has a magnitude of the same The diameters of the magnetic bubble domains
order as the film thickness, and decreases as a con- used in these devices range from 5 μm to less than
stant magnetic eld normal to the sample surface 1 μm.
increases. An ordinary, solitary magnetic bubble The memory performance is based on the oper-
domain is found only when such a field plays a ation of the following elements: magnetic bubble
stabilizing role. In this case, the film magnetiza- domain forwarding circuits, which are often series–
tion is oriented along the magnetic field, while parallel connected for the purpose of increas-
the magnetic bubble domain magnetization is in ing the data exchange rate; switches responsible
the opposite direction. As the field increases to for connections between the shift memory regis-
some critical value Bc (magnetic bubble domain ters formed by elements of forwarding circuits;
collapse field) the domain size decreases, and generator (source) of magnetic bubble domains;
then the domain disappears (magnetic bubble do- detector (sensing element) which takes the read-
main collapse). In the case when the field goes ings from magnetic bubble domains (generally by
to the value B2 (a field of elliptic instability), application of the magnetoresistance effect). In
756 MAGNETIC BUBBLE DOMAIN FORWARDING CIRCUITS
order to carry out the information read-out the a domain. Using permalloy, circuit densities
magnetic bubble domain must be expanded into ∼106 MBD/cm2 are achieved. Still higher densi-
a film strip (see Magnetic lms), so the data read- ties are reached through ion-implanted structures
out module occupies a rather substantial part of the (see Ion implantation). Clock frequencies are re-
film surface. A magnetic bubble domain annihila- stricted by the domain wall peak velocity, and also
tor eliminates unnecessary domains. The domain by technical problems associated with the control
micro-assembly, as well as the domain forwarding field. For the first and third types of circuits the
circuits, are characterized by a region of stable op- tact frequency is, commonly, ∼200 kHz, for cur-
eration located near the functional units mentioned rent conductors it is well above this.
above. The magnetic bubble domain memory is The most important feature of the MBD for-
nonvolatile, involves low energy consumption, is warding circuits is the region of stable operation
insensitive to external actions, and has high re- on the “shift field versus control field (current)”
liability. It has been demonstrated that memory phase plane. This region should be within a nar-
devices may be created on the basis of magnetic row interval of the control field (current) value,
bubble domain lattices, which provides a nearly and at the same time, it should have a maximum
tenfold increase of recording density compared to width with respect to the shift field value. The the-
systems utilizing ordinary forwarding circuits. oretical construction of the stable region requires
an examination of the closed system of equations
MAGNETIC BUBBLE DOMAIN FORWARDING for the magnetic bubble domain and the scheme
CIRCUITS elements.
System of operating elements located on a film
surface, which induces energy wells for magnetic MAGNETIC BUBBLE DOMAIN LATTICE
bubble domains (MBD). Also it is intended for (magnetic cylindrical domain lattice in Russian
the bubble domain storage and transport. As a terminology)
rule, an MBD comprises an associated system A periodic two-dimensional magnetic domain
of memory shift registers, wherein the informa- structure of a thin magnetic lm, its lattice sites
tion is contained as code sequences: the logic unit having magnetic bubble domains. In a film with
“1” indicates the presence, and “0” denotes the the easy magnetization axis perpendicular to its
absence of the MBD storage. The forwarding cir- surface a hexagonal lattice of magnetic bubble do-
cuits are fabricated by photolithographic methods, mains can form. The equilibrium diameters of the
and characterized by a variety of spatial shapes domains of this lattice, and its period, depend on
and types. Schemes of three types (and their com- the value of the external magnetic eld B per-
binations) are used: those in applications of soft pendicular to the surface of the sample. Since the
magnetic materials (commonly, permalloy), cur- generation and annihilation of a separate magnetic
rent conductors, and ion-implanted structures. In bubble domain lattice are often impeded by the
the first and third cases the MBD shifting per sin- presence of an energy barrier, this lattice is often
gle space period of the scheme is handled through not in equilibrium, and in particular ranges of B
the magnetic reversal of the scheme elements, or values its period remains constant. The dynamical
by means of a magnetic field (operating field). properties of a magnetic bubble domain lattice are
This field rotates at a fixed (tact or clock) fre- determined by the set of natural oscillations of the
quency on the plane of the magnetic lm as the magnetization, which are accompanied by small
field turns through an angle 360◦ . In the second displacements of the bubble domains from their
case the operating currents perform the analogous equilibrium positions at the lattice sites, and also
function. by variations of their dimensions and shape. In the
Magnetic bubble domain forwarding circuits spectrum there are two activation-free branches
occupy the major part of the film area defining of vibrations of the acoustic phonon type corre-
the density of information recording. The magne- sponding to displacements of the magnetic bubble
tostatic repulsion among MBDs limits the scheme domain lattice from the equilibrium sites, and
spacing period to 4–5 average diameters of there is also a series of branches of vibrations
MAGNETIC COMPENSATION 757
whose frequencies have finite values of the order reluctance which limits the flux, and mmf is the
of 108 Hz in the limit of infinite wavelengths. The magnetomotive force which generates the flux. In
propagation velocities of the lattice vibrations are this circuit mmf is the analogue of the applied volt-
usually of the order of 104 to 105 cm/s. These vi- age or emf; Φ is the analogue of the current I , and
brations may be excited by an alternating magnetic Rm is the analogue of the resistance (impedance).
field. Kirchhoff’s laws are applicable to magnetic cir-
cuits. However, a fundamental difference between
MAGNETIC BUBBLE DOMAIN, RIGID magnetic and electric circuits is the absence of dis-
See Rigid magnetic bubble domain. sipation of electromagnetic energy (Joule heat) in
a magnetic circuit with magnetic flux Φ that is
MAGNETIC CHARACTERISTIC LENGTH constant in time. The magnetic circuit approach
See Characteristic length of magnetic material. is useful in designing permanent magnets, elec-
tromagnets, magnetic amplifiers, relays, electric
MAGNETIC CHARGE, magnetic pole strength measurement instruments, etc.
An auxiliary concept introduced in magneto-
statics to calculate the demagnetization eld (ana- MAGNETIC COLLISION-FREE DAMPING of
logue of concept of electric charge generating an waves
electrostatic field). In contrast to electric charges, The principal mechanism of damping of elec-
magnetic charges do not actually exist (see Mag- tromagnetic oscillations propagating in metals.
netic monopole), since according to classical mag- Like Landau damping, magnetic collision-free
netism theory the magnetic field has no special damping is associated with the transfer of the elec-
sources besides electric currents. For media ex- tromagnetic field energy to particles (e.g., elec-
hibiting a magnetization M one can introduce the trons) that move in phase with the wave. Yet,
concepts of a volume density, ρ m , and of an area magnetic collision-free damping is not determined
density, σ m , of magnetic charge. The former is de- by the electrical force eE, but rather by the mag-
fined as the divergence of the magnetization, i.e. netic force ∇(μB) with which the alternating field
ρ m = div M, and the latter as a jump of the normal of the wave B acts upon an electron rotating in an
component of the magnetization at a discontinu- external field B ext (μ is the mean magnetic mo-
ity interface: σ m =
M n (e.g., σ m = M n at a ment of a rotating electron). Therefore, magnetic
medium–vacuum boundary, where M n is the nor- collision-free damping is often called magnetic
mal component of the surface magnetization). See Landau damping. Due to the Larmor rotation of
also Magnetic pole. electrons in the magnetic field (see Larmor pre-
cession), the scenarios of Landau damping and
MAGNETIC CIRCUIT magnetic collision-free damping differ from each
A series of connected magnetic substances other. Landau damping involves particles whose
conducting magnetic ux. If the flux in a magnetic velocity projections on the wave propagation di-
circuit is generated by a permanent magnet then rection coincide with the phase velocity, whereas
such a circuit is called magnetized circuit. A mag- in the magnetic collision free case projections of
netic circuit without permanent magnets is known the particle velocity and of the wave phase veloc-
as a neutral circuit, with the magnetic flux in it ity along the direction of B should coincide.
generated by an electric current flowing through
coils enclosing part of the circuit. Depending on MAGNETIC COMPENSATION
the source of flux there are magnetic circuits of the Vanishing of the resultant overall magnetiza-
permanent, alternating, or pulsed types. Because tion of a ferrimagnet when at least one of the
of the formal analogy between electric and mag- magnetic sublattices has a finite magnetization in
netic circuits, the same mathematical formalism is the absence of an applied magnetic eld. It oc-
applicable to both. For instance, a counterpart of curs at a temperature, T c , called the compensation
the Ohm’s law emf = I R is the equation mmf = point. The appearance of magnetic compensation
ΦR m , where Φ is the magnetic flux, R m is the is due to the nonequivalence of the magnetic ions
758 MAGNETIC CONCENTRATION EFFECTS
(atoms) belonging to different sublattices, and to One can control the rates of recombination and
the resulting different temperature dependences generation by changing the surface recombination
of the sublattice magnetic moments in their an- rates, I , at the surfaces involved. For I0 < Id , and
tiparallel orientation. The occurrence of magnetic a current density jx > 0, the specimen is enriched
compensation is typical of most rare-earth iron with current carriers and its resistance R drops; for
garnets, of some orthoferrites, of metallic glasses jx < 0, the specimen is depleted of carriers and R
based on d- and f -metal alloys, and also of a increases. The emergence of such a rectification
large number of mixed (or dilute) ferrites with di- property is termed the magnetic rectifying effect,
verse structures and compositions. Anomalies of or magnetic barrier effect. The deflection of the
thermodynamic and kinetic quantities such as sus- injected current carriers (see Injection) by the mag-
ceptibility, speci c heat, coercive force, etc., are netic field, and the change of their recombination
observed at the compensation point. In addition, rate, is called the Suhl effect (H. Suhl).
the temperature interval near T c is particularly The intrinsic magnetic field of a high current
advantageous for studying noncollinear magnetic flowing through a sample (e.g., wire) draws the
states due to a dramatic decrease in the critical electrons and holes into the central part of the
magnetic field at magnetic phase transitions into specimen, which is known as the pinch effect. The
such states. magnitude of the current causing a marked cur-
rent carrier redistribution is often evaluated based
MAGNETIC CONCENTRATION EFFECTS on the Bennett condition (W.H. Bennett, 1934)
Effects governed by matched changes of the borrowed from gas plasma physics: it involves a
electron and hole concentrations in a bipolar parity between the kinetic pressure produced by
plasma in a semiconductor placed in crossed elec- electrons and holes, and the pressure of the mag-
tric E and magnetic B fields. Such effects were netic field. As distinct from the gas plasma case,
discovered and studied in plasmas of Ge, Si and the surface generation of carriers has a signifi-
InSb. The electrons and holes drifting in the field cant bearing on the pinch effect in semiconductors.
E (see Current carrier drift) are deflected by the A longitudinal magnetic field B suppresses the
Lorentz force F = e(E + v × B), which results in pinch effect due to the developing Kadomtsev–
their being in excess on one surface of the mate- Nedospasov instability which is employed to di-
rial and depleted on the other surface (see Fig.). agnose the latter. Flows of electrons and holes
If the recombination rate of electrons and holes across the specimen may appear owing to spon-
on the first side exceeds the rate of their gen- taneous or deformation-induced anisotropy of the
eration on the second side, then the specimen is current carrier mobility. In this case, the deflection
depleted of current carriers, whereas it is enriched of electrons and holes toward one surface of the
with them otherwise. The specimen resistance, R, specimen is termed electric pinch. The above ef-
changes in both cases; this effect is termed the fects are used to make magnetic field and pressure
galvanomagnetic recombination effect, or Welker gauges, and radiation modulators; they also occur
effect (H. Welker, 1951). in magnetic diodes and triodes.
MAGNETIC COOLING
See Adiabatic demagnetization cooling.
MAGNETIC DEFECTOSCOPY
See Defectoscopy.
MAGNETIC DIPOLE
A particle such as an unpaired electron which
possesses a magnetic moment whose interactions
with other magnetic moments in a medium takes
Magnetic concentration effects. place via the magnetic dipole interaction. Magnetic
MAGNETIC DOMAIN CHARACTERIZATION 759
dipoles present in crystals are: nuclei (with non- MAGNETIC DOMAIN CHARACTERIZATION
zero spins), paramagnetic ions (see Magnetic ion), Experimental studies of the size, shape, po-
free radicals, and triplet excitons. sitioning, and restructuring features of magnetic
domains in magnetic materials with a magnetic
MAGNETIC DIPOLE INTERACTION, magnetic domain structure. Techniques of characterizing
dipole–dipole interaction magnetic domains are based on various physical
Interaction between two magnetic moments μ1 phenomena. Historically the first such technique
and μ2 with each magnetic moment attracted by (1919) was based on the Barkhausen effect, which
the magnetic field of the other. The resulting en- consisted in recording Barkhausen jumps that
ergy arising from this interaction is given by the result from remagnetization of separate domain
equation: groups. In the 1930s the technique of magnetic
μ0 powder patterns was introduced (F. Bitter, 1931;
E mdi =
4π N. Akulov, M.D. Detgyar, 1932; W.C. Elmore,
1936), making it possible for the first time to
(μ1 · μ2 )|R 12 |2 − 3(μ1 · R 12 )(μ2 · R 12 )
× , identify the size and shape of domains in vari-
|R 12 |5
ous ferromagnets. It involved a special treatment
where R 12 is the radius vector connecting the two of the specimen surface, including chemical and
magnetic moments. If magnetic dipole moments mechanical polishing. These techniques made it
are associated with particles with spins S 1 and S 2 , possible to judge the direction of the tangent to
the Hamiltonian terms for the magnetic dipole in- the projection plane of the domain magnetization:
teraction is obtained by the direct generalization a scratch normal to that projection becomes richly
of the above equation through the replacement of decorated with the powder, while one parallel to it
μ1 and μ2 by their quantum analogs, γ1 S 1 and is not so affected (scratch technique). K.I. Sixtus
γ2 S 2 , where γ1 and γ2 are the gyromagnetic ra- and L. Tonks (1931) studied the magnetic domain
tios. wall dynamics in ferromagnetic wires by detecting
a signal in an electromagnetic induction solenoid
MAGNETIC DOMAIN, ferromagnetic domain (induction technique). After the synthesis in the
A region a ferromagnet (or other magnet) 1950s of magnetic materials transparent in the IR
magnetized uniformly to saturation. A ferromag- and visible regions, techniques based on various
netic sample in a sufficiently weak external field. magnetooptical effects came into wide use.
is irregularly partitioned into magnetic domains
Currently the leading methods for delineating
present in thermodynamic equilibrium below the
magnetic domains are magneto-optic techniques
Curie point. The concept of magnetic domain was
that employ the Faraday effect and the Cotton–
introduced by P. Weiss (1907) (see Weiss theory).
Mouton effect (see Voigt effect) for transparent
In the absence of any present or prior external
magnetic field, the magnetization vectors of each materials, and the Kerr effect for opaque ones.
domain are oriented differently, so that the overall The powder pattern method is often used in com-
magnetization of a ferromagnetic crystal is close bination with magnetooptical techniques, making
to zero. The shape and the dimensions of ther- it possible to study both the bulk and the sur-
modynamically equilibrium magnetic domains are face properties of magnetic domains. Electron mi-
determined by the condition of minimum free en- croscopy based on differences between the Lorentz
ergy of the ferromagnet. They depend on the shape force deflections of electron beams in various do-
and dimensions of the body, and are significantly mains with differing magnetizations is also used
larger than the magnetic domain wall thickness be- (Lorentz microscopy), as well as X-ray and neu-
tween them. In magnetic thin films strip domains tron scattering. To study the dynamics of domain
and magnetic bubble domains, etc., can occur. In wall motions various induction and magnetoopti-
large-size samples of ferromagnets the domain cal methods may also be used, including instanta-
shape can be more complicated. neous and sequential superrapid photography.
760 MAGNETIC DOMAIN, LABYRINTHINE
Dependence of the domain wall velocity in yttrium orthoferrite YFeO3 on the external magnetic field.
of the magnetic after-effect (see also Néel equa- B2 the frequency ω2 , of the elliptical deforma-
tion, Magnetic viscosity). tions of the cylindrical cross-section, also reduces
At B = 0 Eq. (1) describes the free attenuating to zero. The characteristics of the forward motion
e xural waves of the domain wall with frequency of the domain due to the non-uniform field arise
ω = (σ/m∗ )1/2 |k ⊥ |, where k ⊥ = (kx , ky , 0). If from the presence of the gyroscopic force F G of
the wall is not one-dimensional (see Domain wall the domain F G = A[M0 × V c ], where M0 is the
twisting), then its dynamics are more complex. For magnetization matrix, V c is the domain velocity,
domain walls with vertical Bloch lines the equa- the coefficient A = (2πh/γ )S, the topological in-
tions for coordinates of the wall and the Bloch dex S is equal to the number of complete rotations
lines must take into account their mutual influence of the magnetization vector in the center of do-
via the gyroscopic force F G . The analysis may be main wall with the turning of the side surface of
performed on the basis of the Slonczewski equa- the domain, h is the thickness of the film, and γ is
tions for the wall coordinate u and the angle ψ the gyromagnetic ratio. The action of F G deviates
that the magnetization makes with the domain wall the bubble domain trajectory from the direction
plane. The wall dynamics may become multidi- of the field gradient, making its dynamics similar
mensional for high values of the field B, or within to that of a charged particle in crossed magnetic
the region of negative differential wall mobilities. and electrical fields. At high domain velocities the
The domain wall dynamics of the boundary of number and shape of the Bloch lines may change,
a bubble domain are associated with deformations, leading to the dynamic transformation of magnetic
and with the forward movement of the domain. bubble domains.
The small wall oscillations caused by non-uniform
radial deviations
r(ϕ) of the bubble domain ra- MAGNETIC DRIFT, unsymmetrical hysteresis
dius with respect to the angle ϕ are conditioned loop drift
by deformations. These oscillations are character- An effect occurring during the magnetic rever-
ized by a discrete set of the natural frequencies ωn sal of magnetic materials in an external magnetic
which depend on the component of B that is e ld that periodically varies between certain val-
normal to the surface of the sample, and on the ues B1 and B2 , with B1 = −B2 , i.e. the magnetic
parameters of the magnetic lm. In the field Bk of reversal cycle is asymmetric. The magnetic drift
a magnetic bubble domain collapse, the frequency occurs because at the end of the cycle, when the
ω0 corresponding to the uniform oscillations re- applied field has returned to its initial value B1 , the
duces to zero; in the field of an elliptical instability magnetization M has grown somewhat, instead of
764 MAGNETIC FIELD
coming back to it initial value. Then the mag- Φ = BA. If the field is not uniform then B must
netic hysteresis loop is not closed on the M(B) be integrated over the area to obtain the flux.
plane, but permanently drifts in the direction of A magnetometer is used to measure the
increasing M. In some magnetic substances such strength of a magnetic field (see Magnetometry).
a process may persist over a great number (up
to 106 ) of magnetic reversal cycles. The drift re- MAGNETIC FIELD, CRITICAL
sults from minor random irreversible changes in See Critical magnetic elds.
the atomic magnetic structure of the specimen dur-
ing the cyclic switching of the magnetic field, and MAGNETIC FIELD PENETRATION DEPTH
it appears most prominently if the field magnitude See Penetration depth of magnetic eld.
varies near the value of the coercive force. Such
changes in M may amount up to 0.01 mT for the MAGNETIC FIELD, QUANTIZED
magnetic drift in ferromagnets. See Quantizing magnetic eld.
Fig. 1. Magnetic films. Perpendicular anisotropy case: (a) striped domain structure for unipolar twisted domain walls and
q
1, (b) corresponding case with q
1 for heteropolar domain walls, and (c) unipolar domain walls with q 1.
constant of bulk specimens). Magnetic films that the normal takes place in the domain centers on
are not too thick and are uniform in thickness pos- the film surfaces (Fig. 1(a)). There is no such de-
sess a passing-through magnetic domain structure; viation in the structure with heteropolar domain
incomplete domains may exist in non-uniform walls (Fig. 1(b)). The structure with heteropolar
films. The good permeability of the magnetic domain walls possesses a higher energy compared
films, in combination with their high specific Fara- to that of the structure with unipolar boundaries,
day rotation, permits employing magneto-optical with the energy difference for these two structures
techniques for the visual observation and study of tending to zero at q → ∞. Since the films with
their domain structure in transmitted light; other EMA n exhibit rotational invariance, they also
methods of magnetic domain characterization are may contain a labyrinthine magnetic domain struc-
also used. The domain structures of magnetic ture, isolated magnetic bubble domains, as well
films are differentiated by the orientation of the as mixed type structures composed, e.g., of ran-
easy magnetization axis (EMA), and by a parame- domly arranged fragments of the stripe and bubble
ter known as the quality factor q = K u /2πM 2 , domains. The bubble domains containing an even
where M is the magnetization of the film: number of pairs of vertical Bloch lines, or contain-
(a) The “perpendicular” anisotropy case ing no such lines at all (Fig. 2(a)), are composed
(EMA n, where n is normal to the magnetic film of the striped domains with heteropolar domain
surface). With q
1, a striped domain structure, walls; the striped domains with unipolar domain
or striped structure, is formed with uni- or het- walls display a trend of turning into bubble do-
eropolar twisted domain walls (see Domain wall mains having an odd number of pairs of Bloch
twisting). In the structure with unipolar domain lines (Fig. 2(b)). With the quality factor q decreas-
walls (when the vectors M in the centers of all ing, the energy of the demagnetization eld starts
domain walls in the magnetic film median plane playing the dominant role in the magnetic film en-
are collinear), a small deviation of vectors M from ergy, hence the angles between the vectors M and
766 MAGNETIC FILMS
lines (epitaxial films of mixed ferrite gar- of the magnetic field vector, H , and the magneti-
nets); working media in storage devices with zation vector, M. In the SI system of units we have
thermomagnetic information recording and
B = μH = μ0 (H + M), (1)
magneto-optical reading (amorphous lms of
d-metal and f -metal alloys), etc. where μ is the magnetic permeability, μ0 =
(2) Magnetooptics and integrated optics: con- 4π·10−7 N/A2 (or H/M) is the magnetic perme-
trolled transparencies, light modulators and ability of vacuum, and the magnetization M is the
group switches, transducers of optical modes magnetic moment per unit volume. The latter is di-
(orthoferrite plates and epitaxial films of the rectly proportional to H in weak magnetic fields
bismuth- and praseodymium-containing fer- and isotropic media:
rite garnets), etc.
(3) SHF, UHF and EHF engineering: guides, M = χH , (2)
reciprocal and non-reciprocal devices using where χ is the dimensionless magnetic suscep-
magnetostatic waves or ferromagnetic reso- tibility. Upon substitution of Eq. (2) into (1) we
nance (epitaxial iron–yttrium garnet films and obtain B = μ0 (1 + χ)H and μ = μ0 (1 + χ). (In
hexaferrites). the CGS system μ0 = 1 and B = H + 4πM .) The
(4) Household electronics: magnetic film heads, units of magnetic flux density are tesla (T) in SI
magnetooptical laser-player disks (permalloy and gauss (G) in CGS; 1 T = 104 G.
and amorphous films of d-metal and f -metal
alloys), etc. MAGNETIC HARDNESS
The property of magnetic materials whereby
MAGNETIC FLUX, ux of magnetic eld the application of a very strong, reversed direction,
The flux Φ of the vector of magnetic ux den- magnetic field is required to bring the material
sity (magnetic field) B through a given surface back to its unmagnetized state (see Demagnetiza-
area S. An element, dΦ of magnetic flux through tion). The extent of magnetic hardness is character-
an infinitely small surface element dS is dΦ = ized by the coercive force, H c , and remanence, B r .
Bn dS, where Bn is the projection of B onto the Materials with large values of H c and B r , or those
direction of the unit vector n, normal to dS. The with the broadest magnetic hysteresis loops, are
magnetic flux through an arbitrary surface S is de- called hard magnetic materials; they are mainly
fined by the integral Φ = Bn dS. For a closed used for permanent magnets.
surface the magnetic flux is zero, and this can be
expressed mathematically with the $aid of diver- MAGNETIC HYSTERESIS
gence, div B = 0, and the integral, Bn dS = 0, A nonunique irreversible dependence of the
expressions. This situation results from the ab- magnetization M of a magnetically ordered mate-
sence of magnetic charges in Nature (see Magnetic rial, e.g., a ferromagnet, an antiferromagnet or an
monopole). For a uniform magnetic field one can amorphous magnetic substance during a variation
write Φ = BS. The units of magnetic flux are: Wb (increasing or decreasing) of an external magnetic
(weber) = Tm2 , in SI and Mx (maxwell) in CGS, field on its prior history. A general cause of mag-
where 1 Mx = 10−8 Wb. netic hysteresis is the presence within a certain
range of applied field of some metastable states
MAGNETIC FLUX DENSITY (side by side with stable ones), with irreversible
The vector of magnetic flux density, B, is transitions between them.
a dynamic parameter of the magnetic eld in a For a detailed explanation of this phenomenon
medium that averages the fields generated by indi- in ferromagnets, the individual magnetic domain
vidual electrons and other fundamental particles. structure must be taken into account, both its for-
The magnetic flux Φ passing through an area S mation through nucleation and incipient growth,
normal to the direction of a uniform field B is as well as its variation during magnetization and
given by the expression Φ = BS. The magnetic magnetic reversal processes. These changes can
flux density B can usually be expressed in terms proceed via domain wall displacement processes,
768 MAGNETIC HYSTERESIS
(3) special purpose magnetic materials, com- The situation in quantum mechanics is some-
prising those with a rectangular hysteresis loop, what different. Consistent equations of motion for
magnetostrictive materials and thermomagnetic a charged particle moving in the field of a mag-
materials, magnetic recording materials (see netic monopole, and those for a monopole moving
Thermomagnetic information recording and in the field of a charged particle, can only be con-
Magnetic bubble domain devices), materials for structed if the electric charge, e, of the particle and
microwave equipment, and microelectronics, etc. the magnetic charge, g, of the magnetic monopole,
See also Amorphous magnetic materials, are related through the expression:
Anisotropic magnetic materials, Hard magnetic
eg = 2πn = nh,
materials, High-permeability magnetic materials,
Soft magnetic materials. where e is the charge of an electron, and n
is a positive or negative integer. This condition
MAGNETIC MOMENT arises when the particles are represented by waves
A vector quantity characterizing the magnetic in quantum mechanics, and interference effects
properties of closed electric currents in bodies emerge in the motion of particles of one type influ-
or material particles. The magnetic moment, μ, enced by particles of another type. If the magnetic
of a closed circuit carrying a current I equals: monopole were to exist, the above equation stip-
μ = I S, where S is a vector numerically equal to ulates that all the charged particles located in its
the area encompassed by the circuit, and aligned vicinity have a charge e equal to a multiple of the
along the normal to the circuit plane in the quantity h/g. Thus, the electrical charges must be
right-handed sense. The SI unit for the magnetic quantized. Still, it is one of the fundamental laws
moment is A·m2 . The intrinsic magnetic mo- of Nature that all existing electrical charges are
ments of fundamental particles arise from their multiples of the electron charge. Had the magnetic
spin. The magnetic moments of atoms are vec- monopole existed, such a law would have a nat-
tor sums of their spin and orbital magnetic mo- ural justification. No other explanation exists for
ments stemming from the spin and orbital motion the quantization of electrical charge. On assuming
of the electrons, and of the nucleons in their e as the charge of an electron whose magnitude
nuclei. The unit magnetic moments in atomic is defined by the relation e2 /(4πε0 c) = 1/137,
and nuclear physics are, respectively, the Bohr one can derive from the above equation the small-
magneton μB = e/(2me ) = 9.274·10−24 J/T, est possible magnetic charge, g0 , of a magnetic
and the nuclear magneton, μN = e/(2mp ) = monopole, called a Dirac monopole, defined by
5.051·10−27 J/T, where me and mp are, respec- the relation g02 /(4πμ0 c) = 137/4, so g0 has a
tively, the masses of the electron and the proton. value much larger than e. It follows that the track
The magnetization, which is the magnetic moment of a fast moving magnetic monopole, e.g., in a
per unit volume of a material, serves as a measure Wilson chamber, must stand out sharply on the
of the state of magnetization for macroscopic bod- background of other particle tracks, but meticu-
ies. lous searches for them have been without success.
The magnetic monopole is stable and cannot dis-
MAGNETIC MONOPOLE, Dirac monopole appear until it annihilates with another monopole
(P. Dirac, 1931) with a charge equal in magnitude and opposite in
Hypothetical particle possessing positive or sign. Searches for magnetic monopoles generated
negative magnetic charge, in analogy with electric by high-energy cosmic rays continuously incident
charge. on the Earth all turn out negative.
The laws of classical electrodynamics leave
room for particles possessing a single magnetic MAGNETIC NEUTRON DIFFRACTOMETRY
monopole to exist, and satisfy certain field equa- A collection of methods to investigate the
tions and equations of motion. These laws incor- atomic magnetic structure of materials (mostly, of
porate no postulates that forbids the possibility of crystals) based on the measurement of thermal
the existence of magnetic monopoles. neutron diffraction associated with the magnetic
772 MAGNETIC ORDER
An order–disorder type phase transition is as- angle Θ = 0, π/2 with selected axes that depends
sociated with the transition of a material from on the temperature and pressure. For instance,
the paramagnetic state into a magnetically or- a magnetic phase transition associated with the
dered state (see Magnetism). This transition is magnetization turning from the hexagonal axis
usually second order at a particular temperature towards the basal plane in the ferromagnets Co
(Curie point for ferromagnets, Néel point for an- and Gd, and the intermetallic compounds of the
tiferromagnets and ferrimagnets, freezing point for RCo5 type (R is a rare-earth element or yttrium)
spin glasses and other amorphous magnetic sub- takes place by way of two second-order phase
stances). The fluctuation region is usually rather transitions. Also, in some orthoferrites (SmFeO3 ,
broad for such phase transitions, hence the Lan- EuFeO3 , etc.), the reorientation of the weak fer-
dau theory is inapplicable; the similarity theory, romagnetic moment from the a-axis to the c-axis
the renormalization group method, the epsilon ex- proceeds by a second-order transition, and the
pansion, and the 1/n-expansion approach can be angular phase stretches over a 10–20 K region.
employed to describe these transitions. In some Realignment of L in antiferromagnets charac-
materials, e.g., in antiferromagnetic Cr and fer- terized by the Dzialoshinskii interaction may be
romagnetic MnAs, the order–disorder transition accompanied by a weak ferromagnetic moment
occurs as a clearly defined first-order one. This appearing or disappearing, as exemplified by the
is attributed to the order parameter being strongly Morin transition in α-Fe2 O3 (haematite), by the
linked with the lattice strain. The order–disorder first-order transition in DyFeO3 at 42 K, etc. Best
type of phase transition also can be changed by known among induced magnetic phase transitions
the effects of critical uctuations. are spin-flop phase transitions in antiferromagnets
There is a great variety of order–order type (see Antiferromagnetism) occurring in a magnetic
phase transitions. Thus, phase transitions involv- field parallel to the antiferromagnetism axis at a
field magnitude B ≈ (2BA BE )1/2 , where B A is
ing a change of the magnetic ordering variety
the anisotropy field, and B E is the exchange eld.
are known, e.g., the antiferromagnet ↔ ferro-
Spin-flip magnetic phase transitions correspond
magnet first-order type occurring in MnP at T =
to the transformation to parallel alignment of the
50 K and in FeRh at T = 350 K, the first-order
sublattice magnetizations with the applied field
ferromagnetic ↔ modulated magnetic structure
increasing to B ≈ B E , as well as metamagnetic
type in some rare-earth metals (see Magnetism
phase transitions. Induced magnetic phase tran-
of rare-earth metals, etc.), as well as the spin-
sitions in finite specimens may be accompanied
reorientation phase transitions associated with
by the emergence of a particular thermodynam-
changes in spin alignment. Among the latter are
ically stable domain structure: the intermediate
the spontaneous magnetic phase transitions oc- state of a magnetic substance. The fluctuation re-
curring due to temperature or hydrostatic pressure gion of these order–disorder transitions is usually
variations, and the induced magnetic phase tran- not broad, so they can be described by the Landau
sitons due to changes of the external magnetic theory.
field and occurring at a definite magnitude of the Another group of magnetic phase transitions
applied B. A phenomenological description links comprises transformations from a homogeneous
spontaneous magnetic phase transitions with a state (a paramagnetic or a magnetically ordered
change in sign of the effective anisotropy constant one) to a macroscopically non-homogeneous state
and, consequently, with the change of orienta- (having a domain structure or a long-period mod-
tion of the easy magnetization axis (see Magnetic ulated magnetic structure).
anisotropy). The direction of magnetization, M, in
ferromagnets (or M = M 1 + M 2 of sublattices MAGNETIC PHONONS
with antiferromagnetism vector L = M 1 − M 2 , in Acoustic-type excitations of an electron–ion
two-sublattice antiferromagnets and ferrimagnets) system in a metal which exist only in a strong
changes stepwise for the first-order case, while quantizing magnetic eld. The magnetic phonons
for the second-order case what are termed an- appear due to the interaction of crystal lattice vibra-
gular phases emerge, with M or L making an tions with a quantum wave whose phase velocity
774 MAGNETIC PLASMA WAVES
differs least from the sound velocity in the metal. Like an ordinary polaron (electron plus induced
Since the spectrum of quantum waves is linear in lattice polarization strain field) in an ionic crystal,
this case, the divergence of dispersion curves at a magnetic polaron is enveloped by a nonuni-
the interaction of the waves with acoustic phonons form polarization (strain) of the lattice (a phonon
does not take place in the vicinity of a single point “cloud”). Such a configuration can travel over the
(intersection), but rather over the whole of their crystal as an integral whole, and be a current
expanse. As a result, two types of tightly bound carrier (charge carrier) or an energy carrier. The
excitations exhibiting linear spectra are generated effective mass of a magnetic polaron may sub-
in a metal in the general case, termed magnetic stantially exceed that of a free electron, and the
phonons to discriminate them from other quantum self-localization of a magnetic polaron is possible
waves that only weakly interact with lattice vibra- in some cases. Depending on the sign of the ex-
tions. change interaction, the nature of the ground state
of a magnetic polaron will differ. If the exchange
MAGNETIC PLASMA WAVES integral is negative then a magnetic polaron may
A variety of weakly damped electromagnetic exist at zero temperature. Some varieties of a mag-
waves that propagate in metals in the presence of a netic polaron are: ferron, a ferromagnetic region
constant external magnetic field. Magnetic plasma enclosing a charge carrier in an antiferromagnetic
waves occur in a compensated metal when the spa- semiconductor; u ctuon, a region of magnetically
tial dispersion of the conductivity is absent, while ordered phase enclosing an electron in a paramag-
frequency dispersion exists. Underlying the prop- netic metal or alloy in the vicinity of a ferromag-
agation of magnetic plasma waves is the fact that netic phase transition; a magnon bubble which is a
the electrical conductivity in a strong magnetic field singlet state region enclosing an itinerant or local-
dramatically decreases in directions transverse to ized charge in an antiferromagnet, etc.
the field, a metal behaving like an insulator along
MAGNETIC POLE
these directions. There are two types of magnetic
A section of the surface of a magnetized spec-
plasma waves: those similar to the fast magne-
imen (body) where the normal component of the
toacoustic waves, and Alfvén waves in a plasma.
magnetization vector M n has a maximum value.
The spectrum of both types of plasma waves, as
The term North-seeking (N) or positive magnetic
distinct from the spectrum of helicons, is linear,
pole refers to the area where magnetic flux or mag-
with the velocity of the plasma wave proportional
netic induction (B field) lines exit the specimen
to the strength of the applied magnetic field. The (see Magnetic ux density), with flux lines entering
polarization of magnetic plasma waves is linear, the material at the South-seeking (S) or negative
the vector of electric intensity of the wave be- magnetic pole. In analogy with the electrostatic
ing perpendicular to the external magnetic field. case one can ascribe to magnetic poles a non-zero
In experimentally attainable magnetic fields, the surface density of magnetic charge, σ m = M n .
conditions required to observe magnetic plasma The total absence of magnetic charges (see Mag-
waves are present in semimetals with a small cur- netic monopole) in Nature means that magnetic
rent carrier density (e.g., in bismuth). induction B lines can never be discontinuous in-
side a specimen (body), so a magnetic pole of one
MAGNETIC POLARON polarity must always coexist with an equivalent
An elementary excitation of electronic origin magnetic pole of the opposite polarity on a mag-
in a magnetically ordered crystal (ferromagnet or netized specimen.
antiferromagnet) due to the exchange interaction
between the spins of an electron (hole) and of an MAGNETIC POWDER PATTERN, Bitter pattern
atom (molecule) in a crystal. A magnetic polaron (F. Bitter, 1931), Akulov–Bitter pattern
can be considered as a compound quasi-particle A picture formed by the nonuniform distribu-
constituted by an electron, a hole or an exciton tion of small ferromagnetic particles on the surface
enveloped by a region of nonuniform magnetiza- of a magnetic substance with a magnetic do-
tion of a magnetic subsystem (a magnon “cloud”). main structure. The nonuniformity is caused by
MAGNETIC RELAXATION 775
the preferable aggregation of particles in regions (see Quenched (frozen) orbital angular momen-
with maximum magnetic field strength, close to tum). From a group theory perspective the states
where magnetic domain walls reach a sample sur- might be characterized by quantum numbers as-
face, and also of domains which have a projection sociated with the irreducible representations of
of their magnetization which is normal to the sur- point groups. The parameter (index) describing the
face. F. Bitter (1931), and also N.S. Akulov and states within the given representation is an ana-
M.V. Degtyar (1932), suggested the use of powder logue of the magnetic quantum number (mk ) for
patterns formed by applying ferromagnetic parti- a crystal; it may take on 2lk + 1 values, where lk
cles to the surface of a ferromagnet in the form of is the dimension of the associated irreducible rep-
a liquid suspension, for magnetic domain charac- resentation.
terization. A magnetic powder pattern can detect
the distribution of the ends of vortices in super- MAGNETIC RELAXATION
conductors. Process of attaining equilibrium in the spin
subsystem of a body. Because of the weak cou-
MAGNETIC PRESSURE pling of the spins of atoms and subatomic particles
An action exerted by a frozen-in magnetic eld with the particle motion, equilibrium within the
on a conducting liquid (e.g., on a plasma, includ- spin system of a magnetically ordered medium
ing a solid-state plasma) and directed perpendicu- (ferromagnets and antiferromagnets) is reached,
lar to the field lines. The magnetic pressure, pM , as a rule, prior to the entire body coming to the
is proportional to the square of magnetic flux den- state of equilibrium. A spin temperature can be
sity, B, viz., pM = B 2 /2μ (in SI units). It can be assigned to the spin subsystem under such condi-
counterbalanced by the kinetic plasma pressure. If tions (see Level population).
the magnetic pressure exceeds the kinetic one, a The magnetic relaxation becomes more in-
discharge compression (known as the pinch-effect) volved due to the forces of different origin acting
occurs; it was first described by W.H. Bennett between the spins. The exchange forces (see Ex-
(1934). The term was introduced by L. Tonks change interaction), largest in magnitude, cannot
(1937). Magnetic pressure is a crucial factor in the change the mean magnetic moment of the system,
operation of a tokamak nuclear fusion reactor. even if its value is a nonequilibrium one, but these
forces can readjust individual spin directions to
MAGNETIC QUANTUM NUMBER (m, M ) equilibrate the spin subsystem temperature. Spin
A quantum number characterizing eigenvalues interaction forces (see spin–orbit interaction, mag-
of the projection of the angular momentum opera- netic dipole interaction, etc.) are responsible for the
tor J along a magnetic field direction. For orbital relaxation of the mean magnetic moment, so its
motion the projection of the orbital angular mo- individual components can return to equilibrium
mentum L, in the units of , can take on the with different rates.
2L+1 integer values mL = 0, ±1, . . . , ±L, where Magnetic relaxation of the magnetic moment
L is the orbital quantum number. For the case of (magnetization) component perpendicular to the
spins a spin magnetic quantum number, mS , is in- applied magnetic field in paramagnets is associ-
troduced. In the case of an integer spin S, mS may ated with the spin–spin interaction, and is known
equal 0, ±1, . . . , ±S, and for half-integer spin mS as the spin–spin relaxation time τ2 (transverse
takes on the values ±1/2, ±3/2, . . . , ±S/2. relaxation time), while relaxation of the longitu-
The magnetic quantum number is used in op- dinal component (component parallel to the field)
tics, microwave, and radiowave spectroscopy to of the magnetization to the lattice is known as
label energy levels, and to identify transitions the spin–lattice relaxation time τ1 (longitudinal
between them. The hybridization of states with relaxation time). Typically, τ1 τ2 , with differ-
different m values may occur in crystals because ent mechanisms operating, so the magnitudes and
of the atomic level splitting in the crystaline fields, the temperature dependences of the two relaxation
and sometimes there is complete or partial orbital times generally differ from each other. The mag-
angular momentum decoupling or cancellation netic relaxation of nuclear spins is conditioned
776 MAGNETIC RESONANCE
by their relatively weak interactions both with (EPR). Ordinarily EPR resonant frequencies ex-
other degrees of freedom in the solid, and with ceed NMR frequencies by three orders of magni-
each other. Therefore, the time constant of nuclear tude. In magnetically ordered materials we have
magnetic relaxation is generally much longer than ferromagnetic resonance, antiferromagnetic reso-
other magnetic relaxation times such as those due nance and ferrimagnetic resonance, respectively.
to electronic spins (see Nuclear quadrupole relax- Magnetic resonance imaging in medicine is an im-
ation). portant application of NMR.
Magnetic relaxation affects magnetization and
remagnetization processes (see Magnetic viscos- MAGNETIC RESONANCE IMAGING (MRI), MR
ity), and it determines the width of a magnetic tomography, NMR endoscopy (title in Russian)
resonance line, as well as the magnetic suscep- A set of methods for investigating the in-
tibility dispersion. Magnetic relaxation puts limits ternal structure of subjects by utilizing nuclear
to the employment of magnetic materials in en- magnetic resonance. The resonance frequency
gineering and in physical experiments. Since this ω0 and permanent magnetic field B0 in NMR
relaxation (like other relaxation processes) de- are related by the expression ω0 = γ B0 (here
pends on the structure of the solid and on the γ is the gyromagnetic ratio of the resonant nu-
presence of dislocations and other defects (espe- clei). If B0 is spatially nonuniform, i.e. B0 =
B0 (x, y, z), then also ω = ω0 (x, y, z). This means
cially paramagnetic ones), the magnetic relaxation
that each resonance signal with the frequency
time can be controlled by technological treatments
ω0 (x, y, z) arises from a particular volume el-
(alloying, quenching, etc.).
ement dV that produced this signal. However,
according to Maxwell’s equations there exist two-
MAGNETIC RESONANCE
dimensional surfaces for any configuration of
Absorption of electromagnetic waves by mag-
B0 (x, y, z) where the value B0 (and consequ-
netic moments precessing in an external magnetic
ently, ω0 ) does not vary. The reconstruction of
field. The absorption involves transitions
E be- the spatial structure of a subject under study is
tween Zeeman energy splittings in an applied not simple. This problem was first resolved by
magnetic field B, with the resonance condition ex- P.C. Lauterbur (1973) who, using the principles
pressed as
E = ω = gμB B for electron spins, of tomography, developed and implemented the
and as ω = γ B for nuclear spins, where g is the di- method of obtaining NMR images, a procedure
mensionless g-factor, γ is the gyromagnetic ratio, later given the name magnetic resonance imaging
and gμB = γ . The dynamics of the phenomenon (MRI).
is described by the Bloch equations for the time A principle of spatial focusing of magnetic res-
dependence of the magnetization vector M: onance, which in various versions is often applied
dMx Mx in magnetic resonance spectroscopy and for pur-
= γ (M × B)x − , poses of NMR, was first performed by R. Dama-
dt T2
dian (1976). The essence of the focusing was the
dMy My creation of such a configuration of the field B0
= γ (M × B)y − ,
dt T2 that it has sufficient uniformity and satisfies the
dMz Mz − M0 resonance condition only over a limited region of
= γ (M × B)z − , space, with nuclei beyond this region either out
dt T1
of resonance or broadened beyond detection. By
where M0 is the equilibrium magnetization in the scanning over a sequence of these uniform reso-
absence of an applied radiofrequency signal, and nant regions, one can study sequentially the spatial
T1 and T2 are, respectively, the spin–lattice and characteristics of the subject. Later more refined
spin–spin relaxation times. This energy absorption and sophisticated methods of preparing systematic
by nuclei is termed nuclear magnetic resonance magnetic field gradiants, and thereby obtaining
(NMR), and the corresponding absorption by mag- improved NMR images, were developed.
netic moments of unpaired electrons in paramag- The resolution of MRI is determined by the
nets is termed electron paramagnetic resonance line width, and by the magnitude and regularity
MAGNETIC RESONANCE SPECTROMETERS 777
of the gradient of the field B0 . In the first exper- frequency. Spectrometers operating in the 3 cm
iments of Lauterbur, the resolution was 0.2 mm. (X), 1.5 cm (K) and 8 mm (Q) bands find the
MRI brought substantial changes to the methods widest use, although both longer and shorter wave-
of medical diagnostics. The application of MRI to length (down to 2 mm) instruments are available.
solids is limited by the appreciable (compared to These spectrometers are also used to record ferri-
liquids) width of the NMR lines, and this has led magnetic resonance, ferromagnetic resonance and
to the study of model subjects. The concepts of antiferromagnetic resonance in which the sample
the MRI have been adopted by other areas of mag- shape should be specified explicitly (usually a thin
netic resonance spectroscopy (see EPR imaging, rectangular plate or a sphere). For these studies it
Tomography). is advantageous to vary the temperature through
the phase transition point.
MAGNETIC RESONANCE LINE NARROWING In the case of NMR radiospectrometers, the
See Narrowing of magnetic resonance line. analogue of the absorbing cavity is a solenoid in
the generator tuned circuit. where the specimen
MAGNETIC RESONANCE, PHONON
under study is placed. The solenoid is positioned
See Phonon magnetic resonance.
in an external magnetic field B 0 so that the ra-
diofrequency field B 1 is perpendicular to B 0 .
MAGNETIC RESONANCE SPECTROMETERS
During resonance, one records a small voltage
Instruments for recording the spectra of mag-
drop in the tuned circuit. An increase in sensi-
netic resonance spectroscopy. The principal spec-
tivity is attained by modulating B 0 and operating
trometer elements include a source of radiofre-
quency or microwaves, an absorbing cell (coil, in a synchronous detection mode. To study NMR
cavity), and a detector. As a rule, magnetic reso- in solids, a number of high-resolution designs are
nance spectrometers also include sources of mag- used in which the sample is rotated rapidly at the
netic fields. They are most often used to observe “magic” angle. Also in use are multipulse tech-
and study the phenomena of electron paramag- niques to accumulate signals, average over various
netic resonance (EPR), nuclear magnetic reso- NMR broadening mechanisms, and then retrieve
nance (NMR), EPR with optical detection of mag- the spectrum after its Fourier transformation.
netic resonance (ODMR), and electron–nuclear When studying ODMR various optical devices
double resonance (ENDOR). are used to detect the resonance. They measure the
EPR radiospectrometers are usually designed angle of rotation of the polarization plane (Fara-
so that the microwave absorption in a specimen day effect) or the magnetic circular dichroism.
is observed while scanning a modulated magnetic Both effects are sensitive to the passage through
field (or frequency). Various instrumental designs the resonance magnetic field. ENDOR spectrom-
are available. In the simplest direct absorption eters are EPR spectrometers with an additional
EPR spectrometer the power from a microwave radiofrequency field introduced into their cavities
source (a klystron) is fed through a waveguide to to induce transitions between the closer spaced nu-
a resonator (cavity) containing the substance un- clear sublevels (NMR levels). The ENDOR spec-
der study and located in a strong magnetic field, trum is obtained by scanning the NMR frequency
and the wave passing through (or reflected from) band while recording the EPR response. A nu-
the cavity goes to the detector. The resonance is clear double resonance radiospectrometer has an
identified from the change in the detector current. absorbing cell similar to those used in NMR. Its
The spectrometers use magnetic field modulation circuit operates at a tunable frequency, with pulses
at frequencies from tens of Hz to 1 MHz, and fed into the cavity to observe spin echoes. Scan-
often employ synchronous detection to signifi- ning the frequency band and observing the echo
cantly improve the sensitivity. Another type of with the help of superheterodyne signal detection
EPR spectrometer is a superheterodyne one in provides nuclear double resonance spectra. Many
which a separate microwave source of a different radiospectrometers include Fourier analysis and
frequency is introduced and the signal amplifica- signal accumulating units. EPR and NMR are also
tion is carried out at an intermediate difference used to measure relaxation times. For continuous
778 MAGNETIC RESONANCE SPECTROSCOPY
wave (CW) saturation the EPR spectrometer is fit- EPR and NMR by far the most widely utilized. In-
ted with a microwave generator covering a broad struments used for magnetic resonance are called
power range to obtain a series of EPR spectra at magnetic resonance spectrometers.
various levels of saturation. These data provide The high information yield of these measure-
relaxation times. Alternatively, a pulse saturation ments arises from the high sensitivity and resolu-
technique uses high power pulses to saturate the tion of the spectrometers, the capability of mea-
specimen, and relaxation times are obtained from suring spacings between the closely lying energy
the time dependence of the EPR signal that follows levels, and the ability to analyze the line shape
each pulse. One may also use multipulse tech- with high precision. It is possible to identify the
niques to record spin echoes. Relaxation times in local symmetry of paramagnetic centers, and to
NMR spectroscopy are ordinarily measured using deduce the crystallographic site and its neighbor-
multipulse techniques. hood. Hence one may understand the nature and
The sensitivity of radiospectrometers depends define the number of paramagnetic centers and de-
on many factors. EPR spectrometers achieve a fects surrounding them. Thus, magnetic resonance
sensitivity of 1010 –1011 spin at a line width of is a technique for crystal defectoscopy. It is also a
0.1 mT. NMR spectroscopy studies of solids often method for studying phase transitions in materials
deal with rare isotopes reaching only 1% of the that undergo electronic and molecular ordering,
basic nuclei. Many spectrometers are fitted with and to determine the characteristics of paramag-
additional devices to obtain spectra at high and netic centers and their environment (the spin den-
low temperatures, at high pressure, during irradia- sity distribution, effective magnetic and electric
tion, etc. dipole moments, electric eld gradients, nuclear
magnetic moments, etc.). Magnetic resonance can
Other magnetic resonance spectrometers in-
also provide information on band structure para-
clude those used in nuclear quadrupole resonance
meters and the nature of chemical bonds. One may
(NQR), and magnetic resonance imaging (MRI).
study tunneling phenomena in solids with its help.
The latter technique, originally referred to as nu-
Magnetic resonance prompted the development of
clear magnetic resonance imaging, is widely used
masers, and then lasers, which became solid-state
in medicine for diagnosing cancer and other dis-
instruments, and eventually MRI or magnetic res-
eases. It employs a strong constant magnetic field
onance imaging in medicine.
provided by a superconducting magnet, with a
superposed magnetic field gradient which makes
MAGNETIC RESONANCE, SPIN PHONON
the protons in the various parts of the body un-
See Spin phonon magnetic resonance.
der examination come to resonance at slightly
different frequencies. It is capable of providing
MAGNETIC RESONANCE, ZERO-FIELD
three-dimensional images of the brain and other
See Zero- eld magnetic resonance.
body organs.
MAGNETIC REVERSAL, remagnetization
MAGNETIC RESONANCE SPECTROSCOPY of The process of decreasing the magnitude and
solids reversing the direction of the magnetization M of
A part of the physics of solids dedicated to a magnetically ordered sample, which takes place
studying substances by their spectra in the radio at the reversal in direction of the external magnetic
and microwave frequency ranges, 106 –1012 Hz. e ld B app which had previously magnetized the
The majority of these techniques involve elec- material to saturation. In multi-domain ferromag-
tron paramagnetic resonance (EPR), nuclear mag- nets and other materials exhibiting spontaneous
netic resonance (NMR), ferromagnetic resonance, magnetization, the reversal process takes place
antiferromagnetic resonance, ferrimagnetic reso- through the generation and growth of magneticly
nance, nuclear quadrupole resonance, cyclotron reversed regions, followed by the redistribution
resonance, paraelectric resonance, and various of magnetic phases through domain wall displace-
types of double and even triple resonances, with ments. As this occurs, those phases with their
MAGNETIC SEMICONDUCTORS 779
magnetization pointing closest to the B app direc- magnet). One can easily control EuSe proper-
tion grow in volume. The final stage of magnetic ties by external factors: a moderate pressure or
reversal involves the complete alignment of M in a magnetic field transforms the antiferromagnetic
these domains in the B app direction (see Domain ordering into a ferrimagnetic or ferromagnetic
wall displacement processes). Magnetic reversal type (magnetoisotropic metamagnetic, see Antifer-
of small magnetically ordered particles may pro- romagnetism). Unique EuSe properties are due to
ceed through irreversible jumps of M followed by higher-order exchange interaction terms compara-
reversible alignment of M with B app . ble to the bilinear Heisenberg exchange.
The magnetic properties of pure magnetic
MAGNETIC SATURATION semiconductors are, due to a small intrinsic current
The state of a material with the maximum carrier density, n, similar to those of insulators.
possible magnetization value, M ∞ , called the sat- A substantial difference only emerges via doping
uration magnetization, that undergoes no further with electrically active impurities. In non-degene-
changes with increases in the applied magnetic rate ferromagnetic semiconductors, the electrons
field. In the case of ferromagnets, M ∞ is attained of non-ionized donors enhance the local ferro-
at the completion of several processes: (a) the magnetic ordering at T = 0 by way of indirect
growth of the domains with their magnetic mo- exchange between the localized d- or f -moments
ments aligned along an easy magnetization axis as in the neighborhood of defects. As a result, the
a result of the domain wall displacement process, paramagnetic Curie point, Θ, may be consider-
(b) rotation of the magnetization vectors towards ably raised, even though the magnetic disordering
the direction of the applied field (magnetization ro- temperature, T c , does not substantially change.
tation process), and what is called a paraprocess, Donors added to antiferromagnets may even cause
viz., the growth of the number of spins aligned Θ to change sign, although the antiferromagnetic
along the field under the action of a strong applied ordering remains.
field, at the expense of those spins with antiparallel In degenerate ferromagnetic semiconductors,
orientations (see Magnetization curves). In prac- the donor electrons get delocalized to effect indi-
tice, this saturation is obtained in fields from a rect exchange over the whole crystal; thus T c may
few millitesla up to 10 T at 20 ◦ C. For antiferro- be appreciably increased (doubled in EuO and
magnets and ferrimagnets magnetic saturation is EuS at n ≈ 5·1026 m−3 ). Unlike the case of met-
reached at a much higher magnitude of the ap- als, indirect exchange in magnetic semiconductors
plied magnetic field, close to that of the exchange cannot be described by a Heisenberg Hamiltonian
field. For some paramagnets, a state approaching because of the relatively low Fermi energy.
magnetic saturation is reached with a field ∼1 T at Distinctive electrical and optical properties of
temperatures of ∼1 K. magnetic semiconductors are also influenced by
the strong exchange interactions of current carriers
MAGNETIC SEMICONDUCTORS with localized d- or f -moments. Therefore, these
Semiconductor materials having as their main properties are highly sensitive to magnetic order-
components ions with partially filled d- or ing, even being qualitatively different in many
f -orbital levels whose magnetic moments be- cases from the properties of non-magnetic semi-
come ordered at T → 0. Almost all known types conductors. For instance, a huge long-wave (red)
of ordering can occur in magnetic semiconduc- shift of the intrinsic light absorption edge occurs in
tors. For instance, there is ferromagnetic ordering ferromagnetic semiconductors at decreasing tem-
in EuO and EuS, antiferromagnetic ordering in perature or at increasing magnetic field strength B,
EuFe, and helical ordering in HgCr2 S4 . Mag- amounting to 0.5 eV in HgCr2 Se4 . Such a shift
netic polymorphism is found in EuSe: the two- results from a decrease of the current carrier en-
sublattice antiferromagnetic structure is replaced ergy at the onset of ferromagnetic ordering. The
by a three-, and then by a four-sublattice one temperature dependences of the photoconductiv-
with increasing temperature (the latter two are for- ity edge and light absorption edge are similar
bidden in the Heisenberg model, see Heisenberg in ferromagnetic semiconductors. Typically, the
780 MAGNETIC SEMICONDUCTORS
absorption edge temperature shift is weak and the conduction electrons in non-degenerate an-
occurs toward the short-wave side in antiferro- tiferromagnetic semiconductors create, indepen-
magnetic semiconductors. An enormous sponta- dently of each other, ferromagnetic microregions
neous Faraday rotation (see Faraday effect) takes and become self-localized in them. Such regions
place in ferromagnetic semiconductors, exceeding may consist of a few thousand atoms under the
in EuS, e.g., a value of 106 deg/cm. The Fara- most favorable conditions. These quasi-particles
day quality factor also reaches gigantic values in (ferrons) are destroyed by the magnetic field that
the ferromagnetic semiconductors, amounting to converts the whole crystal into the ferromagnetic
104 deg/dB in EuO. The n-type ferromagnetic state. Since the ferrons are immobile, their de-
semiconductors, unlike the non-magnetic ones, struction is accompanied by an increase in the
display, as a rule, a temperature dependent re- crystal conductivity. The existence of ferrons in
sistivity, ρ: with a high ρ peak in the vicinity non-degenerate EuSe and EuFe is proven by ex-
of T c . Ferromagnetic semiconductors exhibit a perimental data on the B dependences of their
large negative magnetoresistance in the neigh- photoconductivity and luminescence.
borhood of this peak. In non-degenerate ferro- Collective ferron states are possible in degener-
magnetic semiconductors, the ρ peak is mainly ate antiferromagnetic semiconductors, consisting
governed by the difference in the temperature de- in the spontaneous division of the crystal into fer-
pendences of the donor level depths and of the romagnetic and antiferromagnetic regions, with
position of the conduction band bottom. Since the the conduction electrons passing from the antifer-
local magnetic ordering at a non-ionized donor romagnetic crystal portion into the ferromagnetic
is higher than that averaged over the crystal, the one. With n not too high, the ferromagnetic part
consists of separate droplets, each containing a
donor level rises slower than the band bottom
considerable number of electrons. With n ris-
with temperature, with the distance between them
ing, the number and dimensions of ferromagnetic
reaching its maximum not far from T c . Accord-
droplets grow, and, starting from a certain den-
ingly, n is at its minimum there.
sity np , the ferromagnetic droplets start touching
The ρ peak in degenerate ferromagnetic semi-
one another to form a simply connected ferro-
conductors is due to the temperature dependent
magnetic region. At n < np , the crystal behaves
scattering of conduction electrons (see Current
like an insulator, as all the electrons are trapped,
carrier scattering) from ionized donors. The ran-
each in its own droplet, and cannot carry current.
domly positioned donors give rise to a non- Electron percolation occurs at np , i.e. the crystal
uniformity in the conduction electron density dis- behaves like a metal for n > np . When converted
tribution, and, consequently, to a non-uniformity to the homophase state (by increasing the temper-
of the indirect exchange interaction mediated ature to make the whole crystal paramagnetic, or
through them. A component of static magne- by increasing the magnetic field to make the crys-
tization fluctuating in space appears, with the tal ferromagnetic), a heterophase insulator (n <
maximum scattering on this component occur- np ) simultaneously undergoes a transition into the
ring near T c . A bend, instead of a peak, at the metallic state. At n > np , the transition into the
Néel point on the dependence of ρ(T ) on the homophase affects ρ relatively weakly. Such ten-
temperature is found for pure antiferromagnetic dencies were observed in both degenerate EuSe
semiconductors, while dρ/dT displays a singular- and EuFe.
ity of the speci c heat type. A related phenomenon was discovered in the
Magnetic semiconductors are the only mate- degenerate ferromagnetic semiconductor EuO in
rials where the heterophase self-localization of a narrow n interval: it displays metallic type
current carriers has been observed; the phenom- conductance at T → 0; with increasing tempera-
enon consists in the conduction electrons forming ture, a transition to an insulating state occurs, the
regions of normal unstable phase and stabilizing jump in ρ amounting up to 17 orders of mag-
them by their localization inside such regions. nitude. The high conductivity of ferromagnetic
That occurs when the electron energy is lower in semiconductors at T → 0 ensues from their inabil-
the unstable phase than in the stable one. Hence, ity to contain ferrons at T = 0. However, when
MAGNETIC SOLITON 781
the temperature rises, with the ferromagnetic or- with an equilibrium magnetization. These objects
dering disrupted, the electrons reveal a trend to determine the contribution of the magnetic de-
self-localization into regions with enhanced ferro- grees of freedom to the scattering of penetrating
magnetic ordering. Therefore, the electrons pass radiation, as well as the thermodynamics of quasi-
from delocalized levels to localized levels on the one-dimensional magnetic substances.
donors, with an enhanced ferromagnetic ordering In many cases a magnetic soliton behaves like
taking place in their neighborhood (see ferrons). a particle with an effective mass. A domain wall
is an analogue of an unstructured particle, and a
MAGNETIC SHIELDING bion is an analogue of a particle having an internal
Protection of a volume of space (or an ob- degree of freedom (precession of the vector M).
ject) against penetration by a magnetic eld (as A quantum analogue of a bion is the bound state
well as of its variations) arising from sources lo- of a large number of magnons, elementary excita-
cated outside this volume (or object). Achieved tions of a magnetic material (see Magnon bubble).
with no expenditure of electric energy, magnetic If E and P are the energy and the momentum
shielding is implemented by isolating the space to of a bion, respectively, and N is the number of
be shielded using special envelopes called mag- magnons localized at it, then,
netic shields. The basic principle behind magnetic
shielding is the compensation (within the pro- dE = v dP + ω dN,
tected space) of the external magnetic field by
an oppositely directed field produced by mag- where v and ω are the velocity and the frequency,
netic poles that develop along the shield edges at respectively, of the bion internal precession. The
the places where the external field enters or ex- bion frequency is always below the limit of lin-
its. This compensation causes the magnetic field ear spin wave frequencies, and is dependent on N .
in the shielded space to be much smaller than it The requirement ω dω/dN < 0 is usually the nec-
is without the magnetic screen. With increasing essary condition for bion stability.
frequency of an applied electromagnetic field the We can distinguish topological magnetic soli-
magnetic shielding transforms into electromag- tons whose stability is governed by the internal
netic shielding because of the energy expended on structure of the magnetization (e.g., a 180◦ do-
generating eddy currents (see Magnetic loss) in the main wall in an easy-axis ferromagnet), and dy-
screen. namic magnetic solitons whose stability depends
on the existence of phase integrals of the vector M
MAGNETIC SOLITON field equations (total energy, momentum, projec-
A soliton of the field of magnetization, M, tion of momentum onto anisotropy axis, etc.). The
in magnetically ordered crystals; a solitary solu- bion is a dynamic magnetic soliton.
tion of the nonlinear Landau–Lifshits equations. A two-dimensional magnetic soliton may ex-
The simplest magnetic soliton in an easy-axis fer- ist in the form of a cylindrical magnon bubble
romagnet is a magnetic domain wall, and in an (dynamic magnetic soliton), or of a magnetic vor-
easy-plane magnetic material, it is a rotation wave tex (topological magnetic soliton) with a topo-
or helical vortex. A magnetic soliton in the most logical charge determined by the circulation of
restricted sense (sometimes called a bion) is asso- the magnetization vector around the soliton axis.
ciated with a uniform distribution of magnetization A magnetic vortex in an easy-plane ferromagnet
well away from the soliton (to the left and to the is analogous to a vortex in superfluid 4 He, while
right of the soliton in the one-dimensional case). that in an easy-axis one has no analogues in fluid
A so-called bion is characterized by two para- physics, exhibiting a specific ω versus N depen-
meters: the velocity of the center of gravity, and dence, and a sensitivity to the topological charge.
the frequency of the internal precession of the A particular type of two-dimensional magnetic
vector M. Domain walls, bions and spin waves soliton is the Bloch line separating the segments of
constitute independent nonlinear modes of mag- a planar domain wall having different directions of
netic dynamics of a one-dimensional ferromagnet the turning vector M.
782 MAGNETIC SPECTROSCOPY
length l of the wave function decay into the bulk the difference χ d − χ rev may serve as a gauge
crystal. of the irreversibility of processes associated with
magnetic hysteresis (see Paramagnetism, Diamag-
MAGNETIC SUSCEPTIBILITY netism and Ferromagnetism).
A physical quantity characterizing the ability The magnetic susceptibility is a complicated,
of a material to change its magnetic state (mag- and sometimes hysteretic, function of the temper-
netization) with a changing applied magnetic eld. ature. For instance, χ rises with T in ferromag-
Static and dynamic magnetic susceptibilities are nets, reaching a maximum (known as the Hop-
known, the latter differing from the former for kinson maximum) near the Curie point, and then
media exhibiting temporal dispersion. The static decreases over the paramagnetic region, follow-
susceptibility, χ = M/H , is the ratio of the mag- ing the Curie–Weiss law. An additional maximum
netization, M, to the internal magnetic field, H . of χ occurs in the neighborhood of various phase
It is related to the magnetic permeability, μ, by transitions, e.g., the reorientation-type magnetic
μ = μ0 (1 + χ) in SI units (μ = 1 + 4πχ in CGS). phase transition in nickel, etc. Magnetic suscepti-
The dynamic susceptibility χ(ω) or χ(q, ω) re- bility is a structure-sensitive quantity in ferromag-
lates, in a similar manner, the Fourier components nets, so its magnitude may vary over a wide range
of the magnetization to the magnetic field (see for one and the same material, depending on its
Susceptibility). Static magnetic susceptibilities can heat pretreatment. For anisotropic media, e.g., for
differ in both magnitude and sign for different magnetically ordered crystals where the magnetic
magnetic substances. For diamagnets (χ < 0) and anisotropy is taken into account, the susceptibility
paramagnets (χ > 0), the susceptibility is small, is a tensor quantity χik depending on the direction
of the order 10−4 –10−6 , and it depends weakly on in the material. The static susceptibility is a sym-
the external magnetic field. In ferromagnets and metric real-valued tensor, and the dynamic suscep-
ferrimagnets, the susceptibility may be as high as tibility χik (q, ω) will, in a nondissipative medium,
104 and strongly dependent on H . For that reason, be described by a Hermitian tensor, with the anti-
the differential susceptibility, χ d = dM/dH , is Hermitian part determining the absorption of the
often introduced to describe properties of magneti- alternating magnetic field energy. The χik (q, ω)
cally ordered materials. This quantity χ d is a prop- tensor components obey general expressions fol-
erty of the material that depends strongly on H ; lowing from Kramers–Kronig relations and the
a dependence that is highly nonlinear for actual Onsager relation (see Onsager method). Compo-
ferromagnets which exhibit a domain structure nents of the χik (q, ω) tensor have singularities
(see Saturation magnetization curve). At H = 0, (poles) at ω values close to natural oscillation fre-
χ d has a certain (nonzero) magnitude called the quencies, ω(q), of the magnetization (spin waves,
initial susceptibility. The differential susceptibil- magnetostatic oscillations, etc.).
ity grows with H , reaches a maximum, and then
drops to a very small value with the onset of sat- MAGNETIC SWITCHING
uration at high fields (saturation effect). In such Process of irreversibly changing the magneti-
a high field, the differential susceptibility differs zation of a magnetic substance, with the change
little from that of a paramagnet, and it is known remaining after the remagnetizing field has been
as the paraprocess susceptibility. The so-called switched off. Magnetic switching is mainly due
reversible magnetic susceptibility, χ rev , is often to jump-like displacements of the magnetic do-
singled out: main walls (see Barkhausen effect) and the re-
ordering of the magnetic domain structure. Each
M
χ rev = lim (
H < 0). Barkhausen jump is an elementary act of magnetic
H →0
H
switching, while the magnetic hysteresis loop is
Within a sufficiently small
H in the vicinity of the overall characteristic of magnetic switching.
any H the value χ rev is determined mainly by re- The basic parameters of magnetic switching (min-
versible processes occurring in the magnetically imal force of switching, B s , total polarization
ordered material with the variation of H . Hence, change,
M s , time of magnetic switching, τ s )
786 MAGNETIC SYMMETRY GROUP
An operator magnetic symmetry group is iso- lattice site to another, thus changing the symmetry
morphic to the crystallographic one in coordinate of the local defect environment symmetry. In ad-
function space, while in spinor space two oper- dition the mechanical defect strain field interacts
ators, differing in sign, are associated with each with the magnetization (magnetostriction). In the
crystallographic operation. latter case of defects diffusing in the lattice over
distances of the order of the domain wall thick-
MAGNETIC TEXTURE ness, the relaxation time τ depends on the rate
A preferential space orientation of the easy of diffusion and the type of diffusion (electronic,
magnetization axes in a polycrystalline ferro- or ionic, etc.), and it rises drastically with decreasing
ferrimagnetic specimen, which is indicative of the temperature; e.g., the carbonyl iron has τ ≈ 103 s
specimen magnetic anisotropy. Magnetic texture at −2 ◦ C, and τ ≈ 10−2 s at +100 ◦ C. The mag-
appears under: (i) application of directional me- netic viscosity can be quantitatively described by
chanical stresses to the specimen, which brings introducing lagging terms (generally, nonlinear
about a preferential crystallite orientation (a crys- in domain wall displacements) into the magnetic
tallographic texture); (ii) heat treatment of the domain wall dynamics equation (Néel equation
specimen below the Curie point in the presence of and Slonczewski equation). The magnetic viscos-
a magnetic field (thermomagnetic treatment), and ity is particularly high in hard magnetic materials
(iii) thermomechanical treatment. The presence that are ensembles of single-domain particles, e.g.,
of magnetic texture can dramatically improve the grains in polycrystalline alloys interacting with
magnetic properties in some magnetic materials. one another. This viscosity is governed by the ir-
Classed into magnetically-textured soft magnetic reversible magnetization rotation of such domains
materials are: crystallographically textured trans- (see Magnetization rotation processes) due to ther-
former steel, perminvar, permalloy, etc. Giving a mal fluctuations. Some materials exhibit quite a
magnetic texture to these materials decreases the range of lag times, τ , that can be determined
magnitude of their coercive force and lowers the from the number and positions of peaks appear-
magnetic loss. In hard magnetic materials (mag- ing on the χ frequency dependence excited by
nico, ticonal, barium and cobalt ferrites, etc.), the a harmonically varying applied magnetic field.
magnetic texture results in an increase of their co- Magnetic viscosity appears in spin glasses and
ercive force, remanence, etc. other nonergodic magnetic materials (see Amor-
phous magnetic substances).
MAGNETIC VISCOSITY
A measure of the time lag of magnetization MAGNETISM of solids (fr. Gr. μαγ νητ ης , of
changes in magnetic substances after variations Magnesia)
in the applied magnetic eld. Due to the mag- Branch of physics that studies magnetic sub-
netic viscosity, the steady-state magnetization is stances or materials containing magnetic mo-
reestablished after a lag in time τ that can range ments, generally arising from their constituent el-
from less than a millisecond to a few hours (see ements (atoms, molecules, ions, conduction elec-
Magnetic relaxation). For example, it takes time trons), including their interactions with each other
for magnetic domains to grow and reorient. Mag- or with external magnetic elds and external elec-
netic viscosity is also observed in magnetic mate- tric currents.
rials with internal degrees of freedom that result Basics. An electric charge e moving in a vac-
from the presence of mobile impurities and other uum at the velocity v in the presence of a magnetic
defects in crystals interacting with the magnetic ux density (sometimes called magnetic induc-
subsystem. There may be, e.g., a relaxation drift of tion) B experiences the Lorentz force F given by
the magnetic susceptibility, χ , in ferromagnets (or F = ev × B, so B is sometimes called the force
ferrites) towards its steady-state value through do- per unit current. In free space we also define the
main walls interacting with reorienting or diffusing associated magnetic field H through the expres-
defects. In the former case the magnetic anisotropy sion B = μ0 H where μ0 = 4π·10−7 N/A2 (or
energy changes due to defects jumping from one H/m) is the magnetic permeability of vacuum.
788 MAGNETISM
In a medium the difference between B and H spin moments of unfilled d- and f -orbital shells
involves the presence of magnetic field sources in atoms of the Fe, Pd or Pt groups, rare earth el-
in the matter, viz. of moving electric charges ements, or actinides) are of major importance for
(atomic electrons, conduction electrons, etc.) that magnetically ordered substances. For example, the
give rise to microscopic currents (distinguishing exchange interaction, of electrostatic origin with
between micro- and macroscopic currents is some- energy E e = k B θ W (where θ W is the Weiss tem-
what arbitrary), and of intrinsic (spin) moments perature, k B is the Boltzmann constant) of the
of the substance particles: electrons and atomic order of 10−20 –10−21 J, i.e. θ W = 1000–100 K,
nuclei. This relation between B and H is given results in magnetic ordering when the spin orien-
by B = μ0 (H + M) in SI units (B = H + 4πM tations of neighboring atoms become correlated.
in CGS units) with the magnetization vector, M, This ordering occurs at a magnetic phase transi-
defined as the mean magnetic moment per unit tion at a temperature T c of the order of E e /k B ,
volume. The theory of magnetism aims at es- causing an atomic magnetic structure to appear
tablishing interrelations between B, H and M in the material associated with a non-zero value
in different magnetic materials under given ex- of the mean atomic spin, even in the absence of
ternal conditions (either steady-state, or time- an external magnetic field, B. There is a great
dependent). The magnetization for a given H or B variety of magnetically ordered materials. Their
is defined by the material equations of a magnetic main feature is the nature of the motion of the
material, which result from (quantum-mechanical magnetic carriers, electrons; the magnetism of lo-
and statistical) averaging of the microscopic equa- calized electrons, and that of delocalized electrons
tions of motion for the sources of the magnetic (see Band magnetism). In the simplest case of fer-
field. More precisely, by virtue of the Bohr–van
romagnetism, the spins of all the atoms (or of
Leeuwen theorem, the magnetic properties of a
all the delocalized electrons contributing to ferro-
substance can be consistently described within
magnetism) acquire parallel orientations to form
a quantum-mechanical framework. Nevertheless,
a spontaneous magnetization vector M s (see also
ordinarily simple phenomenological equations are
Ferromagnetism of metals and alloys). In antifer-
found to be adequate for describing the magne-
romagnets the spins of adjacent atoms acquire an
tization vector as a function of the coordinates
antiparallel orientation, however, linear combina-
and time: M = M(r, t) (see Bloch equations and
tions of the spins (antiferromagnetic vectors L)
Landau–Lifshits equation). In the static (or quasi-
static) case, the relation between M and B (or H ) can still differ from zero. In ferrimagnets, both M
is determined by the magnetization curve M (H ) and L are non-zero. For a two sublattice system
or M (B), which differs in shape for different ma- with individual sublattice magnetizations M 1 and
terials. In a number of cases, M is linear in H : M 2 the total magnetization is M = M 1 +M 2 , and
the antiferromagnetism vector is L = M 1 − M 2 .
M = χH , (3) Magnetic ordering of a more complex type is
where χ is the magnetic susceptibility, a tensor exemplified by spiral (helical) ordering (see Mod-
quantity for anisotropic bodies. But sometimes, ulated magnetic structures). Amorphous magnetic
e.g., for ferromagnets, the linearity of Eq. (1) may materials (e.g., spin glasses) are also classified
become invalid even in very weak fields. among the ordered magnetics. Typical of magnet-
Classification of magnets. The classification ically ordered materials is a complicated shape of
proceeds from the analysis of the nature of the their magnetization curve M(B), which can ex-
atomic carriers of magnetism, and their inter- hibit hysteresis, i.e. depend on the sample history,
actions. Diamagnets and paramagnets are weak and is sometimes indeterminate (see Magnetic hys-
magnetic materials where the interaction between teresis and Nonergodicity); and there is also a
the magnetic moments is negligible in magnitude, possibility of macroscopic nonuniformities being
or has no fundamental importance. Some specific present (see Magnetic domain structure, and Inter-
magnetic properties are exhibited by superconduc- mediate state of a magnetic material).
tors (see Magnetic superconductors). The inter- There exist specific resonances that can arise
actions of atomic magnetic moments (especially from the magnetism of the particles of matter, such
MAGNETISM OF TRANSITION METALS 789
as: nuclear magnetic resonance (NMR) of nuclei, interaction. The indirect exchange interaction is
electron paramagnetic resonance (EPR) of para- effected via conduction electrons, and this is a
magnets, and collective spin system excitations in long-range interaction of an oscillating nature (see
magnetically ordered crystals which exhibit a spe- Ruderman–Kittel–Kasuya–Yosida interaction). The
cial kind of weakly damped modes called spin electrostatic interaction of the crystal e ld with
waves, or magnons (see also Ferromagnetic res- an anisotropic 4f -orbital influences the magnetic
onance and Antiferromagnetic resonance). anisotropy of rare-earth metals; the energy of mag-
Importance for science and technology. Mag- netic anisotropy is comparable to the exchange
netic methods (NMR, EPR, etc.) are employed to interaction energy. These features of the mag-
investigate the structure of materials in physics, netism of rare earths cause complex noncollinear
chemistry, biology, and medicine. Engineering ap- magnetic structures to appear in magnetically or-
plications of magnetism are multifaceted: magnet- dered states, including helical ones (e.g., in Dy
ic materials and permanent magnets are used in at T = 85–178 K), conical (Ho at T < 20 K),
power and radio engineering; methods of magnetic cycloidal (Er at T = 20–53 K), sinusoidal (Tm
defectoscopy are applied with success; devices at T = 25–60 K) and some others (see Modu-
based on the magnetic resonance phenomenon lated magnetic structures). Theory can explain the
and on magneto-optical effects have been intro- emergence of noncollinear structures in rare earth
duced into electronics, optoelectronics and com- metals through changes in the topology of the
puter engineering; magnetoresistance has been Fermi surface under magnetic ordering (see Topo-
utilized; microwave devices using spin waves are logical transition). A variety of noncollinear ferro-
employed, as well as magnetic bubble domain and antiferromagnetic structures are found at low
devices, and other magnetic computer memory el- temperatures, and in strong magnetic fields (e.g.,
ements. in Dy at T < 85 K, and in Sm at T < 106 K). De-
formation and restructuring of magnetic structures
MAGNETISM, AMORPHOUS
is observed in relatively weak magnetic fields.
See Amorphous magnetism. Changes of magnetic ordering are associated with
either first-order or second-order magnetic phase
MAGNETISM AND GREEN’S FUNCTIONS
transitions; the transformation of a helical anti-
See Green’s functions and magnetism.
ferromagnetic structure into a collinear ferromag-
MAGNETISM, BAND
netic one is accompanied by a lowering of the
See Band magnetism. symmetry from hexagonal to rhombic (see Crystal
symmetry). Typical for the heavy rare-earth metals
MAGNETISM OF RARE-EARTH METALS are high values of saturation magnetic moments
A branch of the physics of magnetic phenom- (e.g., for Dy, μ = 10.65μB along the a axis), as
ena which studies, both theoretically and exper- well as uniaxial anisotropy and giant magnetostric-
imentally, magnetic properties of the elements tion (anisotropy and magnetostriction coefficients
of the 57 La–71 Lu series (rare-earth elements or are greater by a factor of 10–100 for rare-earth
lanthanides). The strongly localized electrons in metals than for the 3d-iron group).
4f -orbitals act as carriers of magnetism in the rare
earths, with the population of these orbitals in- MAGNETISM OF TRANSITION METALS
creasing with the atomic number from 0 electrons Study of magnetic properties and phenomena
for La up to 14 electrons for Yb and Lu. For the of the transition metals. The atoms of transition
light rare-earth elements (Ce–Eu), J = L − S, and metals possess partially filled inner electron shells
for the heavy ones (Gd–Tm), J = L + S, where J , (lacking some d, or d and f electrons); which
L and S are the total, orbital, and spin angular mo- makes these substances paramagnets at high tem-
menta, respectively. The radii of the 4f -orbitals peratures.
are about 10 times smaller than the minimum in- Of the 3d-metals, α-Fe, Co (hexagonal) and
teratomic distances in a solid which virtually pre- Ni are ferromagnets, with Co having the high-
cludes the possibility of any direct f –f exchange est Curie point (T C = 1403 K); Mn, γ -Fe and
790 MAGNETISM, SURFACE
for both a diamagnet and a paramagnet yields magnetization direction, see Saturation magneti-
a unique relation M = M(B). For magnetically zation curve). Of key importance for magnetic
ordered media which feature magnetic hysteresis materials with several sublattices is the change
such a relation becomes ambiguous, and the mag- of angles between the magnetizations of magnetic
netization curve depends on the prehistory of the sublattices; this sometimes takes place through
sample. For ferromagnetic media (ferromagnets, magnetic phase transitions, e.g., through a spin-
ferrimagnets, amorphous magnetic substances) flop transition in an antiferromagnet, transitions
one may identify the curve of initial magnetiza- in metamagnetics, etc. Magnetization processes
tion. It gives the dependence M = M(B) for the are significantly influenced by intrinsic magnetic
case when B increases from B = 0 (and M = 0), fields of the sample (see Demagnetization elds).
in contrast to those magnetic curves which de- There is also a dependence of the nature of the
scribe a cycle of magnetic reversal (hysteresis magnetization on the sample shape, and this may
loop). The profile of magnetization points to the lead to the generation of inhomogeneous states (in-
existence of magnetic phase transitions induced termediate state) of magnetic materials. The state
by the applied field (see Phase transitions in an ex- of magnetization of magnetically ordered mate-
ternal eld). These appear as jumps during first or- rials often depends on their prior history (see
der phase transitions, or as discontinuities during Magnetic hysteresis).
second order transitions. Modern physical experi-
ments to determine magnetization curves and their MAGNETIZATION ROTATION PROCESS
theoretical analysis, oftentimes computerized, are The stage of the magnetization process during
an important tool for studying the state of the which the angles between the vectors of the spon-
magnetization of materials, and to define the fun- taneous magnetization M of different individual
damental properties of magnetic materials. domains change in order to align the overall mag-
netization with the direction of the local magnetic
MAGNETIZATION PROCESS e ld B acting within the body. For lower values of
Process of establishment of magnetization M the internal magnetic field the main magnetization
in a material under the action of an external mechanism is domain wall displacement, and the
magnetic eld B = μ0 H . This process in rotation process takes place during the subsequent
diamagnets consists in the generation of mi- stage of magnetization buildup when practically
croscopic (atomic) currents and in metals and all possible domain wall displacement processes
semiconductors in the realignment of the flow have already taken place and only the orientation
of current carriers; with the resulting magnetiza- of domains in different regions of the sample vol-
tion antiparallel to the field. Paramagnets become ume may not yet coincide with the direction of B.
magnetized along the field: magnetic moments of The rotation process occurs in magnetic fields suf-
thermally disordered particles, which constitute ficiently high to overcome the fields associated
a paramagnet, become oriented under the action with the crystal magnetic anisotropy. At the end
of the field (see Paramagnetism). Processes of of the rotation the magnetization of a ferromag-
magnetization in magnetically ordered crystals are net becomes saturated (see Magnetization curve).
more complicated, and under different conditions The contribution of the rotation process to the to-
are determined by different phenomena: not only tal magnetization depends on the magnetic texture,
by orientation of magnetic moments of particles on the perfection of the crystal lattice, and on the
by the magnetic field (see Paraprocess), but also shape and dimensions of the sample.
by rotation of the vector of spontaneous magneti- Rotation processes can be reversible or irre-
zation as a whole (if the direction of magnetization versible. Irreversible rotation processes can in-
does not coincide with an easy magnetization volve jumpwise changes of magnetization (see
axis). Besides that, if a magnetic domain structure Barkhausen effect). If the displacements cannot
exists, the magnetization processes are governed occur in a crystal (e.g., hard magnetic materials),
by its rearrangement (shift of boundaries of do- in a thin ferromagnetic film (see Magnetic lms)
mains, magnetization rotation processes in the or in a powder sample, then the magnetization
792 MAGNETIZATION SATURATION
increase is mainly due to rotations. A similar sit- the vectors k and B are not orthogonal, there will
uation takes place when magnetizing along hard be a drift motion of electrons along the directions
magnetization axes. Irreversible rotation processes of k, and magnetoacoustic resonance phenom-
in magnetically hard materials are the main cause ena occur. Such phenomena are pronounced when
of the magnetic hysteresis of ferromagnets. open electron orbits are present for which the u
values do not depend on the electron momentum,
MAGNETIZATION SATURATION the resonance condition being met for all electrons
See Saturation magnetization curve. simultaneously (see Magnetoacoustic resonance
at open orbits). When moving in closed orbits, the
MAGNETIZING FORCE electrons can drift along the k vector only in the
The same as Magnetomotive force. case k · B = 0. The displacement u in one cy-
clotron period depends on the angle, θ , between
MAGNETOACOUSTIC EFFECTS in metals
vectors k and B and on the projection, p B , of the
The oscillations with, and the resonance depen- electron momentum on the magnetic field direc-
dences on the magnetic field of sound absorption, tion:
and of the sound velocity dispersion in metals, due
1 ∂S
to the cyclotron pattern of the electron motion. u= cos θ,
Magnetoacoustic effects are most prominent in the eB ∂p B
region of a quantizing magnetic eld: Ω
k B T where S is the area of the section of Fermi sur-
(Ω is the cyclotron frequency), where the electron face cut by the plane p B = const. Those electrons
motion is one-dimensional (see Giant quantum for which the quantity ∂S/∂p B displays an ex-
oscillations). In the case of Ω k B T magne- tremum, ∂S/∂p B = (∂S/∂p B )extr , make a major
toacoustic effects are determined by the electrons contribution to the sound absorption This condi-
that move over periodic trajectories in the mag- tion is realized at extremal points on the Fermi
netic field, B, interacting with a periodic (in space surface (where p B attains a maximum) as well as
and time) sound wave field, with the condition at some other p B points, if the Fermi surface is
kl, τ Ω
1 (l and τ are the mean free path of not convex. The input of the above electron groups
an electron and its corresponding transit time, and determines a periodic (versus reciprocal magnetic
k is the wave propagation vector of the sound) field) system of asymmetric maxima in the sound
guaranteeing the ballistic regime of electron mo- absorption coefficient that emerge at the field mag-
tion. In the high-frequency region (ωτ
1), the nitudes
resonance absorption of sound occurs at ω = nΩ
k cos θ ∂S
(n = 1, 2, . . .), assuming that the sound wave B= extr .
changes periodically with time (see Acoustic cy- 2πne ∂p B
clotron resonance). The fact that the sound wave is An additional condition determining the possibil-
periodic determines a number of geometrical size ity of the resonance absorption of sound is the
effects when the distance, u, between the points of presence of points with k · v = 0 on the p B =
effective sound absorption (with k · v = 0, where const trajectory in momentum space. There are
v is the electron velocity) is a multiple to the sound boundary points p b which limit the regions of p B
wavelength: k · u = 2πn. The features of these ef- values where intersection of the lines p B = const
fects depend upon the mutual orientation of the and k · v = 0 on the Fermi surface is impossi-
k and B vectors, as well as on the Fermi surface ble. Corresponding to the input of such points are
topology. For k⊥B in metals with a closed Fermi jumps of the sound absorption coefficient at field
surface, the magnitude of u represents the charac- values
teristic size of a closed electron trajectory (lying
k cos θ ∂S
in a plane perpendicular to B), with the condition B= .
2πne ∂p B b
k · u = 2πn defining oscillations of the sound ab-
sorption and of the speed of sound as functions of Magnetoacoustic effects have been found in many
the reciprocal magnetic field (see Geometrical res- metals, and they underlie the ultrasound reso-
onance). When open electron orbits exist, or when nance spectroscopy method employed to study
MAGNETOACOUSTIC RESONANCE 793
MAGNETOACOUSTIC RESONANCE
A phenomenon arising from the interaction of
spins and elastic waves that results in the gen-
eration of magnetoelastic waves. The coupling is
characterized by a dimensionless parameter,
ωme B2
ζ= ≈ , (1)
ω0 CM 2
where M is the magnetization of the crystal (or
of its magnetic sublattice), C and B are the
elastic modulus (see Elasticity theory) and the
magnetostriction coefficient, respectively, ω0 is
the magnon activation frequency, and ωme is an Fig. 1. Dispersion laws of uncoupled (dashed line) and
coupled (solid line) elastic waves and spin waves, ωp (k)
input due to magnetostriction. Since M ≈ 10–
and ωs (k), respectively.
100 mT, C ≈ 108 –109 J/m2 , and B does not
exceed 103 –104 J/m3 in most cases, Eq. (1)
usually yields ζ 1. Therefore, magnons and the sake of simplicity) along a symmetry axis of a
phonons would be conventionally regarded as sep- high enough order so that the normal waves are
arate non-interacting quasi-particles in the first circularly polarized. The dashed lines represent
approximation, their interaction being considered noninteracting elastic ωp (k) and spin ωs (k) waves
in terms of perturbation theory as a cause of in accordance with the dispersion laws:
magnon–phonon scattering (see Magnon–phonon
interaction). Under ordinary conditions such colli-
ωp (k) = ωD (ak),
sions obey the energy and momentum conserva- (2)
tion laws, but only if three or more particles take
ωs (k) = ω0 + ωE (ak)2 ,
part. That corresponds, in terms of the oscillation where ω0 = γ B is the frequency of the uniform
amplitudes of the magnetization,
M, and of the precession of magnetization (k = 0) in the inter-
elastic strain,
uij , to accounting for the third and nal field B (comprising, in the general case, the
higher power anharmonic terms of
M and
uij effective fields of magnetic anisotropy and mag-
in the energy. There is the singular case of mag- netostatic demagnetization of the specimen, that
netoacoustic resonance involving a magnon and of magnetostriction, etc., aside from the external
a phonon with matching (or close to matching) field); ωD and ωE are characteristic frequencies of
magnitudes of energy and momentum. The inter- the elastic (Debye frequency) and spin (exchange
action may be so strong that the individual nature frequency) subsystems; and a is an interatomic
of each quasi-particle vanishes, and a sort of hy- spacing. The solid lines I and II on Fig. 1 corre-
brid quasi-particle comes forth instead. In terms spond to coupled waves that take into account the
of waves this means that coupled magnetoelastic magnetoelastic interaction.
waves, with
M and
u making equivalent con- The strongest magnetoacoustic coupling (mag-
tributions to the amplitude, become normal waves netoelastic resonance) takes place at the intersec-
to replace the elastic and spin waves with frequen- tion point O of the functions given by Eqs. (2),
cies and wave vectors close to one another (with where the frequencies and wave vectors of the
matching polarization). phonons and magnons become equal to each other.
Fig. 1 shows schematically the essence of mag- In the vicinity of this point the excitation of
M
netoacoustic resonance by plotting the branches oscillations is necessarily accompanied by a
u
of the frequency spectrum ω(k) of a ferromag- excitation, and vice versa, which permits the res-
net as a function of wave vector k directed (for onance generation of sound with an alternating
794 MAGNETOACOUSTIC RESONANCE
Fig. 3. Frequencies of coupled magnetoelastic waves, resonance of closed electron orbits, this effect is
ω, plotted against the wave vector, k, near the point observed even at k⊥B, occurring under the con-
of an orientational phase transition. The dashed lines 1 dition that the electron displacement, u, during
and 2 correspond to magnons and phonons, respectively, one cyclotron period is an integer multiple of the
without taking into consideration the magnetoelastic sound wavelength: ku = 2πn. The quantity u ∼
coupling, and the solid lines correspond to: I , quasi-
b sin θ/(eB) does not depend on the electron mo-
magnons, and II and III, quasi-phonons. The group
velocity of quasi-phonons is zero at k = 0. mentum, and is determined by the period b of the
reciprocal lattice along a direction where the Fermi
surface is open, as well as by an angle, θ , between
parameter, ξ , becomes unity at this point. The this direction and the wave vector k. The sound
dynamic modulus of elasticity for the correspond- absorption coefficient, Γ , is a function of the
ing quasi-phonon mode goes to zero at the phase reciprocal magnetic field (see Fig. 4) with a period
transition (Fig. 3) because of the magnetoelastic
(1/B) ∼ 2πe/(kb sin θ) under the conditions
coupling, and so does the speed of sound defined of magnetoacoustic resonance with open orbits.
by it. The first effect, the magnetoelastic gap, was Since the phenomenon is governed by the space
experimentally discovered in the antiferromagnetic (not time) periodicity of the acoustic wave, it is re-
resonance of haematite, and a giant magnetoelas- alized at low sound frequencies: ωτ 1 (τ is the
tic gap (about 20 K) was found in terbium and mean free path time of the electrons), if the condi-
dysprosium by inelastic neutron scattering. The tion Ωτ
1 holds (Ω = eB/m∗ is the cyclotron
second effect, reduction of the sound velocity, was frequency). At ωτ
1, the resonance condition
observed in haematite (more than 50%) and in takes the form ku = 2π(n ± ω/Ω) because of the
terbium. The impact of magnetoelastic effects at acoustic cyclotron resonance, and each absorption
phase transitions on the general understanding of maximum is split into two. The magnetoacoustic
physics is associated with the fact that underlying resonance at open orbits has been found in many
these phenomena are fundamental properties of metals, and is an effective means for the deter-
systems exhibiting spontaneous symmetry break-
mination of Fermi surface parameters, and of the
ing.
effective masses and mean free paths of electrons.
Similar effects (e.g., 1/B periods) can occur as
MAGNETOACOUSTIC RESONANCE AT OPEN
the de Haas–van Alphen effect and Shubnikov–de
ORBITS
Haas effect.
The resonance dependence on the magnetic
field, B, of the absorption of ultrasound in metals
due to electrons drifting along the wave vec- MAGNETOACOUSTIC RESONANCE,
tor k while moving along open periodic orbits NONLINEAR
in momentum space. Unlike the magnetoacoustic See Nonlinear magnetoacoustic resonance.
796 MAGNETOACOUSTIC SIZE EFFECTS
MAGNETOCALORIC EFFECT
Changes in the temperature of a magnetic sub-
stance with a variation in the applied magnetic
e ld B under adiabatic conditions. The magneto-
caloric effect is a consequence of the first law of
thermodynamics. The change of the internal en-
ergy of the magnetic material equals the work
done adiabatically by the magnetic field. Thus,
Types of surface electron trajectories in a magnetic the temperature of the material changes during
field B applied parallel to the metal surface (no provision adiabatic magnetization if the internal energy is
is made for surface scattering): (a) “sliding”; (b) “mixed” temperature dependent. A small variation in the
case of two-channel reflection from the boundary; and magnetizing field,
B, causes the temperature
(c) orbit touching both specimen boundaries (in a weak variation:
magnetic field). Electrons effectively interact with the
sound wave at sections of trajectories approximately par- T ∂M
allel to the wave front.
T = −
B,
CB ∂T B
MAGNETOELASTIC EXCHANGE INTERACTION 797
this, there are some magnetoelastic effects that do former (static) case results in magnetostriction,
not arise from the magnetoelastic exchange inter- mechanostriction, in the magnetic anomalies in
action, e.g., to describe magnets it is necessary to thermal expansion, in elastic moduli, in piezo-
allow for other types of interaction. The magnetoe- magnetism, etc. The stronger exchange interaction
lastic exchange interaction controls, in particular, associated with magnetic phase transitions may
the volume change that results from magnetic or- result in isomorphic changes of the structure of
dering (this can alter the type of magnetic phase the crystal lattice and its symmetry. In the latter
transition), but does not cause elastic strains to ap- (dynamic) case the magnetoelastic interaction re-
pear when the magnetization rotates with its length sults in magnetoacoustic resonance, spin–lattice
unchanged. relaxation, etc. The magnetoelastic interaction is
a relatively weak one, but its importance grows
MAGNETOELASTIC GAP markedly in the vicinity of a magnetic phase
An excitation in a spin wave spectrum of transition where the magnetic anisotropy energy
a magnetically ordered crystal whose magnetic decreases. Near a second-order phase transition
symmetry would not suggest such an excitation the frequency of the soft mode tends toward the
spectrum (see Goldstone theorem); also a con- value of the magnetoelastic gap rather than to
tribution to a spin wave excitation due to mag- zero. Here, as well, the phonon spectrum becomes
netoelastic interactions (see Magnetoacoustic res- highly distorted which, in turn, has a marked bear-
onance). The magnetoelastic gap appears most ing on the static, dynamic, thermodynamic and
prominently in the vicinity of spin-realignment other properties of the magnetic material.
magnetic phase transitions, as the activation due
to the magnetic anisotropy in the spin wave spec- MAGNETOELECTRIC
trum diminishes. Simultaneously vibrations of the See Magnetoferroelectrics.
elastic subsystem are transformed: the dispersion
law for the sound (acoustic wave) interacting with MAGNETOELECTRIC-ACOUSTIC
the spin waves changes from linear to quadratic at PHENOMENA
the transformation point. Associated with the mag- See Acoustomagnetoelectric phenomena.
netoelastic gap is the invariance in the free energy
of a magnetic substance which is linear in the de- MAGNETOELECTRIC EFFECT (predicted by
viations of the magnetization and strain from their L.D. Landau and E.M. Lifshits, 1956)
equilibrium values, and determines the presence of Development (or change) of magnetization
an intrinsic ferroelastic phase transition in the ma- (electric polarization) in a solid under the action
terial (see Ferroelastic). of an electric (magnetic) field. Several types of
magnetoelectric effects are known, differing in
MAGNETOELASTIC INTERACTION their physical nature. The thermodynamic poten-
Interaction between the magnetization and elas- tial of some magnetically ordered crystals may
tic displacements (of spin and elastic subsystems) include the terms Φ = −αik Ei Bk , where Ei and
in magnetically ordered solids. Underlying the Bk are the electric and magnetic fields, respec-
magnetoelastic interaction is the dependence of tively. Therefore, the electric field induces a mag-
the direct (dipole–dipole, spin–orbit), and the ex- netization in such materials: M = −∂Φ/∂Bk =
change interactions of magnetic atoms and mag- αik Ei , and the magnetic field induces an elec-
netic ions upon the spacing between them. Of tric polarization: P = −∂Φ/∂Ei = αik Bk . The
special importance are the direct magnetoelastic classes of magnetic symmetry for which the tensor
interaction and the exchange magnetoelastic inter- αik is nonzero were found by I.E. Dzialoshinskii.
action. These interactions can be described both A magnetoelectric effect was first observed in the
microscopically, in the terms of a spin–phonon antiferromagnet Cr2 O3 . Besides the magnetoelec-
Hamiltonian, and through a phenomenological ap- tric effects induced by an external field, sponta-
proach to the physics of magnetism, and they neous magnetoelastic effects exist in ferroelectrics,
appear in both static and dynamic situations. The which introduce some peculiar features into the
MAGNETOFERROELECTRICS 799
picture of ferroelectric and magnetic phase tran- centers. This was first experimentally observed
sitions, and into the response of such a system to with Al–O− hole centers in quartz. An essential
electric and magnetic fields. A nonuniform magne- role belongs to the electric-field modulation of
toelectric effect may may appear in the presence the processes of relaxation of electric and mag-
of a macroscopic nonuniformity of the magneti- netic dipoles. One can interpret this magnetization
zation (e.g., one associated with magnetic domain enhancement as a quasi-stationary cooling of the
walls) in magnetically ordered crystals, as well as system of paramagnetic centers down to a spin
in modulated magnetic structures; such phenom- temperature (see Level population) below the lat-
ena may take place in crystals of any symmetry tice temperature.
type. The magnetoelectric effect was also predicted
Magnetoelectric effects based on the terms for non-magnetic conductors lacking an inversion
Φ = −αik Ei Bk are impossible in crystalline para- center, where a magnetization must appear due to
magnets and diamagnets because the αik tensor an electric current flowing: M = αE, and an elec-
vanishes in crystals without a magnetic structure, tric polarization must emerge in a magnetic field.
but whose symmetry operations include time inver- The pseudotensor α includes the relaxation time;
sion, T , by itself (T changes the sign of the mag- therefore, as in all kinetic phenomena, the time in-
netic field and magnetization). Still the thermody- version covariance is not a limiting condition in
namic potential of ferroelectric paramagnetic crys- this case.
tals involves the terms Φ = −ξij k Ei Bj Bk , where Underlying the magnetoelectric effects is the
ξij k is a third-rank tensor with the same symmetry fact that the electric and magnetic characteristics
as the piezoelectric tensor. Placing such crystals of a solid are determined by the same particles,
in a strong magnetic field, Bk , one can induce in namely, electrons that carry, in addition to an elec-
them a magnetization along the j direction with a tric charge, a magnetic moment (spin and orbital).
field Ei , or an electric polarization along the i di- There are interactions (exchange and spin–orbit
types) to interrelate the motion of charges in space,
rection with a field Bj . Such a magnetoelectric
and the mutual alignment of their magnetic mo-
effect was first identified in a NiSO4 ·6H2 O crys-
ments. External forces affect the energy of the
tal.
electric and magnetic subsystems that interact
The magnetoelectric effect is also possible in
both with one another and with their environment
diamagnetic nonmetallic crystals with paramag-
(e.g., with phonons). That results in the reciprocal
netic centers of impurity ions. This effect can be
influence of the electric field upon the magnetic
conveniently described in this case in terms of the
characteristics, and of the magnetic field upon the
spin Hamiltonian of a paramagnetic center. If the
electric characteristics, of a solid.
local symmetry of the paramagnetic center, as well
as the symmetry of the crystal as a whole, is lack- MAGNETOFERROELECTRICS,
ing a center of symmetry then the spin Hamiltonian magnetoelectrics
parameters will be linear in the field E, and so Materials featuring simultaneous magnetic and
will the magnetization of the crystal. Along with ferroelectric (antiferroelectric) ordering. The nat-
the purely static magnetoelectric effects caused ural magnetoelectrics are orthorhombic ericaite,
by a change of spin Hamiltonian parameters, dia- Fe3 B7 O13 Cl and chambersite, Mn3 B7 O13 Cl.
magnetic nonmetallic crystals with paramagnetic However, the first studies of magnetoelectrics
centers exhibit magnetoelectric effects associated were conducted using synthetic crystals of the per-
with an unpaired electron traveling (as a rule, by ovskite type. A large number of both ferroelectric
way of tunneling) across different sites in the crys- and antiferroelectric crystals are known, which
tal, and with the effects of an electric field acting may be either ferromagnetic, antiferromagnetic or
on such a motion. Under this type of magneto- ferrimagnetic. Their ferroelectric and magnetic or-
electric effect, an enormous enhancement of the dering occurs at different temperatures, θE and
magnetization of a specimen (by over an order of θM , respectively (see Curie point). Ordinarily, but
magnitude) is observable, arising from the action not necessarily, θM < θE . A substance is a magne-
of an alternating electric field on the paramagnetic toelectric at T < min(θE , θM ). Magnetoelectrics
800 MAGNETOGYRIC RATIO
measures the changes of a specimen temperature the quantization of the particle angular momen-
at its remagnetization. Most of these methods are tum. For orbital motion the latter is a multiple of
now antiquated, and seldom employed. , for spins it can be a multiple of either or /2.
More widely used in recent years are reso-
nance and quantum methods which depend on MAGNETOOPTICAL EFFECTS
the phenomena of nuclear magnetic resonance,
Phenomena consisting in changes of light po-
electron paramagnetic resonance, Josephson ef-
larization or intensity in a material medium un-
fect (see Superconducting quantum interference
der the action of a magnetic eld, B, or of a
devices), etc. Galvanomagnetic methods make use
magnetization, M, that take place as light passes
of the Hall effect, and the changes of the resistivity
through a medium, or is reflected from its surface.
in a magnetic field (see Galvanomagnetic effects).
Falling into the first group are: the Faraday effect,
Optical methods are associated with the rotation of
or magnetic circular birefringence, the Cotton–
light polarization plane (Faraday effect and Kerr ef-
Mouton effect (see Voigt effect), magnetic circular
fect).
dichroism (differential absorption of waves with
left and right polarization in the Faraday effect
MAGNETOMOTIVE FORCE
geometry), and magnetic linear dichroism (dif-
A scalar quantity characterizing the magnetiz-
ferential absorption of intrinsic waves with linear
ing action of electric currents. The magnetomotive
polarization in the Cotton–Mouton effect geome-
force F equals the circulation of the magnetiz-
try). The changes in the intensity and polarization
ing vector B over a closed circuit L enclosing the
of light on its reflection from the surface of a
electric currents Ii that generate the magnetic field
magnetized medium constitute the magnetooptical
n Kerr effect. It can be polar (M n, where n is a
F B dl = Bl dl = Ii , normal to the surface), longitudinal, or meridional
L L i=1 (vector M parallel to the surface and lying in the
plane of incidence), and transverse or equatorial
where Bl is the projection of B on to the direc-
(vector M parallel to the surface and perpendic-
tion of the element dl of the integration contour,
ular to the plane of incidence). The correlation
and n is the number of conductors (turns) carrying
between the optical and magnetic properties oc-
a current I enclosing the circuit. This concept is
curs not only in static, but also in dynamic, phe-
useful for magnetic circuit calculations.
nomena, e.g., in the scattering of light from spin
waves. Magnetooptical effects are found in ma-
MAGNETON
terials of all types: in insulators, semiconductors
A unit of magnetic moment with the value and metals, in diamagnets and paramagnets; but
μ = e/2mfp , where e is the electron charge they are strongest in magnetically ordered ma-
and mfp is the mass of a fundamental particle. terials: ferromagnets, antiferromagnets, etc. The
Of particular importance are the Bohr magneton, cause underlying the appearance of these effects is
μB = 9.274·10−24 J/T (defined by the mass of
the dependence of the energy spectrum of a mag-
an electron) used for atomic magnetism, and the
netic substance on its magnetic state.
nuclear magneton, μN = 5.051·10−27 J/T, with
mfp equal to the mass of a proton. Thus, electronic
magnetism is much stronger by three orders of MAGNETOOPTICAL EFFECTS IN LIQUID
magnitude (mp /me = 1836) than the magnetism CRYSTALS
of nuclei and nucleons. The magnetic moment of Changes in optical properties of liquid crystals
an elementary particle (as well as that of a sys- resulting from the orienting action of an applied
tem of particles) is quantized, and sometimes it magnetic field upon the director (optical axis of a
is a multiple of the corresponding magneton. The liquid crystal) due to the diamagnetic anisotropy
magneton plays the part of an elementary mag- with χa = χ − χ⊥ > 0 (see Electrooptical ef-
netic moment; which is closely associated with fects).
802 MAGNETOOPTICS
MAGNETOOPTICS MAGNETOPLUMBITE
Study of the interrelations between magnetic A mineral of hexagonal (in the crystallographic
and optical properties of matter. These magne- sense) symmetry. Pb2 Fe15 Mn7 (Al, Ti) O38 . Ox-
tooptical phenomena can be classified into: mag- ide monocrystalline ferrimagnets of composition
netooptical effects (Faraday effect, Kerr effect, etc.) n1 (Pb, Ba, Sr)O·2n2 Fe2 O3 (n1 1, n2 3), iso-
caused by changes in the optical properties of morphous to the above, are referred to as hexa-
materials (dielectric constant and magnetic perme- ferrites of the magnetoplumbite structure. The unit
ability tensors) under the action of a magnetic eld cell comprises four alternating blocks: S blocks
or magnetization (or sublattice magnetization vec- of the cubic spinel structure (two-layered), and
tors in multisublattice magnetic substances); and R blocks of a hexagonal structure (three-layered),
photomagnetic effects arising from the impact of positioned perpendicularly to the c axis according
to the pattern SRS ∗ R ∗ (where the asterisk denotes
optical radiation on such magnetic parameters as
blocks turned through 180◦ ). The space group is
magnetization and magnetic anisotropy (reciprocal 4 ), and the point group has sym-
P 63 /mmc (D6h
Faraday effect, photoinduced magnetic anisotropy,
metry 6/mmm (D6h ). The best known members
etc.). Magnetooptical phenomena appear in all
of the magnetoplumbite family are: PbO·6Fe2 O3 ,
magnetic materials, but are most pronounced in
barium ferrite BaO·6Fe2 O3 and strontium fer-
magnetically ordered media.
rite SrO·6Fe2 O3 . Typical magnetoplumbite para-
meters are: crystal lattice constants a = 0.586–
MAGNETOPLASTIC EFFECT 0.589 nm and c = 2.303–2.319 nm; density
Effects of a magnetic field on mechanical 5.12–5.67 g/cm3 , magnetic moment per unit
properties of ferromagnets and other materials. chemical formula ∼20μB , and the Néel point
Changes occur in the yield limit, ultimate strength, 430–460 K.
permanent set and creep of materials, wear hard-
ness (see Wear), and the durability of parts and MAGNETORESISTANCE
tools. In an alternating field the strain is discon- Changes in resistivity of solid conductors un-
tinuous. The magnetoplastic effect arises from der the action of an external magnetic eld, B.
the interaction of dislocations with magnetic do- There are transverse and longitudinal magnetore-
main walls, or from magnetostrictive stresses (see sistive effects with the electric current, I , running
Magnetostriction). The magnetic domain structure
perpendicular to the magnetic field B, or with
I B, respectively. The underlying cause of this
becomes much finer under strain, which makes
effect is the bending of current carrier trajectories
the interaction of domain boundaries with dislo-
in the magnetic field (see Galvanomagnetic ef-
cations in a weak field more intense. In a stronger
fects). The relative change of the transverse resis-
field, domain boundaries disappear, but the drag
tivity at room temperature amounts to (
ρ/ρ)⊥ ≈
on dislocations due to their disaligning action
10−4 at B ≈ 1 T in metals, with the exception of
upon the spin moments remains. When studying
Bi which has (
ρ/ρ)⊥ ≈ 2 at B = 3 T. The latter
magnetoplastic effects experimentally, it is nec- justifies the employment of Bi to gauge the mag-
essary to take into account specimen heating by netic field (see Magnetometry).
eddy currents in an alternating magnetic field, or at Decreasing the temperature and increasing B
the switching (on/off) of the permanent field. The result in an increase in (
ρ/ρ)⊥ (see Kapitza
magnitude and sign of the magnetoplastic effect law). In a weak field, (
ρ/ρ)⊥ is proportional
depend on many factors: the magnitude and direc- to B 2 . The proportionality factor is usually posi-
tion of the magnetic field; the sequence in which tive, i.e. resistivity increases with magnetic field;
stress and field have been applied; the temperature the exception being ferromagnets (see Kondo ef-
and the rate of deformation; the number of impurity fect, Mesoscopic uctuations). Since resistivity is
atoms, the structural homogeneity of the material sensitive to the presence of impurities and de-
including the grain and subgrain sizes; and mag- fects in the crystal lattice, and to the temperature,
netic characteristics (domain structure, saturation measurements may yield different ρ versus B de-
magnetization, magnetostriction, etc.). pendences. Nevertheless, different data referring
MAGNETOSTRICTION 803
to one and the same metal are found to fit a sin- where μ0 is the magnetic permeability of vacuum,
gle straight line (see Kohler rule). γ is the gyromagnetic ratio, M is the saturation
The magnetoresistive effect is used to investi- magnetization, B is the magnetizing component
gate electronic energy spectra and mechanisms of of the external field (B z), and Nx , Ny and Nz
current carrier scattering, and to gauge the mag- are the demagnetizing factors of the ellipsoid. The
netic field strength. magnetic oscillations are excited with a nonuni-
Larger changes in magnetoresistance called gi- form microwave field by placing the specimen in
ant magnetoresistance (GMR) have been observed a wave guide or resonator.
in films made with alternating layers of a ferro-
magnetic and a non-ferromagnetic material. This MAGNETOSTATICS
effect was first detected in alternating layers of Fe
A division of magnetism and electrodynam-
and Cr, and is more pronounced in Co/Cr layers.
ics that studies the properties of a steady-state (or
Other materials, such as La0.67 Ca0.33 MnOx , ex-
relatively slowly changing) magnetic field (from
hibit resistance changes of 103 in a magnetic field
a permanent magnet or direct electric currents).
of 6 T, an effect referred to as colossal magnetore-
Calculations of the field make use of the equations
sistance.
of magnetostatics, which correspond to a special
MAGNETORESISTANCE, ANOMALOUS case of Maxwell’s equations:
See Anomalous magnetoresistance.
∇ · B = 0, ∇ × H = j,
MAGNETOSTATIC OSCILLATIONS
where B is the magnetic ux density, H = B/μ
Free oscillations of magnetization of a mag-
is the magnetic field, and j is the conduction cur-
netically ordered material with characteristic in-
rent density. The basic field computation methods
homogeneity dimensions of the order of the spec-
in magnetostatics ensue from the Ampère theo-
imen size. The frequency of magnetostatic oscil-
rem, which asserts that the magnetic fields of
lations, ω, must satisfy the inequality ω c/ l
closed direct current loops can be regarded as the
(l is the characteristic specimen size), implying
that one is allowed to neglect the effects of retar- fields of paired magnetic charges corresponding
dation (lag) between different parts of the mate- to magnetic dipoles. Introduction of the magnetic
rial. At the same time, the specimen dimensions charge concept allows the formal application of
should be large compared to the characteristic field computation methods employed in electro-
exchange interaction length, D 1/2 (l
D 1/2 , statics. The magnetostatic field in the absence of
where D ≈ 10−12 –10−14 cm2 ). The magnetic os- conduction currents is a potential field, i.e. the
cillations can be described, for these simplistic work of the forces of a magnetostatic field act-
circumstances, by magnetostatics equations and ing on a probe charge is independent of the path.
by the equations of motion of the magnetiza- The magnetostatic equations describe a varying
tion of magnetic sublattices (see Landau–Lifshits magnetic field adequately enough, provided the
equation). The structure of the magnetic field and characteristic dimension of the field inhomogene-
magnetization, and the spectrum of magnetic os- ity is much smaller than the wavelength of the
cillations, depend on the specimen shape, and the oscillatory changes in B. Satisfying this condition
boundary conditions at its surface. There is a va- is called the magnetostatic limit.
riety of characteristic magnetic oscillations (see
Walker modes), including those corresponding to MAGNETOSTATIC WAVES
uniform precession of the magnetization. The res- See Spin waves.
onant frequency of ferromagnetic resonance for a
ferromagnetic ellipsoid is given by the Kittel equa- MAGNETOSTRICTION
tion (Ch. Kittel, 1948): Strain of a material induced by an applied
1/2
ω = γ B + μ0 (Nx − Nz )M magnetic field, discovered by J. Joule (1842). It
1/2 is characterized by relative changes of the lin-
× B + μ0 (Ny − Nz )M , ear dimensions of the body λ =
l/ l (linear
804 MAGNETOSTRICTION, GIANT
magnetostriction) or of its volume (bulk magne- to the magnetic state changing under the action of
tostriction). The magnetostriction phenomenon is external elastic stresses is termed forced magne-
present in all magnetic materials, with differing tostriction, or mechanostriction.
values of λ in different cases. In ferromagnets Associated with magnetostriction is the recip-
and ferrimagnets, λ ∼ 10−5 –10−4 , in antiferro- rocal process or Villari effect, which is the change
magnets, paramagnets and diamagnets, the mag- in the magnetic induction B of a ferromagnetic
netostriction is ordinarily smaller (10−6 –10−7 ). material resulting from an applied mechanical
Giant magnetostriction with λ ∼ 10−3 –10−2 is stress.
exhibited by rare-earth elements, actinides and
their compounds. The magnitude of λ depends on MAGNETOSTRICTION, GIANT
the orientation of the size change direction rel- See Giant magnetostriction.
ative to the external magnetic field B and the
MAGNETOSTRICTIVE MATERIALS (fr. Gr.
crystallographic axes. One can determine λ along
μαγ νητ oς , magnet, and Lat. strictio, tightening,
the field B (longitudinal magnetostriction) and
narrowing)
perpendicular to it (transverse magnetostriction).
Soft magnetic materials exhibiting a marked
Underlying this phenomenon is the fact that the
magnetostriction. Some basic parameters of mag-
interactions defining the magnetic state of a crys-
netostrictive materials are: the magnetomechani-
tal depend upon the distances between magnetic
cal coupling parameter, k, whose square equals
atoms or ions. Hence, any change in the mag-
the ratio of the transformed energy (mechanical
netic state under the action of an applied magnetic
or magnetic) to the supplied energy (magnetic or
field, temperature, elastic stress, etc., results in the
mechanical, respectively), neglecting losses; the
atoms being displaced from their equilibrium posi-
dynamic magnetostriction parameter, a, defining
tions during the ensuing deformation of the body.
the transducer sensitivity in the projector mode;
Magnetostriction is known as exchange mag- the relative magnetic permeability; the sound ve-
netostriction or direct magnetostriction accord- locity; the saturation magnetostriction determining
ing to the type of magnetic interactions involved the ultimate sound intensity emitted by the magne-
(exchange interaction, or direct interaction such tostriction transducer; and the coercive force and
as spin–orbit or dipole–dipole). The direct mag- electric resistivity responsible for the energy loss
netostriction occurs in the region of saturation through hysteresis and eddy currents, respectively.
magnetization of ferromagnets at B < B s (B s is Employed as magnetostrictive materials are iron–
the saturation magnetizing field) and displays a nickel (permalloy), iron–cobalt (permendur), and
marked anisotropy. At B > B s , i.e. in the so-called iron–aluminum (alfer) alloyed with chromium,
paraprocess region, the exchange magnetostric- vanadium and other metals as well as with some
tion is typical. In this region, the magnetostric- soft ferrimagnetic materials. What should be taken
tion is isotropic in cubic crystals and anisotropic into consideration when choosing magnetostric-
in uniaxial ones. The longitudinal and transverse tive materials are: their corrosion resistance; the
magnetostrictions are usually opposite in sign for plasticity permitting fabrication into thin sheets; an
B < B s in most magnetic materials, and have sim- operational temperature range conditioned by the
ilar signs for B > B s . The value of λ at B = B s temperature dependences of the material charac-
is called the saturation magnetostriction. Magne- teristics, and by its Curie point; the ease of process-
tostriction is an even function of the magnetiza- ing; and the cost. Typical of iron–cobalt alloys is
tion M. the combination of very high values of the mag-
Antiferromagnets exhibit, in addition to the netostriction, saturation magnetization, and Curie
usual variety, a magnetostriction linear in the temperature. For this reason, they are used to de-
magnetization, and a piezomagnetic effect (see sign high-power transducers and high intensity
Piezomagnetism) reciprocal to it. A strain due to projectors. Some permendur types are remarkable
changes of the magnetic state with temperature for their good plasticity. Iron–aluminum alloys
(spontaneous magnetization) is termed sponta- possess large magnetostriction and electric resis-
neous magnetostriction. An additional strain due tivity, and in addition mechanical strength, but
MAGNON 805
they display low ductility and a pronounced ten- (see Adiabatic demagnetization cooling). When the
dency to corrode. The cores of magnetostriction atomic or nuclear magentic moments of a spin
transducers are assembled from sheets of magne- system absorb microwave or radiowave energy
tostrictive materials electrically isolated from one during a magnetic resonance experiment (see elec-
another to reduce eddy current loss. tron paramagnetic resonance, nuclear magnetic
Ferrite magnetostrictive materials exhibit a resonance) they transfer this energy to the lattice
very high corrosion resistance, and they are used, vibrations with a time constant called the spin–
due to their high resistivity and the ensuing low lattice relaxation time. If this relaxation time is
eddy current loss, for the conversion of radio very long the spin system can become heated, and
frequencies, particularly, in magnetostriction fil- temporarily acquire a spin temperature which is
ters. Magnetostrictive materials find applications higher than the lattice temperature.
as projectors, and as receivers of ultrasonic me-
chanical vibrations. MAGNET, POWDER
See Powder magnet.
MAGNETOTHERMAL PHENOMENA
Changes in the thermal states of materials
MAGNET, SAMARIUM–COBALT
that take place during changes in their magnetic
See Samarium–cobalt magnet.
states (at their magnetization or demagnetization).
Singled out are: the magnetothermal phenomena
MAGNON
during an adiabatic magnetization and demagneti-
A quasi-particle corresponding to a spin os-
zation (magnetocaloric effect when the body tem-
cillation wave (spin wave) in a magnetically or-
perature changes), and isothermal magnetother-
dered crystal. The number of different magnon
mal phenomena associated with the evolution or
types coincides with that of spin wave branches;
absorption of heat. In principle, magnetothermal
phenomena are observable in any material, since hence acoustic magnons, optical magnons, nu-
underlying them are general laws of thermody- clear magnons, etc., are distinguished. The density
namics. A change of the internal energy of a body of magnons at thermal equilibrium depends on
(see Thermodynamic potentials) is associated with the temperature, T , and decreases as T → 0.
a change in its magnetic state. Magnons manifest themselves in the thermody-
Most prominent are magnetothermal phenom- namic properties of magnetic substances, viz.,
ena in ferromagnets, antiferromagnets and ferri- they determine the low-temperature behaviour of
magnets. The features of the phenomena in these the magnetization, M(T ) (see Bloch law), con-
materials stem from the nature of their magnetiza- tribute to the speci c heat, thermal conductivity
tion processes: shifting of domain boundaries (see and other thermal parameters, and also feature
Domain wall displacement processes); rotation of prominently in the microwave and kinetic proper-
magnetic moments of domains (see Magnetization ties of matter. Describing three-dimensional mag-
rotation processes); paraprocess; or processes in- netic materials in terms of a gas of weakly in-
volving of the decay or induced build-up of a teracting magnons is valid, as a rule, up to a
noncollinear atomic magnetic structure (in anti- temperature of the order of half the temperature of
ferromagnets and ferromagnets). Particularly im- the transition to the paramagnetic state (see Para-
portant are thermal effects accompanying the two magnetism). Taking into consideration, in addition
latter processes. Closely linked by thermodynam- to magnons, nonlinear excitations such as mag-
ics with magnetothermal phenomena occurring netic solitons, magnon bound states and magnon
during magnetization are the anomalies of the spe- bubbles, may be necessary in low-dimensional
cific heat found in the vicinity of the Curie point, magnetic materials, as well as under the conditions
Néel point, and other magnetic phase transition of strong excitation. Direct investigations of the
points (e.g., near the switch point of the non- magnon dispersion law have been carried out us-
collinear magnetic structure in ferrimagnets). The ing the methods of inelastic neutron scattering and
magnetothermal phenomena in some paramag- the scattering of light, as well as with parametric
nets are used to achieve temperatures below 1 K excitation of spin waves.
806 MAGNON BOUND STATE
whereas γ -Mn and δ-Mn cannot be fixed on cool- modulus is ≈175 GPa at 294 K, shear modulus
ing to room temperature. α-Mn has a complex is 76.52 GPa (at 298 K); Poisson’s ratio is 0.24
cubic crystal structure with a = 0.8904 nm at (at 298 K). Rockwell hardness is 70 HRC for
298 K, space group I 4̄3m (Td3 ). The conventional α-Mn and 20 HRC for γ -Mn. At room temper-
unit cell contains 58 Mn atoms, whereas the prim- ature, α- and β-Mn are hard and brittle; there-
itive cell contains 29. Packing fraction is 54.8%. fore not subject to plastic deformation. γ -Mn is
The structure is considered as a system of inter- plastic and easily deformed until it transforms
penetrating body-centered cubic lattices similar to into α-Mn. Linear low-temperature electronic mo-
α-W containing four types of Mn atoms with dif- lar heat capacity: α-Mn ≈15.3 mJ·mol−1 ·K−2 ,
ferent coordinations and sizes, assumed to be in γ -Mn 9.20 mJ·mol−1 ·K−2 . Electric resistivity
different oxidation states with magnetic moments (free of residual resistance): α-Mn 1390 n·m (at
in different directions. β-Mn also has a com- 295 K), β-Mn 910 n·m, γ -Mn 350–400 n·m
plex cubic structure, space group P 4̄3m, (Td1 ), (at room temperature); resistivity temperature coef-
a = 0.6470 nm at 1023 K. For non-equilibrium ficient: α-Mn (2–3·104 K−1 , β-Mn 0.00136 K−1 ,
β-Mn, a = 0.6302 nm at 298 K. Conventional γ -Mn 0.0055–0.0065 K−1 . Work function of
unit cell has 20 Mn atoms in two types of sites polycrystal 3.83 eV. Manganese is paramagnetic
with different coordinations. Packing fraction is but at low temperatures α-Mn and β-Mn are
67.6%. Modifications of α-Mn and β-Mn, unlike metastable, whereas γ -Mn and δ-Mn turn into
other metals, have no center of symmetry. Crys- antiferromagnetic state; Néel temperatures: α-Mn
tal structure of γ -Mn is treated as face-centered 98 K, γ -Mn ≈515 K, δ-Mn 625 K. Entropy
cubic with a = 0.3859 nm at T = 1370 K with changes of antiferromagnet–paramagnet transfor-
4 atoms in the unit cell and packing fraction is mation amount to 0.21 J·mol−1 ·K for α-Mn and
74.05%, space group F m3̄m, (Oh5 ). There is evi- < 4.187 J·mol−1 ·K for nonequilibrium γ -Mn.
dence that, even on rapid cooling to room temper- β-Mn remains paramagnetic until extremely low
ature, γ -Mn attains a pseudo-cubic face-centered temperatures (1.5 K). Molar magnetic susceptibil-
tetragonal structure, slightly deformed along one ity of α-Mn is +527·10−6 (at 293 K), of β-Mn
of the cube edges, space group I 4/mmm (D4h 17 ),
+483·10−6 (at 293 K), of γ -Mn +467·10−6
a = 0.3780 nm, c = 0.3520 nm, packing fraction (at 1367–1407 K). Nuclear magnetic moment of
68.7%. Crystal structure of δ-Mn is body-centered 55 Mn is 3.461 nuclear magnetons. Manganese is
cubic, space group I m3̄m, (Oh9 ), containing two one of the most industrially important metals. It
Mn ions, a = 0.3081 nm at 1410 K. Packing frac- is used in the manufacture of steel and cast iron
tion is 68.2%. Density of α-Mn is 7.471 g/cm3 , as a deoxidizing and desulfurizing agent which
of β-Mn 7.245 g/cm3 (at 298 K), of γ -Mn provides wear resistance, strength, corrosion resis-
6.335 g/cm3 (at 1368 K), of δ-Mn 6.239 g/cm3 tance, as well as in manufacturing of nonferrous
(at 1407 K); Tmelting = 1517 K, Tboiling = 223 K, metal alloys. Some industrially-used manganese
heat of melting 14.9 kJ/mole, heat of evapo- compounds are toxic.
ration 228.9 kJ/mole, bond energy of α-Mn is
−2.98 eV/atom (at 0 K). Transition heats: α-Mn MANY-CONFIGURATION APPROXIMATION
→ β-Mn ≈2.54 kJ/mole, β-Mn → γ -Mn 2.28 A natural extension and refinement of the
kJ/mole, γ -Mn → δ-Mn 1.8 kJ/mole; heat of sub- Hartree–Fock method. Whereas the latter treats
limation 224.8 kJ/mole. Specific heat of α-Mn is the wave function of many particles as a product
0.477 kJ·kg−1 ·K−1 , of β-Mn 0.649 kJ·kg−1 ·K−1 or determinant that is composed of one-particle
and of γ -Mn 0.502 kJ·kg−1 ·K−1 at 298 K. Debye wave functions (the single-configuration approx-
temperature is 410 K; linear thermal expansion imation), the many-configuration approximation
coefficient in temperature range 273 to 293 K is represents the many-particle wave function as a
22.3·10−6 K−1 for α-Mn, 24.9·10−6 K−1 for linear combination of several determinants that are
β-Mn, 14.75·10−6 K−1 for γ -Mn; thermal con- constructed from different sets of single-particle
ductivity is 66.57 W·m−1 ·K−1 (at 298 K) and wave functions. Each determinant corresponds
5.86 W·m−1 ·K−1 (at 95 K). Isothermal bulk mod- to a certain configuration of electrons distrib-
ulus is 59.65 GPa (at room temperature); Young’s uted over the single-particle functions. Hence, the
808 MANY-VALLEY SEMICONDUCTORS
MANY-VALLEY SEMICONDUCTORS
Isoenergetic ellipsoidal surfaces in the conduction band
Semiconductors with several minima (max- of Si.
ima) of electron energy (treated as a function of
the electron quasi-momentum p) in the neighbor-
hood of the bottom of the conduction band (top of speaking, these quantities fail to form a tensor.
valence band). If the electron energy has an ex- Isoenergetic surfaces in momentum space near
tremum at point p 1 = 0, then there will be similar the band edge are ellipsoids, ellipsoids of revo-
extrema at all points p α that result from apply- lution when m2 = m3 , and spheres when m1 =
ing operations of point group of symmetry. Such m2 = m3 . If p α = 0, these surfaces are ellipsoids
extrema are called equivalent. Since the concen- even in a cubic crystal. Such a situation is realized
tration of current carriers in semiconductors is not for germanium and silicon (see Fig.). The isoen-
high, they only fill those states that are very near ergetic surfaces of these same semiconductors in
extremum points, and such groups of states are the conduction band are ellipsoids of revolution
called valleys. If there is no degeneracy of the with effective mass ratios 19.3 and 5.16, respec-
energy spectrum at the extreme point p α , the elec- tively. The response of current carriers to external
tron energy is expressible as a series in powers actions depends on the relative positions of the
of the deviation of the quasi-momentum p from valleys and their orientation relative to external
valley center p α . With an accuracy up to the first forces. For instance, with the application of a uni-
nonzero terms, we obtain axial crystal strain, or strong applied magnetic
or electric fields, the relative populations of var-
1 (pi − pαi )(pj − pαj )
E(p) = E(pα ) + ious valleys can be significantly changed, and the
2 mij electrical conductuctivity of the crystal is thereby
ij
(1) affected. Intervalley redistribution is responsible
for the Gunn effect, the Sasaki effect, and the gen-
The quantities 1/mij are components of the recip- eration of intervalley noise.
rocal effective mass tensor. This tensor is symmet-
ric; and when written in the principal axis system MANY-WAVE APPROXIMATION
Eq. (1) assumes the form Method of dynamic radiation scattering theory
(X-ray, neutron, or electron scattering), which al-
(p − pα1 )2 (p2 − pα2 )2
E(p) = E(pα ) + 1 + lows for multiple scattering of an incident wave
2m1 2m2 into several diffracted waves and vice versa, pro-
(p3 − pα3 )2 vided the Bragg law is completely or approxi-
+ . (2) mately satisfied for several systems of crystallo-
2m3
graphic planes. As a result, a unified wave field
The quantities mij (m1 , m2 , m3 ) are called effec- arises that is a coherent superposition of incident
tive mass tensor components, although, strictly and diffracted waves with nodes and antinodes
MARTENSITE 809
(standing wave); its interaction with atoms of the temperature of cold brittleness T CB ). That is, the
crystal is quite different from the ordinary case. In margin of plasticity for body-centered cubic met-
particular, the effects of an anomalous passage of als is larger, the greater the difference T − T CB ,
X-rays, indirect excitation of forbidden reflections, where T is the testing or operating temperature.
etc. take place. For X-rays, thermal neutrons, and When a coarse-grained recrystal structure (see Re-
slow electrons, this occurs only if a crystal is crystallization) with perfect grain boundaries is
aligned relative to the incident beam with a high present, the margin of plasticity determined by the
degree of accuracy. In this case, wave amplitudes relative elongation at a temperature T > T CB is
and refractive indices are found as eigenvectors larger than that with a fine-grained or dislocational
and eigenvalues of a certain matrix, with elements cellular structure, but the plasticity in the second
that describe processes of individual scattering case is retained down to lower temperatures.
of waves one by another. The description of the
passage of high-speed electrons or ions along a MARGIN OF SAFETY
crystallographic symmetry axis (high-resolution The relation between allowed and actually act-
electron microscopy, channeling) always requires ing stresses. This is limited by the safety factor
the many-wave approximation, and the number which is usually determined as the ratio of per-
of waves taken into consideration here is rather missible stress (yield limit, ultimate strength, en-
high. This approximation is also used in solving durance limit, etc.) to the maximum stress that
problems of X-ray optics. See also Two-wave ap- occurs during the course of construction. The
proximation. safety factor takes into account the possibility of
random overloads, the influence of crystal struc-
MANY-WELL POTENTIAL ture defects (pores, inclusions, dislocations, etc.),
Adiabatic potential of a polyatomic system, and the nonuniformity of their distribution. The
with two or more minima equivalent in energy selection of a value for the safety factor involves
and symmetry. In accordance with the theory of vi- the necessity of conserving material and creating
bronic interactions, the latter are the only source of constructions of minimum weight. Minimum val-
instability of highly symmetric configurations, and ues of the safety factor are adapted for expendable
the formation of several minima of the many-well products with short-term applications, and maxi-
potential (see Jahn–Teller effect, Renner effect, Vi- mum values apply to constructions of long-term
bronic instability). The assumption of a many-well duration, especially under dynamic loads.
potential (in particular, of a double-well potential)
is used in various descriptions of molecules, glassy MARTENSITE (named after A. Martens)
systems, crystals, metastable states, and also in the 1. In some crystalline materials: a low-tem-
theory of order–disorder structural phase transi- perature polymorphous structural modification. It
tions, the theory of noncentral impurities in crys- results from a diffusion-free phase transition in-
tals, in magnetic resonance spectroscopy, etc. volving a lattice shift (see Martensitic transfor-
mation, Polymorphism), and has the appearance
MARGIN OF PLASTICITY of individual crystallites that are smaller than the
Capability of a material to undergo plastic de- initial crystal grains. There is a strict crystallo-
formation without failure. The margin of plasticity graphic interrelation between the crystal lattices of
is measured by the plastic strain before failure. the original and the martensitic phases (so-called
For metals with a body-centered cubic lattice, the orientation relations), which is responsible for the
margin of plasticity correlates with the ratio of the parallelism of some close-packed planes and di-
values of breaking stress σ F to yield stress (see rections of both lattices. As a consequence of the
Yield limit) σ Y . The greater the ratio σ F /σ Y the cooperative and natural (regular) shifting of atoms
larger the margin of plasticity. As σ Y sharply in- during the martensite crystal formation, there oc-
creases with a decrease in temperature the margin curs a macroscopic shift of the specimen surface
of plasticity sharply drops upon approaching the (surface relief arises in the form of ridges and
temperature of the viscous–brittle transition (to the valleys). Martensite crystals have the shapes of
810 MARTENSITE TWINNING
needles, wafers and lenses. The internal structure phases possess close-packed structures, therefore
is more complex, and sometimes includes a high they dissolve more C than the body-centered cu-
density of dislocations (1011 cm−2 ) such as twins, bic lattice. Interstitial carbon atoms cause less
stacking faults, and sets of thin plates. distortions in ε - and ε-martensite, so the hard-
2. In steel: a structural component that is ness and strength of these steels are less than half
an oversaturated solid solution of carbon in a those of hardened carbon steel. In the process of
polymorphous modification of alpha iron. Since deformation, both ε - and ε-martensite transform
austenite dissolves up to 2.0% C at the hardening into α-martensite in the order: γ → ε → ε → α.
temperature, and the body-centered cubic α-iron At 100 and 200 ◦ C, respectively, ε - and ε-mar-
lattice dissolves less than 0.002% C at room tem- tensite transform into austenite, i.e. the reverse
perature, all the carbon goes into martensite as ε → γ and ε → γ martensite transformations
a result of a diffusion-free rearrangement of the take place. In other metals and alloys, which con-
face-centered cubic lattice into the body-centered tain no interstitial atoms of C or N, the properties
cubic one. The carbon atoms are arranged along of martensite and the original phase differ very
the c axis between Fe atoms in octahedral pores, little. In nonferrous alloys (Cu–Al–Ni, Au–Cd)
and form a sublattice that is ordered with respect the possibility of thermoelastic martensite crystal
to C. In carbon steel martensite, however, filling formation exists. On cooling, the dimensions of
all pores along the c axis never occurs (to accom- such martensite crystals increase and, on heating,
plish this, the concentration of C would exceed they decrease, which is due to forward and reverse
its ultimate solubility in austenite). As a result martensite transformations taking place. The for-
of the inclusion of C atoms, the body-centered mation of martensite in steel is accompanied by
cubic Fe lattice becomes significantly distorted, volume increases of 4–5%. As a result, consid-
i.e. a tetragonal or rhombic lattice arises, with erable internal stresses arise, leading to warping
unit cell parameters (a, b, c) that differ from each and failure of products. Besides, internal stresses
another. The degree of tetragonality (c/a ratio) in- result from directional cooperative shear, which
creases linearly with the rise in C concentration, contributes to crystal growth. These stresses are
and when the carbon content reaches 1.8%, c/a partially lowered due to the formation of internal
equals 1.08. Martensite crystal lattice distortions twins (see Twinning of martensite) and dislocations
promote high strength and hardness (high-carbon inside martensite, as well due as to the formation
martensite hardness is about 9.8·108 Pa), and ex- of especially shaped martensite crystals and their
tremely low plasticity of hardened steel. distribution in the elastic matrix of the original
When steel is heated, C leaves the crystal phase.
lattice, forming disperse particles of iron car-
bides. As a result, the degree of tetragonality MARTENSITE TWINNING
is reduced, crystal lattice distortions decrease, See Twinning of martensite.
hardness and strength also decrease, whereas the
plasticity increases. Martensite with such prop- MARTENSITIC POLYTYPES
erties is called tempered martensite. The reverse Polytypes that arise at martensitic transforma-
martensitic transition in carbon steels is not ob- tions (see Polytypism) in a diffusion-free manner,
served because C, on heating, manages to leave and are martensitic phases (see Martensite). Un-
the martensite lattice, and its reverse transition to like equilibrium polytypes, martensitic polytypes
austenite goes through diffusion. In carbon-free al- are nonequilibrium (metastable) and turn under
loys (Fe–Ni), α- and γ -transformations proceed plastic deformation into other polytype structures.
without diffusion. Besides α-martensite (see Al- Martensitic polytypes are generated in metal-
pha phase) with a body-centered cubic lattice, lic alloys with low energies of stacking faults
ε-martensite (see Epsilon phase) with hexago- (γ < 30 mJ cm−2 ), and possess close-packed
nal close-packed lattice and ε -martensite with structures. In alloys with infinite solubility of the
an 18-layered rhombohedral lattice form in man- alloying element (see Alloying) (Fe–Mn, Co–Ni,
ganese steels (>10% Mn). Like austenite, these Co–Cr, Co–Mn), high-temperature face-centered
MARTENSITIC TRANSFORMATION 811
cubic and low-temperature hexagonal close- through nucleation and nuclei growth; therefore,
packed polymorphic modifications are polytypes upon completion of the transformation, the solid is
(see Polymorphism). In alloys with limited solubil- composed of a mixture of martensite crystals with
ity of the alloying element (Fe–Mn–C, Fe–Mn–N, different orientations.
Fe–Mn–Cu, Co–C, Co–Ti, Co–Al, Co–Nb, Co–Ta, The necessary condition for a martensitic trans-
Co–W), there arises a variety of martensitic poly- formation is an increase in the system free energy.
types with various multilayer crystal lattices that However, the transformation, indicated by a down-
contain up to (and over) 100 close-packed layers in ward pointing arrow on the figure, does not begin
unit cells. A variety of martensitic polytypes arise on cooling to the point of phase equilibrium T0 ,
on cooling nonferrous Cu-, Au- and Ag-based where the free energies of original and martensite
alloys as a result of a martensitic transforma- phases are equal, but rather at a lower temperature
tion, and from elastic deformation of Cu–Al–Ni called the martensitic point Ms , and the trans-
monocrystals oriented along the [100] direction. formation is completed at the temperature point
As the external applied stresses are either in- Mf (see Fig.). On heating, the reverse marten-
creased or decreased, transitions between various sitic transformation begins at a temperature As
martensitic polytypes take place, accompanied by and ends at Af , both higher than T0 : the gain in
an abnormally large elastic strain (see Superelas- free energy must compensate for the energy out-
ticity). Martensitic polytypes are generated also in lay (in the form of elastic and surface energy)
a Ti–Ni alloy (nitinol). at the generation of martensite crystals. Marten-
sitic transformations proceed over a wide range of
MARTENSITIC TRANSFORMATION, martensitic temperatures: with the initial stage of the process
phase transition at Ms , and the final stage at Mf . The temperatures
First-order phase transition in solids, which re- of initial and final stages for reverse martensitic
sults in a martensite phase being formed from transformation are As and Af , respectively. The
the original phase in a diffusion-free manner (see kinetics of a martensitic transformation may be ei-
Diffusion-free transformation, Martensite). The ther athermic or isothermal. In the first case some
martensite phase differs by crystal lattice type and martensite (up to 80%) is obtained at the tempera-
by physical and mechanical properties. This trans- ture Ms during a short period of time, but further
formation is one of the most common types of holding at this temperature does not lead to growth
phase transition; it is a special case of a polymor-
phic transformation. According to the G.B. Kur-
dyumov theory a martensitic transformation in-
volves a regular reorganization of a crystal lattice
in which the atoms do not exchange places, but
only displace over a distance which does not ex-
ceed an interatomic spacing. An example is a
transformation between a hexagonal close-packed
and a face-centered cubic structure by the sliding
of close-packed planes. There is a close crys-
tallographic interrelationship between the lattices
of martensite and the original phase (involving
so-called orientation relations); cooperative and
directional atom displacement lead to the gener-
ation of relief on the specimen surface; in most
cases, there is a high formation and growth rate of
martensitic crystals (∼10−7 s); a rapid decay of
the transformation on reaching the specified tem- Diagram of martensitic points, regions where transfor-
perature; a reverse martensitic transformation into mations proceed (diagonally hatched) between the alloy
the original phase on heating. Similar to other first- (A) and martensite (M) phases, and temperature versus
order phase transitions, a martensitic one proceeds alloying element concentration, as explained in the text.
812 MASER
of new martensite crystals. To obtain additional re- MASS SPECTROMETRY, mass spectroscopy
gions of martensite it is necessary to reduce the A technique for investigation of substances,
sample temperature. An isothermal transformation based on determination of masses and quanti-
runs at a constant temperature with a varying rate, ties of ionized atoms and molecules from the
so the relationship between initial rate and temper- material under study. A plot of mass abundance
ature is in the form of a curve with a maximum. versus the mass values is called a mass spec-
Martensitic transformations were first discov- trum. The essence of the technique is as fol-
ered in steels, and then in pure metals (Co, Li, lows: positively or negatively charged ions that
Ti, Zr), metal alloys (Cu–Al, Ag–Cd, Ni–Ti), min- have different masses and move in a vacuum
erals, organic crystals, and in other crystalline are separated according to their mass-to-charge
substances. The product may be thermally stable ratio under the effect of electric and magnetic
phases (e.g., in pure metals, in certain Ti-, Zr-, fields of a mass-analyzer. To make use of the
Co-based alloys) or metastable phases (martensitic technique, the investigated sample is first ionized
steels, phases of Cu–Al, Ag–Cd and Ni–Ti alloys, except for the cases when the substance under
etc.). This transformation finds its greatest practi- study exists in a plasma state (e.g., glow discharge,
cal application in steels (if hardened, steel gains ionosphere). Mass spectrometry exhibits a very
high strength and hardness), and it underlies the high absolute sensitivity (10−12 –10−13 Pa for
shape memory effect and the superelasticity phe- gases, 10−11 –10−13 g for solids) and relative sen-
nomenon. sitivity (10−5 –10−8 % for solids). Mass spectrom-
etry was initially used to determine the isotopic
MASER (acronym for Microwave Ampli ca tion by
composition of atoms, and to make precise mea-
Stimulated Emission of Radiation)
surements of atomic masses. Later, it developed
Quantum generators and amplifiers of the radio
into an analytical method, used in nuclear power
frequency and microwave bands. The term laser
engineering, experimental physics, chemistry, ge-
(Light Amplification by Stimulated Emission of
ology, geological chemistry, biology, archeology,
Radiation) came into use by analogy with maser.
medicine, environmental monitoring, technology
See Solid-state quantum electronics, Quantum am-
processes, space research, etc. Mass spectrome-
pli er.
try allows one to find ionization, excitation, and
MASS, EFFECTIVE
dissociation energies of molecules, heats of evap-
See Effective mass. oration and free formation energies of compounds,
absolute age of geologic samples, and composi-
MASS OPERATOR OF QUANTUM tions of gas mixtures. Mass spectrometry underlies
MECHANICS, intrinsic energy part a wide variety of investigation and analysis meth-
Part of the Dyson equation for interacting ods which are used in solid state physics, and
fermions and bosons (see Polarization operator) differ in the manner of the sample ionization.
that is irreducible (i.e. containing no bare connect- They are: mass spectrometry with surface ioniza-
ing lines in Feynman diagrams) involving Green’s tion, atomic probe method (ionization in strong
functions (propagators) of particles and inter- electric field), spark mass spectrometry (ionization
actions. The quantum-mechanical mass operator in spark discharge), electronic-probe mass spec-
(denoted by Σ or Π ) determines the renormaliza- trometry (ionization of an electronic-beam evapo-
tion of the quasi-particle spectrum, in particular, rated sample by electron impact), secondary ion
of the effective electron mass in normal metals, mass spectrometry (ionization by ion bombard-
due to the Coulomb interaction and the electron– ment and sputtering of solids), mass spectrometry
phonon interaction, as well as the spectrum of in rapid-atom bombardment, laser mass spectrom-
phonons, magnons, and other quasi-particles. The etry (ionization during laser probing of a sur-
anomalous quantum-mechanical mass operator Δ face), thermodesorption mass spectrometry, and
in the Gorkov–Dyson equations determines the the method of electron- and irradiation-induced
energy gap in the elementary excitation spectrum desorption of ions. Methods are also developed on
of superconductors. the basis of mass spectrometry of neutral atoms,
MASURIUM 813
evaporated or sputtered along with ions from the which causes pore healing (sintering). Foreign in-
surface of a solid and ionized by electron im- clusions may move as a whole in external fields
pact, in glow and high-frequency discharges, or such as an electric field E (∇μE = ZeE, where
by means of resonance and nonresonance multi- Ze is the effective atomic charge), a temperature
photon laser ionization. These methods find their field (∇μT = θ∇T /T , where θ is the thermodif-
use in the general analysis of elemental, isotopic, fusion coef cient), or an inhomogeneous vacancy
and molecular composition of the surface and bulk distribution field (∇μv = kB T (∇Cv /Cv ), where
of a solid, in analysis of impurities (dopants) in Cv is the vacancy concentration). For stresses that
gases and other substances, in local and layer- exceed a threshold the volumetric mass transport
by-layer microanalysis of high spatial and layer is determined by directional motion of dislocation
resolution, in studies of processes of diffusion, ensembles, which may be responsible for crystal
segregation, evaporation, adsorption, desorption, creep: ε̇ = ρbV (σ ) (ρ is the mobile dislocation
oxidation, corrosion, catalysis, formation of thin density, V (σ ) is the stress-dependent dislocation
lm s, sputtering, and other aspects of the interac- motion rate, b is the Burgers vector).
tion between energetic beams of atomic particles Surface mass transport may be carried out
and solids. Of most widespread use in solid-state through the mechanism of surface self-diffusion
mass spectrometry are mass-analyzers with sector and through material transport in a near-surface
magnetic and electric fields, quadrupole magnets, gas layer. In the first case, mass transport is de-
and time-of-flight mass separation. scribed by the equation
n D
MASS SPECTOMETRY VIA SECONDARY IONS JS = − S S ∇S μ,
See Secondary ion mass spectometry. kB T
where ∇S μ is the surface gradient of atomic chem-
MASS TRANSFER, ANOMALOUS ical potential, DS is the coefficient of surface
See Anomalous mass transfer. self-diffusion, and nS is the surface atom den-
sity. For a gas mechanism surface mass transport
MASS TRANSPORT may run in an evaporation-condensation mode in-
Process of directional transport of material in volving diffusion in a near-surface gas layer. The
a field that is external or internal with respect to corresponding flux equations have the following
a crystal. Volumetric and surface types of mass form:
transport are recognized. Mass transport occurs
P (k) − P0
during creep, sintering, changes in shape of pores Je-c = ,
and cracks, etc. (2πmkB T )1/2
Volumetric mass transport may be of either a n D
nonthreshold (diffusion) or threshold (dislocation) Jg-d = − G G ∇μG ,
kB T
type. Diffusive mass flux in volumetric mass trans-
fer is defined as where Je-c is the evaporating-condensing current,
Jg-d is the gas diffusion current, P (k) is the pres-
N D
J0 = − 0 0 ∇μ, sure of the vapor in equilibrium with the portion
kB T of surface with local curvature k, P0 is the equi-
where N0 is the number of electrons per unit librium vapor buoyancy, DG , NG , ∇μG are the
volume, D0 is the coefficient of volumetric self- diffusion coef cient, number of atoms per unit vol-
diffusion, T is the temperature, and ∇μ stands for ume, and chemical potential gradient in the gas
the atomic chemical potential gradient produced layer, respectively; m is the atomic mass.
by the field to induce the flux. For a curved sur-
face which bounds a hollow pore or a gas-filled MASS TRANSPORT, SURFACE
cavity we can write ∇μk = αω∇k where α is the See Surface mass transport.
surface tension, k is the surface local curvature,
and ω the atomic volume. The gradient ∇μk di- MASURIUM
rects the flux of crystal material into a pore cavity See Technetium.
814 MATERIALS SCIENCE
excimer molecules ArO, ArN, KrF, XeBr, XeCl2 , MAXWELL RELAXATION, dielectric relaxation
etc., and complicated biomolecules. A number Relaxation of a space charge and its atten-
of compounds have been synthesized which had dant electric field as a result of interactions of
never been observed as a gas, e.g.: MO2 (M = Li, the charged particles that produce it. Some of the
Na, K, Ca, Sr), M(CO)n (M = Ni, Cu, Ag, Au, charge carriers must be mobile and satisfy the con-
U), M(N2 )n (M = Ni, Pd, U). The main meth- dition ωp τ 1, where τ is the carrier mean free
ods of investigations are spectroscopic: photoab- path time, and ωp is the plasma oscillation fre-
sorption, luminescence, Raman scattering of light, quency. The characteristic field change time for
NMR, EPR, etc. These methods allow determin- Maxwell relaxation can be expressed in different
ing the energy spectrum of particles, to diagnose ways: τM = ε/σ , where ε is the dielectric con-
processes of dissociation recombination, diffusion, stant, σ the electrical conductivity; τM = RC,
formation of clusters (see Cluster), and accumu- where R is the sample resistance and C is the
lation of particles, and to study interactions of capacitance; τM = l 2 /D, where l is the Debye
particles with one another and with elementary screening radius, and D the charge carrier dif-
matrix excitations. fusion coef cient. During Maxwell relaxation the
electric field is established by applying a voltage,
MATTHIAS RULE (B.P. Matthias) and the charge results from fluctuating electron
An empirically found regular dependence of thermal motion. The relaxation time τM = ε/σd ,
the critical temperature of superconductors in crys- where σd is the diffusive conductivity, is intro-
talline materials on the number of electrons per duced in the case of a nonlinear current–voltage
atom. A bell-shaped Collver–Hammond curve characteristic. If σd < 0, τM < 0, the charge grows
(M.M. Collver, R.H. Hammond) is an analogue gradually with time, corresponding to a strong
of the Matthias rule for amorphous superconduc- field regime. If conditions opposite to those of
tors. Maxwell relaxation hold then the characteristic
field change time is 1/ωp .
MAXIMUM OVERLAP
See Pauling’s maximum overlap principle. MEAN FIELD APPROXIMATION
The common mean field approximation as-
MAXWELL DISTRIBUTION (J. Maxwell, 1859) sumes that each magnetic ion in a material or
The probability that a large number of identical crystal experiences a local magnetic field that is
classical particles in equilibrium at a tempera- proportional to the magnetization that is present.
ture T , have particle velocity components in the See Molecular eld, Random phase approxima-
range vi → vi + dvi (i = x, y, z): tion, Coherent potential method, Self-consistent
e ld, Hartree–Fock method, Self-consistent eld
f (v) dvx dvy dvz method in superconductivity theory.
3/2
m
= MEAN FREE PATH of particles
2πkB T The average distance which a particle (quasi-
particle in solids) traverses between two succes-
m(vx2 + vy2 + vz2 )
× exp − dvx dvy dvz , sive collisions with particles (or obstacles) of a
2kB T
medium. In the case of relatively rare collisions,
where m is the particle mass, and kB is the its motion can be specified by giving values of
Boltzmann constant. The Maxwell distribution the energy and momentum which change abruptly
describes statistical ensembles of various quasi- at the instant of collision. The path length be-
particles subject to classical statistics, provided tween subsequent collisions depends on the par-
their dispersion law is quadratic (see Boltzmann ticle energy; and to find its mean value it should
distribution). This distribution was corroborated by be averaged over the particle energy distribution
O. Stern (1920) in his molecular beam experi- function. It can involve elastic, inelastic, or ioniz-
ments. ing collisions. In the case of inelastic scattering,
816 MECHANICAL HEAT TREATMENT OF METALS
the interaction is accompanied by the generation creep by 1–2 orders of magnitude. Depending on
or absorption of various elementary excitations the operating conditions, mechanical heat treat-
(see Quasi-particle) in solids. The mean free path l ment requires optimization of preliminary strain in
is related to the cross-section σ for the corre- order to preclude any recrystallization.
sponding process by the expression l = 1/(σ N), For materials with complicated alloying con-
where N is the concentration of scattering parti- ditions, mechanical heat treatment may be com-
cles (impurity atoms, defects, phonons, plasmons, bined with other types of thermoplastic treatments,
and so on). The parameters which determine the e.g., thermomechanical treatment that forms the
efficiency of transport (such as electrical conduc- hardened state (see Hardening of materials) as a
tivity, thermal conductivity, viscosity coefficient, result of an accumulation of dislocation-saturated
etc.) depend on the corresponding mean free paths structures at polymorphic and phase transitions,
of mobile particles. In the case of an electron
ultrasonic treatment, and subsequent aging (see
energy ≈75 eV, the mean free path of the elec-
Alloy aging). Under these, creep resistance, fa-
trons entering the solid before a collision event
tigue strength (see Fatigue), impact strength and
has a minimum value of about 0.5–1 nm. This
resistance to corrosion under stress are increased.
allows the use of electron beams with this en-
A mechanical heat treatment is usually carried
ergy for the purpose of solid state surface analysis
out using standard pressing, rolling and draw-
(see Low-energy electron diffraction, Auger elec-
tron spectroscopy).
ing equipment, or using special equipment de-
veloped for new technologies. The application
MECHANICAL HEAT TREATMENT OF METALS of a mechanical heat treatment for the harden-
Strengthening treatment of metals, which in- ing of products made of nickel alloys, titanium
cludes a slight (10–15%) plastic deformation and alloys, materials containing a metastable β-phase,
polygonization stabilizing annealing. Mechanical molybdenum alloys, tungsten alloys, heat-resistant
heat treatment is an independent type of ther- austenite steels (see Austenite), or rolled aging alu-
moplastic technology; it brings about the im- minum alloys, increases the economic efficiency of
provement of a complex of mechanical properties the use of these materials due to the improvement
(strength and plasticity) through formation of spe- of material properties, the increase of reliability,
cific dislocation substructures that limit the mean and the saving in alloying elements.
free path and mobility of linear defects.
The cause of the strengthening effect of a MECHANICAL PROPERTIES OF SOLIDS
mechanical heat treatment is the formation of a
Ability of solids to resist strain and failure,
higher density of intragrain boundaries in a cellu-
combined with the capacity to undergo elastic de-
lar structure, or a polygonal fragmented structure
formation and plastic deformation under the action
(see Fragmentation), which interact with dissolved
of external forces. Mechanical properties to char-
impurity atoms and mobile linear defects. For
acterize the elasticity of solids include elastic mod-
single-phase metals and alloys that undergo no
uli, proportionality limit and elastic limit. Strength
polymorphic or phase transitions, mechanical heat
treatment is the only effective method for vary- is described by yield limit, ultimate strength, hard-
ing technological and operating properties over a ness, and breaking stress. Plasticity is determined
broad range by choosing corresponding disloca- by relative elongation (shortening) and lateral con-
tion distributions and substructures. traction (spreading) in static testing for tension
In order to increase the workability of brittle (compression), as well as by impact strength in
materials a cellular substructure is formed; while dynamic testing (see also Plastic limit). In cyclic
a polygonal (see Polygonization) substructure that testing (see Cyclic strength), fatigue limit and fa-
contains an excess of dislocations of one sign is tigue strength (see Fatigue) are determined. The
formed when resistance of high-temperature creep mechanical properties at elevated temperatures in-
is to be increased. A polygonal substructure is clude creep limit, long-term strength, long-term
thermally resistant under stress even up to pre- hardness and relaxation resistance. At low tem-
melting temperatures, which decreases the rate of peratures, the mechanical properties of a solid
MECHANICAL TESTING OF MATERIALS 817
with metallic bonds depend on the type of crys- they are due to the generation of radiation-induced
tal lattice. Thus, metals and alloys with a face- defects and to irradiation doping (at large irradia-
centered cubic lattice (Cu, Ag, Au, Al, Pb, Ni, tion doses). Mechanical properties of a solid in a
austenite steels, etc.) on lowering the temper- metastable state change under irradiation because
ature (down to 4.2 K) retain plasticity with a of the progress of radiation-induced processes,
certain increase of strength. At low temperatures, in particular, of radiation-induced diffusion. The
these metals exhibit creep, its rate slightly depend- environment has an effect on the mechanical prop-
ing on temperature. Relative elongation, lateral erties. Thus, testing brittle materials, e.g., NaCl,
contraction, and impact strength of many met- in a saturated water solution causes a noticeable
als and alloys with a body-centered cubic lattice increase in strength (Ioffe effect), whereas dehard-
(W, Mo, Cr, α-Fe, carbon and low-alloy steels) as ening takes place in the presence of surface-active
well as of those with a hexagonal close-packed agents (Rebinder effect, see Strength reduction
lattice (Be, Mg, Zn) tend to zero on cooling be- through adsorption). High pressures increase the
low the viscous–brittle transition temperature Tb elastic moduli and strength of solids. Brittle solids
(cold brittleness). On refining grains as well as exhibit a significant increase of the plasticity limit
on purifying the above metals of impurity atoms, (see Hydroextrusion). The increase of plasticity is
in particular from interstitial impurities, Tb de- especially substantial when a phase with covalent
creases. Typical of mechanical properties of solids or ionic bonding transforms under high pressure
with covalent bonds are high hardness and brit- into a phase with metallic bonding owing to a r st-
tleness (see Brittle failure) even at the high purity order phase transition.
pertaining to many semiconductors; their plastic-
ity manifests itself at temperatures higher than 0.7
MECHANICAL STRESS
of the absolute melting temperature Tm . Mechani-
The force acting on an arbitrarily chosen unit
cal properties of solids with ionic bonds (TiC, WC)
region inside a body; a measure of internal forces
are also typical for brittle materials, but plastic-
arising during strain. Mechanical stresses are quan-
ity manifests itself already at temperatures above
titatively characterized by the stress tensor. They
0.5T m . Mechanical properties of polymeric ma-
stem from interatomic forces, whose range is of
terials have certain specific features: an elastic
the order of interatomic distances. Forces that
strain that arises nearly instantly upon applying
cause internal stresses are referred to in elastic-
an external stress is rather small; a large re-
ity theory as short-range forces which act between
versible deformation (high-elasticity state), and
irreversible viscous-fluid deformation take place neighboring points only. Forces that are exerted on
during a period from a fraction of a second to a part of a body by surrounding regions act di-
many hours, depending on the temperature and rectly through the interface between them. This
polymer structure (see Rubbers). The alloying of does not apply if the deformation of the body is
metals leads to the generation of substitutional accompanied by the generation of space charges
solid solutions (see Substitutional alloys), intersti- and macroscopic electric or magnetic fields in
tial solid solutions (see Interstitial alloys) and/or a it (semiconductors, pyroelectric materials, piezo-
heterogeneous system. Such alloys are stronger electric materials, magnetostrictive materials).
than pure metals, but their plasticity is, as a rule,
lower. Certain metal alloys with clearly defined MECHANICAL TESTING OF MATERIALS
cold brittleness provide an exception, e.g., alloys Tests consisting of determining, by mechani-
of tungsten and molybdenum with rhenium, their cal means, those properties of materials that are
strength growing along with plasticity. Heat treat- responsible for the ability to resist deformation
ment and thermomechanical treatment enhance and failure, together with elastic and plastic be-
mechanical properties of solids that experience havior, under the action of external forces. The
phase transformations (see Polymorphism, Alloy tests are carried out at room temperature, lowered
ordering, Alloy aging). Mechanical properties of and elevated temperatures. Mechanical testing of
solids change under irradiation. Irradiation hard- metals is divided into static types (sample loaded
ening and irradiation embrittlement are observed; slowly and smoothly, or load remains constant
818 MECHANICS
during a long period), dynamic types (sample is failure. The following schemes of loading are used
loaded rapidly, i.e. with grippers moving faster to carry out tests for fatigue: pure bending at rota-
than 1.6·10−4 m/s), and cyclic types (sample un- tion; lateral flexure at rotation; the same operation
dergoes repeated loads that vary in magnitude and in plane; tension–compression; variable torsion;
possibly also in direction). The level of mechani- and internal pressure. In addition, tests with com-
cal properties of a certain material depends on its bined loading are often used. Sometimes, along
nature, and on the nature of the applied stresses with tests for ordinary many-cycle fatigue, those
(type of the state of stress). Therefore, mechanical for few-cycle, high-frequency, impact, thermal,
tests are also classified by types of loading (ten- thermomechanical and corrosion-mechanical fa-
sion, compression, shearing, flexion, torsion, etc.) tigue are used.
that permit testing in a linear, planar or spatial state
of stress. Static testing evaluates such properties MECHANICS
as elasticity, resistance to primary plastic defor- See Analytical mechanics, Continuum mechan-
mation, resistance to major plastic deformation, ics, Physicochemical mechanics, Quantum me-
plasticity. These properties are determined by tests chanics.
for tension, compression, e xure, and torsion. The
MECHANOCALORIC EFFECT (J.G. Daunt,
hardness of a specimen is determined by pressing
K. Mendelssohn, 1938)
into it a rigid tip (indenter) in the shape of a ball,
Heating of liquid helium 4 He in a vessel during
cone or pyramid that experiences no appreciable
a rapid transfer of He II from one vessel into an-
deformation during the test. The state of stress thus
other through a narrow capillary or a slot (∼1 μm)
attained makes it possible to use the hardness de-
at a temperature below that of the transition to
termination method for materials that are brittle in
the superfluid state (lambda point, 2.19 K at stan-
other kinds of testing. Micromechanical testing is
dard pressure), and the cooling of the escaping
used for assessing properties of a small amount of
helium. The mechanocaloric effect has been ob-
material (generally of a metal; e.g. welded joints,
served below 0.4 K (P.L. Kapitza). For a slight
local quenching or strain hardening, degree of
temperature difference
T the flow-over process
inhomogeneity, degree of anisotropy of crystals,
stops at the pressure difference
P = ρS
T ,
etc.). Prolonged static loads are used to determine
where ρ is the helium density, and S is its en-
the resistance to plastic deformation (creep limit), tropy. The mechanocaloric effect is a macroscopic
and resistance to failure (long-term strength); these example of quantum liquid properties. The inverse
tests are carried out under special conditions (ex- phenomenon also occurs: spouting of He through
posure to elevated temperatures, hostile media, a capillary at the addition of heat (thermome-
etc.). Fracture toughness (short-term crack resis- chanical effect). Both effects are explained by the
tance) as an estimate of the resistance of a material quantum theory of super uidity. The increase of
to brittle failure is determined through static tests temperature in a vessel raises the concentration of
for tension or flexure on samples with purpose- quasi-particles inside it. The behavior of He II is
fully induced fatigue crack (see Fatigue). approximately described with the help of two sets
The principal objective of dynamic impact test- of equations: one involving the normal component
ing is to provide the simplest and least labor- and the other the superfluid component of the in-
consuming determination of the hazardous brittle- terpenetrating liquids. The latter has zero entropy
ness of metals for comparison purposes. Dynamic and does not interact with the normal component.
tests evaluate technical properties of metals at in-
creased rates of deformation, when it is necessary MECHANOCHEMISTRY
for structures to experience high-rate loading un- A branch of chemistry concerned with chemi-
der operating conditions. cal and physicochemical transformations that take
Cyclic testing allows the determination of fa- place when a material is exposed to external
tigue resistance, because most working parts of mechanical action: polymer degradation and syn-
devices experience variable loads that cause the thesis of polymers, chemical reactions with fric-
generation of fatigue cracks, their propagation and tion, strain and failure of a solid (tribochemistry),
MELTING TEMPERATURE 819
reactions that take place under exposure to ul- MEISSNER EFFECT (W. Meissner,
trasound and high pressures, etc. Methods of R. Ochsenfeld, 1933)
mechanochemistry are used in plasticization of Phenomenon of absolute diamagnetism of su-
raw rubbers, hydrolysis of cellulose, production perconductors (susceptibility χ = −1 in SI units)
of building materials, ultrasonic manufacturing of involving complete expulsion of the magnetic field
medicinal preparations, for prevention of chemical from the entire bulk superconductor except for
processes that cause aging and failure of assem- a thin subsurface layer (see Penetration depth of
blies and working pieces of machinery, etc. magnetic eld), in fields B not in excess of the
thermodynamic critical magnetic eld Bc (T ) for
type I superconductors, and the lower critical eld
MECHANOSTRICTION Bc1 (T ) for type II superconductors. The Meissner
Generalized term for a phenomenon related to effect is due to the shielding of the external field
strain of a body under the action of external elec- B by undamped surface super currents that exist
tric or magnetic fields. According to the type of via macroscopic quantum coherence and stability
applied field, electrostriction and magnetostriction of the wave function of a superconducting con-
are distinguished. Electrostriction is observed in densate of Cooper pairs, with their lower energy
solid, liquid and gaseous insulators as a result of state separated from the upper excited one-particle
polarization of insulator in an electric field (dis- states by an energy gap of 2Δ (see Landau super-
placement of ions of electric dipoles, or change u idity criterion). Type II superconductors in the
of their orientation). Electrostriction applies to all range of fields Bc1 (T ) < B < Bc2 (T ) exhibit an
solid dielectrics regardless of their structure and incomplete Meissner effect in which the magnetic
flux partially penetrates into the bulk supercon-
symmetry, as distinct from piezoelectric materi-
ductor in the form of Abrikosov vortices (see Mixed
als where the anisotropic structure (see Anisotropic
state, Shubnikov phase). In type I superconductors
medium) is of primary importance. Magnetostric-
of an arbitrary shape and orientation relative to the
tion is strongly pronounced in ferromagnets and
external field B, partial field penetration can occur
in certain ferrites, whereas it is very weak in through the coexistence of normal and supercon-
antiferromagnets, and only rarely occurs in dia- ducting regions (normal domains, see Intermediate
magnets and paramagnets. The deformation of state).
bodies in a magnetic field is also called the Joule
effect (J. Joule, 1842). An inverse phenomenon MELTING
also occurs that involves the change of magnetic The transition of a substance from the crys-
properties of a body under its forced deformation. talline solid state to the liquid one (see States of
In this case, an electromotive force arises in a coil matter). The process of melting involves absorp-
that is wound, e.g., on a ferromagnetic core. The tion of heat (latent heat), and is a rst-order phase
direct and inverse effects are used in sonic and ul- transition.
trasonic emitters and receivers, respectively. The
Wiedemann effect (G. Wiedemann, 1858) consists MELTING HEAT
in the twisting of a current-carrying conductor in a See Heat of melting.
longitudinal magnetic field.
MELTING, INCONGRUENT
See Incongruent melting.
MEDIUM, ACTIVE
See Active medium. MELTING TEMPERATURE, melting point
The temperature Tmelting of the phase tran-
sition of a crystalline solid to the liquid state
MEGA. . . (fr. Gr. μεγ ας , large) under a constant external pressure; a particular
Prefix for a physical unit to obtain a 106 mul- example of a rst-order phase transition. As the
tiple of the original unit, with the symbol: M. melting arises from the breakage of some inter-
Example: 1 MW (megawatt) = 106 W. atomic (intermolecular) bonds, Tmelting depends
820 MEMBRANE
Table 1. Melting temperatures (in ◦ C) of various high melting materials at atmospheric pressure
on the type of interatomic bond, the coordination fixed membrane edge, and initial conditions ξ =
number, and other characteristics of the crystal lat- f (x, y), (∂ξ/∂t) = g(x, y) at t = 0. The natural
tice. The melting temperature of a metal depends frequency spectrum of a rectangular membrane is
on the concentration of the electrons that deter- given by the formula
mine the energy of the metallic bond. See Table 1. 2 1/2
c m 2 n
νik = + ,
MEMBRANE 2 a b
Three types of membranes are as follows:
1. A flexible thin stretched lm that serves (m, n = 1, 2, . . . , a, b are membrane dimensions).
as a sensor surface for a number of acoustic Such an approach leaves out the scattering (emis-
devices. The membrane elasticity arises from ex- sion) of energy, but this does not affect the calcula-
ternal forces that hold it stretched. In this regard tion of natural frequencies. Allowing for emission
it differs from a plate, since the elasticity of the loss is necessary only when calculating amplitudes
latter is a function of its material and width. Ac- of forced vibrations with frequencies that are close
cording to the shape of their border where the to resonance.
tension is applied, membranes are classified as 2. A membrane in engineering is a partition
rectangular, circular, etc. Natural vibrations of a or diaphragm which exhibits selective penetrabil-
membrane are systems of standing waves with par- ity for various compounds, and provides for the
ticular patterns of nodal lines. Different systems separation of those compounds, perhaps with the
of two dimensional standing waves correspond to help of chemical catalytic processes. The mem-
different normal modes, the totality of the latter brane does not clog after long use, as is the case
making up the discrete spectrum of membrane for filtration. Separation of a mixture flux into two
natural frequencies. Transverse vibrations of a ho- components (one passing through the membrane,
mogeneous, uniformly stretched membrane are the other being retained) is due to the selective
described by solutions of the wave equation penetrability. The flux is caused by the application
2 2 of gradients of pressure, concentration, electric
∂2ξ 2 ∂ ξ +∂ ξ ,
= c potential, or temperature to the membrane.
∂t 2 ∂x 2 ∂y 2 3. A membrane in biology is a thin shell, of-
where ξ is the lateral displacement of a point ten a bilayer which is two (perhaps elongated)
(x, y), t is time, c = (T /σ )1/2 is the velocity of molecules thick, confining a cell or subcellular
propagation of the vibrations, T is the tension, formation. It provides sites for enzymes; ensures
and σ is the membrane surface density. Solutions intercellular contacts; participates in processes of
must satisfy the boundary condition ξ = 0 at a motion, secretion, absorption, protein synthesis,
MESIC ATOM 821
and cell division; exhibits electrical activity (neu- shells is 4f 14 5d 10 6s 2 . Ionization energies are
rons) at the propagation of nerve impulses; detects 10.43, 18.752, 34.3, 45.98 eV. Atomic radius is
and transmits information on minor environmen- 0.150 nm; radius of Hg2+ ion is 0.110 nm. Oxi-
tal changes via special molecule-receptors and dation state is +1, +2. Electronegativity is ≈1.6.
molecule-mediators, and facilitates storage of this Mercury is silvery-white liquid metal, the only
information. metal which remains liquid at low temperatures. It
has a rhombohedral crystal lattice; a = 0.2999 nm,
MEMBRANE MATERIALS α = 70◦ 45 . Density is 13.5951 g/cm3 (at 273 K)
Polymeric film (see Polymeric materials), a thin and 13.5459 g/cm3 (at 293 K). Tmelting =
metallic, ceramic or glassy plate, a layer of liquid, 234.3 K, Tboiling = 630 K. Heat of melting is
gas or vapor. Membrane materials are obtained 2.352 kJ/mole; heat of evaporation is
through pressing, rolling, extrusion (stamping), 58.38 kJ/mole; volume thermal expansion coeffi-
leaching, spraying, electrophoresis, polymeriza- cient of solid mercury is 12.5·10−5 to
tion, casting, etching. Preferential and selective 17.1·10−5 K−1 (in temperature range 183.2 to
properties of membrane materials are governed 233.7 K), temperature coefficient of liquid mer-
by the structure of factors that arise during their cury is 1.823·10−4 to 1.889·10−4 K−1 (in tem-
formation: channels, slots, nets. The main func- perature interval 234.3 to 623 K). Specific heat
tion of membrane materials consists in separating is 0.1419 kJ·kg−1 ·K−1 (at 234.3 K), 0.1404
the components of mixtures based on physical kJ·kg−1 ·K−1 (at 273 K), 0.1396 kJ·kg−1 ·K−1
processes: diffusion, evaporation, ultra-filtration, (at 293 K), 0.1371 kJ·kg−1 ·K−1 (at 373 K),
reverse osmosis, etc. Membrane materials are 0.1357 kJ·kg−1 ·K−1 (at 623 K). Electrical re-
widely used for purification and desalination of sistivity is 9.17 n·m (at 273 K), 9.5833 n·m
water, technological separation of gases, in chem- (at 293 K). Mercury is diamagnetic with spe-
ical, food and microbiological industries, in medi- cific magnetic susceptibility −0.168·10−6 CGS
cine. units (at 293 K). Viscosity of liquid mercury is
0.001544 kg·m−1 ·s−1 (at 293 K). Mercury vapor
MEMORY, SOLID-STATE at low temperatures consists mainly of atoms, with
See Solid-state memory. increasing temperature the degree of association
increases, and at the critical temperature it almost
MEMORY, STRUCTURAL totally consists of diatomic molecules. Pressure of
See Structure memory. saturated mercury vapors is 0.00193 Pa (at 203 K),
27.273 mPa (at 273 K), 173.05 mPa (at 293 K),
MENDELEVIUM, Md 1786.49 mPa (at 323 K). Adiabatic elastic moduli
See Transuranium elements. of mercury monocrystal: c11 = 36.0, c12 = 28.9,
c13 = 30.3, c14 = 4.7, c33 = 50.5, c44 = 12.9
MENISCUS (fr. Gr. μηνισ κoς , half-moon, (in GPa) at 83 K. Mohs hardness of frozen mer-
crescent) cury is 1.5. Compressibility of liquid mercury is
Shape of surface of liquid in a narrow vessel or 40.51 pPa−1 (at 303 K and pressure from 105 Pa
a tube. Wetting liquids form a concave meniscus, to 5·106 Pa); compressibility coefficient of solid
and non-wetting liquids form a convex one (see α-Hg is 35.2 pPa−1 (at 200 K and normal pres-
Wetting). sure). Mercury is widely used in the chemical
industry, electronics, lighting engineering, etc.
MERCURY (Lat. hydrargyrum ), Hg
A chemical element of Group II of the pe- MESIC ATOM, muonic atom
riodic system of elements; atomic number 80, Atom of some chemical element where one of
atomic mass 200.59. Natural mercury has 7 stable the electrons is replaced with a negatively charged
isotopes: 196 Hg, 198 Hg, 199 Hg, 200 Hg, 201 Hg, muon (μ− ) which, like an electron, is also a lepton
202 Hg and 204 Hg; 23 radioactive isotopes are and behaves like a heavy electron. When the sub-
known. Electronic configuration of filling outer stance is irradiated with fast muons, mesic atoms
822 MESONIC CHEMISTRY
form as the muons are decelerated: μ− is trapped allows identification of states of hydrogen in var-
at a high-excitation state and replaces one of the ious media. Methods of X-ray spectrum analysis
atomic shell electrons. The transition to lower have been used for mesic atoms. Channeling be-
energy states is accompanied by radiation of char- havior of π + -mesons and μ+ -muons have been
acteristic gamma quanta, or by the Auger effect. observed.
Intensities of certain lines of characteristic radia-
tion by a mesic atom depend on the nature of its MESOPHASES
bonding to nearest neighbors, which allows the de- Thermodynamic phases of liquid crystals char-
termination of the type of chemical compound that acterized by specific ranges of temperature, pres-
is involved. As the transition of the muon to the sure, and concentration. In thermotropic liquid
K-shell takes place, the electron shell of a mesic crystals, mesophases are found between the iso-
atom with atomic number Z becomes similar to tropic fluid (I ) and solid crystal phases (S). The
that of an ordinary atom with number Z − 1 be- normal sequence of mesophases follows the de-
cause of the small Bohr radius of the bound muon. crease of degree of ordering with increase of
The small sizes of mesic atoms with hydrogen temperature: S ↔ smectic B(H ) ↔ smectic C ↔
nuclei (radius of 10−11 cm) allow them to pen- smectic A ↔ nematic N ↔ (cholesteric-Ch) ↔ I .
etrate into electron shells, and even into nuclei of In certain liquid crystals and their mixtures there
other atoms. This results in the generation of mesic can be deviations from this sequence, e.g., C ↔
molecules, and causes reactions that involve nuclei N ↔ A ↔ N ↔ I . Melting points of mesophases
and muons (muon catalysis). range between −60 and +400 ◦ C, and the width
of mesophases varies from 0.01 to ∼100 ◦ C. Tran-
sitions S ↔ N are rst-order phase transitions
MESONIC CHEMISTRY
with
H ∼ (2–15)·103 cal/mole. Phase transi-
Chemistry of compounds containing mesic
tions N ↔ I and N ↔ A are also first order
atoms. Replacement of one of the electrons with a
with
H ∼ 10−1 –10−2 and 1–10−2 kcal/mole,
negatively charged meson (muon, π - or K-meson)
respectively. In lyotropic liquid crystals, thermal
may take place as these particles are decelerated
transitions N ↔ I are first order phase transitions
in the material. Since masses of mesons are hun-
with
H ∼ 10−3 kcal/mole. Transformations of
dreds of times that of an electron, the Bohr radius
mesophases are usually second-order phase tran-
of a mesic atom appears to be significantly smaller
sitions.
than that of an ordinary atom. This results in the
capability of the simplest mesic atom (mesonic MESOSCOPIC FLUCTUATIONS
hydrogen) to penetrate inside atomic shells of or- Scatter of observed parameters over an en-
dinary atoms and initiate nuclear transformations, semble of specimens that exhibit identical macro-
formation of mesonic molecules, etc. scopic characteristics (dimensions, shape, impu-
rity concentration, etc.) but differ in the specific
MESONS (fr. Gr. μεσ oς , average, intermediate) configurations of static disorder.
Unstable elementary particles that possess zero It is reasonable to describe mesoscopic fluc-
or integral spin. The first to be discovered and tuations in statistical terms. This approach seeks
the best studied are π - and K-mesons that have neither to depict atomic level motion of a par-
masses between those of a proton and an elec- ticular specimen (microscopics, see Microscopic
tron; see also muons (at first erroneously called description of solids), or to limit itself to comput-
μ-mesons). Intense meson beams are obtained in ing ensemble-averaged observables (macroscop-
proton accelerators, the so-called meson factories. ics, see Macroscopic description of solids). Rather,
This allows using mesons to study solids. The its aim is to study the distributions of mesoscopic
method of charge exchange of π -mesons on hy- (intermediate length scale) fluctuations. The ir-
drogen nuclei is known; the probability of charge regular dependence of these fluctuations on an
exchange depends on the atomic number of the external field allows one to observe them despite
nucleus that is bound to hydrogen, as well as the fact that it is a particular specimen under ex-
on the nature of the chemical bond. This method perimental investigation, rather than the overall
METAL–INSULATOR–SEMICONDUCTOR STRUCTURE 823
METAL, AMORPHOUS
See Amorphous metals and metallic alloys.
METAL, ANTIFERROMAGNETIC
See Antiferromagnetic metals.
applied to the gate, then all the electrons under low melting points, high forgeability and plasticity.
it in p-Si will become attracted to the thin oxide On the other hand, transition elements possessing
layer, thereby creating a semiconducting inversion several d-electrons (up to 8–9, as in Ni, Pd, Pt)
layer of n-type in it. This results in the generation exhibit a high metallic bond energy; and they are
of a current-carrying channel between the source refractory and brittle, like, e.g., W. They have a
and the drain. Such a system is equivalent to a large covalent bond contribution from highly lo-
vacuum triode (source is cathode, drain is anode, calized d-electrons. Ions that differ very little in
gate is grid). It may also serve as a memory unit. size easily replace one another, and this facili-
A two-layer insulator is used for this purpose: thin tates the formation of alloys over a wide range of
layers of SiO2 and of silicon nitride. Electrical composition, e.g., alloys of gold with copper and
charge brought into Si may be transferred from silver, bronzes, brasses, alloys of tin with lead, and
Si to traps on the oxide–nitride interface. These many others. Intermetallic compounds InSb, InBi,
traps remain charged for a relatively long time af- HgTe have both metallic and covalent bond contri-
ter removing the voltage between the gate and the butions. Tin can form different phases: a zero-gap
substrate (memorization). This state may be read covalent crystal α-Sn, and metallic β-Sn.
out by the change in properties of the subsurface
area of the substrate. The MOS structure is a basic METALLIC CORROSION
element of solid-state electronics; it is also used See Corrosion of metals.
for studying surface properties of semiconductors
METALLIC GLASSES
(near the boundary with an insulator).
Amorphous metallic alloys that are obtained
through ultra-high-speed melt cooling. The rate of
METAL–INSULATOR TRANSITION
cooling must be about 106 K/s to prevent crystal-
See Anderson localization, Mott metal–insulator
lization of the melt; this is attainable if at least one
transition, and Peierls transition.
dimension of the specimen to be hardened is small.
Therefore, metallic glass specimens have the form
METALLIC BOND of foils, strips, wires with thickness of tens (more
Type of chemical bond in materials exhibiting rarely, hundreds) of micrometers.
metallic properties. Most atoms of the first three Types of metallic glasses. Melts capable of
groups of the periodic table, as well as transition turning into glasses are called glass forming, and
elements, have their electronic structure signifi- this capability depends on the melt composi-
cantly modified when included in crystal lattices tion. Several families of two-component metal-
of metals and alloys. Their weakly bound outer lic glasses are: transition metal–polyvalent metal
s- and p-electrons lose their affiliation with their (e.g., Fe80 B20 , Pd80 Si20 ); polyvalent metal–poly-
“own” atoms and move more or less freely through valent metal (e.g., Mg70 Zn30 ); “heavy” transition
the whole crystal. The most typical example is metal–“light” transition metal (e.g., Ni60 Nb40 );
Group I metals Li, Na, K as well as Cu and Ag, rare-earth metal–polyvalent metal (e.g.,
which serve as univalent cations immersed in an La70 Al30 ); rare-earth metal–monovalent noble
electron fluid. Compared to an isolated atom with metal (e.g., La80 Au20 ); transition metal–“light”
a wave function that decays rapidly with distance, transition metal (e.g., Cd70 Ni30 ); uranium–tran-
a nearly free electron has a wave function of a sition metal (e.g., U70 Cr30 ). Multicomponent
smoother character, corresponding to a kinetic en- metallic glasses include atoms of more than one
ergy far below the Fermi energy. It is this energy metal from a given group instead of only one of
decrease, along with the Coulomb energy of an ion them; e.g., the alloy Fe40 Ni40 P14 B6 belongs to
in a roughly homogeneous electron fluid, that is re- the transition metal–polyvalent metal family.
sponsible for the metallic bond energy. Typical of Structure. Integral and partial radial distri-
metals are highly symmetric body-centered cubic, bution functions of atoms in metallic glasses,
face-centered cubic and hexagonal close-packed obtained through various diffractometry analysis
lattices with high coordination numbers 8, 12 and methods, indicate the presence of short-range or-
12, respectively. This explains their comparatively der (see Long-range and short-range order) in
METALLIC GLASSES 825
the atomic arrangement extending over several in- may be several types of local order which may
teratomic distances. Mössbauer spectroscopy and alternate randomly. Cluster boundaries are locally
other investigation methods indicate the presence disordered and contain tunneling states responsi-
of one or more types of local atomic ordering, ble for low-temperature anomalies in amorphous
and studies of surface structure of some metal- solids.
lic glasses using the scanning tunneling micro- Properties. The density of metallic glasses is,
scope (see Scanning tunneling microscopy) and as a rule, 1–2% lower than that of the crystal al-
the high-resolution electron microscope reveal or- loy of the same composition. Metallic glasses owe
dered areas up to 2.5 nm in size. With the help their high strength to their lack of topological or-
of transmission electron microscopes, a domain- der (ultimate strength approaches the theoretical
type structure was discovered in a number of value). The bulk modulus is nearly the same and
glasses, the domain sizes being of the order of the shear modulus is 10–30% smaller than that of
the crystal analogue. Plasticity is about 1% at tem-
10 nm. A number of models of metallic glass
peratures below 0.7Tg (where Tg is the vitri ca tion
structure have been suggested. The random close-
temperature) and up to several tens of percent at
packing model, supplemented by concepts of free
temperatures close to Tg . Plastic deformation is
volume (see Free volume model), dates back to
uniform and is described via Newton’s equation
Bernal’s liquid model, but fails to allow for in-
at T ≈ Tg . At low temperatures, plastic deforma-
termetallic bonds and the existence of local topo- tion is nonuniform; and slip bands about 20 nm in
logical and compositional order. In polycrystal thickness are generated, separated by up to several
models an amorphous body is considered as a μm. Uniform plastic deformation in polyclusters is
dislocation- and disclination-disordered crystal, or a diffusely viscous flow controlled by diffusion and
a fine-grained polycrystal. The distances between slip over cluster boundaries. At low diffusion mo-
extensive defects in paracrystals must extend over bility of atoms, this flow switches to a nonuniform
several interatomic distances, but as yet there is deformation mode accompanied by the generation
no experimental proof of their existence in metal- of slip bands. In the free-volume model, plastic de-
lic glasses. A polycluster model (see Polycluster formation is associated with realignment of units
amorphous solids) treats a metallic glass as an ag- containing free volume.
gregation of locally matched, regular clusters with Due to the lack of thermodynamic equilibrium,
each atom in the cluster locally ordered. There metallic glasses undergo structural relaxation and
826 METALLIC HYDROGEN
crystallization with a typical crystallization en- Many iron-, cobalt- and nickel-based metallic
thalpy of 1–5 kJ/mole. The rate of structural glasses of the “transition metal–polyvalent metal”
relaxation and crystallization is determined by the type, which contain about 80% transition met-
speed of diffusion displacements of atoms, and als, are magnetically ordered media (see Amor-
the rate of formation of stable configurations and phous magnetic materials, Amorphous magnetic
crystalline-phase nucleating centers by atoms. In- substances). Their Curie point may be changed
stead of compositional phase diagrams, structural over a wide range (up to approximately 1000 ◦ C)
transformation diagrams in temperature–time co- by varying the composition of the alloy. Corrosion
ordinates are constructed for metallic glasses. The properties of metallic glasses are, as a rule, notice-
figure shows the structural equilibrium diagram of ably better than those of their crystal analogues.
an alloy, with the thin line on the left correspond- Because of their extraordinary mechanical, elec-
ing to the curve of the melt cooling regime which tric and magnetic properties, metallic glasses find
ensures the formation of a metallic glass. various practical applications. Their applicabil-
Diffusion coef cients of intrinsic atoms in
ity is limited by comparatively low crystallization
metallic glasses are, as a rule, noticeably greater
temperatures, insufficient thermal stability, and
than those in crystals at the same temperature. Mi-
small thickness of the materials that are obtained
croscopic mechanisms of diffusion are not com-
in the form of strips and wires.
pletely clear. In a free-volume model, the diffusion
occurs along free cavities, the volumes of the latter
being comparable to atomic volumes. In poly- METALLIC HYDROGEN
clusters two-dimensional diffusion occurs along A hypothetical high-pressure phase of solid
cluster boundaries, while volume diffusion in clus- hydrogen which exhibits metallic properties. Ac-
ters occurs mostly along vacancies. cording to general concepts of a metal–insulator
The electrical resistivity of metallic glasses at (Peierls) transition, the change of electric prop-
temperatures near the Debye temperature is close erties is due to the broadening of energy bands
to that of the melt. The thermal resistance co- and the disappearance of the band gap between
efficient may have a small positive, negative or metal and insulator, with strong compression of
zero value, depending on the composition of the the material. Hydrogen becomes a conductor but
glass. The Mooij correlation (J.H. Mooij, 1973) retains the hexagonal close-packed structure of a
between the resistivity and the thermal resistance
molecular crystal by the gradual overlapping of the
coefficient is observed. Some amorphous alloys
gap between the valence band and the conduction
are superconductors (see Superconducting alloys).
band. The further increase of pressure promotes
The transformation temperatures of amorphous al-
the transition to the atomic metal phase with a
loys into the superconducting state are, as a rule,
body-centered cubic structure that is typical of
lower than those of their crystal analogues, and
alkali metals. Then, the appearance of one par-
usually do not exceed 9 K. Yet, amorphous super-
conductors find practical application because of tially filled band to replace two overlapping ones
their narrow range of transformation temperatures provides high electrical conductivity. The exper-
(about 0.05 K) and their high plasticity. Metal- imental evidence indicates only a tendency for
lic glasses feature low-temperature anomalies in such a rearrangement into an atomic phase: as
the heat capacity, heat conductivity, velocity of evidenced by Raman scattering of light, the vi-
propagation and absorption of sound, which are bration frequency of H2 decreases under pressure
typical of various types of glasses, and result from above 0.3 Mbar, i.e. the strength of the bond
characteristics of their atomic structure, in partic- decreases. The formation pressure of metallic hy-
ular, from the existence of two-level systems. At drogen is theoretically estimated to be of the order
low temperatures, features of the resistance and of 2–7 Mbar. An important question is the stability
magnetoresistance temperature dependences are of metallic hydrogen after removing the applied
found, which result from weak localization of elec- pressure. There is particular interest in the prob-
trons. lem of metallic hydrogen because of its possible
METALLOPORPHYRINS 827
onset of superconductivity with a high critical tem- during the preparation of the specimen. Metallo-
perature, perhaps Tc ≈ 100 K. graphic analysis uses a wide range of techniques
and procedures. For bright-field illumination, the
METAL-LIKE COMPOUNDS specimen is positioned perpendicular to the optical
Compounds of metals with nonmetals (hydro- axis of microscope. Oblique illumination requires
gen, bromine, carbon, nitrogen, silicon, phospho- either displacement of the light source, or illumi-
rus, sulphur, germanium, arsenic, selenium, and nation of the specimen from under the objective
tellurium). The nature of the chemical bond in lens; this allows increasing the contrast of the
metal-like compounds determines their proper- picture, but is difficult to implement when high
ties. The presence of electrons delocalized over magnification is needed. Dark-field illumination
the crystal is responsible for their crystal chem- is achieved by placing the specimen in the ray
istry characteristics: face-centered cubic, body- that bypasses the objective lens and is directed
centered cubic, hexagonal close-packed or more by special reflectors. The contrast achieved by
complex structures, considerable electrical con- using dark-field illumination is opposite to that
ductivity (with high negative temperature coeffi- of bright-field illumination, i.e. those structure
cient), thermal conductivity, and high melting tem- details, which appear dark under bright-field il-
peratures. Unlike metals, metal-like compounds lumination, appear bright under dark-field illumi-
exhibit high hardness, small values of the ultimate nation, and vice versa. The microscope resolution
strength and plastic limit at low temperatures, com- does not decrease with dark-field illumination.
paratively low values of the linear temperature ex- Special methods of metallographic analysis
pansion coefficient, and low resistance to thermal include the phase contrast method, interference
shock. Some metal-like compounds are actually method, those of high-temperature and low-tempe-
superconductors. Most metal-like compounds are rature analysis, those based on the application of
phases of variable composition, whose properties polarized light, etc. Metallographic analysis is one
depend on the ratio of metal and nonmetal compo- of the main techniques for structure studies in the
nents. Transformation into semiconducting phases physics of metals, metal science, materials sci-
may also occur. ence, and heat treatment.
is due to numerous chemical modifications of heat treatment, and other kinds of rolled stock
the molecules which do not effect the system treatment, cold plastic deformation, applica-
of π -electrons. Metalloporphyrins of most met- tion of protective, decorative and other kinds
als (excepting alkalis, alkaline earths, lanthanides) of coatings, and so on.
are extra-stable complexes (stability constant in
glacial acetic acid is 1035 –1045 ). Materials pro- At present, the principal process is the two-
duced on their basis exhibit a wide range of useful stage (coke/blast-furnace) method of metal mak-
features. They are sensitive and selective cata- ing that is based on the production of cast iron and
lysts for many chemical processes; when included
reprocessing it into steel. A radical innovation was
into a thin film of polymer they act as photo-
the development of continuous feed casting ma-
electrochemical energy converters. “Organic met-
als” exhibit properties of semiconductors; elec- chines, which led to abandoning the reprocessing
trochromic materials have a long working life; of ingots into feeds (blooming-slabbing mills) and
photosensitive materials are used in phase three- reduced the consumption of rolling steel. How-
dimensional holography, etc. ever, there are disadvantages inherent in this two-
stage method (high coke consumption with limited
METALLURGY (fr. Gr. μετ αλλoυργ ια , reserves of coking coal; minimal prospects of
metallurgy) equipment productivity growth; unfavorable eco-
Study of metals, including processes of extrac- logical aftereffects, etc.) that inspired a renewed
tion from ores and other materials, and subsequent interest in direct processing metallurgy. This in-
treatment in liquid and solid states, to provide volves the direct production of iron according to
them with specified shapes and properties. The
the scheme: shaft metallization furnace–electric
most important metallurgic products are hot- and
furnace–continuous feed casting machine–rolling
cold-rolled bar and sheet products, pipes, metal ar-
mill, without any blast-furnace or accompanying
ticles. By-products include chemical compounds,
fertilizers, building materials, etc. Ferrous and coke-chemical and agglomeration process stages.
nonferrous metallurgy are known. Ferrous metal- It may also be possible, through improved
lurgy covers the manufacture of steels and iron- methods of iron ore dressing which prepare prac-
based alloys (cast iron, steel, ferrous alloys), while tically pure iron oxides with economically accept-
nonferrous metallurgy involves nonferrous metals able performance, to employ deep reduction to
and alloys based on them. Powder metallurgy in- obtain iron powders for the manufacture of work
volves the preparation of ferrous and nonferrous pieces, or to press compact feeds for rolling. This
metal powders that are transformed into finished could provide finished products through bypassing
parts by pressing and subsequent sintering.
the blast-furnace and steel-making process stages,
The production system of modern metallurgy
without liquefying the metal (so-called solid-state
includes mining complexes and concentrating
mills, by-product coke plants, factories specializ- metallurgy).
ing in ferroalloys and refractory materials, metal-
lurgical and metal-working combines and enter-
METALLURGY, ION
prises. There are four process stages:
See Ion metallurgy.
(1) metallurgical production proper includes
blast-furnace involvement,
(2) steel-making stage (open-hearth furnaces, METALLURGY, POWDER
converters, electric furnaces),
(3) rolling-mill procedures (roughing and billet See Powder metallurgy.
mills, sheet, plate and section mills),
(4) there is an ever-increasing significance and
wide application of operations of heat treat- METAL MODIFICATION
ment, thermomechanical treatment, chemical See Modi cation of metals.
METALS 829
influence various thermodynamic characteristics resistance of pure metals (e.g., W, Ga, Mg) is 105
of metals, e.g., terms linear in temperature are times smaller than their resistance at room tem-
added to the specific heat and thermal expansion perature. The electron mean free path l in these
coefficient. Direct proof of the electronic nature metals at T → 0 is as long as several centimeters.
of metallic conduction is provided by electrical At low temperatures samples of the same metal,
inertia experiments. Most crystals composed of differing in impurity content, follow the relation-
identical atoms are metals, with only 19 elements ship ρ = ρresidual + ρideal called the Matthiessen
forming nonmetallic crystals under equilibrium rule (L. Matthiessen, 1864) where ρideal is the re-
conditions. Under changes of temperature, pres- sistance due to scattering of electrons by thermal
sure and magnetic field, crystals may undergo lattice vibrations. At T
ΘD (ΘD is the De-
structural phase transitions (so-called polymor- bye temperature) ρideal ∝ T , while for T ΘD
phism). Crystals that are not electrical conductors the temperature dependence has the typical form
may, in this process, become metals. For instance, ρ ∝ aT + bT 5 . Here, a and b are constants that
under high pressure the semiconductors silicon measure the scattering of electrons by electrons
and germanium become metals, while metallic and thermal lattice vibrations (thermal phonons),
crystals may become nonmetals (e.g., white tin
respectively. Highly disordered metallic systems
is a metal, and gray tin a semiconductor). Under
(amorphous metals) may experience Anderson lo-
elastic deformations that have no effect on the lat-
calization that leads to a metal–insulator phase
tice symmetry, electronic topological transitions
transition through quantum interference phenom-
(see Lifshits transition) caused by abrupt change in
ena, the latter arising due to repeated scattering
dynamical properties of conduction electrons are
observed in a number of metals. The concept of of electrons by static defects. Superconductivity
metallic structure as an ionic skeleton surrounded occurs in 25 pure metals and numerous metallic
by an “electron fluid” that compensates for the alloys.
forces of repulsion between ions and binds them The superconducting state in metals is due to
in a solid is a fairly good representation of reality. the formation of a condensate of Cooper pairs
Electrical and thermal properties of metals. in the degenerate electron gas as a result of the
According to the electronic theory of metals, elec- electron–phonon interaction. The values of the su-
trons are capable of moving through an ideal perconducting transition temperature Tc for pure
lattice without any resistance. The appearance of metals range from 0.026 K for Be to 9.25 K for
electrical resistance is due to electron scattering by Nb; the highest Tc for a classic metallic compound
faults in the lattice due to atomic thermal motion, being 23 K for Nb3 Ge. Cuprate high tempera-
the presence of impurity atoms, vacancies, dislo- ture superconductors have transition temperatures
cations, and other static defects. Conduction elec- as high as 133 K. The transition into the super-
trons in metals are charged low-energy excitations conducting state in an external magnetic field is
(quasi-particles) of an electronic Fermi liquid, and accompanied by expulsion of the field from the
form a slightly imperfect Fermi gas with degener- bulk metal (Meissner effect). The superconducting
acy temperature T0 = εF /k B ∼ 105 K (εF is the state is destroyed by an applied magnetic field that
Fermi energy). These values of T0 correspond to exceeds a certain critical value Bc . For pure met-
characteristic electron velocities ∼108 cm/s. With als, Bc ≈ 10–100 mT; for alloys, Bc ≈ 1–10 T.
decreasing temperature, the amplitude of atomic Electrons participate in the transfer of electric
oscillations decreases with the result that the elec- charge as well as heat.
tron mean free path l increases. This behavior of There is a simple relationship
l causes an increase of conductivity with the de-
crease of temperature, which is typical of metals. K π2 kB 2
At T → 0 the electric resistance of actual metals =
σT 3 e
saturates, that is it tends to a limit referred to as the
residual resistance ρresidual which is a measure of between the thermal conductivity K and the elec-
the lattice perfection. The criterion for this satu- trical conductivity σ called the Wiedemann–Franz
ration is kF l > 1 (Ioffe–Regel rule). The residual law. This law is valid in temperature regions where
METALS 831
inelastic processes are negligible in electron scat- emission) takes place. The number of depart-
tering. Metals display a number of thermoelectric ing electrons is proportional to exp[−U/(kB T )],
phenomena. where U is the work function (Richardson equa-
An external magnetic field bends electron tra- tion). The escape of electrons from a metal may
jectories, and thus affects transport processes in also take place without heating, through the tun-
metals (galvanomagnetic effects and thermomag- nel effect that arises when an electric field is
netic phenomena). This effect is largest when the applied ( eld electron emission). Substantial field
condition rL l (here rL is the ion Larmor radius) emission is observed in fields E ≈ 107 V/cm.
is satisfied. At T = 4 K, the resistance of some The magnitude of the field-emission current is an
pure metals increases in a magnetic field by a fac- exponential function of E, and is practically inde-
tor of 104 to 105 . Studying galvanomagnetic and pendent of temperature. The electrical resistance
thermomagnetic phenomena in monocrystalline of metals changes abruptly on melting (increases
metals at rL l provides the topology of the two- or threefold); three known exceptions being
Fermi surface, while the concentration of charge Bi, Sb and Ga: whose resistance decreases at melt-
carriers in a metal is determined by the Hall effect. ing.
A magnetic field makes possible a nonuniform Optical properties of metals; X-ray spectra.
current distribution in a metal across the spec- The characteristic luster of metals is caused by a
imen cross section (static skin-effect). Interband high reflectivity (exceeding 99% in many metals)
tunneling of electrons in a magnetic field (mag- for electromagnetic waves with frequency ω lower
netic breakdown) was detected in many metals. than the plasma frequency ωp (radio waves, IR,
Under conditions of magnetic breakdown, the ki- visible region). At ω > ωp , metallic optical prop-
netics of electrons in pure metals is governed by erties are close to those of insulators. In this case,
quantum interference of electron waves that arise of key importance are quantum effects – electron
in band-to-band tunneling. Such interference may transitions between energy levels with the absorp-
transform a pure metal into an incommensurate tion of light quanta, and γ -quanta of X-radiation
system with absolute quantum localization of a (inner level photoemission effect). For instance,
certain group of conduction electrons. Due to the the reflectivity of silver is 35% in the visible re-
high value of the metallic conductivity, a high- gion and decreases to 4.2% in the UV (reflection
frequency electromagnetic wave cannot penetrate by glass). In the X-ray region metals differ from
far into the metal depth, i.e. practically all the insulators in optical properties by the structure of
high-frequency current flows in a narrow skin- the absorption band edge. In a metal, the edge of
depth layer near the specimen surface (skin-effect). the soft X-ray absorption band is sharply cut off,
In typical metals, the depth of the skin-layer is δ ∼ which reflects the degeneracy of the electron dis-
10−6 cm at the frequency ω ∼ 108 Hz. Depend- tribution in the conduction band. X-ray absorption
ing on the ratio between the values of l and δ, two bands contain information on the energy spectrum
types of skin-effect are distinguished: the normal of a metal, i.e. they allow determining the overlap
skin-effect (δ
l) and the anomalous skin-effect of energy bands by impurities
(δ l). Two types of phenomena are observed Mechanical properties of metals. Not only do
in a magnetic field under the conditions of the free electrons define the electronic and thermo-
anomalous skin effect. The first is cyclotron reso- physical properties of metals, but they also notice-
nance that arises when the Larmor frequency of ably affect the cohesive forces of atoms in a lattice,
some group of electrons coincides with the fre- as well as the elasticity, strength and plasticity.
quency of the applied high frequency field. The A salient feature of a metallic bond consists in the
other includes various kinds of slightly decaying decrease of the kinetic energy of a valence electron
electromagnetic waves that cause an anomalous compared to its value in an isolated atom. The spa-
penetration of the high-frequency field into the tial directionality of p- and d-electron atomic or-
metal to depths considerably exceeding δ. bitals leads to substantial anisotropy of elasticity,
At high temperatures, the evaporation or emis- strength, and plasticity properties of monocrystals.
sion of a small number of electrons (thermionic Elastic properties of metals depend only slightly
832 METAL SCIENCE
on the temperature. The role of conduction elec- lattice parameters. At low temperatures in mag-
trons makes itself evident in the change of me- netic fields, metal single crystals exhibit compli-
chanical properties of metals at a transition into cated oscillating relationships between the mag-
superconducting states, e.g., the elastic limit of Pb netic moment and the field (see De Haas–van
becomes 5% lower, i.e. the process of deharden- Alphen effect), which is due to the Landau quan-
ing takes place. According to the nature of the tization of the energy spectrum of conduction
thermal change of plasticity metals can be divided electrons in a magnetic field (see Landau dia-
into two groups: those that remain plastic down magnetism). After investigating the de Haas–van
to liquid-helium temperatures, e.g., Al, Cu, Ni, Alphen effect along various crystallographic di-
Pb, and those that become brittle under cooling, rections, one can reconstruct a Fermi surface, and
the so-called cold brittle metals (see Cold brit- find the effective mass of an electron in each of its
tleness), e.g., α-Fe, Mo, Cr. Plastic deformation extremal cross-sections.
of crystals occurs in the form of slip, twinning
of crystals, faulting. The carriers of the elemen- METAL SCIENCE, physical metallurgy
tary act of plastic deformation are dislocations. Comprehensive study of the composition,
The theoretical strength of metals is observed structure and properties of metals (alloys), as well
on perfect (dislocation-free) crystals; the actual as their changes in response to thermal, chemical,
strength is smaller by factor of 100–1000 because mechanical, and other types of influences; a sub-
of structural imperfections. Mechanical properties field of materials science. It provides the scientific
of metals depend on a number of external fac- basis for synthesizing metallic materials (e.g. al-
tors (temperature, force application rate, type of loys) with controlled properties. Metal science is
closely related to the physics of metals.
stressed state) and internal factors (chemical com-
position, microstructure). Application of alloying
METALS, CREEP
and various kinds of heat treatment and ther-
See Radiation-induced creep.
momechanical treatment (hardening, programmed
hardening and other methods) substantially ex-
METAL–SEMICONDUCTOR JUNCTION
tend the range of practically important mechanical
The area around the contact or junction be-
properties of metals.
tween a metal and a semiconductor which in-
Magnetic properties of metals. Essential to cludes a transition layer with a space charge
magnetic properties of metals is the division of region. Together with a semiconductor junction
metals into two groups. Weak magnetic materials and a metal–insulator–semiconductor structure the
form a group of metals exhibiting either diamag- metal–semiconductor junction is one of the main
netism or paramagnetism. Strong magnetic mate- elements of semiconductor electronics, and is
rials exhibit paramagnetism and, with decreasing widely used as an ohmic contact.
temperature, attain magnetic ordering that results The properties of this junction arise either from
from the cooperative interaction of electrons from the difference between the work functions ϕM
inner incomplete d- and f -orbitals of an atom and ϕS of the metal and semiconductor, or are
(see Ferromagnetism of metals and alloys and An- determined by the surface states on the phase in-
tiferromagnetism). In magnetic fields that exceed a terfaces. In the first case the surfaces in contact
certain, so-called critical value Bc , antiferromag- become oppositely charged due to the difference
netic metals may turn into either ferromagnetic in the movement of electrons from the metal
or paramagnetic states. The magnetic structure of and semiconductor, and a space charge region is
an antiferromagnetic metal may be either com- formed. This results in the energy band bending
mensurate or incommensurate with their crystal- in the neighborhood of the contact (see figure).
lographic structure. Rare-earth metals (see Mag- An n-type semiconductor with an upward turning
netism of rare-earth metals) exhibit complicated band curvature has its space charge area depleted
atomic magnetic structures: spiral (screw) struc- of majority carriers, has increased resistance, and
tures with a pitch that is incommensurate with is referred to as a gate layer. For downward band
METAL–SEMICONDUCTOR JUNCTION 833
of metal–semiconductor contact, fast optical de- relative dielectric constant ε = 9.7–10.0 for black
tectors, semiconductor solar cells, charge coupled mica, 6.0–8.0 for muscovite, 6.7 for lepidolite.
devices, etc. (see Semiconductor devices). The Magnetic susceptibility χ = (3–80)·10−6 cm3 /g
advantages of these devices are operating speed, for black mica, (3–5)·10−6 cm3 /g for muscovite.
technological simplicity, small size of active re- Cohesion is rather perfect within layers. Micas are
gion, etc. used in radio and electrical engineering as thermal
and electrical insulating materials.
METALS, HEAT TREATMENT
See Mechanical heat treatment of metals. MICHELL PROBLEM (J.H. Michell, 1900)
The problem involves describing the state of
METAMAGNETISM stress of a prismatic rod with a uniformly loaded
See Antiferromagnetism. lateral surface. With the acceptance of the Saint-
Venant principle (effects of a system of forces that
METASTABLE STATE
add vectorially to zero applied to a region of a
State of incomplete thermodynamic equilib-
solid are only felt within that region), the prob-
rium of a macroscopic system, corresponding to
lem reduces to the Saint-Venant problem involving
a local minimum of its free energy, with a long
forced e xure and rod twisting, and also to a certain
lifetime before transformation to the equilibrium
plane problem of elasticity theory.
state. This transition usually occurs via the for-
mation of nucleating centers of the new phase
MICRO. . . (fr. Gr. μικρoς , small)
(see Nucleation). These centers may be, e.g., frag-
Prefix for a physical unit, which allows obtain-
ments of polar phases in ferromagnets and ferro-
ing a multiple unit, equal to 10−6 of the initial
electrics. In quantum systems at low temperatures
unit. Symbol: μ. Example: 1 μs (microsecond) =
the metastable state may collapse by activation, or
10−6 s.
because of macroscopic quantum tunneling.
(γ0 is the true specific surface energy, B is a from the following expressions: R = 3ṽ/ S and
coefficient of order 1). Relaxation of energies of ñ =
S/(4πR 2 ). In the case of metals with a body-
randomly distributed dislocations located in the centered cubic or face-centered cubic lattice, the
region of dislocation cell boundaries, can com- area of a molar surface S amounts to thousands of
pletely compensate for the increase of surface square meters per mole, and the microcrystallite
energy when the crack is being opened, which radius ranges from 10 nm (alkali metals) to 1.5 nm
explains the strati ca tion-type failure. The cracks (metals with high melting temperatures Tmelting );
can occur in equilibrium with different disloca- the product RTmelting is thus roughly constant and
tion structures (e.g., with pile-up of dislocations). amounts to ∼3·10−4 cm·K.
Changing the external conditions can shift this The mean intercrystalline distance is a +
a,
equilibrium either toward the healing (see Heal- where a is the mean interatomic distance of the
ing of defects) or toward the growth of fractures lattice,
a =
ṽmelting /(1/2) S (
ṽmelting is the
(see Fatigue crack growth). At high temperatures volume increment due to melting). The ratio
a/a
the elastic diffusion microcrack-dislocation inter- is nearly the same for different metals and amounts
action is significant (see Diffusion), and it can to ≈0.28. The distance between microcrystallites
result in the healing of cracks with the formation increases with increasing temperature:
a/a =
of a limited dislocation wall. B exp(−Tmelting /T ), where B is a dimension-
less numerical factor. The average value of B for
MICROCRYSTALLITE PARAMETERS monatomic materials (including alkali metals, no-
The physical quantities: surface energy, molar ble gases in solid state, platinum, gold) and water
surface, radius, content per mole, spacing between is 0.78. Molecular solids may exhibit deviations
microcrystallites in amorphous bodies. of B from 0.78 toward both higher (1.3 for O2 and
The surface energy of microcrystallites is not PH3 ) and lower (0.58 for CS2 ) values.
equal to the surface energy of a crystal in vacuo
since microcrystallites are in contact with one MICROCRYSTALLITES
another, and the external bonds of their surface The smallest individual blocks or constituents
atoms are partially saturated by their coupling with of a polycrystal, with sizes of the order of several
the surfaces of neighboring microcrystallites. The interatomic distances, which retain the structure of
molar surface energy is given by the equation a crystal. In accordance with one of the microcrys-
σ̃ms = σ̃1 − σ̃2 , where σ̃1 = [(kv − ks )/kv ]U is tallite models of the structure of the amorphous
the total molar surface energy of microcrystallites state of a solid, glasses (see Vitreous state of mat-
(kv and ks are coordination numbers of atoms ter) are an aggregation of microcrystallites with
in the volume and at the lattice surface, U is random relative orientations. A modification of the
the molar lattice energy). The quantity given by microcrystallite model, known as the paracrys-
the equation σ̃2 = f N1/3 ṽ 2/3 σ (coefficient f de- tallite model, assumes that there are no distinct
A
pends on the surface layer structure, N A is the boundaries between individual microcrystallites;
Avogadro number, ṽ is the molar volume, σ is the short-range distortions (see Long-range and short-
surface energy) is the part of the energy required range order) are not localized in the neighborhood
by a certain region of a microcrystallite surface. of boundaries, but rather are smoothly spread
The surface energy of microcrystallites in metals over the entire bulk of microcrystallites (see Fig.).
exhibiting body-centered cubic and face-centered In the limiting case this model is equivalent to
cubic lattices is close to the value of the surface the random continuous network model of bonds,
energy of the corresponding melts. or random close-packing model of spheres with
The molar surface (the total surface occu- small uctuations of short-range order parameters.
pied by one mole of microcrystallites) is given The existence of microcrystallites in the major-
by the equation S = q̃/σms , where q̃ is the heat ity of amorphous solids has not been confirmed
of melting, σms is the surface energy density of by direct (diffraction) methods, since diffraction
microcrystallites. The microcrystallite mean ra- lines broaden as crystals become smaller in size.
dius R and content per mole ñ are determined Broadening of lines may possibly result, not from
836 MICROELECTRONICS
MICROSCOPIC THEORY OF FERMI LIQUID for zero-sound vibrations, and the role of the
Non-interacting particles obeying Fermi–Dirac kernel in the equation is played by the function
statistics form a Fermi gas, and when these par- ∞ (p, p ) (where z is the renormaliza-
zp zp Γαβ,γ δ p
ticles interact with each other in such a way that tion constant of the quasi-particle Green’s func-
their properties remain close to those of a Fermi tion), which allows one to identify it with the
gas they constitute what is called a Fermi liquid. Landau correlation function
The microscopic theory of a Fermi liquid (a
liquid whose particles obey Fermi–Dirac statis- fαβ,γ δ (p, p ) = zp zp Γαβ,γ
∞ (p, p ).
δ (5)
tics) provides an explanation for the appearance
of zero-sound vibrations in a neutral Fermi liq- As in the case of Γ ω and Γ k , the function f ∗ is
uid, and it also establishes a connection between connected by an integral relation to the physical
the Landau correlation function and the forward forward scattering amplitude:
scattering amplitude of quasi-particles at the angle gαβ,γ δ (p, p ) = zp zp Γαβ,γ
k
δ (p, p ). (6)
zero. These considerations are based on the inves-
tigation of the properties of the vertex function The connection of the Landau correlation func-
tion with the vertex function allows one to obtain
Γαβ,γ δ (p1 , p2 ; p3 , p4 ) ≡ Γαβ,γ δ (p1 , p2 ; k), a sum rule for the Landau parameters Al and Bl
(1) of a charged Fermi liquid, taking into account the
where k = p1 − p2 = p4 − p3 , k ≡ (k, ω) for long-range Coulomb interaction:
neighboring values of the pairs of variables (k → 0
Al Bl 1
for forward scattering). Γαβ,γ δ is obtainable by (2l + 1) + =− .
the summation of diagrams of the “ladder” ap- 1 + Al 1 + Bl 1 + A0
l
proximation which have a singularity at k → 0 (7)
that, in its turn, leads to the appearance of a singu- In the case of a neutral Fermi liquid the sum rule
larity in $αβ,γ δ : namely that the result of taking is obtained by equating the left side of Eq. (7) to
the limit depends on the relation between ω and zero.
|k|. By this the function The isotropic model of a Fermi liquid is strictly
ω
Γαβ,γ applicable only for the alkali metals Na and K. The
δ (p1 , p2 ) application of the Fermi liquid theory for describ-
= lim Γαβ,γ δ (p1 , p2 ; k) (2) ing the properties of other metals is complicated
k→0, |k|/ω→0
by the sharp anisotropy of the Fermi surface. Ex-
corresponds to the nonphysical limiting case of the perimental values of the parameters A2,3 , B0,1,2
scattering of quasi-particles on the Fermi surface for Na and K are small (A2 ≈ (3–5)·10−2 , A3 ≈ 0;
with a small transfer of energy and a transfer of B0 ≈ −0.2 to 0.3; B1 ≈ 5·10−2 , B2 ≈ 0). The
momentum strictly equal to zero, and the function only appreciably large Landau parameter is
k
Γαβ,γ A0 ∼ 1. However, due to the complications in-
δ (p1 , p2 )
volved in separating the electronic terms, it turns
= lim Γαβ,γ δ (p1 , p2 ; k) (3) out that A0 , A1 cannot be determined experimen-
k→0, ω/|k|→0
tally. Experimental Fermi liquid effects have been
corresponds to real physical processes of the scat- observed in 3 He, where they start to appear at
tering of quasi-particles on the Fermi surface with T < 0.1 K. The values of the Landau parameters
a change of momentum, but no change of en- A0.1 , B0 depend on the pressure P , thus: A1 =
ergy. The functions Γ ω and Γ k are connected 2.1, A0 = 10.8, B0 = −0.67 at P = 3·104 Pa;
by an integral relation. The poles of the function and A1 = 4.8, A0 = 75.6, B0 = −0.72 at P =
Γαβ,γ δ (p1 , p2 ; k) characterize acoustic Bose ex- 2.8·106 Pa.
citations of a Fermi liquid. The equation describ-
ing these excitations coincides with the equation MICROSCOPY, ACOUSTIC
See Acoustic microscopy, Thermoacoustic mi-
∂n
(kv − ω) δ n̂ − kv 0 η(p, p ) δn δτ = 0 (4) croscopy.
∂ε
MICROSPALLING RESISTANCE 841
the formation of the electron–hole plasma. Such class of microwave semiconducting devices (see
a drawback is absent in generators built on Gunn Acousto-electronics). See also Photocontrolled mi-
diodes that typically have lower output power but crowave semiconductor devices.
high-quality spectral characteristics of their output
signal. However, they have a relatively low fre- MIGRATION
quency threshold of negative conductivity (up to Outward movement of objects away from loca-
80–90 GHz in GaAs and 140–160 GHz in InP). To tions in space; the physics of the condensed state
raise this upper frequency threshold, since the fre- of matter usually deals with migrations of small
quency is being limited by the inertia of electron particles of microscopic (including atomic) sizes,
transfer, one may operate Gunn diodes in a bihar- and also of quasi-particles. Migration of interstitial
monic mode, so the microwave power is derived atoms, vacancies, of autolocalized electron states
from the second or even from the third harmonic (see Self-localization) of small radius, etc. repre-
of the fundamental frequency. In the mid-1980s sents a sequence of randomly oriented jumps with
GaAs Gunn diodes were putting out up to 80 mW length of the order of interatomic distances; the
at 94–96 GHz (power obtained from second har- activation energy of a jump is determined by the
monic), while the third harmonic yielded 3–5 mW height of a potential barrier. Migration of conduc-
at 140–150 GHz. Indium phosphide Gunn diodes tion electrons and holes, Wannier–Mott excitons,
yield up to 150 mW at 94–96 GHz (fundamen- etc. may be described as a motion of a particle
tal frequency) and up to 10 mW at 180–200 GHz with a quasi-momentum that changes when the
(second harmonic). particle interacts with phonons, defects in crys-
Field-effect transistors made of GaAs are tals, current carriers, etc. At low temperatures, the
mainly used for low-noise amplifiers in the mil- mean free path of a particle between two succes-
limeter range. For example, a noise level of 4.5 dB sive collisions generally considerably exceeds an
is achieved at the frequency of about 10 GHz us- interatomic distance. The mechanism of migration
ing these devices. The most powerful amplifiers of particles with small masses may also involve
are built on silicon bipolar transistors. Modern ten- tunneling of particles through a potential barrier
dencies in building transistor amplifiers consist in (see Tunneling phenomena in solids). Mobile par-
merging several amplifying cascades into a mono- ticles in a material experience random collisions
lith integrated circuit. with atoms and electrons of the medium as well
The principal materials used to produce micro- as with one another, and, with no preferred di-
wave semiconducting devices are silicon, GaAs, rection present, execute random walks (stochastic
and InP. Devices operating on the basis of GaAs, motion). One should distinguish macro- and mi-
InP and solid solutions of these materials feature croscopic aspects of migration. If no reactions
a high efficiency and low noise level. Electron involving mobile identical particles take place, the
inertia, however, affects the functioning of such initial macroscopic uniform spatial distribution of
devices in the millimeter range. The time required the latter is retained. But the microscopic distri-
for electrons to reach their required distribution bution changes continually; and if the medium
in microwave semiconducting devices is compa- has some centers capable of undergoing reactions
rable to the microwave oscillation period, and this with particles (e.g., trapping at traps for mobile
limits the potentially achievable frequencies and particles, recombination), the migration leads to
output power of the devices. Generally, microwave the actualization of such reactions as the particles
semiconducting devices are widely used to build approach the centers. If the macroscopic distri-
various microwave instruments. bution of centers is nonuniform, this will cause
In parallel with perfecting semiconducting de- nonuniformity in the distribution of mobile parti-
vices of the lumped type, solid state electronics is cles. If the initial macroscopic spatial distribution
developing toward microwave devices functioning of particles is nonuniform, then migration leads
in the distributed operational mode. Dangling ul- to equalization of the concentration. The change
trasonic waves that interact with the electron flux of concentration of particles n(r, t) per unit vol-
resulted in the design and production of a new ume about point r at instant of time t that results
844 MILLER EFFECT
from diffusion via a random walk is described by second harmonic frequency; growing in intensity
the equation ∂n(r, t)/∂t = ∇ 2 Dn (where ∇ 2 is by a factor of hundreds. The Miller effect finds ap-
the Laplacian operator, D is the diffusion coef - plication in the nonlinear optics of solids.
cient). Besides a diffusion component, migration The term Miller effect also refers to the effec-
includes a drift component that arises when forces tive capacitance between terminals in a vacuum
act on the particle, e.g., those due to macroscopic tube.
electric, magnetic, elastic fields, or to the field of a
defect of atomic scale. The force −∇U (r) (where MILLER INDICES
∇ is the gradient operator, U (r) is the particle po- See Crystallographic indices.
tential energy) causes an increase of the hopping
rate along the field, and its decrease in the opposite MILLER RULE (R.C. Miller, 1964)
direction, when the migration occurs via a “hop- A rule in nonlinear optics, which allows one
ping mechanism”. If migration proceeds without to evaluate approximately the components of the
activation, this force causes oppositely directed nonlinear quadratic susceptibility tensor dij2ωk that
accelerations along and against the field. Drift re- are responsible for the efficient second harmonic
sults in a nonuniformity of particle distributions, generation in anisotropic piezoelectric crystals
or the establishment of a stationary current, de- and other media lacking a center of symmetry. The
pending on the boundary conditions. In the general said components are determined from the com-
case, ∂n/∂t is the sum of diffusion and drift terms, ponents of the linear susceptibility χij through a
2ω , the Miller tensor,
third-rank tensor δij
as well as terms that describe the generation and k
annihilation of particles.
Migration of a particle involves transport of dij2ωk = χii2ω χjj
ω χ ω δ 2ω .
kk ij k
mass; charge, spin and energy may also be si- This relationship may be extended to the case of
multaneously transferred in this process. Energy sum-frequency generation ω3 = ω1 + ω2 :
transport is also possible without mass transport, ω ω ω ω ω
if the migrating particle is an excitation wave in dij 3k = χil 3 χj m1 χkn2 δlmn
3
.
a crystal, e.g., exciton, phonon, plasmon, etc. In Since the components of the tensor δij k differ
particular, excitation of a certain region of a non- from one material to another to a considerably
metallic crystal under light absorption may cause lesser extent than the components of dij k (in
luminescence from the dark region, induced by
practice one can set δij k ≈ const), an order-of-
resonance transfer of an excited state between magnitude estimate of δij k for any material may
identical atoms. Migration of Wannier–Mott ex- be found from the value of its refractive index n at
citons, their subsequent trapping by defects of a given frequency (χ = n2 − 1).
various types, and radiative recombination may
cause luminescence at different frequencies. En- MILLI. . . (fr. Lat. mille, thousand)
ergy transfer in condensed media may also occur Prefix for a physical unit, which gives the frac-
directly by photons. tional part equal to 10−3 of the initial unit. Sym-
bol: m. Example: 1 mA (milliampere) = 10−3 A.
MILLER EFFECT (R.C. Miller, 1965)
Abrupt increase of the intensity of the second MINERALS
harmonic signal during the process of a nonlinear Naturally occurring chemical compounds
interaction between a laser beam and a ferro- (more rarely, native elements), which are prod-
electric crystal. It is observed in certain many- ucts of various physicochemical processes taking
domain crystals with domain sizes comparable to place in the Earth’s crust. A typical feature of
the length of the coherent interaction. If domains minerals is their comparative homogeneity that re-
are inclined at an angle of 90◦ to one another, as, sults from a definite chemical composition and
e.g., in KDP crystals (potassium dihydrogen phos- regular arrangement of atoms in a structure. The
phate, KH2 PO4 ), then the beam, in passing from physical properties and characteristic geometri-
one domain to another, changes its phase at the cal shapes of minerals are related to the specifics
MIXED STATE 845
of their composition and structure. Most geolo- isovalent and heterovalent, are distinguished. At
gists consider only solid compounds as minerals, isovalent mixing one atom is replaced by an-
but Vernadsky included some liquids and gases other atom, or one ion by another, and true
as minerals. There is disagreement on including solid solutions are formed that are described by
certain compounds of organic origin (e.g., fos- thermodynamic equations such as that of Gibbs–
sil resins), silicate glasses, and other compounds Duhen and others (e.g., K(Rb)Cl, KCl(Br), and
among minerals. Underlying the classification are so on). At heterovalent mixing it is possible
differences in types of compounds, crystal struc- to have two cases: without change of common
tures, and features of the spatial arrangement atom numbers in formula units, but with different
of these structures (e.g., insular, chain or lami- charge states of atoms (e.g., ZnS–GaAs, BaSO4 –
nated structures). About 3000 types of minerals KMnO4 , 2SnO2 –FeNbO4 ), or with a change in
are known: silicates (25%), oxides and hydroxides the number of atoms. This case leads to vacancy
(12%), sulfides and analogs (13%), phosphates, formation (in CdCl2 added to NaCl there is one
arsenates, vanadates (18%), halides, carbonates, vacancy per Cd2+ ion); insertion of ions at in-
sulfates and other oxygen-containing salts, native terstices (see Interstitial atom) (in CdF2 –YF3 one
elements. Silicates include 6 structure types: Neso F− ion is changed by one Y3+ ion), and filling
(individual SiO4 tetrahedra), Soro (double tetra- of already existing vacancies (in YO1.5 –PrO1.5+x
hedra Si2 O7 ), Cyclo (rings (SiO3 )N , N = 2, 3, with x oxygen atoms (x < 0.5) filling vacancies
4, 6), Ino (single (SiO3 )N or double (Si4 O11 )N in the oxygen sublattice of Y2 O3 ). Such mixed
chain), Phyllo (layered (Si2 O5 )N ), and Tekto crystals have certain features: point defects, ir-
(three-dimensional (SiO2 )N network). regular charge distribution, quasi-chemical com-
Diagnostic attributes of minerals are: the shape plexes, ordering phenomena. Mixed crystals are
of the crystal and precipitation, color, density, op- increasingly used: ruby (Al2 O3 with 0.05% Cr),
tical, mechanical, electrical, magnetic etc. proper- LiY(Nd)F4 and others are active elements in
ties. Products of laboratory or industrial processes solid-state lasers; YO1.5 –PrO1.5+x is used for
with the composition and structure of natural min- magnetohydrodynamic generators (due to electri-
erals, are classed as artificial minerals. The latter cal conductivity and thermal resistance), Na(Ti)I
may differ somewhat in their properties from nat- monocrystals are scintillators. Ferrites, and some
ural minerals. ferroelectrics, are ordinarily mixed crystals (e.g.,
Fe3 O4 is Fe2 O3 plus FeO).
MIRROR GLIDE PLANE
See Glide-re ection plane. MIXED STATE of superconductors, vortex state
A condition involving a type II superconduc-
MISHIMA ALLOYS (after T. Mishima) tor with a heterogeneous magnetic flux distribu-
High-coercivity alloys of the Fe–Ni–Al system tion and an average internal field B = nΦ0 when
(see Alni), which are used to manufacture perma- present in an applied field B0 with a value be-
nent magnets. tween the lower critical eld Bc1 and the upper
critical e ld Bc2 . This mixed state is a result of
MIXED CRYSTALS the penetration of Abrikosov vortices into the su-
Phases of variable or nonstoichiometric com- perconductor. Each vortex has one ux quantum
position (berthollides) formed by two or more Φ0 = h/2e, and they are distributed with the den-
materials. Mixed crystals are widespread, and sity n(B), with the value 0 below the field B0 =
are found in many mineral and rock compo- Bc1 and the maximum value at the field B0 = Bc2
sitions. Mixed crystals may be separated from (nmax ∝ 1/ξ 2 , where ξ is the coherence length).
gas, liquid and solid phases. Mixing can vary The dependence of the superconductor magneti-
over a wide range (from fractions to 100%). In zation M = (B − μ0 H )/μ0 on the applied field
the simplest cases it can be predicted from cal- B0 has the form depicted in Fig. 1. The values of
culations based on thermodynamic potentials of the critical fields Bc1 and√Bc2 are given by the√ex-
mixed crystal formation. Two types of mixing, pressions: Bc1 = [Bc /(κ 2)] ln κ, Bc2 = Bc κ 2,
846 MOBILITY EDGE
also Anderson localization, Amorphous semicon- sion of simultaneous electron drift in all valleys,
ductors. which takes place due to the symmetrical location
of electrons in reciprocal space (k -space), leads
MOBILITY OF CURRENT CARRIERS to isotropy of the mobility. A change of the rel-
Average rate of current carrier drift in an elec- ative contributions to the current from different
tric field of unit intensity; defined by the equa- valleys, which takes place, e.g., at the Sasaki ef-
tion μ = |v|/|E|, where v is drift rate, E is fect or at strain, results in anisotropy of mobility of
electric field intensity. The value of the mobil- current carriers. A low-symmetry crystal exhibits
ity μ is determined by the balance between the anisotropy of mobility of current carriers.
acceleration of current carriers due to the elec- The temperature dependence of μ is deter-
tric field and the scattering of the carriers. The mined by the mechanism of current carrier scat-
mobility μ of current carriers is related to the tering. The mobility μH , which is obtained from
diffusion coef cient D by the Einstein relation measurements of electromotive force or Hall cur-
μ = eD/(k B T ). In the absence of degeneracy of rent (Hall mobility), differs from μ: μH = rμ,
the carriers it follows from the solution of the where the Hall factor r is a numerical parameter
Boltzmann equation that the electron (hole) mo- which depends on the mechanism of current car-
bility is given by μ = (e/mn )τp (ε), where e is rier scattering, structure of energy bands and the
charge, mn is effective mass, ε is energy, τp (ε) magnitude of the magnetic field (see Galvanomag-
is the relaxation time of the quasi-momentum
netic effects). The difference between μ and μH is
of scattered particles, the operator . . . indicates
related to the fact that μ describes the response of
averaging of current carriers over energy states
current carriers to the force −eE, which has an
with the help of the distribution function. If elec-
equal action on all current carriers, whereas μH
trons (holes) are strongly degenerate, then μ =
describes the response to the Lorentz force, which
[e/mn ]τp (εF ), where εF is Fermi energy, because
depends on the particle velocity, and therefore ex-
by virtue of the Pauli exclusion principle only
erts a different action on every current carrier.
electrons with ε ≈ εF are scattered. Anisotropy
For Hall mobility determinations the quantity μ
of energy spectra of electrons and phonons leads
is sometimes called conduction mobility (mobil-
to anisotropy of the effective mass and relaxation
time, which are described by the reciprocal effec- ity in terms of conduction) or drift mobility. The
tive mass tensor and the relaxation time tensor, latter term is also used in reference to mobility,
respectively. If the crystal exhibits high micro- which is determined from the time of drift of
scopic crystal symmetry then, in spite of this current carriers between two separated points of
anisotropy, the mobility of current carriers is a the sample. Ambipolar drift (see Ambipolar diffu-
scalar quantity, which depends on components of sion) is described in terms of ambipolar mobility;
the above tensors. The combination of recipro- mobility in terms of magnetoresistance is deter-
cal effective mass tensor components, which is mined from magnetoresistance. Surface mobility of
included in the equation for the mobility instead current carriers and field-effect mobility are intro-
of mn , is sometimes called conduction effective duced at consideration of the motion of current
mass, Ohmic effective mass, or optical effective carriers at a semiconductor surface. Surface mo-
mass. The last name owes its origin to the fact bility of current carriers is the mean mobility of
that the above combination of tensor components current carriers, which are situated in channels
is included in the high-frequency mobility, which at different distances from the surface; it takes
is introduced at the description of the motion into account the surface scattering of current car-
of current carriers in an ac field of frequency riers. The field-effect mobility is the derivative of
ω
1/τp . Often this is the region of visible the current which flows through a sample of unit
radiation. length and width, with respect to the charge of the
In many-valley semiconductors with cubic sym- field-effect electrode. The value of this quantity
metry (germanium, silicon) the electrons of every depends on the trapping of current carriers by elec-
valley have an anisotropic mobility. The inclu- tron surface states.
848 MOBILITY OF DOMAIN WALL
in the cubic crystals MnSi and FeGe which lack a a periodic change of the concentration of alloying
center of inversion (space group P 21 3, T 4 ). This elements (see Alloying) along one of the alloy’s cu-
structure owes its origin to the invariant M∇ × M bic directions. In diffuse phase transitions (aging,
(V. Baryakhtar, E.S. Stefanovsky, 1968) that de- alloy ordering, eutectoid decomposition), a modu-
scribes the nonuniform Dzialoshinskii–Moriya in- lated structure is an ensemble of domains where
teraction (see Dzialoshinskii interaction). the precipitates of the second phase form a macro-
lattice. One-, two-, and three-dimensional modu-
MODULATED STRUCTURES lated structures are known. One-dimensional ones
Regularly and quasi-periodically arranged re- have the highest thermal stability in aging alloys
gions that are, e.g., either coherent precipitates as they correspond to a minimum of the elastic en-
in aging (see Alloy aging), ordered, eutectically ergy of the system. The formation of modulated
decomposing alloys (see Eutectic alloys, Alloy de- systems is accompanied by a significant change in
composition), individual (or group) stacking faults their physical and mechanical properties.
in one-dimensionally disordered alloys; or else
antiphase domains in antiferromagnets and ferro- MODULATION SPECTROSCOPY
electrics. Complicated atomic or molecular com- Investigation of optical spectra under condi-
plexes in minerals, in chemical organic or in- tions of modulation of material (sample) parame-
organic structures may also be considered as ters (internal modulation), or of the light beam
such regions. Modulated structures were first (external modulation). Used for producing mod-
discovered by the X-radiographic investigation ulation can be electric and magnetic fields, illumi-
(see X-radiography) of an aged cunife alloy: the nation intensity, temperature, uniform or uniaxial
X-ray spectrum of this alloy has additional reflec- pressure. One can modulate such light beam pa-
tions (satellite lines) observed near the main Bragg rameters as wavelength λ, angle of incidence,
reflections. The effect is explained by the model of polarization, etc. Since it is the change of light
850 MODULATOR, ELECTROOPTICAL
σN = Sij vi vj ,
MODULATOR, ELECTROOPTICAL
σS2 = Sij Sj k vi vk − (Sij vi vj )2 .
See Electrooptical modulators.
Here, summation is meant to be taken over re-
peated indices.
MODULUS OF ELONGATION
See Young’s modulus. MOLARITY
Measure of concentration, number of moles of
a solute in 1 liter (1 dm3 ) of solution. Alternate
MODULUS OF RIGIDITY name: volumetric molar concentration. Measured
See Shear modulus. in mole/liter (often denoted by M).
MOLECULAR ENERGY DIAGRAM 851
levels reflect their group theory classification, as in the neighborhood of a phase transition temper-
a rule involving the irreducible representations of ature, one-dimensional systems, two-dimensional
the symmetry point group of the molecule, and XY -models and Heisenberg model, dilute mag-
thus they provide the (orbital) degeneracy of the netic materials with short-range exchange in the
levels. For multielectron molecular terms, the mul- neighborhood of flux thresholds, etc.), the mole-
tiplicity of the spin degeneracy of the state is also cular field approximation yields qualitatively in-
indicated. In analogy with electronic levels, sim- correct results. In addition this approximation fails
ilar diagrams are used for rotational, vibrational,
to describe collective spin excitations (magnons),
and more complicated states, e.g., for electronic-
and some low temperature features in Heisenberg
vibrational states (vibronic states), etc.
magnets.
MOLECULAR FIELD (B.L. Rozing, P. Weiss,
1907)
MOLECULAR INTEGRALS
One of the most widely used approximations
for explaining magnetic properties in a solid is The same as Quantum chemistry integrals.
to invoke a molecular field (self-consistent eld,
average field). The molecular field approxima-
tion assumes that the interaction of a certain spin MOLECULAR MASS DISTRIBUTION OF
Si with N neighboring spins is approximated by POLYMERS
introducting a molecular field B MF that is pro- Distribution of macromolecules of a given
portional to the magnetization in a ferromagnet: polymer sample expressed in molecular mass units.
gμB B MF = λM. Here, λ is the molecular field This molecular mass discussion pertains to poly-
constant, g is the Landé g -factor, μB is the Bohr mers only, and results from the random nature
magneton, and M is the magnetization. In the
of reactions of polymer generation and polymer
presence of an external magnetic field B eff =
degradation. The concept of molecular mass dis-
μ0 H , M satisfies the self-consistent equation
tribution loses its meaning for three-dimensionally
M = 2NgμB SBS [(λM + gμB B)/(k B T )], where
B S is the Brillouin function, and N is the number linked polymers (see Polymer cross-linking), when
of nearest neighbors. At B = 0, this equation de- the whole polymer body behaves as if it were a
scribes the behavior of spontaneous magnetization single “molecule”. Two types of molecular mass
that disappears in the paramagnetic state above the distribution are known: proper distribution ρn (M)
Curie point. In a microscopic approach, the local of macromolecules in the number of chains with a
molecular field is expressed in terms of parameters given molecular mass M (numeric molecular mass
Jij of the exchange Hamiltonian
and mean values distribution) and distribution ρω (M) in masses
of spins: gμB BMF (i) = j Jij Sj . This allows of chains with a given molecular mass (mass
one to perform self-consistent calculations of spin molecular mass distribution). These are related
behavior both in ferromagnets, and in more com- through the expression ρω (M) = ρn (M)MM n ,
plicated atomic magnetic structures. This approxi- where M n is the so-called number-average mole-
mation is an asymptotically exact solution in mod- cular mass given by M n = 0∞ ρn (M)M dM.
els with an infinite exchange interaction radius R,
There is also a mass-average molecular mass
and hence, it serves as a zero-order approximation
M ω = 0∞ ρω (M)M dM to characterize molecu-
in the perturbation theory of the thermodynamics
of spin systems, which is based on an expansion in lar mass distribution. The inequality M ω > M n is
terms of 1/R or 1/N . The molecular field is also always true, and the deviation of the ratio M ω /M n
used to describe other phase transitions: structural from unity serves as a measure of the molecular
and ferroelectric types, those from the normal state mass distribution width (typically, 1 < M ω /M n
to the superconducting state, and from the normal < 2). The molecular mass distribution of a poly-
state to the superfluid state. When the role of fluc- mer is most often determined by sedimentation,
tuations is significant (regions of critical behavior fractionation, and chromatography experiments.
MOLYBDENUM 853
is ≈464 K; coefficient of linear thermal expan- (3) Joint alloying with an element (titanium, zir-
sion is ≈6.0·10−6 K−1 for unannealed state conium, hafnium, vanadium, niobium, tanta-
(at 298–2273 K). Thermal conductivity is lum) and carbon (nitrogen) to increase high-
124–67 W·m−1 K−1 (at 1200–1900 K), 145 temperature strength through the mechanism
W·m−1 ·K−1 at 290 K and 183 W·m−1 ·K−1 of precipitation hardening (see Precipitation-
at 90 K. Adiabatic elastic moduli: c11 = 440.8, hardened materials).
c22 = 172.4, c44 = 106.8 GPa at 298 K. Young’s (4) Heavy alloying with rhenium (up to 50%) or
modulus is ≈340 GPa, shear modulus is with tungsten.
119.7 GPa, Poisson ratio is ≈0.33. Breaking
strength of wire 6.12 mm in diameter is 525– The production of most molybdenum alloys in-
2500 MPa and depends on the degree of defor- volves the simultaneous use of several alloying
mation; that of arc melted ingot is 571 MPa and methods. Molybdenum alloys are melted in arc
that for monocrystalline molybdenum is 310 MPa. furnaces and electron beam furnaces, or produced
Hardness of molybdenum polycrystal is 1592 MPa through methods of powder metallurgy. Molybde-
at 293 K, 3730 MPa at 457.5 K and 126 MPa at num alloys find use in electronics, rocket engineer-
2130 K. Relative elongation 10–20%. Coefficient ing, nuclear power industry, and in other areas.
of self-diffusion is 3.1·10−15 m2 /s at 2120 K.
Electrical resistivity is 51.4 n·m at 293 K. MOMENT OF INERTIA
Temperature coefficient of electrical resistance is Sum of products of masses mi of points in
479·10−5 K−1 at 293–2893 K. Critical temper- a solid that rotates about a fixed axis and the
ature of transition into superconducting state is squares of their distances Rl from the axis: I =
2
0.94 K. Critical magnetic field 9.65 mT. Work i mi Ri . Knowledge of the moment of inertia al-
function is ≈4.3 eV. Reflection factor for optic lows expressing the angular momentum L of the
waves with λ = 5 μm is 97.2%. Thermal neu- body with respect to the fixed axis in terms of
tron capture cross-section is 2.7 barn per atom. its angular velocity: L = I ω; the rotational ki-
Molybdenum is a paramagnet, with molar mag- netic energy is expressible in terms of the moment
netic susceptibility +5.4·10−6 CGS (at 291 K). of inertia and the square of the angular veloc-
Nuclear magnetic moment is 0.9 nuclear magne- ity: K = (1/2)I ω2 . In the general case of an
tons. arbitrarily directed
axis through the origin of coor-
Molybdenum is used in the production of spe- dinates, I = 3α,β=1 Iαβ cos θα cos θβ , where θα
cial alloys and alloyed steels. Molybdenum metal is the angle of this axis to
the αth coordinate axis.
is a construction material for vacuum electronic The quantities Iα,β = i mi (Ri2 δαβ − xiα xiβ )
industry and nuclear reactors. Liabilities are its form moments of inertia tensors
and allow ex-
high oxidizability and relatively high temperature pressing L and K as Lα = 3β=1 Iαβ ωβ ; K =
of cold brittleness. Molybdenum is one of the most (1/2) 3α,β=1 Iαβ ωα ωβ . See also Principal mo-
promising metals for creation of high-temperature ments of inertia.
alloys (see Molybdenum alloys).
MOMENTS OF SPECTRAL LINES
MOLYBDENUM ALLOYS
Characteristics of shapes of spectral (or some
Alloys based on molybdenum. Four main meth-
other) curves. For a resonance line (curve) that is
ods are singled out in molybdenum alloying.
described by a normalized (to unit area) line shape
(1) Introduction of highly active elements, ex- function f (ω), the nth moment Mn with respect to
hibiting large affinity to interstitial impurities ω0 is defined by
(rare earth metals, etc.).
(2) Alloying with a small quantity (about 1%) Mn = (ω − ω0 )n f (ω) dω.
of transition metal to achieve competitive
segregation with interstitial elements at crys- If f (ω) is symmetrical with respect to ω0 then all
tal structure defects and to increase high- odd moments vanish. If all Mn are known, then
temperature strength. f (ω) may be recovered. However, knowledge of
MONOCRYSTAL 855
only the first few moments provides important in- (in magnetic resonance or optical light absorption,
formation on the shape of a resonance curve and, respectively). This function depends on the system
in particular, on the rate of its decay at shoulders Hamiltonian term V that is responsible for the line
away from ω0 . To illustrate, we consider the two broadening. Expanding the autocorrelation func-
most typical shapes of resonance curves, the sym- tion in a series in powers of V , one can obtain
metrical Gaussian shape and Lorentzian shape. closed form equations for moments of lines (e.g.,
The Gaussian shape is described by the normal- the second moment equation is of the form M2 =
ized function −Tr{[V , Sx ]2 }/TrSx2 ). This procedure, referred to
as van Vleck’s theory of moments, was applied
1 (ω − ω0 )2
f (ω) = √ exp − , to magnetic resonance in calculating line widths
Δ 2π 2Δ2 caused by the magnetic dipole–dipole interaction.
for which M2 = Δ2 , M2n = 1·3·5 · · · (2n−1)Δ2n , J.H. Van Vleck calculated M2 and M4 for a reg-
ular lattice of identical spins with dipole–dipole
and M4 /M22 = 3. The half-width δ of the line at
broadening and showed the resonance band shape
half-height is δ = Δ(2 ln 2)1/2 ≈ 1.18Δ. Thus, the to be close to Gaussian. In the case of diluted para-
value of the second moment for a Gaussian curve magnets, the ratio M4 /M22 increases significantly,
provides a satisfactory approximation for δ. The while the decrease of concentration of particles
other common line shape is Lorentzian described makes the shape of the line closer to Lorentzian.
by the normalized function The exchange interaction has no effect on M2 , but
δ 1 increases M4 . The calculation of higher order mo-
f (ω) = . ments is a cumbersome procedure; and describing
π δ + (ω − ω0 )2
2
a the shape of a line in terms of its moments is
Since all even moments for a Lorentzian line, rarely done in practice.
starting with the second one, are infinite, a more
physically adequate approximation is that of a MOMENTUM SPACE
“cut-off” Lorentzian curve with f (ω) = 0 at fre- Space of points that correspond to momen-
quencies |ω − ω0 | = α
δ. Then, within the tum values of masses or particles of a system. By
accuracy to terms of order δ/α, we may write momentum space in solid state physics one under-
2αδ stands quasi-momentum space (i.e. k -space).
M2 = ,
π MONOCLINIC SYSTEM (fr. Gr. μoνoς , one,
M4 πα single, and κλινω , bend)
=
1,
M22 6δ Crystallographic system that is characterized
2 1/2 by the presence of two right angles between co-
δ π M2 ordinate axes (α = γ = 90◦ = β) with unequal
√M = √ M 1,
lengths (a = b = c). The angle β that is not a
2 2 3 4
right angle is called the monoclinicity angle, and
i.e. δ of a Lorentz curve is much smaller than
√ the axis b is aligned along the only twofold axis,
M2 , which goes along with the slow decay of or is perpendicular to the plane m. The so-called
the shoulders. On the other hand, the assumption second setting is sometimes used: the c axis is
of a Gaussian curve may seem reasonable if the along 2 or 2, c < b < a, γ = 90◦ . This sys-
ratio M4 /M22 is close to 3. tem includes 2 Bravais lattices (primitive and body
The advantage of the method of moments is centered), 3 point groups, and 13 space groups.
that moments of lines may be calculated on the
basis of general principles without finding the MONOCRYSTAL, single crystal
eigenvalues of a system Hamiltonian. In accor- Homogeneous crystal with a unified crystal lat-
dance with linear response theory, a function of tice and anisotropic properties. The outward shape
shape f (ω) is expressible in terms of its auto- of a monocrystal results from its atomic struc-
correlation function (see Correlation function) of ture and crystallization conditions: under equilib-
magnetic or electric dipole moments of the system rium conditions, a monocrystal acquires a clearly
856 MONOCRYSTAL GROWTH
pronounced natural crystal faceting. Examples of surface a crystal of large dimensions might grow
well-cut natural crystals are quartz, halite, Iceland with all its atoms at regularly ordered lattice posi-
spar, diamond, ruby. Some monocrystals may lack tions. Such a single crystal is called a monocrystal.
a regular cut (e.g., artificially grown rounded ruby Monocrystals are usually obtained with relatively
crystals, silicon monocrystals). small deviations from equilibrium in the source
Many monocrystals exhibit special physical phase, since otherwise the addition of particles to
properties: diamond and boron carbide are very the growing crystal surface occurs in a disordered
hard, sapphire and fluorite are transparent over manner, with the resulting formation of a polycrys-
a wide wavelength range; quartz is piezoelectric tal. To obtain a monocrystal, spatially uniform in
(see Piezoelectricity); the strength of thread-like its physical properties and structure, the crystal-
crystals of corundum is unmatched. Monocrys- lization conditions must be stable.
tals are capable of changing their properties under Crystallization conditions, nature of the mater-
exposure to external actions (light, mechanical ial, and interface structure determine the different
stresses, electric and magnetic fields, radiation, mechanisms of monocrystal growth. Depending
temperature, pressure); hence, their use as trans- on the concentration of surface defects (steps,
ducers of various kinds in radio- and quantum etc.), the interface may be atomically smooth (see
electronics, acoustics, etc. As a result, there is a Atomically clean surface) or rough (see Surface
need for monocrystals of various sizes and shapes: roughness). The last case is realized when the
from microcrystals, films and thread-like crystals distance between defects approaches unit cell di-
several milligrams in mass to large specimens that mensions. In the case of an atomically-smooth
weigh tens of kilograms. Initially, use was made interface, the formation of a two-dimensional nu-
of natural monocrystals, but reserves of these in cleus on its surface is necessary for monocrys-
tal growth, and the probability of its appearance
nature are limited. The need for artificial ones
(as also for a three-dimensional nucleus) is as-
grew; and many synthesized monocrystals (see
sociated with the energy outlay that depends on
Monocrystal growth) have no natural analog (see
overcooling (oversaturation) of the source phase
Synthetic monocrystals).
at the interface. Since this probability is low for
According to their degree of perfection, mono-
small overcooling, the mechanism of bidimen-
crystals are classified into non-mosaic (nearly per-
sional nucleation may be noticeably apparent at
fect) and mosaic crystals. The most perfect con-
a comparably high deviation from equilibrium at
tain only point defects (see Defects in crystals); the the interface. The relation of the rate of monocrys-
less perfect mosaic ones exhibit dislocations and tal growth v to overcooling
T is determined by
stacking faults. Mosaic crystals with small block v = A exp(−B/
T ), where A and B are con-
dimensions (10−5 cm and less) scatter X-rays stants, weakly depending on the temperature T .
in accordance with the theory of elastic scatter- The presence of a screw dislocation at the in-
ing of radiation. Increase of disorientation angles terface leads to the appearance of an echelon of
between blocks up to a magnitude of tens of de- steps, at which particles from the source phase
grees results in the polycrystalline state (see Poly- may build. In this case, the dislocation mecha-
crystal). Many monocrystal properties (electrical, nism of monocrystal growth is realized, which can
optical, strength, etc.) depend on the degree of per- proceed at small overcooling (oversaturation). The
fection, and the impurity content. The most perfect dependence of v on T under this mechanism is
ones are used in semiconductor engineering, quan- quadratic: v ∝
T 2 . If the crystal surface is rough,
tum electronics, computer machinery, etc. then practically every point of it is an active site
for the addition of particles. The mechanism of
MONOCRYSTAL GROWTH, single crystal crystal growth with a rough interface of phases is
growing called normal. The relation of the growth speed to
Types of crystallization processes leading to the overcooling (oversaturation) for the normal mech-
formation of monocrystals. If within the volume of anism is linear.
the source phase there is only one nucleus capable The presence of impurities and inclusions in
of growing, then by the addition of particles to its the source phase affects the monocrystal growth.
MONOPOLE, MAGNETIC 857
where N(0) is the electron density of states at not observed in liquids and gases). The underlying
the Fermi level, VC is the screened Coulomb cause of the effect is that the nucleus recoil en-
interaction matrix element averaged relative to ergy ratio ER = E/(2Mc2 ) (here E is the energy
the momentum transfers, ωD is the Debye fre- of the γ -quantum, M is the mass of the nucleus)
quency, Em is the maximum cutoff interaction is small relative to phonon energies, so the crystal
energy which equals in order of magnitude the lattice as a quantum system is not excited, and the
Fermi energy EF or the plasma oscillation en- recoil is not absorbed by an individual nucleus but
ergy ωp of electrons; thus, ln[Em /(ωp )]
1 rather by the crystal as a whole. The probability
(see Bogolyubov–Tolmachyov logarithm). of the Mössbauer effect is f = exp[−4πx 2 /λ2 ],
where λ is the γ -quantum wavelength, and x 2
MORIN TRANSITION (F.J. Morin, 1950)
is the mean square of amplitude of nucleus oscil-
Spontaneous (with no external magnetic eld)
lations in the γ -quantum direction. The absence
magnetic phase transition from a weak ferromag-
of excited phonons causes the emission (absorp-
netic state to an antiferromagnetic state. It occurs
tion) line width to be determined solely by the
in antiferromagnets that the exhibit the Dzialoshin-
width of the corresponding nuclear levels. The
skii interaction, and involves a change of spin
latter is much smaller than the energies of the
orientation. The Morin transition results from the
magnetic dipole interaction and electric quadru-
change of the equilibrium direction of the antifer-
pole interaction of a nucleus with electrons, and
romagnetism vector L = M 1 − M 2 with changing
this permits studying the hyper ne structure of
temperature. The temperature where the Morin
transition takes place is called the Morin point. solids (see Mössbauer spectroscopy). Owing to
This transition was first seen in haematite (α- the Mössbauer effect, the energy of γ -quanta is
Fe2 O3 ) at T M = 258.5 K, and has been detected measured with an accuracy 10−13 –10−16 (ratio
in a number of orthoferrites and other antiferro- of line width to energy of γ -quantum) that is an
magnets. all-time high for electromagnetic radiation. The
Mössbauer effect finds its use in chemistry, solid
MOSAIC CRYSTALS, mosaic structure (fr. Ital. state physics, metallurgy, geology, and biology.
mosaico, picture, ornament)
Crystals that are aggregations of small mono- MÖSSBAUER EFFECT, OPTICAL ANALOG
crystalline blocks, slightly disoriented relative to Generation of narrow phononless lines in opti-
one another. Practically every real crystal is, to a cal spectra of imperfect crystals. Narrow absorp-
greater or lesser extent, a mosaic crystal. The sizes tion peaks in both γ -spectra and optical phonon-
of mosaic blocks range from several hundreds less lines result from quantum transitions between
to several thousands of crystallographic lattice energy levels of defects, which involve neither
constants, with disorientations of several angular emission nor absorption of phonons; from this
minutes. For the larger size ranges the blocks are point of view they are much alike. However,
considered to be grains having ordinary low-angle there are also important differences in the physical
grain boundaries. Thus, there is a continuous range mechanisms for generating these spectral lines, in
of ordered states between mosaic structures and the conditions of their observation, as well as in
those characterized by ordinary small-angle grain applications of these phononless spectra.
boundaries. The mosaic structure forms during
crystal growth (see Crystallization, Block structure) MÖSSBAUER ELECTRON SPECTROSCOPY
as a result of the appearance of various defects in Spectroscopy based on the detection of in-
the crystal (see Defects in crystals). ternal conversion electrons and Auger electrons
generated as a result of resonance absorption of
MÖSSBAUER EFFECT, nuclear gamma ray γ -quanta (Mössbauer effect). This technique per-
resonance (R. Mössbauer, 1958) mits the determination of the chemical state of
Emission and absorption of gamma quanta by Mössbauer atoms in surface layers of thickness
atomic nuclei bound in a solid. These feature al- ranging from several nm to hundreds of nm. Infor-
most zero “recoil” energy of a nucleus (an effect mation in the form of the spectra is obtained with
860 MÖSSBAUEROGRAPHY
the resonance value of the γ -quantum energy and to 10−9 eV) relative resolution of the transferred
the energy width at resonance differ from the en- energy, and the fact that it is not necessary to
ergy and width of a Mössbauer level of an isolated have Mössbauer nuclei in the sample. The ma-
nucleus. The collective nature of the interaction jority of studies are with single crystals, and are
manifests itself also in the suppression of inelastic devoted to determining the Debye–Waller factor, to
channels of nuclear reactions (Kagan–Afanasyev the identification of thermal diffuse scattering, to
effect). Under conditions of Bragg diffraction by the separation of static and dynamic diffuse scat-
a perfect crystal, γ -radiation penetrates abnor- tering, and to the determination of the nature of the
mally deeply compared to in a mosaic crystal, or quasi-elastic, so-called central peak, in the diffrac-
to the absence of diffraction. This phenomenon tion study of structural phase transitions. Inelastic
is a nuclear analogue of the Borrman effect (see scattering from crystal lattice distortions around
Anomalous passage of X-rays). defects, relaxation motion in crystals with struc-
Diffraction of Mössbauer radiation by crystals tural disorder, scattering in solid electrolytes, etc.
that contain no Mössbauer nuclei also exhibits a is investigated. In diffraction studies of biological
specific feature. The use of extremely high en- systems important results have been obtained in-
ergy resolution of Mössbauer detectors makes it volving the influence of water on the dynamical
possible to separate elastic and inelastic scattering properties of proteins, etc. The low intensity of
in the region of a Bragg maximum. This per- the diffracted beam, and therefore the low reso-
mits the study of the dynamics of atomic motion lution of the momentum being transferred, is the
with variations of the crystal temperature by ob- main disadvantage of this technique. These diffi-
serving some of the elastically scattered quanta. culties are partly overcome by the application of
Kinematic and dynamic theories of the diffraction coordinate-sensitive detectors, and also of ring de-
of Mössbauer γ -radiation by crystals have been tectors, and by synchronizing the motions of the
developed, and the main physical processes of in- source and absorber of the gamma resonance radi-
teraction of this radiation with matter have been ation. See also Mössbauerography.
studied. Mössbauerography experiments are un-
der way in all the above directions. A significant
MÖSSBAUER SPECTROSCOPY, nuclear
obstacle to the widespead applicability of Möss-
gamma resonance spectroscopy
bauerography is the unavailability of powerful
Method based on the Mössbauer effect of in-
sources of Mössbauer γ -quanta. There is hope for
vestigating hyper ne interactions between nuclei,
creating nondecaying sources of resonance scat-
and electric and magnetic fields, and hence various
tering through synchrotron radiation, as well as
physical properties of solids. Mössbauer spec-
through resonance γ -radiation, which arises when
charged relativistic particles interact with crystals. troscopy exhibits a remarkably high accuracy of
The creation of a Mössbauer γ -laser would pro- measuring the energy of γ -quanta, which is de-
vide a powerful source for mössbauerography. fined by ratio of the resonance band width Γ0
(10−10 –10−8 eV) to the energy E0 of a nuclear
MÖSSBAUER RADIATION RAYLEIGH transition Γ0 /E0 ≈ 10−10 –10−16 . This makes
DIFFRACTION it possible to observe shifts of nuclear energy
Nuclear gamma resonance radiation (Möss- levels (10−4 eV) caused by hyperfine inter-
bauer radiation) Rayleigh diffraction takes place actions: electrostatic (Fig. 1(a)), electric quadru-
at electronic shells of atoms and ions in condensed pole (Fig. 1(b)) and magnetic dipole interactions
media. It is usually carried out in accordance with (Fig. 1(c)) between a nucleus and surrounding
the X-ray diffraction scheme: gamma resonance electrons by recording Mössbauer spectra (nuclear
radiation is scattered by the sample with sub- gamma ray resonance spectra). They represent
sequent energy analysis of the diffracted beam the dependences of the number N of detected
with the help of a Mössbauer absorber. As a γ -quanta that pass through an absorber (transmis-
rule, nuclei of Fe57 with a γ -quantum energy of sion mode, Fig. 2(a)), or were scattered by the
14.4 keV are used as the source. The main ad- sample (scattering mode, Fig. 2(b)), or of X-rays
vantage of this procedure is the very high (∼10−8 and electrons that were emitted by the sample
862 MÖSSBAUER SPECTROSCOPY
Fig. 1. Nuclear gamma resonance absorption spectra of 57 Fe, their energy levels, and their parameters.
due to internal conversion, on the extent of mis- (b) Line (level) width Γ e , where Γ e > 2Γ0 due
match between the energies of the emitter and to broadening caused by unresolved hyperfine
absorber (scatterer). The method of modulating structure, diffusive motion of atoms, defects
γ -quanta energy, which results in the mismatch, of crystal structure, and self-absorption effects
is the most widespread version, and it is based on in the case of thick samples. Depending on
the Doppler effect. The γ -radiation source is set the broadening mechanism, Γ e yields infor-
in motion past the absorber (scatterer) at a veloc- mation on the inhomogeneity of composition,
ity relative to the absorber (scatterer) v = Γ0 c/E0 , the structural and magnetic states of solids,
so that the γ -radiation energy changes by
E = and the presence of defects.
±E0 v/c. Nuclear gamma resonance absorption (c) Shift δ = δi + δT of nuclear gamma resonance
spectra (similar to those of scattering) of conver- spectra relative to a standard. The isomer shift
sion electrons exhibit hyper ne structure, which δi is caused by the interaction of the nucleus
depends on the type of hyperfine interaction, and with the electrostatic field arising from the
is described by the following parameters (Fig. 1): surrounding electrons, and depends on the dif-
(a) The magnitude of the Mössbauer effect h and ference in the charge density of s-electrons
area S depend (in general, in a nonlinear man- of the absorber and emitter nuclei. A temper-
ner) on the number of nuclei n and on the ature shift δT is associated with the second
Mössbauer effect probability f (probability order relativistic Doppler effect, and charac-
of emission or absorption of γ -quanta with- terizes vibratory motion of atoms in a solid.
out recoil; f is expressed via the mean square The quantity δ is an important characteristic
atomic oscillation amplitude, it characterizes of atom bonding in solids; it depends on the
the interatomic interaction in solids, and is a composition, structure, magnetic state, tem-
function of the temperature). perature, and pressure.
MOTT–DAVIS MODEL 863
Fig. 2. Diagram of Mössbauer spectrometer: (a) transmission mode; (b) scattering mode. 1, modulator of γ -quanta energy
with a radiation source; 2, modulation signal generator; 3, absorber or scatterer under study; 4, radiation detector; 5, pulse
recording and analyzing unit; 6, collimator.
(d) Quadrupole splitting 2E d caused by the inter- and structural transformations, atomic order-
action of the nuclear quadrupole moment Q ing, relaxation phenomena, texture, magnetic
(Q = 0 for nuclei with spin I > 1/2) with anisotropy, phase analysis, etc.
the electric field gradient. It splits nuclear
levels into two or more sublevels. For in-
MOTT–DAVIS MODEL of density of states
stance, nuclei of 57 Fe (or 119 Sn, 125 Te) have
two sublevels, and accordingly, there are dou- (N.F. Mott, E.A. Davis, 1970)
blets in nuclear gamma resonance spectra (see Model of electronic density of states of amor-
Fig. 1(b)). The magnitude of the quadrupole phous semiconductors. According to the Mott–
splitting yields information on the asymmetry Davis model, tails of localized states of the va-
of the volume charge distribution, distortion of lence band and conduction band are relatively
the crystal lattice, the presence of impurities. narrow, and spread into the band gap to a depth
(e) Magnetic splitting B typical of nuclear gamma of several tenths of an eV. There is a narrow band
resonance spectra of magnetically ordered of compensated levels in the center of the gap
materials. The interaction of the magnetic di- which owes its existence to defects in a random
pole moment of a nucleus μ with the magnetic continuous network of atoms; it “fixes” the Fermi
field B at a nucleus leads to the Zeeman effect. level to the center of the gap. It follows from the
The nuclear gamma resonance spectrum in Mott–Davis model that the electrical conductivity
this case is a set of lines; their number equal- has three components, each of them dominating in
ing the transitions between sublevels that are its temperature region. At very low temperatures,
allowed by the selection rules. For instance, the conductivity is due mainly to thermally acti-
the nuclear gamma resonance spectrum of vated jumps of carriers between states at the Fermi
57 Fe exhibits six lines (see Fig. 1(c)) with the level, and is realized through a hopping mecha-
magnitude B measured from the distance be- nism (see Hopping conductivity). With a rise in
tween the extreme hyperfine components (for temperature, the conductivity increases because of
57 Fe, there is a sextet of lines corresponding the excitation of carriers into localized states in the
to transitions ±3/2 → ±1/2). The magnetic tails of the corresponding energy bands. These car-
hyperfine structure is sensitive to composi- riers also take part in charge transport through a
tion, temperature, pressure, and applied mag- hopping mechanism. On a further increase of tem-
netic field, and is used to study magnetic perature, the main mechanism responsible for the
864 MOTT LAW
conductivity arises from carriers that are excited in the lower band. If the concentration n of car-
into delocalized states. See also Amorphous semi- riers is low, they form a bound state, an exciton.
conductors. The electron gas screens the Coulomb interaction
by decreasing the energy of an electron–hole pair.
MOTT LAW (N.F. Mott, 1969) At n > nc , the generation of excitons is no longer
Temperature dependence of electrical conduc- possible because the electrons delocalize and the
tivity due to jumps or hopping of current car- system becomes a metal. Thus, strong electron–
riers over impurity- and defect-related localized electron correlation leads to an abrupt change
states near the Fermi level. A model of local- of free carrier concentration at the Mott metal–
ization centers randomly distributed in space ex- insulator transition. The value of nc is determined
plains why the temperature dependence of hopping by the distance d0 between electron and hole. The
conductivity is weaker than in an ordinary ac- Mott transition occurs in materials with carriers
tivation case. The Mott law may be written as of the sp-type, e.g., transition-metal compounds
σ (T ) = σ0 (T ) exp[−(T0 /T )n ], where σ0 (T ) is a (V2 O3 , NiS, etc.) and doped semiconductors. The
weak function of the temperature, T0 is a con- last impurity band in the insulator state is split
stant of the order of 108 K, the exponent n is by the Coulomb interaction U . The Density of
always smaller than 1 and depends on particu- states at the Fermi level N(EF ) becomes finite
lar physical properties of the object (relationship when the lower Hubbard band of width B (see
between density of states at the Fermi level; de- Hubbard model) overlaps the upper one: B U .
gree of electron–phonon interaction; presence of This inequality roughly determines the condition
correlation effects; system dimensionality, etc.). for the Mott metal–insulator transition. Disorder-
In the three-dimensional case when the density ing effects may blur the upper band boundary,
of localized states is independent of energy and and form a density of states “tail”. As only local-
the electron-phonon interaction is negligible, we ized states contribute to this “tail”, the transition
find n = 1/4 theoretically. Reported experimen- takes place at considerable overlap of both bands
tal studies of hopping conductivity of noncrys- and finite N(EF ). Experimental data from many
talline solids confirm the validity of the Mott types of semiconductors fit the universal relation
1/3
law, but with values of n in the wider range: nc a0 ≈ 0.27.
0.2 n 0.55. Mott (1978) showed that an energy gap de-
crease
E can result in an abrupt change of n
MOTT–LITTLETON APPROXIMATION (N.F. Mott, from zero to nc , the position of the energy min-
M.J. Littleton, 1938) imum of the electron gas. Since this minimum
A method of calculation of a static configura- equals the sum of the kinetic energies of electrons
tion of defects in a crystal lattice (often ionic) that and holes (∼n2/3 ) plus the potential energy of
takes into account long-range Coulomb forces and their attraction ∼(−n1/3 ), it is attained at Emin <
polarization forces. The physical significance of 0. Thus, spontaneous generation of an electron-
the approximation is that the electric dipole mo- hole gas becomes energetically favorable before
ments of ions, which are induced by a charged the two bands overlap at
E −Emin . In most
defect and located far from it, are determined by cases, this gas has a lower energy than the alterna-
the macroscopic polarization. tive phase of a Bose condensate of excitons.
Since generation of a free electron and hole
MOTT METAL–INSULATOR TRANSITION requires a finite energy, the Mott transition must
(N.F. Mott, 1949) be accompanied by an abrupt change of free
Phase transition in an electron gas, which energy F . Strong enough disorder cancels such
brings about an abrupt change of electrical con- jumps of n and F . This results from the fact that
ductivity. In the dielectric (insulator) state there is if an N(E) “tail” exists, then the kinetic energy
a gap between completely filled and empty energy of the electron–hole gas depends less on n, which
bands. An electron transition results in the appear- results in Emin > 0 (Mott and Kaveh, 1983).
ance of a free electron in the upper band and a hole On further disordering, localization of electron
MULLIKEN–WOLFSBERG–HELMHOLZ METHOD 865
Z + Δμ
A = −I A −
Hμμ B ,
μ
RAB in the figure, where second-order phase transitions
B(=A)
are indicated by dashed lines, first-order ones are
where IμA is the electron energy of atom A in indicated by solid lines; and the symbol I desig-
its one-particle state, taking into account the total nates regions of phases which possess the highest
charge and configuration of other electrons in the symmetry.
valence shell, ZB is the effective Mulliken charge Tricritical points also include points where
of atom B, corrected by Δμ for electric charge curves of first-order phase transitions turn into
screening, and RAB is the interatomic distance. curves of continuous phase transitions. These
Off-diagonal elements are estimated from the for- points are also called critical points of second-
mula order transitions, or Landau critical points.
1 A material can be brought to a multicritical
Hμν = − k(Hμμ + Hνν )Sμν , point by changing more than one external para-
2
meter. These usually are temperature, pressure,
where k = 1.67–2.0; other equations have also
electric or magnetic field, concentration of one
been proposed. An important feature of the Mul-
of the components of a solid solution. Multicriti-
liken–Wolfsberg–Helmholz method is the proce-
cal points have been found in compositional phase
dure for self-consistency in charges and configu-
depend on. This procedure diagrams of many ferromagnets and antiferromag-
rations that IμA and ZB
uses the results of the preceding iteration to de- nets, ferroelectrics, and crystals with various types
termine the total charge on each atom, while IμA of structural ordering.
values for the next iteration are calculated through
interpolation from known ionization potentials of MULTIPHOTON PROCESSES
the atoms and corresponding ions. The Mulliken– Interaction processes between electromagnetic
Wolfsberg–Helmholz calculations allow determi- radiation and atoms and molecules of matter when
nation of binding energies, ionization potentials, several photons simultaneously participate in each
energies of electronic transitions, distributions of event (absorption, emission). The greater the num-
charge and spin density, as well as conformations ber of photons that participate the less probable
of complexes. and the more difficult to observe are the processes.
Multiphoton processes may either have no thresh-
MULTIAXIAL MAGNET old, or exhibit a threshold value of radiation inten-
See Magnetic anisotropy. sity where they appear.
Raman scattering of light is a two-photon
MULTICRITICAL POINT, polycritical point process involving the simultaneous absorption of
A point on the compositional phase diagram a photon of one frequency and the emission of
of a substance, where regions of several different a photon of another frequency in a crystal. The
phases come in contact, these regions (not nec- involvement of a phonon provides the energy bal-
essarily all of them) being separated by second- ance. This process has no threshold; neither does
order phase transition lines. Bicritical points, triple two-photon light absorption. The coefficient of
critical points and tetracritical points exemplify two-photon absorption is proportional to the wave
multicritical points. Schematic representation of intensity. Therefore, the larger the incident wave
corresponding regions of phase diagrams are given amplitude, the more rapid the attenuation of the
MULTIPOLE 867
wave in the crystal. In strong fields of laser ra- initial position of the segment is recovered. Pro-
diation, nonlinear absorption of higher order is vided σ > σs the Frank–Read source will contin-
observed. uously generate dislocation loops until its action
Three-photon processes include second har- is blocked by the long-range stresses produced by
monic generation, excitation of waves at sum and loops already formed. The action of other sources
difference frequencies, parametric light scatter- follows the same principle. However, in the case of
ing, and parametric light generation. The proba- a double transverse slip the pinning positions of
bility of multiphoton processes is proportional to a dislocation segment are the ends of sections of a
product of the densities of photons that take part in screw dislocation that has executed its cross slip
the absorption and emission, i.e. the product of the into a nearby plane, and then returned to a slip
intensities of waves of different frequencies. Be- plane parallel to the initial one. In a pole source,
sides that, this probability depends on the structure only one end of the segment is fixed, and the other
of the energy levels of the material plus the posi- can eventually reach the crystal surface.
tion of combination frequencies relative to these The Bardeen–Herring sources base their func-
levels, and it increases sharply if one or several tion on the processes of crawl of a fixed disloca-
combination frequencies are close to those of in- tion segment that proceed via the absorption of
termediate transitions. The increase of the incident point defects, and this results in the formation of
radiation intensity also causes the probability of helical dislocations and prismatic dislocations in
multiphoton processes to increase. When passing the crystal bulk.
from single-photon processes to multiphoton ones
the selection rules change. Thus, in centrally sym- MULTIPOLE
metrical crystals a transition between states of like An aggregation of electric charges (ei ), which
parity involves an even number of photons, and is characterized by a multipole moment. It may
that between states with opposite parity involves be introduced through
the potential of a sys-
an odd number (see Mutual exclusion). Therefore, tem of charges e
i i /|R ± r i | at distances R
transitions between different energy states that are from this system that far exceed the distance
forbidden for a certain set of photons are allowed r i to any ith charge (origin of coordinates is
for another number of photons. Investigations of inside the multipole). For R
ri , expanding
multiphoton absorption spectra can involve study- |R − r i |−1 in a power series yields: d0 = i ei ,
ing transitions forbidden for single-photon excita- zeroth-order moment or monopole moment; d 1 =
tion. This method finds wide use in the nonlinear r i , first-order moment or dipole moment;
i ei
optics of solids, and multiphoton processes ac- d2 = i eI (3riα riβ −δαβ ), second-order moment
companied by photon emission are used in optical or quadrupole moment (see Quadrupole), etc.,
frequency converters. where riα and riβ are projections of the vector r i
on the coordinate axes, and δαβ is the Kronecker
MULTIPLICATION OF DISLOCATIONS delta. In the general case one speaks about the nth
An increase of the total length of dislocation order moment.
lines in the bulk of a crystal developing during the Not only does a multipole moment determine
course of plastic deformation. It results from the fields at a distance R, but it also characterizes the
action of several sources, the main one being the behavior of the multipole in external fields and, in
Frank–Read source. To picture it one may con- particular, there is the multipole interaction. Thus,
sider a section of dislocation of length L fixed at the expression d 1 ·E 0 defines the interaction en-
its ends in the slip plane. It bends out under the ap- ergy between a dipole and the uniform part of the
plied stress σ , and upon reaching a certain critical electric field E 0 at the dipole. Also, the expres-
configuration at σs = Gb/L (G is the shear mod- sion [(d 1 ·d 1 )R 2 − 3(d 1 ·R)(d 1 ·R)]/R 5 defines
ulus, b is the Burgers vector), determined by the the dipole–dipole interaction between two dipoles
balance between the external force and the elas- d 1 and d 1 spaced a distance R apart that far
tic stress force, the dislocation segment loses its exceeds their sizes (r i ). Also known are dipole–
stability. This loss of stability results in the forma- quadrupole, quadrupole–quadrupole and other in-
tion of a closed dislocation loop, within which the teractions of higher order.
868 MULTIWELL POTENTIAL
By analogy with electrical multipoles, mag- parity is not conserved in this decay, the angular
netic, elastic and other kinds of multipoles can be distribution of positrons (or electrons) is asymmet-
introduced. The concept of multipole is used in rical relative to the direction of the muon spin.
solid-state physics for investigating processes of Positrons are emitted mostly in the same direction
absorption (emission) of electromagnetic waves, as the muon spin. This underlies the application
phase transitions, broadening of resonance lines, of stopped muons μ+ for the investigation of ma-
etc. The concept of multipole has given rise to terials through the method of muon spin rotation.
scientific disciplines and new concepts, such as Muons are moderated in the material, and either
quadrupole resonance, dipole–dipole reservoir, di- become localized or diffuse. Observation of μ+
pole instability (see Vibronic instability), etc. (by detecting the decay of positrons) is feasible in
the time interval from 10−7 to 10−5 s.
MULTIWELL POTENTIAL An external magnetic field B applied to the
See Many-well potential. sample causes the muon spin to precess about the
direction of the field. As this takes place, the num-
MUONIUM ber of positrons which escape in specific directions
An atom (μ+ e− ), sometimes referred to as varies with time in a regular manner. The preces-
a light (“exotic”) isotope of hydrogen, which is sion frequency is ωμ = eB/mμ with the value
formed when a positively charged muon μ+ forms 13.55 MHz/T, where mμ is the muon mass. The
a bound state with an electron e− , an event that precession amplitude is proportional to the degree
sometimes occurs in semiconductors and insula- of polarization of the ensemble of muons. Depo-
tors. Muonium is unstable, and has a lifetime of larization caused by interaction of the muon spin
2.2 ms. All electrical and chemical interactions and magnetic moment with atoms of the medium
of the atom μ+ e− in nonmetals are the same as results in a decrease of the precession amplitude.
those of a hydrogen atom; therefore, the investiga- This fact is used for studying properties of solids:
tion of the dynamics of the behavior of muonium phase transitions in magnetic materials and super-
in matter yields information on hydrogen behav- conductors, and phenomena of quantum diffusion
ior. Besides the normal state (μ+ e− ), there is also of muons in metals. When stopped in a mater-
the state of anomalous muonium (μ+ e− )∗ ; in it, ial negatively charged muons (μ− ) form mesic
the hyper ne interaction is weaker than in μ+ e− , atoms. Muons μ+ form muonium atoms (μ+ , e−
it is anisotropic, and the precession frequency in bound state) in semiconductors and insulators. The
a magnetic field is different. The (μ+ e− )∗ is a phenomenon of channeling in monocrystals was
bound system located in the neighborhood of a observed for μ+ .
lattice atom along a 111 axis in silicon. Unlike
(μ+ e− )∗ , normal muonium (μ+ e− ) is an atom MUTUAL EXCLUSION, alternative exclusion
which rapidly diffuses in matter. A selection rule for transitions involving the
parity conservation law in quantum systems.
MUONS According to mutual exclusion, electric-dipole
Lepton elementary particles with spin 1/2, transitions are only possible between states of dif-
charge +e or −e (where e is the unit elec- ferent parity. Magnetic-dipole and electric-quadru-
tric charge), rest energy 105.66 MeV, magnetic pole electron transitions only occur between states
moment Mμ = 3.18μp (μp is magnetic mo- of the same parity. Mutual exclusion holds in the
ment of proton). The former (incorrect) name was presence of a center of symmetry, relative to which
μ-meson. Intense beams of muons are generated the states are divided into odd and even types.
by accelerators through the decay of π -mesons A corollary of mutual exclusion is the phonon op-
π ± → μ± + ν, where ν is a neutrino. This decay tical spectrum rule whereby transitions symmetric
produces muons that are polarized along their mo- relative to the center of symmetry are forbidden
mentum direction. The lifetime of a muon at rest is in IR absorption spectra, while those that are anti-
∼2.2·10−6 s; a muon decays into an electron, neu- symmetric are forbidden in spectra of the Raman
trino and antineutrino: μ± → e± + ν + ν. Since scattering of light.
Nn
NANO. . . (fr. Gr. νανoς , dwarf) the shear modulus is 25–30% lower. Internal fric-
Prefix for a physical unit name, which des- tion in nanocrystals is noticeably higher than in
ignates a multiple unit equal to 10−9 initial crystals. Diffusion of atoms takes place mostly
units. Symbol: n. Example: 1 nm (nanometer) = at boundaries, with an activation energy of self-
10−9 m. diffusion 3–4 times smaller than that in crystals.
The relatively high plasticity is due to diffusive
NANOCOMPOSITES transport of atoms and slippage along the bound-
Combined or composite materials with sizes of aries. High-strength ceramic nanocrystals are ob-
particles of an imbedded phase much smaller than tained by oxidation of metallic samples. Proper-
the wavelength of visible light. Nanocomposites ties of nanocrystals are close to those of amor-
consisting of an optically isotropic matrix contain- phous materials, so a possible method for their
ing particles of high quadratic or cubic nonlinear production is a rapid crystallization of amorphous
electric susceptibility (dielectric constant) exhibit bodies. They can be synthesized with sizes down
nonlinear optical properties. This allows their use to 2 nm through methods of colloidal chemistry.
in devices of integrated optics for the control of the Nanocrystal materials may be used as catalysts,
passage of light beams, including modulation, de- high-strength construction materials, etc.
flection and optical radiation frequency changes.
NANOELECTRONICS
NANOCRACK
Subfield of semiconductor physics that deals
See Submicrocracks.
with artificial devices having dimensions or su-
NANOCRYSTALS perstructure components (see Semiconductor su-
Polycrystalline materials with crystallite sizes perstructure) equal to or smaller than 102 nm in
of 5–10 nm (some authors say 2–50 nm). Nano- order of magnitude. Such devices, also known as
crystals are obtained through pressure compacting nanostructures, can be prepared by high resolution
of ultradisperse powders, which form during vapor lithography, molecular beam epitaxy, and self as-
condensation in an inert gas atmosphere, and by sembly.
chemical methods. The density of nanocrystals is Some variants to the general name “nanostruc-
10–15% lower than that of crystalline samples; the tures” are: quantum dots, single-electron transis-
density of boundaries is about 600 m2 /cm3 , while tors, Coulomb islets, and finally artificial atoms.
about 20% of the atoms belong to boundary or Most artificial atoms are created on the basis of
surface layers. Structure studies show that bound- heterojunctions of the AlGaAs–GaAs type. As the
ary layers (about two atomic diameters thick) are relative dielectric constant of gallium arsenide is
significantly disordered and contain many cav- ε = 13, and the value of the electron effective mass
ities comparable with atomic sizes, as well as m∗ comes to only 7% of the free electron mass m,
larger cavities (10–100 atomic volumes in size). the effective Bohr radius a ∗ is εm/m∗ times
A nanocrystal is a convenient object for the study greater than a hydrogen atom radius, and equals
of solids with inclusions of locally disordered ar- ≈10 nm, whereas the effective Rydberg constant
eas. The speci c heat of nanocrystals is approxi- Ry∗ (effective Rydberg) is m∗ /mε2 times smaller
mately 1.5 times higher than that of crystals, and than its standard value, and equals ≈6 meV. As
869
870 NANOTECHNOLOGY
a result artificial atoms, like ordinary ones, pos- NANOTECHNOLOGY, submicrometer technology
sess quantum properties at liquid helium temper- A complex of technological processes involved
atures, since their sizes are close to a ∗ in order in the development and manufacture of semicon-
of magnitude, and thermal diffusion is small by ductor integrated circuits or other structures with
virtue of Ry∗
kB T . The latter statement is con- dimensions much smaller than a micrometer. Im-
firmed by numerous optical investigations in the portant processes are X-ray or electron lithogra-
IR region of the spectrum, as well as by methods phy, ion implantation doping, molecular epitaxy,
of tunneling spectroscopy. In order to enhance the ion etching, and plasma etching. A pattern on a
effect, one often works with a large number (up to semiconductor substrate is obtained by exposing
108 ) of artificial atoms, embedding them at reg- a mask, i.e. by irradiating the film made of a
ular sites of a two-dimensional “crystal lattice”, special resist by either a parallel beam of soft
which makes systems under consideration akin X-ray radiation (synchrotron radiation), or an elec-
to artificial superlattices. The positions of spec- tron beam that is scanned over the resist surface in
tral levels and, accordingly, of IR absorption lines, accordance with a given program. The polymer-
may be changed by continuously varying the volt- ization of the photoresist film is used for its fur-
age at the artificial atom gate, as well as by plac- ther removal, and for forming a mask needed ei-
ing nanostructures into external electric and mag- ther for adding an insulator or metal layer, or for
netic fields. In particular, when periodic superlat- ion doping. The formation of conductive, semi-
tices were placed into a uniform electric field E conducting, or insulating layers is carried out with
directed transversely to the layers it was discov- the aid of particle beams (molecules or ions)
ered that an electron band split into a periodic set in a high vacuum. Related to nanotechnology is
of quantum levels, a phenomenon known as the molecular technology that uses molecular-ordered
Wannier–Stark ladder (see Quantum mechanics in structures of organic molecules, e.g., Langmuir–
solid state physics); an energetic step of this ladder Blodgett lms.
is equal to |e|EL, where e is the electron charge,
and L the superlattice constant. Such a spectrum
results from the quantization of Bloch oscillations, NARROW-GAP SEMICONDUCTORS
which were predicted by F. Bloch at the dawn of Semiconductors which have a narrow band gap
the development of solid state theory. width. The lower limit to the size of the band gap
Besides a discrete energy spectrum, another is represented by semiconductors with a vanishing
quantum property of matter manifests itself in ar- gap (gapless semiconductors), and the nominal
tificial atoms: the discrete nature of charge. This upper limit is conventionally taken to be ≈0.3 eV.
property better matches the alternative name of The best known narrow-gap semiconductors are
an artificial atom: single-electron transistor. The the binary compounds InSb, PbTe, PbSe, binary
point is that the addition of charge to an artifi- alloys of the general formula Bi1−x Sbx , as well
cial atom increases its energy by ∼e2 /C, which, as two classes of isovalent solid solutions based on
due to the low value of the capacitance (C) of chalcogenides of Groups II and IV: Cdx Hg1−x Te,
the device, coincides in order of magnitude with and also Pb1−x Snx Te and Pb1−x Snx Se. The al-
the distance between energy levels of an artificial loy Cdx Hg1−x Te behaves as a zero-gap semi-
atom. Thus, if there is no voltage at the gate, the conductor for x < 0.16; the width of the forbid-
transfer of an electron from emitter to collector at den zone increasing with increasing x for x 0.
kB T e2 /C is forbidden, since this would en- Narrow-band gap semiconductors may be classi-
tail the climbing of the electron over the barrier, fied as an individual group of semiconductor ma-
which is ∼e2 /C in height; thus the device carries terials owing to their unique physical properties,
zero current (the so-called Coulomb blockade). and their areas of practical application. As a con-
The blockade is pierced only at particular gate sequence of the narrowness of the band gap, the
voltages, which is accompanied by a tremendous electron effective mass (holes in case of PbTe,
increase of conductivity (by 103 to 104 times). See PbSe, both pure and alloyed with Sn) is abnor-
also Solid-state quantum electronics. mally low, whereas the mobility is extremely high.
NARROWING OF MAGNETIC RESONANCE LINE 871
Unlike semiconductors with wide gaps, the spec- NARROWING OF MAGNETIC RESONANCE
trum of free current carriers in narrow-gap semi- LINE
conductors is highly nonparabolic, with the degree Line width decrease due to modulation of the
of nonparabolicity increasing with the decrease of paramagnetic center resonance frequency by dy-
the width of the band gap. The electronic proper- namic processes within the spin system. A classi-
ties of AII BVI and AIV BVI compound types are cal example of this line narrowing is the strong de-
highly sensitive to crystal lattice defects involving crease of the magnetic dipole line width in liquid
deviations from stoichiometric composition, i.e. to and gaseous samples compared to their values in
excess quantities of metal or chalcogenide atoms. solids at the same concentration of magnetic mo-
The predominance of metal vacancies results in ments. This kind of narrowing is called motional
p-type electrical conductivity, whereas dominating narrowing since it is due to the rapid motion of
chalcogen vacancies give rise to n-type conductiv- atoms in the environment. A rapid random motion
ity. of particles produces fluctuations of the internal lo-
The widespread interest in studies of narrow- cal electric elds that act on every spin. As a result,
gap semiconductors is explained by the fact that within a time greater than the fluctuation duration,
these materials are used as infrared (IR) devices only the mean value of the local field is effec-
such as photodetectors, laser diodes, and semi- tive, a value much smaller than the instantaneous
conductor radiators operating over the wave range value, so that line narrowing occurs. To produce
4–30 μm. Narrow-band semiconductors with a motional narrowing of the line, the field fluctua-
predetermined band gap may be adjusted by vary- tions must be rapid compared to the Larmor pre-
cession in the local field. Since the precessional
ing the ratio of the alloy components. This al-
frequency in the local field is of the order of mag-
lows the construction a device for a particular
nitude of the resonance line width δω (with static
wave band. Binary compounds of elements of
surrounding spins), the condition for line narrow-
Groups III–V (InSb, InAs) are used for produc-
ing is δω·τc 1 (here τc is the local field cor-
ing liquid-nitrogen cooled photodetectors that ex-
relation time). The width
ω of the line narrow-
hibit high performance over the range of 3–5 μm,
ing can be qualitatively estimated as the inverse of
which corresponds to the first atmospheric spec-
the time interval
t during which two spins, pre-
tral window. Photodetectors based on polycrys-
cessing in corresponding local fields which are ini-
talline layers of lead selenide may also be used tially in phase, become out of phase by the order of
over this operating range. The latter photodetec- π/2, and the width of the narrowed line becomes
tors, which either do not require cooling at all or
ω ∼ (δω)2 ·τc .
need slight thermoelectric cooling for their opera- The modulation of local fields which leads to
tion, are the simplest in design, and readily avail- narrowing of the line can also account for the rapid
able for the IR spectral region. The early inves- reorientation of neighboring spins. Such a type
tigations of mixed crystals HgCdTe and PbSnTe of narrowing appears, e.g., in NMR pulse exper-
were aimed at creating refrigerated IR detectors iments, and this is called pulse line narrowing.
for the wave band 8–14 μm, which corresponds to The essence of pulse narrowing consists of excit-
the second atmospheric spectral window, and this ing the sample by a repeating sequence of radio-
is of considerable interest for communication. All frequency pulses which turn the spins in various
the above-listed photodetectors may be used for directions through large angles. This process, like
measuring temperature gradients in the environ- the narrowing of a line due to molecular motion,
ment, or for medical purposes, for their spectral averages out the dipole–dipole interaction and re-
range corresponds to the thermal radiation max- duces the dipolar width.
imum in the neighborhood of room temperature. Rapid reorientations of neighboring spins
Tunable diode lasers based on lead chalcogenides which induce line narrowing can also take place
are widely used for producing spectral devices that as a result of internal interactions in the spin sys-
feature high resolving power, and are used in sci- tem. Such a phenomenon occurs, in particular, in
entific investigations, as well as for analyzing gas the presence of a strong exchange interaction be-
mixtures in industry and ecology. tween spins which induce exchange narrowing.
872 NEARLY-FREE ELECTRON APPROXIMATION
The exchange interaction operator is expressed in conditional strain of the neck formation is given
the form of a sum of scalar products of spin op- by εmax ≈ n(σB − σS )/σB , where σB is the ulti-
erators; therefore, in the case of a system of iden- mate strength, and σS is the yield limit of the ma-
tical spins this does not contribute to the second terial. The appearance at the neck of the uniax-
moment of the line M2 (see Moments of spectral ial tension leads to the change of a uniform state
lines), the usual indicator of line broadening. How- of stress to a triaxial one. At ε > εmax (second
ever, owing to the presence of spin operators of the stage) the plastic deformation becomes rather het-
type Si+ Sj − (i and j denote different spins) in the erogeneous. For conditions of increasing nonuni-
exchange interaction which induces flip-flop tran- formity of stress distribution, it is accompanied
sitions (see Antiferromagnetism), a random mod- by the hardening of the neck material, and by the
ulation of the dipole and hyperfine local fields dense local occurrence of pores. At the third stage,
takes place that results in the narrowing of the line. in the central part of neck at the point of maxi-
Quantitatively, this exchange interaction narrow- mum triaxial tension resulting from the merging
ing brings about an increase of the fourth moment, of the pores, a microcrack is formed (see Crack)
so the observed magnitude of the width appears which subsequently increases and brings the de-
considerably smaller than (M2 )1/2 , in contrast to formation process to an end by failure (tough fail-
the case of the pure dipole–dipole interaction. ure, quasi-brittle failure). In the neck a texture of
the axial type arises.
NEARLY-FREE ELECTRON APPROXIMATION The word neck is also used to designate narrow
Method of band theory of solids based on the connections in adjacent Fermi surfaces.
use of perturbation theory with the crystalline
potential (weakly-bound electron approximation) NEEDLE-SHAPED DEFECT
considered to be a small perturbation compared to
A linear defect in crystals corresponding to a
the free particle kinetic energy. This method finds
closed Burgers contour. If u(x1 , x2 , x3 ) (where
its explanation in the pseudopotential method. It
x1 , x2 , x3 are Cartesian coordinates of all crystal
is used for determining the Fermi surfaces of met-
points except those on the defect line) is the dis-
als.
placement vector due to a linear defect, and dl is
the element of a sufficiently large closed contour
NECK
drawn around the linear defect, then the integral
Local narrowing in a sample appearing under
and so on. A needle-shaped defect can be exem- where F {. . .} is a certain functional depending on
plified by chains composed of vacancies or im- the type of domain wall, the nature of the induced
purity atoms. The stability of such chains is low anisotropy, etc. The third term in the Néel equa-
as they easily break up into point defects under tion defines the magnetic field pressure on the
atomic diffusion. Also classified as needle-shaped wall. Thus, the Néel equation describes the do-
defects are thin needle-like precipitates of phase main wall equilibrium condition, which is the con-
components in alloys. dition for zero total pressure affecting the wall.
The Néel equation applies when the intrinsic re-
NÉEL EQUATION (L. Néel, 1952) laxation of the magnetic vector to equilibrium pro-
A phenomenological equation describing the ceeds much faster than the relaxation of the in-
displacements of a magnetic domain wall (its duced anisotropy, and when the inertia (mass) of
quasi-stationary dynamics) in a magnetic sub- the wall can be neglected (the latter is permissi-
stance during its magnetic reversal in the presence ble at low wall speeds). The Néel equation is used
of magnetic viscosity effects: to describe the dynamics of domain walls in ferro-
P (u, t) + R(u) + (M 1 − M 2 )B(t) = 0, (1) electrics, which involves making the replacements
M → P , B → E in Eq. (1), where P is the polar-
where B(t) is the external magnetic field; M 1 and ization vector and E is the electric field.
M 2 are magnetization vectors in adjacent domains
separated by the domain wall between them; u is NÉEL POINT (L. Landau, 1933)
the displacement of the wall; R(u) is the pressure Temperature at which magnetic ordering ap-
on the wall produced by defects and imperfec- pears in antiferromagnets and ferrimagnets. They
tions in the material plus that from demagnetiza- lose their specific magnetic properties above that
tion e lds; and P (u, t) is the pressure on the wall point and behave as paramagnets whose suscepti-
due to induced magnetic anisotropy and magnetic bilities follow the Curie–Weiss law. A maximum
viscosity effects. In contrast to R(u), the pres- of the magnetic susceptibility, as well as anom-
sure P (u, t) is explicitly time dependent. More- alies in certain other quantities (such as conduc-
over, beside its dependence on the displacement u tivity, sound absorption factor, etc.), are observed
at the instant t , P (u, t) also depends on the ear-
in the vicinity of the Néel point. See also Curie
lier displacements at t < t . To account for this,
point, Magnetic phase transitions.
P (u, t) is expressed as
t NÉEL WALL
P (u, t) = F u(t), u(t ), t, t dt , A type of magnetic domain wall in a thin mag-
−∞ netic lm, in which its magnetization rotates in a
Schematic diagram of Bloch wall (a) and Néel wall (b), where δ is the domain wall width. Arrows A and B show
the magnetization M directions in central planes parallel to xz in the Bloch wall and parallel to xy in the Néel wall,
respectively; crosses indicate the magnetostatic poles related to the exit of M to the film surface and domain wall surface,
respectively.
874 NEGATIVE CRYSTALS
plane parallel to the film surface. Physically, the the absolute value of the luminescence radiation
reason for the formation of a Néel wall, instead of power attains its ultimate magnitude which equals
a Bloch domain wall, is a decrease of magnetosta- the power of the background radiation absorbed
tic energy. Indeed, consider a domain wall sepa- by the system. Negative luminescence in systems
rating two adjacent domains and let the magneti- with a discrete energy spectrum is usually ex-
zation in it change in a plane parallel to the surface cited by pumping. In the fundamental absorption
of the wall itself (xz plane in Fig. (a)). Magnetic band of semiconductors, negative luminescence
charges will then appear on opposite surfaces of arises through the depletion of electrons and holes.
the film in bands of width δ. An additional mag- The magnetic concentration effect, and other meth-
netostatic energy of the crystal is associated with ods are used to determine this. Negative lumines-
these charges, and it grows while the film thick- cence in semiconductors serves as the foundation
ness, d, decreases. For d < δ that energy may be for a new class of optoelectronic devices, such
lowered if the magnetization vector M rotates in as high-temperature sources of IR radiation, op-
a plane parallel to the film surface (xy plane in tical temperature-sensitive elements, etc. (see Op-
Fig. (b)). Thus, there is a critical film thickness d0 toelectronics).
below which the Néel wall is energetically more
favorable (see Fig. (b)); a Bloch wall should exist NEGATIVE MAGNETORESISTANCE
above d0 . The value of d0 depends on the nature A decrease of conductor resistance in a mag-
of the film, and may be as large as hundreds of netic eld. The effect is observable in the low-
angstroms. temperature range for a wide class of disordered
solids, and manifests itself most clearly in systems
NEGATIVE CRYSTALS
with reduced dimensionality (in metal lm s and
A type of three-dimensional defect in crystals,
wires, two-dimensional electrons of heterostruc-
which represents polyhedral cavities with a sym-
tures, etc.). The explication of this phenomenon
metry corresponding to the microscopic crystal
involves some difficulties, since both the classi-
symmetry. Such cavities nucleate during crystal
cal theory of magnetoresistance (which allows for
growth (in particular, in nature) or at the dissolu-
the electron trajectory bending in a magnetic field)
tion of crystals. During the course of growth neg-
and its treatment under hopping conductivity (due
ative crystals can capture ambient material, which
to the magnetic field contracting the wave func-
results in the formation of solid, liquid, or gaseous
inclusions. tions of localized states) predict an increase of
resistance in the magnetic field. The mechanism
NEGATIVE LUMINESCENCE suggested in the 1960s is related to scattering by
The decrease of the radiation by a system com- magnetic impurities (when the magnetic field sup-
pared to its equilibrium thermal radiation in the presses the process of electron spin- ip scattering,
same spectral region. Negative luminescence re- see Kondo effect) and proves to be inapplicable to
sults from the decrease in population of a series most of the systems under consideration, where
of excited energy levels of the system, and the the concentration of such impurities is small. In
corresponding decrease of the number of radiative the early 1980s a theory was put forward to relate
transitions at particular frequencies. In this situ- the negative magnetoresistance to suppression of
ation the absorption of background thermal radi- the interference quantum correction to the conduc-
ation by the system predominates over its spon- tivity of a disordered system (see Weak localiza-
taneous radiation, and the resultant radiation en- tion). This approach explains all specific qualita-
ergy flux in the indicated frequency range comes tive features of negative magnetoresistance: (i) the
from outside to the system surface. The lumines- appearance of the effect in weak fields and its
cence power, defined as the difference between square-root (or logarithmic in the two-dimensional
the powers of spontaneous radiation and of ab- case) dependence on the magnetic field strength;
sorption, proves to be negative; which is the rea- (ii) its independence of the angle of the current to
son for the term “negative luminescence”. Un- the field in the three-dimensional case (while in
der total suppression of spontaneous radiation, the two-dimensional case the effect is determined
NEMATIC LIQUID CRYSTAL 875
only by the magnetic field component normal to field strength Ey (see Galvanomagnetic effects) is
the layer); (iii) its strong temperature dependence greater than the strength Ex of the electric field
caused by suppression of interference during elas- producing the current flowing through the sample
tic scattering (on account of the electron–phonon along the positive x-axis (see Fig.). The injected
and electron–electron processes). A quantitative holes enter the sample and drift in the field Ey ; the
description of the experiments calls for, as a rule, Lorentz force shifts them along the OX axis in the
taking into account the factors that produce a pos- direction opposite to that of the force eEx acting
itive contribution to magnetoresistance (spin-flip on them, and this produces the negative photocon-
scattering, quantum corrections due to electron in- ductivity.
teraction, and so on). See also Galvanomagnetic
effects. NEGATIVE RESISTANCE
Materials consisting of alternating nanometer- Electric resistance with a negative sign.
thick layers of a ferromagnetic material and a non- (a) For alternating current, it means a nega-
magnetic metal experience a very large decrease in tive magnitude of the real part of the complex im-
resistance in a magnetic field called giant magne- pedance Z(ω) = R(ω) + iX(ω). Under harmonic
toresistance. Examples of such materials are alter- action with the frequency ω across a subcircuit
nating layers of iron and chromium, and multilay- containing both resistive and reactive impedance,
ers of cobalt and copper. Even larger changes in a phase shift δ appears between the current and
magnetoresistance called colossal magnetoresis- voltage values, which depends on the circuit para-
tance have been found in the perovskite-like ma- meters and the frequency. In this case, R(ω) (in-
terial LaMnO3 which has manganese in the Mn3+ volving the current amplitude I and voltage V in
valence state. Replacing some of the La3+ ions by the relation I R(ω) = V cos δ) can be negative over
divalent Ca, Ba, Sr, Pb or Cd ions produces the a certain frequency range. Then, dynamic nega-
mixed-valence Mn3+ /Mn4+ system which has tive resistance is spoken about. Across subcircuits
many mobile charge carriers. The particular com- that have dynamic negative resistance, the release
bination La0.67 Ca0.33 MnOx displays more than a rather than the absorption of energy takes place.
thousand-fold change in resistance in a 6 T mag- (b) In the static regime, there is a negative dif-
netic field. ferential resistance R = dV /dI that can be nega-
tive over a limited part of the current–voltage char-
acteristic. The appearance of the negative value
NEGATIVE PHOTOCONDUCTIVITY in a
can be due to the dependence of the mobility of
magnetic eld
current carriers or their concentration on the elec-
Weakening of photoconductivity, which takes
tric field. The negative differential resistance leads
place under injection of charge carriers into the
to instability of the electric circuit, and its sponta-
central part of a long semiconductor sample placed
neous transition to another state, or to a regime of
in a strong magnetic eld B. For instance, it
oscillations.
may be an n-type semiconductor where the Hall
The negative resistance characteristic is uti-
lized for the operation of various semiconductor
devices; see Gunn diode, Generation type semi-
conductor diodes, Barrier injection transit time
diode, Microwave semiconductor devices.
molecular compounds in water or other solvents 4f 4 5d 0 6s 2 . Ionization energy is 5.51 eV. Atomic
(lyotropic nematic liquid crystal). The structural radius is 0.1816 nm, Nd3+ ion radius is 0.102 nm.
units of the latter are micelles comprising hun- Oxidation state is +3. Electronegativity is ≈1.12.
dreds of molecules, instead of individual mole- In free form, neodymium is a silvery-gray
cules (see Lyotropic liquid crystals). Nematic liq- metal. It exists in two modifications (α-Nd and
uid crystals are also formed by liquid-crystal poly- β-Nd). α-Nd is stable below ≈1133 K with a
mers, in which the preferred direction is specified double close-packed lattice (alternation of hexag-
by ordering either the links of the basic chain or onal atomic layers to ABAC. . .), space group is
its side branches. Orientational ordering may per- P 63 /mmc (D6h 4 ); a = 0.36582 nm, c = 1.17966
tain to either all three principal molecular axes, to nm under normal conditions. Stable above 1128 K
two of them (biaxial nematic liquid crystal), or to and up to Tmelting = 1291 K is β-Nd with body-
only one of them (uniaxial nematic liquid crystal). centered cubic lattice, space group is I m3̄m (Oh9 );
The biaxial type is rarely found. Among the uni- a = 0.413 nm (extrapolated value) after harden-
axial types are calamitics and discotics, with struc-
ing. There are contradictory data that the transition
tural elements having the shape of rods and discs,
α-Nd → β-Nd is preceded by its transformation
respectively. Orientational ordering of molecules
into a (possibly, metastable) allotropic modifica-
results in a strong anisotropy of the physical prop-
tion with face-centered cubic lattice having condi-
erties of nematics. They are birefringent, usu-
tional unit cell constant a = 0.480 nm after hard-
ally uniaxial media (see Order parameters in liq-
ening, space group F d 3̄m (Oh5 ). The density is
uid crystals). Elastic properties of nematic liquid
crystals are defined by their orientational strains, 7.010 g/cm3 at 293 K. Tboiling ≈ 3340 K. Bind-
which occur when a director (prescribing the pre- ing energy of neodymium is −3.35 eV/atom at
dominant orientation direction of the molecules) is 0 K. Heat of melting heat is 7.16–10.9 kJ/mole;
locally forced out of its equilibrium position (see heat of sublimation is 288.9 kJ/mole, heat of
Elasticity of liquid crystals). evaporation is 296 kJ/mole; Debye temperature
Combining a strong anisotropy of their optical is ≈152 K. The linear heat expansion coeffi-
and electromagnetic properties with a capability cient of neodymium polycrystal is 6.7·10−6 K−1
to easily reorient their director when affected by (at 298 K), coefficient of thermal conductivity is
weak external fields, nematic liquid crystals en- 12.98 W·m−1 ·K−1 (at 299–303 K). Isothermal
joy various uses in systems of recording, process- bulk modulus is 32.6 GPa (at room temperature
ing, and presentation of optical data (see Fred- and zero external pressure); Young’s modulus is
eriks transition, Twist effect, Guest–host effect, 37.9 GPa (at room temperature); ultimate tensile
Electrooptical effects in liquid crystals). Nematics resistance is 0.171 GPa (in a cast sample at 293 K);
owe their name to disclinations – linear defects in relative elongation is 11%. Vickers hardness (at
the director field. Microscopic observations show 293 K) is 35 for cast neodymium and 76 HV for
these to be flexible threads found both within the forged Nd. High purity neodymium has high cold
bulk and at the surface. deformability (up to 70% reduction) without fail-
ure. Application of intermediate annealing allows
NEODYMIUM, Nd obtaining thin strips (0.5 mm thick and less). Ef-
A chemical element of Group III of the peri- fective thermal neutron trapping cross-section is
odic system with atomic number 60 and atomic 44.6 barn. Electric resistivity of α-Nd polycrystal
mass 144.24; it belongs to lanthanides. Natural is 643 n·m (at 298 K), temperature coefficient
neodymium consists of isotopes 142 (27.11%), of electric resistance is 0.00164 K−1 . Work func-
143 (12.17%), 144 (23.85%), 145 (8.30%), 146 tion of neodymium polycrystal is 3.3 eV. Magnetic
(17.22%), 148 (5.73%), and 150 (5.62%), the iso- first-order phase transition of the order–disorder
tope 144 Nd is weakly radioactive with half-life type occurs in neodymium at the Néel point
of about 5·1015 years, and 17 additional radioac- TN ≈ 19.5 K: the magnetic moments of atoms
tive isotopes are known with mass numbers from in neighboring hexagonal layers become antipar-
129 to 154. Outer electronic shell configuration is allel. At the temperature 7.5 K, the ordering of
NETWORK, PLANATOMIC 877
the magnetic moments of atoms with cubic near- 9.6 kJ/mole, heat of evaporation is 394.8 kJ/mole.
est neighbor symmetry takes place (with a change Debye temperature of α-Np is 188 K. For neptu-
in the effective number of atoms with ordered nium of 99.97% purity the temperatures of poly-
magnetic moments). The antiferromagnetic struc- morphic phase transitions: α → β at 555 K, β →
ture of hexagonal layers with cubic environmen- α at 542 K, β → γ at 832 K, Tmelting = 870 K
tal symmetry in α-Nd is similar to the structure of (start); the linear thermal expansion coefficient
layers with hexagonal nearest neighbor symmetry, grows from 11.1·10−6 K−1 (at 100 K) up to
but with magnetic moment directions in the for- 32.7·10−6 K−1 (at 525 K) for α-Np, and from
mer turned by 30◦ relative to those of the latter. 42.8·10−6 K−1 (at 550 K) up to 78.1·10−6 K−1
The magnitudes of magnetic moments of atoms in (at 850 K) for β-Np. Calculated value of isother-
the layers of both kinds vary as the sine. Magnetic mal bulk modulus is 68 GPa (at room tempera-
susceptibility of neodymium is 5600·10−6 CGS ture). Electric resistivity of α-Np polycrystal is
units; nuclear magnetic moment of isotope 143 Nd 35.5 n·m at 4.2 K and 1176.5 n·m at 295 K.
is 1.25 nuclear magnetons. There are suggestions The 5f -electrons in neptunium are delocalized
that the nonequilibrium face-centered cubic mod- (partially full 5f -shell is somewhat expanded).
ification transforms into the ferromagnetic state at Accordingly, neptunium may have a tendency to
the Curie point TC = 29 K. Neodymium is used superconduct for Tc < 0.4 K, and Pauli paramag-
as an alloying addition in light alloys (e.g., mag-
netism is present in it. On the other hand, as
nesium alloys), as a component for mischmetal, in
the density of states of 5f -electrons at the Fermi
laser materials, etc.
level is not high, and the criterion of band ferro-
magnetism is not met, neptunium is an exchange-
NEPTUNIUM, Np
coupled band paramagnet with magnetic suscepti-
An actinide chemical element with atomic
bility about 5·104 cm3 /mole. The longest-living
number 93 and atomic mass 237.048. There are no
stable isotopes, and 14 known radioactive isotopes isotope 237 Np (α-radiator) with lifetime T1/2
with mass numbers from 228 to 241. about 2.14·106 years is obtained in nuclear reac-
Outer shell electron configuration is 5f 4 6d 1 tors. It finds application in the production of the
7s 2 . Atomic radius is ≈0.1525 nm; radius of nuclide 238 Pu, which is used in space exploration,
Np3+ ion is ≈0.0998 nm, radius of Np4+ is nuclear micropower engineering, medicine etc.
≈0.093 nm, radius of Np7+ is 0.071 nm. Oxida-
tion state is +5, more rarely +2, +3, +4, +6, +7. NERNST–ETTINGSHAUSEN EFFECT
Electronegativity is ≈1.25. See Thermomagnetic phenomena.
In the free form neptunium is a heavy silvery
metal. It exists in α-, β-, γ -modifications. Sta-
NERNST GLOWER (W. Nernst, 1900)
ble at temperatures below ≈550 K is α-Np with
an orthorhombic crystal lattice (8 atoms in unit A rod made from a mixture of refractory metal
cell; space group P nma, D2h 16 ). At 293 K, a = oxides (mainly zirconium with magnesium) heated
0.4722 nm, b = 0.4887 nm, c = 0.6662 nm. Sta- by an electric current (T ∼ 1500 ◦ C) that can be
ble above 542 K is β-Np with tetragonal crystal used as an IR range radiation source. However it
lattice (4 atoms in unit cell, space group P 421 2, is rarely used for this purpose, since it is only effi-
D42 ). At 586 K, a = 0.4896 nm, c = 0.3387 nm. cient in the short-wave band (λ < 10 μm). More-
At temperature ≈830 K there occurs a polymor- over, because of its poor electrical conductivity at
phic transformation of β-Np to γ -Np, stable up room temperature it needs preliminary heating to
to Tmelting = 910 K. γ -Np has a body-centered ∼800 ◦ C. The latter circumstance is related to the
cubic lattice (space group I m3̄m, Oh9 ); at 873 K, ionic nature of the electrical transport in ZrO2 ,
which is a solid electrolyte.
a = 0.3525 nm. The density is 20.45 g/cm3 for α-
Np, 19.36 g/cm3 for β-Np, and 18.0 g/cm3 for γ -
Np; Tboiling = 4174 K. Bonding energy of neptu- NETWORK, PLANATOMIC
nium is −4.55 eV/atom at 0 K. Heat of melting is See Planatomic networks.
878 NETWORK, RANDOM
results in nuclear doping of the solid. Such nu- of the electrons that generate the magnetic mo-
clear reactions are also used in neutron activation ment (mainly d and f unpaired electrons). The
analysis. Thermal neutrons may be scattered by magnetic scattering form factor, which depends on
the atoms of solids, a phenomenon which under- the scattering vector (angle), describes such scat-
lies the neutron diffractometry techniques. tering.
In contrast to X-rays and electrons, neutrons Neutron scattering may be classified accord-
are scattered by the atomic nuclei. The basic char- ing to several criteria: (a) the change in the en-
acteristics of interactions of neutrons with atomic ergy of scattered neutrons (elastic versus inelas-
nuclei are the effective scattering amplitude and tic); (b) the degree of coherence of scattered neu-
the cross-section. Both parameters have been thor- tron waves (coherent versus incoherent). Coherent
oughly studied and tabulated. Neutron scattering elastic scattering is structure-sensitive, and takes
by nuclei may be of either a potential or a reso- place when the momentum of the neutron is trans-
nance nature. These two types of scattering govern ferred to the crystal solid as a whole (scatter-
the amplitude, f , of the scattering of slow (ther- ing vector equals site vector of reciprocal lattice),
mal) neutrons which is given by the Breit–Wigner resulting in the thermal neutron diffraction. This
formula (G. Breit, E. Wigner, 1936): phenomenon underlies the techniques of structural
Γn (2k) neutron diffractometry, and magnetic neutron dif-
f = fn (k) − . fractometry. If, during the process of scattering,
(E − E0 ) + iΓ /2
the neutron exchanges energy with one of the
The first term characterizes potential scattering
atoms, generating or absorbing either a phonon
due to the direct interaction of neutrons with the
or a magnon in the process, there takes place co-
force field of the nucleus. The second term de-
herent inelastic neutron scattering. In that case the
scribes resonance scattering that involves an inter-
scattering vector differs from that of the site of the
mediate excited nucleus formed during the course
reciprocal lattice by the wave vector of the phonon
of the nucleus–neutron interaction. Here fn is the
or the magnon. Neutron spectroscopy techniques
amplitude of the potential scattering, k is the wave
vector of the incident neutron, E is its energy, are based on this effect. Techniques for study-
E0 is the resonance energy, Γ is the width of the ing the dynamics of atom groupings in solids and
excited nuclear energy level (proportional to the fluids, which all use inelastic incoherent scatter-
reciprocal of the excited state lifetime); Γn is the ing of neutrons, also belong to the same category.
width of the resonance energy level (proportional Incoherent elastic neutron scattering takes place
to the probability of composite nucleus decay dur- in atomically and magnetically ordered systems,
ing elastic neutron scattering). The amplitude of and is used correspondingly rarely. Beside the nu-
potential scattering is always positive, and propor- clear and the magnetic interactions, there are other
tional to A1/3 (A is the mass number of the scat- weaker (by 4 orders of magnitude) interactions be-
tering nucleus). For resonance scattering, the value tween neutrons and atomic charges (J. Schwinger,
of f depends on the incident neutron energy, and L.L. Foldy).
on the position and half-width of the resonance To conduct experiments on the interactions of
level of the nucleus. These two inputs combine to neutrons with solids, atomic reactors and pulsed
produce different amplitudes for the scattering of sources are used. To slow down fast neutrons pro-
thermal neutrons by different isotopes of the same duced during fission reactions, slowing materi-
element, so that it may be either positive or nega- als (moderators) are used. Upon exiting the reac-
tive, real or imaginary. tor, the white (broad wavelength range) neutron
A neutron has a magnetic moment and inter- radiation is monochromated by reflection from
acts with the electronic magnetic moment of the a monocrystal (constant wavelength technique).
atom. The atom can also have a nuclear magnetic Modern sources produce monochromatic neutron
moment, but its much lower value makes it dif- beams with intensities up to 106 –107 N·cm−2 s−1 .
ficult to observe by neutron scattering. The scat- Fast neutron pulse reactors generate bursts of
tering of neutrons by atomic electron shells de- white neutron radiation (approximately 100 μs
pends on both the spatial and energy distributions long). Electron and proton accelerators are also
880 NEUTRON SCATTERING, INELASTIC
used to produce neutron fluxes by generating brak- and acoustic phonon spectra over the entire range
ing γ -quanta (Bremsstrahlung) that strike a tar- of momenta (wavelengths).
get containing heavy nuclei, and release neutrons
through photonuclear transformations. The result- NICKEL ALLOYS
ing neutron bursts vary within 1–5 μs in duration. Alloys of nickel, alloyed by one or several ele-
When operating pulsed neutron sources, the time- ments. Nickel forms substitutional solid solutions
of-flight method is employed, where the spectral with Fe, Cr, Cu, Mn, Co, Mo, W and other metallic
expansion along the flight path is carried out using elements over a wide temperature range. The solu-
a time analyzer and fixing the angle of scattering tions have a face-centered cubic lattice imparting
(or prescribing a set of discrete angles). them with some plasticity. The alloys harden be-
cause of a decrease of the energy of stacking faults,
NEUTRON SCATTERING, INELASTIC reduction of the diffusion mobility of atoms, and
See Inelastic neutron scattering. changes of electronic structure in the course of al-
loying. Structural nickel alloys (except for Monel
NEUTRON SPECTROSCOPY metal) are low concentration alloys known for
A set of techniques for studying atomic and their high emission properties, enhanced mechani-
molecular vibrations, as well as magnetic excita- cal strength and corrosion resistance. They are fa-
tions (magnons) in condensed matter, which are vorable for all types of machining, even in the cold
based on inelastic scattering of neutrons (see Neu- state. Monel metal (Ni with 1.5% Mn, 2.5% Fe,
trons in solid state physics). An interaction of a 28% Cu) is characterized by high grade mechan-
neutron beam with an atom vibrating at a crystal ical properties and corrosion strength in water,
lattice site may be treated formally as a collision acids, strong alkali. In air it retains its strength and
of a neutron with a phonon, so the conservation plasticity at very low temperatures and upon heat-
laws of energy and momentum may be applied to ing up to 400 ◦ C. Thermal-electrode and (ohmic)
that process. Thus one may describe the change resistance alloys which are classified as electri-
in energy and momentum of neutrons during the cal engineering nickel alloys possess high elec-
collision and plot a system of dispersion curves tromotive force, electrical resistance and thermal-
for monocrystal samples, or retrieve the density of environment resistance. Chromium–nickel alloy
states of the phonons for polycrystals. (Ni with 9.5% Cr, called chromel) and alumel
Experimental studies of the spectrum of in- (Ni with 1.2% Si, 2.3% Mn, 1.8% Al) are thermal-
elastic neutron scattering involve measuring the electrode alloys, used for manufacturing thermo-
change in neutron momentum (wavelength) as a couples for measuring temperatures up to 1000 ◦ C
function of sample orientation with respect to the (in long term) and up to 1300 ◦ C (in short term).
neutron beam, and the angle of scattering. There Nichrome, ferronichrome and also alloys, addi-
are many schemes available for such measure- tionally alloyed with Al, Cu, Mn, Mo, etc., have
ments involving both stationary and pulsed neu- high ohmic resistance; and are used for manu-
tron sources. The most popular among them em- facturing resistors, thermoresistors, tensoresistors
ploys a three-crystal spectrometer, a monochro- and electrical heaters. Nickel alloys with specific
matic incident beam, and a crystal analyzer posi- physical properties include soft magnetic materials
tioned behind the sample. By tuning the analyzer (permalloy) and corrosion-resistant alloys, e.g.,
to different wavelengths in the scattered emission nickel–molybdenum alloys (27% Mo, 1.5% V),
one measures the so-called phonon reflections, and nickel–chromium–molybdenum alloys (15%
their maxima relating to points on the dispersion Cr, 16% Mo, 4% W). They also include alloys
curve, while their half-widths relate to lifetimes with a preset temperature coefficient of linear ex-
of the phonons. Magnon spectra may be similarly pansion (invar), alloys with high elasticity prop-
studied. Unlike optical spectroscopy, which only erties over a wide temperature range (e.g., 10%
yields information on the long-wave part of the Cr, 10% Mo, 5% W, 4.5% Al), alloys for thermal-
spectra of optical phonons in solids, neutron spec- bimetallic parts. Some elements, such as Al, Ti,
troscopy makes it possible to study both optical Nb, Ta, Be, Hf, etc. have limited solubility in Ni.
NIOBATES 881
Complex nickel alloys containing these elements electrical resistance 292·10−5 K−1 . Hall constant
are hardened during the decay of oversaturated is 36·10−11 m3 /C. Polycrystal work function is
solid solutions due to phase precipitation and to 4.61 eV.
the associated coherent stresses that are caused by Nickel is ferromagnetic with the Curie temper-
the mismatch of the matrix and hardening phase ature 631 K. Atomic magnetic moment is 0.604
lattices. Besides, the antiphase boundaries and Bohr magnetons, nuclear magnetic moment of
stacking faults of the hardening phase somewhat 61 Ni is 0.746 nuclear magnetons. Total effec-
assist hardening. Hardened-matrix nickel alloys tive high-energy neutron adsorption and scattering
containing precipitations of intermetallic (see In- cross-section is 3.2 barn (3 to 10 meV), for low-
termetallic compounds), carbide, boride, and other energy neutrons it is 22.1 barn (0.25 eV).
phases, are useful as high-temperature materials.
Nickel is mainly used for the production of al-
Nickel alloys find their application in different en-
loys (with iron, chromium, copper and other met-
gineering areas such as structural and electrical en-
als), distinguished by valuable mechanical, anti-
gineering alloys with specific physical and chem-
corrosion, magnetic and electrical properties, by
ical properties, and as thermal-environment resis-
tant materials.
high temperature strength and heat resistance (see
Nickel alloys).
NICKEL, Ni
Chemical element of Group VIII of the peri- NIOBATES
odic table with atomic number 28 and atomic mass Metallic oxide compounds of niobium, includ-
58.69. Natural nickel contains stable isotopes 58 Ni ing the NbO3 group with metals, e.g., LiNbO3 ,
(67.89%), 60 Ni (26.16%), 61 Ni (1.13%), 62 Ni KNbO3 , Ba2 Na(NbO3 )5 (“banana”), Srx Ba1−x
(3.66%), 64 Ni (1.16%). Outer shell electron con- (NbO3 )2 (“bastron”), etc. They possess nonlin-
figuration is 3d 8 4s 2 . Successive ionization ener- ear optical and acoustical properties. They are
gies are 7.6, 18.15, 35.16 eV. Atomic radius is obtained by the synthetic Czochralski method
0.125 nm; Ni2+ ion radius is 0.072 nm. Oxidation (see Synthetic monocrystals). Uniaxial trigonal
state is +2, more rarely +1, +3, +4. Electroneg-
monocrystals of lithium niobate are ferroelectrics
ativity is 1.8.
which serve as a piezoelectric material in acousto-
In free form nickel is a silvery-white metal.
electronics devices for surface acoustic waves, and
It has face-centered cubic lattice, space group
as an electrooptic and optically nonlinear mate-
F m3̄m (Oh5 ), a = 0.35243 nm (at 298 K). Den-
rial used in quantum electronics in electrooptical
sity is 8.897 g/cm3 ; Tmelting = 1728 K; Tboiling ∼ modulators, light gates, second harmonic genera-
3100 K. Heat of melting is 16.9 kJ/mole, heat
tors of neodymium laser radiation (see Nonlinear-
of evaporation is 370 kJ/mole. Specific heat is
optical crystals), and in near infrared optical para-
445 kJ·kg−1 ·K−1 (at 298 K). Linear thermal ex-
metric oscillators. They possess photorefractive
pansion coefficient is 13.1·10−6 K−1 (at 293 to
properties (see Photorefraction); they exhibit the
373 K). Thermal conductivity is 92.0
photochromic effect and the bulk photovoltaic ef-
W·m−1 ·K−1 . Debye temperature is 465 K. Adi-
fect; they are used in optical holographic stor-
abatic elastic moduli of crystal: c11 ≈ 248, c12 ≈
163, c44 ≈ 113 GPa at 298 K. Young’s modu- age devices with high capacity and short ac-
lus is ≈224 GPa (at 298 K), shear modulus is cess time (see Optical techniques of information
≈86 GPa (at 298 K). Poisson ratio is ≈0.30. recording). Tetragonal potassium niobate with the
Tensile strength is ≈450 MPa, elastic limit is perovskite structure is an analog of barium ti-
80 MPa, yield point is 120 MPa, relative elon- tanate, but with Tc = 708 K instead of 403 K.
gation is 40%. Brinell hardness of cast nickel Monocrystals with some application in nonlin-
is ≈700 MPa (at 293 K). Self-diffusion coeffi- ear optics and acousto-electronics are rhombic
cient is 1.749·10−21 m2 /s (at 700 K). Coefficient barium–sodium niobate which requires complex
of linear term in low-temperature molecular spe- mechanical untwinning, as well as lead–potassium
cific heat is 7.02 mJ·mole−1 ·K−2 . Electric re- niobate and some of their analogs, crystallizing in
sistivity 7·10−8 μ·m, temperature coefficient of the potassium–tungsten bronze structural type.
882 NIOBIUM
structure occur at the formation of the solid solu- wear resistant parts of instruments (e.g., compass
tions and chemical compounds, and at phase tran- needles) are made of Os and Ir alloys. The alloy
sitions. As a result, some properties of compo- of Pt (90%) with Ir was used to manufacture the
nent noble metals may improve and new prop- standards of the meter and kilogram.
erties may appear. For instance, the alloys of Pd
and Ir in the middle part of their phase diagram NOISE in semiconductors
have the coloration from light yellow to golden Fluctuations of currents I (or voltages) in semi-
and then to pink-violet. Some alloys of paramag- conductors which are due to the random nature of
nets Pd and Rh are ferromagnets (see Ferromag- current carrier motion, fluctuations of carrier num-
netism). The alloys Au–Cu–Ag are stronger than ber, and also the discrete character of the electrical
each individual component (solid-solution harden- conductivity process. Because of their frequent col-
ing), while the decomposition of the solid solu- lisions with phonons and lattice defects, electrons
tions may result in a severalfold increase of their and holes move like Brownian particles. Their ve-
yield limit. The use of noble metal alloys is con- locity fluctuations result in inducing electromotive
ditioned by their unique physical and chemical forces associated with thermal noise. Often, in ad-
properties: high stability to oxidation, corrosion re- dition, shot noise occurs in semiconductors. For
sistance, high-temperature strength and thermal- example, it is observed when current carriers sur-
environment resistance, beautiful color and oth- mount the barriers in semiconductor junctions. The
ers. Noble metal alloys are used in engineering individual barrier crossings occur discretely, inde-
(for electrical contacts, critical-duty resistors and pendently of each other, and lead to uncorrelated
potentiometers, for thermocouples operating in an current impulses if an external voltage is applied
oxidizing medium, for permanent magnets, for to the barrier.
the supports of axes of measuring devices, for their
Generation–recombination noise has a similar
elastic elements, etc.). These alloys are often more
character when the number of current carriers, and
efficient than pure noble metals as catalysts and
therefore, the crystal resistance δR fluctuates. In
construction materials.
this case there appears electromotive force noise
NOBLE METALS
δE = I δR. In many-valley semiconductors, inter-
A group of metals with especially low chemi- valley noise is observed. It results from the oc-
cal activity. casional electron transitions from one valley to
These are Ag and Au, and metals of the plat- another. The total number of current carriers is
inum group Ru, Rh, Pd, Os Ir and Pt that belong invariant, however, their effective mobility fluctu-
to Group VIII of the periodic table. Ag and Au ates, because the mobility of carriers differs in in-
are highly plastic; and the other noble metals are equivalent valleys.
notable by their refractory properties (T melting ≈ The resistance of semiconductors depends also
1800 ◦ C or higher). Most of the noble metals form on a number of additional physical quantities, the
solid solutions at mutual melting, e.g., Au and Ag factors which induce the fluctuation of δR, and
(face-centered cubic lattice), Os and Ru (hexag- the excitation of electromotive force noise. Often
onal lattice). Good electrical conductivity, corro- these are referred to as current noise, since in con-
sion resistance, high melting temperature and high trast to thermal noise, they exist only for I = 0.
reflectivity of noble metals and their alloys de- Flicker noise, telegraph noise (popcorn noise), and
termine their wide applications including various also noise due to the fluctuations of current car-
contacts, resistors with low thermal coefficient and riers or of the crystal lattice temperature, involve
thermoelectromotive force (paired with copper). this type of noise. These varieties of noise are
Coatings of Au 0.01–0.02 mm thick are applied to spatially nonuniform. Their evolution is described
the outside surface of spaceships and satellites to by a diffusion-type equation. Its solution and the
improve the reflection of the solar electromagnetic spectral density of noise depend on the boundary
radiation. High quality mirrors are made of Ag. conditions, and the system dimensions. Also there
Pure platinum and its alloys are used in thermome- are local carrier concentration fluctuations in the
try (resistance thermometers, thermocouples). The bulk semiconductor, as well as in the density of
884 NOISE IN ELECTRONIC DEVICES
the two-dimensional gas adsorbed by its surface, the conductor. When a conductor is in thermody-
and these have a diffusive character. The investiga- namic equilibrium with a thermostat at tempera-
tion of noise is useful for semiconductor diagnos- ture T , then T = Tn . Eq. (1) for Sω represents the
tics, for determining their parmeters, for designing Nyquist formula (H. Nyquist, 1928). Consider an
low-noise devices, and for selecting optimum op- ideal resistor with a resistance equal in value to
erating conditions. that of the matched load of a system. The noise
For different varieties of noise, see also Bark- temperature Tn is the temperature to which the
hausen noise, Generation–recombination noise, ideal resistor must be raised to generate the same
Telegraph noise, Flicker noise, Intervalley noise, noise as the load. At this temperature the spec-
Shot noise. tral density of the ideal resistor noise equals the
spectral density of the load resistor noise under
NOISE IN ELECTRONIC DEVICES study. This fact is often used for the measurement
Voltage (or current) uctuations in electronic of noise temperature. A departure of noise tem-
devices due to their internal structure and proper- perature from the ambient temperature T demon-
ties. There are also occasional disturbances from strates the nonequilibrium character of the fluctua-
external sources, referred to as interference. Noise tions. The value of Tn may be used when estimat-
arising in sensitive amplifiers or converters re- ing the extent of the warm-up of current carriers
stricts the attainment of maximum sensitivity. (see Electron heating).
Noise may also appear in devices with a high sig-
NONCENTRAL IONS, off-center ions
nal level, such as amplifiers or generators. Here
Substitutional impurity ions with their equilib-
the noise may show up as amplitude and phase
rium positions displaced off lattice sites. Such a
fluctuations of the main signal. The influence of displacement usually occurs along a crystal sym-
interferences, and also noise due to the instabil- metry direction, and is often accompanied by the
ity of device parameters, may be decreased by appearance of a dipole moment oriented along the
selective improvements. At the same time noise impurity atom displacement. Thermal or tunnel
such as the thermal type, shot noise (see Noise), jumps between different equivalent positions are
generation–recombination noise, which are due to possible, and there appears an inversion splitting
fundamental random-type processes, affect device between the vibrational levels of the impurity. The
operation. As a rule, it is impossible to diminish simplest possible example is the inversion splitting
the amplitude of these noises without changing in a double well potential (see Fig.). The value of
the device working characteristics. The search for such a splitting, Δ, usually lies within the energy
low-noise devices led to the choice of those based range corresponding to Δ = 20 K. Similar to dipo-
on the Josephson effect, on principles of quantum lar molecules such as OH− , noncentral ions may
electronics (masers), on parametric processes, etc. orient themselves in both external and internal
Such devices have a high sensitivity for measure- electric fields. The presence of noncentral ions sig-
ments of magnetic fields, voltages, currents. Also nificantly alters the properties of crystals (anom-
they can amplify extremely weak signals. See also alies arise in speci c heat, thermal conductivity,
Superconductor electronics.
NOISE TEMPERATURE
Physical quantity Tn , which is a measure of the
noise power in conductors (resistances). It is de-
fined by
Sω = 2kB Tn R(ω), (2)
where Sω is the spectral density of the electro-
motive force u ctuation in an electrical conductor,
kB is the Boltzmann constant, R(ω) is the resis- Inversion splitting, Δ, of the vibration levels in a double-
tive component of the differential impedance of well potential (±x0 are the positions of the minima).
NONCRYSTALLINE CLUSTERS 885
Among the noncrystalline clusters are, e.g., a regu- recording the measured parameter data either dur-
lar tetrahedron consisting of four atoms (Fig. (a)), ing direct contact of the sensor element with the
an icosahedron containing 12 atoms (Fig. (b)), sample (contact nondestructive testing technique),
a dodecahedron containing 20 atoms (Fig. (c)), or while there is no such contact (noncontact non-
a Mackay icosahedron (A.L. Mackay, 1952), con- destructive testing technique). The capabilities of
taining 55 atoms, a Buckyball or fullerene C60 these techniques depend on their sensitivity, reso-
containing 60 carbon atoms, and many others. lution, adequacy of data, and reliability of instru-
mentation and procedures.
NONDESTRUCTIVE TESTING TECHNIQUES Nondestructive testing types can be classed
Techniques for testing metals, as well as semi- as to the methods (algorithms) used to calibrate
finished and finished items, for their compliance the monitored parameter against a prescribed ref-
with established specifications to be sure they are erence; physical phenomena or laws underlying
suitable to be used for their designated purposes, these techniques; the kind of sensors and trans-
without harming them. The nondestructive testing ducers used; the type of interaction between the
methods that are often used are as follows: monitoring device and the sample; the methods
(1) a magnetic technique based on analyzing a used to recover the values monitored; the kind
magnetic field interaction with the sample, of instrumentation used; the information parame-
(2) an electric technique based on recording para- ter of the measured data signal; the properties
meters of an electric field interacting with the of the sample used for the purpose. Nondestruc-
sample, or such a field generated in it under tive testing also includes defectoscopy, structur-
the effect of some external factor, oscopy and structurometry, thickness gauging, en-
doscopy, etc. Nondestructive testing is used in
(3) an eddy current technique based on analyz-
ing the interaction of the electromagnetic field various industrial technological processes and re-
produced by an eddy current transformer with search practice.
that of eddy currents induced in the sample,
NONEQUILIBRIUM SUPERCONDUCTIVITY
(4) a radio wave technique based on recording the
A special state of a superconductor developed
changes in the parameters of radio band elec-
under the action of external fields on one of its sub-
tromagnetic waves interacting with the sam-
systems – either the electron, the phonon, or the
ple,
Cooper pair condensate. The structure and prop-
(5) a thermal technique based on recording
erties of nonequilibrium superconductivity are de-
changes in temperature fields of the sample,
termined by both the type of external action that is
(6) an optical technique based on recording the
applied, and the physical parameters and geometry
parameters of optical radiation interacting
of the superconductor. To induce this phenomenon
with the sample,
the following types of external action are com-
(7) a radiation technique based on recording and
monly used: ultrasonic or electromagnetic fields,
analyzing penetrating ionizing radiation after
tunnel injection or extraction of monoenergetic
it interacts with the sample,
quasi-particles, electric current running through
(8) an acoustic technique based on recording the
either the normal metal–superconductor interface,
parameters of elastic waves generated by or
or through thin superconducting threads. When the
originating in the sample, and
fields are weak enough one may apply the follow-
(9) a penetration technique based on the penetra-
ing kinetic equations for the nonequilibrium distri-
tion of particular substances into the defect
bution functions of the electrons fε and phonons
cavities of the sample.
Nω :
Both active and passive nondestructive testing ∂fε
techniques are used. In the former case the sample = Iep {ε, fε , Nω } + Iee {fε } + Iei {ε, fε }, (1)
∂t
is exposed to the action of an external (auxiliary)
∂Nω
energy source which produces the data reporting = Lpe {ω, Nω , fε } + Lpi ω, N(ω)
signal, and in the latter case the origin of this sig- ∂t
nal is the sample itself. These techniques involve + Lpp {ω, Nω }, (2)
NONEQUILIBRIUM SUPERCONDUCTIVITY 887
where Iep , Iee , Iei are electron–phonon, electron– the relaxation time of the charge imbalance has
electron, and electron–impurity collision integrals; the form τQ = 4kB Tc τE /(πΔ). For T → Tc , the
and Lpe , Lpp , Lpi are phonon–electron, phonon– quantity τQ diverges as (1 − T /Tc )−1/2 , thus
phonon, and phonon–impurity collision integrals, impeding the transfer of the charge from quasi-
respectively. This system of equations is subject to particles to the condensate.
the self-consistency condition for the nonequilib- The nonequilibrium distribution function for
rium energy gap, Δ, electrons f (Ek ) may be evaluated experimentally
1 − 2f (Ek ) using the S–I–S tunnel junctions and measuring the
I = VC , (3)
2Ek current–voltage characteristic I (V ):
k
∞
where Ek = (ξk2 + Δ2 )1/2 ; VC is the standard 2
electron–electron attraction potential. In a layer I (V ) = ρ(E)ρ(E + eV )
eR
of width ωD (ωD is the Debye frequency), the
latter has a constant positive value in the vicin-
× f1 (E) − f2 (E + eV ) dE, (4)
ity of the Fermi surface; ξk is the kinetic en-
ergy of an electron with momentum P referred to where R is the junction resistance, ρ(E) is the
the Fermi level. The parameter of the energy gap, density of states in the superconducting electrodes
which, according to Eq. (3), depends on the func- S of the junction. We have for an isotropic super-
tion f (Ek ), may increase as a result of the exter- conductor ρ(E) = |E|/(E 2 − Δ2 )1/2 . Usually a
nal action. The effect of broadening the energy gap double tunnel junction of the type S1 –I –S2 –I –S3
(as well as increasing other superconductor criti- is used, where one of them (e.g., S1 –I –S2 ) is
cal parameters) experienced under the action of ei- low-ohmic and is called the injector (or genera-
ther photons or phonons has been experimentally
tor), while the other one is a high-ohmic detector.
observed, and was named the Eliashberg effect.
When a constant bias voltage V is applied to the
When the external field is switched off, the energy
S1 –I –S2 junction, this results in an intense injec-
gap relaxes back to its equilibrium value within
tion of monoenergetic electrons (or holes) into the
the time τΔ = 1.2τE /[I − (T /Tc )]1/2 (here τE
superconducting electrode S2 resulting in a non-
is the time of the electron energy relaxation,
equilibrium change in their distribution function
and Tc is the critical temperature of superconduc-
f2 (E). The latter is then recorded at the S2 –I –S3
tors). For T → Tc , the quantity τΔ diverges as
(1 − T /Tc )−1/2 . When charged particles are in- junction by measuring its current–voltage charac-
jected into the superconductor (e.g., through the teristic. The explicit form of ρ(E) is determined
tunnel barrier I of a superconductor–insulator– using Eq. (4). In a similar way the electron distri-
superconductor tunnel junction), the energy gap bution function is determined when the supercon-
width does not change, only the phase of the or- ductor film is bombarded by photons. Additional
der parameter changes. The total charge of the su- information about nonequilibrium superconduc-
perconductor remains unchanged since the change tivity is obtained by studying the effects that result
in the charge of quasi-particles, Q∗ is compen- from the nonequilibrium properties of the super-
sated by a change in the charge of the condensate. conducting order parameter, including its relax-
Essentially, nonequilibrium superconductivity re- ation behaviour. Techniques used to evaluate or-
sults from an imbalance between the populations der parameter relaxation times include subjecting
of the electron and the hole branches of the spec- the superconductor to pulses of current, to injec-
trum, when the pairs of electron excitations re- tion of quasi-particles, and also to laser and ther-
combine into a superfluid Bose condensate, form- mal pulses. For example, one may exploit the fact
ing Cooper pairs and hole excitation pairs. The that the maximum value of the current Idc that
quantitative characteristic of such a charge im- may run through a superconducting sample with-
balance is given by the difference between the out bringing it back to its normal state depends on
chemical potentials of quasi-particles μn and the the angular frequency ω and the amplitude of the
condensate, μS , respectively. The expression for applied alternating current, Iac . Measurements of
888 NONERGODICITY
the temperature dependence of Idc made it possi- tions along its propagation path. The absorption of
ble to evaluate the relaxation time of the order pa- such a wave is much stronger than that of a linear
rameter. The value of τΔ thus found for aluminum wave (nonlinear sound absoption). Interactions
samples at temperatures close to Tc appeared to between several elastic plane waves (phonons) re-
be proportional to (I − T /Tc )−1/2 , which agrees sult, apart from the self-action effect of each indi-
well with theoretical models. vidual wave, in the appearance of waves at combi-
nation frequencies. If synchronism conditions are
NONERGODICITY satisfied in a solid (energy and quasi-momentum
Discrepancy between the results of averaging of phonons conserved), both longitudinal L and
various physical characteristics of a system over transverse T waves propagating at different veloc-
time and over the Gibbs distribution. Nonergodic-
ities interact with each other. Examples of non-
ity is typical of spin glasses and other amorphous
linear acoustic effects in solids are Raman scat-
magnetic materials; it is found in systems featur-
tering of sound by sound; generation of harmon-
ing continuous spectra of their relaxation times
ics of T waves (“forbidden” by nonlinear elasticity
when the duration of an experiment is shorter than
theory but appearing because of dislocations and
the longest relaxation time. A continuous spec-
trum of relaxation times is due to the infinite internal mechanical stresses); nonlinear polariza-
degeneracy of the ground state of a spin glass, tion effects; generation of higher harmonics of sur-
when averages must be calculated in two stages. face acoustic waves, etc. As far back as in 1937,
First, averaging is done over each of the degen- Landau and Rumer constructed a theory of high-
erate states, and then over their ensemble, taking frequency sound absorption (∼109 Hz) at low
account of the probability of each. Many phys- temperatures (∼20–70 K) in insulators satisfying
ical variables (e.g., magnetic susceptibility, spe- the condition kl
1 (k is the sound wave vector,
ci c heat) become non-self-averaging with their l is the mean free path of thermal phonons), based
fluctuations characterizing the degree of the sys- on nonlinear interactions of coherent phonons
tem nonergodicity. That is why, when allowing for (sound), and thermal phonons subject to synchro-
the nonergodicity of spin glasses, they are to be nism conditions. Akhiezer did the same in 1938
described by an infinite number of the order pa- for ultrasonic frequencies at room temperature
rameters that define the overlap of local frozen when the condition kl 1 is satisfied. The first
magnetic polarizations in various states. Noner- experiments on coherent phonons to prove the ex-
godicity results in a large number of irreversible istence of a three-phonon interaction, and in par-
phenomena in spin glasses, in some other amor- ticular the generation of harmonics in solids, were
phous magnetic materials with their lack of equi- reported in the early 1960s. In piezosemiconduc-
librium, and also in memory and aging phenom- tors (e.g., CdS), a strong electroelastic nonlin-
ena, etc., all of them experimentally observed.
earity may develop under certain conditions due
to the piezoelectric effect (see Piezoelectricity),
NONLINEAR ACOUSTIC EFFECTS
while magnetically ordered crystals (e.g., antifer-
Phenomena originating in an acoustic field
romagnets, haematite) may display a strong mag-
with high sonic amplitudes. Among these phe-
nomena are, for example, distortions of the sine netoelastic nonlinearity due to magnetostriction.
profile of longitudinal elastic waves, L, near their These “mixed” nonlinearities exceed the lattice
source in solids, and the generation of harmonics. (elastic) nonlinearity by 3–4 orders of magnitude,
This phenomenon (self-action of acoustic waves) and considerably improve the performance of var-
takes place due to the anharmonic vibrations of ious nonlinear devices that process radiosignals in
a crystal lattice, and resultant deviations from acousto-electronics and magnetoacoustics (corre-
Hooke’s law. If there is no sound velocity disper- lators, convolution devices, filters, solid-state ul-
sion and the sound absorption remains weak then trasonic delay lines). Such devices play a signif-
the nonlinearity, although small in itself, will re- icant role in developing new measurement tech-
sult in the accumulation of wave shape distor- niques in solid state physics.
NONLINEAR MAGNETOACOUSTIC RESONANCE 889
NONLINEAR ELASTICITY THEORY are elastic moduli of order 2 and 3. The defining
Part of elasticity theory treating large elas- equations for isotropic materials in the second or-
tic strains. Within nonlinear elasticity theory, the der of elasticity theory have the form:
strain of a medium is characterized by the displa-
cement vector of its elements, ui , and by the strain σik = λull δik + 2μuik
tensor,
λ 2 ν
+ hlm + 1 − λ u2ll + ν2 u2lm δik
2 2
1 ∂ui ∂u ∂u ∂u
uik = + k + l l , + 2(λ − μ + ν2 )ull uik + μhil hkl
2 ∂xk ∂xi ∂xi ∂xk
+ 4(μ + ν3 )uil ukl ,
while its state of stress is described by the stress
tensor, σik . The energy density of the elastic strain where hik = ∂ui /∂xk ; λ, μ, ν1 , ν2 , ν3 are the
W is considered to be a function of the compo- elastic moduli. Nonlinear elasticity theory is used
nents of the tensor uik (the free energy density to calculate the state of stress of highly deformed
or some other thermodynamic potential is used bodies (e.g., shells). It is also used in the physics of
for W at finite temperatures). By virtue of the solids to analyze the propagation of shock waves
requirement of invariance, W depends on alge- and strong ultrasonic waves, and to calculate non-
braic invariants composed from uik (see Elas- linear effects in crystals containing dislocations
ticity theory invariants). In isotropic media (see (e.g., dilatations), etc.
Isotropy), such invariants are the coefficients I1 ,
I2 , I3 of the polynomial P (x) = |xδik − uik |
NONLINEARITY, MAGNETODYNAMIC
= x 3 − I1 x 2 + I2 x − I3 . The defining expressions
that relate strains to deformations have the form See Magnetodynamic nonlinearity.
σik = I −1/2 (∂W/∂uik ), where I = |δik + 2uik |.
These expressions may be written explicitly if NONLINEAR MAGNETOACOUSTIC
one chooses some form for the function of invari- RESONANCE
ants W . The following simplified defining equa- A first-order parametric phenomenon, in which
tions are sometimes used in mechanics for nonlin- uniform oscillations of the magnetization occur-
ear elastically isotropic media: σ = f (u), σik =
ring in a magnetically ordered medium under the
g(u )uik . Here σ = (1/3)σll , u = (1/3)ull , σik
action of external pumping, as a result of a non-
and uik are the stress deviator and the strain devia- linear magnetoelastic interaction, excite a spin
tor, respectively. Further, u = (2/3)(I12 − 2I2 )1/2 wave with the frequency ωk , as well as an elas-
is the intensity of shear strain, while the func- tic vibration of the sample with the frequency ωa .
tions f (u) and g(u ) depend on the nature of the The pumping frequency, ω, is ω = ωk + ωa , with
medium and are evaluated empirically (in a lin- ωk ∼ ω
ωa . The effect occurs in any clean sys-
ear case, we have f = 3K, g = 2μ, where K tem, such as a two-dimensional electron gas in
and μ are the bulk modulus and shear modulus). the quantum Hall effect regime. It is observed in
In the physics of solids, W is usually constructed high-quality ferrite monocrystals at a pumping fre-
by the successive approximation method: W = quency near ≈10 GHz. The elastic frequencies
Wlin + W + W + · · ·, where Wlin corresponds to of a body depend on its dimensions, and usually
the linear elastic case, and W , W , . . . are the cor- are in the range of MHz, with threshold pumping
rections to Wlin of increasing order in uik . In that power as high as few mW for perfect monocrys-
case, the relation between σik and uik may also be tals of yttrium–iron garnet. Other magnetic oscil-
written down directly from symmetry considera- lations may serve for pumping, such as paramet-
tions. Nonlinear defining equations for crystals are rically excited spin waves (nonlinear secondary
often used in the form W = (1/2)ciklm uik ulm + magnetoacoustic resonance, see also Parametric
(1/6)ciklmnp uik ulm unp , where ciklm and ciklmnp excitation of spin waves).
890 NONLINEAR MEDIUM
Table 1. Crystal system (syngony) and applications of some common nonlinear-optical crystals
Note: SHG, second harmonic generator; PG, parametric generator; M, electro-acoustic modulators and deflectors; IO, in-
tegrated optics; OBD, optical bistable devices; SMRS, stimulated magnetic Raman (“spin-flip”) scattering.
bonds, ∼107 –108 V/cm. Laser fields reach the and related optical nonlinearities in solids are pro-
levels of E ∼ 106 –107 V/cm and higher, com- nounced due to the large concentration of parti-
ing close to Ei . Stimulated oscillations of a bound cles in a unit volume. They sharply increase as the
electron affected by such a field become notice- frequency of the electromagnetic wave approaches
ably anharmonic, which results in a nonlinear de- either an exciton, or a band-to-band resonance fre-
pendence of P on E. For E/Ei < 1, one may rep- quency. In addition, there are other mechanisms
resent this dependence as a power series: responsible for the nonlinearity of refraction of
(1) (2) (3) light in crystals, not related directly to P nl , such as
Pa = χab Eb + χabc Eb Ec + χabcd Eb Ec Ed + · · · electrostriction, laser heating of the medium, input
from free electrons, etc. Free carriers excited dur-
= Pal + Panl , (1)
ing either single-photon two-photon absorption,
(1) produce a strong nonlinearity of n(E) in semicon-
where Pal = χab Eb ; Panl are higher-order terms:
(2) (3) ductors. Diffusive redistribution of charges in pho-
χabc , χabcd are quadratic, cubic etc. nonlinear torefractive crystals, such as LiNbO3 , Bi12 SiO20 ,
susceptibilities. As for their order of magnitude, BaTiO3 , etc. results in the generation of an elec-
χabc ∝ 1/Ei , χabcd ∝ 1/Ei2 . Typical numerical trostatic field that produces a change in the refrac-
values for insulators in the transparency region tive index via the linear Pockels effect.
are: χ (1) ∼ 1, χ (2) ∼ 10−11 –10−12 m/V; χ (3) ∼ Applications of nonlinear phenomena in solids
10−21 –10−22 m2 /V2 . These are tensor quantities include light frequency multiplication, tunable op-
characterizing the properties of a crystal medium. tical parametric oscillators, control of spatial and
By symmetry 3rd, 5th and higher odd-rank tensors temporal characteristics of laser beams, etc. Bire-
vanish for crystals with a center of symmetry, and fringent nonlinear crystals can efficiently match
χ is nonzero for every rank when there is no center the phase velocities of both fundamental and
of symmetry. frequency-doubled waves, and several tens of per-
A nonlinear dependence of the polarization of cent efficiency can be achieved in frequency dou-
the medium on the field strength results, accord- bling using the crystals potassium- and ammon-
ingly, in a nonlinear dependence of the electric ux ium-dihydrogen phosphate, as well as LiNbO3 ,
density, D on the field E (in CGS units):
LiIO3 , BaNaNb5 O15 , ZnGeAs2 , CdGeAs2 ,
Da = Ea + 4πPa ArGaSe, and others.
The enhanced nonlinearity of refraction in
= Ea + 4π(χab Eb + χabcd Eb Ec Ed + · · ·). semiconductors and photorefractive crystals (see
(2) Photorefraction) makes possible highly effective
recording of dynamic phase holograms (see Real-
This, in its turn, brings about dependences of the time holography), for wave front reversal, and for
dielectric constant ε and refractive index n on E: producing bistable optical elements (see Optical
bistability). Other applications include devices to
n(E) = n0 + n2 E 2 + · · · . (3)
control the parameters of laser radiation, and to de-
The dependence n(E) results in phenomena such velop purely optical communication systems and
as self-focusing and self-defocusing of a light computers.
beam in a nonlinear medium. Terms in Eq. (1),
nonlinear in E, describe basic nonlinear optical NONLINEAR PYROELECTRIC
processes. Thus, the term χabc Eb Ec is respon- A pyroelectric substance (see Pyroelectric ma-
sible for generation of the second harmonic (fre- terials) featuring a spontaneous reversible polar-
quency doubling); χabcd Eb Ec Ed for the third ization by an electric field (see Ferroelectricity).
harmonic, i.e. parametric oscillation, and these The temperature dependence of this polarization
provide the dependence n(E). In fact the term is nonlinear at the phase transition point. Ferro-
χabcd Eb Ec Ed also contains an intensity-depend- electrics which display a strong spontaneous po-
ent absorption correction accounting for two- larization temperature dependence are called non-
photon light absorption. Nonlinear susceptibilities linear pyroelectrics. Above the phase transition
892 NONLINEAR SCHRÖDINGER EQUATION
layer: amorphous selenium, chalcogenide vitreous changes of the refractive index, are the basis of
semiconductors, crystalline photoconduction ma- amplitude–phase recording of images, holograms,
terials in a polymeric host, and organic semicon- and optical data in a bit-by-bit mode in thin layers
ductors. Progress has been achieved in the de- and bulk plates. It is possible to perform record-
velopment of nonsilver photography on the ba- ing with a resolution ν > 103 mm−1 in real time.
sis of hydrogenated amorphous silicon. Also used The sensitivity is low, but the materials are ap-
in nonsilver photography are photothermoplastic plicable in laser systems for the manufacture of
materials where a development of the latent elec- memory elements and integrated optics elements.
trostatic relief is achieved either due to a strain in Selective etching of an exposed layer determines
the polymeric layer that is also a photoconductor, the applicability of the layers of chalcogenide vit-
or due to extra layers of inorganic semiconductors. reous semiconductors as high-resolution inorganic
The sensitivity of these materials is high and com- photoresists.
parable to that of ordinary photomaterials. The phenomenon of photorefraction, or photo-
In microphotography, as well as when copying
induced alteration of the refractive index in elec-
and duplicating technical and text documents, dia-
trooptical crystal ferroelectrics underlies the appli-
zotype materials and vesicular materials are used.
cation of these crystals to phase recording of holo-
The image is produced by light-scattering bub-
grams.
bles of the gas released in a polymer layer under
photochemical decomposition of a light-sensitive The photoinduced phase transformations take
diazocompound introduced into the layer. There place in crystallizing layers of chalcogenide vitre-
is a possibility of recording images via photo- ous semiconductors, in TeOx and magnetooptical
chemical reactions in the layers of a semiconduc- materials MnBi, and offer prospects of producing
tor (of PbI2 , BiI3 type) as well as via stimulated bit-by-bit recording carriers. Nonsilver photogra-
metal diffusion in structures containing a metallic phy also uses thermophotographic materials and
layer (Ag, Cu) and a semiconductor (e.g., PbI2 ). systems that are sensitive to a temperature change
Their sensitivity is not high but the process can be on account of the energy absorbed during the ex-
used to record optical elements and optical stor- posure. These are liquid crystals, luminophors with
age disks. Widely used in nonsilver photography thermal extinction, and evaporating metallic or or-
are also photopigments where the image record- ganic layers. An extension to the IR region is
ing is based on the phenomenon of photochro- possible by using the semiconductor photographic
matism of organic and inorganic substances (see processes based on the intrinsic photoeffect.
Photochromic effect). In view of the irreversibil-
ity of light-stimulated changes in coloration, these NORMAL DISTRIBUTION
materials are suitable for repeated image record- The same as Gaussian distribution.
ing with high resolution ν > 103 mm−1 , e.g., in
holography. New materials with similar proper-
NORMAL VIBRATIONS (normal modes)
ties are bacteriorhodopsin layers in a polymeric
host. The processes of photopolymerization (pho- Small natural vibrations of a system of sev-
toresists), photopretanning (gelatin), and isomer- eral or many particles (including those in a crystal
ization (stilbene) are used in nonsilver photogra- lattice) described by such single-frequency time-
phy in the fields of printing and microelectronics. dependent harmonic functions (normal coordi-
To record the phase relief of holograms, reoxan is nates) that their number coincides with the number
used in which the oxidation of the anthracene in of the vibrational degrees of freedom of the sys-
a polymeric host (polymethylmethacrylate) takes tem. An arbitrary vibration may be represented in
place under the action of light. The sensitivity of the form of a superposition of normal vibrations.
this method is low; and fixation of the image is Normal vibrations in a crystal lattice are specified
needed to produce a stable phase relief. by their quasi-wave vector k, and the number of
Reversible and irreversible photostructural their vibrational branch, α. Together they uniquely
modifications of the chalcogenide vitreous semi- define the frequency of such normal vibrations, ω,
conductors, which are accompanied by significant and the polarization vector, e. The displacement of
894 NOTCH
the sth atom in the nth unit cell positioned at r n is higher this ratio, the more sensitive the material is
written as to the notch. To arrive at comparable results, sam-
ple shapes and testing techniques have been stan-
usn = Qα (k)es (k, α) exp(ikr n ),
dardized.
kα
where the normal coordinates Qα (k) satisfy the NOTTINGHAM EFFECT (W.B. Nottingham, 1941)
harmonic equation Heating of the eld-emission cathode and cath-
ode cooling during thermal electron eld emission.
d2 Qα (k)
+ ωα2 (k)Qα (k) = 0. The energy distribution of electrons at low temper-
dt 2 atures is practically identical to the Fermi–Dirac
When the equations of motion are linear (harmonic distribution at 0 K (see Fermi–Dirac statistics).
approximation), normal vibrations are mutually in- Therefore, the average energy of autoelectronic
dependent, and the system energy may be repre- emission electrons tunneling through the potential
sented as a sum of the energies of individual har- barrier in a strong electric field applied to the sur-
monic oscillators corresponding to separate nor- face of the cathode remains lower than the Fermi
mal modes. When the anharmonic vibrations are energy in the cathode, and lower than the energy
taken into account the normal mode vibrations in- of electrons reaching the cathode from the exter-
teract with each other, so the above expressions nal electric circuit. This energy difference results
become only approximate. In the quantum theory in the heating of the cathode. In the case of ther-
of crystal lattice vibrations the normal vibrations mal electron field emission, the electrons escap-
correspond to phonons, the elementary excitations ing from the cathode have energies exceeding the
of the crystal. Fermi level of the cathode. Their replacement by
electrons from the external circuit having average
NOTCH energies lower than that of the escaping electrons
Angular cut or indentation serving as a geo- results in the cooling of the cathode.
metric stress concentrator (see Stress concen-
tration), used to decrease the coefficient of soft- NUCLEAR DOPING, irradiation doping
ness of the mechanical testing of materials. Most Method of introducting doping additives into
widely used tests, which involve application of material which is based on nuclear transforma-
stress concentrators, are tests of samples with a tions of host nuclei to daughter impurity atoms
notch for stretching and e xure. Samples of circu- under the action of charged particles (protons p,
lar cross-section are often notched circularly; the α-particles, deuterons d, and so on), gamma rays
preparation of rectangular samples for stretching or neutrons n (in the latter case one speaks of neu-
requires making a symmetrical notch along two tron doping).
sides. Samples for tests of tension are notched A principal difference between nuclear doping
along one of the sides of a section, perpendicu- and routine methods of embedding impurities in
lar to the longitudinal axis. The stress concentra- crystals consists in the direct formation of the im-
tion factor, the ratio of maximum stress to nomi- purity in the bulk sample from the atoms of the
nal stress, is used for a quantitative assessment of dopant, instead of introducting the impurity in the
the notch rigidity. The factor is determined by the semiconductor from outside, i.e. from the melt or
geometry of the notch: the deeper and sharper the the gaseous phase, or by diffusion. The production
notch, the higher the factor, and the easier it is to of impurity atoms at any point of a bulk sample at
achieve brittle failure. For practically every metal, a distance smaller than the mean free path distance
even very plastic ones (see Plasticity), one can find is equally probable, hence the resulting impurity
an appropriate shape and dimensions of the sample distribution is essentially isotropic.
and the notch to produce brittle failure under given The reactions that take place from incoming
test conditions. The criterion of notch sensitivity d, α, and p particles of the types (α, p), (α, n),
of a material is the ratio of ultimate strength val- (p, α), (p, d), etc. are induced by the very high
ues for intact and notched samples (σB /σBn ): the energy of the incident particles which is required
NUCLEAR DOPING 895
to overcome the Coulomb barrier when approach- the donor impurities Ge and Se are formed. In
ing a nucleus, transfer to it the energy needed to semiconductors of the type of InSb, the donor im-
compensate for the breaking of chemical bonds, purity Sn prevails.
and again overcome the Coulomb barrier by the The mean free paths of thermal neutrons in var-
charged particle that flies out of the nucleus. This ious semiconductors are the following (in cm):
barrier reaches approximately 10, 20, and 30 MeV
Si Ge GaAs InSb
for light, intermediate, and heavy nuclei, respec-
tively. This process was implemented by doping 125 4.0 2.8 0.14
silicon with phosphorus, e.g., through the reac- The mean free path determines the volume of the
tions 28 Si (α, p) 31 P; 29 Si (α, 2p)31 P; as well as doped material within which the impurity is spread
by doping germanium with the impurities Ga, As, uniformly. The indirect negative phenomena that
and Se and utilizing the actions of deuterons and accompany the intentional formation of necessary
α-particles. The nuclear reaction 28 Si (α, p) 31 P impurities are induced (bulk and surface) radioac-
corresponds to the silicon isotope 28 Si absorb- tivity and radiation defects. The bulk radioactivity
ing an α particle to form the excited state sulfur is due to the radiative decay of an unstable isotope
nucleus 32 S, which then transforms to the stable to an atom of the doping impurity. To compensate
phosphorus isotope 31 P by emitting a proton p. for this the exposure of the doping material during
Due to the high rate of braking (decelleration) of 10 half-life periods is needed. The source of sur-
charged particles in matter, the thickness of the re- face radioactivity is the adsorption of radioactive
sulting doped layer is small, and the distribution of elements from the medium of the nuclear reactor
the introduced impurity is nonuniform. by the surface of irradiated ingots. This is elimi-
Nuclear reactions which take place under the nated by washing the ingots in special solutions,
action of γ -quanta are called photonuclear reac- or by etching the surface layer. Radiation defects
tions, and reactions of the types (γ , n), (γ , p), are annealed during the heat treatment.
(γ , 2n), and (γ , pn) are referred to as threshold The technology of the nuclear doping of semi-
reactions. A particular feature of these reactions conductors consists of the following stages: ob-
is the dependence of the reaction cross-section σ taining the initial monocrystals, irradiation, de-
on the γ -quantum energy E in the range E ≈ 10 to contamination, heat treatment, control of electri-
25 MeV, where σ (E) has a pronounced maximum, cal physical parameters, and output radiological
the so-called giant resonance. The main impurity monitoring. The main advantages of the nuclear
that is formed in silicon as a result of photonuclear doping of semiconductors include the high degree
reactions is aluminum. In germanium the gallium of uniformity of the introduced impurity distrib-
atoms are predominantly formed, and the impuri- ution (see Fig. 1), the accuracy of the regulation
ties of Se, Ge, and Zn are formed in GaSe. Gamma
quanta can penetrate deeply into a semiconduc-
tor, so photonuclear reactions are capable of dop-
ing large volume regions of the material. Also of
interest is nuclear doping through nuclear reac-
tions with thermal neutrons (n, γ ). The energy of
thermal neutrons E ∼ 0.5–5·10−3 eV is compa-
rable with the energy of the atomic thermal vi-
brations of a moderator, and their velocity is v ≈
2·105 cm/s. The nuclear reaction cross-section is
inversely proportion to v (“1/v law”). The main
sources of thermal neutrons are nuclear reactors
and nuclear fuel, that is 235 U. A final product of
the (n, γ )-reactions in silicon is phosphorus. In
germanium an acceptor impurity and donor impu-
rity, respectively As and Se, are formed. Three ac- Fig. 1. Distribution of specific resistance in initially un-
ceptor atoms are formed per donor atom. In GaAs doped (1) and nuclear-doped silicon (2).
896 NUCLEAR ELECTRIC QUADRUPOLE MOMENT
an oscilloscope or a recorder, or for transmission and the local symmetry at the position of the nu-
to a data processor, or a computer, etc. cleus is lower than cubic;
A particular method for observing NMR is its
(k) γ 2 h2 −3
detection under the conditions of the electron– Wss = rik 3 cos2 θik − 1
2
nuclear double resonance, where the detection i>k
of the NMR signal is performed indirectly, i.e.
× Iˆi Iˆk − 3Iˆzi Iˆzk ,
through the observation of an electron paramag-
netic resonance (EPR) signal. In the vicinity of where rik is the distance between the kth nucleus
electron paramagnetic centers, the hyper ne inter- under consideration and the ith nucleus, θik is the
action of electrons and nuclei results in a common angle between vectors r ik and B 0 ; WP and WE
system of spin levels. Thus, the nuclear-spin tran- have forms similar to that of WQ with a single
sitions can affect the difference in those level pop- difference in the sense of spin operators: in the
ulations between which an electron-spin transition case of WP these operators relate to P , and in
takes place. This leads to a change in the EPR sig- the case of WE they relate to the electric field E.
nal that serves for detection of the NMR signal. The term WE which is linear in respect to E, is
Hamiltonian. Being a purely quantum effect, nonzero only for nuclei in sites without a center of
NMR is completely described with the help of symmetry. The terms WP and WE can be looked
a Hamiltonian that is called a spin Hamiltonian upon as terms of extra-induced quadrupole inter-
because the Hamiltonian operators all depend on actions. When considering the whole system of
spin variables. This Hamiltonian W has a partic- nuclei, the expression (1) should be supplemented
ular form in every individual case, but in general by a summation which takes into account all the
it can be represented in the manner that takes into nuclei; this is of particular importance in the case
of solids. Each term in Eq. (1) can be additively
account all possible NMR-involved interactions of
divided into static and dynamic parts. The static
the nucleus with the environment:
part determines the line position in the spectrum,
W = W Z + WQ + W ss + WP + WE , (1) while the dynamic part determines the probabil-
ities of quantum transitions (line intensities) and
where W Z is the interaction with the applied DC provides for relaxation processes. In this way, not-
magnetic fields B (Zeeman interaction) WQ is the ing that B = B 0 + B 1 , the Zeeman term W Z =
interaction with the crystal eld gradient (quadru- W Z0 + W Z1 involves two main interactions, a
pole interaction), W ss is the interaction with mag- static one W Z0 with the large applied field B 0
netic moments of surrounding nuclei and electrons which provides the line positions, and a dynamic
(spin–spin interaction), WP is the effective inter- one W Z1 with the alternating radio-frequency
action with applied external pressure P (strains), field B 1 which induces the transitions. The Zee-
WE is the effective interaction with external elec- man term also includes interactions with local
tric fields E. Each interaction in Eq. (1), in its turn, magnetic field inhomogeneities (static) and fluc-
consists of one or more terms whose form depends tuations (dynamic) that account for the line width
on the nature of the interaction, the number and and saturation mechanisms.
magnitudes of the nuclear spins, and the local sym- Spectrum. The NMR spectrum (i.e. the number
metry of their sites. The most commonly encoun- of lines, their intensities and angular dependence)
tered expressions can be written in the following is completely determined by the spin Hamil-
way: tonian (1). In the case of a nucleus with spin I in a
permanent magnetic field B 0 (term W Z0 ), 2I + 1
WZ = −γ (I B 0 ), equidistant energy levels εm = −γ I mB0 appear
QVzz ˆ2 (here m is the quantum number of operator Iˆz
WQ = 3I − I (I + 1)ê , which takes on values −I, −I + 1, . . . , I − 1, I ).
4I (2I − 1) z
The radio-frequency field B 1 (time-dependent
where Q is the electric quadrupole moment, Vzz is term W Z1 ) induces the magnetic dipole transi-
the z component of the electric eld gradient ten- tions between adjacent levels (
m = ±1) un-
sor, ê is the unit matrix; WQ = 0 unless I > 1/2, der the condition B 1 ⊥B 0 (selection rule), so that
900 NUCLEAR MAGNETIC RESONANCE
there will appear only a single isotropic line in affect the quantum transitions. Hence the electric
the NMR spectrum since the transition frequency component of the radio-frequency field E 1 can
ν = (εm − εm−1 )/ h does not depend on m. In (as well as B 1 ) induce transitions between mag-
fact, owing to presence near the nucleus under netic sublevels (for the sample positioned between
consideration of electrons and other nuclei, the capacitor plates), but with special selection rules
magnetic field in atoms, molecules, and condensed (e.g.,
m = ±2 at E 1 B 1 ).
matter will differ from the external field B 0 due Line broadening. As a rule, crystals are not
to the polarization of electron shells and the di- perfect. Therefore, NMR is often used for inves-
rect nuclear magnetic dipole–dipole interaction. As tigating disordered systems (powders, glasses, so-
a result, there appear many incremental additions lutions, crystals with diverse defects). A particu-
B 0 to B 0 which, in the general case, are not par- lar feature of such systems is the inequivalence
allel to B 0 . These lead to line broadening and also (nonuniformity) of the conditions at the positions
anisotropic shifts of the NMR lines: chemical shift, of nuclei, due to: (i) differences of magnetic and
Knight shift (see Hyper ne interaction), and also electric fields, and deformations induced by struc-
shifts due to an indirect nuclear spin–spin interac- tural distortions of the crystal and defects of di-
tion. In many cases the nuclei in a solid are situ- verse kinds; (ii) the orientational inequivalence in
ated at two or more chemically inequivalent sites, regard to applied external fields; (iii) the nonuni-
each of which is characterized by its own specific formity of the external fields, etc. All these lead
chemical shift. As a result, there appear particular to dispersion of the parameters of the spin Hamil-
lines in the spectrum which arise from different tonian (1), and as a consequence to inhomoge-
nuclear groups. The level splittings from equiva- neous broadening of the NMR lines. Despite the
lent sites are still equally spaced, but broadened. existence of a number of methods for describing
Other terms of the spin Hamiltonian (1) break such broadening (statistical theory of line shape,
the rule of an equidistant arrangement of energy moment method (see Moments of spectral lines),
levels. Thus, the operator WQ adds to a Zeeman theory of randomly or partially disoriented cen-
level εm an extra term which is proportional to ters, etc.) innomogeneous broadening of all types
m2 − I (I + 1)/3. As a result, the lines in the NMR can be treated from a common viewpoint. In this
spectra become split into 2I lines ( n e structure). case, the general expression for the line shape
In the case when B0 = 0, the system of levels is takes the form:
determined by operator WQ (there appear two lev-
els if I = 3/2), and transitions between them in- I (ω) = ϕ ω − ωP (F ) V (F )ρ(F ) dF, (2)
duced by the radio-frequency field B 1 lead to a
zero field NMR that is often called nuclear quadru- where ϕ is the initial line shape without taking per-
pole resonance. A specific line splitting arises un- turbing factors into account, V (F ) is the transi-
der action of operator WE . There sometimes exist tion probability per unit time, ρ(F ) is the distribu-
inversion-inequivalent positions in crystals which tion function of the parameters F (F1 , . . . , Fk ) that
the nuclei of the same type can occupy with equal characterize the broadening mechanisms (com-
probability. Since the magnetic field is insensitive ponents of fields, deformations, angles). So in
to the inversion operation, the field does not dis- the case of randomly oriented centers (powders,
tinguish these positions, and the related lines in glasses) one would interpret F as the Euler angles
the NMR spectrum are superimposed. An electric that characterize the orientation of a powder par-
field E 0 applied to a crystal, due to the mutual ticle with respect to the coordinate system related
inversion of these inequivalent positions, will be to the external fields.
oppositely directed for each of them. As a result, Relaxation. NMR is accompanied by relax-
the frequency corrections linear in E 0 correspond- ation processes which act to restore the initial
ing to different positions will have different signs, Boltzmann distribution equilibrium state which had
and the shift of these two groups of lines in op- been disturbed by the transitions induced by the
posite directions will produce an additional split- applied electromagnetic radiation. Such relaxation
ting in the NMR spectrum. The term WE can also processes are due to the interactions between the
NUCLEAR MAGNETIC RESONANCE 901
nucleus and the lattice, as well as between the nu- TS < 0, stimulated development of double reso-
clei themselves. Therefore, a distinction is made nances based on NMR. These are characterized
between spin–lattice relaxation and spin–spin re- by resonance transitions at two frequencies, which
laxation. If the transitions stimulated by the radio- in the presence of a multilevel system occur simul-
frequency field are dominant, a saturation effect taneously or in a certain sequence. The purpose of
(equalization of level populations) takes place, this is to increase the intensity of the absorption by
and manifests itself in the decreased intensity increasing the difference in population (this princi-
and broadening of the NMR signal. The relax- ple is used in electron–nuclear double resonance),
ation processes are characterized by their relax- and to obtain a source of electromagnetic radiation
ation times, and are described by kinetic equations through the establishment of a larger population at
for level populations. In the simplest case of two the upper level than at the lower one (a maser uses
levels i and j , the equations for populations ni and this principle).
nj take the form: Approximation approaches. The complexity of
an analytical description of NMR is due to the
ṅi = −ni a + nj b, ṅj = −ni a − nj b, enormous number of interacting lattice nuclei.
This circumstance complicates diagonalizing the
where a = Vij0 + Vij ; b = Vj0i + Vj i ; the terms
spin Hamiltonian (1), calculating and simulat-
Vij0 = Vj0i and Vij are the probabilities of tran- ing NMR spectra, and studying the kinetics of
sition from level i to level j per unit time under spin systems (relaxation, saturation, etc.). In this
the action of radio-frequency field and relaxation connection, approximate approaches and methods
mechanisms, respectively. The relaxation time is were developed in order to describe the diverse de-
determined by the expression T r = (Vij + Vj i )−1 tails of the behavior of the spin system in NMR.
and characterizes the rate of establishing equilib- First of all, it is worthy to mention the Bloch equa-
rium. These relaxation processes shorten the life- tions (F. Bloch, 1946) which describe the dynam-
times of particles at spin levels, which leads to ics of nuclear magnetization that were introduced
their broadening by affecting the width and shape at the very beginning of the NMR era. Although
of the NMR lines. This broadening, which is the these equations are based on a classical description
same for all nuclei, is referred to as homogeneous of magnetic moment precession, and hence are not
broadening. It can, in particular, determine the adequate for explaining all experimental data, they
shape of ϕ in Eq. (2). played a major role in establishing magnetic reso-
Spin temperature. For a description of spin sys- nance as a main branch of science. Subsequently
tems the concept of “spin temperature” TS can more refined theories were developed which are
be introduced. To do this, the relation between grounded on the quantum kinetic equations for a
the level population and the temperature which density matrix (Bloch–Red eld theory, Provotorov
determines the Boltzmann distribution is general- theory, and so on). An enormous amount of ex-
ized for the case of nonequilibrium populations. perimental data have been successfully explained
The assumption being made is that the spins are within the frameworks of these theories.
in equilibrium with each other, but not with their Rotating coordinate system. A widely used ap-
surroundings (lattice vibrations for a solid and proach for analyzing NMR spectra is to transform
Brownian motion for a liquid). Hence, for arbi- the motion to a rotating coordinate system, i.e. to
trary ratios between the populations of upper (nu ) examine the NMR resonance in an terms of an ef-
and lower (nl ) levels, it follows that fective magnetic field B eff = iB1 + k(B0 − ν/γ )
εu − εl acting upon the nuclear spins, in a coordinate sys-
TS = − . tem which rotates with the frequency ν about the
k B ln(nu /nl )
permanent magnetic field B 0 z, i.e. in the direc-
For nu = nl (saturation), TS = ∞; for nu > nl , tion of the nuclear precession, where the radio-
TS < 0. The opportunity to produce nonequilib- frequency field B 1 is applied normally to B 0 .
rium populations, including the particular case The frequency ν eff corresponding to B eff is given
with TS = ∞ as well as negative spin temperatures by ν eff = γ B eff = [(ν0 − ν)2 + γ 2 B12 ]1/2 , where
902 NUCLEAR MAGNETIC RESONANCE, ACOUSTIC
ν0 = γ B0 . This method not only clarified the ex- adsorption of atoms and molecules on solid sur-
planation of many phenomena in the NMR, but it faces; studies of the attributes of phase transi-
allowed Redfield to predict, and then to observe, tions; determination of spatial ordering in liquid
a so-called NMR in a rotating coordinate system. crystals, investigating weakly coupled molecular
The essence of this phenomenon involves the fol- groups in polymers and organic crystals; establish-
lowing. If an additional alternating field B 2 B 0 ing the character of collective motions of atoms
with the frequency ν eff is applied to the spin sys- (molecules) of the lattice; revealing mechanisms
tem, then its perpendicular component B eff will of chemical reactions; magnetic resonance imag-
induce a new resonance, i.e. NMR in a rotating ing. Finally, worthy of mention is the past and
coordinate system. present role of NMR in studies of the structure and
High-resolution NMR. Due to the large num- properties of such recently discovered materials as
ber of nuclei which influence each other via the fullerenes and fullerites.
dipole–dipole interaction, the width of NMR lines In magnetically ordered materials where NMR
in solids is abnormally large (about 10−4 T or tens is observable in the effective field of a hyper ne in-
of kHz), which hinders the observation and inves- teraction between the nuclear and electronic spins
tigation of such fine structure effects as chemi- of a magnet, NMR studies yield data on the atomic
cal and Knight shifts, indirect internuclear inter- magnetic structure, its dynamics, the interaction of
actions, etc. In gases and low-viscosity liquids the nuclear and electron spin waves (Suhl–Nakamura
thermal motion of atoms and molecules causes the interaction), etc. (see Indirect nuclear spin–spin
coupling).
“natural” averaging of the dipole–dipole interac-
tion, diminishing it close to zero. This circum-
NUCLEAR MAGNETIC RESONANCE,
stance makes possible the high resolution of the
ACOUSTIC
NMR method which is employed in the majority
See Acoustic nuclear magnetic resonance.
of the physicochemical and analytical applications
of NMR. This branch of NMR designed for detect-
NUCLEAR OPTICS
ing and investigating very narrow lines is referred
Branch of the theory involving the interaction
to as high-resolution NMR. For solids, it was nec-
of nuclear radiations with matter where the radi-
essary to develop special methods for artificially
ation propagation processes are interpreted using
narrowing NMR lines by suppressing the dipole–
the formalism of optics. In the case of neutron
dipole interaction. Included among them are mul-
waves, the term neutron optics is accepted. In nu-
tipulse methods of various types, and magic angle
clear optics an index of refraction n for a “hard”
spinning (MAS) of √ the sample at the magic an- electromagnetic or electron wave (i.e. high-energy
gle (θ = cos−1 (1/ 3) = 54◦ 45 ). These methods
wave) is introduced by the relation
have expanded considerably the utility of NMR,
particularly in its applications for solids. 2πρf (0)
n=1+ , (1)
Applications. The NMR method is widely em- k2
ployed in various branches of physics, chemistry, where f (0) is the forward scattering amplitude
biology, medicine, and other areas of science. It (elastic coherent scattering amplitude at zero an-
has achieved particular success in studies of con- gle), ρ is the density of scattering particles in mat-
densed matter where the need for information is ter, and k is the wave number of the incident wave.
quite extensive, and it involves a variety of di- For the high energy (small wavelength) condition
verse interactions of nuclear systems with the en- |n − 1| 1 the geometrical optics laws of refrac-
vironment and other factors. Several examples of tion (Snell’s law) and total internal re ection are
areas of application are: establishing the atomic valid. For the case when f (0) depends on the po-
structure of crystals (determination of atomic co- larization state of the neutron or photon, there ap-
ordinates); determination of the electronic struc- pears a difference in refractive index for right-hand
ture of crystals and the nature of chemical bonds; and left-hand circular polarization which yields
establishing the nature of chemical and physical the Faraday effect. Eq. (1) is applicable for the
NUCLEAR QUADRUPOLE INTERACTION 903
description of matter consisting of randomly dis- the action of high internal local fields ∼10–100 T
tributed scatterers, and for radiation with a wave- relative to the hyper ne interaction on the nuclei
length much less than the distance between scat- of paramagnetic ions. Dynamical methods apply
terers. When absorption is taken into account the magnetic resonance in a straightforward manner;
refractive index n has a complex form. high electron spin polarization is “pumped over”
to the nuclear spin polarization with the help of
NUCLEAR ORIENTATION the microwave saturation of “allowed” or “forbid-
A collection of methods for producing non- den” electron spin transitions. These methods in-
randomly oriented nuclear spins. The extent of clude the Overhauser effect, solid effect, electron–
their orientation is characterized by two quantities, nuclear cross-relaxation, and dynamic cooling (see
Dynamic nuclear polarization) of the nuclear spin
Iz
P1 = subsystem which is in (direct or microwave field-
I induced) thermal contact with the electron spin–
and spin reservoir that is disturbed from thermal equi-
librium by the incomplete saturation of the elec-
3Iz2 − I (I + 1)
P2 = , tron paramagnetic resonance line. Examples of
I (2I + 1) dynamic methods are those of chemical and opti-
called the nuclear polarization and the nuclear cal nuclear polarization. Nuclear orientation meth-
alignment, respectively. Here I is the spin of the ods are employed mainly for obtaining the polar-
nucleus, and Iz is its projection along the z axis; ized nuclear targets needed for fundamental stud-
. . . means averaging over the ensemble. P1 is ies in elementary-particle physics. Another area
proportional to the nuclear magnetic polarization of application of nuclear orientation is the study
(i.e. magnetization), it determines the value of of nuclear magnetic ordering at the temperature
the nuclear magnetic resonance absorption sig- of the reservoir of nuclear spin–spin interactions
nal, and in the case of I = 1/2 it is the differ- TI I 10−6 K.
ence in the number of spins directed along and
against (antiparallel to) the magnetic eld direc- NUCLEAR POLARIZATION
tion. Obviously, −1 P1 1. For randomly ori- See Nuclear orientation.
ented nuclear spins, P1 = 0. The quantity P2 is
a measure of the excess of the number of spins NUCLEAR QUADRUPOLE INTERACTION
aligned along the z axis (parallel or antiparallel) Interaction of nuclei of a crystal with the elec-
over those aligned transverse to it. One can see that tric field gradients of their environment caused by
polarization is accompanied by alignment (when the presence of nuclear electric quadrupole mo-
P1 = 0, then P2 = 0) but not vice versa (when ments and a spatial nonuniformity of the fields.
P2 = 0 and P1 = 0). The angular distribution of The part of the total crystal Hamiltonian which
γ - and α-radiation of radioactive nuclei with par- characterizes this interaction is called the quadru-
ity conservation depends on P2 and perhaps even pole Hamiltonian. Without taking into account
on higher-order terms. In the case of β-decay par- the internuclear interactions in the crystal (as a
ity is not conserved, and the distribution of radi- rule, they are very small), the overall quadrupole
ation depends on P1 , P2 and related higher-order Hamiltonian is the sum of the quadrupole Hamil-
terms, both even and odd. tonians of the individual nuclei, each of which is
Methods of nuclear orientation can be subdi- a convolution of the nuclear electric quadrupole
vided into static and dynamic types. The static moment tensor QT and the tensor q T of the elec-
method is a “brute force” one since, to attain the tric eld gradient at the site of the nucleus. By ex-
maximum values of nuclear orientation, extreme pressing the components of tensors QT and q T in
conditions are used such as very high external the principal axis system and defining the scalar
fields and very low temperatures. Another exam- quadrupole moment Q and the electric field gra-
ple of a static method is the Gorter–Rose method dient factor q as the components along the main
(C.I. Gorter, M.E. Rose, 1948) which is based on z axis (Qzz and qzz = −e∂Ez /∂z = ∂ 2 V /∂z2 ),
904 NUCLEAR QUADRUPOLE RELAXATION
the quadrupole nuclear Hamiltonian HQ is re- metastable positions); the modulation mechanism
duced to the form: (various movements of particles neighboring those
eqQ 2 containing the nuclei under consideration). Com-
HQ = 3I − I 2 + η Ix2 − Iy2 , parison of results of experimental and analytical
4I (2I − 1) z
investigations of the temperature dependence of
where e is the electron charge, I is the nuclear the relaxation times is employed in the physics
spin, Ix,y,z are the components of the nuclear and chemistry of solids for studying the internal
spin vector in the principal axes system of ten- dynamics and the phase transitions in crystals.
sor qT , the quantity eqQ is called the quadru- A more recent approach to the computation of re-
pole interaction constant (sometimes qzz is de- laxation times based on the use of the Einstein–
fined as −∂Ez /∂z and the quantity e2 qQ is called Fokker equations is much more general than the
the quadrupole interaction constant). The asym- earlier conventional approach in the sense that
metry parameter η in this equation indicates the it does not assume that the thermal motions are
deviation of the quadrupole moment from axial small, so perturbation theory methods do not ap-
symmetry (0 η 1, η = 0 for the axial case). ply. The term nuclear quadrupole relaxation is also
The nuclear quadrupole interaction which deter- employed to denote quadrupole interaction mech-
mines a “pure” nuclear quadrupole resonance (that anisms for the relaxation of nuclear magnetic po-
is without any applied magnetic field) also influ- larization in an applied external magnetic field.
ences nuclear magnetic resonance (NMR): it pro-
duces shifts, splittings, and broadening of lines in NUCLEAR QUADRUPOLE RESONANCE (NQR)
NMR spectra. Resonance absorption of the energy of a radio-
frequency electromagnetic field by a nucleus with
NUCLEAR QUADRUPOLE RELAXATION a nuclear spin I > 1/2 and an electric quadru-
Establishment of equilibrium between a pole moment Q situated in a microscopically in-
nuclear quadrupole spin system and a crystal lat- homogeneous crystalline electric field. This crys-
tice. As a rule, the related processes are studied tal field produces splittings of the nuclear mag-
with the help of the Pauli equations for the level netic sublevels which are proportional to the value
populations associated with the nuclear quadru- of the quadrupole moment Q, and to the gradi-
pole Hamiltonian (see Nuclear quadrupole interac- ent of the electric field at the nuclear site in the
tion) which characterize the interaction of nuclei crystal (see Nuclear quadrupole interaction). From
with the electric field of their electronic environ- another viewpoint, nuclear quadrupole resonance
ment without taking into account the internal inter- (NQR) involves observing magnetic dipole transi-
nuclear interactions of nucleons with each other. tions between nuclear spin sublevels of the type
The roots of the characteristic equation which en- studied in nuclear magnetic resonance for nuclear
ter a set of Pauli equations are the relaxation times spins I > 1/2, but NQR involves the case when
of energy level populations (commonly called nu- the quadrupole splitting of the spin sublevels is
clear quadrupole spin–lattice relaxation times). larger than the Zeeman splitting of these sublevels
The interaction of the nuclear quadrupole spin sys- (see Zeeman effect). In this case a nuclear quadru-
tem with the lattice which brings about the nuclear pole resonance can be observed in the absence of
quadrupole relaxation is due to the thermal motion an applied magnetic field B. When the magnetic
of molecular groups and particles which, in turn, field B = 0 and the crystalline electric field E has
is the cause of the random time dependence of the axial symmetry the quadrupole energy levels Em
nuclear quadrupole Hamiltonian terms. The most are given by
frequently considered relaxation mechanisms in-
e2 qQ 2
volving thermal motion are the following: the li- Em = 3m − I (I + 1) ,
bration mechanism (rotational vibrations of mole- 2I (2I − 1)
cular groups containing the nuclei under con- where q = ∂E/∂z is the field gradient, and the
sideration); the reorientational mechanism (angu- magnetic quantum number m has the range of
lar jumps of such groups or particles between values −I m I . Another seldom-used name
NUCLEAR RADIATION DETECTORS 905
for this zero field experiment is called “pure- and in the case of monocrystals this leads to either
quadrupole resonance”. a Stark effect shift or a line splitting due to the
The main parameters which characterize nu- nuclear quadrupole resonance dependence on the
clear quadrupole resonance are the following: the crystal field symmetry, and the localization of the
quadrupole interaction constant which determines nuclei involved in the resonance (see Stark effect).
the spectral frequencies; the asymmetry parame- This type of information is useful in studies of the
ter η of the electric eld gradient which mea- nature of chemical bonds.
sures deviations from axial symmetry; the spec-
tral line width; the spin–spin and spin–lattice re- NUCLEAR QUADRUPOLE SPIN SYSTEM
laxation times (see Nuclear quadrupole relaxation). A subsystem of a magnetic resonance spin sys-
These parameters are highly sensitive to the chem- tem in a crystal whose properties are determined
ical and physical structure of solids, the crystal lat- by the isolation of the quadrupole part of the over-
tice dynamics, and external actions such as me- all Hamiltonian. The word “spin” is used in this
chanical, thermal, magnetic, electric, and radia- term since the nuclear electric quadrupole Hamil-
tion actions, as well as to the presence of diverse tonian can be expressed in terms of the nuclear
impurities and lattice defects. Owing to this, nu- quadrupole moment, the nuclear spin operator I ,
clear quadrupole radiospectroscopy is an effec- and the characteristics of the electric eld gradi-
tive technique for investigating the chemical and ent Vij = −∂Ei /∂xj of the electronic environ-
physical state of a solid. This technique is em- ment of a nucleus. For example, the quadrupole
ployed in nuclear physics to determine quadru- Hamiltonian can be written in tensor notation as
pole and hexadecapole nuclear moments, and it H = I QI , where I is the nuclear spin vector op-
is applied in the chemistry and physics of solids erator, and the quadrupole tensor has the compo-
for studying intramolecular and intermolecular in- nents Qij = eQVij /[2I (2I + 1)], e is the elec-
teractions, molecular dynamics and phase transi- tronic charge, Q = ρ(r)(3z2 − r 2 ) dV is the
tions, molecular structure, and the nature of chem- scalar quadrupole moment, and ρ is the charge
ical bonds. There are also various special appli- density. Nuclear spins I > 1/2 have quadrupole
cations of nuclear quadrupole resonance (among moments (see Nuclear quadrupole interaction).
them a thermometry based on nuclear quadrupole
resonance). NUCLEAR RADIATION DETECTORS
Tables of nuclear quadrupole resonance fre- Devices for detecting charged nuclear particles
quencies of a variety of chemical compounds are and gamma rays. In a solid-state nuclear radia-
given in monographs devoted to this method, and tion detector the space charge region of a p–n
in handbooks. or p–i–n junction (i is a region of intrinsic con-
The sensitivity of the nuclear quadrupole reso- ductivity, see Semiconductor junction) is the sensi-
nance method is less than that of NMR. As a rule, tive part. The ionizing particle forms electron–hole
to observe the signal the sample must be of a few pairs within the space charge region. An energy
units or even tens of cm3 in volume. ε equal to several widths of the band gap (e.g.,
A weak magnetic field induces an extra Zee- in silicon ε ≈ 3.6 eV, in germanium ε ≈ 2.9 eV)
man splitting (in addition to the quadrupole split- is spent in the formation of each pair, indepen-
ting), with the appearance of additional spectral dently of the type of radiation. Electrons and holes
lines. By varying the orientation of this field with are separated by the electrical field, and at the ca-
respect to the orientation of a monocrystal one can pacitance of the n–p junction the charge ∝ E/ε
establish experimentally the directions of the prin- (E is the particle energy) accumulates. Thus, the
cipal axes of the electric field gradient tensor with voltage pulse, recorded by the electronic circuit,
respect to the crystallographic axes. In a strong will be proportional to the particle energy if its
magnetic field an ordinary NMR transition takes path is within the space charge region. This is the
place with a superposed quadrupole splitting (Zee- operating principle of the so-called E-detector. If
man splitting > quadrupole splitting). An external the path of the particle is longer than the space
electric field causes a shift of quadrupole levels, charge region width then the signal of the detector
906 NUCLEAR RADIATIONS
is proportional to MZ 2 /E, where M and Z are the the proton yield upon irradiation by fast neutrons,
mass and charge of the detected particle (channel- serves as the detector. Another type of detector of
ing detector, detector of specific energy losses or fast neutrons is based on the radiative damage in-
E-detector). Multiplication of the signals from flicted on a silicon p–n junction. With increasing
E- and
E-detectors provide information about radiation dose the silicon resistance grows and the
the mass of the particle. In order to determine the measured resistance of the p–n junction provides
location of the particle passage, position-sensitive the density of fast neutron flow, and the adsorbed
detectors are used, which simultaneously with the dose within the range 5 to 5000 rad is measured.
energy measurement provide information about
the position of the particle entry into the detector. NUCLEAR RADIATIONS
The charge created by the particle within the de- Fluxes of nuclear particles and gamma rays.
tector is distributed between two contacts located Nuclear radiations result from nuclear reactions
at opposite ends of the rear side, and it depends and the decay of radioactive isotopes. The prod-
on the resistances Rx from the zero contact to the ucts of nuclear reactions are neutrons, protons,
point of the particle’s penetration, and Ry from the deuterons, α-particles, nuclear fission fragments,
penetration point to the second contact. The sig- high-energy charged atomic nuclei, β-particles in-
nal, taken from the front end of the detector, deter- cluding negatively charged β − (electrons) and
mines the total value of the energy E, and the ratio positively charged β + (positrons), as well as
of these signals determines the location of the par- gamma quanta. Gamma rays appear as a result of
ticle penetration. nuclear capture reactions when a nucleus trans-
The surface-barrier E-detectors have a fast re- forms from an excited state to its ground state, and
sponse (∼10−9 s) and they are used for the deter-
when high-energy electrons brake (decellerate) in
mination of the mass of particles by their time of
the fields of heavy nuclei.
flight from the target to the detector. For the spec-
At their interaction with matter, nuclear radia-
trometry of nuclear particles Si-detectors are used,
tions bring about ionization and thereby produce
and for the spectrometry of γ - and X-radiation Ge
positively charged ions and negatively charged
and Si diffusion-drift detectors are used operating
electrons. As a result, nuclear radiations are of-
at the temperature of liquid nitrogen. The energy
ten called ionizing radiations. These are divided
resolution of these detectors is some fractions of
into the ionizing radiation itself which consists
one percent, which is much better than that for
other types of detectors such as ionization cham- of charged particles with sufficient kinetic energy
bers, scintillators, etc. to ionize atoms at collisions (electrons, positrons,
It is possible to detect high-energy (100 to protons, α-particles, etc.), and indirectly ionizing
1000 GeV) nuclear particles by recording their radiation which consists of uncharged particles
transition radiation which appears at the passage (neutrons) or gamma rays and X-rays.
of the particle through the interface of two media At an elastic collision of a nuclear particle with
with different refractive indices (e.g., gas–solid). an atom, when the kinetic energy exceeds a certain
In order to increase the sensitivity nuclear radi- value atoms can be displaced from lattice sites,
ation detectors are used which contain 103 gas– with the resultant production of other types of ra-
solid layers. The intensity of the forward transi- diation defects.
tional radiation is proportional to the incident par- At an inelastic collision when the energy of the
ticle’s energy, and this permits its determination. nuclear particle is greater than the nucleon binding
Surface-barrier silicon detectors are also used energy of the nucleus, then nuclear reactions can
for detecting neutrons. For recording slow neu- accompany the nuclear radiation output.
trons the so-called converter (e.g., the layer of 6 Li,
10 B), within which the (n, α) reaction takes place, NUCLEAR SPIN DIFFUSION
is located at the surface-barrier nuclear radiation Process of transport of nuclear magnetization
detector which records the exiting α-particles. For Mz (nuclear spin polarization Pz ) caused by di-
detecting fast neutrons a hydrogen-containing tar- pole–dipole (d–d) flip-flop transitions of spins of
get (e.g., polyethylene film), which is the source of neighboring nuclei. As long as the mutual spin
NUCLEAR SPIN RELAXATION IN SUPERCONDUCTORS 907
flips do not change the overall Zeeman energy (see ∇ 2 is the Laplacian operator, and c is a known
Zeeman effect) the process of nuclear spin dif- constant. The parameter D is called the nuclear
fusion takes place without involving the lattice. spin diffusion coefficient. In the general case, D is
One should stress that nuclear spin diffusion does a symmetric second-rank tensor whose compo-
not involve any spatial movement of the nuclei; nents depend on the orientation of the external
it only concerns a deviation of the local nuclear magnetic field B 0 relative to the crystallographic
magnetization from its value averaged over the axes. Only for the cases of cubic monocrystals,
sample. In many cases (and always for spin I = polycrystalline samples, and a powder belonging
1/2) one can introduce the concept of a spatial- to an arbitrary crystal system, is D reduced to a
coordinate-dependent nuclear spin temperature Ts scalar. An exact derivation of Eq. (1) was given by
(see Level population), and speak about the equal- L.L. Buishvili and D.N. Zubarev (1965). At suffi-
ization of Ts over a sample through the process ciently low temperatures, the nuclei near paramag-
of spin diffusion. The concept of nuclear spin dif- netic impurities have strongly shifted resonant fre-
fusion was introduced by N. Blombergen (1949) quencies, and this hinders the flip-flop processes
when studying nuclear magnetic relaxation in non- between them. Therefore, there is a diffusion bar-
metallic diamagnetic crystals with small concen-
rier of radius δ near every impurity paramagnetic
trations of paramagnetic impurity ions (see Mag-
ion, inside the barrier D ≈ 0, and outside the bar-
netic ion). In such samples the nuclear spin–lattice
rier D ≈ const. It has been proven analytically and
relaxation at low temperatures is mainly due to the
experimentally that there is a possibility of break-
magnetic dipole–dipole interaction of the nuclear
ing the diffusion barrier under the influence of flip-
magnetic moments of the nuclei with the much
flop transitions of neighboring nuclei which are
stronger (by a factor of 103 ) electronic magnetic
moments of the impurity ions. Since the nuclei of induced by the spin lattice relaxation or the spin–
the host are at various distances from the impurity spin relaxation of the paramagnetic ions.
ions, the nuclei closest to paramagnetic impurities
relax at the highest rate. As a result, during the re- NUCLEAR SPIN RELAXATION IN
laxation process there arises a transient variation SUPERCONDUCTORS
of the degree of magnetization (e.g., by saturation A characteristic time is associated with the
of nuclear magnetic resonance, NMR) throughout change of non-equilibrium nuclear spin magne-
the sample, and the gradient of nuclear polariza- tization in a superconductor. The so-called spin–
tion (spin temperature) Mz becomes position de- lattice and spin–spin relaxation times, T1 and T2 ,
pendent. Thus one might expect a distribution of respectively, are distinguished (see Nuclear spin
nuclear relaxation times, whereas experimentally relaxation time). Nuclear spin relaxation in metals
a single nuclear relaxation time T1n is observed, is caused mainly by the coupling between the nu-
that is the nuclei behave as a coherently intercon- clear and electronic subsystems via the hyper ne
nected system of spins. This process is explained interaction. The dependence of T1 on the tempera-
by the efficiency of the process of nuclear spin ture calculated in the Bardeen–Cooper–Schrieffer
diffusion in nonmetallic diamagnetic crystals con- theory has a non-monotonic character. As the tem-
taining paramagnetic impurities, and it is quantita- perature is lowered below the superconducting
tively described by the Blombergen equation transition temperature T c , the relaxation time T1
∂Mz at first decreases, reaches a minimum, and then
= D∇ 2 Mz − c |r − r m |−6 (Mz − M0 ),
∂t monotonically increases, due to the exponentially
m
(1) decreasing concentration of normal excitations in
where the first term describes the nuclear spin the superconductor. The initial decrease of T1 is
diffusion directed toward establishing a uniform caused by the increase (compared with a normal
magnetization Mz , and the second term describes metal) of the density of electron states at the edge
the relaxation itself which brings about transient of the superconductor gap. A high concentration
local nonuniformities of the distribution; r m is the of impurities and a strong magnetic field smear
radius-vector of mth impurity paramagnetic ion, out the singularity in the superconductor density
908 NUCLEAR SPIN RELAXATION TIME
of states, and lead to the smoothing out or disap- NUCLEAR ZEEMAN EFFECT
pearance of the minimum in T1 . Type II supercon- Splitting of energy levels and spectral lines
ductors in the mixed state exhibit non-exponential of emission or absorption by atoms, molecules,
nuclear spin relaxation caused by different values or crystals under the effect of an external mag-
of T1 within and outside the vortex core. netic field B 0 , arising from the interaction of a
nuclear magnetic moment with B 0 . The com-
NUCLEAR SPIN RELAXATION TIME plete theory of the nuclear Zeeman effect is pro-
A characteristic time for the return of a non- vided by quantum mechanics. The nucleus with
equilibrium nuclear spin magnetization back to its spin I has the related dipole magnetic moment
equilibrium configuration. When a system of nu- μI = γI I (γI is the nuclear gyromagnetic ra-
clear spins in a magnetic field absorbs a pulse of tio). In the field B 0 , the nucleus acquires the Zee-
radio-frequency energy at the Larmor frequency man energy −μI B 0 = γI B0 m, where the pro-
a non-equlibrium distribution of nuclear magne- jection m of I along the direction B 0 can as-
tization is established in the system. The spins sume the 2I + 1 values m = −I, −I + 1, . . . , I .
return to thermal equilibrium by passing this ex- As a result, the nuclear spin energy level is split
cess energy to the random Brownian motion in into 2I + 1 equidistant Zeeman sublevels with
the case of liquids, to the lattice vibrations in the the spacing ωI , where ωI = γI B0 is the nuclear
case of insulators, and to the conduction electrons Zeeman Larmor frequency. Nuclear magnetic res-
in metals (Korringa relaxation). The time con- onance arises from quantum transitions between
stant for this transfer of excess energy is called the nuclear Zeeman sublevels which are induced by a
spin–lattice (or longitudinal) relaxation time T1 . radiofrequency field perpendicular to B 0 , with the
The nuclear spins reestablish thermal equilibrium frequency ω ≈ ωI .
among themselves in the characteristic time T2
called the spin–spin (or transverse) relaxation
NUCLEAR ZEEMAN PSEUDOEFFECT
time. For low viscosity liquids T1 ∼ T2 , and in typ-
ical solids T1
T2 . Phenomenon of splitting of the neutron (n) spin
S = 1/2 energy level into two sublevels upon pas-
NUCLEAR SPIN WAVES sage of a flux of slow neutrons through a polarized
Variety of spin waves due to oscillations of nuclear target. In addition to a weak dipole–dipole
nuclear spin directions. The most convenient ob- interaction, there is a stronger magnetic interac-
jects for observation of nuclear spin waves are the tion between the neutron magnetic moment and
easy plane and cubic Mn-based antiferromagnets the target nuclei, arising from the nuclear forces.
(MnCO3 , CsMnF3 , RbMnF3 , MnO, etc.). The in- The interaction with the target nuclei can be con-
direct coupling of nuclear spins, which gives rise sidered as arising from an effective pseudomag-
to a very strong spin–spin interaction, takes place netic field B ∗ = αP acting upon n from the po-
in these materials through the intermediary of larized target, where P is the polarization of the
magnons, which have a very high effective range nuclei (see Nuclear orientation), and α is a func-
r0 of coupling that reaches 104 a (a is the aver- tion of the scattering cross-section of the neutron
age distance between spins). Due to this high value by the nucleus. As a result, the neutron in a con-
of r0 , the nuclear spin waves may exist at helium stant magnetic field B 0 experiences the effect of
temperature (T 4K), which far exceeds the tem- the total field B 0 + B ∗ , and its Zeeman frequency
perature of nuclear spin ordering (about 10−2 K ωI∗ = γI (B0 + B ∗ ) (assuming B ∗ B 0 ) lies in the
for these materials). Experimental investigations UHF range, where B ∗
B0 , corresponding to an
of nuclear spin waves involve studying nonlinear appreciably large polarization P . Quantum tran-
effects caused by these waves using the technique sitions between Zeeman pseudo-levels (so-called
of nuclear magnetic resonance. pseudomagnetic resonance) can be induced by an
alternating magnetic field. In this case, the effec-
NUCLEAR TARGET, POLARIZED tive alternating field is due mainly to the pseudo-
See Polarized nuclear targets. magnetic contribution.
NUCLEATION 909
911
912 OHMIC CONTACT
the one-dimensional disordered metals may ei- are localized. The simplest quantity to character-
ther be related to randomly located defects of ize the localization is the localization length l that
the crystal lattice, or pertain to a particular com- equals the inverse increment (Lyapunov index)
pound. For example, in a unit cell of crystals of the envelope of the squared magnitude of the
of the KPC type (abbreviated term for the salt wave function. This localization of states drasti-
K2 Pt(CN)4 Br0.3 ·3H2 O), the number of equiv- cally changes the kinetic properties of the system.
alent positions for halogen atoms exceeds the At zero temperature, the low-frequency conduc-
number of these atoms, and the positions are ran- tivity tends to zero along with the frequency. The
domly occupied. The elastic scattering of one- most typical form of such a dependence (Mott law)
dimensional electrons by random lattice inhomo- is σ (ω) ∼ (ωτ )2 ln2 (ωτ ) (with τ ∼ l/vF , where
geneities leads to the phenomenon of localization vF is the Fermi velocity). In one-dimensional dis-
of all electron states in a one-dimensional disor- ordered systems of finite length L, the static elec-
dered metal, and its transformation into a zero-gap trical conductivity is exponentially small in the pa-
dielectric with a high degree of polarizability. The rameter L/ l.
charge transport in such systems is effected by the
hopping of electrons between localized states with ONE-DIMENSIONAL SUPERCONDUCTIVITY
energies differing by the magnitude of a quan- See Superconductivity in quasi-one-dimen-
tum of the external field of frequency ω. The low- sional systems.
frequency conductivity of localized electrons at
zero temperature is proportional to ω2 ln2 (ωτ ), ONSAGER METHOD (L. Onsager)
where τ is the mean free path time. At a fi- Method for calculating the statistical sum of
nite temperature, by virtue of inelastic scatter- the Ising model for a planar N × M lattice in the
ing by phonons, the electronic states become absence of an external field. The Onsager method
delocalized, and nonzero static conduction ap- is based on
the representation of the statistical sum
pears. Examples of organic molecular compounds as Z = σ (V N )σ σ , where the transfer matrix
with quasi-one-dimensional metallic type electri- V = V1 V2 . The matrix V1 describes the spin in-
cal conductivity are salts of 7, 7, 7, 8-tetracyano- teractions of a lattice row with a neighboring one:
p-quinodimethane, called TCNQ for short. Salts
of bis(ethylenedithia)tetrathiafulvalene, called V1 = exp K σn,m σn+1,m ,
BEDT-TTF for short, exhibit low-dimensional m=1,...,M
electrical conductivity. and the matrix V2 describes the interaction of ad-
jacent spins in one and the same row:
ONE-DIMENSIONAL MODELS OF
DISORDERED STRUCTURES V2 = exp K σn,m σn,m+1 .
Theoretical models that allow one to write m=1,...,M
closed form equations for one or another dynamic
quantity, and thus obtain equations (Smoluchowski The quantity Z is expressed through
the eigenval-
equation or Fokker–Planck equation) for probabil- ues λi of the matrix V , i.e. Z = λN i ≈ λmax
N
ity densities of these quantities. Such equations where λmax is the largest eigenvalue. To calcu-
can be solved or investigated in detail for many late λmax , the commutation relations of the ma-
important cases. trices V1 and V2 are used. Due to the presence of
The richness in the physical content of one- translational symmetry along the rows, the prob-
dimensional models of disordered structures is due lem simplifies through the application of a Fourier
to the fact that these models express so clearly transform, and the following relation is found
the principal difference between disordered sys- for λmax :
3 4
tems and ordered ones, namely, the presence of π dq
λmax = (2 sinh 2K) M/2 exp − εq ,
a macroscopic number of localized states of an
−π 4π
essentially non-Bloch type (see Bloch theorem):
all states in one-dimensional disordered structures cosh εq = cosh 2K coth 2K − cos q.
914 ONSAGER THEORY
ONSAGER THEORY
A theory in the physics of insulators, which es- Diagram of the polarization directions of two cases of
tablishes relations between the static εs and high- opposed domains.
frequency ε∞ dielectric constant of a medium
containing polar molecules (with an intrinsic di-
pole moment). Account is taken of the differ- in magnetic substances they can be compensated
by strong nonuniform external magnetic fields.
ence in the dipole moment of a molecule in a
medium and in vacuo, resulting from interactions OPTICAL ABSORPTION SPECTROSCOPY
with neighboring dipoles, and the difference be- Measurement of the absorption or reflection of
tween the electric field in the medium before and electromagnetic radiation in the infrared (IR), vis-
after applying an external electric field E. A mole- ible, and ultraviolet (UV), frequency ranges to es-
cule is treated as a sphere with an isotropic polar- tablish the energy level structure of materials.
izability. The Onsager equation is as follows: The absorption spectra of solid insulators and
semiconductors in the UV, visible, and near IR
4πμ2v N0 3εs ε∞ + 2 2
ε s − ε∞ = , correspond to interband and impurity electron
3kB T 2εs + ε∞ 3 transitions, while in the middle and far IR they
(1)
arise from crystal lattice vibrations, and possi-
where N0 is the concentration of polar mole-
bly from the presence of free carriers and im-
cules. The Onsager theory disregards deviations
purities. Metals are opaque in the visible and
from a macroscopic description of the interac-
transparent in the UV with the cross-over be-
tion between nearest neighbor molecules. Accord- tween these two regions given by the plasma fre-
ingly, the correction reduces to adding the factor quency ωp = [ne2 /(ε0 m)]1/2 which is typically
(1 + zcos γ ) into the right-hand side of Eq. (1).
∼1015 Hz. Absorption spectra are investigated ex-
Here, z is the average number of nearest neighbors
perimentally with the aid of spectrometers, spec-
of the molecule, cos γ is the mean value of the
trophotometers, spectrographs, monochromators,
cosine of the angle between neighboring dipoles.
etc. Fourier-transform spectrometers (see Fourier
Eq. (1) with the added factor (1 + zcos γ ) is
spectroscopy) which employ a Michelson inter-
called the Kirkwood formula (see Kirkwood ap- ferometer are widely used in the infrared range.
proximation).
Examples of detectors are photoplates, photocells,
photomultipliers, photoresistors, thermoelements,
OPPOSED DOMAINS bolometers, and acoustic-optical receivers. Exam-
Domains (magnetic, ferroelectric, etc.), charac- ples of radiation sources are lasers, heated fil-
terized by a jump of the polarization (magnetic, aments, gas-discharge and arc-discharge lamps,
electrical, etc.) component, normal to the domain and laser sparks arising under the action of pow-
wall, e.g., polarization can reverse direction in op- erful laser pulses on a material. Cherenkov ra-
posed domains (see Fig.). Usually such domains diation and synchrotron radiation are also avail-
are energetically unfavorable because of the ap- able. The transmission spectrum of the sample
pearance of demagnetizing (depolarizing) fields. under study is often compared to that of a refer-
However in some cases they can form, e.g., in fer- ence spectrum. The absorption generally follows
roelectrics with high conductivity, so that the the the Beer–Lambert law of exponential decay with
bound charges at the walls, creating the high fields, material thickness. Typical spectra plot absorp-
can be compensated by incoming free charges; and tion versus radiation frequency (or wavelength). If
OPTICAL BISTABILITY 915
the absorption spectrum is measured over a wide medium. The feedback arises during the reflection
enough frequency range then the optical conduc- of light from the interferometer mirrors. The inter-
tivity can be determined by the Kramers–Kronig ferometer transmittance in the absence of absorp-
relations. tion is determined by the Airy function:
OPTICAL ACTIVITY IT 1
T= = , (1)
See Rotation of light polarization plane. I0 I + F sin2 Φ/2
Optical bistability. (a) Graphical solution of Eqs. (1) and (2); (b) hysteresis of optical memory mode: the “on” pulse
switches the optically bistable element from state 1 to state 2, whereas the “off” pulse switches from 2 to 1 (I is the
intensity of memory storage); (c) optical transistor mode showing signal amplification.
resonance dynamical nonlinearities in semicon- cal processes that determine the bistable behavior
ductors, which are determined by specific fea- of systems: the feedback provides the possibility
tures of absorption and refraction of laser radiation to detect a nonlinear process under relatively low
at resonance transitions, at resonance excitation excitation levels.
of excitons, biexcitons, bound excitons, impurity
states, and so on. These nonlinearities (called giant OPTICAL BREAKDOWN, light breakdown,
ones) are several orders of magnitude larger than optical discharge, laser spark
the corresponding classical nonlinearities in non- Transformation of matter to a plasma state as
absorbtive media. The latter nonlinearities de- a result of intense ionization under the action of
termine such processes as the second harmonic electromagnetic fields at optical frequencies. In
generation, parametric generation (see Optical optical breakdown a spark seen as a bright flash
parametric oscillator), and so on, and manifest emerges in the focal point of a lens, and is accom-
themselves in intense light fields comparable to panied by a strong pulse of sound. In irradiated
intratomic ones. When using giant resonance non- optically transparent solids the optical breakdown
linearities, one should take into account the dy- can occur at the entrance or exit surface, or in the
namical properties and the relaxation (lifetime) of sample bulk, depending on the lens focal point.
excited states, which control the speed of response The latter case requires the highest intensity of fo-
of devices based on the optical bistability effect. cused radiation. The plasma of a spark in the bulk
The switching-on time of a device is related to the expands in space in such way that the glow front
time τ c of the field relaxation in the interferometer drifts towards the path along the normal to the sur-
(if L = 10 μm, τ c < 10−12 s), and to the intensity face at a speed of 107 cm/s. The glow spectrum
of the exciting light I0 (i.e. to the value of the in- of the plasma formed in the bulk of the condensed
duced nonlinearity), while the switching-off time matter is continuous, and that near the surfaces
is determined by the lifetime of excited states. is continuous with a superimposed line spectrum
Optical bistability is of interest both from the emitted from regions of lower plasma density.
practical viewpoint: for the production of optical Optical breakdown in condensed matter where
transistors, ultraspeed selectors, storage elements, power laser radiation propagates can be the cause
and other optoelectronic devices (see Optoelec- of a material failure. The evaporation of the
tronics) where the light beam replaces an elec- solid-state target surface for the case of sym-
tric current; and for investigating nonlinear opti- metrical irradiation stimulates uniform compres-
OPTICAL DETECTION OF MAGNETIC RESONANCE 917
sion and heating of the material to high temper- point source, and its shadow image is examined on
atures, with the formation of a high-temperature a remote screen. The sensitivity is very high, and
(≈107 K) dense plasma where thermonuclear re- allows detecting a heterogeneity that produces a
actions might be possible. path difference of 0.01λ at the crystal boundaries.
Optical inhomogeneities are easily detectable
OPTICAL COMPENSATOR by interference methods in Michelson, Rayleigh,
A device for introducing a certain optical path- and Jamin interferometers. The essence of inter-
length difference for two coherent light beams, or ference methods is the fact that the refractive in-
for reducing an existing path-length difference to dex of any inhomogeneity differs from that of the
zero, or to some constant value. As a rule, an op- host crystal. The refractive index difference intro-
tical compensator is equipped with a reading de- duces an optical path difference between the in-
vice to facilitate its use as a path-length difference terfering beams, causing the interference lines to
meter. There are two main types of optical com- bend. The magnitude of the bending allows one
pensators. Interference optical compensators are to judge the geometrical dimensions of inhomo-
used in double-beam interferometers for equaliz- geneities, and to determine the refractive index of
ing the optical path-length difference of interfering impurities. In crystals with microscopic refraction
beams, e.g., by adjusting the inclination of parallel index inhomogeneities, the latter are observed us-
plates in the beam path to a certain angle. Polar- ing microinterferometers.
izing optical compensators are used for the analy- Polarization methods are especially sensitive
sis of elliptically polarized light. The path-length to stress of any kind in a crystal lattice. To ob-
difference in this case is adjusted by varying the serve stress-induced birefringence, a crystal is po-
thickness of plates cut from a birefringent crystal sitioned between crossed polarizers.
with differing velocities of the ordinary ray and the
extraordinary ray. Optical compensators are used OPTICAL DETECTION
to study stresses in transparent materials with the The same as Pockels effect.
help of polarized light; in crystal optics in combi-
nation with the polarizing microscope, and also to OPTICAL DETECTION OF MAGNETIC
change and analyze light beam polarization. RESONANCE
Methods for detecting magnetic resonance op-
OPTICAL DEFECTOSCOPY tically during microwave saturation of spin sub-
Application of optical methods to determine levels. Applied to solids, this technique is similar
crystal structure perfection. As distinct from to the optical double magnetic resonance method.
X-ray defectoscopy (see Defectoscopy) that estab- There are two variations of optical detection tech-
lishes the nature and concentration of defects at niques. One of them uses incident circularly po-
the atomic level, optical defectoscopy yields data larized pumping light, and detects effects that are
on the quality of the crystal lattice as a whole. in some way related to the circular polarization of
The investigation methods used for this purpose the emergent radiation (optical pumping method).
are diverse: luminescence, Raman scattering of Another variety does not necessarily involve the
light, light absorption, as well as shadow, polariza- light polarization: these techniques are based on
tion, and interference methods. By analyzing the the spin-dependent nature (see Spin-dependent ef-
intensity, line width, and frequency of the lines fects) of the optical properties of solids (e.g., re-
observed in spectra, and by comparing the crys- combination luminescence). When applying the
tals under study with standard samples, one can process of optical detection, a sample under study
judge the perfection and block (granular) struc- is placed in a resonator of an electron paramag-
ture of a crystal lattice. The crystal blocks (grains), netic resonance spectrometer, and is subjected to
the traces of slip and other weak local inhomo- the simultaneous action of optical pumping and
geneities with clearly pronounced boundaries are microwave saturation. As a rule, the microwave
easily detected using the shadow method, whereby power is amplitude-modulated, and the optical de-
a crystal is inserted in a divergent beam from a tection is performed at this modulation frequency.
918 OPTICAL EXTINCTION
An advantage of optical detection relative to elec- radiation lines arriving from outer space. Among
tron paramagnetic resonance is the significantly the most widely used optical filters the best para-
higher sensitivity (approximately by the ratio be- meters are possessed by interference filters with
tween the light and microwave frequencies, i.e. multilayer dielectric mirrors that constitute a half-
∼104 ). This provides the possibility of studying wavelength layer, in which the transmission band
the radio spectroscopic characteristics of excited is formed as a result of the interference of the
states of paramagnetic centers, and excited triplet light.
states involving exceedingly small quantities of
short-lived intermediate radicals, i.e. photochem- OPTICAL FREQUENCY CONVERTER
ical reaction products. In the context of semicon-
Device based on nonlinear-optical crystals
ductors, the optical detection method allows one to
where laser radiation is converted into light of
identify unambiguously the pairs of particles (e.g.,
another frequency. Crystals without a center of
donors and acceptors) that are responsible for var-
symmetry exhibit a nonlinear polarization compo-
ious bands of recombination luminescence.
nent proportional to the square of the light wave
The addition of an extra radio frequency source
electric field amplitude. The three-photon interac-
transforms the optical detection technique into that
tion occurring in such crystals (ω3 = ω1 + ω2 )
of electron–nuclear double resonance. This makes
allows one to perform light frequency doubling,
it possible to detect transitions between nuclear
and the generation of sum and difference waves.
spin sublevels in an optical channel with a sensi-
The most efficient transfer of laser radiation en-
tivity characteristic of the optical detection tech-
ergy takes place if the condition of phase synchro-
nique, and a resolution characteristic of the nu-
nism k3 = k1 + k2 is met (see Second harmonic
clear magnetic resonance technique.
generation). More than 90% of the laser energy is
converted in frequency doublers (optical harmonic
OPTICAL EXTINCTION
generators). Nonlinear-optical generators of sum
See Luminescence quenching.
and difference waves find use as coherent radia-
tion sources in the IR and UV regions of the op-
OPTICAL FILTERS
tical spectrum. A pride of place among them be-
Devices that change the intensity of radiation
passing through them, and the distribution of in- longs to optical parametric oscillators that provide
tensity over the spectrum. There are absorption, a smooth frequency conversion.
interference, and interference-polarization optical When a nonlinear crystal is used to mix laser
filters, depending on how the light interacts with radiation with IR radiation that carries information
the working elements of the solid-state filters. The on a subject or a process, there results an upward
two last types are multilayer systems where the in- frequency conversion. This can be used to trans-
terference of natural or polarized radiation takes form the infrared image to a visible spectrum so it
place and, as result, a narrow spectral band is iso- is recorded with the help of ordinary methods. Us-
lated. ing several successive nonlinear crystals, one can
Solid-state absorptive optical filters can be repeat the operations of frequency doubling, addi-
neutral-gray (reduce light beam uniformly over tion, and subtraction, and thus obtain sources of
a broad spectral region) and selective (cut-off coherent radiation at combination frequencies, in-
or band type). The cut-off optical filters ab- cluding high harmonics.
sorb a wide short-wave or long-wave part of In centrally symmetric crystals with a cubic
the spectrum and pass adjacent regions, while nonlinearity, the frequency conversion is achieved
selective filters isolate a certain narrow spec- via four-photon interactions. Such crystals are use-
tral band. High selectivity is a salient feature ful in the direct generation of third-harmonic light
of interference-polarization optical filters whose waves, and the two-photon pumping of paramet-
bandwidth can be 10−2 nm. Such optical filters ric amplifiers and generators. To convert the fre-
are labor-consuming to manufacture, and find use, quency of intense light waves, use can be made
mainly, in astrophysics for isolating narrow atomic also of the process of induced light scattering.
OPTICAL NUCLEAR POLARIZATION 919
sample, and on its orientation. This system con- prove to be diffusive because of viscosity. The
sists of a crossed polarizer and analyzer with the scattering becomes still greater by a few orders
sample under investigation placed between them. of magnitude during the emergence of turbulent
Usually it is not the absolute value of the inten- flows in a regime of dynamic light scattering.
sity of the light transmitted by the system that is • Giant optical nonlinearity (roughly 9 orders of
determined, but rather its variations relative to the magnitude higher than that in carbon bisulfide)
center of the interference pattern (see Conoscopy), is due to director rotation under illumination.
or a study is made of the interference picture
which displays the distribution of internal stresses Unique optical properties arise in LC with a he-
of the object under study. With the aid of the com- lical structure (cholesterics, see Cholesteric liquid
pensator of birefringence, the rotation angle of crystal, and chiral smectics C): (i) enormous opti-
the analyzer up to the transmission minimum is cal activity (about 104 deg/mm), and (ii) selective
measured, and using this angle the birefringence light reflection (also seen in blue phases) due to
is calculated. The optical polarization method is Bragg diffraction from the helical structure. Since
utilized in electro-, piezo-, magneto-, acousto- the helix pitch is temperature sensitive, this phe-
optical, measurements for investigating stresses in nomenon finds wide use in thermography in engi-
transparent materials. neering and medicine.
(2) Time constant – time of signal growth up to form a new class of multichannel analyzers with
a specified value, characterizes the rapidity of improved sensitivity, dynamic range, and fast re-
the response of the optical radiation detector. sponse in real time compared to the photochemi-
(3) The transformation ratio determines the value cal multichannel optical radiation detectors in the
of the output signal, normalized to the value form of different photosensitive layers, as used in
of detected input radiation flux; it is measured photography.
in volts per watt or in amperes per watt.
(4) The spectral sensitivity is the dependence of
OPTICAL RADIATION, SCANNING
the output signal of the detector on the wave-
See Scanning of optical radiation.
length of incoming radiation. Thermal optical
radiation detectors usually have a broad spec-
tral sensitivity range, and they are called non- OPTICAL RADIATION SOURCES
selective receivers of optical radiation. Pho- Devices for conversion of various kinds of en-
tonic optical radiation detectors operate, as a ergy into electromagnetic irradiation in the opti-
rule, in narrow ranges of the spectrum, and are cal range within the conventional wavelength lim-
very selective. its (in vacuo) λ = 100 nm to 1 mm (3·1015 –
(5) Dynamic range is the region of values of 3·1011 Hz). Solid state optical radiation sources
recorded input optical radiation flux, where can be classified into three groups according to
the output signal is proportional to the input. their mode of generating radiation:
Thermal optical radiation detectors and photo-
diodes in the current generator mode possess (1) The temperature (thermal) mode converts the
a wide dynamic range. heat motion of particles of a solid directly
(6) Operating temperature of the sensing element. into electromagnetic radiation energy. The
commonest are heat sources of optical radia-
The main thermal optical radiation detectors tion, viz., incandescent lamp, globar, Nernst
are bolometers, thermocouples (thermoelements glower, platinum-ceramic sources heated by
and thermostacks), pyroelectric radiation detectors
an electric current, and electrodes raised to
and optico-acoustic (pneumatic) receivers (see
a white glow in a gas discharge (such as a
Acousto-optic devices); they have the threshold
carbon arc). As a rule, thermal optical ra-
sensitivity 10−8 to 10−12 W/Hz1/2 and a time
diation sources can generate both a discrete
constant 1 to 10−3 s. Superconducting bolome-
and a continuous spectrum with the equiva-
ters are 100 to 1000 times superior according
lent brightness temperature of a black body
to their sensitivity and fast response. For mea-
T b 4000 K.
surements of very high fluxes of optical radiation
calorimeters are used (see Calorimetry). The fol- (2) The luminescence mode uses the recombina-
lowing devices are photonic optical radiation de- tion radiation of electronic and intramolecu-
tectors: photocells, photodiodes, photomultiplier lar vibrational states of crystals and glasses
tubes, photoresistors with threshold sensitivity (luminophors). These states can be excited ei-
up to 10−12 W/Hz1/2 and time constant up to ther with electric current (see, e.g., light emit-
10−10 s. Tunnel photodiodes with microanten- ting diodes) or a flux of various particles:
nae may operate with a time constant 10−14 s. γ -quanta, photons (photoluminescence), elec-
Photoelectronic multipliers and avalanche pho- trons (cathodoluminescence), etc. Lumines-
todiodes have their threshold sensitivity up to cent optical radiation sources can generate
10−17 W/Hz1/2 . both discrete and continuous spectra, and op-
There are multichannel optical radiation de- erate in both continuous and pulse modes us-
tectors based on arrays of photodiodes combined ing single or periodic pulses of flashes with
into shift registers with charge coupling (charge- durations longer than several nanoseconds.
coupled devices), and also receivers of the televi- The equivalent brightness temperature of a
sion tubes type, such as super silicon diode array black body for continuous luminescent optical
cameras. Multichannel optical radiation detectors radiation sources can reach 10,000 K.
OPTICAL RECORDING MEDIA 923
The light sensitivity for direct recording on op- or light pipes, lm s, etc.), and in resonant media.
tical media is lower by several orders of magnitude The existence of optical solitons in optical single-
than that of silver-halide materials with chemi- mode light pipes was predicted by F. Tappert and
cal development. Only in photothermoplastic ma- A. Hasegawa (1973); and was observed by L. Mol-
terials (see Thermoplastic materials) and photore- lenauer et al. (1980). The optical soliton in a light
fractive materials (see Holography) recorded in an pipe corresponds to the so-called envelope soli-
electric field (enhancement of the primary reac- ton of a light wave. Soliton modes may appear
tion), does the sensitivity approach that of ordinary in a light pipe when there is mutual compensa-
photographic media (see Table 1). tion for the effect of an increase of slope of the
Optical recording media find use in recording wave front because of nonlinearity, and there is
holograms and discrete information in real time. wave packet spreading related to dispersion (see
Controlled spatial light modulators have been de- Wave packet representations of wave functions).
veloped for optical filtration (see Optical lters). The nonlinear properties of light pipes are related
Optical recording media are also applied to the to the deformation of the electronic shells of glass
molecules (see Vitreous state of matter) under the
production of holographic optical elements such
action of the electric field of the wave. The dy-
as interference filters, holographic lenses, dif-
namics of propagation of the envelope soliton is
fraction gratings, etc. Amorphous chalcogenides
described by the nonlinear Schrödinger equation,
are applied as master optical storage disks when
in which the role of time is played by the co-
manufacturing secondary disks (so-called com-
ordinate along the fiber direction. Dark solitons
pact disks) for reproduction of audio- and video- arise, if the light pulse propagates over a region
recordings. Also, reversible optical disks have of normal dispersion of light; bright solitons oc-
been developed for computers (to replace mag- cur if the propagation is over a region of anom-
netic media in ROM systems), and for playback alous dispersion of light. Dark solitons are dips in
devices for audio and video recordings. intensity of quasi-continuous pumping, or a long
pulse; bright solitons are individual localized wave
OPTICAL SOLITON packets. Typical parameter values for the genera-
Self-localized nonlinear excitation (see Soli- tion of a single bright optical soliton in a light pipe
ton) of an electromagnetic field. An optical soli- are: wavelength λ = 1.55 μm, input pulse duration
ton occurs in media that exhibit a linear relation- 70 ps, light guide cross-section S = 100 μm2 , peak
ship between the refractive index and the inten- power 1 W. Optical solitons show promise for ap-
sity of the traveling wave (in fiber-optics tubes plications to fiber-optic information cables with
OPTICAL SPECTROSCOPY OF SURFACES 925
extra-high transmission rates, logic gates, soliton There are both direct band-to-band transitions
lasers. and indirect band-to-band transitions. The former
arise without the participation of phonons, and
OPTICAL SPECTROSCOPY OF SOLIDS are characterized by high values of the absorp-
The branch of spectroscopy concerned with the tion coef cient (104 –105 cm−1 ). The latter ap-
study of light absorption, re e ction of light, lumi- pear with the participation of either phonons or de-
nescence, Raman scattering of light, and photo- fects, and have absorption coefficients lower by 2
conductivity optical spectra of solids, the deter- to 3 orders of magnitude. The interaction of ex-
mination of their energy level configurations, and citons with photons results in the formation of
their suitability for various applications. The the- photoexcitons (polaritons). The edge and exciton
ory underlying the analysis of the above spectra absorption, luminescence, reflection, and photo-
is the quantum theory of solids and group the- conductivity spectra of semiconductors and insu-
ory. The latter takes into account the symmetry lators occupy the regions of near-IR, visible, and
properties of a crystal lattice (and, consequently, UV radiation. Electronic transitions in molecular
of wave functions) and establishes selection rules crystals, and in particular dyes, are found in the
for quantum transitions. Spectroscopic data pro- same spectral region.
vide an opportunity to judge the nature of the in- For electronic transitions with the participation
teraction between a light wave and a crystal, the
of impurity states formed by atoms of rare-earth,
energy migration in a crystal, the nature of defects,
transuranium and transition elements, very nar-
the centers of radiative recombination of electrons
row absorption and luminescence bands can ap-
and holes, the presence of photochemical transfor-
pear even at high temperatures. They have a great
mations, the influence of external factors (temper-
number of components due to the splitting of de-
ature, pressure, electric and magnetic fields) on the
generate states in the crystal e ld, and are related
energy bands and discrete levels, and the role of
with the inner (e.g., 4f n for rare-earth elements)
surface states. Under high (laser) levels of exci-
electronic shells of impurity atoms.
tation, multiphoton processes and nonlinear opti-
In addition to quantum transitions between the
cal effects, including stimulated Raman scattering
and stimulated radiation emission, become feasi- electronic states that result from vibrations of
ble. Optical spectroscopy methods are quite in- atoms, ions, or molecules that compose a crystal,
formative for the spectral analysis of solids. Any there can be a vibration-induced change of the di-
spectrum depends on the chemical composition, pole moment and the polarizability of the unit cell.
the crystal structure, the presence of defects in a Also, IR absorption and Raman scattering may ap-
crystal lattice, their nature, etc. pear. For crystals with a center of symmetry in the
In accordance with the band theory of solids, unit cell some of the lattice vibrations appear in
quantum transitions between the allowed bands, the IR spectrum, and others are found in the Ra-
intraband transitions, those with participation of man spectrum (see Mutual exclusion), as permitted
defects of various kinds, and quasi-particles, are by selection rules. The lattice spectra of molecular
possible. Band-to-band transitions produce broad crystals reveal the rich structure of intramolecular
absorption spectra with a sharp cutoff (intrinsic vibrations. In the case of anisotropic lattices, all
light absorption edge) on the long-wavelength spectra are polarized.
side. Adjacent to these spectra are impurity-band,
exciton absorption, and luminescence spectra. The OPTICAL SPECTROSCOPY OF SURFACES
latter result from the electron–hole Coulomb in- Studies of solid surfaces and phase boundaries
teraction. Typical of excitons with a large radius by optical methods. The optical spectroscopy of
(Wannier–Mott excitons) is a set of narrow bands surfaces is used mainly to investigate the molecu-
forming a hydrogen-like (Rydberg) series. Its edge lar structure of the near-surface layer, its changes
coincides with that of intrinsic absorption, while during the course of chemical reactions, and the
the most outlying band is at a distance correlated adsorption process at the surface (IR absorption,
with the binding energy of an exciton, and deter- IR reflection, multiple total internal re ection);
mines its ground state. specific features of surface quasi-particles such as
926 OPTICAL SPIN ORIENTATION
surface plasmons, surface polaritons, and excitons duction of audio and video data or recordings
(attenuated total internal re ection, Raman scat- (Fig. 1(a)). The digital master record is usually
tering of light, surface laser photoluminescence, made with a semiconductor laser. The diameter
reflection spectroscopy); the rate of surface re- of a single bit (a dot) of a record on an optical
combination, surface state parameters, and the storage disk is about 1 μm (Fig. 1(b)), which al-
planar band potential (IR absorption, electrore ec- lows the recording of about one billion data bits
tion, photoluminescence); the structure of energy on a disk 12 cm in diameter (equivalent to 50,000
levels in heterojunctions, superlattices, and quan- pages of text, or a 20 min color video recording
tum wells (IR absorption, Raman scattering, pho- with sound). Master recording on the disk usually
toluminescence). makes use of optical recording media designed for
As a general rule, optical spectroscopy of a real-time operation. The master disk is copied on
surface is not associated with a strong interaction to secondary ones that are plastic plates with a sys-
there, and it often yields the most direct data on tem of small holes (about 1 μm in diameter). Sub-
surface parameters (e.g., energy levels in quantum strates for optical storage disks are mostly plastic
wells and superlattices). This presages the wide or glass. Each disk has a required protective coat-
use of optical spectroscopy in investigations of ing. Those used in computers are 36 cm in diame-
physical and physicochemical processes near the ter. The development of read–write optical storage
surface. disks began in 1983–1985 (see Information record-
ing in solids).
OPTICAL SPIN ORIENTATION
Appearance of anisotropy of the spatial dis- OPTICAL TECHNIQUES OF INFORMATION
tribution of spins of charge carriers excited by RECORDING
circularly polarized electromagnetic radiation in Light-induced residual changes of physico-
solids. The incident circularly polarized radiation
chemical properties of materials (see Optical re-
possesses a finite momentum, and by virtue of its
cording media). Optical techniques of information
conservation the charge carriers that emerge af-
recording are applied to the recording of images,
ter its absorption in a semiconductor should have
discrete (digital) information, holograms and so
the same momentum. Since the angular momen-
on, and can be implemented in two-dimensional
tum of free electrons in a semiconductor is re-
and three-dimensional media. The limit of the
lated to the presence of spin, the spins of the light-
recording density is set by the conditions of dif-
generated electrons will have a prevailing orienta-
fraction and is as high as λ−2 bit/cm2 in two-
tion along the radiation propagation direction. The
dimensional media and λ−3 bit/cm2 in three-
spin–orbit interaction can also result in the orien-
dimensional ones (here λ is the wavelength).
tation of holes. A similar effect arises when po-
The theoretical limit of recording density has
larized light is absorbed by a system of atoms,
been attained in two-dimensional systems (at
and the resultant angular momenta of the atoms
λ = 632.8 nm it is 2.5·108 bit/cm2 ).
are oriented. Experiments reveal an optical spin
orientation owing to the presence of circular po- Optical techniques of information recording
larization of the recombination radiation of spin- use changes in optical constants (absorption in-
oriented electrons and holes. In GaAs, the maxi- dex κ, refractive index n) or profile
(dn) (d is
mum polarization achieved is 90%. This value de- the thickness) of the medium (see p. 928, Fig. 2).
creases as a result of relaxation of the oriented Depending on the relation between changes in
electrons and holes during the time delay between optical constants, one can distinguish amplitude
excitation and emission of the radiation (see Para- |κ1 − κ|
|n1 − n| and phase |n1 − n|
Fig. 1. (a) Optical storage disk, and (b) details of localized region A.
The design of optical information recording are constructed as complicated multilayer systems
techniques is linked to the production of lasers and containing a light-sensitive layer, various sublayer
optical storage disks. Here, the recording process coatings, and a substrate. Of interest is the so-
itself runs in real time (1 to 100 ns for RAM), called bubble recording when the light absorbed
whereas the information recorded in ROM must by the medium also causes heating of the sub-
be stored for years. These optical techniques are layer with subsequent gas liberation (formation of
used for both one-time and repeated (reversible) a bubble). This makes it possible to enhance the
recording. In the latter case, optical erasure is used photosensitivity and improve the detection para-
for a new recording. As a rule, recording media meters.
928 OPTICAL THEOREM
OPTICAL WAVEGUIDE
A dielectric structure along which electromag-
netic radiation of an optical spectrum can propa-
gate with low loss. The commonest types involve
ber optics, such as fiber light guides, and various
kinds of optical waveguides for integrated optics
(planar, channel, etc.).
OPTICO-ACOUSTICS
See Acousto-optics.
OPTICS, NONLINEAR
See Nonlinear optics and Nonlinear-optical crys-
tals.
OPTICS, NUCLEAR
Fig. 2. Three types of optical information recording grat- See Nuclear optics.
ings: (a) amplitude type with thickness d, separation s,
and absorption indices κ1 and κ, (b) phase type with in-
OPTICS OF METALS
dices of refraction n1 and n0 , and (c) profile-phase type.
See Metal optics.
OPTICS OF SOLIDS
OPTICAL THEOREM
Part of physical optics studying processes of
A relation connecting the imaginary part of
light propagation and light interaction with solids.
the scattering amplitude f (k, Θ) for a particle or
This includes polarization of light, birefringence,
quasi-particle of wave vector k at a zero angle
interference, diffraction of light, optical activity
(Θ = 0), with the total (elastic or inelastic) scat- (see Rotation of light polarization plane), disper-
tering cross-section σ tot of the particle, σ tot = sion of light, light absorption, magnetooptics, elec-
(4π/k)Imf (k, 0). trooptics, etc. The basis of the phenomenologi-
cal theory of the optics of solids, which reflects
OPTICAL VIBRATIONS of crystal lattice the wave aspect of radiation, is Maxwell’s equa-
Branches of normal vibrations that exist in crys- tions. The coefficients of these equations are the
tals containing more than one atom per unit cell. material constants of matter: the relative dielec-
Long-wave displacements of vibrating atoms (or tric constant ε and magnetic permeability μ which
molecules, or ions) occur with no shift in the cen- are related to the crystal lattice structure, and de-
ter of gravity of individual unit cells (with wave termine its refractive index n = (εμ/ε0 μ0 )1/2 ,
OPTOELECTRONICS 929
where ε0 and μ0 are the free space (vacuum) val- plasma oscillations, starting at frequencies much
ues. All three parameters, n, ε, and μ, are func- higher than those of highly-doped semiconductors.
tions of frequency and wave vector. Most crystals The exciton spectra that are usually adjacent to the
are anisotropic, so these three parameters depend intrinsic light absorption edge are in many cases
on the light propagation direction (see Anisotropy superimposed on the electron band-to-band spec-
of crystals). In nonmagnetic solids μ ≈ μ0 . tra. The electronic spectra of molecular crystals re-
All crystals are characterized by their intrin- produce, to a great extent, the spectra of the indi-
sic absorption spectra which are determined by vidual molecules that they contain, and are situ-
the structure of the energy bands of electrons ated predominantly in the UV and visible regions.
and phonons. The electronic absorption spectra Very rich spectra of intramolecular vibrations are
are related to the resonance band-to-band transi- found in the IR.
tions of electrons as well as to intraband transi-
tions and, as a rule, are situated in the UV, visi- OPTICS, VIBRONIC EFFECTS
ble and near-IR regions. The absorption of phonon See Vibronic effects in optics.
radiation takes place in the middle and far-IR re-
gions. The strongest absorption over broad optical OPTOELECTRONICS
spectral regions pertains to metals owing to their A technological scientific discipline where a
high free-carrier concentration. The band-to-band combination of electronic and optical methods
transitions of metals are, in most cases, in the UV and means is used for processing, transmission,
and perhaps visible regions, so many good met- recording, storage, and presentation of informa-
als are opaque in the visible. A high transparency tion. The main feature of optoelectronics is the
of metals is found at frequencies above that of presence of an optical photon connection to bring
Fig. 1. Diagrams of optoelectronic transformations of: (a) an electrical signal ϕ(J, V ) which is a function of current J and
voltage V , and (b) an optical signal ϕ(Phν , λ) which is a function radiation density Phν and wavelength λ. The systems
include a radiation source (RS), a photon detector (PD) and an optical medium (OM) with electrical ϕ0 (J, V ) and optical
0 , λ) input reference signals. The indicated transformations are electron-to-photon (EP), photon-to-electron (PE),
fc0 (Phν
electron-to-optical (EO) and photon-to-photon (PP). The intermediate transformed signals are denoted by f and ψ .
930 OPTOELECTRONICS
about an optoelectronic (either electron–photon extent, the term “optoelectronics”. The connec-
or photon–electron) transformation. A character- tions shown can be broken in one or several places.
istic of optoelectronics is charge neutrality, low Thus there are several independent branches in op-
noise feature, and the accessibility of the informa- toelectronics: devices for the display and presen-
tion carrier that is a photon for visual perception. tation of information, photoelectronics and photo-
Besides, optoelectronics provides the possibility electronic automatic equipment, ber optics, inte-
of excluding the effect of a frequency-dependent grated optics, acousto-optics, magnetooptics, and
wave impedance from the circuit matching, of sup- so on. The block diagram of an optoelectronic
pressing reactances from communication chan- transducer can contain any number of electronic
nels. It can also ensure almost complete isola- elements (e.g., amplifiers, generators, logical iso-
tion and autonomy of the channels by combining lators, etc.) in its electrical part. This introduces no
quantum mechanisms for generating radiation and changes into the algorithm of the optoelectronic
current carriers in the material with electronic transformation, its essence being unambiguously
methods for controlling these mechanisms. The determined by the basic elements of optoelectron-
application of the optoelectronic transformation to ics: radiation sources, photodetectors, and optical
electrical (a) and optical (b) signals is illustrated media. Simple optoelectronic transducers are ex-
in Fig. 1. The diagrams reflect the duality of op- emplified by optoisolators (optrons) based on an
toelectronic circuits, which determined, to a great optically connected radiation source and photode-
ORBACH–AMINOV–BLUME PROCESS 931
tector located in the same enclosure (like inte- from the light-emitting diode to the photodetector.
grated circuits); fiber-optical communication lines Optoisolators or optrons are used for amplifying
where the optical medium has the form of a long and converting electrical and optical signals, for
glass-fiber cable; solid-state amplifiers; and radi- switching, for modulation, etc. To increase the cur-
ation image converters where the photodetector rent transformation ratio I2 /I1 , which is typically
and radiation source are distributed structures, in- about 10−3 , by tens of times, a phototransistor is
tegrated technologically in a single device. The used in the optoisolator instead of a photodiode.
functional possibilities of an optoelectronic trans- See also Optoelectronics.
ducer are determined by the nature of the radia-
tion used. Coherent radiation provides more infor- OPTRON
mation content, and has a higher Q. The sources See Optoisolator.
of coherent radiation in optoelectronic transducers
are lasers and light emitting diodes. The function- OPW METHOD
ality of incoherent optoelectronics stems, mainly, See Orthogonalized plane waves.
from the photoelectric transformation. Used to this
end in optoelectronics are photoresistors, photodi- ORBACH–AMINOV–BLUME PROCESS
odes, phototransistors, and other photoelectric de- (R. Orbach, M.K. Aminov, M. Blume)
vices and their arrays. The generation, transfor- Establishment of an equilibrium distribution of
mation, and detection of the radiation are in most populations (see Level population) Em and Em
cases carried out by semiconductor elements and of spin levels of a magnetic ion with the near-
structures that lend themselves to microminiatur- est excited state energy El lower than the maxi-
ization and integration. The advantages of opto- mum phonon energy (see Fig. 2). The process re-
electronics that determine its role in solving some sults from absorption of a phonon of frequency
technical and scientific tasks, and its relation with ωph =
/ = (El − Em )/ and the spontaneous
various fields of physics, are clearly seen on the or induced radiation of a phonon with energy
diagram of Fig. 2. El − Em . This process, called resonant phonon
fluorescence, as a mechanism of spin–lattice relax-
OPTOISOLATOR, optron ation involving levels m and m was considered
A semiconductor device containing a light by Orbach (1961) and investigated independently
source (light emitting diode) and a detector (photo- by Aminov (1962). The rate τ −1 of this relax-
diode, phototransistor) without direct electrical ation process is characterized by its exponential
contacts between input (control) and output sig- dependence on the reciprocal temperature, τ −1 ∝
nals (see Fig. 1). The signal input current I1 trig- exp[−Δ/(k B T )], and it is inversely proportional
gers the generation of radiation by the light- to the level width of E l determined by the spon-
emitting diode. This radiation is detected by the taneous phonon emission. The process was ob-
photodiode and then transformed into an elec- served experimentally with Ce3+ ions in cerium–
tric current I2 to be applied to the load resistor magnesium nitrate, as well as with some ions of
R L . Fig. 1 shows a photon hν during its transit the iron group. Blume and Orbach (1962) studied
the spin-relaxation process (see Paramagnetic re- be written in the form of a Slater determinant:
laxation) for the ground state multiplet of a mag- 1
netic ion, i.e. when the value of Δ is very small. ψ=√ ×
N!
Such a process results in a temperature-dependent ⎡ ⎤
relaxation rate τ −1 ∝ T 5 . ψ1 (1)α(1) ψ1 (1)β(1) . . . ψn (1)α(1) ψn (1)β(1)
⎢ ψ1 (2)α(2) ψ1 (2)β(2) . . . ψn (2)α(2) ψn (2)β(2) ⎥
⎢ ⎥
det ⎢
⎢ . . . . . ⎥.
⎥
ORBITAL ANGULAR MOMENTUM, orbital ⎣ . . . . . ⎦
. . . . .
moment ψ1 (N)α(N) ψ1 (N)β(N). . .ψn (N)α(N) ψn (N)β(N)
The angular momentum of a microparticle aris-
The above representation of the wave function for
ing from its motion in a spherically symmet-
N electrons (N = 2n) ensures its antisymmetric
ric force field. According to quantum mechanics,
properties, with the Pauli principle automatically
the orbital angular momentum L is quantized;
satisfied.
namely, its magnitude as well as its projection
upon an axis, arbitrarily chosen in space (z axis),
ORDER
can take on only certain discrete values: L2 =
See Long-range and short-range order.
2 l(l + 1), Lz = m. Here, l = 0, 1, 2, . . . is the
orbital (azimuthal) quantum number, m = l, l − ORDER, ATOMIC
1, . . . , −l is the magnetic quantum number. The See Atomic order.
classification of the states of microparticles in ac-
cordance with their values of l is used in the theory ORDER DEGREE
of atoms and atomic nuclei, and also in the theory See Degree of order.
of scattering and collisions.
ORDER–DISORDER LAYERED STRUCTURE
ORBITAL, HYBRID (K. Dornberger-Schiff, 1964)
See Hybridization of orbitals. Structures of so-called order–disorder (OD)
layers, their alternation exhibiting some charac-
ORBITALS teristic features of order–disorder. The OD layers
Single-electron wave functions for construct- can be adjacent to each other in several equiva-
ing the wave function of a many-electron system. lent ways, depending on the symmetry of individ-
In the single-electron approximation, the state of ual layers and neighboring layer pairs. Layers in a
every electron of the system is described by its pair are related by operations of partial symmetry,
own wave function, an orbital, reflecting the ex- which are valid for a given pair, but not necessar-
istence of an individual state of an electron in a ily for the system as a whole. Owing to the vari-
certain effective field of the other electrons. In ation of partial symmetry for different pairs, there
atoms, the individual electron states are described could be, in principle, a diversity of OD structures
by atomic orbitals, and in molecules they are de- formed from the same layers (of one or several
scribed by molecular orbitals. kinds), and symmetrically equivalent pairs of lay-
A single-electron wave function dependent ers (so-called adjacency condition), both involv-
solely on spatial coordinates of a given electron ing disordered and various strictly periodic alter-
is called an orbital; or, taking into account spin nations of layers.
variables, a spin orbital. The sequence of single-
electron orbitals ψm corresponds to the spec- ORDER–DISORDER PHASE TRANSITION
trum of single-electron energy levels εm (m = Phase transition involving an ion (or atom) sit-
1, 2, 3, . . .). If each ith electron of the system uated in a potential well with two or more min-
is described by its individual spin-orbitals (i.e. ima, which determines by its position the value of
ψm (i)α(i) and ψm (i)β(i), allowing for two spin the order parameter. These minima are separated
directions: up (α) and down (β)), and its principal by maxima with energy
U , considerably higher
quantum number m, then the single-electron ap- than the interaction energy J between the different
proximation lets the overall system wave function ions of this type (see Fig.).
ORDER PARAMETER 933
ORDER IN ALLOYS
See Alloy ordering.
Single-particle potential showing the two minima where
ORDERING, ISOTOPIC
the ion is located at the order–disorder transition (Q is a
generalized coordinate, and
U
J ). See Isotopic ordering.
ORDERING TEMPERATURE
Above the order–disorder phase transition tem- The temperature of a rst-order phase transi-
perature (Tc ) both minima of the single-particle tion or of a second-order phase transition to a
potential are populated with equal probability. In phase of lower symmetry (ordered phase, see Or-
some systems above the transition point the double dered system).
minimum potential reduces to a simple parabolic
form characteristic of only one minimum. Below ORDER PARAMETER (introduced by
Tc the interaction J causes an asymmetry of the L.D. Landau, 1937)
distributions in the potential wells, and an order A generalized quantity, which is a measure of
parameter different from zero appears. At T = 0 the change of structure (change of internal symme-
all ions occupy equivalent minima. Sometimes, try) of a condensed many-particle system (a solid,
upon the variation of the external conditions, the in particular) in passing through a phase transi-
correlation between
U and J becomes reversed, tion point. The order parameter differs from zero
and a displacive type phase transition takes place only in the ordered phase (i.e. in the phase of low-
(
U < J ). Transitions of alloy ordering, many ered symmetry). It undergoes an abrupt change
ferroelectric phase transitions (e.g., KH2 PO4 and at a rst-order phase transition and is continuous
NaNO2 ), structural phase transitions, etc., are re- at a second-order phase transition, where its first
lated to the order–disorder type. derivative with respect to a certain thermodynamic
variable (temperature, pressure, density, etc.) has
ORDERED PHASE a discontinuity. The nature of the order parameter,
See Ordered system. its tensor dimension, and the number of its com-
ponents, are determined by the physical character-
ORDERED SYSTEM, ordered phase istics of the phase transition. In the case of a tran-
From the viewpoint of solid state physics, an sition to the ferro- or antiferromagnetic state, the
ordered system is one exhibiting a certain type of role of the order parameter is played by the vector
long-range order (see Long-range and short-range of spontaneous magnetization (ferromagnets) or
order), such as a crystal. In terms of the physics the antiferromagnetism vector (antiferromagnets).
of order–disorder phase transitions, the term “or- In the case of a ferroelectric phase transition and
dered system” is used in reference to that of two ferroelastic transitions, the polarization vector and
thermodynamic phases, one of which (usually the strain tensor act as order parameters, whereas in
934 ORDER PARAMETER PHASE
parameters with respect to different coordination of biaxiality may be of importance also in uniaxial
spheres: liquid crystals. In the case of an uniaxial crystal,
LL (ρ ) = P LL (ρ ) − P L P L , the tensor order parameter is of the form
εαα l αα l α α
1
α, α = A, B, C, . . . , (3) Qαβ = S nα nβ − δαβ , (1)
3
L, L = 1, 2, . . . , Q, where S is the degree of order or order para-
LL (ρ ) is the probability of occupation meter modulus, n is a unit vector. The physical
where Pαα l meaning of uniaxiality is as follows: there ex-
of a site of type L by an atom of type α and si-
ists an axis n about which the tensor Qαβ re-
multaneous occupation of a site of type L (at the
mains invariant under rotation. For instance, in
distance ρl in lth coordination sphere) by an atom
the case of n = {0, 0, 1}, Qαβ is diagonal and in-
of type α ; PαL , PαL are a priori probabilities of variant under rotation about the z-axis. It can be
occupation of sites L and L by atoms α and α seen from Eq. (1) that replacement of n with −n
respectively. In the case of alloys, which are de- leaves Qαβ unchanged, i.e. the states that are de-
scribed by Eq. (2), scribed by Qαβ involving n or −n, are equiva-
σ = η2 + 4εAB
12 (ρ ), (4) lent, from which it follows that the medium un-
1
der consideration is nonpolar (if molecules of the
i.e. the short-range order parameter σ is deter- medium exhibit permanent dipole moments, then
mined in the ordered state not only by the corre- equal quantities of the dipoles are directed up-
lation parameter, but also by the degree of long- wards and downwards). The name “order parame-
range order η. In the case of disordered A–B al- ter” is often given to the quantity that characterizes
loys (η = 0), the following short-range order para- the degeneracy of the equilibrium state of a sys-
meters: tem. In this case, such a quantity is the unit vec-
PAB (ρl ) tor n, two directions of which (n and −n) are con-
σl = 1 − , (5) sidered to be equivalent (such a vector is called the
cA cB
director). The order parameter Qαβ is the quan-
are often used. These short-range order para-
tity, in terms of which the thermodynamic poten-
meters are related to the correlation parameters
tials are expanded in the theory of phase transi-
εAB (ρl ) = PAB (ρl ) − cA cB by the equation
tions. For example, the nematic-to-isotropic liquid
αl = εAB (ρl )/(cA cB ) (the indices L and L are
transition is described using Qαβ . In the case of
omitted in these equations, because all sites of a
cholesteric liquid crystals, the order parameter (as
disordered alloy are of the same type).
given by Eq. (1)) is meaningful only locally, and
in the general case consideration must be given
ORDER PARAMETERS IN LIQUID CRYSTALS
to biaxiality, which is related to the helicity of
Parameters which are representative of orienta-
the director (the auxiliary optical axis). The or-
tional order and of partial translational order. Ori-
der parameter space of a nematic (the domain of
entational order is exhibited by the long axes of
values of n) is a sphere with equivalent diamet-
rod-shaped molecules, and by the normals to the
rically opposite points. The order parameter space
planes of disk-shaped molecules. Partial transla-
of a cholesteric is found to possess a more intricate
tional order occurs in smectic liquid crystals and
shape, which is why the cholesteric exhibits, e.g., a
columnar phases of discotics. In the general case,
wider diversity of defects and textures. A periodic
orientational order is described by the symmetri-
modulation of the density ρ takes place in smec-
cal traceless tensor Qαβ , which may be used as
tics (as a consequence of one-dimensional transla-
a measure of a biaxial medium. In the case of,
tional order). For a smectic A the density modula-
e.g., a plank-shaped molecule, the ordering of both
tion is given by the equation
major and minor axes can readily be imagined.
This type of ordering may occur in nematics and 1
ρ(r) = ρ0 1 + √ ψ(r) exp(iq0 nr) + c.c. ,
in smectics C (biaxial lyotropic nematics, see Ne- 2
matic liquid crystal). Furthermore, the fluctuations (2)
936 ORDER, RADIATION-INDUCED
where 2π/q0 is the distance between smectic lay- nature of the electron spectrum, as well as a strong
ers, and the complex modulation amplitude ψ(r) anisotropy of crystal electronic properties in di-
is the order parameter for describing the nematic– rections along and transverse to the conducting
smectic A phase transition. In this case, the free chains (see Quasi-one-dimensional crystals). Due
energy is set up by invariants, which are derived to the quasi-one-dimensional electron spectrum,
from Qαβ and ψ . In the case of a smectic A– at low temperatures these materials are usually
smectic C transition, the role of the order parame- insulators. In the organic compound TTF-TCNQ
ter is played (at given laminated structure) by the (tetrathiofulvalene-tetracyano-quinodimethane) as
complex function ψ = ω(r) exp{iϕ(r)}, where ω well as in a number of other organic conduc-
defines the inclination of the director with respect tors, the insulating state arises by virtue of a
to the normal to smectic layers in the C-phase, and metal–insulator Peierls transition. In another class
the phase ϕ defines the direction of the inclina- of organic conducting compounds (TMTSF)2 X
tion (C-director). Hence, this phase transition is (TMTSF is tetramethyl-tetraselenium-fulvalene,
similar to the transition of 4 He to the superfluid and X are anions of various types), metal–insulator
state (“helium analogy”). Other mesophases and phase transitions are also observable. In com-
the phase transitions between them are considered pounds (TMTSF)2 X with low-symmetry anions
in a similar manner. For instance, the density is a X = ReO4 , BF4 , NO3 , this transition is related to
two-dimensional periodic function of coordinates the anion ordering that leads to a superstructure
for the columnar phases of discotics. As in the with a longitudinal wave vector component Q =
case of smectics, the role of the order parameters 2kF (here kF is the Fermi momentum). In crys-
is now played by the corresponding amplitudes of tals with symmetric anions AsF6 , PF6 , SbF6 and
the given modulation. TaF6 , the insulating state is related to spin density
waves. In both cases, the metal–insulator phase
ORDER, RADIATION-INDUCED transition is caused by the quasi-one-dimensional
See Radiation-induced order. character of the electron spectrum. As in the
case of a Peierls transition, the insulating state in
ORDINARY RAY, ordinary wave (TMTSF)2 X corresponds to the appearance of a
One of the two light waves arising as a result gap in the electron spectrum at the Fermi surface
of the refraction of light at the boundary of a uni- (or some part of it). The insulating character of
axial crystal, that is normal to the principal sec- the electron spectrum hinders the superconductiv-
tion of the crystal (see Crystal optics). The ordi- ity. The insulating state can be suppressed by re-
nary ray obeys ordinary refraction laws: it lies in ducing the distance between the chains (e.g., using
the plane of incidence, and its speed of propaga- external pressure), since the probability of elec-
tion and, consequently, refractive index n0 , are in- tron transitions between neighboring chains in-
dependent of the direction of propagation. creases in this case, and the electron spectrum be-
comes three-dimensional. Studies of compounds
ORGANIC CONDUCTORS AND (TMTSF)2 X under pressure showed that pressure
SUPERCONDUCTORS can be instrumental in suppressing the transition
A series of organic materials has been syn- to the insulating state, which is related to the gen-
thesized that exhibits at high temperatures metal- eration of spin density waves or anion ordering in
lic type electrical conductivity proceeding mainly these compounds. Here, the superconducting state
along molecular chains. The crystals of such or- appears under pressure P ≈ 1.1 GPa at T Tc ≈
ganic conductors, as a rule, consist of planar mole- 1 K. In the compound (TMTSF)2 ClO4 supercon-
cules packed in zigzag stacks that form chains. ductivity appears at Tc ≈ 1.3 K under pressures
The extensive overlap of the electron wave func- as low as one atmosphere. This is attributed to
tions of neighboring molecules in a stack provides the belief that the so-called chemical contraction
for metallic conduction along the chain. The over- has taken place in the given compound. Essen-
lapping of wave functions of neighboring chains is tially the idea involves the substitution of octahe-
small enough to result in a quasi-one-dimensional dral anions PF6 , AsF6 for tetrahedral anions ClO4
ORGANIC POLYMERS 937
of smaller size. This is equivalent to the action of or nitrogen serve as heteroatoms (e.g., in poly-
pressure, and results in enhancing the degree of esters and polyamides). The backbone of an or-
three-dimensionality of the electronic spectrum. ganic polymer can have a linear or a branching
(TMTSF)2 X compounds exhibit superconductiv- network structure. Linear polymers exhibit some
ity and spin density waves coexisting in a narrow specific properties, including their ability to form
range of pressures and temperatures. anisotropic highly oriented fibers and films, and
In addition to (TMTSF)2 X, superconductiv- to undergo extensive reversible deformations, so-
ity has been observed in another class of organic called highly-elastic deformations. In the progres-
conductors, (BEDT-TTF)X (here BEDT-TTF is sion from linear polymers to branched and cross-
bis(ethylenedithiol)tetrafulvalene, X is any an- linked types, their specific properties becomes
ion). The structure of compounds of this class is less pronounced. Cross-linked polymers (see Poly-
close to that of (TMTSF)2 X compounds, but the mer cross-linking) are insoluble, infusible, exhibit
electronic spectrum in the former is more two- no plasticity, and are often brittle (e.g., phenolic
dimensional. The compound (BEDT-TTF)2 I3 su- resins). Some polymers are amorphous and some
perconducts with Tc ≈ 6–8 K. are crystalline. In addition to the highly elastic
Apparently the phonon mechanism of super- state, amorphous polymers exist in vitreous and
conductivity is operative in organic superconduc- visco-elastic states. Polymers that transform from
tors. The possible realization of the exciton mech- highly elastic to vitreous below room tempera-
anism of superconductivity, suggested by W. Lit- ture are classed as elastomers; and those with
tle (1964) for obtaining high-temperature super- higher transition temperatures are plastics (see
Polymeric materials). The ability to crystallize is
conductivity in organics, is problematic. Still, the
great potentialities of organic synthesis sustain the characteristic of polymers with regular sequences
of monomers, and a minimal amount of branch-
hope of significantly raising their critical tempera-
ing. Thus, “low-density” polyethylene with 20–50
tures.
branches per 1000 atoms has a 55–70% degree of
crystallinity, and “high-density” polyethylene with
ORGANIC POLYMERS 5–15 branches per 1000 atoms has 80–90% crys-
High molecular weight organic compounds tallinity. The density and strength characteristics
whose macromolecules consist of a great number are higher for the latter. Crystalline polymers are
of repeating units called monomers. For example, similar to plastics in their behavior. The proper-
the polymer polyvinylchloride (–CH2 –CHCl–)n ties of organic polymers are also determined by
is a derivative of the monomer vinylchloride intermolecular interactions due to the presence of
CH2 =CHCl which has a double bond that be- polar groups. For example, polymethylmethacry-
comes a single bond at polymerization. The de- late has a linear structure, and its transition to
gree of polymerization n, called the multiplicity, the highly-elastic state occurs at only 100 ◦ C; cel-
can be as high as several hundred thousand. If lulose is formed from rigid linear chains, and
polymer chains contain several kinds of repeat decomposes before reaching the highly elastic
units, the compound is called a copolymer (e.g., state transformation temperature. The substitution
butadiene–styrene rubber). According to the type of fluorine atoms for hydrogen produces poly-
of backbone chain, organic polymers are classi- mers (e.g., polytetrafluoroethylene, brand name
fied into homochain and heterochain types. The Teflon) with excellent mechanical and electri-
backbone of carbochain polymers consists solely cal insulation properties, as well as high thermal
of carbon atoms. Most carbochain polymers con- and chemical resistance. Organic polymers can be
tain saturated (single bonded) polymeric chains, produced by polymerization reactions (mostly,
but there are also polymers with unsaturated (mul- carbochain polymers) and polycondensation (het-
tiple) bonds in the chains, such as raw rubbers erochain polymers).
and polyacetylene. In heterochain polymers the The mechanical strength, elasticity, and elec-
backbone consists of carbon-containing segments trical insulation properties of organic polymers
joined by heteroatoms. In most cases, oxygen in combination with a low molecular weight and
938 ORGANIC SEMICONDUCTORS
low production cost are responsible for their wide that exhibit significant hole conductivity or elec-
application. Polymers are the basis for manufac- tronic conductivity, and positive temperature co-
turing plastics, chemical fibers, rubbers, paint- efficients of electrical conductivity. They include
work materials, glues, sealing compounds, ion- molecular crystals of organic compounds, organic
exchange resins. Biopolymers are the foundation dyes, charge-transfer complexes, ion-radical salts,
for all living organisms; they participate in all vi- and polymers with conjugated bonds. Electri-
tal function processes, and examples of them are cal and photoelectric properties of organic semi-
enzymes, nucleic acids, polysaccharides, etc. conductors are due to their system of conju-
Nowadays, the so-called conducting polymers gated bonds; their current carriers are excited
are intensively investigated. These are carbochain π -electrons associated with these conjugated sys-
organic polymers with a backbone of unsatu- tems. Organic semiconductors exist as monocrys-
rated polymeric chains containing alternating sin- tals, polycrystals, amorphous lms and powders.
gle and double bonds (polyvinylenes). Owing to The resistivity at room temperature varies from
their backbone-bond conjugation, these polymers 1018 (naphthalene, anthracene) to 1022 ·cm
exhibit unique electrical properties. The simplest (ion-radical salts).
of them is polyacetylene (–CH=CH–)n which is a Charge carriers in organic semiconductors with
semiconductor with conductivity 10−3 (·m)−1 . weak interactions of the van der Waals type (see
Introducing impurities in small quantities, either Van der Waals forces) are able, owing to the pro-
donors or acceptors (e.g., Li and I2 , respectively), nounced localization of individual molecules, to
results in an increase in electrical conductivity that interact with electron and nuclear subsystems of
can reach metallic values. For example, the max- their environment, and form polarons. The con-
imum conductivity of iodine-doped polyacety- duction bands in organic semiconductors are nar-
lene is 1.5·107 (·m)−1 . Conducting polymers row (∼0.1 eV), and the mobility of current carri-
are promising materials for the creation of cur- ers is typically as low as ∼1 cm2 ·V−1 ·s−1 . In
rent sources, gas-filled capacitors operating on the addition to the band mechanism of conduction,
hopping conductivity takes place. The temperature
principle of electrochemical doping, phototrans-
formers, solar cells, etc. dependence of the mobility of current carriers in
model systems of organic semiconductors exhibits
A copolymer is composed of polymer chains
a low-temperature behavior similar to that in inor-
containing two or more types of repeat units (A)m
ganic semiconductors, namely, T −n with n > 1.
and (B)n . In many cases one polymer component
The trapping centers (traps) of charge carriers
(A)m is water soluble, and the other (B)n is not.
in organic semiconductors can be of either struc-
A polymer nanosphere is formed when one long
tural or impurity origin. The structural defects that
polymer (A)m coils up in a spherical shape and
produce a local clustering of molecules in a crystal
develops cross links between adjacent lengths of
lattice become traps due to the accretion of elec-
strands to give rigidity to the sphere. Strands of
tron polarization energy of the crystal by a carrier
the other component (B)n attached to the surface
in the defect region. Due to the strong localiza-
form what is called a “hairy” nanosphere copoly-
tion of charge carriers at individual molecules in
mer if they are short compared to the diameter of organic semiconductors, structural defects such as
the sphere, and they form a so-called “star” poly- dislocations, stacking faults, orientation and con-
mer if they are long compared to the nanosphere formational defects, etc., play a more significant
diameter. A “brush” copolymer consists of a sub- role in processes of charge carrier transfer in com-
strate with long strands of soluble (A)m projecting parison to inorganic semiconductors. The impurity
on one side, and long strands of insoluble (B)n on traps of charge carriers in organic semiconductors
the other side. are produced by impurity molecules with a greater
electron af nity or lower ionization potential than
ORGANIC SEMICONDUCTORS those of host molecules. There are charge carrier
A class of materials that belongs, because of traps localized in the vicinity of an impurity mole-
their type of bonding, to molecular compounds cule even when the impurity molecules themselves
ORIENTATIONAL ORDER 939
ORGANOPLASTICS
Plastics (see Polymeric materials) reinforced
with synthetic fibers. Organoplastics possess rel-
atively low density and low thermal conductiv-
ity, high strength, good electrical insulation prop-
erties, as well as high atmospheric, water, and
chemical resistance; they are readily machinable.
Organoplastics are used in radio-engineering and
for construction materials in the aircraft, automo-
bile, and other industries. Absorption line shape (1) and its derivative (2) in the
case of a spin Hamiltonian H
with orthorhombic sym-
metry of the form H = μB i=x,y,z gii Si Bi , where
ORIENTATIONAL LINE BROADENING
Bri = hν/(μB Sgii ) (μB is the Bohr magneton, gii are
Inhomogeneous broadening that arises due to
g-tensor principal value components, Si are spin opera-
different orientations of paramagnetic centers rel- tors, and ν is the microwave frequency).
ative to an external magnetic field B. Randomly
oriented paramagnetic centers appear in powders,
glasses, frozen solutions, and crystals with a mo- of each individual line in a spectrum from a para-
saic structure (see Mosaic crystals). In many cases magnetic center in a monocrystal.
the population density of these centers (glasses,
powders) directed along any solid angle is the ORIENTATIONAL ORDER
same. There is an additional condition for orienta- A characteristic of a system where, in princi-
tional broadening to appear. The resonance value ple, the spontaneous breaking of rotational sym-
of the magnetic field B r should depend on the an- metry at phase transitions is possible (see also
gles specifying the direction of the field B rela- Spontaneous symmetry breaking). To describe
tive to the principal axes of the paramagnetic cen- orientational order, the concept of order parame-
ters, i.e. the spectral lines of monocrystals should ter is introduced. As a rule, short-range orien-
exhibit an angular dependence of spectra. In this tational order appears at temperatures above the
case the variously oriented paramagnetic centers phase transition point, and long-range orienta-
will have different values of B r , resulting in ori- tional order appears below this point (see Long-
entational broadening with an asymmetrical line range and short-range order). For example, ferro-
shape that depends on the type of spin Hamil- magnets, ferroelectrics, and nematics (see Nematic
tonian (see the Figure). In some cases of a non- liquid crystal) display long-range orientational or-
monotonic angular dependence of B r , extra ab- der with the corresponding order parameters mag-
sorption peaks appear in electron paramagnetic netization, polarization, and director. In terms of a
resonance spectra, which facilitate their analysis. qualitative description, the orientational order in
On the whole, the interpretation of spectra with a ferromagnet and a ferroelectric can be thought
orientational broadening is a much more difficult of as the aligning of magnetic and electric dipoles
task than the analysis of the angular dependence along some (particular) axes over the whole (in-
940 ORIENTATIONAL PHASE TRANSITION
finite) space. In nematics, the elements ordered the resultant beam density are concentrated in in-
along the director are the long axes of rod-shaped terfacial regions of the crystal (see Anomalous
molecules or the normals to planes of disk-shaped passage of X-rays).
molecules. True long-range orientational order ex-
ists when the average over all orientations of, e.g., ORIENTATION RELATIONS
magnetic moments or long axes of molecules, dif- A way to express the geometrical correspon-
fers from zero. However, this can not always be dence of two conjugate crystal bodies or phases.
achieved. For instance, in two-dimensional mag- The orientation relations were established by Hei-
nets or nematics, true long-range orientational or- dinger in 1827 for intergrown minerals of hematite
der is impossible, but a quasi-long-range one can and magnetite; the basal plane (001) in hematite is
exist. The latter is understood as the falling-off of parallel to the (111) plane in magnetite. The ori-
the corresponding correlation functions as a power entation relations (OR) are expressed in terms of
of distance in contrast to the short-range orienta- parallel or near-parallel close-packed planes and
tion order (e.g., paramagnets or isotropic phase of directions of the crystal lattices of these bodies or
nematics) when the correlators fall off exponen- phases. These relations are found with the help of
tially. X-ray, electron diffraction, electron-microscopic,
and optical methods. A single relation corresponds
ORIENTATIONAL PHASE TRANSITION to a few crystallographically equivalent orienta-
A phase transition where the orientation of the tions of crystals of some phase with respect to
order parameter vector η (magnetization, polariza- a crystal of another phase. The Nishiyama ori-
tion, and so on) changes as the temperature, pres- entation relation (Z. Nishiyama, 1934–1935) cor-
sure, or external field changes. responds to 12 equivalent orientationally variant
martensitic crystals. For example, a variant of this
ORIENTATION EFFECTS in irradiated relation is (001)α (111)γ , [01̄1]α [2̄11]γ ; and
monocrystals the remaining 11 variants involve permutation of
A sharp dependence of the efficiency of various Miller indices (see Crystallographic indices). The
processes during the passage of charged particles Kurdyumov–Sachs orientation relation (G.V. Kur-
or high-energy photons through monocrystals on dyumov, G. Sachs, 1930) gives the directional
the angle ψ of the particle (photon) beam direc- interrelation of crystal lattices of austenite (A)
tion with respect to the directions of atomic chains and martensite (M) in steel and iron alloys. In an
or planes. The following orientation effects are ob- austenite monocrystal, the martensite lattice can
servable in the channeling of positive ions: abnor- be oriented in 24 ways, e.g., (111)A (101̄)M ,
mally deep penetration of ions into monocrystals; [101]A [111]M , where the close-packed planes
reduction of the nuclear reaction efficiency (e.g., and directions of austenite and martensite are par-
of protons or alpha particles with atoms of the allel; in this case twinning structures may form. An
crystal); decrease of characteristic X-ray radiation intermediate OR between those of Nishiyama and
intensity; and decrease of backscattering of ions Kurdyumov–Sachs is the Greninger–Troiano ori-
when the the angle of incidence ψ is limited to entation relation (A.B. Greninger, A.R. Troiano,
values below the critical angle ψk . Other orienta- 1949), according to which 24 variants of mutual
tional effects include the channeling of ion beams orientations of austenite and martensite exist, e.g.,
and the related shadow effect, the anisotropy of (111)A (011)M , [5, 12, 17]A [7, 17, 17]M . Ori-
ion–ion and ion–electron emission and cathode entation relations are found in practically every
sputtering, and the appearance of Vener spots dur- diffusion and diffusion-free transformation in solids,
ing sputtering (see Focuson). During the chan- oxidizing reactions, directional and eutectic crys-
nelling of electrons and positrons intense hard ra- tallizations. Orientations often tend to spread (de-
diation can appear. An orientational effect at the viations from planar and direction parallelism) due
incidence of a beam of electrons or X-rays at the to crystal structure defects and external and in-
Bragg angle is their anomalous passage through ternal stresses. The strictest adherence to orienta-
monocrystals due to interference between trans- tions (within an error of ∼1◦ ) is found in marten-
mitted and scattered beams, when the maxima of sitic transformations. For the transition from a
OSCILLATIONS, KHAIKIN 941
face-centered cubic (FCC) to a body-centered cu- chemical formula is RFeO3 , where R is yttrium or
bic (BCC) (or body-centered tetragonal) lattice, a rare-earth element. The chemical unit cell coin-
depending on the composition of the alloys, the cides with the magnetic one, and contains four for-
Kurdyumov–Sachs, Greninger–Troiano or Nishiy- mula units. The space group is P bnm (D2h 16 ). The
ama OR is valid. (The latter appears in iron–nickel 3+ 3+
R and Fe ions are in crystallographic site po-
alloys and differs from the Kurdyumov–Sachs OR sitions 4c and 4b, respectively. The system of Fe
by a turn of the martensite lattice through 5◦ 16 ions experiences magnetic ordering at T ∼ 600 K.
about the direction [001]α .) A similar transition The type of ordering is antiferromagnetic, usually
in thin lms leads to the Bain orientation rela- with weak ferromagnetism. The system of rare-
tion (001)α (001)γ , [100]α [110]γ (E.C. Bain, earth element ions is ordered at liquid helium tem-
1924). At the transition of a BCC phase to a peratures. With the temperature lowered, sponta-
hexagonal close-packed (HCP) phase, Burgers neous spin-reorienting magnetic phase transitions
orientation relation (W.G. Burgers, 1936) applies occur in many orthoferrites.
most often: (110)BCC (0001)HCP , [1̄10]BCC
[112̄0]HCP . ORTHOGONALITY, STRONG
Orientation relations are used for establishing See Strong orthogonality.
the crystalline geometrical mechanism of the re-
arrangement of one crystal lattice into another dur-
ORTHOGONALIZED PLANE WAVES, OPW
ing a solid state transformation. For example, the
method
rearrangement of a FCC phase into a BCC (or
A band structure computation method that uses
BC-tetragonal) phase with an invariant plane in-
plane waves, orthogonalized to wave functions of
volves the intrinsic (Bain) deformation to re-
the atomic core, as the basis functions. When the
arrange the initial lattice to the final one; a defor-
basis of the OPW method is over-determined, use
mation with an invariant lattice by means of shear
is often made of its modification, the completely
and twinning of the crystal; and, finally, the rota-
orthogonalized plane wave method. In the latter
tion of the rearranged region through an angle that
case, as many orthogonalized plane waves are re-
ensures the adaptation of the rearrangement mech-
moved from the basis set as is necessary to elim-
anism to the external conditions, in particular, with
inate over-determination. The OPW method was
the elastic action of the surrounding initial phase.
historically the foundation for the pseudopotential
It is this rotation that determines the orientation
method.
relation. No rotation of the rearranged region is
present at the transition of the FCC phase to close-
packed martensite phases (in particular, those with ORTHORHOMBIC SYSTEM (fr. Gr. oρθoς , right,
the HCP lattice) that are formed by means of and ρoμβoς , rhombus)
the shear of close-packed planes (111). In this Crystallographic system determined by the
case, the ORs are determined by the plane and presence of three second-order symmetry axes
the shear direction (e.g., (111)FCC (0001)HCP , in crystals; it belongs to the orthorhombic crys-
[101̄]FCC [21̄1̄0]HCP ). For transformations with tal class. The rectangular basis coordinate system
a predominance of surface energy, those orienta- axes of this system 2 (or 2̄) are directed along
tion relations are present that correspond to the the edges of the unit cell with parameters a =
minimum at the cross-section of a polar diagram b = c, α = β = γ = 90◦ . The orthorhombic sys-
of the surface energy, and provide for the best in- tem includes 4 Bravais lattices, 3 point groups, and
terlinking of the lattices of two crystal phases over 59 space groups.
the plane of their contact.
OSCILLATIONS, FRIEDEL
ORTHOFERRITE (fr. Gr. oρθoς , straight, and See Friedel oscillations.
ferrite)
Ferrites with rhombically distorted variants of OSCILLATIONS, KHAIKIN
the perovskite (CaTiO3 ) structure. The general See Khaikin oscillations.
942 OSCILLATIONS, SELF-
OSCILLATIONS, SELF- 1960). The latter is caused by the Hall drift of the
See Self-oscillation. plasma in mutually perpendicular magnetic and
electric fields, which arises as a result of the shift
OSCILLATIONS, ZERO SPIN of electron and hole screw perturbations of density
See Zero spin oscillations. in a constant electric field. The equality of the Hall
and diffusion flows corresponds to the threshold of
OSCILLATOR STRENGTH excitation of the instability. Therefore, the insta-
Dimensionless quantity describing the prob- bility is excited in sufficiently strong electric and
ability of a quantum transition during a radia- magnetic fields: ∼30 V/cm and ∼1 T. A related
tion process (see Radiative quantum transition) and phenomenon is the oscillation of a bar of semi-
light absorption. For electronic transitions between conducting material such as germanium when sub-
levels of energies Ei and Ej arising from a change jected to a magnetic field and a direct current.
of the system dipole moment the oscillator strength
fij is given by OSMIUM, Os
A chemical element of Group VIII of the peri-
8π 2 m 1
fij = − 2
νij |p̄iα,jβ |2 , odic system with atomic number 76 and atomic
3he gi mass 190.2. Natural osmium consists of stable
αβ
isotopes 184 Os (0.018%), 186 Os (1.582%), 187 Os
where m, e are the electron mass and charge, re-
(1.64%), 188 Os (13.27%), 189 Os (16.14%), 190 Os
spectively, gi is the statistical weight of the initial
(26.38%) and 192 Os (40.97%). Outer electronic
state, νij = (Ei − Ej )/ h, and piα,jβ is the ma-
shell configuration is 4f 14 5d 6 6s 2 . Successive
trix element of the electric dipole moment operator
ionization energies are 8.7, 17.0 eV. Atomic radius
between states (iα) and (jβ). For Ei > Ej (radi-
is 0.133 nm; radius of Os4+ ion is 0.065 nm, of
ation), the oscillator strength is negative, and for
Os6+ is 0.069 nm. Electronegativity is 1.52. Oxi-
Ei < Ej (absorption) it is positive
dation state is +8, +6, +4; less often, +3, +2.
gj In free form, osmium is a bluish-gray metal
fij = − fj i .
gi (heaviest known metal), very hard and brittle. It
has a hexagonal close packed lattice with parame-
The oscillator strength obeys the sum rule ters a = 0.273445 nm and c = 0.431745 nm at
j fij = N , where N is the number of elec- 4 ). The den-
trons in the quantum system. The term oscilla- 300 K, space group P 63 /mmc (D6h
tor strength arises from the fact that the contribu- sity is 22.6 g/cm3 , T melting = 3318 K, T boiling ∼
tion of transition i → j to the polarizability can 5300 K. Binding energy is −8.11 eV/atom
be considered as the polarizability of an oscilla- at 298.15 K. Heat of melting is 29.3 kJ/mole;
tor multiplied by fij . Therefore, this term not only heat of evaporation is 678 kJ/mole, and heat of
applies to electronic but also to other types (in par- sublimation is 728.5 kJ/mole. Specific heat is
ticular, oscillatory) of transitions. The oscillator 0.129 kJ·kg−1 ·K−1 at 298 K; Debye temperature
strength is determined by measuring the integrated is 278.5 K; the linear thermal expansion coeffi-
absorption, by curves of anomalous dispersion of cient is 5.6·10−6 K−1 along the main axis 6 and
light, and by other methods. 3.9·10−6 K−1 transverse to it (at room temper-
ature), the coefficient of thermal conductivity is
OSCILLISTOR 87.1 W·m−1 ·K−1 at 273 to 373 K. Adiabatic bulk
Oscillations of current in the electron– modulus is 375 GPa, Young’s modulus is 559 GPa,
hole plasma of a semiconductor (see Solid-state rigidity is 223 GPa, Poisson ratio is 0.25 at 298 K;
plasma) placed in parallel electric and magnetic Vickers hardness is 2.9 to 6.6 GPa for annealed
fields. This effect was first detected in germa- osmium at 298 K and 1.4 to 4.0 GPa at 1473 K
nium. The phenomenon of current instability re- (range due to anisotropy of elasticity). Osmium
sults from the Kadomtsev–Nedospasov screw in- metal is practically impossible to deform and dif-
stability (B.B. Kadomtsev and A.V. Nedospasov, ficult to volatilize in high vacuum (saturated vapor
OXYGEN 943
pressure at 2773 K is 0.0133 Pa). Ion-plasma fre- magnetic field with Zeeman frequency ωS
ωI
quency of osmium is 16.127 THz; coefficient of (ωI is the splitting of nuclear spin levels). If the
the linear term in low-temperature electronic spe- electron–nuclear relaxation is controlled by the
cific heat is 2.26 mJ·mole−1 ·K−2 ; electric resis- anisotropic part of the hyperfine interaction with
tivity is 96.6 n·m at 298 K; temperature coeffi- selection rules
M +
m = ±2, then pm equals
cient of electrical resistance 0.0042 K−1 from 273 P0 /2.
to 373 K; superconducting transition temperature
is 0.655 K, critical magnetic field is 6.5 mT (at OVERHEATING, METAL
0 K). Osmium is paramagnetic with magnetic sus- See Metal overheating.
ceptibility +13.1·10−6 cm3 /mole; nuclear mag-
netic moment of isotope 189 Os is 0.651 nuclear OXIDATION, oxide coating
magnetons. Formation of a solid oxide phase on a metal or
Owing to the high hardness, corrosion and semiconductor surface. It proceeds via the electro-
wear resistance of osmium, as well as its natural chemical interaction of ambient oxygen molecules
and artificial alloys with other platinum group met- with atoms of the substrate. The rate of formation
als, it is employed in various instruments. Osmium of oxide decreases as the thickness of the oxide
(and its compounds) is a good catalyst for many film grows; therefore, the steady-state thickness of
kinds of synthesis. the oxide is determined by its transparency to pen-
etration by oxygen molecules. The formation of
a dense film (Al2 O3 ) promotes surface passiva-
OVERHAUSER EFFECT (A.W. Overhauser,
tion. Loose oxide films (Fe2 O3 , CuO etc.) hinder
1953)
further oxidation, thereby facilitating surface cor-
Dynamic nuclear polarization caused by spin
rosion. The controlled coating of a semiconductor
relaxation (see Paramagnetic relaxation) of an un-
surface with oxides is widely used in semiconduc-
paired electron center coupled to nuclear spins
tor engineering to produce insulating and passivat-
by the hyper ne interaction. In the case of the
ing coatings. To this end, thermal and anode oxi-
Overhauser effect, dynamic nuclear polarization
dation is applied.
is achieved owing to the electron spin subsys-
tem that was disturbed from thermodynamic equi- OXIDATION, THERMAL
librium by external alternating fields. The elec- See Thermal oxidation.
tron spin relaxation involves nuclear spin forbid-
den relaxation transitions with a certain proba- OXIDE MAGNET
bility. Since under nonequilibrium conditions the A magnetically hard ferrite; a permanent mag-
balance of relaxation paths is disturbed, the sys- net made of these materials. Oxide magnets are
tem is dominated by relaxation processes char- manufactured from Fe, Ba, Co, and Sr oxides, and
acterized by a certain change
M of the projec- most often from barium ferrite BaO·6Fe2 O3 in
tion M of the electron spin S. The selection rules two modifications, isotropic and anisotropic (the
relate the change
m of the projection m of nu- latter obtained by pressing powder in a magnetic
clear spin I with
M; this leads to a predomi- field). Some properties of oxide magnets are given
nantly nonequilibrium population of particular nu- in Table 1 of Hard magnetic materials.
clear spin sublevels. In the case of the Overhauser
effect, the subsystem is saturated by the resonance OXYGEN, O
oscillating field, while the electron nuclear relax- Chemical element of Group VI of the peri-
ation is caused by the contact interaction of the odic system with atomic number 8 and atomic
S · I type that does not change the resultant pro- mass 15.9994. Naturally occurring oxygen is com-
jection of spins:
M +
m = 0. The maximum posed of three stable isotopes: 16 O (99.762%),
possible dynamic nuclear polarization in the case 17 O (0.038%), and 18 O (0.200%). The tetrahe-
S = I = 1/2 is pm = −P0 = tanh[ωs /(2k B T )], dral atomic radius is 0.066 nm (coordination num-
which is practically coincident in value with the ber 4), the octahedral radius is 0.074 nm (coor-
equilibrium polarization P0 of electron spins in a dination number 6), and the ionic radius of O2−
944 OXYGEN
is 0.136 nm. The electronic configuration of the sible for a very close coordination of the mag-
outer shell is 2s 2 2p4 . Successive ionization ener- netic and lattice properties of solid oxygen, as
gies are: 13.618, 35.117, and 54.90 eV. The elec- well as for a very rich spectrum of collective ex-
tron affinity is 1.467 eV, and the oxidation state citations: acoustic phonons, librons, magnons, vi-
is −2, more rarely, −1. brons, excitons, biexcitons, and various combina-
Oxygen in the free state usually forms the tions thereof (multi-particle excitations). Due to
molecule, O2 , more rarely, ozone, O3 (10−4 %). the unique character of its properties solid oxy-
O2 is one of the few triplet ground state mole- gen has been a popular material for study in low-
cules (spin S = 1); the dissociation energy is temperature physics. At the saturation vapor pres-
493.6 kJ/mol (at 0 K), and the distance be- sure solid oxygen exists in three crystallographic
tween the nuclei is 0.120735 nm. The specific modifications.
heats of O2 are: cp = 29.27 J·mol−1 ·K−1 , cv =
The high-temperature cubic γ -phase (space
20.5 J·mol−1 ·K−1 (at 273 K). The density of group P m3n, Oh3 ) has a high degree of dis-
gaseous oxygen is 1.42897 g/dm3 (at 273 K order of its molecular orientations. The orien-
and standard pressure), that of liquid oxygen is
tationally ordered rhombohedral (trigonal, R3m,
1.1321 g/cm3 (at T boiling), and that of solid 5 ) β-phase (T
D3d γ β = 43.80 K) and monoclinic
oxygen is 1.46 g/cm3 (at 20.3 K); T melting = 3 ) α-phase (T
54.43 K, and T boiling = 80.15 K. The triple point
(C2/m, C2h βα = 23.88 K) have a
similar pattern of molecular packing since both
parameters are: Ttriple = 54.36 K and ptriple =
146 Pa; the heat of vaporization is 6.82 kJ/mol, phases have layered structures. The O2 mole-
and the heat of fusion is 0.443 kJ/mol. At 273 K, cules in each layer form densely packed planes
the thermal conductivity is 24 mW·m−1 ·K−1 , the with the molecular axes perpendicular to these
dielectric constant is 1.000547, the viscosity is planes. The monoclinic lattice distortion at the
18.9 μPa·s, and the susceptibility χ (at 293 K) is β → α transition arises from the magnetic or-
107.8·10−9 . dering. The α-O2 modification is a quasi-two-
Solid oxygen belongs to the group of mole- dimensional antiferromagnet. The interlayer ex-
cular crystals or cryocrystals. It combines, in a change in β-O2 results in a strong antiferro-
unique fashion, the properties of a molecular crys- magnetic short-range order, even though there is
tal and of a magnetic substance. Solid oxygen is no long-range magnetic order in this phase. The
the only naturally occurring homomolecular mag- γ -phase and liquid O2 are paramagnets. Stud-
netic compound with a direct exchange interac- ies of solid oxygen under pressure have revealed
tion, this interaction comprising a substantial por- several new phases with an orientational structure
tion of the total lattice energy. This fact is respon- similar to those of α- and β-O2 .
Pp
PAIRING, TRIPLET wave functions ψi of the ground state can be ex-
See Triplet pairing. pressed in terms of one-electron functions χi lo-
calized at two-center bonds. There is a unitary
PAIRWISE ADDITIVE APPROXIMATION transformation to express the functions ψi as a lin-
One of the methods of approximately calcu- ear combination of basis functions χi localized at
lating the total energy of molecular systems. The the bonds, so the total system energy and charge
pairwise additive approximation is used in situa- distribution remain invariant under the transition
tions where it is possible to construct the model to the new basis.
of the pairwise interaction of atoms, in which
PAIRWISE INTERACTION OF ATOMS
the many-atomic interactions are effectively taken
into account. In its general form the total poten- An approximation sometimes used to describe
the interaction between atoms (ions) in solids, as
tial energy of system of N atoms is presented by
well as in liquids and gases. It assumes that the
the sum of double, triple, etc., individual potential
potential energy of a solid (adiabatic potential) U
energy terms
within the adiabatic approximation may be repre-
sented in the form of the sum of potential energies
1 1
N N
U= ϕij + ϕij k + · · · (potentials) of pairs of atoms (ions), the coordi-
2! 3! nates of which are R l and R l :
i=j i=j =k
1
1
N
U= ϕ(R l , R l ).
+ ϕij ...m , 2
m! ll
i=j =···=m
This ignores interactions of three or more atoms
which in some cases can be written in the form (ions), a neglect which can not always be justi-
fied. Besides that, the approximation of pairwise
1 ∗
N
interaction of atoms fails to take into account en-
U= Eij ,
2 ergy terms which are not dependent on the solid
i=j
structure, but which are functions only of its vol-
where ume. Such a neglect is not warranted in the case of
a metal, although it does provide good results for
∗ =ϕ 1 ϕij k crystals with ionic bonds, with covalent bonds and
Eij ij 1 + ;
3! ϕ with van der Waals chemical bonds (see Van der
k=i,j ij
Waals forces). In the simplest form of the pairwise
this includes averaged interactions of higher or- interaction approximation, the potential ϕ is as-
ders. Introduction of the pairwise additive approx- sumed to be a function of the interatomic distance
imation is especially useful in cases where notions |R l − R l | only. This approximation restricts the
of isolated chemical bonds are valid, as, e.g., in interatomic interaction potentials to central forces
covalent structures. The picture of two-centered only.
bonds corresponds to real singularities of electron The pairwise interaction approximation is valid
behavior in saturated systems, where the basic when analyzing the properties of solids in the
945
946 PALLADIUM
context of the harmonic approximation. The inter- dium noticeably. Thus, 50% reduction increases
atomic interaction potential in this case is incon- the ultimate tensile strength to 325.5 MPa; the
sistent with a simple two-particle potential, but value of hardness increases by a factor of 2–
takes into account the contributions of many-ion 2.5. The annealing, which follows the reduction,
potentials. In this connection, the concept of ef- softens the metal again. Alloying elements, par-
fective pairwise potential is introduced, the para- ticularly ruthenium and nickel, also increase the
meters of which are determined experimentally. value of palladium harness. Palladium exhibits
This interatomic potential can account for the the following values of adiabatic elastic moduli
prominent features of an interatomic interaction in (GPa): c11 = 227.10, c12 = 176.04, c44 = 71.73
general: short-range repulsion and comparatively at 300 K. The characteristic feature of palladium is
long-range attraction. its ability to reversibly absorb hydrogen, whereby
up to 900 volumes of the gas are taken by a
PALLADIUM, Pd single volume of metal; absorption of hydrogen
An element of Group VIII of the periodic ta- causes considerable increase of volume and em-
ble with atomic number 46, atomic weight 106.42. brittlement of palladium. On moderate heating in
Natural palladium consists of isotopes with fol- vacuum the metal releases the absorbed hydro-
lowing mass numbers: 102 (1.02%), 104 (11.14%), gen.
105 (22.33%), 106 (27.33%), 108 (26.46%), 110 Pure palladium is used for the manufacture of
(11.72%). Electronic configuration of the outer chemical equipment; gold-, platinum- and radium-
shell is 4d 10 . Sequential ionization energies (eV): bearing alloys are used in temperature controllers,
8.336, 19.428 and 32.92. Atomic radius 0.137 nm. thermocouples, etc. The chief application of
The radii of Pd2+ and Pd4+ ions are respec-
palladium is in medicine, particularly in dental
tively 0.080 nm and 0.065 nm. Oxidation states:
work.
+2, +4, less often +3. Electronegativity value
≈2.2.
Free palladium is a silvery-white soft metal. PAPKOVICH–NEUBER SOLUTION
It exhibits a face-centered cubic structure, a = (P.F. Papkovich, 1932; H.Z. Neuber, 1934)
0.38902 nm (at 296 K). Density 12.02 g/cm3 at A general solution of the equations of elas-
293 K, Tmelting = 1827 K, Tboiling ≈ 3210 K. tic medium equilibrium, which is expressed in
Heat of melting 16.71 kJ/mol, heat of evaporation terms of the scalar function ϕ and vector func-
353 kJ/mol; specific heat 0.2445 kJ·kg−1 ·K−1 tion ψ . If the equilibrium equation is written as
(at 298 K); the coefficient of thermal conductiv- μ∇ 2 u + (λ + μ) grad div u + ρF = 0, where u is
ity 71.2 W·m−1 ·K−1 . Coefficient of linear ther- the displacement vector, λ, μ are the elastic Lamé
mal expansion 11.1·10−6 K−1 (at 273–373 K). coef cients, ρ is the density, F is the mass den-
Resistivity 99.6 n·m (at 298 K). The tem- sity of the bulk force, then the Papkovich–Neuber
perature resistance coefficient 0.00379 K−1 (at solution is of the form u = grad(ϕ + r · ψ) −
273–373 K). Electron work function 4.99 eV. 4(1 − ν)ψ , where ν is the Poisson ratio, and the
Palladium is paramagnetic with the specific mag- functions ϕ and ψ satisfy the equations
netic susceptibility 5.231·10−6 at room temper-
ρ
ature. Saturation vapor pressure is 13.33 MPa 4(1 − ν)∇ 2 ϕ + r · F = 0,
(at 1427 K) and 1333 MPa (at 1773 K). An- μ
nealed palladium exhibits the elastic modulus of
121.128 GPa, the shear modulus of 49.98 GPa, ρ
ultimate tensile strength ≈189 MPa, relative 4(1 − ν)∇ 2 ψ = F.
μ
elongation from 25 to 40%. The value of Brinell
hardness 480.2 MPa. Palladium is amenable to It follows from the latter equations that the scalar
forging, punching, rolling to thinnest sheets, function ϕ and the components of the vector func-
stretching to thin wires, it lends itself readily to tion ψ are harmonic functions in the absence of
polishing and welding. Cold strain hardens palla- bulk forces (F = 0).
PARAELECTRIC RESONANCE 947
PARALLELEPIPED, PRIMITIVE
See Primitive parallelepiped.
PARAMAGNET
Dependence of the derivative of the ultrahigh-frequency A material which becomes magnetized propor-
power absorption on the applied electric field E at the tional to and in the direction of an applied external
noncentral ion 7 Li in KCl (ν klystron = 63 GHz, T =
magnetic eld B, but does not exhibit magnetic or-
4.2 K, N Li = 7·1016 cm−3 ): (1, 3) forbidden transitions,
dering. Diamagnetism that may be present is dom-
(2) allowed transition, χ is the imaginary part of the
inated by paramagnetism in paramagnets. Classi-
electric susceptibility.
cal paramagnets (Langevin paramagnets), which
have a temperature dependence of the susceptibil-
given in Fig. as an example. The locations of the ity which satisfies the Curie law, include a wide
resonance lines provide the magnitudes of the tun- group of materials with intrinsic magnetic mo-
nel splitting Δ, and the electric dipole moment d ments: gases (O2 , NO, metallic vapors), liquids
of the impurity. For 7 Li+ in KCl Δ ≈ 32 GHz, and solids (rare-earth salts, salts of transition met-
d [111], |d| ≈ 18.5·10−30 C·m (without taking als, solutions of these salts). Many paramagnets
into account the correction for the Lorentz eld). are Langevin types, and yet exhibit stronger inter-
Paraelectric resonance may be also observed actions of their magnetic moments with the matrix
in the absence of the applied electric field, a sit- and with one another (salts and oxides of transi-
uation called zero-field paraelectric resonance. In tion ions, certain rare-earth ions, including those
this case the resonance frequencies, which corre- magnetically ordered above the Curie point or Néel
spond to the peaks of observed bands, coincide point, even gases O2 and NO under elevated pres-
with the tunnel splitting frequencies. For Li+ in sure), do not obey the Curie law, but instead fol-
KBr they are 62, 65, 141 GHz (for 7 Li) and 88, low the Curie–Weiss law. The majority of non-
88, 118 GHz (for 6 Li). The 40% difference be-
transition metals (all alkali metals; alkaline-earth
tween magnitudes of the tunnel splitting of the
metals, except for Be and Al, etc.) are paramag-
isotopes 6 Li and 7 Li is consistent with the results
nets of the Pauli type with a temperature inde-
of measurements of thermal conductivity and spe-
pendent susceptibility χ = μ2B g(E F ) arising from
ci c heat, and confirms the tunneling nature of the
the conduction electrons, where g(E F ) is the den-
noncentral ion transport. The study of the loca-
sity of states at the Fermi level. Transition metals,
tion, shape and width of the resonance lines, and
of the environmental influences on these factors, which do not exhibit magnetic order, are clearly
provides information on the noncentral ions, their defined paramagnets, exhibiting high values of
interaction with one another and with the lattice, magnetic susceptibility χ and rather complicated
as well as on the properties of the solid under con- temperature dependences. This specific group of
sideration. paramagnets consists of strong band paramag-
nets (see Strong band paramagnetism). There ex-
PARAELECTRICS ist also nuclear paramagnets, nuclear spin systems
Insulators which exhibit a linear (and hyster- with spin I > 0, such as 3 He at low temperatures
esis-less) dependence between the polarization P (T < 0.1 K).
PARAMAGNETIC RELAXATION 949
time” were introduced by F. Bloch (1946). Lon- also be determined from the experiments on pulse
gitudinal relaxation is related to the change of or continuous saturation of electron paramagnetic
Zeeman energy of a paramagnet in an external resonance lines. In systems which exhibit slow
field B0 , and therefore, may take place only in the spin diffusion, information on the spin relaxation
presence of a thermal reservoir (e.g., vibrating lat- times may be gained through the method of elec-
tice, Brownian motion); longitudinal relaxation is tron spin echo.
spin–lattice relaxation, and in solids it arises from
lattice vibrations. These vibrations are responsi- PARAMAGNETIC RESONANCE, also called
ble for the change of the intracrystalline electric electron spin resonance
field, which acts upon the paramagnetic center, Resonance absorption of microwaves by a
and influences its magnetic moment through the paramagnet; a particular case of magnetic reso-
spin–orbit interaction. Transverse relaxation is not nance. From the standpoint of physics, the reason
related to the change of energy of the spin sys- for the occurrence of paramagnetic resonance is
tem, and may take place both through the spin– that the application of a constant magnetic eld
lattice interaction, and through the spin–spin in- to materials which contain paramagnetic parti-
teraction. The principal mechanism of spin–spin cles gives rise to an energy level system El , with
relaxation in dilute paramagnets is the magnetic each level characterized by a certain orientation of
dipole interaction. Spin–spin relaxation generally the magnetic moment of a particle. The transition
takes place in a shorter time interval than spin– from the j th level to the ith level, which involves
lattice relaxation, and the spin system reaches a reorientation of the magnetic moment and absorp-
quasi-equilibrium state during the spin–spin relax- tion of an energy quantum hνij , takes place under
ation; a state which may be described in terms of action of an electromagnetic wave of frequency
a spin temperature, which is distinct from the lat- νij = (Ei − Ej )/ h. Zero-field paramagnetic reso-
tice temperature. In this case, the process of spin– nance, i.e. paramagnetic resonance in the absence
lattice relaxation brings about the equalization of of external magnetic field, is much less common.
the temperatures of the spin system and the lattice. In this case, the energy level scheme arises from
One of the techniques for determining the lon- the interaction between the magnetic moments of
gitudinal relaxation time, which was proposed by the particles and the intracrystalline electric field.
C. Gorter (1936) prior to the discovery of the Depending on the nature of the paramagnetic par-
paramagnetic resonance phenomenon, consists in ticles, electron paramagnetic resonance and nu-
measuring the nonresonance absorption of the clear paramagnetic resonance are distinguished
variable magnetic field energy for a parallel mu- (the latter is ordinarily referred to as nuclear mag-
tual alignment of the constant field and the vari- netic resonance). Frequencies and intensities of
able field. Nonresonance absorption is described paramagnetic resonance lines are sensitive to vari-
by an equation, which is similar to the Debye ous fields, which surround the magnetic moments
formula for dielectric loss in dielectrics, and is of the particles, and hence, paramagnetic reso-
characterized by the time parameter τ , which is nance is applied in studies of local properties of
the relaxation time of the resultant magnetic mo- solids.
ment of the system. The parameter τ may dif-
fer from the values of the longitudinal relaxation PARAMAGNETIC RESONANCE, ACOUSTIC
time, which are measured by paramagnetic res- See Acoustic paramagnetic resonance.
onance experiments (pulsed or continuous sat-
uration), when nonequilibrium magnetization is PARAMAGNETISM (fr. Gr. παρα , near, by, and
induced by changing the populations of a par- magnetism)
ticular pair of spin levels. All methods of deter- The property of materials that possess no
mining the longitudinal relaxation time provide atomic magnetic structure (paramagnets) to ac-
equal values only in the case of systems with an quire magnetization, which partially align their
equidistant Zeeman energy spectrum, or when the magnetic moments when placed in an external
spin S = 1/2. The transverse relaxation time may magnetic eld. Paramagnetism is the opposite of
PARAMAGNETISM 951
diamagnetism. Paramagnetism was first described and in describing them the interaction of magnetic
by M. Faraday (1847). From the standpoint of moments with one another is usually ignored.
microphysics, paramagnetism is characterized by If the orientation of the magnetic moments of
certain particles of paramagnetic substances which the particles in the magnetic field is unrelated to
possess uncompensated magnetic moments m (an their motion, then the phenomenon is referred to as
exception is provided by Van Vleck paramag- orientation paramagnetism or Langevin paramag-
netism). The distinction between paramagnetism netism (see Langevin function). In this case, para-
and other types of magnetism involving uncom- magnetism is determined by independent orienta-
pensated magnetic moments is that in the case of tion of magnetic moments in the external field B,
paramagnetism in the absence of an applied mag- and is described on the basis of the Gibbs distrib-
netic field these moments are oriented randomly ution for the projections of the magnetic moments
and are independent of one another, i.e. the mean mz on the direction of the field B (on the z direc-
value of the m components of each particle is tion): mz = gμjz . With allowance made for quan-
equal to zero (in contrast to spin glasses); in ad- tization, the jz projection assumes 2j + 1 values,
dition the magnetization of the body (in contrast and the magnetic moment of N atoms is deter-
to ferromagnets) and the mean magnetic moment mined by the Brillouin function:
of any sublattice of the crystal (in contrast to an-
tiferromagnets) are also equal to zero. Typical of mmax B
M = Nmmax Bj , (2)
paramagnetism is the linear relation between the kB T
magnetic polarization M and the magnetic inten-
where mmax is the maximum value of the projec-
sity H = B/μ0 : at low values of the field intensity
tion of magnetic moment of the particle (for elec-
M = χH , where χ is the magnetic susceptibility
tronic magnetic moment mmax = gμB J ). Hence
(χ may be a tensor quantity in crystals). Another
it follows that at mmax B k B T there exists a lin-
typical feature of paramagnetism is the probabil-
ear relation between M and B, and the magnetic
ity of magnetic saturation M → Ms when increas-
susceptibility is governed by the Curie law:
ing B to values which exceed the so-called sat-
uration field Bs . The value and the temperature c N (j + 1)m2max
dependence of χ , as well as the value of Bs , are χ= , c= . (3)
T 3 j kB
determined by the properties of the particles of
nonzero m (of electronic or nuclear nature). An It is a common practice to deal with the suscepti-
electronic magnetic moment is produced by elec- bility of a unit volume, or the molar susceptibility;
trons of unfilled shells of atoms and molecules, by in these cases N is the number of magnetic mo-
collective electrons of conductors, etc.: ments per unit volume or per mole of a material,
respectively.
m = gμB J (1) The magnetic saturation (M → Ms = Nmmax )
where J is the angular momentum (of a con- is observed at B > Bs , where Bs = k B T /mmax .
duction electron, electronic shell, etc.) in units of Eqs. (2) and (3) adequately describe the paramag-
Planck constant ; g is the Landé g -factor ; μB is netism of classical paramagnets, namely, liquids
the Bohr magneton. As a rule, m is different from and gases, the molecules of which possess mag-
zero in a condensed medium in the case of inner netic moments, as well as the paramagnetism of
(d and f ) unfilled shells. Since the orbital mo- localized magnetic moments in solids on the con-
ment of the shells in crystals is usually equal to dition that the interactions between these localized
zero (“frozen” or “quenched” because of the spin– moments may be neglected. Interactions between
orbit interaction), the electronic magnetic moment the magnetic moments is responsible for devia-
of an atom in a crystal is determined for the most tions from the Curie law (see Curie–Weiss law),
part by the spin. Magnetic moments of nuclei are and may lead to magnetic ordering at low temper-
close in value to the nuclear magneton, and con- atures.
siderably less than electronic magnetic moments. Pauli paramagnetism is common for a degen-
Several types of paramagnetism are distinguished, erate Fermi gas or Fermi liquid, e.g., electrons in
952 PARAMAGNETISM, STRONG BAND
metals, liquid 3 He, etc. The onset of magnetic po- PARAMETRIC EXCITATION OF SPIN WAVES
larization (i.e. generation of unequal quantities of Excitation of spin waves in magnetic materials
fermions with different values of spin projection) (ferro-, ferri- and antiferromagnets), which are ex-
in such systems results in an increase of the mean posed to the action of an electromagnetic pumping
kinetic energy of the gas by virtue of the Pauli wave; parametric excitation is observed only when
exclusion principle. In the case of Pauli paramag- the amplitude h of the magnetic field in the wave
exceeds a certain threshold value hth . Spin waves
netism the magnetic susceptibility χp does not de-
arise in the sample at h > hth . The frequencies ω1
pend on the temperature, and for the electron gas
and ω2 and wave vectors k 1 and k 2 of these spin
is given by an equation waves are related through parametric resonance to
the frequency ω and wave vector k of the pumping
χp = μ2B D(EF ), signal as follows:
ω1 + ω2 = nω, k 1 + k 2 = nk, (1)
where D(EF ) is the density of states at the Fermi where n is the order of the parametric process.
surface. In the case of a spherical Fermi surface, The amplitude of the oscillating magnetization of
χp = 3nμB 2 /(2EF ), where EF is the Fermi en- the initial spin wave exceeds the thermal level by
ergy, n is the electron density. When these equa- six or more orders of magnitude, reaching several
tions are compared with Eq. (3), it is apparent that tenths of a percent of the saturation magnetiza-
gas degeneracy results in a decrease of suscepti- tion M0 of the material under consideration. In-
bility (by the factor k B T /EF ). Magnetic satura- dividual phases of the excited spin waves are ran-
dom variables, and it is only the sum of phases
tion may arise only in very strong fields (Bs ∝
that is single-valued with respect to the pumping
EF /μB ). Paramagnetism of the current carriers in wave phase. The experimentally obtained values
nondegenerate semiconductors is characterized by are ω1,2 ∼ 1010 –1011 s−1 , n = 1 or 2; the length
an exponential temperature dependence, which is of the pumping wave often noticeably exceeds the
due to the temperature dependence of the den- dimensions of the sample, i.e. the pumping is uni-
sity of carriers. Van Vleck paramagnetism or po- form, and hence k = 0, k 1 = −k 2 ; the value of
larization paramagnetism is related to quantum- the wave vector of the excited waves k 1,2 de-
mechanical corrections, which are due to the ad- pends on the pumping frequency and on the in-
mixture of excited states with nonzero magnetic tensity of the magnetic biasing field H in which
moments into the ground (nonmagnetic) state of the sample is located, and may range within 103 –
the atom. This kind of paramagnetism has only a 106 cm−1 .
weak dependence on temperature. The study of the The first-order parametric excitation of spin
waves (n = 1) results in the excitation of waves
paramagnetism of solids and its high-frequency
of a frequency that is equal to half the pump-
manifestations (electron paramagnetic resonance
ing frequency: ω1 = ω2 = ω/2. Depending on
or nuclear magnetic resonance) is an important the orientation of h with respect to H , transverse
method of investigation of the properties of solids. pumping (h ⊥ H ) (H. Suhl, 1957) and longitudi-
The phenomenon of paramagnetism is also used nal pumping (h H ) (E. Schlömann, 1960) are
for reaching extremely low temperatures (see Adi- distinguished. In the former case the electromag-
abatic demagnetization cooling). netic field first excites uniform oscillations of mag-
netization, which induce the paramagnetic excita-
tion of spin waves; in the latter case, the imme-
PARAMAGNETISM, STRONG BAND diate cause of the parametric relation between the
See Strong band paramagnetism. waves is the longitudinal pumping magnetic field.
The threshold value of longitudinal pumping in an
isotropic ferromagnet is given by the expression
PARAMAGNETISM, VAN VLECK
Hk ω
See Van Vleck paramagnetism. hth = , (2)
ωM sin2 θk
PARAPROCESS 953
where
Hk = γk /γ , γk is the spin wave damping scattering event is related to generation of para-
decrement; ωM = 4πγ M0 , γ is the gyromagnetic metric quasi-particles, which are known as bipho-
ratio; θk is the angle between the wave vector of tons. The sum of the frequencies of the scattered
the spin waves and H0 . The threshold value of the waves is equal to the incident light frequency:
field in perfect monocrystals of yttrium iron garnet ω1 + ω2 = ω3 . Conservation of momentum of
(see Iron garnets) amounts to several microtesla at three photons k 1 + k 2 = k 3 allows for the effec-
room temperature. Since the uniform precession tive accumulation (parametric gain) of biphotons
of magnetic polarization contributes to the process along coherent directions which are determined
of excitation of spin waves, the threshold of the by the dispersion and anisotropic properties of
first-order transverse pumping depends heavily on the nonlinear crystal. As a result, the scattered
the degree of closeness of the pumping frequency waves assume a specific frequency–angular scat-
ω to the ferromagnetic resonance frequency ω0 . tering pattern. This circumstance sets the para-
If ω = ω0 , then the threshold is particularly low metric light scattering apart from ordinary lumi-
and amounts to ∼10−8 T in yttrium–iron garnet nescence. Another distinction between the two
monocrystals. Away from the ferromagnetic reso- phenomena is that the process of parametric light
nance frequency, its value is close to the longitudi- scattering ceases within several periods of light
nal pumping threshold value. Second-order para- oscillations after the passage of the principal
metric excitation of spin waves (ω1 = ω2 = ω) wave ω3 . The correlation between photons, which
is observed only in the proximity of the ferro- constitute a single biphoton, allows one to obtain
magnetic resonance frequency ω = ω0 provided information on the photon ω1 by detecting pho-
there are no spin waves of frequency ωk = ω/2 ton ω2 . Parametric light scattering is used in the
in the spectrum, which excludes the possibility nonlinear spectroscopy of crystals, and may pro-
of first-order parametric spin-wave processes. Its vide a basis for absolute measurements of light in-
threshold amounts to several nanotesla in these tensity. In the case of a high intensity of primary
garnet monocrystals. Experimental measurement radiation ω3 , parametric light scattering goes over
of the longitudinal pumping threshold hth is the into parametric superluminescence. The latter is
most convenient way of determining the damp- characterized by an exponential increase in the in-
ing decrement of spin waves of wave vectors of tensity of waves ω1 and ω2 with an increase in the
nonlinear crystal axial length, and the power of the
the order of 104 –106 cm−1 by the application of
wave ω3 .
Eq. (2). Parametric excitation of spin waves has
found practical application in the production of
PARAMETRIC OSCILLATOR
ferrite power limiters: the damping provided by
See Optical parametric oscillator.
these devices increases abruptly at h > hth be-
cause of the onset of extra losses of signal energy PARAPROCESS, true magnetization
due to the parametric excitation of waves. The Increase of the absolute value of the magnet-
maximum operating powers of devices like filters, ization M s of ferromagnets and ferrimagnets un-
phase shifters, phase-differential circulators, etc., der the action of an external magnetic field B,
are determined by the occurrence of the parametric which takes place on completion of the engi-
excitation of spin waves: the damping provided by neering magnetization processes (see Saturation
these devices increases for h > hth as in the case magnetization curve), i.e. when the magnetic sub-
of the limiters. stance has passed into the single-domain state.
The paraprocess is related to the orientation of
PARAMETRIC LIGHT GENERATION, parametric elementary carriers of magnetism (spin and or-
luminescence bital magnetic moments of atoms or ions) in the
A phenomenon that involves the decay of in- field B. It is the final stage of magnetization, dur-
cident photons (of frequency ω3 ) in a nonlinear ing the course of which M s tends toward its ab-
crystal into pairs of photons of lower frequen- solute saturation value M 0 (i.e. the magnetization
cies ω1 and ω2 , which are reemitted by the crys- of a ferromagnet at absolute zero, and in the ab-
tal. The production of a pair of photons in a single sence of spin zero-point vibrations) with increased
954 PARENT PHASE
of partial dislocations with high-energy stacking used for the purpose of particle turning. Unlike or-
faults are possible, while in diamond-type crys- dinary channeling, surface-channeled ions of low
tals nonequivalent sets of partial dislocations ex- and average energy lose little energy by reflec-
ist, which may be described as having a structure tion from the surface producing the beam bending.
of the face-centered crystal type with a basis com- Computer simulation confirmed the high efficiency
posed of two atoms at each site of the lattice. Spe- of particle beam bending via surface channeling.
cific partial dislocations of a superdislocation type
are observed in ordered crystals.
PARTICLE EMISSION
Partial dislocations are essential for the expla-
Emission of electrons, ions, or neutral particles
nation of the twinning of crystals, phase transfor-
into a vacuum or gaseous medium by solids or liq-
mations, hardening due to the dislocation barriers
uids. In most cases the term particle emission is
that occur at intersecting dislocations, creeping,
used in names of various types of electron emis-
and cross slip processes.
sion. As a rule, the kind of action is also added:
eld electron emission (strong electric field), ex-
PARTICLE BENDING in a curved crystal
plosive electron emission (rapid, very strong elec-
Planar channeling involves a forward moving
tric field), secondary electron emission (electron
positively charged particle executing stable oscil-
bombardment), ion–electron emission (ion bom-
lations between two crystallographic planes, expe-
riencing alternating “total reflections” from each bardment), thermionic emission (heating), photo-
of them. It is possible to turn the direction of electron emission (illumination), X-ray photoelec-
a charged particle beam by flexing the crystal tron emission (X-ray irradiation), exoelectronic
planes. The radius of curvature R of a curved crys- emission (mechanical action), autophotoelectronic
tal may be evaluated from the classical equation of emission (strong electric field and illumination),
motion: and so on. The emission of molecules from heated
solids is called sublimation, and from liquids it is
mv 2 ∂U called evaporation. In adsorption a term “evap-
=− ,
R ∂r oration” is used when describing the particulate
where m and v are, respectively, the mass and ve- emission of adsorbent material; the emission of
locity of the channeled particle, and U is the av- adsorbed particles is called desorption (thermal
eraged potential of the plane. In deciding on a desorption, electron-stimulated desorption, pho-
particular radius of curvature, it is necessary to todesorption, field desorption, and so on). The
take into account that the length of the curved emission of ions by incandescent solids has an ad-
channel
l = R
θ (where θ is angle of beam ditional name surface ionization. The emission of
bending) must be much less than the dechanneling ions and neutral particles under bombardment of
distance l dec (for relativistic particles the dechan- the surface by fast ions and atoms accompanied by
nelization distance increases proportional to the surface erosion is called cathode sputtering (see
energy). However, particle beam bending for ax- Sputtering); its particular case which is used in the
ial channeling offers several advantages over beam diagnostics of the surface chemical composition is
bending with planar channeling: (1) the accep- called ion-ion emission (Secondary ion emission).
tance angle of axial channeling exceeds that of
planar channeling; (2) the number of particles in-
PARTICLE, SMALL
volved in the axial channeling mode is greater
See Small particles.
than the number accepted by the planar channel-
ing mode; (3) the dechanneling distance for axial
channeling is greater by approximately an order PARTITION FUNCTION, statistical sum
of magnitude than that for planar dechanneling. It Quantity, which constitutes the reciprocal of
is difficult to bend low and average-energy parti- the normalization factor of the Gibbs canonical
cle beams because of the strong particle dechan- distribution (see Gibbs distribution) in quantum
neling propensity. Surface channeling can also be statistical physics, and is expressed as the sum of
956 PASSIVATION
occupation possibilities over quantum states (Zus- resulting austenite is rapidly overcooled to 450–
tandsumme) Z: 550 ◦ C; at this temperature the austenite under-
goes decomposition, yielding thin-laminated sor-
En
Z= exp − , bite. The steel with such a structure is substan-
n
kB T tially hardened under strain, reaching the ultimate
where En is the energy of system in the nth quan- compression strength of 5000 MPa at total press-
tum state, and T is the absolute temperature. The ing of 99%.
summation is taken over all allowed n (includ-
ing states of equal energy). The partition func- PATH INTEGRAL, functional integral, Feynman
tion allows the calculation of all thermodynamic path integral (R. Feynman, 1948)
potentials. In particular, the Helmholtz free en- Presentation of the quantum-mechanical time
ergy given by F = −kB T ln Z can be calculated evolution operator and its matrix elements in the
as a function of temperature, volume, and num- form of an integral along a number of classical tra-
ber of particles depending on the particle interac- jectories.
tion potential. Given the system Hamiltonian H , Suppose that Ut = eit H/ is the unitary oper-
Z = Tr exp[−H /(kB T )]. In the case of an ideal ator for the time evolution of a quantum system
gas with a finite number of degrees of freedom (H
is the Hamiltonian), q is a point in configuration
εi ni
Z= exp − i . space, i.e. a set of coordinates, Ut (q1 , q2 ) is a ma-
kB T trix element of Ut in the coordinate representation.
{ni }
The integral over the paths has the form
The summation in this equation
is taken subject
to the additional condition that i ni = N , where Ut (q1 , q2 )
N is the total number of particles. The summa-
tion in the argument of the exponential function is = eiS dq(τ ),
made over all single-particle quantum states i of {q(τ ): q(0)=q1 , q(t )=q2} 0τ t
energy εi ; {ni } is the ensemble of possible values
of ni , over which the summation is taken. In the where q(τ ) is the classical trajectory, 0 τ t ,
case of Bose–Einstein statistics, ni = 0, 1, 2, . . . , the quantity S
and in the case of Fermi–Dirac statistics, ni is ei- t
ther 0 or 1. In the formalism of classical statisti- 2
S = S q(τ ) = q̇ (τ ) − V q(τ ) dτ
cal physics the partition function corresponds to
the statistical integral, the integral analogue of the 0
above discrete summations. is the classical action for the fixed path q(τ ),
V is the potential energy; integration is carried out
PASSIVATION along all paths with a fixed beginning and end.
See Surface passivation. To evaluate the integral the interval [0, t] is di-
vided into sections: 0 < t1 < t2 < · · · < tn < t ,
PATENTING the continuous path q(τ ) is replaced by a discrete
The method of heat treatment of medium- one with vertices at the points of subdivision, the
carbon low-alloy steel (carbon content 0.45– integration is performed sequentially for the vari-
0.85% C), which involves heating above the trans- ables q(t1 ), i = 1, 2, . . . , n, so the result is finite,
formation temperature, with subsequent controlled then the limit of n → ∞ is taken, corresponding
cooling. It is used in order to obtain work pieces of to a continuous succession of τ values. The inte-
high plasticity. This circumstance allows the use of gration is performed with respect to the functions
patenting prior to the production of cold-formed q = q(τ ), whence the origin of the term “func-
wires, strips and other articles, which require tional integral”.
deep drawing at room temperature. The process In the middle 1950s functional integrals were
of patenting involves heating to 870–920 ◦ C. The introduced into the quantum theory of systems
PEARLITE 957
with an infinite number of degrees of freedom, i.e. the bonding orbitals of the molecule. Maximum
into quantum field theory, orbital overlaps of two neighboring atoms provide
lowering of the energy of the electron of one atom
Ut = eiS dϕ(x), in the field of the other atom; in the case of parallel
spins of two atoms the maximum orbital overlap
where x is a point of four-dimensional space–time, provides the highest value of the exchange integral
ϕ(x) is a classical field, S = S[ϕ(x)] is the action. (involved in the equation for the negative energy of
The path integral method is used in the theory of the molecule).
solids for the description of phase transitions, po-
larons, solitons, and superconductivity. PAULI PARAMAGNETISM
See Paramagnetism.
PAULI MATRICES (W. Pauli)
Two-by-two square matrices σx , σy , σz ,
PEACH–KÖHLER EQUATION (M.O. Peach,
0 1 0 −i J.S. Köhler, 1950)
σx = , σy = ,
1 0 i 0 The equation defining the force f that is ap-
plied to a dislocation of unit length by the field of
1 0 b , where i = 1,
internal stresses: fi = eilm τl σmk
σz = , k
0 −1 2, 3. Here eikl is the unit antisymmetric tensor of
which are operators for the spin magnetic moment rank three; τ is the unit vector of the tangent to the
of particles of spin 1/2; the magnetic moment op- dislocation line, which is aligned with the sense
erator is S = (μB /2)σ , where μB is the Bohr of rotation of the dislocation line, b is its Burgers
vector, and the tensor quantity σmk is the strain de-
magneton. The Pauli matrices satisfy the commu-
tation relations viator. The Peach–Köhler equation defines a force
that is always directed perpendicular to the dislo-
[σj , σk ] ≡ σj σk − σk σj
cation line, and is independent of the hydrostatic
= 2iεj kl σl , j, k, l = x, y, z, pressure in the crystal. The Peach–Köhler equa-
tion was refined by J. Weertman (1965). The equa-
where the Levi–Civita symbol εj kl = +1 for a
tion is most commonly used for determining the
cyclic permutation of the indices (xyz, yzx, zxy),
value of the force that acts in the slip plane of the
εj kl = −1 for an anticyclic permutation of the
dislocation, and is defined as f = ni σik bk , where
indices (xzy, yxz, zyx), and εj kl = 0 if any two
n is the unit vector normal to the slip plane. In
indices are the same (e.g., xxy). In addition the
the case of a planar dislocation loop, this force is
Pauli matrices have the cyclic permutation prop-
determined by a single component of the σik ten-
erty σj σk = iσl , and they anticommute:
sor (provided two coordinate axes are aligned with
σj σk + σk σj = 2δj k I, j, k = x, y, z, the vectors n and b). For the case of an edge dis-
where δj k is the Kronecker symbol, and I is the location which is parallel to the z-axis, and has
2 × 2 unit matrix. Each Pauli matrix has a deter- the Burgers vector aligned with the x-axis, the
minant of −1, and the eigenvalues of the Pauli ma- equation for the force f assumes the simpler form
trices are equal to ±1. The Pauli matrices have f = bσxy .
the unusual property of being both unitary and
Hermitian. These matrices are generators of the PEARLITE
double-valued representations of the simple uni- A laminar micro-constituent of steels and cast
tary group in two dimensions SU(2). irons; mechanical mixture of ferrite and cementite.
Under close to equilibrium conditions, pearlite is
PAULING’S MAXIMUM OVERLAP PRINCIPLE formed by the eutectoid decomposition (see Eu-
(L. Pauling) tectoid alloys, Alloy decomposition) of austenite
A rule which allows one to predict energeti- on slow cooling according to the reaction: aus-
cally the most favorable configurations of atoms in tenite → ferrite + Fe3 C, where Fe3 C is cementite.
molecules or complexes, and to select the atomic The decomposition proceeds at constant tempera-
orbitals which make the maximum contribution to ture T = 996 K. The leading phase at this pearlite
958 PEARLSTONE
generation is cementite, the nucleation centers of generation of a lattice deformation (charge density
which appear at the boundaries of austenite grains. wave) with the wave vector Q = 2pF / (pF is the
The austenite region in the neighborhood of the Fermi momentum), which is caused by the inter-
first cementite plates becomes depleted of carbon, action between the electrons and crystal lattice vi-
and as a result, the plate-like ferrite layer devel- brations. Peierls instability is accompanied by the
ops. As this takes place, the carbon content of creation of an energy gap of width Δ at the Fermi
the neighboring austenite regions increases and a surface, and the transformation of the system to
new nucleation center of cementite appears. All the insulating state (see Peierls transition).
the carbon is concentrated in cementite, for fer-
rite contains practically no carbon at room tem- PEIERLS–NABARRO MODEL (R.E. Peierls,
perature. Granular pearlite and lamellar pearlite 1940; F.R.N. Nabarro, 1947)
are distinguished. The ferrite and cementite of Approximate model which describes the ac-
lamellar pearlite are plate-shaped, the spacing be- tion exerted upon a dislocation by forces arising
ing 0.5–1 μm. Pearlite consists of cementite up to from the discreteness of the crystal lattice struc-
12%, the balance being ferrite. The ratio of thick- ture, and from the atomic nature of the disloca-
ness of ferrite laminas to that of cementite ones is tion core structure (Peierls–Nabarro forces). The
approximately 40/3. The structure of pearlite de- tangential stress τ , which describes the interaction
pends to a certain extent on the cooling rate (the between atomic layers, lying on different sides of
higher the cooling rate, the higher the dispersity of the dislocation slip plane, is considered to be a
structure). Fine-grained modifications of pearlite periodic function of the local displacement δ of
are sorbite and troostite. The hardness of pearlite, atoms from their positions in an ideal crystal lat-
which is obtained under slow cooling, is 3·105 Pa. tice; Peierls drew on the simplest periodic rela-
Construction steels of pearlite structure exhibit ad- tion τ = G sin[2π(δ/b)], where b is the magni-
equately high values of strength and plasticity, and tude of the Burgers vector, the coefficient G is of
readily lend themselves to machining. the order of magnitude of the crystal shear mod-
ulus. Nabarro applied this model to the evaluation
PEARLSTONE
of the Peierls stress σP , which is required for the
Eruptive volcanic rock of an acidic nature, vari-
motion in the slip plane, and derived the following
ation of volcanic glasses; exhibits concentrically
equation: σP = G exp(−2πω/b), where ω is the
conchoidal separation and spheroidal (pearlitic)
dislocation width, whose magnitude is of the or-
structure, which is formed due to compression un-
der cooling: the rock is as a whole glassy, and der of b. According to this model, the value of σP
consists of pearl-like balls 1–15 mm in diame- lies within the range from 10−5 to 10−2 G. The
ter, which are either embedded in the glass indi- Peierls–Nabarro force, which acts upon the dislo-
vidually, or constitute the whole rock. The rock cation in its slip plane, is a periodic function of the
is light gray in appearance, occasionally tinged position of the dislocation center. In the context
with blue or yellow, the luster resembles that of of a continuum description of dislocations (see
wax, enamel, or silk. The rock is brittle. The den- Continuum theory of defects), the Peierls–Nabarro
sity 1.3–1.6 g/cm3 varies with the porosity. It force, which acts upon every element of the dislo-
contains up to 3–5% water of hydration. Crushed cation line in the slip plane, is introduced phenom-
pearlstone rock swells on rapid heating to 1000– enologically as a periodic function of the displace-
1200 ◦ C, increasing up to 10–20 times in vol- ment of the dislocation line η in the direction of its
ume. This modification of pearlstone is a sound- motion. This force may be compared to the peri-
absorbing, heat-insulating, and refractory material odic shift-dependence of the line density of dislo-
which is used as filler for concrete, plaster, and cation potential energy (Peierls relief)
paints.
1 2π
U (η) = U0 + UP 1 − cos ,
2 bη
PEIERLS INSTABILITY (after R.E. Peierls)
The phenomenon of the instability of the lat- where U0 is the line density of the intrinsic energy
tice of a one-dimensional metal with respect to the of the dislocation without regard for the discrete-
PEIERLS TRANSITION 959
which involves a partially insulating behavior of pendula. The effect of pendular oscillations of in-
the electron spectrum, is observed in NbSe3 ). tensity is used for exact determination of structure
The Peierls transition involves, along with re- factors (see Structure amplitude) of crystals and of
arrangement of the electron system, substantial the Debye–Waller factor for real monocrystals con-
changes in the phonon spectrum. The collective taining defects.
modes ω± (q) which arise below TP are related
to the amplitude and phase oscillations of the su- PENDULUM
perlattice. The low-frequency Fröhlich collective An ideal (mathematical) pendulum is a mass
mode ω_(q), which is related to phase oscilla- point connected by a weightless rod (filament) to
tions, is optically active and contributes apprecia- a fixed point of suspension O, and is in motion
bly to the system polarizability at low frequencies. due to the force of gravity. A physical pendulum is
At q → 0, this mode corresponds to the motion a solid body suspended by a rod at a certain point.
of the charge density wave and superlattice along The vertical plane motion of an ideal pendulum
the chains. In a constant current, the contribution is described by the equation ϕ̈ + (g/ l) sin ϕ = 0,
of the Fröhlich mode may be rather substantial in where ϕ is the angle of deviation of the rod from
electric fields that exceed a certain threshold value the vertical, l is the length of the rod, and g is
ET , which is determined by the pinning of charge the gravitational acceleration. For a sufficiently
density waves due to structural defects, or effects small ϕ, when sin ϕ ≈ ϕ, the material point per-
of commensurability with the original lattice (see forms simple harmonic oscillations with period
Fröhlich conductivity). T = 2π(l/g)1/2 . A similar motion is exhibited
In a quasi-one-dimensional Peierls insulator by a physical pendulum according to the equation
with a commensurate charge density wave, there I ϕ̈ + Mgl sin ϕ = 0 (I is moment of inertia with
may exist excitations, along with ordinary elec- respect to the suspension point, and l is the dis-
tronic excitations over the Peierls gap, which are tance between the suspension point and the cen-
related to solitons of the charge density wave am- ter of mass). The constancy of the oscillation pe-
plitude and phase. An example of a system, in riod of a pendulum makes it convenient to use it as
which these solitonic excitations play an impor- a regulator in timepiece mechanisms. A torsional
tant role, is the polymeric material (see Polymers) pendulum is a wheel, whose center point is con-
trans-polyacetylene (trans-(CH)x ). nected to a spiral spring that twists and untwists
as the wheel rotates back and forth. If angles of
PEKAR WAVES twist are not too large, the moment of elastic forces
The same as Additional light waves. may be considered proportional to the angle of ro-
tation. Thus, the motion of the torsion pendulum
PELTIER EFFECT is also of a simple harmonic nature with period
See Thermoelectric phenomena. T = 2π(k/I )1/2 , where I is the moment of inertia
of the wheel, and k is the coefficient of elasticity
PENDULAR OSCILLATIONS OF INTENSITY of the spring.
This effect of dynamic radiation scattering of
X-rays, neutrons, or electrons in monocrystals PENETRATION DEFORMATION, bursting
consists of periodical variation of intensities of di- A plastic deformation effected by a concen-
rect and diffracted beams either when the crys- trated load that penetrates through the depth of
tal thickness increases or when the angle of in- a crystal within its limited area. A strain of this
cidence of the beam onto the crystal varies close kind produces protuberances at the opposing sur-
to the Bragg angle (see Bragg law). The reason is face, normal to, or oblique with respect to the crys-
interference of two waves with slightly different tal surface undergoing deformation. Such protru-
phase velocities. In a weakly absorbing crystal, sions are called bursting figures. They feature reg-
the total intensity of the radiation is periodically ular geometric shapes that are defined by the el-
pumped from the direct beam to the diffracted one ements of slip in the deformed crystal, and by the
and back, similarly to oscillations of two coupled orientation of its faces. Bursting was first observed
PENETRATION DEPTH OF MAGNETIC FIELD 961
by M. Bauer (1882) in galena (PbS) crystals. The D = (1/3)v F li is the electron diffusion coef cient.
faster the crystal is loaded, and the lower its tem- In the vicinity of T c , the value of λE varies as
perature, the clearer is the shape of the bursting λE (T ) = λE (0)/(1−T /T c )1/4 , and it determines
figures. Penetration deformation has been stud- the specific dimensions of the region of inhomo-
ied most completely in CsCl type crystals. Stud- geneity in the resistive state of films.
ies combining a set of comprehensive techniques,
such as optical polarization method, X-radiography, PENETRATION DEPTH OF MAGNETIC FIELD
surface decoration, and selective etching demon- into superconductor
strated that bursting involves such mechanisms as A specific temperature-dependent length λ(T )
common translation slip, prism slip, faulting, and that determines the spatial scale of the penetration
displacement of point defects. The actual mecha- of an external magnetic field B(r) into a supercon-
nism depends on the crystal itself, and the condi- ductor. Under the assumption λ(T )
ξ(T ) (here
tions of deformation. ξ(T ) is the coherence length), the density of the
supercurrent (see Ginzburg–Landau theory of su-
PENETRATION DEPTH OF ELECTRIC FIELD perconductivity) in the bulk, with the exception of
into superconductor a thin surface layer of thickness ξ , is given by
The distance λE associated with branch im-
4e2
balance relaxation, which defines the space scale Js = − |Ψ0 |2 A, B = ∇ × A, (1)
of nonequilibrium processes in superconductors mc
(see Nonequilibrium superconductivity), is called where 2e is the charge of a Cooper pair. Accord-
the electric field penetration depth. The simplest ing to the Ginzburg–Landau–Abrikosov–Gorkov
case is the passage of current through a galvanic (GLAG) theory the modulus |Ψ0 | of the spatially
contact between a normal metal and a supercon- uniform, coherent superelectron wave function of
ductor. In the normal metal the current is carried the superconduction condensate Ψ0 (i.e. order pa-
by electrons, and in the superconductor it is car- rameter) is determined at T < T c by the equation:
ried by Cooper pairs. The transformation of nor- k B ξ0 #
mal electrons to Cooper pairs takes place inside |Ψ0 | ∼
= 3.06 mN(0)Tc (Tc − T ), (2)
the superconductor within the distance λE from
where N(0) is the electron density of states at the
the interface with the normal metal. In this region
Fermi surface, and ξ0 is the coherence length at
the electric field E is non-zero, so λE can be re-
T = 0. Substitution of Eq. (1) into the Maxwell
garded as the penetration depth of the longitudi-
equation and taking into account Eq. (2) leads to
nal “dissipative” electric field into the supercon-
the following expression for the penetration depth:
ductor, in contrast with its counterpart the usual
nondissipative penetration depth λ of the trans- mc2
verse magnetic field (see Meissner effect). At a λ(T ) =
16πe |Ψ0 (T )|2
2
distance close to λE , in order of magnitude, a non-
equilibrium superconducting state appears which ∼ λL (0)
=√ , (3)
is characterized by an imbalance of the popula- 2(1 − T /Tc )
tions of the electron-like (P > P F ) and hole-like
where λL (0) = c/ωp is the London penetration
(P < P F ) branches of the spectrum, where P F is
depth, and ωp is the plasma frequency. The ex-
the Fermi momentum, and P is the excitation mo-
pression (3) is valid in pure type II superconduc-
mentum. The imbalance relaxation is associated
with inelastic electron–phonon collisions charac- √ Ginzburg–Landau parameter κ =
tors when the
λL√/ξ0 > 1/ 2. In type I superconductors (κ <
terized by the mean free path lε = v F τε . These
1/ 2) the relationship between J s and A is non-
collisions have low effectiveness at low temper-
local (A. Pippard, 1953), and in the extreme limit
atures, leading to quite large values of λE , up
λL ξ0 it takes the form
to 10 μm. The theory provides the value λE =
(DτE )1/2 , where τE = (4T /(π
))1/2 τε is the e 2 ns λp
time constant of the imbalance relaxation, and Js ≈ − A, (4)
mc ξ0
962 PERCOLATION EFFECTS
where ns is the superconductor electron concen- where ξp ∝ |p − pc |−νp is the correlation length of
tration, with the Pippard penetration depth given percolation theory, νp is the critical index, and pc is
by the critical fraction of the superconducting phase.
2/3 The nature of the temperature dependence of the
1/3 λL
λP ≈ λ2L ξ0 = ξ0 ξ0 . (5) critical superconducting current Ic of the whole
ξ0 composite varies in accordance with the relation-
In “dirty” superconductors with a short electron ship between ξ s (T ) and ξp . More particularly, for
mean free path l < ξ0 , the coherence and the local ξ s < ξp , Ic ∝ (1 − T /Tc )4/3 and Ic is independent
correspondence between J s and A are violated at of (p −pc ), whereas for ξ s > ξp , Ic is proportional
the distance l, so that to (1 − T /Tc )3/2 and depends on (p − pc ) (Tc
e 2 ns l is the critical temperature of the superconducting
Js ≈ − A. (6) transition). The decrease of the size of the granules
mc ξ0
of a granular superconductor causes its transfor-
In this case the effective penetration depth be- mation into an amorphous superconductor. In the
comes λeff ≈ λL (ξ0 / l)1/2 > λL , and for the con- latter case, percolation effects give way to effects
dition l λ2L /ξ0 it significantly exceeds ξ0 , so related to the weak localization of current carriers
that in disordered alloys and amorphous supercon- in superconductors.
ductors κ = λeff /ξ0
1.
PERCOLATION THEORY, seepage theory (fr.
PERCOLATION EFFECTS in superconductors Lat. percolatio, ltration)
Particular properties of macroscopically uni- There are two main ways to define percolation
form superconducting composite materials which theory: (1) it is part of probability theory dedicated
depend on the volume fraction p of the supercon- to studying problems of the formation and statis-
ducting phase. Percolation effects manifest them- tical properties of infinite bonded clusters on lat-
selves in mixtures of superconducting materials tices; (2) it is transport theory (of mass, charge, en-
with nonsuperconducting ones, or in the case of ergy) over a random set (cluster), either discrete or
the existence of superconducting grids above the continuous. In both cases percolation theory con-
percolation threshold (see Percolation theory), or siders bonding of a very large (macroscopic) num-
in the case of superconducting regions linked by ber of elements on the condition that the bond of
weak Josephson junctions (see Josephson junc- each element with a neighbor is of a random na-
tion), as can occur in granular superconductors. ture, but involves a randomness prescribed in quite
In the latter case, percolation effects are related to a specific manner (e.g., by a random number gen-
the random scatter of system parameter quantities erator with specific properties).
(in particular, of energies Jij (T ) of weak coupling Percolation theory describes well many sys-
between granules i, j ). Taking into account perco- tems in which a geometric phase transition takes
lation effects, electromagnetic properties are de- place, provided a certain parameter reaches its
scribed on the basis of solutions of the Ginzburg– critical value in such a system. An example is the
Landau equations (see Ginzburg–Landau theory of metal–insulator transition in mixtures of conduct-
superconductivity) for grids of thin wires. The se- ing and insulating regions, where the concentra-
lection of grid structures is based on the infinite tion of metal phase is lower than a certain critical
cluster model of perturbation theory. The fractal concentration, pc . Of a similar nature is the de-
dimensionality (see Fractal) of the infinite cluster velopment of metallic conductivity in a semicon-
results in the fact that the properties of the grid ductor when its doping exceeds a critical level. An-
of an impure superconductor are defined by the other example is given by rock fracturing when the
quantity ξ s (T ) = [ξs (T )]2/(2+θ) (where θ ≈ 0.8 forces applied to a rock sample exceed a critical
for the two-dimensional grid and θ ≈ 1.7 for the value, and microcracks merge into a single ma-
three-dimensional grid), instead of the ordinary jor crack. Percolation theory is used to describe
coherence length ξs (T ). Percolation effects are filtration in porous media, hopping conductivity
important in the irregular mode, when ξ s < ξp , in doped semiconductors, Anderson localization
PERCOLATION THEORY 963
in disordered systems, wave propagation through of model, and are only defined by the space dimen-
media with a spatially inhomogeneous refractive sionality. That statement, tested by a very large
index, elasticity of polymer gels, etc. number of numerical experiments, serves as the
Many phenomena described by perturbation basis of the theory. The system behavior is con-
theory fall into the class of critical phenomena. trolled by the ratio of two spatial scales, the min-
They are characterized by a critical point at which imum length a (the lattice constant, the average
the properties of a system change sharply. In per- jump distance, the characteristic size, U (r) in con-
colation theory a change of system properties tinuum problems, see below), and the characteris-
takes place when an infinite cluster forms from oc- tic spatial dimension of the system – its correla-
cupied sites bonded together. tion length Lc ∝ a(p − pc )−ν . System properties
The basic statements of percolation theory are fluctuate quite strongly over distances shorter than
usually formulated for lattice problems. Consider Lc . At scales far exceeding Lc the system appears
a space lattice of nodes or sites, with bonds avail- homogeneous. It may then be viewed as consist-
able for linking together all nearest neighbor sites. ing of blocks of size Ldc (d is the dimensionality
In site-percolation theory each site is occupied of space into which the percolation system is “im-
with a probability p, or empty with a probability mersed”). One may speak about the volume Ldc as
(1 − p), and all nearest neighbor occupied sites the representative volume. This means that prac-
are connected together by bonds. Bonds only exist tically all the possible combinations of empty and
between nearest neighbor occupied sites. A clus- occupied sites are realized within the scale Lc , and
ter is a group of occupied sites such that each self-averaging of the specific characteristics of the
such site is bonded to at least one other occupied system, such as the infinite cluster density ρ∞ ,
site in the cluster. For small values of p all clus- electrical conductivity, density of states, etc., takes
ters are small, each containing only a few occu-
place there.
pied sites. An unlimited cluster of occupied sites is
When the system approaches the percolation
called an infinite cluster. The percolation thresh-
threshold, the size of the region in which self-
old, pc , is a particular value of p such that the
averaging takes place increases. However, as func-
probability for the existence of an infinite clus-
tions of Lc the properties of the system do not
ter P∞ (satisfies the conditions P∞ (p < pc ) = 0,
P∞ (p > pc ) = 0. A geometric phase transition change. For |p − pc | 1 we have Lc
a, and
takes place when p increases to the point pc , and there exists a range of intermediate asymptotic
isolated clusters merge to form an infinite cluster. behaviour a L Lc . In the L range all the
For p > pc the infinite cluster grows in area (or characteristics of those percolation clusters with
volume) until it fills the entire lattice at p = 1. sizes larger than L are similar to the characteris-
Around pc many physical properties of the sys- tics of the infinite cluster as a whole. This means
tem exhibit critical behavior. Examples are the in- that when averaged over many blocks of the size
finite cluster density, ρ∞ (p) ∝ (p − pc )β , its spe- Ld their properties depend on Lc , and moreover,
cific electric conductivity, σ∞ (p) ∝ (p − pc )t , follow the same laws that ρ∞ , σ∞ , . . . follow on
its diffusion coef cient for a particle randomly Lc . In that sense, no self-averaging takes place
walking through the infinite cluster, D∞ (p) ∝ within the Ldc volume and the specific charac-
(p − pc )θν (see Random walk). The infinite clus- teristics display a universal dependence on the
ter density ρ∞ (p) is an analogue of the order size of the averaging range (the power law). The
parameter in second-order phase transitions. The power indices (β, t, ν, θ) are universal character-
analogue of the susceptibility is the average num- istics of the physical properties of the system for
ber of points in a finite cluster, proportional to |p − pc | 1 for a space of a given dimensional-
|p − pc |−γ . The dynamic characteristics of an in- ity. The reason for such a universality lies with the
finite cluster (σ∞ , D∞ ) have no analogues in the statistical self-similarity (fractality) of geometric
thermodynamic theory of phase transitions. objects at L scales. In that sense, percolation the-
The dimensionless critical indices β, γ , t, ν, . . . ory is a particular case of the theory of physical
are universal, i.e. they do not depend on the choice properties of fractals.
964 PERCUSSION CENTER
The universality principle may be violated in welded set apart. The major variables of percus-
those systems whose physical properties are de- sion welding are the temperature and energy of the
termined by the local microgeometry at scales percussion, which provide the required high-rate
smaller than a. Among such systems are those de- strain of the near-contact regions of the metals be-
scribed by continuum percolation theory. Continu- ing welded. The duration of the percussion weld-
ity problems arise when studying the motion of a ing process is of the order of a fraction of a second.
particle in the field of a random potential U (r). The total welding cycle time involves a period of
The transport is only possible within the spatial preheating to welding temperature, a period of an-
range in which the particle energy E exceeds the nealing at the welding temperature to ensure uni-
potential U (r). The lower limit at which transport form heating of components in the welding area,
through the whole system is possible is the perco- and a time of deformation, which includes weld-
lation level Uc in a continuous problem. The spe- ing and cooling time.
cific electric conductivity in such systems is usu-
PERCUS–YEVICK APPROXIMATION
ally defined by the range around the saddle points
(J.K. Percus, G.S. Yevick, 1958)
of the potential relief, U (r), so that U saddle ≈ Uc .
One of the most accurate approximations for
The characteristic size of such ranges is b a.
the determination of the radial distribution function
A similar situation is encountered in different
and structure factor (see Structure amplitude) of
classes of continuous problems of percolation the-
simple Newtonian liquids, which exhibit central
ory (so-called Swiss-Cheese, Blue-Cheese), where
pairwise interactions between atoms V (r). The ra-
the resistance of the whole system is defined by dial distribution function g(r) is defined from the
the size of the constriction which appears in the following relations:
conducting matrix when it is saturated with non-
conducting inclusions of size a. The characteristic n(2) (r12 ) = ρ 2 g(r12 ),
transverse constriction size depends on (p − pc )
N(N − 1)
and on the condition b a. n(2) (r12 ) = d3 r3 . . . d3 rN
ZN
PERCUSSION CENTER 1
× exp − V (rij ) ,
See Center of percussion. kB T
i=j
Sk appears in the expression for the cross-section thermodynamic quantities are functions of the pa-
for diffuse scattering of X-rays (neutrons) by liq- rameter η only. Thus, the equation of state takes
uids, which makes it possible to determine this the form
factor directly from experiment.
pν 1 + 2η + 3η2
The direct correlation function is introduced = 1 + 4ηy(d) = .
through the agency of the Ornstein–Zernike law
Nk B T (1 − η)2
(L.S. Ornstein, F. Zernike, 1914) for the purpose The Percus–Yevick approximation in the case of
of analyzing the radial distribution function: hard spheres is employed to advantage as the ba-
sis system in the thermodynamics of simple liq-
h(r12 ) = C(r12 ) + ρ d3 r3 C(r13 )h(r32 ), uids in the context of Gibbs–Bogolyubov thermo-
dynamic perturbation theory. This approximation
where h(r) = g(r) − 1. The characteristic radius is closely related to the so-called mean sphere ap-
of C(r) is smaller than that of h(r), and the proximation, which deals with systems with a po-
Fourier-transformed image of C(r) is related to tential called the “hard spheres plus an attractive
the structure factor by a simple equation Sk = tail”: g(r) = 0 (r < d), C(r) = −βV (r) (r > d).
1/(1 − ρCk ). In the limit of low densities, the ra- The mean sphere approximation is identical to the
dial distribution function and its relation with C(r) Percus–Yevick approximation for r
d, and per-
are determined by application of the Mayer di- mits the derivation of analytical solutions for a
agram technique which leads to a cumulant ex- wide class of potentials, including the Coulomb
pansion in powers of the Mayer function f (r) = type.
exp[−βV (r)]−1. The simplest diagrams obtained
through this technique correspond to the Percus– PERFECT CRYSTAL
Yevick approximation C(r) = f (r)y(r), where A crystal of a perfect shape with crystallo-
y(r) = g(r) exp[βV (r)]; the Percus–Yevick equa- graphically equivalent faces developed to an equal
tion is of the form: degree (see Crystal faceting). Crystals close to per-
fect can grow suspended in a gently stirred solu-
y(r12 ) = 1 + ρ d3 r3 f (r13 )y(r13 ) tion. See also Ideal crystal.
× y(r23 ) exp −ρV (r23 ) − 1 . PERFECT DIAMAGNETISM
A large body of calculations shows that the Per- See Absolute diamagnetism.
cus–Yevick approximation describes the radial
distribution function very accurately over broad PERFECT DISLOCATION
ranges of parameters. This approximation is wide- A dislocation in which the Burgers vector is
ly accepted, since there exists an analytical solu- one of the vectors of a crystal translation. Accord-
tion of the Percus–Yevick equation for the phys- ing to the energetic Frank criterion (F.C. Frank) a
ically important case of a hard-sphere liquid (see perfect dislocation with Burgers vector b1 splits
Hard-sphere model), namely: into perfect dislocations b2 and b3 , if b12 >
b22 + b32 ; therefore, energetic stability is exhibited
C(r) = 0, r > d, only by dislocations with one (or, sometimes, two)
3 possible shortest Burgers vectors.
r ηa 2 r
C(r) = −a 2 + 6η(a + b)2 − ,
d 2 d
PERFECTLY RIGID BODY
1 + 2η An idealized concept used in physics and engi-
a= ,
(1 − η)2 neering when the strain of a body can be neglected,
e.g., in the assumed absence of the bending of a
3η
b=− , rod subjected to torques. Formally, the perfectly
2(1 − η)2 rigid body model corresponds to the real body
where d is the hard sphere diameter, and η = limit with the values of its elastic moduli tending
πρd 3 /6 is the degree of packing. In this case, the to infinity. The model of a perfectly rigid body is
966 PERIODIC POTENTIAL
PERIODIC POTENTIAL
The potential of a probe particle, which is a
periodic function of its coordinate r . The poten-
tial of an ideal crystal is periodic as seen by an
“excess” electron (or hole) moving throughout the
crystal. Its period is determined by the transla-
tion vector of the crystal lattice. Within the single-
electron approximation, the description of the mo-
One-dimensional periodic potential energy of an elec-
tion of an itinerent electron of a crystal in the self- tron in the Kronig–Penney model; the ordinate V0 is the
consistent eld of other electrons also leads to the electron potential energy.
concept of a periodic potential. The motion of one
of the itinerent electrons (the states of other elec-
trons being unaltered) proceeds in the total peri- the parameter γ = 2m(V − E)1/2 /2 (m is elec-
odic potential, and the potential of the hole (ei- tron mass, E is energy) tends to infinity in such
ther free or bound by a defect), which is left by a way that the quantity p = γ 2 ab/2 remains fi-
this electron. At long distances from the hole, the nite, (p sin βa)/βa + cos βa = cos ka, where β =
potential may be treated as approximately peri- (2mE)1/2 /, a is the width of the valley between
odic (interaction with the hole is a perturbation, the barriers, V0 is the barrier height. This relation
which results in scattering). The ideal crystal po- provides the k-dependence of the energy E. Its
tential is fully periodic with respect to the mo- graphical solution defines the system of allowed
tion of the center of gravity of the electron–hole and forbidden energy bands. Another model peri-
pair (i.e. the exciton). The fundamental solutions odic potential is the sine potential. The Hill equa-
of the Schrödinger equation in the periodic poten- tion with a sine potential reduces to the Mathieu
tial (called the Hill equation in mathematics) are equation (E.L. Mathieu, 1868) which has the stan-
Bloch functions (see Bloch theorem): dard form:
d2 ψ
ψkλ = exp(ikr)ukλ (r), + b − g 2 cos2 x ψ = 0.
dx 2
where ukλ (r) exhibits translational spatial sym- Its solutions have been studied in detail. A peri-
metry, and k is the quasi-momentum of the probe odic potential for a period far exceeding the crys-
particle. The energy ελ (p) of an electron corre- tal lattice constant may be produced by artificial
sponds to this eigenfunction, where λ is the in- means (see Superlattices).
dex of the energy band. The energy is a periodic
function of the quasi-momentum with the period b PERITECTIC ALLOYS (fr. Gr. περιτ ηκω ,
of the reciprocal lattice ελ (p) = ελ (p + b). This melt around)
property (periodicity) is a general property of any Alloys in which crystallization from a liquid
quasi-particle in the crystal. The particular form (the melt) proceeds through precipitation of the
of ukλ (r) and the energy spectrum of the probe solid phase α, followed by formation of the sec-
particle are determined by the explicit form of the ond solid phase β from the α-phase and the re-
periodic potential. maining liquid. In the case of three-component
A simple model periodic potential is the one- peritectic alloys, one liquid phase and one solid
dimensional Kronig–Penney potential (R. Kronig, phase produce two solid phases, or a single solid
W.G. Penney, 1931) (see Fig.), which admits an phase is produced from one liquid phase and two
exact solution. The form of this solution is rel- solid phases. The alloys are referred to as peri-
atively simple, if the barrier width b → 0 and tectoid alloys, if the starting phase is not a liquid,
PERMITTIVITY 967
but rather a solid phase, which passes into another demagnetization curve, viz., the portion of the hys-
solid phase on cooling (see also Phase diagram). teresis loop delimited by the B r and H c points.
An important characteristic of permanent magnet
PERITECTIC REACTION materials is the maximum value of the product
Phase transformation in which a liquid phase, of the coordinates on a demagnetization curve,
upon cooling, combines with a solid phase to form (BH )max , known as the maximum energy prod-
a new solid phase. It occurs in compounds which uct. The external magnetic field established by a
change their composition when they melt. permanent magnet acquires its maximum energy
when the coordinates of its operating point (B r ,
PERMALLOY (abbr. of permeable alloy) H c ) correspond to (BH )max . Permanent magnets
Iron–nickel based alloy. See High-permeability have (BH )max values amounting to 320 T·kA/m.
magnetic materials. Permanent magnets find their application in
a diversity of fields. The earliest device utiliz-
PERMANENT MAGNET
ing a permanent magnet was a compass. Perma-
A hard magnetic material which serves as a
nent magnets are employed in electric motors, in
source of a constant magnetic eld. In the simplest
electronics, automatics, and robotics, for magnetic
cases, a permanent magnet is a material (shaped as
couplings, clutches and bearings, in clock-making,
a horseshoe, a bar, a washer, etc.) that has under-
in household equipment (e.g., in loudspeakers, for
gone an appropriate heat treatment and been mag-
refrigerator door sealing, etc.) and in many other
netized up to saturation. In more complex cases,
cases. The total industrial output of permanent
a permanent magnet is a component of a system
designed to provide a constant or variable mag- magnets runs into thousands of tons per year.
netic field. Two parameters associated with the
material’s magnetization curve or hysteresis loop PERMEABILITY
are: its remanence B r or the magnetic flux den- See Magnetic permeability.
sity when the applied field H vanishes, and the
coercive field H c , which reduces B to zero (see PERMENDUR (abbr. of permeable and durable)
Fig.). As a source of magnetic field, a permanent Fe–Co alloy. See Soft magnetic materials.
magnet operates under the impact of its own de-
magnetization eld, H d , and its residual magnetic
PERMINVAR (abbr. of permeable and invariable)
flux density, B d , is always smaller than the mater-
ial remanence, B r . The quantity B d determines the A soft magnetic material based on Fe–Ni–Co
field in the air gap. It depends upon the permanent alloys (with Cr and Mo additions), which is distin-
magnet shape, on the configuration of the overall guished by its weak field dependence of the mag-
magnet system, on the values of H c and B r char- netic permeability μ over a wide range of mag-
acteristic of the magnetic material, the shape of the netic intensities, and low magnetic losses due to
hysteresis. Cooling perminvar in a magnetic field
results in induced magnetic anisotropy due to di-
rectional ordering of atoms. This results in a right-
angle hysteresis loop along the field, and a line
hysteresis loop in the transverse direction. If per-
minvar is cooled in the absence of an external field,
then the hysteresis loop assumes a specific shape,
which is characterized by twisting near the origin
of the coordinates, and by a low value of rema-
nence (“perminvar loop”).
PEROVSKITES (after the mineral perovskite with carried by these cations, and by the temperature.
the formula CaTiO3 ) The decrease of symmetry with decreasing tem-
Natural and artificial chemical compounds, perature at phase transitions occurs in follow-
which belong to the class of mixed metal ox- ing stages: cubic (m3m) → tetragonal (4mm) →
ides, halides and sulfides. The general formula of orthorhombic (mmm) → trigonal (3m). This se-
perovskites is Am B n Xp (m + n = −3p), where quence is realized completely only in BaTiO3 and
A and B are atoms of metals and Xp is F− , KNbO3 . Only a part of the transformation se-
Cl− , Br− , O2− , S2− . Mixed oxides, with va- quence is realized in the case of other compounds
lences which satisfy the equality m + n = 6, may (PbTiO3 , PbZrO3 , NaNbO3 ). The change in sym-
exist in the form of perovskites of composition metry of the perovskite structure is accompanied
A+ B 5+ O3 , A2+ B 4+ O3 and A3+ B 3+ O3 . Mixed by a size change; the volume of the unit cell of
halides of the perovskite structure are represented
the newly-generated phase is either slightly dif-
by the formulae A+ B 2+ X3 (X is F, Cl, Br) (see
ferent from the volume of initial phase unit cell,
Table 1).
or is a multiple of it. The perovskites of formu-
The perovskite structure is arranged as follows:
lae LnBO3 (Ln = lanthanide or Y; B = Al, Cr,
atoms A (of larger radius) and X form extremely
close packing by assuming positions of coordina- V, Fe, Sc, Ga), CaRuO3 and SrRuO3 are char-
tion number 12 (face-centered cubic). Atoms B acterized by an orthorhombic distortion due to
(of smaller size) occupy the octahedral positions. adaptation of the octahedral frame to cubic octa-
Octahedra BX6 are united by the sharing of ver- hedra with small-sized Ln3+ , Y3+ , etc. The de-
tices, thereby forming a three-dimensional octahe- gree of orthorhombicity increases with increasing
dral frame (see Fig.), the large cavities of which lanthanide atomic number. The high-temperature
contain atoms A. The stability of the perovskite BaTiO3 may exhibit an hexagonal structure which
structure is determined (in the case of oxides) by is close to the perovskite structure. Such a struc-
the tolerance factor t , ture is observed in BaMnO3 , BaNiO3 , LiSbO3 .
d(A–O) The undistorted octahedral perovskite frame is
t=√ , observed in ReO3 and WO3 . Filling the cavi-
2d(B–O)
ties of the frame of Ti-, Nb-, W-octahedra with
where d are interatomic metal–oxygen distances. atoms of Li, Na, K, Ca, Sr, Ln results in the
The perovskite structure is realized, if 0.8 t structure of cubic bronzes, which may be consid-
1.0. The perfect arrangement of A and B atoms ered as cation-deficient perovskites of general for-
with an undistorted “oxygen environment” ex- mula Ax BO3 (A = atom of alkali, alkaline-earth
hibits cubic symmetry (P m3m). The distortion metal, lanthanide, Cu, Al; B = Ti, Nb, Ta, W).
from cubic symmetry is determined by the ra- There exist oxygen-deficient perovskite phases
tio of the sizes of cations A and B, by charges ABO3−x (SrTiO2.5 , SrVO2.5 , BaFeO2.5 ), with
Table 1. Perovskites
A Na, K, Rb, Ca, Sr, Ba, Cd, Sc, Y, Ln∗ , Pu, Li, Na, K, Rb, Cs, Ca, Sr, Ba, Eu
Cs, Tl Pb, Eu Bi NH4 , Tl, Ag, Hg
B Nb, Ta, W, Ti, V, Mn, Fe, Co, Al, Sc, Ti, V, Cr, Mg, Ca, Sr, Ba, Ti, V, Sn, Zr,
Bi Zr, Mo, Ru, Sn, Mn, Fe, Co, Ni, Zn, Cd, Hg, Cu, Hf, Ta
Ce, Pr, Hf, Nb, Ga, Y, Rh, Bi Cr, Mn, Fe, Co, Ni
Pb, Th, U, Pu
∗ Ln = La, Ce, Pr, Nd, Sm, Eu, Gd, Dy, Ho, Er, Yb, Lu.
PERSISTENT CURRENT 969
in practice are, as a rule, thermodynamically stable the kind of phase transition occurring in the sys-
(i.e. are at thermodynamic equilibrium), yet non- tem, the coexistence of several phases at the curve
equilibrium metastable phases are also observed (or at the point) of phase transition may be al-
in a number of cases. Examples of thermody- lowed (e.g., rst-order phase transition) or forbid-
namic phases are: solid, liquid, and gaseous states den (second-order phase transition). The number
of matter, or paramagnetic (see Paramagnet) and of phases which may coexist in equilibrium is
magnetically ordered states of a magnetic mater- given by Gibbs’ phase rule.
ial, or states of the same substance characterized Usually, it is possible to cross the curve of
by different crystal structures or different electri- a first-order transition is such a manner that the
cal properties (metal, insulator, superconductor). phase transition does not have enough time to oc-
In the majority of cases a thermodynamic phase cur, and the phase remains in a metastable state
is spatially homogeneous, yet there exist phases in (metastable phase, see Metastable state) until its
which the order of homogeneity is modulated by a state variables assume those corresponding to a
wave, the wavelength of which is in general case complete lack of stability (e.g., supercooled liq-
not a multiple of the principal structural constant uid). In the case of first-order transitions, the for-
of the substance. The latter phases are referred to mation of a new phase involves the appearance
as incommensurate phases, or incommensurable of its nuclei (see Nucleation at phase transitions).
ones (see Incommensurate structure). Examples of The presence of factors which permit maintaining
incommensurate phases are those involving spin the metastable state results in hysteresis (e.g., dur-
density waves, charge density waves, waves mod-
ing magnetization reversal). Under the conditions
ulating atomic displacements from initial periodic
of phase equilibrium, first-order phase transitions
positions, a number of modulated magnetic struc-
are accompanied by heterophase uctuations. The
tures, etc.
characteristics of spatial phase boundaries in mul-
Each of the possible phase states of a sub-
tiphase systems are determined for specified val-
stance is an equilibrium and thermodynamically
ues of state variables, subject to the condition of
stable state over a certain range of values of
minimal thermodynamic potential with regard to
various quantities (temperature, pressure, com-
surface energy and the energy of interfaces.
position, electric and magnetic field intensities,
In a number of cases (e.g., a ferromagnet,
etc.). When these state variables take on val-
ues from allowed intervals, the thermodynamic or a ferroelectric with uncompensated surface-
potential of the phase becomes minimal relative
bound charges), the character of the order para-
to other phases. Multidimensional (in accordance meter is such that long-range fields arise (from
with the number of descriptive state variables) surface magnetic moments, or surface charges),
geometrical representations of the ranges of val- which increase the energy of the homogeneous
ues that correspond to equilibrium states of pos- phase. In these cases, the requirement that the to-
sible phases of a substance form compositional tal energy of the sample and its fields be mini-
diagrams, or phase diagrams. In more common mal may result in the generation of special ranges
use are cross-sections (plane sections) of these di- of values of extensive state variables (like, e.g.,
agrams, which relate a certain pair of state vari- magnetic field strength or its direction relative to
ables: e.g., two-dimensional compositional dia- crystal axes), within which the generation of in-
grams relating temperature and pressure, or tem- homogeneous multiphase states is favored. The
perature and magnetic field intensity. The regions latter states are bounded by equilibrium periodic
of thermodynamic equilibrium of a phase on such interfaces near first-order transitions across the ex-
a diagram may be closed, i.e. enclosed by a sur- ternal field (magnetic or electric). The regions of
face (line), the intersection of which (by varying a homogeneous phase are referred to as domains.
any of the state parameters) causes a phase tran- When a phase transition involves a “change in di-
sition, or it may be open or unbounded. In the lat- rection of the vector order parameter” in the neigh-
ter case, the line of phase transition on the dia- borhood of zero external field, then phases of dif-
gram ends with a critical point, at which the dif- ferent domains exhibit one and the same type of
ferences between phases disappear. Depending on ordering, but the directions of magnetization and
972 PHASE ANALYSIS
Multiphase diagrams of various types: (a) single-component pressure versus volume plot, (b) and (d)–(i) two-component
(binary) temperature versus composition plots, and (c) three-component (ternary) compositional phase diagram. In these
plots L denotes liquid, α denotes solid of component A, β is solid of component B, α + β is a solid solution, Am Bn is
a particular compound of components A and B.
important parameters are the temperature T and assumes the simplified form f = c + 1 − ϕ. Ac-
pressure P . The number of the internal para- cording to the number of components c it is sim-
meters is determined by the number of compo- ple (unitary), binary, ternary, etc. Depending on
nents. The relation between the number of com- the degrees of freedom f there are the following
ponents c, phases ϕ and degrees of freedom or types of equilibrium: invariant points at f = 0,
variances f (the number of independent variables monovariant lines at f = 1, divariant surfaces at
or coordinates which determine the geometry of
f = 2, etc. The phase diagram of a simple system
the phase diagram) is given by Gibbs’ phase rule:
is plotted in standard coordinates P versus T on
f = c + 2 − ϕ. The phase diagram of a condensed
Fig. (a). The lines of this diagram show the influ-
system is generally formed from isobaric cross-
sections of temperature–composition–pressure di- ence of external factors on the state of matter, and
agrams at the standard pressure P = 101.3 kPa = also the structure of the solid state if the variation
1 atm. A small deviation of P from 101.3 kPa of T or P leads to polymorphous transformations
does not essentially influence the phase compo- between solids α and β (see Structural phase tran-
sition of condensed systems, since no gas phase sition). Lines of a magnetic phase transition may
is present. If the pressure is fixed, the phase rule be shown on a phase diagram.
974 PHASE DIAGRAM
In order to show the composition of poly- transformation or the formation of a liquid and an-
component systems (c 2) uniform coordinates other solid phase from a solid phase β ↔ L + α
are used, where the sum of the concentrations of (Fig. (i)). In the solid state we have the eutectoid
components is equal to a constant number. Thus, or the low-temperature decomposition of a solid
e.g., the composition of the alloy of the binary solution into two solid solution phases, δ ↔ β + γ
system formed by the two components α and β (Fig. (d)); a peritectoid whereby upon cooling two
(Fig. (b)) is determined by the position of a point p solid phases form a single solid phase, β + γ ↔ α
along the horizontal length AB, aligned with the (Fig. (d)); and a monotectoid transformation α ↔
abscissa axis of the diagram, whose length is se-
β + α may occur. These differ from eutectic, peri-
lected equal to 1 or 100. The β-component con-
tectic and monotectic types since there is no liquid
centration of the alloy is proportional to the value
crystallization but rather solid state disassociation.
of the horizontal length Ap, with the percent com-
position X = (Ap/AB) · 100%. The alloy compo- The curve 1–2–3 where the crystallization
sition in the ternary system consisting of the com- starts (Fig. (d)) is called the liquidus curve, and
ponents α, β and γ is given by a point within the curve 1–4–2–5–3 where it ends (beginning of
the equilateral triangle ABC (Fig. (c)). There the melting) (Fig. (d)) is called the solidus curve. In
property is used, that the sum of distances from three-component systems we have, respectively,
any point p (composition point) within the triangle the liquidus surface and the solidus surface. The
to its three sides is a constant value, equal to the temperature gap between these curves (surfaces)
height of the triangle. The concentration of the α- of liquidus and solidus is called the crystallization
component in the alloy at point p varies inversely gap.
with the distance from the point p, to the side BC, The phase diagram of a binary system with liq-
opposite to the vertex A. The composition of a
uidus and solidus curves only is called a melting
four-component system is designated by a point
diagram. In ternary systems the projection of the
within a regular tetrahedron, the vertices of which
solidus surface at the concentration triangle is also
are labeled by the components. The content of a
component is proportional to the distance from the called a melting diagram. In the case of systems
composition point to the face opposite to this ver- with one component at room temperature and at-
tex. In order to represent equilibrium in the system mospheric pressure in the liquid state, the term
of five components it is necessary to use a pen- “solubility diagram” is used, interchangeably with
tatope, or a regular figure with five vertices equally the term “phase diagram”. The phase diagram of a
separated from each other in four dimensional system with c 3 can be displayed in the form of
space. A system of (n + 1) components is repre- isothermal cross-sections, or as polythermal cut-
sented by a polytope formed by (n + 1) equally away views at fixed concentrations of several com-
separated vertices in an n-dimensional space. ponents.
The basis for a phase diagram classification lies Most phase diagrams are plotted experimen-
either in (1) the character of phase formation in tally using data from thermal, X-ray structural,
the solid state, i.e. diagrams with unlimited sol- resistometric, local X-ray spectral, dilatometric,
ubility (Fig. (b)), with boundary solid solutions
magnetic, optical microscopy, electron micro-
(Fig. (d)), with intermediate compounds (Fig. (e)),
scopy and other methods of analysis. The complex
or (2) the type with invariant equilibrium: eutec-
method of investigating phase balances by com-
tic transformation or the formation of two phases
L ↔ α + β from a liquid; syntectic transforma- bining convential experimental studies with ther-
tion or the formation of a solid phase from two modynamic calculations is widely used. It is es-
liquids L1 + L2 = Am Bn (Fig. (h)); peritectic pecially helpful for the study of polycomponent
transformation, or the formation of a solid phase systems. A phase diagram can be the basis for the
from a liquid and another solid phase L + β ↔ α development of new materials prepared with pre-
(Fig. (f)); monotectic transformation or the for- determined properties since it identifies equilib-
mation of a solid phase plus a liquid from a liq- rium states of physico-chemical systems as
uid of another composition (Fig. (g)); metatectic a function of internal and external parameters.
PHASER 975
the source of coherent hypersound, constituting a phase difference in the resistive state of supercon-
coherent phonon generator (see Phonon genera- ducting films and whiskers occurs at the centers
tion). This is achieved by filling a high quality fac- and along the lines of such a slip.
tor Q acoustic resonator (rod-shaped crystal with Phase difference centers (PDCs) are micro-
parallel optically flat faces and its axis aligned scopic quasi-normal regions in a superconducting
along a high symmetry crystallographic direction) channel (a film or a whisker), at which the phase
with active paramagnetic centers. The amplifica- difference of the order parameter occurs in the re-
tion increases with higher power pumping. Upon sistive state. PDCs spontaneously develop as the
reaching the threshold level, when losses are com- current grows and provide the electric resistance
pensated by amplification, the phaser generates of the channel. The order parameter in a PDC is
coherent phonons at the resonant frequencies of close to zero, and the electric field is at its maxi-
the acoustic resonator. Single-mode phonon gen- mum, decreasing in the bulk superconducting re-
eration occurs only if the threshold is slightly gions over large lengths, lE . Such a decay is due
exceeded. For pumping far beyond the thresh- to relaxation of the electric imbalance of the elec-
old level, multimode generation of phonons takes tron and the hole branches of the spectrum (that
place. Phaser amplification and generation are is, due to mixing). The above length prescribes the
realized by application of paramagnetic centers size 2lE and the resistance 2l E ρn (ρn is the normal
Cr3+ and Ni2+ in a crystal of Al2 O3 , as well resistivity) of the discrete resistive domain, asso-
as Ni2+ and Fe2+ in MgO. The phenomena of ciated with the PDC. A nonstationary Josephson
phaser generation and amplification of coherent effect is observed in the vicinity of the PDC.
phonons are applied in hypersonic acoustic para- Phase difference lines (PDLs) constitute a gen-
magnetic resonance spectroscopy to study the dy- eralization of the concept of PDC to resistive states
namics of electron–nuclear spin systems in strong in wide superconducting films. Such a generaliza-
hypersonic fields, and to study nonlinear processes tion is based on the experimental observation of
related to ultralong wave propagation of sound in specific features of current–voltage characteristics
of such films that are similar to features found in
crystals containing active paramagnetic centers.
the narrow channels. For slightly supercritical cur-
rents the resistance of wide films results from the
PHASE SLIP in superconductors motion of Abrikosov vortices across the film. At
A jump (or change) in the time rate of growth a certain critical vortex velocity PDLs form along
of the phase χ of the complex order parameter the vortices, and a resistivity mechanism similar to
in superconductors. It is one of the manifestations that of channel resistance switches on.
of macroscopic quantum coherence of supercon-
ductors in their nonequilibrium state. The spatial PHASE TRANSFORMATION
gradient ∇χ of that phase determines the super- See Phase transitions.
u id velocity and the Meissner effect, while its time
derivative ∂χ/∂t prescribes the chemical potential PHASE TRANSITION HEAT
of the Cooper pairs in the condensate. This chem- See Heat of phase transition.
ical potential (1/2)∂χ/∂t adds to the electric
potential ϕ to form the electrochemical potential PHASE TRANSITION HYSTERESIS
Φ of the superconductor: Φ ≈ (/2)∂χ/∂t + ϕ. See Hysteresis in phase transitions.
The relationship between the phase difference
χ
and the difference between the electric potentials, PHASE TRANSITION INDUCED PLASTICITY
∂(χ1 − χ2 )/∂t ≈ 2e(ϕ2 − ϕ1 ), follows from the See Plasticity induced by phase transitions.
equilibrium condition ∇ 2 Φ = 0 outside the region
occupied by the field. Due to the proximity effect PHASE TRANSITION, PHENOMENOLOGICAL
the phase change in a junction between two super- THEORY
conductors separated by a thin layer of dielectric See Phenomenological theory of phase transi-
generates a nonstationary Josephson effect. The tions.
PHASE TRANSITIONS AT A SURFACE 977
PHASE TRANSITIONS IN AN EXTERNAL FIELD in an external magnetic field are observed in mag-
Changes of the characteristics of materials netic materials that have several sublattices (anti-
(magnetization, polarization, volume, compressibi- ferromagnets, ferrites). The strong fields give rise
lity, symmetry, etc.) brought about by an action of to phase transitions of the type of a spin-flip tran-
a field (electric or magnetic field, elastic stress, sition (involving the collapse of sublattices, see
gravitation, etc.) When compared with the zero Antiferromagnetism) and magnetic ordering break-
field case, phase transitions in an external field ex- down.
hibit a number of specific features. An external In the case of ferroelectrics, an applied electric
field may cause a phase transition to die out, may field brings about the collapse of a first-order tran-
shift the transition temperature Tc , or bring about sition and a shift of the boundaries of a second-
the generation of new states. Magnetic ordering order transition. Since the transitions between fer-
in a ferromagnet can take place as a second-order roelectric and antiferroelectric phases are first-
phase transition. At the application of an external order, one may notice certain parallels between the
magnetic eld B this phase transition can die out, processes that induce a ferromagnetic phase by a
abrupt changes of characteristic thermodynamic magnetic field, and those that induce a latent fer-
quantities can become smeared since at B = 0 the roelectric phase by an electric field.
magnetization is nonzero, and the symmetry of the
PHASE TRANSITIONS IN A RADIATION FIELD
magnetic subsystem is the same at various temper-
The change of phase composition of a sub-
atures. However, in ferromagnets exhibiting strong
stance under irradiation by charged or neutral par-
magnetostriction (see Magnetostriction), magnetic
ticles, or high-energy photons. Irradiation of met-
ordering often occurs as a rst-order phase tran-
als and alloys with neutrons, electrons, ions, or
sition. This phase transition does not die out at
gamma rays causes the internal energy of sys-
B = 0, but rather it is characterized by a num-
tem to increase by an amount of the order of
ber of specific features: hysteresis with respect
108 kJ/mol due to the generation of radiation
to temperature and pressure is unusually high (up
defects and electronic excitations. This extra en-
to 200 K and 700 MPa respectively; the transi- ergy may bring about phase transitions at tem-
tion temperature depends strongly on the applied peratures lower than those specified by an equi-
field: ∼10 K/T); the stability region of the fer- librium phase diagram. Besides that, the radia-
romagnetic phase (see Phase) may change con- tion field may give rise to a phase transformation
siderably in strong fields; the action of a mag- that does not proceed under equilibrium condi-
netic field within the limits of the intermediate re- tions. The radiation-induced segregation of impu-
gion separating the initial and final states allows rity atoms produces a local increase of their con-
the onset of an irreversible ferromagnetic phase; centration at the sites of point defects to the extent
the inducing fields are temperature- and pressure- that the generation of a new thermodynamic phase
dependent, and may be controlled by changing becomes possible. At the same time, segregation
the sample composition. The phenomenon of ir- stimulates additional dissolution of the impurity
reversibly inducing metastable phases is related atoms by decreasing their content in the bulk of the
to the effect of striction blocking of nuclei of the crystal. Ion bombardment of a crystal (implanta-
nonuniform state (see Nucleation at phase transi- tion) brings about the generation of solid solutions
tions). There is a possibility of complete block- over a range of concentrations that far exceeds
ing of nuclei, when the generation of these nu- the limits allowed by the compositional diagram
clei becomes thermodynamically forbidden over for the temperatures at which the irradiation takes
the whole stability region of the initial phase. place. Irradiation of crystals can also cause order-
Magnetic ordering in antiferromagnets is, as a ing or disordering of the alloy crystal lattice (de-
rule, a second-order phase transition, which may pending on experimental conditions), decomposi-
occur also at B = 0. The Néel point turns out to tion of oversaturated solid solutions, amorphiza-
be a function of the magnetic field intensity. Vari- tion of crystal structures, change of stability of the
ous magnetic phase transitions related to a change amorphous state of completely disordered struc-
of orientation of the magnetization of sublattices tures.
PHENOMENOLOGICAL THEORY OF PHASE TRANSITIONS 979
PHASE TRANSITIONS UNDER PRESSURE sures, yet it may be generated under atmospheric
Phase transitions under pressure may be ei- pressure by the action of electric field pulses.
ther first order or second order. The application
of pressure shifts the phase transition tempera- PHASE VELOCITY
ture. In the case of a rst-order phase transition, The velocity of the motion of the phase of a
the change of Tc under the action of pressure is harmonic wave that propagates without changing
related to the latent heat of phase transition L by its shape. The value of the phase velocity is equal
a simple equation: L = Tc (dTc /dP )−1 (V2 − V1 ), to the ratio of the angular frequency ω to the mag-
where dTc /dP is the value of the derivative of nitude of the wave vector k = 2π/λ, i.e. vp = ω/k.
temperature with respect to pressure at the phase For crystals, in the general case, the phase velocity
equilibrium curve, and the difference V2 − V1 is is an anisotropic quantity for quasi-particles.
the change of volume of the material at the transi-
tion. PHASON
The influence of pressure on the phase tran- A quasi-particle corresponding to the wave of
sition of ferromagnets and antiferromagnets can vibrations of the order parameter phase in an in-
be appreciable, and in the case of ferroelectrics, commensurate structure. The notion was intro-
the values of dTc /dP are of the same order of duced by A.W. Overhauser (1971). Fluctuations
magnitude. The type of phase transition may also cause the amplitude and phase of the stationary
be changed under the action of pressure. Thus, periodic distribution of the order parameter corre-
the process of magnetic ordering in the alloy sponding to the incommensurate structure to vary,
Mn0.91 Ti0.09 As is second order at pressures be- and these variations may be distributed in a wave-
low 600 MPa, and first order at higher pressures. like manner within the crystal. The density of the
The critical pressure, at which the type of phase thermodynamic potential of the system depends
transition is changed, is 150 MPa for the alloy quadratically on the phase derivatives, but not on
MnSb0.12 As0.88 . It follows from thermodynamic the phase itself. Therefore, in an ideal crystal the
considerations that the action of a magnetic field B phason spectrum does not have a gap, and its en-
is equivalent in the case of the first-order transi- ergy depends linearly on the variation of the wave
tion to the shift of the boundaries at the phase dia- vector with respect to the stationary incommensu-
gram relating temperature and pressure (see Phase rate structure.
diagram), the isotherms being given by the ex- A particular case of a phason is a u ctuon
pression P ∼ B(M2 − M1 )/(V2 − V1 ). Depend- which is formed upon self-localization of an elec-
ing on the sign of the volume change at the trans- tron near the fluctuation region of the second
formation, the boundary is shifted into the region phase. Reduction of the electron energy upon lo-
of either positive or negative pressures. Boundary calization makes a phason stable under conditions
shifts of more than 1 GPa have been observed ex- when it would be unstable in the absence of the
perimentally. Within the limits of the intermedi- electron. Phasons may appear in some tempera-
ate region of the pressure–temperature diagram, ture ranges near rst-order phase transition points
the field causes the irreversible onset of the fer- in nonmetallic systems, and may be 1 nm in size.
romagnetic phase. Materials which exhibit a spon-
taneous ferromagnetic phase generation boundary PHENOMENOLOGICAL THEORY OF PHASE
in a region of negative pressures, and ferromag- TRANSITIONS, thermodynamic theory of phase
netic phase decomposition boundary in a region transitions
of positive pressures, should be assigned to a sep- Theory of second-order phase transitions, and
arate group. In this case, even a short-time ap- of phase transitions close to first order, based on
plication of a magnetic field brings about the ir- examining the symmetry of the initial and final
reversible generation of a magnetic phase, which phases, and on the assumption that the free en-
cannot be induced in any other manner. For many ergy and other thermodynamic potentials of the
antiferroelectrics, the ferroelectric phase turns out system in the vicinity of the transition point may
to be concealed within the region of negative pres- be expressed as analytical functions of the order
980 PHONON AVALANCHE
parameter and of other thermodynamic variables. narrow spread in frequency. The rate of the relax-
The phenomenological theory of phase transitions ation is limited by the fact that these so-called hot
describes well many features of magnetic phase phonons are unable to readily transfer the exces-
transitions, ferroelectric phase transitions, struc- sive energy, obtained from the excitations, to the
tural phase transitions and superconducting phase remaining vibrations of the phonon spectrum. As
transitions, but it is not adequate for describing the a result there appears an excessive number of hot
critical region near the transition point. The Lan- phonons compared to the thermal phonons in the
dau theory of second-order phase transitions is an normal distribution of the phonon density of states
example of a phenomenological theory. at the temperature of the experiment (there is lo-
cal overheating of the hot phonons, hence their
PHONON AVALANCHE name), and the process of relaxation is retarded.
Phenomenon of a sharp (avalanche) increase The spin system and the hot phonons are thus
of the number of phonons upon their interaction raised to temperatures much higher than that of
with a strongly nonequilibrium (population in- the thermal phonons (lattice vibrations) which are
verted) physical system. The avalanche process in direct contact with the thermal reservoir or heat
takes place when the probability of phonon gen- bath (e.g., liquid helium).
eration is proportional to the number of phonons, An example of phonon relaxation occurs when
and radiation exceeds absorption. Such a propor- a high-power pulse of radio or microwave fre-
tionality is a common feature of induced radia- quency ω raises nuclear or electron spins to an
tion (in our case – of phonons). In fact a phonon excited energy state in a magnetic resonance ex-
avalanche is a realization of the idea of a laser periment, and the spins then pass their excess en-
based on phonons (phonon maser). ergy on to the lattice vibrations via the interm-
The conditions for phonon avalanche develop- mediary of phonons. At cryogenic temperatures
ment may be created when a physical system with (e.g., 4 K) the direct process of spin–lattice relax-
a finite number of energy levels (a state easily al- ation involves the excited spins transferring their
tered by external effects) is brought into contact energy to the hot phonons which are at resonance
with a system of phonons. To achieve effective (that is at the same frequency ω) with the spins,
contact it is desirable to have the energy gaps of and then this energy is transferred it to other (non-
the system coincide with the phonon frequencies. resonant) phonons which eventually pass it on
The higher the extent of the population inversion to the surrounding heat sink which has a high
of the energy levels and the lower the number of speci c heat (e.g., a liquid helium bath). If the
phonons, the more favorable the conditions are for phonons “on speaking terms with” or at reso-
the development of a phonon avalanche. For ex- nance with the spins cannot disperse their energy
ample, a crystal with paramagnetic impurities lo- to other phonons rapidly enough then the relax-
cated in an external magnetic field has an equi- ation path becomes blocked, in other words there
librium distribution of Zeeman energy levels (see is a phonon bottleneck. This can occur because
Zeeman effect). A rapid reversal of the field direc- the time constant τs-hp for transferring the spin
tion creates an inverted level population (opposite energy to the hot phonons is much shorter than
to the equilibrium one). Contact with phonons ex- the time constants τhp-tp for the hot phonons to
ists due to the spin–phonon coupling. Low temper- transfer their energy to the other nonresonant ther-
ature, high magnetic atom concentration and fast mal phonons, and is also much shorter than the
switching of the field are the necessary conditions time constant τhp-hr for the hot phonons to transfer
for the appearance of the avalanche process in the their energy directly to the heat reservoir. In other
phonon system. words: τs-hp < τhp-tp and τs-hp < τhp-hr . There
is a severe bottleneck when τs-hp τhp-tp and
PHONON BOTTLENECK τs-hp τhp-hr . Eventually the combined system
Situation which takes place during processes of (excited spins + resonance phonons) relaxes to
relaxation of an excitation in a solid via phonons, the reservoir temperature with a longer time con-
and involving a small group of phonons with a stant of the order of τhp-tp or τhp-hr , whichever
PHONONLESS SPECTRAL LINE 981
is shorter. For a mild bottleneck the overall spin– diffraction of light on sound (see Acousto-optic
lattice relaxation time T1 is a combination of the diffraction). For reception of the acoustic waves
three times τs-hp , τhp-tp and τhp-hr , and for a severe acoustic wave transducers are used.
bottleneck it depends mainly on τhp-tp and τhp-hr .
If one of these two time constants is much shorter
than the other then T1 will be approximately equal PHONON–HELICON RESONANCE
to the shorter of the two. See Helicon–phonon resonance.
PHONON ECHO
PHONONLESS SPECTRAL LINE
See Electroacoustic echo.
A narrow line in optical spectra of imperfect
PHONON GENERATION crystals due to electron transitions between defect
An excitation in a medium of nonequilibrium levels, which are not accompanied by the creation
phonons acted upon by an external influence. Dur- or destruction of quanta of crystal lattice vibra-
ing the generation process both coherent phonons tions (i.e. phonons).
(acoustic waves) and incoherent phonons (noise) The existence of a phononless spectral line
appear. The main mode of phonon generation is was predicted by M.A. Krivoglaz and S.I. Pekar
their excitation with the help of either a constant,
(1953). Depending on the strength of the inter-
a radio-frequency, or an optical electric field. High
action between the electronic transition and the
power monochromatic sound sources are called
phonons, the phononless line appears in the spec-
acoustical generators. In addition to phonons that
trum as either a rather isolated peak with weak
propagate in the bulk of the material, surface
acoustic waves can also be generated. phonon wings on its sides (weak interaction) or as
For the generation and the amplification of a relatively low-intensity peak on the background
broad-band acoustic noise, an acousto-electronic of a wide electron-vibrational band (appreciable
effect is used. This is the excitation of sound waves interaction). The width of the phononless line is
in piezosemiconductors by electrons whose drift considerably less than a characteristic phonon fre-
velocity under an applied electric field exceeds the quency. At very low temperatures for RF transi-
sound phase velocity. tions, the width approaches the natural line width
Many methods of generation are based on the determined by the probability of radiative decay
light- (laser-)induced excitation of sound. In ther- of an excited state per unit of time. With increas-
mal methods, the sound is emitted by portions of ing temperature, the phononless line intensity de-
the medium subject to an intensity-modulated or creases, its width increases, and the position of
surface scanning laser beam. High power coher- its peak shifts. The nature of these changes is de-
ent hypersound appears during Brillouin scatter- termined by the mechanism of the interaction of
ing induced by the action of a high power laser the electron transition with phonons, and by the
pulse, with a single spectral component of thermal type of phonon spectrum. Therefore, studies of
acoustic noise arising simultaneously. the phononless line provide important information
At low (liquid helium) temperatures, a phaser
not only on the electron structure of defects (en-
(phonon analogue of a laser) can be used for co-
ergy spectrum, symmetry, and so on) but also on
herent phonon generation. Its working principle is
the electron–phonon interaction. For doped disor-
the stimulated radiation of phonons by a two-level
system of spins with an inverse population estab- dered systems, phononless lines are investigated
lished under the conditions of acoustic paramag- with the help of selective laser spectroscopy meth-
netic resonance. ods to eliminate the inhomogeneous broadening.
The recording of phonons is based on the same The phenomenon of phononless lines has wide ap-
principles as their generation. Hence, at low tem- plications, in particular it underlies the operation
peratures, bolometers, avalanche diodes, and de- of solid-state lasers based on transition and rare-
tectors using acoustic paramagnetic resonance are earth element impurities, as well as the Mössbauer
employed. Optical methods are represented by the effect.
982 PHONON MAGNETIC RESONANCE
and compounds, degenerate semiconductors and zone (see Van Hove singularities). Due to the inter-
semimetals, and low-dimensional organic conduc- relationship between the absorption and radiation
tors and superconductors. The role of the phonon probabilities of materials, processes of photon–
mechanism in the superconducting metal-oxide phonon interactions contribute to the spectral de-
cuprate compounds (see High-temperature super- pendence of crystal radiation in the IR region.
conductors) has been discussed in the light of their In the phonon optical spectra arising from Ra-
unusually weak isotope effect. man scattering of light there are also resonances as-
sociated with the creation and annihilation of long-
PHONON OPTICAL SPECTRA wavelength (k ≈ 0) optical phonons, and the struc-
Optical spectra of solids arising from the inter- ture arising from many-phonon processes. Due to
action of electromagnetic radiation with phonon the difference between the selection rules of IR
excitations. Such an interaction reveals itself in absorption and Raman scattering transitions, they
the spectral dependence of the light absorption, re- generally complement each other. Phonons may
e ction of light, light radiation and light scatter- also be involved as the third participant in the in-
ing processes in the solid. Features of IR spec- teraction of photons with an electron system in in-
tra of ionic crystals in the range of frequencies trinsic light absorption edge spectra, and in lumi-
corresponding to the intrinsic transverse νTO and nescence spectra. See also Polariton.
longitudinal νLO optical phonons are often re-
PHONON–PHONON INTERACTION
ferred to as phonon optical spectra. The varia-
Interaction of vibrational excitations (phonons)
tion of the dipole moment associated with trans-
in condensed matter, caused by anharmonic vibra-
verse vibrations determines the possibility of a
tions. The phonon–phonon interaction in crystals,
resonance interaction with electromagnetic radi-
as a rule, is weak, and the frequencies obtained in
ation. According to the quantum theory this in-
the harmonic approximation are weakly renormal-
teraction leads to the disappearance of the pho-
ized by such an interaction. But due to this interac-
ton which entered the crystal, and the generation
tion there appears a dependence of the phonon fre-
of a transverse optical phonon of frequency νTO quencies, sound velocity, and elastic moduli on the
with a very small momentum (like the photon) temperature, together with a series of singularities
kphon = kphot ≈ 0. This interaction produces an in the behavior of the Debye–Waller factor, etc.
intense absorption band at the frequency νTO , and The phonon–phonon interaction plays a funda-
in reflection there appears a band at frequencies mentally important role in the relaxation processes
in the range νTO < ν < νLO (residual ray band). which lead to the spatial and temporal attenua-
If the unit cell of the crystal contains more than tion of phonons excited by an external action, thus
two ions there may appear several resonances. In in many cases influencing the value of their life-
crystals with any type of chemical bonding two- times. As a result, this interaction determines: at-
phonon interaction processes with light can occur, tenuation of elastic waves (sound), lattice thermal
whereby one photon disappears and two phonons conductivity, and the broadening of spectral lines
are simultaneously created, or one phonon is gen- associated with the scattering and adsorption of
erated and the other one is annihilated. The inten- various types of radiation when these processes
sity of such processes is several orders of magni- are accompanied by the generation or adsorption
tude below that of single-phonon resonances, and of phonons. By taking account of the phonon–
three-phonon processes of absorption are even less phonon interaction it is usually possible to re-
probable. For two-phonon processes the conser- strict consideration to several of the lowest order
vation of momentum law allows the participation terms in the anharmonic potential (the first of such
of phonons with any value of momentum, there- terms determines processes with the participation
fore, the adsorption spectrum extends from ν = 0 of three phonons, etc.). A phonon–phonon inter-
to double the maximum frequency of the crystal action in an ideal crystal conserves the phonon
lattice phonons. Maxima in the spectrum are as- wave vectors within an arbitrary reciprocal lattice
sociated with the participation of phonons corre- vector b. For b = 0 the processes are called nor-
sponding to high-symmetry points of the Brillouin mal processes, and for b = 0 they are referred to
984 PHONONS
as Umklapp processes. The latter can provide the PHONONS (fr. Gr. ϕωνη , sound)
main limiting factor of the thermal conductivity co- The quanta of crystal lattice vibrations which
ef cient of an ideal lattice. appear as a result of quantization of the normal vi-
In relaxation phenomena accompanied by scat- brations. Phonons are quasi-particles which play
tering processes, or by processes of generation and the role of elementary excitations of a vibrating
annihilation of actual phonons, the energy con- crystal. As a quasi-particle each phonon possesses
servation law is always satisfied. The laws of en- the energy ε = ω, the momentum p = k, and
ergy and momentum conservation lead to specific the velocity v = ∂ε/∂p = ∂ω/∂k, where k is the
quasi-wave vector, ω is the frequency of the cor-
restrictions on allowed processes. Three-phonon
responding normal vibration, and the frequency
processes (which determine the attenuation of
ω(k) is a periodic function of k with the reciprocal
acoustic branches) in which all three modes of vi-
lattice period. Phonons are Bose particles (bosons)
bration belong to the same branch of the spectrum,
(see Bose–Einstein statistics).
appear to be forbidden to a first approximation,
In the harmonic approximation the energy of
but they become possible if the broadening of the
the crystal excited state is represented as a sum of
highest frequency phonons exceeds the sublinear independent oscillator energies
deviation of the dispersion law (in this case they
are sometimes called Simons processes). Another E = E0 + ωα (k)Nα (k), (1)
restriction lies in the fact that at a breakup one kα
phonon should belong to a lower energy branch where E0 is the energy of zero-point vibrations,
than the source one, and at a scattering event one and the integer Nα (k) = 0, 1, 2, . . . denotes the
phonon is generated in a higher energy branch than number of phonons. The index α differentiates the
that of the disappearing ones. branches of acoustic phonons (see Acoustic vibra-
The probability of relaxation W of phonons tions) and of optical phonons (see Optical vibra-
caused by the phonon–phonon interaction depends tions). Thus the weakly excited state of a crystal
on the phonon frequency ω, and on the temper- is equivalent to an ideal gas of phonons, and the
ature T . For long-wave acoustic vibrations W ∼ ground state of the crystal is a phonon vacuum
ωT at high temperatures and W ∼ ωn T 5−n at (Nα (k) = 0). Upon thermodynamic averaging the
low temperatures (below the Debye temperature), energy of Eq. (1) provides the lattice part of the
where the coefficient n (1 n 5) depends on internal energy of the crystal, namely E = E,
various factors. For the lowest-frequency vibra- whereas in the state of thermal equilibrium the
tions (ω less than the inverse lifetime of the average number of phonons is determined by the
phonons taking part in scattering) W ∼ ω2 , and Bose–Einstein distribution
−1
it decreases when the temperature is lowered. ωα (k)
Nα (k) = exp −1 . (2)
For long-wavelength optical vibrations the scat- kB T
tering processes of a single acoustic phonon are
Taking into account the anharmonic vibrations of
forbidden; the decomposition into two (if al-
the lattice leads to the phonon–phonon interaction.
lowed by energy conservation) or several acoustic
See also Ballistic phonons, Biphonon, Interval-
phonons is possible, as is also a four-phonon
ley phonons, Magnetic phonons, Virtual phonons.
scattering process with the participation of two
acoustic modes. Similar processes also determine PHONON SCATTERING
the broadening of the lines of local vibrations in a Processes during which quantized lattice vibra-
crystal. tion particles of a condensed medium (phonons)
In crystals with impurities or other defects the undergo a change in the direction and perhaps also
quasi-momentum of the phonon–phonon interac- in the frequency of the wave vector due to their in-
tion may not be conserved, which leads to other teraction with each other (see Phonon–phonon in-
dependences of the relaxation process probabili- teraction), with lattice defects, with electrons, with
ties. spin excitations (magnons) in magnetic materials,
PHONON SPECTRUM 985
etc. Phonon scattering is the determining mecha- ωk on the wave vector k for different acoustic
nism of phonon relaxation resulting in their spatial and optical branches arising from the vibrations
and temporal damping, in broadening of the corre- of atoms or molecules interacting with one an-
sponding spectral lines, etc. other in a crystal lattice. The electrons and nuclei
During phonon scattering from defects, only which form the atoms and molecules in the crys-
the direction of the phonon wave vector changes, tal are two strongly interacting subsystems, whose
while its frequency, ω remains unchanged (elastic motions are coupled with each other. However,
scattering). Moreover, mode conversion is possi- due to the large difference between their masses,
ble in which the incoming and outgoing phonons their motions may to some extent be treated sepa-
belong to different branches. The probability of rately on the basis of the adiabatic approximation.
scattering W of acoustic phonons (see Acoustic For each random configuration of nuclei (with co-
vibrations) from point defects, such as interstitial ordinates symbolically designated by the multi-
or substitutional atoms, is proportional to ω4 and dimensional variable R) it is possible to introduce
is temperature-independent. When the phonon fre- the electronic wave function ϕ(r, R), which para-
quency approaches the characteristic frequency of metrically depends on the variable R (coordinates
the defect itself (in particular, when the wave- of electrons are designated by r). Therefore, the
length is close to the size of the defect), there energy of the electrons will depend on the nuclear
takes place a resonance enhancement of W . A no- coordinates R, as if they were constant, i.e. the
ticeable role in scattering from defects is played electron states adiabatically follow the displace-
by dislocations, grain boundaries, the crystal sur- ments (motions) of the nuclei. Further, it is also
face, etc. The probability of scattering from dis- possible to write down the classical or quantum-
locations is proportional to ω in the long-wave mechanical equations, which describe the motions
limit. More complex dependences are obtained for of the nuclei. In these equations the electron en-
phonon scattering from dislocation clusters. Col- ergy E(R) plays the role of the potential energy
lisions with grain boundaries and the crystal sur- of the nuclear subsystem. It is called an adiabatic
face may be significant in pure enough crystals at potential, and it differs for different states of the
low temperatures. The corresponding probability electronic subsystem. The expansion of E(R) in
is W ∝ L−1 where L is the characteristic size of terms of the extent of the displacement of the nu-
the grain, or of the crystal as a whole. Phonon scat- clear coordinates from their equilibrium positions
tering from electrons via the electron–phonon in- u = R − R 0 is taken into account up to fourth or-
teraction is accompanied by the absorption of the der. In the most important case, the so-called har-
phonon. In ideal crystals, when the conservation monic approximation which is quadratic in u, the
law of quasi-momentum holds, such a process is quantum vibrational equations lead to the expres-
observed only if the phonon wave vector does not sion for the crystal energy
exceed twice the wave vector of the electron (the
E= ωμ (k)Nμ (k) + E0 ,
corresponding wave vector of an electron is on
μ,k
the Fermi surface in a metal). For long-wavelength
acoustic phonons we have W ∝ ω, provided the where k is the wave vector of the vibrations,
phonon wavelength is shorter than the electron and μ designates the branches of the frequency
mean free path, and W ∝ ω2 otherwise. As for dependence on the wave vector k. The individ-
imperfect crystals, the wave vector is not necessar- ual terms of this sum are the energies a system
ily conserved during the scattering of phonons off of non-interacting linear oscillators. In addition,
electrons, which explains certain features of the this expression for E may be interpreted as the
probability of this type of scattering. sum of energies of separate quasi-particles, that
is phonons (ωphon ) with momenta k. The de-
PHONON SPECTRUM pendence of ω on k is the dispersion law of these
The set of natural vibrational frequencies of a quasi-particles, and the factor Nμ (k) is the num-
crystal. In ideal crystals phonon spectra are de- ber of phonons of the given type. Phonons obey
scribed by the dependence of phonon frequencies Bose–Einstein statistics, so their numbers Nμ (k)
986 PHONON SPECTRUM
depolarization field: increase of the field at the sur- photoelastic tensor are independent of the ampli-
face and its decrease in the bulk of the crystal. The tude for a slight deformation of the object, and as
increase of the near-surface field accelerates the a rule are determined exclusively by the internal
domain breakup processes. The effect exerted by structure of the body under consideration, and by
the nonequilibrium carriers on the domain struc- the electromagnetic wave frequency. The depen-
ture is determined by the ratio between the thick- dence of the photoelastic tensor components on
ness of the surface layer and the screening length. the deformation amplitude appears under severe
The photodomain effect was experimentally stud- strain (nonlinear elasticity effect). Typical values
ied in BaTiO3 , SbSI, and other ferroelectrics.
of photoelastic tensor components of solids fall in
the range from 0.05 to 0.2. The effect of a change
PHOTOEFFECT
of dielectric constant of materials during strain re-
Redistribution of electrons among their en-
sults in the generation of acousto-optic diffraction.
ergy levels, which takes place at the absorption
of electromagnetic radiation by a solid. A pho-
toemissive effect and a photoconductive effect PHOTOELECTRIC PHENOMENA
are distinguished. During the course of photoe-
Changes of the electronic properties of mate-
mission the electrons, which have absorbed a
photon, emerge through the solid surface into
rials which take place during exposure to light.
the surrounding medium (see Photoelectron spec- Photoelectric phenomena are brought about by the
troscopy). The photoconductive effect is responsi- transfer of energy from absorbed photons to the
ble for photoconductivity through a photoelectro- electrons of a material. The absorption of a photon
motive force (see Photovoltaic effect), and a pho- raises an electron to a higher energy state (photoef-
todielectric effect. The photoconductive effect may fect). If the excitation results in the emission of an
play an important role in nonmetallic crystals. The electron from the material, then the process is re-
photoelectric effect may involve electrons, which ferred to as photoemission (or photoelectron emis-
are located at various quantum levels of the solid, sion). The energy spectrum of the electrons emit-
ranging from loosely bound conduction electrons ted under irradiation provides information on the
to electrons bound at inner atomic shells. Hence energy band structure of the sample (see Photo-
the necessary condition for the photoelectric ef- electron spectroscopy). If the electrons excited by
fect is that the photon energy exceeds the binding the light remain inside the sample, then the process
energy of the electron at the ground level (see Ein- is referred to as the photoconductive effect. This
stein relation). The photoemissive effect requires effect is responsible for the phenomenon of pho-
additional energy equal to the work function of toconductivity. If the photon energy is insufficient
the solid. This condition determines upper limits
for the production of free electrons, and electron
to the wavelengths, which result in photoemission
transitions to bound excited states take place, then
from various electron energy levels in a solid.
a photoinduced change of the dielectric constant
PHOTOELASTIC TENSOR, elastooptic tensor
called the photodielectric effect) is observed. Un-
The tensor, which establishes the linear re- der certain conditions (if the semiconductor con-
lationship between the change of dielectric con- tains inhomogeneities, or if inhomogeneities are
stant ε, which takes place in a body under slight generated by the exposure to light) a photoelec-
strain, and the magnitude of the deformation: tromotive force arises, and a potential difference
is produced (photovoltaic effect). If the circuit is
0 −
εik = εik 0 ε0 P
εil km lmnj Unj , closed then electric current can flow in the ab-
l,m,n,j sence of a voltage source. These effects are also
where εik is the dielectric constant tensor of the observed for uniform illumination of a regular but
0 is the same tensor in the ab-
body under strain, εik unipolar crystal; and abnormally high photoelec-
sence of a strain, Pimnj is the photoelastic tensor, tromotive force values can be reached in this case
Unj is the strain tensor. The components of the (see Anomalous photovoltaic effect).
994 PHOTOELECTRIC THRESHOLD
PHOTOELECTRIC THRESHOLD
The maximum wavelength for which radiation
can still induce a photoeffect. The existence of a
red limit follows from the quantum nature of the
phenomenon: the energy required for the photon
absorption. This absorption results in a transition
between two electronic energy levels, so the pho-
ton energy cannot be less than the separation be-
tween the two states (at zero temperature). Pre-
cise experimental measurements of this red limit
remain difficult to make since other factors which
contribute to photoexcitation, such as thermal mo-
bility and electric fields, serve to blur the red limit.
Therefore its exact position is often extrapolated
from the dependence of photoeffect output on the
wavelength in the neighborhood of this limit.
PHOTOELECTRON EMISSION
See Photoelectron spectroscopy.
PHOTOELECTRON SPECTROSCOPY,
photoemission spectroscopy
The branch of spectroscopy which deals with
the spectrum of kinetic energies of electrons which
are generated as a result of the interaction of elec-
tromagnetic radiation with the material, and are
emitted from the solid into a vacuum or another
medium. Ultraviolet, and synchrotron or X-ray va-
rieties of radiation are often used as the sources of
quanta. Under exposure to electromagnetic radia-
tion, the energy absorbed by the material can be The processes of photoelectron emission from a solid.
transferred to electrons of both the valence band (a) Sketch representing initial and final (excited) energy
and the inner core levels (Fig. (a)), and if the re- states of electrons in the solid: valence band VB; in-
sulting energy of the electrons approaching the ner core levels LI,II,III and K; adsorbate energy EAD ,
surface is sufficient to overcome the potential bar- Φ = hν0 is the work function or ionization potential of
rier, then they may escape from the material and the highest energy electron of the valence band; χ is the
generate a flow of free charges in the vacuum. electron affinity; Eg is the width of the band gap. For
The general laws of photoelectron emission metals VB is the conduction band and Eg = 0. (b) Spec-
are: trum of photoelectrons in vacuo; VB is signal of valence
band electrons; AD are electrons at the adsorbate level;
(1) Each material exhibits a certain threshold LI,II,III and K are inner core levels; AU 1 , AU 2 , AU 3
value of energy: the electron binding en- are lines and bands of Auger electrons, which appear
ergy Eb or minimum photon energy above as a result of relaxation of inner core excited states (the
which photoelectron emission takes place background signals belong to scattered primary and sec-
(threshold energy). ondary electrons; Ek is photoelectron kinetic energy).
PHOTOEXCITON 995
(2) The main equation, which describes the pro- inner core electrons with a high degree of accu-
cess, is Einstein’s equation ω = Eb + KE racy (10−2 eV for UV and 10−1 eV for X-ray
which connects the energy of incident quan- radiation), and to identify the presence and con-
tum ω with Eb and the electron kinetic en- centrations of particular atoms. The use of mono-
ergy KE. If the object under investigation is a chromators makes it possible to obtain a resolution
solid, then it is necessary to include a correc- of 0.5–0.2 eV, and to investigate various specific
tion for the work function, and for charging the features of the structure of a solid. The method
sample (the latter is of highest importance for permits the study of thin surface layers 2–4 nm
non-conducting samples). thick, and using synchrotron radiation it is pos-
(3) The number of ejected photoelectrons is di- sible to investigate layers 0.8–1 nm in thickness,
rectly proportional to the irradiation intensity and to pass continuously from the surface into
(Stoletov’s law). the bulk. Photoelectron spectroscopy is used as
The absorption of luminous energy results in a method of analysis of the surface and the bulk
the emission of electrons which are excited in the of a sample. The distinctive feature of photoelec-
bulk of the sample (bulk photoelectric effect), the tron spectroscopy is that this method provides not
excitation and emission of electrons from the near- only the quantities of the atoms of a given type,
surface layer (surface photoelectric effect), and but also the valences of these atoms. By taking
the ejection of excited electrons from the surface into account the sharp dependence of the pho-
of the sample (emission from surface states). The toionization cross-sections on the energy of the in-
number of emitted photoelectrons depends on the cident quanta, which differs in the UV and syn-
density of occupied and empty electron states of chrotron spectral regions, one may determine not
the solid and its surface, on the probability of opti- only the overall density of states in the valence
cal transitions, on the probability of electron scat- band, but also the partial densities of states. Ex-
tering in the course of motion towards the sur- ploration into the angular dependences of the pho-
face, and passing through the potential barrier. The toelectron spectra is a principal method for inves-
mean free path of photoelectrons depends on their tigating of the dispersion law of solids. The ana-
energy, and determines the depth of emission. One lytical application of photoelectron spectroscopy
may control the contributions of the bulk photo- is often called electron spectroscopy for chemical
electric effect and surface photoelectric effect to analysis (ESCA) by chemists.
the process of photoelectric emission by varying
the excitation energy hν. Measurable character-
istics of photoelectron emission are (1) quantum PHOTOELECTRON SPECTROSCOPY
yield, i.e. the number of emitted photoelectrons EXPERIMENTAL METHODS
for a single photon incident upon the surface or See Experimental methods of photoelectron
absorbed by it; (2) the dependence of the quan- spectroscopy.
tum yield on the photon energy (spectral charac-
teristic); (3) the energy and angular distribution of PHOTOEMISSION SPECTROSCOPY,
electrons, and (4) the extent of their spin polariza- ULTRAVIOLET
tion. Photoelectrons exhibit a continuous spectrum See Ultraviolet photoemission spectroscopy.
from zero energy to ω − ω0 (Fig. (b)). The fine
structure of the spectrum is determined by specific
features of the density of electron states and the PHOTOEXCITON
nature of the bulk and surface scattering of elec- A quasi-particle resulting from the admixture
trons. In this connection, photoelectron emission of states of a photon and an exciton when they
plays a crucial role in the spectroscopy of elec- have a strong resonance interaction with each
tronic energy levels of solids, and in the investiga- other (see Polariton). The term “photoexciton” was
tion of electron scattering mechanisms. introduced to describe the actual elementary exci-
The method of photoelectron spectroscopy al- tations in a crystal. Meanwhile the exciton is un-
lows the determination of the binding energy of derstood as an idealized excited state of a crystal
996 PHOTOFERROELECTRICS
the process becomes in this case extremely low. pattern, through the agency of which a noise holo-
At the first stage of this classical photoprocess graphic grating is formed in the crystal (see Dif-
(in which an emulsion containing silver halide is fraction grating). The weak scattered wave may be
used) the liquid emulsion is subjected to chemical amplified by virtue of the diffraction of the inci-
nucleation (warming-up in the presence of chemi- dent wave by this lattice (holographic light am-
cally active impurities), which leads to the genera- plification). An increase of intensity of the scat-
tion of impurity centers (deep electron traps) at the tered wave brings about an increase of contrast
surface of microcrystals of AgHal. At the second of the interference pattern, and raises the diffrac-
stage, the exposure of the sample to light produces tion efficiency of the lattice, which leads to further
small clusters of Ag (latent photographic image) amplification of the scattering, i.e. the process is
generated at the first stage impurity centers. The nonlinear. Depending on the kind of holographic
third stage (development) involves specific reduc- amplification, which causes photoinduced light
tion of only those microcrystals which contain the scattering, the scattering process may be either sta-
clusters, and which form the latent image; thus the tionary or nonstationary. The properties of pho-
process of reduction is selectively catalyzed. For torefractive crystals manifest themselves in the
several atoms of Ag, which are produced under anisotropy of processes for recording and read-
the influence of light, there are up to 1010 chem- ing the data of noise holographic gratings; the ef-
ically generated Ag atoms, so the gain factor of ficiency of diffraction by these lattices depends
the primary photochemical product is very large, on their wave vector, and the polarization of the
109 . It is a common practice to put most of incident wave. Such anisotropy results in spe-
the silver halide layers through spectral sensitiza- cific angular dependences of the scattered light.
tion at the first stage of the photoprocess. Spec- The far zone region may exhibit either com-
tral sensitization is the process whereby the sensi- paratively large-angle scattering (characteristic of
tivity (naturally limited by the edge of the AgHal photoinduced light scattering, related to nonlo-
fundamental absorption band in the blue–violet re- cal mechanisms of photorefractive response and
gion of the visible spectrum) of a photographic nonstationary types of such light scattering) or
layer is extended to longer wavelengths (short- cross- and ring-shaped scattering patterns (char-
wavelength infrared) by the adsorption of sen-
acteristic of parametric amplification processes at
sitizing dyes (spectral sensitizers) on the AgHal
arbitrary local or nonlocal response conditions).
surface. Besides this development, the third stage
Isotropic and anisotropic photoinduced light scat-
(visualization) also includes the removal of unex-
tering are distinguished. The scattered light in the
posed microcrystals from the emulsion layer (pho-
isotropic process retains the state of polarization
tographic fixing).
of the incident wave; while in the case of the
anisotropic process the scattered light contains a
PHOTOGRAPHY, NONSILVER
component with orthogonal polarization.
See Nonsilver photography.
In certain photorefractive crystals (BaTiO3 ,
LiNbO3 , Ba–Sr niobate, Te, etc.) photoinduced
PHOTOINDUCED LIGHT SCATTERING
light scattering takes up most of the energy of the
Quasi-elastic nonlinear light scattering in pho-
incident beam, for a typical sample thickness of
torefractive crystals (crystals in which the index of
refraction depends on the light, see Holography). several millimeters; the increments in this process
The monochromatic light wave incident upon the may exceed 10, which can involve the amplifica-
crystal induces in it a multitude of noise waves, tion of a primary weak noise wave by a factor of
which are scattered by static optical (bulk and 105 or more.
surface) inhomogeneities. Each elementary com-
ponent of the scattered light is coherent with re- PHOTOLUMINESCENCE
spect to the corresponding elementary component Luminescence stimulated by optical excitation.
of the incident light; elementary components of By virtue of the diversity of processes of energy
scattered and incident light form an interference transfer and conversion, mechanisms of excitation
998 PHOTOMAGNETOELECTRIC EFFECT
of photoluminescence in solids may differ, de- rule, the effect is not observed even at 100–200 ◦ C.
pending on the wave length of the exciting radi- It also does not take place at high microhardness
ation. The spectrum of photoluminescence excita- indentations (high loads on indenter); this circum-
tion (dependence of luminescence yield on wave- stance indicates that the photomechanical effect
length of exciting radiation) exhibits a red-shifted may occur only in the surface layers of the mater-
edge located, in accordance with the Stokes’ rule, ial. The depth of the layer, which is plasticized by
in the spectral region before the longest wave- the incident light, is 1–3 μm. Therefore, the photo-
lengths of the photoluminescence. Vavilov laws mechanical effect is sometimes called the surface
also hold for photoluminescence. photoplasticization effect.
The nature of the photomechanical effect seems
PHOTOMAGNETOELECTRIC EFFECT to be related to photochemical transformations of
The same as Kikoin–Noskov effect. hydrogen- and oxygen-containing impurity groups
(perhaps OH− ) in the surface layers of crystals,
PHOTOMECHANICAL EFFECT although various speculations have been made on
The appreciable decrease (by 30–70%) of this subject at various times. All types of treat-
microhardness of a material, which takes place un- ments which remove these groups from the sur-
der exposure to light (discovered by L.C. Kuczyn- face layers (e.g., vacuum annealing) result in the
ski, R.H. Hochman, 1957). The photomechanical disappearance of the photomechanical effect. Hy-
effect is observed in monocrystals of germanium, drogen annealing, especially with slight oxidation
silicon, various binary and more complex semicon- of the samples, as well as prolonged standing in a
ducting materials, in certain metals (antimony, tel- damp environment, restores the effect. On the ba-
lurium), alkali-halide crystals, crystals of yttrium sis of these results the photomechanical effect is
iron garnets. Studies have been made to determine classified among the groups of phenomena which
the influence of electrically active and gaseous are known as the Rebinder effect, or the strength
impurities in crystals which exhibit the photome- reduction through adsorption during exposure to a
chanical effect. Electrically active impurities re- gaseous or liquid medium.
duce the magnitude of this effect in germanium The probability of finding practical applica-
and silicon. The photomechanical effect can even tions for the photomechanical effect is related to
vanish for concentrations of electrically active im- the deformation of articles of small sizes, pro-
purities ∼ 1025 m−3 . The same effect arises from duced in brittle materials, and to speeding up the
the incorporation of various impurities into lat- polishing of semiconducting samples. The pho-
tices of alkali-halide crystals (with the valence of tomechanical effect plays a significant role in
the impurity differing from that of the host lattice processes involving friction and wear.
ions). Very pure monocrystals also fail to exhibit
a photomechanical effect. Neutral gaseous impu- PHOTON
rities (nitrogen, helium) also have no influence on Neutral particle of unit spin, which is the quan-
the magnitude of the effect. The spectral range of tum of the electromagnetic field. The photon en-
the photomechanical effect is in the IR, and its ergy is given by E = ω, and its momentum is
magnitude increases with increasing IR light in- p = k = (ω/c)n, where = h/(2π) is the re-
tensity; until at an illumination ∼ 1 W/m2 the mi- duced Planck constant, c is the speed of light in
crohardness versus illumination curves reach sat- vacuo, ω = 2πf is the frequency of the electro-
uration. The exposure time needed to produce the magnetic radiation, and n is unit vector directed
photomechanical effect is between 0.02 and 0.5 s, along the wave vector k. If the number of pho-
although it is possible that this time interval does tons is sufficiently high, then the ensemble of them
not actually characterize the induction time for forms a transverse classic electromagnetic wave
generating the effect, but rather the ability to mea- E = eA exp[−i(ωt − kr)], where e is the polar-
sure the changes of hardness. A particular prop- ization vector of the wave, and A is its amplitude,
erty of this effect is its pronounced decrease in |k| = 2π/λ = ω/c. The following types of elec-
magnitude with an increase in temperature. As a tromagnetic waves are distinguished: radio waves,
PHOTON ECHO 999
Pulse–echo sequence, where I, II, and III are optical pumping pulses, S1 is the primary photon echo, and S2 is the
stimulated photon echo.
with each other. After the time interval τ a sec- wave of the photon echo propagates in the direc-
ond microwave pulse II turns around the direction tion of the second pumping beam (k e = k 2 ), and
of the precessions, so the spins reverse their mo- at k 3 = −k 2 the case of a reversed photon echo is
tion and gradually get closer together in phase. Af- realized (k e = −k 1 ).
ter another τ interval of time they are all again in Applications of photon echo are: spectroscopy
phase with each other, and so they radiate a sig- of extra-high resolution (determination of dipole
nal S1 called an echo. This sequence of events is moments of transitions, longitudinal and trans-
shown on the left side of the figure. verse relaxation times, etc.), space–time transfor-
A material which exhibits photon echos must mations of wave fields in real-time holography
possess phase memory, i.e. retain information of (multiplication of phase disturbances, reversal of
optical parameters of the pumping fields over a light wave fronts, optical processing of informa-
period of time which exceeds the duration of the tion).
pumping pulses, and the time intervals between
these pulses.
PHOTOPIGMENTS, photochromic materials
Photon echo has its origins in matching the
Class of light-sensitive materials which expe-
phases of oscillations of dipoles with slightly dif-
rience a reversible change of color or of opti-
ferent natural frequencies. The moment of phase
cal density under exposure to activating radia-
matching determines the moment of initiating the
photon echo signal, which depends on the pump- tion of suitable wavelengths (see Photochromic ef-
fect). Photopigments may be gaseous (aerosols),
ing conditions (widths of absorption and pumping
spectral lines, pulse intensity, etc.). This depen- liquid, vitreous, or crystalline. Depending on their
dence takes its simplest form in the case of very function, photopigments are produced in various
short and intense pulses, when the delay time τ states: in the form of liquid solutions, films, coat-
of the photon echo is equal to the interval be- ings, fibers; their sizes are not limited. Solid-state
tween two initial impulses (see Fig., left). Unlike photopigments are silicate glasses, which contain
spin echo, the photon echo exhibits sharply de- crystals of silver halides, activated alkali-halide
fined directionality, which results from the prop- crystals, oxides, and salts of alkaline-earth metals
agation and interference of light waves. The direc- (CaF2 :La, Ce), polymers, etc. The principal char-
tion of the photon echo wave is determined by the acteristics of photopigments include the absorp-
wave vector k e , and is subject to the conditions of tion spectra of initial and photoinduced states, the
vector alignment (synchronism). In the case of a spectral region and the value of the light sensitiv-
primary photon echo S1 , this condition (given by ity, rates of darkening processes, reversibility of
the equation k e = 2k 2 − k 1 ) is satisfied only in the photochromic cycle.
the case of low angles between k 1 and k 2 (pump- Most photopigments undergo photochromic
ing vectors I and II shown in Fig.), i.e. at k 2 = transformations under exposure to UV and short-
k 1 . In the case of a stimulated photon echo S2 , wavelength visible radiation. Application of vari-
brought about by a third light pulse III, the align- ous sensitizers or preliminary excitation with UV
ment k e = k 3 + k 2 − k 1 and phase matching may irradiation (optical sensitization) facilitates the ex-
be observed in several directions (in particular, in tension of the light sensitivity to longer wave-
Bragg directions). For example, at k 3 = k 1 the lengths (spiropyranes, photochromic glasses).
PHOTORESISTORS 1001
In theory the limiting light sensitivity of pho- of
n by virtue of the linear electrooptical ef-
topigments is 0.25 m2 /J, but actual values of their fect. This mechanism of photorefraction is the
light sensitivity are smaller by 2 to 3 orders of prevailing one in the case of ferroelectrics like
magnitude. The time constant for the transition of LiNbO3 :Fe.
a photopigment from its initial state to the pho- (2) Optical recharging of impurity centers, which
toinduced state is determined for the most part by results in a photoinduced change of the spon-
the intensity of the exciting radiation, and by the taneous polarization (see Ferroelectricity).
duration of the photochemical processes, the latter (3) Generation of local electric fields via nonuni-
being 10−13 –10−7 s for most photopigments. The form illumination of the crystal (diffusion
lifetime of the photoinduced state of a photopig- mechanism); this mechanism is also observed
ment varies over a wide range depending on the in nonferroelectric electrooptical photocon-
type of material; and at room temperature it may ductors.
range from several microseconds to many months (4) Pyroelectric mechanism, according to which
and even years. The limits of variations of the life- the field arises through the agency of heating
time of a given photopigment are temperature de- via the pyroelectric effect.
pendent. The change of birefringence
n, which takes
Certain photopigments exhibit fatigue, i.e. a de- place at illumination of an electrooptical crystal
gredation or loss of photochromic properties dur- in the external field, is also attributed to photore-
ing the process of storage and operation, yet most fractive phenomena. In this case the photorefrac-
inorganic photochromic crystals survive more than tion is caused by the screening of the external field
104 darkening–decoloration cycles with no change applied to the crystal by nonequilibrium charge
of properties. carriers. This kind of photorefraction has found
an application in the production of electroopti-
PHOTOREFRACTION, photorefractive effect cal recording of information through the process-
Photoinduced change
n of the refractive ing of crystals of the sillenite group (Bi12 GeO20 ,
index n in ferroelectric and piezoelectric mate- Bi12 SiO20 ).
rials (i.e. photorefractive materials); it is some-
times called “optical distortion”. From the PHOTOREFRACTIVE CRYSTALS
practical point of view, the phenomenon of pho- See Holography.
torefraction consists in the reversible change of
birefringence during local illumination of a piezo- PHOTOREFRACTIVE EFFECT
electric or ferroelectric crystal with intense trans- See Photorefraction.
mitted light (focused laser beam). The magni-
tude of this change reaches 10−4 –10−3 for certain PHOTORESISTIVE EFFECT
crystals (LiNbO3 , LiTaO3 ); the memory time of See Photoconductivity.
the effect ranges over fairly broad limits (from
milliseconds for BaTiO3 to months in the case of PHOTORESISTORS
LiNbO3 ). Photorefraction has found an applica- Optoelectronic devices (see Optoelectronics),
tion in the recording of three-dimensional holo- whose operation is based on the change of resis-
grams. The recording of a hologram is carried out tance of the active element at exposure to light (see
through volume modulation of
n, which corre- Photoconductivity). Unlike electromechanical po-
sponds to modulation of the recording beam, and tentiometers, photoresistors do not have moving
the resolution is quite high (see Holography). mechanical contacts, so there is a sharp increase
Photorefraction may be produced by a number of the speed of response, a raising of the thresh-
of mechanisms: old of sensitivity, and an improvement of the re-
liability. Along with linear photoresistors, which
(1) A bulk photovoltaic effect leads to the gener- have a uniform distribution of resistance, there ex-
ation of a strong electric field in an isolated ist widely used functional photoresistors; and to
ferroelectric, and this field induces the change assure that these latter devices have a favorable
1002 PHOTOTHERMOPLASTIC MATERIALS
the surface temperature Ts . At room temperature potential for separation at an electrode, polarizabil-
τ ∼ 10−9 s, and with the reduction of Ts to a value ity and diamagnetism, light absorption spectrum,
of the order ∼100 K, the lifetime τ increases up uorescence spectrum, magnetic moment of nu-
to times of the order of ∼1 s. At equilibrium the cleus and the electron shells, etc. All the methods
rates of adsorption and desorption are equal. An based on these properties, are, in essence, physical
equilibrium coating, prepared under ordinary cir- ones. They may be called physicochemical when
cumstances, amounts to fractions of a monolayer. a chemical reaction is included in the preparation
For an adsorbent surface temperature Ts close to or transformation of the material. For example, the
the adsorbate condensation temperature, and also transformation of the material to the gaseous state
under high pressure, the equilibrium coating may for mass-spectrometric analysis (see Mass spec-
increase up to a monolayer, or even to a multilayer. trometry); the attachment of fluorescent markers to
The dependence of the monolayer coating on the specific molecules for fluorescent analysis.
pressure under constant temperature conditions is
described by the Langmuir adsorption isotherm. PHYSICS OF METALS
Division of physics studying the nature, struc-
PHYSICOCHEMICAL MECHANICS
ture, composition and physical properties of
A branch of physical chemistry which studies metals and alloys, as well as the processes that
the relationships between the properties of multi- take place when metals and alloys are exposed
phase (dispersed) systems, and the physicochem- to treatments and other influences. Since physical
ical properties of the interface surfaces between
concepts and physical methods of investigation are
the phases. Methods of physicochemical mechan-
widely employed in metal science and metallurgy,
ics are used to prepare dispersed systems with
the physics of metals plays an important role in
pregiven properties (strength, durability, porosity,
preparing alloys with preset properties.
etc.).
A major focus of present day research is the
study of the electronic structure and characteris-
PHYSICOCHEMICAL METHODS OF ANALYSIS
tics (electronic, magnetic, spectral, etc.) of met-
Methods used for the qualitative and quan-
als and alloys, their properties associated with the
titative determination of the composition of a
arrangement and motion of atoms (elasticity, plas-
substance (see Qualitative analysis, Quantitative
ticity, thermal, diffusion properties, etc.), as well as
analysis). They are based on the specifics of the
special features and kinetics of phase transitions.
individual physical properties of the separate com-
Of particular importance is investigating the dis-
ponents or phases of the material. The goal of
ruption of crystalline regularity that results from
physicochemical methods of analysis is to com-
pletely ascertain material compositions over a defects (see Defects in crystals) and imperfectly
broad range of concentration. In complicated sys- ordered arrangements of atoms at alloy crystal lat-
tice sites. The theory of metals is based on ap-
tems this goal is seldom reached by any one
method of analysis, but the combined application plying the quantum theory of solids, statistical
of several methods allows one to approach it. It is physics, and physical kinetics. Of wide use in the
customary to select some specific property of the experimental studies of metals are electron diffrac-
component under investigation, and to construct tion analysis, neutron diffractometry, X-ray struc-
a process for bringing about its quantitative de- ture analysis, as well as electron microscopy, elec-
termination, or even, in some cases, to separate tron resonance, electrical, magnetic, and other in-
it from the other components. Typical properties vestigation methods.
on which physicochemical methods of analysis are
based, are: thermal conductivity, speci c heat, heat PICO. . .
of combustion, latent heat of phase transition, dif- Prefix for a physical unit name, which allows
fusion coef cient of particles, mass of molecule obtaining a multiple unit, equal to 10−12 initial
or ion, size of molecule, adsorption energy, ion- units. Symbol: p. Example: 1 pF (picofarad) =
ization potential, electron af nity, electrochemical 10−12 F.
1004 PIEZOELECTRIC CERAMICS
Functional Material ε33 /ε0 kp dij ·1012 , C/N gij ·103 , V·m/N Loss tangent QM Tc , ◦ C
group type
d31 d33 g31 g33 E < 25 E 30
kV/m kV/m
1 PZT-19 1725 ± 325 0.4 100.0 200.0 10.6 24.0 0.035 – 50 290
PZTNV-1 2250 ± 560 0.45 165.0 400.0 – – 0.015 – 70 240
RPC-1 700 0.62 – – 16.5 38 0.02 – 90 355
2 PZT-23 1075 ± 225 0.43 100.0 200.0 – – 0.0075 0.04 200 275
PZTB-3 2300 ± 500 0.45 125.0 320.0 – 15 0.012 0.035 200 180
RPC-6 2300 ± 500 0.64 195.0 440.0 – 22 0.004 0.01 1100 230
Functional Material ε33 /ε0 kp d31 ·1012 , g31 ·103 , Loss tangent, QM δfθ /fr Tc , ◦ C
group type C/N V·m/N E = 225 kV/m
−60 to −10 to
+85 ◦ C +50 ◦ C
3 PZT-22 800 ± 200 0.2 27.0 – 0.025 400 0.435 0.18 320
NBC-3 1800 ± 400 0.2 45.0 – 0.020 300 1.45 0.60 250
PZT-35 1000 ± 200 0.38 70.0 – 0.025 550 0.50 0.20 300
DPC-329 320 ± 50 0.43 40.0 – – 1000 50 ± 10 – 330
RPC-15 950 0.45 75.0 – 0.006 2500 0.30 0.10 325
4 PZT-21 550 ± 150 0.2 26.6 5.5 0.025 100 2.90 1.20 400
RPC-24 480 0.53 64.0 15.1 – 200 3.80 – –
3. Materials featuring high durability and tem- of the system. Hence, the piezoelectric effect is
perature stability of their resonance frequency, observed in crystalline materials belonging to 20
used to build filtering and other selective de- of the 32 symmetry point groups. Among them
vices operating with volume waves. Materi- are nonopolar piezoelectrics (e.g., quartz, potas-
als with high electromechanical bonding fac- sium chlorate and bromate, nickel sulfate), and po-
tors are also preferable for this application; the lar piezoelectrics or pyroelectrics (see Pyroelec-
width of their transition band and their me- tricity), e.g., tourmaline, lithium sulfate, potassium
chanical qualities are responsible for the sharp tartrate. A subgroup among the pyroelectrics is
peak of the amplitude-frequency characteristic comprised of ferroelectrics: monocrystals (e.g.,
within the transition band. Rochelle salt, barium titanate, ammonium phos-
4. High temperature materials to produce ele- phate) and polarized polycrystals – the piezoelec-
ments that remain operable at temperatures in tric ceramics; piezosemiconductors (e.g., cadmium
excess of 300 ◦ C and feature higher stability of sulfide, zinc oxide, antimony sulfoiodite); organic
their piezoelectric parameters in the prescribed solid piezoelectrics (e.g., wood) and piezoelectric
temperature and mechanical stress ranges. The polymers (such as polyvinylidenefluoride), and
parameters of some of materials from groups 3 electrets featuring a preferential direction in their
and 4 are presented in Table 2. piezoelectric effect, i.e. the so-called piezoelectric
texture found in them.
PIEZOELECTRICITY (discovered in 1880 by The mechanism of direct piezoelectricity is re-
P. and J. Curie in quartz and other crystals) lated to the appearance or change of an electric
A phenomenon consisting in the appearance moment, i.e. to the spatial redistribution of electric
or change of electric polarization as a result of charge in the structural elements of solids (e.g.,
mechanical stress (direct piezoelectric effect), or crystal lattice unit cells) that develops under the
of mechanical strain under the effect of an elec- effect of mechanical stresses produced by com-
tric field (inverse piezoelectric effect) in certain pression, extension, shear. When the electric field
anisotropic insulators and semiconductors (piezo- affects elementary charges in a structure they suf-
electrics). The inverse piezoelectric effect should fer displacements, the average distances between
be distinguished from the electrostriction that is them change, and a deformation takes place (in-
observed in all materials. In the piezoelectric ef- verse piezoelectric effect).
fect the strain depends linearly on the electric Piezoelectricity in various materials is charac-
field, while it is quadratic in that field for elec- terized by the values of their piezoelectric con-
trostriction. The necessary condition for the ex- stants which are components of third-rank ten-
istence of a piezoelectric effect is the absence of sors, di,kl , gi,kl , ei,kl , hi,kl (i = 1, 2, 3 and k, l =
a center of symmetry in the structural elements 1, 2, 3) that determine the mutual relation between
1006 PIEZOELECTRIC MATERIALS
the components of electric polarization Pi (elec- significant interest are also piezopolar films and
tric ux density Di ) and those of mechanical stress piezoelectric composite structures offering a num-
σkl (e.g., Pi = kl di,kl σkl ), of strain ukl and
ber of construction advantages, as well as piezo-
the electric field Ei (e.g., ukl = i di,kl Ei ) in semiconductors characterized by a strong electron–
the corresponding equations of state. The ten- phonon interaction.
sor components di,kl are symmetric in the last
two indices k and l. These components, which
PIEZOELECTRIC MATERIALS
are most often used in physics and technology,
are called piezomoduli (e.g., quartz has d1,11 = Materials used to manufacture piezoelectric el-
ements for various functional devices in radio-
2.31·10−12 C/N, the BaTiO3 piezoelectric ceram-
electronic technology: ceramics, composite mate-
ics have d3,33 = 10−10 C/N). Important charac-
rials and polymeric materials. The material may
teristics of piezoelectricity are the electromechan-
be textured, polarized, made single-domain by su-
ical bonding factors, ki,kl = di,kl /(εii Snn )1/2 ,
where εii and Skk are, respectively, components perposing a constant electric field or by way of
of the tensors of the dielectric constant and elas- a corona discharge or mechanical deformation.
tic compliance. They define the ratio of the energy The operations of sound and ultrasound recorders,
of strain due to the piezoelectric effect, to the total adapters, microphones, deformation sensors, me-
energy of the electric field acting on the piezoelec- chanical-electric generators of electric energy are
tric (e.g., k1,11 = 0.095 for quartz, and k3,33 = 0.5 based on the direct piezoelectric effect, while
for BaTiO3 ). the operations of sound and ultrasound emit-
Currently more than a thousand piezoelectrics ters, piezoelectric motors, atomic force micro-
are known. Among them are natural materials scope, scanning tunneling microscope, piezoelec-
(i.e. found in the natural environment, such as tric transformers of voltage and current, are all
quartz, tourmaline) and artificial piezoelectric ma- based on the inverse piezoelectric effect (see Pie-
terials which are synthesized in the laboratory or zoelectricity).
processed electrically or mechanically to produce The operation of resonance frequency stabiliz-
a piezoelectric texture. Such are, e.g., piezoelec- ers, sensors of pressure, humidity, chemical com-
tric ceramics, piezoelectric polymers, and piezo- position, etc., and piezoelectric filters is based on
electric composite structures (the latter combine the piezoelectric resonance effect. Piezoelectric el-
crystalline piezoelectrics with polymer materials), ements are manufactured preoriented in their crys-
and also electrets. tallophysical directions in such a way as to provide
Among materials belonging to the 20 groups a quality factor (QF), optimal for this or that par-
without centers of inversion the most powerful ticular application. The value of the QF is given in
piezoelectric properties are displayed by ferro- relative units with respect to piezoquartz (QF = 1).
electrics: single crystals and polarized polycrys- In the reception mode the QF for piezoelectric ce-
talline compounds (piezoelectric ceramics) be- ramics reaches 2–4; its value is 8 for lithium nio-
longing to the group of pyroelectrics (10 such bate; 10 for lithium tetraborate; 10–12 for polar-
groups are known). Such crystal compounds dis- ized polyvinylidenefluoride; 20 for Rochelle salt;
play abnormally strong piezoelectricity because 3–30 for piezoelectric composites. In the emission
of a ferroelectric phase transition which differ- mode, the QF for quartz is 1; it is 13 for Rochelle
entiates them from the more common, so-called salt; 15 for lithium niobate; 50–150 for piezo-
linear pyroelectrics (see Ferroelectricity, Electro-
electric ceramics. The best materials for piezo-
mechanical effects). Some examples of these are
electric resonators and other devices using them
Rochelle salt, antimony sulfoiodide, piezoelec-
are piezoquartz, lithium tantalate, selenites, zinc
tric ceramics based on barium titanate or lead
oxide.
titanate – zirconate. Piezoelectric materials find
many technological applications, in particular in
electro-acoustics as active elements in piezoelec- PIEZOELECTRIC MODULI
tric transducers (see Acoustic wave transducer). Of See Piezoelectricity.
PIEZOSEMICONDUCTORS 1007
fect (see Piezoelectricity, Piezoelectric materials), compression) of oriented thin plates. The maxi-
acoustic waves in crystals excite an electric field at mum permissible stresses are determined by the
a frequency coinciding with that of these waves. mechanical properties of the materials. Specimens
Such a field affects free current carriers, elec- are cooled to helium or nitrogen temperatures to
trons and holes. Piezosemiconductors play the narrow their spectral bands, and thereby simplify
role of active elements in devices used in electro- the detection of their deformational splitting. Such
acoustics and acousto-electronics, e.g., to amplify, a detection is conducted by direct recording of
generate and modify the amplitude and phase the spectra in polarized light, or by way of vari-
of the acoustic waves (in particular, the surface ous sensitive techniques from the arsenal of mod-
acoustic waves, SAW). They are also used to ulation spectroscopy. The basic characteristics of
build electromechanical filters and nonlinear ele- the piezospectroscopic effect that are measured are
ments for radio-electronic circuitry. The presence the number of components in the split band, their
of the inner photoeffect makes it possible to apply intensity, their states of polarization, and the fre-
piezosemiconductors to tasks of acousto-optics. quency shifts of their components.
Among various semiconductors this piezoelectric By comparing experimental characteristics of
effect is more strongly expressed in CdS, CdSe, piezospectroscopic splittings of electronic transi-
ZnO, ZnS (hexagonal symmetry crystals), and tions with the phenomenological theory based on
also in ferroelectric semiconductors and photofer- group theoretical calculations (see Group theory in
roelectrics such as SbSI. solid state physics) and with perturbation theory,
one may obtain definitive data on the fundamen-
PIEZOSPECTROSCOPY tal symmetry properties of both the impurity and
Branch of the optical spectroscopy of solids the intrinsic crystal electronic states in the crystals
involving the study of optical spectra during the under study. Among them are the point local sym-
directional (anisotropic) elastic strain of crys- metry of impurity centers and defects, the sym-
tals. The usual objects of such studies are de- metry of their electronic states, their position in
formations of electron transitions (as manifested k -space, and the symmetry of the extrema of exci-
in absorption, reflection and excitation spectra, ton and electron bands. Also one may obtain data
photoconductivity or luminescence). In a general on the multipole character of optical transitions,
case the directional elastic deformation lowers the and on the phenomenological numerical parame-
crystal lattice symmetry, splits degenerate elec- ters that characterize the effect of electric defor-
tron energy levels in crystals (both intrinsic and mations on electron states (in band structure cal-
impurity types), and shifts nondegenerate lev- culations strain potentials play the role of such
els. This entails the reversible splitting (displace- parameters). Such determinations were performed
ment) of the bands of the corresponding spec- for a large number of impurity centers and de-
tral transitions, this being the principal spectro- fects (color centers) in ionic crystals, of deep lev-
scopic effect of a directional deformation. Another els and shallow levels in semiconductors, excitons
piezospectroscopic effect consists in a reversible and band extrema in semiconductors belonging to
change of band intensity and polarization (includ- groups IV, III–V, II–VI, etc.
ing “excitation” of forbidden transitions) that re- During piezospectroscopic studies of electron–
sults from the mixing of electronic states during an phonon band transitions (indirect excitations, vi-
anisotropic deformation of the lattice. Piezospec- bronic structure of impurity atoms) one may also
troscopy studies are most informative in the case determine the symmetry of the vibrations taking
of cubic crystals, since due to their high symmetry part in those transitions. Separate observations are
such crystals feature the greatest amount of degen- also available of direct deformation splittings of
eracy in their energy states. bands in the vibrational spectra of crystals (such
A static uniform anisotropic elastic deforma- as local vibrations in IR spectra; vibrations active
tion is usually produced by the uniaxial compres- in Raman scattering of light), and the resulting data
sion of a single crystal specimen along its [100], may be utilized to evaluate anharmonic interaction
[111], [110] axes, or by the planar extension (or constants.
p–i–n DIODE 1009
Studies of the spectra of crystals undergo- total number N of dislocations of either sign in a
ing anisotropic deformation make it possible to dislocation pile-up is
numerically predict the shift of bands during
L/2
isotropic deformation, i.e. those observed during (1 − ν)Lσ
experiments with bulk (hydrostatic) compression. N= n(x) dx = .
μb
The changes in the absorption spectra that re- 0
sult from anisotropic electric deformation yield The force acting upon a unit length of the head
changes in the values of the crystal refractive in- dislocation is F = πNbσ . For a single-sided pile-
dex, that is, changes in the photoelastic constants up of dislocations clamped to the obstacle by the
of the crystal. 0 the dislocation density n(x) is
shear stress σxy
L1 /2 + x 1/2
PILE-UP OF DISLOCATIONS 0
n(x) = 2(1 − ν)σxy (μb)−1
A group of dislocations which accumulate near L1 /2 − x
an obstacle when exposed to an applied stress. and
The obstacle for the piling-up of dislocations may
be a grain boundary, coherent or incoherent in- N1 = πN.
clusions, or barriers formed by extended disloca- The length of piled-of dislocations is
tions. In this case the main dislocation pile-up is
μbN1
fixed because of the interaction with the control- L1 = 0 (1 − ν)
.
ling field of the obstacle whose gradient σ of in- πσxy
ternal stresses is usually rather large, so that at an
The shear stress σxy near the piling-up head is
increase of σ the dislocation displacement is infin-
1/2
itesimal, i.e. it may be considered fixed. At every 0 L1
σxy ≈ σxy .
piled-up dislocation there are present not only ex- x − L1 /2
ternal stresses, but also interaction forces with
The results of the model of continuous distribu-
other dislocations of the pile-up. As a result, there
tion of dislocations in pile-up are not valid for dis-
is a stresse concentration, and the main disloca-
tances smaller than those separating main disloca-
tion is acted upon by a force proportional to Nσ ,
tions. Pile-up analysis is important for examining
where N is the number of dislocations in the pile-
the origin and spreading process of plastic defor-
up. If N is large enough then at the pile-up dis-
mations, brittle cracks and tough cracks, and also
location head there may appear stresses equal to
for strain hardening theory.
the theoretical strength that can initiate plastic flow
(see Plastic deformations) or induce cracks near
PINCH-EFFECT
the obstacle. The process of dislocation pile-up is
See Magnetic pressure.
well approximated by the continuous dislocation
distribution approximation. The dislocation den- p –i –n DIODE
sity n(x) is determined by the condition of force A semiconductor diode that consists of two
equilibrium resulting from external stresses and low-resistivity regions having n- and p-types of
interaction forces in the dislocations of the pile- conductivity, and a base region (intrinsic, or
up. In the particular case of a two-sided pile-up, i-region) of higher resistivity, which exhibits in-
when both sides are symmetrically aligned, there trinsic conductivity. The base region is so lightly
is a pile-up of dislocations with opposite signs, doped that on application of even a small re-
which leads to verse bias it becomes overlapped by the space
2(1 − ν)σ x charge region of the semiconductor junction. The
n(x) = , devices referred to as p–i–n-diodes are com-
μb[(L/2)2 − x 2 ]1/2
monly p–ν–n- or p–π –n-diodes, where the let-
where b is the magnitude of Burgers vector, μ is ters ν and π designate a lightly doped material
the shear modulus, and ν is the Poisson ratio. The having, respectively, n- or p-type conductivity.
1010 PINNING
A material with intrinsic conductivity may be ob- also occurs in solid solutions, which have passi-
tained by precisely compensating with oppositely vating films that exhibit lowered protective abil-
charged dopants, but this complicated technologi- ity, and are prone to adsorption of corrosion-
cal process is rarely used. activating anions. Corrosion protection involves
Applications of p–i–n-diodes are based on electrochemical methods (cathodic, anodic, and
employing the following properties of the non- galvanic protection), introduction of corrosion in-
equilibrium electron–hole plasma in the p–i–n hibitors, introduction of corrosion-resisting addi-
structures: abrupt change of resistance under tives to alloys, e.g., alloying of stainless steels with
change of sign of the applied voltage (power recti- chromium, silicon, molybdenum. The highest re-
fiers, detectors, charge-storage diodes, UHF power sistance to corrosion in chlorine-containing media
switches); absorption of electromagnetic waves by is shown by titanium.
electron–hole plasma (absorptive attenuators for
controlling the UHF power, optical modulators of PLANAR DEFECT
IR radiation); injection luminescence of electron– An inner surface of a crystal lattice which sepa-
hole plasma (light emitting diodes); change of the rates crystal regions of different orientation or dif-
lifetime of charge carriers under action of temper- ferent phase state. Examples of planar defects are
ature, pressure, irradiation (temperature and pres- grain boundaries in a polycrystal, in blocks, cells
sure transducers, nuclear and X-ray radiation de- and fragments, dislocation walls, twin boundaries
tectors); change of mobility of carriers under the (see Twinning of crystals) and phase interfaces.
action of a magnetic field (magnetic diodes, mag- The separation of planar defects into an indepen-
netic field transducers). dent class of defects is due to the fact that many
properties of planar defects are independent of the
boundary structure, and are determined solely by
PINNING
the uniform distortion, which transforms the mate-
See Vortex pinning.
rial on one side of the boundary into the state of
the material on the another side. The role of such
PINNING OF FERMI LEVEL
a distortion may be played by rotation, twinning,
See Fermi level pinning.
etc. A theory of planar defects as independent im-
perfections of a crystal lattice has been developed.
PIPE DIFFUSION
In specific cases, the equations of the theory go
See Tubular diffusion. over into the relations of the theory of dislocations,
disclinations and dispyrations. The continuum the-
PITTING ory of planar defects (see Continuum theory of de-
The destruction of a metal under attack by se- fects) allows one to effectively solve various prob-
vere point corrosion involving the formation of lems of solid state physics: calculating the plasto-
small, but sharp, localized cavities (see Corrosion elastic fields, analyzing the processes of twinning
of metals). Materials subject to pitting failure are and of phase transitions, determining the values of
metals and alloys like iron, carbon and stainless stresses in the neighborhood of precipitates, etc.
steels, aluminum alloys, nickel alloys, titanium al-
loys, alloys of zirconium and other alloys in cor- PLANAR STATE OF STRESS
rosive media containing oxidizing agents (oxygen, The state of an elastic body that does not ex-
nitrates, nitrites, chromates) and activators (chlo- perience the action of normal and shear stresses
rine, bromine, and iodine anions). Pitting corro- (see Shear) at elements of area perpendicular to
sion occurs most commonly in sea water, in HCl- a certain direction (z-axis): σxz = σyz = σzz = 0
containing water solutions of alcohols, in cool- (σik is the stress tensor), so stresses act only in
ing brines, in circulating water supply systems of the x, y plane. The state of planar stress is real-
the chemical industry, and other enterprises. Pit- ized, e.g., in the case of a thin plate, which is sub-
ting arises at grain boundaries, nonmetallic inclu- jected to surface forces and bulk forces that are
sions (especially sulfide ones), in regions charac- perpendicular to the z-axis and uniformly distrib-
terized by a lowered chromium concentration. It uted along it. In the general case, the z-dependence
PLANATOMIC NETWORKS 1011
of the tensor σαβ (α, β = 1, 2) is quadratic. If or bonds with other neighboring atoms. The net-
σ1 , σ2 are the principal values of σαβ , the peak work is said to be a combinatory regular network,
value of the tangential stress equals |σ1 − σ2 |/2. if for each pair of its nodal points there exists a
The important specific cases of the uniform state combinatory topologic transformation, which con-
of plane stress are: simple shear (σ1 = −σ2 ), uni- verts the nodes into one another and the network
form compression in the x, y plane (σ1 = σ2 = 0), into itself. All possible combinatory differing net-
uniaxial state of stress (σ1 = 0, σ2 = 0).
works were found by A.V. Shubnikov from con-
PLANATOMIC NETWORKS (A.V. Shubnikov, sideration of the Euler equation for a finite simply
1916) connected complex F − E + N = 1 where F, E,
Planar atomic networks in which every internal and N are respectively the numbers of faces, edges
atom (node) has an equal number k of linkages (bonds), and nodes (vertices). This relation may be
presented in the form of the following equation: This Planck formula reflects the nature of the
1 1 1 k electromagnetic radiation and, in particular, the
+ + ··· + = − 1, fact that the thermal radiation may be treated as an
α1 α2 αk 2
ideal gas of light quanta (photons). It was in this
which relates the number of bonds k, which meet Planck formula that the Planck constant, which re-
at each internal node (k 3) to the number of ver- lates the photon energy ε to the light frequency ν
tices αi of the polygons which meet at that node. (ε = hν), was first introduced. The Planck for-
The solution of this equation for k = 3, 4, 5, 6 mula is sometimes written in terms of the density
(k cannot exceed 6) leads to 11 different combina- of heat radiation energy per unit wavelength inter-
tory planatomic networks (see Fig.). For example, val.
for network No. 1 we have k = 6, F = 24, E = 42,
N = 19, and all six αi values equal 3.
PLANE PROBLEM OF ELASTICITY THEORY
Any combinatory regular network may be
The plane problem of elasticity theory is con-
transformed without disruption of linkages into a
cerned with the stress-deformed state of a slab
metrically regular network. Therefore, any group
or plate, which is loaded symmetrically about its
associated with combinatory topologic transfor-
mid-plane (see Shells), or of a long (in theory, in-
mations of any combinatory regular network is
finite) cylinder (prism), which is loaded by forces
isomorphic with the related two-dimensional
normal to its lateral surface and invariant along its
space group (called Fedorov symmetry group in
length. Mathematically, the plane problem of elas-
the Russian literature); in other words, the basis
of two-dimensional crystallography is purely topo- ticity theory is reduced to a boundary value prob-
logical. No analogous statement has yet been ob- lem for the biharmonic equation (anisotropy adds
tained for the case of three-dimensional space. complexity to this equation) under corresponding
boundary conditions. The general solution of these
PLANCK CONSTANT, quantum of action, h equations is expressed in terms of two analyti-
(M. Planck, 1900) cal functions of complex variables. In the case of
One of the fundamental physical constants, dynamic loads (steady-state vibrations, abrupt ap-
which is characteristic of the discrete (quantum) plication of load), the plane problem reduces to
nature of objects of matter. The energy of a quan- solving what can be called metaharmonic equa-
tum particle (quasi-particle) ε is related to its fre- tions.
quency ν by the equation ε = hν, with the cor-
responding de Broglie wavelength λ given by the PLASMA-CHEMICAL TECHNOLOGY
equation λ = hc/ε, p = h/λ (c is the speed of The collection of methods for forming mate-
light). The term “reduced Planck constant” is of- rials by the use of chemical reactions, in which
ten used in reference to the quantity = h/(2π), the processes occurring in a low-temperature plas-
ε = ω (ω is the angular frequency, ω = 2πν). ma are employed. The specific features of these
The accepted value for h is 6.626076·10−34 J·s; processes are: high reaction rates, which favor use
hence, the value for h is 1.0545727·10−34 J·s. of miniature equipment; the possibility of employ-
ing easily available starting materials of little value
PLANCK FORMULA, Planck radiation formula, (e.g., air); the possibility of obtaining materials,
Planck distribution (M. Planck, 1900) which do not exist at ordinary temperatures (e.g.,
The spectral distribution of the density of elec- Al2 O, AlO, clusters consisting of 2–9 carbon
tromagnetic radiation energy in thermal equilib- atoms, etc.). Plasma-chemical technology involves
rium with a material is given by the expression: the following stages: production of plasma with
−1 plasma generators (electric arc, high frequency
8πhν 3 hν
dEν = exp −1 dν, and extra-high frequency (microwave) ones), in-
c3 kB T troduction of reagents into the plasma, halting the
where ν is the radiation frequency, and T is the reaction at the required stage by abruptly lowering
temperature. the temperature (quenching). The reaction time is
PLASMA REFLECTION 1013
where N and m∗ are respectively the concentra- face purification. The etching according to a pre-
tion and effective mass of free charge carriers; determined pattern (so-called dry plasma etching)
ε∞ is the dielectric constant due to the contribu- is one of the tools used in techniques for pro-
tion of band-to-band transitions. In metals, N ∼ ducing integrated circuits with a high degree of
1022 cm−3 , and in most metals the plasma fre- integration. The plasma precipitation is used for
quency lies far in the UV region, which is respon- forming thin dielectric layers for integrated cir-
sible for the high reflectivity of metals for visible cuits and film coatings, which offer a broad spec-
and infrared radiation. The concentration of free trum of useful features and applications. The latter
charge carriers in semiconductors is lower by at trend, which gained wide acceptance in the 1980s,
least several orders of magnitude, even under the permits multiply increasing the surface hardness,
conditions of heavy doping or intense optical exci- wear resistance, resistance to chemicals, etc., by
tation. Therefore, the plasma frequency and corre- application of thin films consisting of refractory
sponding increase of the re e ctance are observed intermetallics, and particularly of “diamond-like”
within the IR spectral region. For n-Ge featuring carbon (see Plasma polycondensation, Diamond-
N ≈ 1019 cm−3 , e.g., the region of intense plasma like carbon lm s). These films are applied onto
reflection begins at λ ≈ 10 μm, and extends to the semiconductor surface at a moderate temper-
longer wavelengths. ature (down to room temperature). The plasma
The abrupt change of the reflectance in the fre- treatment is carried out in vacuum and continu-
ous plasma reactors, as well as by application of
quency region ω ≈ ωp is related to the contribu-
“plasma torches”, which inject the plasma dis-
tion made by free electrons (or holes) to the crys-
charge jet into the atmosphere. Besides etching
tal dielectric constant ε, which determines the re-
and deposition of coatings, it is possible in prin-
fractive index n and the reflectance R. Neglecting
ciple to perform the diffusion (implantation) treat-
the attenuation of plasma oscillations and the ab-
ment of a solid surface in a plasma. The plasma
sorption of radiation by the crystal, we obtain at
discharge is also used for processing raw materi-
normal incidence als and recycling wastes. In this case, the principal
2
ωp2 role is played by the high-energy reactions, which
ε(ω) = n (ω) = ε∞ 1 − 2 , (2) take place in the plasma.
ω
n−1 2 PLASMA WAVES, MAGNETIC
R(ω) = . (3) See Magnetic plasma waves.
n+1
Hence it follows that as ω leaves the high- PLASMON
frequency region and approaches ωp , the value A quasi-particle, the quantum of plasma oscil-
of R decreases and reaches zero as n(ω) be- lations of characteristic energy εk = ωp (at small
comes 1, and then abruptly increases again and be- values of the quasi-momentum k), where ωp is
comes equal to unity at ω = ωp . The study of the the plasma frequency. In multicomponent (many-
plasma reflection spectrum is used for determining band) systems, or systems of low dimensionality
ωp and, hence, N or m∗ . (laminated, chain-type ones), the dispersion law
for plasmons may assume the quasi-acoustic form,
PLASMA, SOLID-STATE the plasmon energy εk tending to zero as k → 0
See Solid-state plasma. (see Acoustic plasmon). See also Plasmon in su-
perconductors.
PLASMA TREATMENT PLASMON, ACOUSTIC
The physicochemical treatment of a solid sur- See Acoustic plasmon.
face (see Ion-plasma treatment). Of primary im-
portance for solid state physics are two types of PLASMON MECHANISM OF
processes: etching and deposition of thin lms . The SUPERCONDUCTIVITY
isotropic etching in inert or chemically active me- A hypothetical phononless mechanism of su-
dia (e.g., in freons) is an effective method for sur- perconductivity, which is based on the assumption
PLASMONS IN SUPERCONDUCTORS 1015
of Cooper pairing (see Cooper pairs) of “light” etc. It has been suggested that a plasmon mecha-
degenerate current carriers (conduction electrons, nism might play a role in explaining the supercon-
holes) due to the exchange of virtual quanta of ductivity in cuprate compounds with a perovskite-
collective excitations (charge density oscillations) type structure. This hypothesis is supported by
in the system of “heavier” carriers, e.g., d- or the fact that the isotope effect in superconduct-
f -shell electrons of transition (rare-earth) met- ing metal-oxide compounds is generally unusually
als, electrons or holes of higher effective mass weak.
in many-valley (many-band) degenerate semicon-
ductors (semimetals), or in layered (heteroge- PLASMON–PHONON RESONANCE
neous) semiconducting structures. H. Fröhlich and The closeness of the natural frequency of the
E.A. Pashitsky proposed this mechanism (1968) plasma oscillations of long-wavelength free charge
independently of one another. The spectrum of carriers ωp = [e2 N/(ε0 m∗ )]1/2 (ε∞ is the high-
collective plasma oscillations of relatively low fre- frequency dielectric constant; m∗ is the effective
quency is an acoustic one at vFl
vFh and ωpl
mass of carriers; N is concentration of carriers)
ωph (where vFl and vFh are Fermi velocities, ωpl to the frequency of longitudinal optical phonons
and ωph are plasma frequencies of respectively ωLO . In the region of plasmon–phonon reso-
light and heavy fermions), ωq = q(vFl vFh /3)1/2 nance, an important role is played by the inter-
at q → 0 (see Acoustic plasmon), and slightly de- action of corresponding oscillations; this interac-
cays within the frequency range qvFh ω ωph tion occurs through longitudinal oscillating elec-
(q is the wave number). With regard for the effects tric fields, which are associated with both excita-
of retardation (dynamic over-screening) due to the tions. The contributions of plasmons and phonons
exchange of virtual plasmons, the matrix element to the crystal dielectric constant are additive:
associated with the screened Coulomb interaction 2 − ω2
between electrons may be represented as ωp2 ωLO TO ,
ε(ω) = ε∞ 1 − +
ω2 2 − ω2
ωTO
4πe2 (ωq + iγq )2
VC (q, ω) = 1 − , where ωTO is the frequency of transverse optical
q 2 + κl2 (ωq + iγq )2 − ω2
phonons.
where κl−1 is the radius of sceening by “light” The frequencies of the coupled oscillations,
carriers, and γq is the plasmon damping factor which result from the resonance, are determined
due to the decay into electron–hole pairs (see from the condition ε(ω) = 0, and have the form
Landau damping) and current carrier scattering 2 = 1 2 2
ω± ω + ωLO
by the crystal lattice defects. It follows from this 2 p
equation that the retarded interaction is attrac- 2 2 2 − 4ω2 ω2 1/2 .
± ωp + ωLO p TO
tive by nature in the region qvFh ω ωq =
qωph (q 2 + κl2 )−1/2 , Re VC (q, ω) < 0. The ef- The plasmon–phonon resonance modifies the IR
ficiency of the electron–plasmon interaction in- spectra of solids, changing the reflection in the
creases in polar (ionic) semiconducting crystals frequency region of the residual ray band (Re-
due to the hybridization of acoustic plasmons strahlen). The frequencies of the plasmon–phonon
with optical phonons, which results in a broad- modes and their dependence on the concentration
ening of the region of attraction up to the energy of free charge carriers clearly manifest themselves
given by (ωph 2 + ω2 )1/2 , where ω in the spectra of Raman scattering of light.
LO LO is the fre-
quency of the longitudinal optical phonon. The in-
terest in the plasmon mechanism of superconduc- PLASMONS IN SUPERCONDUCTORS
tivity has been stimulated by the discovery of the Quasi-particles which are associated with the
phenomenon of high-temperature superconduc- collective excitations of a system of conduction
tivity in metal-oxide compounds like La2−x (Ba, electrons, and which arise from electron–electron
Sr)x CuO4 ; YBa2 Cu3 O7−δ ; Bi2 Sr2 CaCu2 O8+δ , Coulomb repulsion. The energy of a plasmon is
1016 PLASMON, SURFACE
given by ε = ωp , where ωp is the plasma os- levels: translation at one level is accompanied by
cillation frequency. The value of the plasma fre- rotation at a higher level, and vice versa. Depend-
quency ωp in homogeneous superconductors co- ing on the type of elementary deformation carriers,
incides with its value in a normal metal, ωp = the dislocation model of plastic deformation and the
(ne2 /ε0 m)1/2 , where n is the concentration of disclination models of plastic deformation are dis-
electrons, and m is the electron mass. Excep- tinguished, as well as plastic deformation due to
tions are provided by superconductors with heavy the transport of point defects (vacancies, intersti-
electrons (heavy fermions), anisotropic supercon- tial atoms, crowdions).
ductivity in quasi-one-dimensional systems, and
quasi-one-dimensional crystals featuring a quasi- PLASTIC DEFORMATION, DISCLINATION
acoustic dispersion law (see Acoustic plasmon), as MODELS
well as by thin lms , in which the plasma oscilla- See Disclination models of plastic deformation.
tions (“mini-plasmons”) are of low frequency. In
Josephson tunnel junctions (see Josephson junc- PLASTIC DEFORMATION LOCALIZATION
tions), there exists the possibility of the propaga- See Localization of plastic deformation.
tion of low-frequency plasma oscillations, the so-
called Josephson plasmons, the excitation energy PLASTIC DEFORMATION OF AMORPHOUS
of which is ωJ = (2eIc /C)1/2 , where Ic is the METALLIC ALLOYS
critical Josephson current, and C is the junction The plastic deformation of amorphous metal
capacitance. Typical values of ωJ range from 109 alloys is appreciably different in nature in the tem-
to 1011 s−1 , whereas the plasma frequency of typ- perature intervals above and below the tempera-
ical metals is ωp ∼ 1016 s−1 . The low value of the ture equal to 0.7Tv , where Tv is the vitri ca tion
frequency of Josephson plasma oscillations is due temperature. At low temperatures plastic deforma-
to the low density of charge carriers in the barrier tion is very localized (inhomogeneous deforma-
(see Josephson effect). tion), and is concentrated in the shear bands, the
width of which is 10–40 nm, according to data
PLASMON, SURFACE provided by electron microscopy, the height of the
See Surface plasmon. steps at the surface being 100–200 nm. Hence it
follows that the severity of the deformation within
PLASTIC DEFORMATION such a band is very high. On the macroscopic
The irreversible change of the dimensions and scale, at the same time, the degree of strain prior
shape of a solid caused by exposure to external to failure is very small. Another specific feature of
actions. From the standpoint of the geometry of plastic deformation is either the complete absence
deformation, the following types of plastic defor- of, or the rather small magnitude of, strain hard-
mation should be distinguished: slip, twinning (see ening. One of the mechanisms of strain hardening
Twinning of crystals), rotation and viscous flow. is the segregation of the intrusion elements in re-
Twinning takes place in crystals; sliding, rota- gions of increased free volume concentration (see
tion, and viscous flow or diffusion creep are ob- Free volume model). At high enough temperatures
served in all solids. Each type of plastic deforma- (above 0.7Tv ), the nature of the deformation of
tion manifests itself in different forms depending amorphous metallic alloys (see Amorphous met-
on the temperature, the severity and rate of defor- als and metallic alloys) changes; the inhomoge-
mation, and each type has a different influence on neous deformation goes over into a homogeneous
the resulting structure. The difference of paths of deformation, no shear bands are generated, and the
plastic deformation in monocrystals and in poly- whole bulk of the material becomes viscously de-
crystals are sometimes more important than the formed. This takes place with the onset of a pro-
change of the deformation conditions. The plas- nounced degree of plastic flow. As the temperature
tic deformation of real materials proceeds by sev- further increases (Tv < T < Tcr , where Tcr is the
eral mechanisms simultaneously, not by a single crystallization temperature), superplasticity arises
one. It proceeds concurrently at several structural in amorphous metallic alloys.
PLASTIC DEFORMATION RATE 1017
By analogy with the mechanisms of deforma- structure on mechanical properties above T ∗ are
tion of crystalline alloys, several dislocation mod- determined by the same general rules as in metals.
els of plastic deformation have been proposed, Below T ∗ , no cellular structure is formed under
among them being the Gilman model, according deformation, and the failure is brittle by nature.
to which sliding is carried out by dislocations with The highest value of T ∗ (∼0.85Tm ) is exhibited
Burgers vectors of alternating magnitude and di- by purely covalent crystals (Si, Ge, diamond). In
rection. In subsequent studies, the dislocation lat- partially covalent crystals (III–V and II–VI com-
tice model and the disclination–dislocation model
pound semiconductors, carbides, borides, nitrides,
were advanced, according to which the structure
oxides, etc.), the value of T ∗ decreases with the
of amorphous metallic alloys may be represented
as a three-dimensional dislocation grid. Although decrease of directionality of the interatomic bond.
the mechanism of plastic deformation still remains Among the group of crystals under consideration,
to be fully explained, the importance of the role the lowest value of T ∗ ≈ 0.2Tm is observed in
of the free volume is well established. The failure transition metals with a body-centered cubic struc-
of amorphous metallic alloys under homogeneous ture.
deformation proceeds by the opening of a crack
along the localized slip band.
PLASTIC DEFORMATION RATE
PLASTIC DEFORMATION OF COVALENT Rate of deformation accompanied by irrevers-
CRYSTALS ible changes of a sample shape. In terms of the
The specific kind of plastic deformation, which dislocation mechanism the rate of plastic defor-
occurs in crystals featuring a very high value mation of crystalline materials dε/dt = ε̇ is de-
of Peierls–Nabarro stress, where the dislocations fined as the density of mobile dislocations ρ
are practically frozen at low temperatures (see times their speed v, that is ε̇ = bρv, where b is
Peierls–Nabarro model). As the temperature is in- the dislocation Burgers vector. For the diffusion
creased, the thermally activated motion of disloca- mechanism of crystal creep the rate of plastic de-
tions is observed. This motion consists in the gen-
formation in the elementary case is defined by
eration of a double inflection at the dislocation line
the gradient of the vacancy concentration c times
(see Kinks on dislocations), due to which the dis-
location becomes drawn into a neighboring equi- their diffusion coef cient D, that is ε̇ = Aa 3 D∇c,
librium position. The low mobility of dislocations where A = const, and a 3 is the atomic volume.
under the conditions of the above strain mecha- With the increase of ε̇ the resistance to plastic
nism is responsible for high values of the yield deformation σ , as a rule, also increases. A decrease
limit and ultimate strength, which decrease with in- of the rate of plastic deformation can accompany
creasing temperature, as well as for the occurrence the development of strain aging processes, as ob-
of a wide temperature interval near 0 K where the served, e.g., in metallic crystals supersaturated by
failure is brittle by nature (see Brittle failure). The vacancies. The quantity σ is not a single-valued
starting point of the abrupt increase of the yield function of ε̇ at a particular moment of deforma-
point (with decreasing temperature) is determined tion, but rather depends on earlier rates of defor-
by the characteristic deformation temperature on mation due to the influence of ε̇ on the disloca-
the homologous temperature scale t ∗ = T ∗ /Tm ≈
tion structure. The increase of ε̇ promotes an in-
0.22[U/(kB Tm )]1/2 , where U is the activation en-
crease of the homogeneity of plastic deformation
ergy of dislocation motion, and Tm is the melting
temperature. Above T ∗ , the mechanism of plas-
(see Uniform strain), and the formation of a more
tic deformation of covalent crystals is similar to fine grained but less flexible structure. There is an
the mechanism of deformation of plastic metals. increase in the number of active slip planes and
In this case the deformation involves the forma- slip systems, fine shear lines, and deformation slip
tion of a cellular dislocation structure, and the oc- bands, and a decrease in the cell sizes of the dislo-
currence of recovery and recrystallization is pos- cation structure. At very high rates of deformation
sible. The strain hardening and the effect of the inertial effects play a pronounced role.
1018 PLASTIC DEFORMATION TEMPERATURE CONDITIONS
on dislocations; the theoretical assessments also of the martensite of quenched steel upon vigor-
predict that this force is reduced by several times ous heating to the tempering temperature. During
at the normal-to-superconducting transition of a the process of steel tempering, the yield limit turns
metal. The marked role of conduction electrons out to be strongly decreased, whereas the value
in the macroscopic plasticity, which is determined of plasticity is considerably increased. The trans-
by the conditions of the dislocation motion in the formation plasticity is characteristic of all kinds
crystal lattice, was first shown using the effect un- of phase transformations: both martensitic and
der consideration as an example. The presence of diffusion-types, forward and reverse ones. In the
this increase of material plasticity is of great im- case of martensitic transformations, however, the
portance in various applications associated with transformation plasticity most often occurs only at
the phenomena of fatigue, friction and wear, as well the cooling stage, e.g., at the stage of forward tran-
as in the active application of superconducting ma- sition, whereas only the shape memory effect oc-
terials in cryogenics. curs during the heating half-cycle, i.e. during the
course of a reverse transformation (TiNi, CuMn,
PLASTICITY INDUCED BY PHASE CuAlNi alloys, etc.). The latter effect prevails even
TRANSITIONS, transformation plasticity under conditions of extremely strong counterac-
The occurrence of inelastic (irreversible) me- tion. In the FeMn and DyCo alloys the transfor-
chanical deformation induced by phase transi- mation γ ↔ ε features plasticity during both heat-
tions. Transformation plasticity manifests itself ing and cooling. The phenomenon of the trans-
under various conditions: in the first place, it oc- formation plasticity is attributed to the action of
curs during the course of plastic flow of metals, several factors. The main such factor is the dis-
the phase transformation of which may be caused tortion of the transformation: in accordance with
by mechanical stresses. In this case, the reaction the Clausius–Clapeyron principle, it brings about
of the transformation produces additional strain in a greater deformation along the force and a lesser
the direction of the applied force. This kind of one in the opposite direction. In other cases, trans-
transformation plasticity is commonly referred to formation plasticity is initiated by the oriented dif-
as transformation-induced plasticity. In the sec- fusion or drift of other defects in the direction
ond place, transformation plasticity occurs during which favors the propagation of the deformation
the heating or cooling of stressed crystals, the tem- along the applied force direction, by microstresses
perature being changed over the temperature range which arise due to incomplete matching of crystal
of the phase transition. In this case, the transfor- habits to each other, or by the difference between
mation plasticity is detected by an abrupt increase the specific volumes of the parent phase and the
of inelastic deformation within the temperature reaction product, etc. This plasticity phenomenon
range of the phase transition, and sometimes also is widely used in physical experiments on creating
beyond this range, but it is always caused by the special types of steels, and in metalworking tech-
phase transition. The resulting deformation is usu- nology.
ally proportional to the stresses applied, which is
ruled out in the physics of plastic deformations. PLASTICITY MARGIN
This is due to the fact that transformation plastic- See Margin of plasticity.
ity is determined by forces of a chemical nature,
the role of stress reduces to that of a minor dis- PLASTIC LIMIT
turbing parameter, and the expansion in a power The highest value of plastic deformation pre-
series only involves terms of odd power. Thirdly, ceding failure. The plastic limit is small for mate-
the transformation plasticity may be detected un- rials with covalent bonds (see Plastic deformation
der conditions of deformation of a material, when of covalent crystals). The plastic limit is higher for
a change of a certain factor, e.g., an abrupt change materials with a face-centered cubic lattice than
of temperature, causes the immediate completion it is for those with a body-centered cubic lattice.
of the transition from one phase into another. Metals with a hexagonal close-packed lattice have
The typical illustration of this type of transfor- intermediate values. For a given material plastic-
mation plasticity is the immediate decomposition ity the limit depends on the temperature and on the
1020 PLASTICS
strain conditions. Thus, e.g., it grows with a transi- of specific collective forms of motion in the dis-
tion from tensile deformation to compression, and location ensemble. These collective effects in-
especially to hydroextrusion. See also Plasticity. volve the generation and movement within the
material of partial disclinations. Thus, twisting
PLASTICS modes of plasticity are often called disclination
See Polymeric materials. modes. A partial disclination moving in an origi-
nally monocrystalline region leaves behind bound-
PLASTICS, CONSTRUCTION aries, or points where several boundaries come to-
See Construction plastics. gether (joint disclinations). Defects of the discli-
nation type are concentrated in the regions of
breaks, or according to their nature, are sources of
PLASTICS, GLASS-REINFORCED
strong long-range internal stresses (internal elas-
See Glass-reinforced plastics.
tic fields). Thus to bring about the reduction of
latent (stored) energy, disclinations join together
PLASTIC TWISTING DEFORMATION into groups of defects of opposite sign, e.g., they
Process of plastic deformation associated with adapt dipole and perhaps quadruple configura-
irreversible change of sample shape or of volumes tions. The motion within the dipole material of
present in it due to the twisting or turning of partic- partial wedge disclinations leads to a variation
ular regions of crystallographic directions. The ex- of the shape of the sample, and to the formation
perimental effect of the plastic twisting or bending of ribbon-like structures and fault bands. Discli-
distortion is detected by an electron-microscopic nation internal stresses play an important role in
investigation (see Electron microscopy) of the de- strain hardening at later stages of deformation, and
fect structure of deformed materials. Within the in the tough failure of materials.
defect (dislocation) structure it is possible to iden-
tify those formations which cause disorientations PLATES, laminae
in adjacent regions of the material. The disorien- Solid elastically deformed bodies having the
tation angle may reach tens of degrees. The so- shape of a right-angle prism or a right-angle cylin-
called cellular structures are observed, with dis- der, with a thickness h which is much smaller
orientations of adjacent cells, fragmented regions then the transverse dimensions. A plate is said
(see Fragmentation), ribbon-like structures, and to be thin, if h does not exceed one fifth of the
their various modifications. Information on rota- least dimension of the base. Plates are divided into
tional modes of plasticity may be also obtained by classes according to their state of stress. A plate
X-ray methods, and with the help of optical mi- is called rigid if the stresses prevailing in it un-
croscopy (through analysis of the surface relief). der transverse loading are flexural stresses, and the
Plasticity distortions are characteristic of large tension–compression stresses (see Flexure) in the
strains or the presence of a nonuniform state of median surface (membrane stresses, see Shells)
stress within the sample. Typical manifestations may be neglected. By convention, the deflection
of twisting plasticity include faulting or the appear- of plates classified as rigid does not exceed h/5.
ance and development of interlayers with a reori- If membrane stresses are comparable in magni-
ented crystal lattice. The thicknesses of interlayers tude to tension stresses then the plate is said to be
vary over a wide range: from fractions of microm- elastic. An absolutely elastic plate (a membrane)
eters for mesofaults in metals to meters for faults in is one for which the tensile stresses may be ne-
rocks. A typical size of fragments in a fragmented glected in comparison with the membrane ones;
structure is 0.1 to 0.5 μm. These plasticity defor- plates featuring a deflection 5h are convention-
mations are present in practically all plastic crys- ally classed as membranes. The contour of the de-
talline materials, and also in polymers and com- flection depends on the distribution of load, and
posite materials. on the boundary conditions (free, built-in, sim-
Microscopic mechanisms of twisting strain are ply supported, or elastically supported edges). The
associated with the appearance and development value of the maximal deflection of a stiff plate
PLEOCHROISM 1021
under action of normal loads F satisfies (in or- resistance temperature coefficient 0.003927 K−1
der of magnitude) the law: f ∝ F l 2 /D, where (at 273–373 K). The electronic work function is
l is the characteristic transverse dimension of the 5.23 eV.
plate, D = Eh3 /12(1 − ν 2 ) is the so-called cylin- Platinum is paramagnetic; magnetic suscepti-
drical rigidity of the plate, E is Young’s modulus, bility is 0.971·10−6 CGS units at room temper-
and ν is the Poisson ratio. In the case of elastic ature. Saturation vapor pressure is 133.3 μPa (at
plates and membranes, the dimensionless deflec- 1773 K) and 13,333 μPa (at 2023 K). The elas-
tion ζ = f/ h is described by an equation of the tic modulus of annealed platinum is 147 GPa,
form c1 ζ 3 + c2 ζ = σ ∗ , where c1 and c2 are con- rigidity is 72.4 GPa (at 273 K), ultimate tension
stants, and σ ∗ = (σ/E)(l/ h)4 , σ ∝ F / l 2 is the 137 MPa. Mohs hardness 4.3; Vickers hardness
normal pressure; hence in the case of a strong de- 37–48 HV (according to different authors). Pure
flection (f
h) we have f ∝ F 1/3 l 2/3 . platinum is one of the most malleable and duc-
If the plate is loaded with compressive forces tile metals; readily lends itself to forging, may
lying in its plane, then at a certain critical value of be rolled into foil (to 0.0025 mm in thickness)
loading the plate loses stability and buckles (buck- or drawn into wire (0.001 mm in diameter); is
ling). In order of magnitude, σk ∝ D/ hl 2 . See amenable to polishing and welding. Incorporation
also Vibrations of rods and plates, Plane problem of impurities, even in small amounts, causes a de-
of elasticity theory. crease of plasticity and an increase of hardness.
Application of cold strain hardens platinum no-
PLATE VIBRATIONS ticeably: relative elongation of cold hardened plat-
See Vibrations of rods and plates. inum decreases to 1–2%, whereas the value of
HV increases to 90–95. Subsequent annealing re-
PLATING stores the softness and plasticity of the metal. All
See Chrome plating, Electroplating, Lead plat- alloy additives, when added within the solid so-
ing, Zinc plating. lution region, harden platinum: the hardness in-
creases by the factor of 2–2.5. Platinum is one of
PLATINUM, Pt the most corrosion-resistant materials (see Corro-
Chemical element of Group VIII of the pe- sion resistance). Platinum is widely used in vari-
riodic table with atomic number 78 and atomic ous fields of engineering. The metal (and its alloys
weight 195.08. Natural platinum consists of iso- with rhodium and iridium) are used for produc-
topes 190 Pt (0.0127%), 192 Pt (0.78%), 194 Pt tion of chemical industry ware. Platinum is one of
(32.9%), 195 Pt (33.8%), 196 Pt (25.2%) and 198 Pt the most widely used catalysts. Platinum is used
(7.23%), of which 190 Pt and 192 Pt are radioactive: for the production of thermocouples, resistance
half-lives 6.9·1011 and 1015 years, respectively. thermometers, electric contacts, wires for winding
Electronic configuration is 4f 14 5d 10 6s 0 . Succes- high-temperature furnaces, medical instruments.
sive ionization energies are (eV): 8.96, 18.54,
28.5. Atomic radius 0.138 nm, ionic radii: Pt 2+ PLEOCHROISM (fr. Gr. πλεoν , full and χρωμα ,
0.080, Pt4+ 0.065 nm. Oxidation states +2, +4, color)
more rarely +1, +3, +6. Electronegativity ≈2.2. Change of color of a material exposed to trans-
Free platinum is a silvery-white metal. Ex- mitted light, which is related to the direction of
hibits face-centered cubic lattice; lattice constant propagation and polarization of this light. The
a = 0.39233 nm (at 297 K). Density 21.45 g/cm3 phenomenon of pleochroism is one of the man-
(at 293 K); Tmelting = 2042 K, Tboiling = 4073 K. ifestations of optical anisotropy. Pleochroism is
Melting latent heat is 21.84 kJ/mole; heat brought about by the anisotropy of absorption,
of evaporation 470.4 kJ/mole; specific heat which in its turn depends on the polarization of the
0.1327 kJ·kg−1 ·K−1 (at 298 K); coefficient of radiation (see Polarization of light), the wavelength,
thermal conductivity is 73.047 W·m−1 ·K−1 in and the direction of wave propagation. The phe-
temperature range 273–373 K; coefficient of lin- nomenon of pleochroism is most often observed in
ear thermal expansion is 9.1·10−6 K−1 (at 276– crystals. In the case of uniaxial crystals, two “prin-
373 K). Electric resistivity is 103 n·m, electric cipal” (main) color variations are distinguished,
1022 PLUTONIUM
source materials for synthesis of transuranium ele- a welding point contact is formed using a short-
ments (Am, etc.). Pu and its compounds are highly time current pulse which leads to the formation of
toxic. a microscopic bridge that shunts the tunnel con-
nection. The point contact resistance with a direct
p –n JUNCTION connection is determined by the Sharvin formula
See Semiconductor junction. (see Sharvin resistance) if the contact size d is
much smaller than the electron mean free path l.
POCKELS EFFECT (F. Pockels, 1894)
In the opposite extreme (d
l), the contact re-
A linear electrooptical effect, which consists in
sistance is of the order of magnitude of the ratio
a change of refractive indices under the action of
ρ/d (ρ is the metal resistivity). For typical met-
an electric field; a change proportional to the field
als such as Na, Cu, Al, etc., a contact spot di-
intensity. The effect was first observed in 1883 in
ameter of 30–3 nm corresponds to a resistance of
quartz crystals. The effect is possible only in ma-
terials without a center of symmetry, and occurs 1–100 . The point contact current density can
in piezoelectric materials. A piezocrystal, which is reach 109 –1010 A/cm2 , which involves a notice-
initially isotropic, becomes a uniaxial crystal under able deviation of the current–voltage characteris-
the action of the electric field; birefringence, if it is tic from Ohm’s law. Point contacts are employed
present in the material, becomes changed. The na- in microcontact spectroscopy (see Tunnel spec-
ture of the Pockels effect is related to the effect of troscopy) as local injectors and collectors of non-
an electric field on the structure of a crystal lattice equilibrium electrons in semiconductors at low
and the electric polarization. temperatures in experiments on the electron focus-
Extensive studies of the Pockels effect began ing in metals by a magnetic field.
when this phenomenon was observed in crystals Superconducting point contacts are widely
of KH2 PO4 , and it was found that it may be used used, i.e. the electric contacts formed by touch-
for virtually infinitely fast control over laser ra- ing of a pair of bulk superconductors at a small
diation. It is used for high-frequency amplitude area of interface. The properties of these contacts
and phase modulation, deflection and scanning of depend on the ratio between the transverse di-
light, optical systems for transmission and record- mension d of the contact, and microscopic length
ing of information, and light locators in control- parameters of the materials such as the coher-
lable narrow cut filters. The reverse Pockels effect ence length ξ and the electron mean free path l.
consists in optical detection, i.e. in generation of In the case d < ξ , the superconducting point
static polarization of noncentrosymmetric crystals contacts exhibit the Josephson effect (see also
under the action of laser radiation. Josephson junction). Owing to the Josephson ef-
fect and the energy gap Δ in the spectrum of
POINT CONTACTS
quasi-particles, the current–voltage characteristic
Small-area mechanical, electrical, and heat
of these point contacts is essentially nonlinear. For
contacts between bulk conductors which are
superconducting point contacts with direct con-
formed in the vicinity of where they touch. Point
contacts are produced by pressing together elec- duction there is a substantially larger current at
trodes of a special kind: needles and planes, sharp V
Δ/e whose magnitude is of the order of
edges, etc. As a rule, the area of a mechanical con- Δ/(eR0 ) (here R0 is the point contact resistance
tact is much greater than the spot through which in the normal state), and it depends weakly on
the electric current passes, due to an oxide layer the bias voltage. These superconducting contacts
with insulator properties that covers the electrode are incorporated into superconducting quantum in-
surface. The electric current passes through the terference devices (SQUIDs) which are widely
contact either due to electron tunneling (see Tun- used for highly precise, supersensitive measure-
nel effect) in those places where the insulator layer ments of magnetic fields. Owing to their small ca-
thickness does not exceed a few nanometers, or pacity, these superconducting point contacts are
directly through regions with a direct contact in- used for the detection and transformation of high-
terface between the two metals. In experiments frequency electromagnetic radiation.
1024 POINT-CONTACT SPECTROSCOPY