Sei sulla pagina 1di 20

Soviet Physics Uspekhi

REVIEWS OF TOPICAL PROBLEMS Related content


- ABSORPTION OF LIGHT IN
MAGNETIC FIRST ORDER PHASE ANTIFERROMAGNETIC DIELECTRICS
V V Eremenko and A I Belyaeva
TRANSITIONS - QUANTUM THEORY OF
THERMOMAGNETIC PHENOMENA
INMETALS AND SEMICONDUCTORS
To cite this article: N P Grazhdankina 1969 Sov. Phys. Usp. 11 727 P S Zyryanov and G I Guseva

- ELECTRON THEORY OF TRANSITION


METALS. I
S V Vonsovski and Yurii A Izyumov
View the article online for updates and enhancements.

Recent citations
- Strong magnetoelastic effect in
CeCo1xFexSi as Néel order is suppressed
V. F. Correa et al

- Reliable thermodynamic estimators for


screening caloric materials
Nikolai A. Zarkevich and Duane D.
Johnson

- Application of the exchange-striction


model for the calculation of the FeRh
alloys magnetic properties
Eduard Valiev et al

This content was downloaded from IP address 115.145.140.43 on 10/02/2020 at 02:12


S OVIET PHYSICS USPEKHI VOLUME 11, NUMBER 5 MARCH-APRIL, 1969

536.76 + 538.2
MAGNETIC FIRST ORDER PHASE TRANSITIONS

N. P. GRAZHDANKINA

Institute of Metal Physics, USSR Academy of Sciences, Sverdlovsk


Usp. Fiz. Nauk 96, 291-325 (October, 1968)

1. INTRODUCTION spin ordering. In addition, we consider data on the influ-


ence of high pressure and strong magnetic fields on the
JM.AGNETIC transformations of ferromagnets at the temperature of the magnetic transformations. In the
Curie point or of antiferromagnets at the Neel point are conclusion we shall report on papers in which other
usually, from the thermodynamic point of view, second model explanations, not connected with the exchange-
order phase transitions. In the last 5—6 years, however, striction mechanism, are proposed for the explanation
much experimental material has been accumulated, indi- of magnetic first-order phase transitions.
cating that there exist also magnetic systems in which It must be noted that from the formal point of view
these transitions are of first order. Such systems one can classify as magnetic first-order transitions of
include many compounds and alloys, and also a large the order-order type also the transformations of anti-
group of rare-earth metals. In spite of the great variety ferromagnets at the Morin point, i.e., transitions due to
of magnetic systems in which these transitions have the reorientation of the spins relative to the crystal
been observed, and also in spite of the different charac- axes (e.g., the low-temperature transition in a-Fe2O3).
ter of the transformations themselves, a common fea- As shown by Dzyaloshinskii [l23: , the existence of tran-
ture of all the first order magnetic phase transitions is sitions of this type can be explained within the frame-
that the principal role in the transformation is played work of the Landau theory (see C5J ). This class of mag-
by the change of the magnetic (exchange) interactions, netic transformations is not connected with the inversion
whereas the changes of the volume and of the symmetry of the exchange and is not considered in the present re-
of the crystal lattice are only concomitant phenomena. view.
All magnetic first-order phase transitions can be
conditionally subdivided into two groups: order-disorder
transitions and order-order transitions. The order- 2. MAGNETIC FIRST-ORDER PHASE TRANSITIONS OF
disorder transitions are called transitions with inver- THE ORDER-DISORDER TYPE
sion of the exchange interaction. This term was used by
Kittel in a theoretical analysis of magnetic transforma- 1) Classification of Magnetic Phase Transitions
tions connected with the change of the character of the
spin ordering and arising in ferromagnetism-antiferro- Historically, the first theoretical analysis of mag-
magnetism transitions, and also in transitions of sub- netic phase transitions of ferromagnets was the statis-
stance with complicated helical magnetic structures into tical-thermodynamic theory of Weiss. The temperature
the ferromagnetic or antiferromagnetic state. Transi- dependence of the magnetization calculated on the basis
tions of the order-disorder type are due to destruction of this theory actually describes, in general outline, the
of the spin order and are observed in ferromagnetism- experimentally observed character of this phenomenon,
paramagnetism transformations, and also in antiferro- i.e., the continuous transition from the ordered to the
magnetism-paramagnetism transformations. A theor- disordered state of the spin system. After Ehrenfest
etical analysis of this class of magnetic transitions was introduced the classification of phase transitions, mag-
first performed by Bean and Rodbell. netic transformations at the Curie point of ferromagnets
were classified as second-order phase transitions, in
According to the theory of Kittel u : l , and also of Bean accordance with the entire set of the then available ex-
and Rodbell. m , the main cause of magnetic first-order perimental data.
phase transitions is the strong change of the exchange
interactions as a function of the interatomic distances, As is well known, according to Ehrenfest, a phase
and also the associated change in the elastic energy of transition is characterized by the order of those deriva-
the material. This magnetic-transformation mechanism tives of the thermodynamic potential *, which experi-
is called the exchange-striction mechanism of magnetic ence finite changes during the transformation. First-
phase transitions. One of the necessary conditions for order phase transitions include those in which the first
realization of large volume changes of the exchange derivatives of the thermodynamic potential with respect
integrals is a sufficiently high compressibility of the to the temperature T and the pressure P change jump-
material. In this connection, great interest attaches to wise, and therefore jumpwise changes occur in such
investigations, at high hydrostatic pressures, of ferro- quantities as the entropy S and the volume V:
and antiferromagnets having the indicated properties to a®
a lesser or greater degree. In this review we shall con- ~dT
sider in greater detail these two quantitative treatments ~dP~
Second-order phase transitions are those in which the
of magnetic phase transitions ll'2:, in light of which we first derivatives of * are continuous, but its second
discuss the available experimental data concerning in- derivatives are discontinuous, namely the specific heat
vestigations of magnetic, electric, and elastic proper- Cp, the compressibility K, and the coefficient volume
ties of various magnetic systems with different types of thermal expansion a:

727
728 N. P . GRAZHDANKINA

Lp tions can be of first order was R i c e a i ] . In an investiga-


= ~' {-ar*~)p' K~ ~T\~dP*)T' a
^y\dP~dfj- tion of the general thermodynamic properties of a mag-
According to this definition, in the case of magnetic netic Ising lattice, Rice has shown that the magnetic
second-order phase transitions, besides the sharp transformation connected with the disordering of the
changes of Cp, K, and a, there should be discontinuities spin system can be a first-order transition if the ex-
of the magnetic susceptibility, the volume magnetostric- change interactions change with changing interatomic
tion, and of the magnetocaloric effect at the transforma- distance, and as a result of these changes the lattice be-
tion point. With further development of both theoretical comes deformed. These questions were considered in
and experimental investigations connected with problems greater detail by Bean and Rodbell 12 ' 123 , who determined
of ordering, remarks were made that the identification the concrete conditions for the occurrence of magnetic
of magnetic transformations as second-order phase first-order phase transitions and who predicted new
transitions calls for an additional and more detailed phenomena connected with the influence of high pressure
analysis. This concerned primarily the questions con- on the temperature of magnetic transformations.*
nected with character of the anomalies and with the
finite nature of the "jumps" of different physical quan- 2) The Bean and Rodbell Theory of Magnetic Phase
tities in second-order phase transitions. Theoretical Transitions
calculations of the spontaneous magnetization, based on
theoretical calculations in accordance with the Ising Li their theory, Bean and Rodbell [2] considered on
scheme113-1 (two-dimensional plane lattice of magnetic the basis of the molecular field model the interaction
moments), have shown that in this case the temperature between an assembly of magnetic dipoles, and the entire
dependence of the magnetization has a very sharp drop attention was focused on the change of the character of
at the Curie point, and the specific heat tends to infinity. the phase transition in the presence of a strong volume
Interest in the Ising model increased particularly in dependence of the exchange interaction. It was shown
connection with the work by Onsager, who pointed out the that in substances having large compressibility and a
possibility of a rigorous analytic analysis of the order- strong shift of the Curie temperature, changes of pres-
ing phenomena for a two-dimensional lattice. Experi- sure can realize conditions under which the magnetic
mental investigations of magnetic phase transition have transition connected with the destruction of the magnetic
shown that in many cases (principally in magnetic trans- structure, i.e., an order-disorder type of transition, is
formations of antiferromagnets) the temperature depen- a first-order phase transition. The physical picture of
dence of the specific heat is in the form of a X curve, this phenomenon is as follows. If it is assumed that the
thus indirectly confirming the infinite magnitude of the exchange interaction changes strongly when the inter-
jump which, as already noted above, was predicted atomic distance changes, then the decrease of the volume
theoretically for the particular case of a flat lattice. At at a given temperature leads to a change of the spontane-
the same time, a finite jump of the specific heat was ous magnetization a s of the ferromagnet as a result of
observed for most ferromagnets, and the magnitude of a change of the exchange forces. The change of crs will
this jump, as shown by Belov and co-workers (see c4] ), be accompanied by a change of the dimensions of the
could be calculated on the basis of the thermodynamic sample and by the occurrence of elastic stresses. In
theory of second-order phase transitions of Landau and turn, taking into account the strong volume dependence
Lifshitz15-1, the application of which to magnetic phenom- of the exchange integral, striction changes of the sample
ena was first considered by Vonsovskii C6: and Ginz- lead in final analysis to a new equilibrium value of the
burg [ 7 ] . Several attempts were made to find a suitable Curie temperature (T c ). If this change of the effective
theoretical scheme, within the framework of which one temperature is sufficiently large, then a situation can
could identify with assurance the various phase trans- arise wherein a spontaneous magnetization will be ob-
formations among the second-order transitions.* Thus, served above the true Curie point (To) of the "rigid"
for example, Ginzburg tl0] advanced arguments indicating system, i.e., a system in which the exchange interaction
that the singularities of different second-order transi- does not change with the volume, and there is no assoc-
tions can be attributed to different ratios of the volume iated spontaneous lattice deformation. Inasmuch as the
and correlation energies. Verification of the validity of existence of spontaneous magnetization at T c > T o is
these considerations calls for experimental investiga- due to the deformation of the lattice, and is in a certain
tions of exceedingly high precision. Taking into account sense unstable, the change of the elastic stresses with
the very limited accuracy of the available experimental increasing temperature can lead to a sharp decrease of
data, it is impossible at present to discuss exhaustively the spontaneous magnetization, i.e., to a first-order
the behavior of ferromagnets at the Curie point in light phase transition. Whether the magnetic transformation
of the aforementioned theory, so that the identification will have in each concrete case the features of a first-
of this phenomenon as a second-order phase transition order or a second-order phase transition depends on the
is to some degree arbitrary. ratio of the exchange to elastic energies, on the com-
pressibility K, and on the change of the exchange energy
The first to suggest that purely magnetic transforma- /3 upon compression.
Bean and Rodbell proposed that the dependence of the
*Semenchenko [8], on the basis of further development of the Gibbs Curie temperature on the volume is of the form
theory of thermodynamic stability, noted that second-order transitions
are identical with critical phenomena; the magnetic transformations at ^ - ] , (2.1)
the Curie point of ferromagnets were considered as transcritical transitions.
Phase transitions in ferroelectrics were considered from a similar point
of view by Tiza [']. "Later, similar ideas were considered also by Farrell and Meijer [13].
MAGNETIC FIRST ORDER PHASE TRANSITIONS 729

where Vo is the volume of the sample in the absence of w h i c h i s t h e r e s u l t o f s u b s t i t u t i n g ( 2 . 3 ) i n ( 2 . 2 ) a n d r e -

exchange interaction; the parameter /3 can be either p l a c i n g t h e e x p r e s s i o n f o r t h e e n t r o p y b y a s e r i e s i n

positive or negative. The free energy of the system Gy p o w e r s o f a . T h e s u b s e q u e n t g r o u p i n g o f t h e t e r m s w i t h

per unit volume contains five terms: the exchange en- e q u a l p o w e r s o f a y i e l d s E q . ( 2 . 5 ) . B e a n a n d R o d b e l l

ergy, the elastic energy, the entropy, and the energies h a v e s h o w n t h a t w h e n T / T o


< 1 a n d cr i s s m a l l , t h e f r e e

due to the external magnetic field and the pressure: e n e r g y i s d e t e r m i n e d m a i n l y b y t h e t e r m c o n t a i n i n g a 2


,

w h i c h l e a d s t o a d e c r e a s e o f t h e f r e e e n e r g y w i t h i n -

2
c r e a s i n g cr, s i n c e t h e c o e f f i c i e n t o f cr i s n e g a t i v e . W i t h

f u r t h e r i n c r e a s e o f a , t h e c o n t r i b u t i o n o f t h e t e r m s

4 8
c o n t a i n i n g a a n d a i n c r e a s e s , a n d t h i s l e a d s t o a n i n -

here a is the relative magnetization, i.e., the ratio of the c r e a s e o f t h e f r e e e n e r g y . T h e r e f o r e t h e p l o t s o f

specific magnetization at a given temperature to the G / N k T o a g a i n s t a h a v e m i n i m a a t d e f i n i t e v a l u e s o f a

magnetization at T = 0°K. Minimization of the free en- a n d T . A t t h e C u r i e p o i n t , t h e c o e f f i c i e n t o f a 2


v a n i s h e s ,

ergy with respect to the volume makes it possible to i . e . ,

determine the equilibrium value of the volume:


T c
^ r 0
( i — P K $ ) . ( 2 . 6 )

v — v B
(2.3) T h e c h a r a c t e r o f t h e p h a s e t r a n s i t i o n i s d e t e r m i n e d b y

which shows that the volume changes are consequences


4
t h e s i g n o f t h e c o e f f i c i e n t o f a . T h e m a g n e t i c t r a n s -

of the influence of the pressure and the magnetization. f o r m a t i o n w i l l b e a s e c o n d - o r d e r p h a s e t r a n s i t i o n i f t h i s

Substitution of (2.3) in (2.2) and minimization of the en- c o e f f i c i e n t i s p o s i t i v e , i . e . , i f

ergy with respect to the magnetization make it possible ( 2 . 7 )


-f->r\.
to determine (in implicit form) the temperature depen-
dence of a: O n t h e o t h e r h a n d , w h e n T / T o < TJ, t h e n a t t e m p e r a t u r e s
above T the negative contribution from the t e r m con-
arcthcr c

t a i n i n g cr 4 l e a d s t o a m i n i m u m o f t h e f r e e e n e r g y a t
a s* 0 , a n d t h i s , a s a l r e a d y n o t e d , c a u s e s a first-order
Relation (2.4) contains the parameter TJ, the absolute transition. T h e f o r e g o i n g i s i l l u s t r a t e d b y p l o t s of
values of which determine the character of the magnetic ( G o - G a ) / G o a g a i n s t a, c a l c u l a t e d f o r TJ = 1 . 2 a t P = 0
phase transition in each concrete case. r\ > 1 for first- a n d d i f f e r e n t v a l u e s of T / T o ( F i g . 2 ) .
order transitions and TJ < 1 for second-order transitions. A more general criterion determining the character
This can be verified by examining the cr = f (T/To) depen- of t h e p h a s e t r a n s i t i o n u n d e r c o n d i t i o n s of c o m p r e s s i o n
dence expressed by Eq. (2.4). Figure 1 shows the tem- of m a t t e r w a s d e r i v e d o n t h e b a s i s o f ( 2 . 6 ) a n d (2.7),
perature dependence of the magnetization for different namely
values of the parameter TJ. For the case r\ = 0 and P = 0,
we have the usual functional dependence of a on T/T o , Pxf, > 1 - (2.8)
determined by the Brillouin function. When r\ < 1, as The r e l a t i o n (2.8) i n d i c a t e s that an e x t e r n a l p r e s s u r e
seen from the curves of Fig. 1, the transition is of sec- c a n c h a n g e t h e c h a r a c t e r of t h e m a g n e t i c transformation.
W h e r e a s , for e x a m p l e , at a t m o s p h e r i c p r e s s u r e (for t h e
c a s e p > 0) t h e m a g n e t i c t r a n s f o r m a t i o n w a s a f i r s t -
o r d e r t r a n s i t i o n , at high p r e s s u r e s the t r a n s f o r m a t i o n
can become a second-order transition.
In c o n c l u d i n g t h i s s e c t i o n , l e t u s d w e l l o n t h e q u a l i t a -
t i v e a n a l o g y b e t w e e n t h e d e t e r m i n a t i o n o f t h e t y p e of t h e
m a g n e t i c p h a s e t r a n s i t i o n b y t h e t h e o r y of B e a n a n d

0 ff.t HZ Q3 B 0,5 Of ffj Of OS Iff V 12 V №


T/Ta
FIG. 1. Temperature dependence of the reduced spontaneous magneti-
zation a, calculated for different values of the parameters T) in accordance
with the theory of Bean and Rodbell [ 2 ] .

o n d o r d e r , w h e r e a s w h e n TJ > 1 t h e t r a n s i t i o n a t t h e
C u r i e p o i n t i s of f i r s t o r d e r . The dashed CT(T/T0) curve
for rj = 4 s h o w s t h e n o n - e q u i l i b r i u m s t a t e s of t h e m a g -
n e t i c o r d e r i n g , c o r r e s p o n d i n g t o t h e m a x i m u m of t h e
f r e e e n e r g y ( s t a t e s w i t h s u c h v a l u e s of a c a n n o t b e r e a l -
ized). T h e a n a l y s i s of t h e d e p e n d e n c e of t h e f r e e energy
on the magnetization i s b a s e d on the e x p r e s s i o n
-5 •
IT . „ 2HoBo
NkT0
F I G . 2 . P l o t s o f t h e free e n e r g y a g a i n s t a as c a l c u l a t e d b y R o d b e l l
2
a n d B e a n [ ] for t h e c a s e j = 1/2 a n d 7? = 1.2. T h e n u m b e r s at t h e c u r v e s

( 2 . 5 ) c o r r e s p o n d t o t h e ratio T/T o
.
730 N. P . GRAZHDANKINA

Rodbell and the general thermodynamic theory of phase starting from relation (2.4). The second criterion is the
transitions of Landau and Iifshitz 153 . This analogy was sign of the coefficient b, which is determined from mag-
emphasized by Banerjee [14] . As is well known, the netic measurements only, in accordance with Eq. (2.10).
theory of Landau and Iifshitz is based on general The available experimental data, which will be discussed
thermodynamic relations with allowance for the symme- later, offer evidence that the relation (2.4) is satisfac-
try properties of the crystal. It is assumed here that torily valid.
the thermodynamic potential *(T, P, | ) can be expanded
in a series about the transformation point, where the
ordering parameter | is small*. Vonsovskii [6: and 3) Investigation of Magnetic Transformations in Man-
Ginzburg [7] used this theory to study phase transitions ganese Arsenide and Comparison of the Experimental
at the Curie points of ferromagnets, using the magne- Data with the Theory
tization of the material (f = cr) as the ordering param- The theory developed by Bean and Rodbell t2] was used
eter. When account is taken of the energy of the external by them, and also by Rodbell and de Blois cl5] to explain
magnetic field, the expansion of the thermodynamic po- the magnetic transformations in manganese arsenide.
tential is of the form In spite of the many investigations of the magnetic trans-
formations in MnAs, the nature of these transformations
— <sH, (2.9) is not yet fully clear. The experimental investigations
where a and b are coefficients that depend on P and T. of Guillaud[16] and Serres t l 7 ] have shown that at tem-
At equilibrium, near the Curie point, 8*/3a = 0; on this peratures 313 and 399°K the magnetic properties of
basis, Eq. (2.9) can be rewritten in the form MnAs experience a change. At low temperatures, MnAs
has ferromagnetic properties; near 313°K a sharp drop
aa + ba^H. (2.10) of the magnetization takes place, accompanied by a
A thermodynamic analysis of the properties of the co- change of the volume and by release of the latent heat of
efficients a and b indicates that at the Curie point a = 0 the transformation. In the temperature interval from
and b should always be positive for second-order phase ©i = 313 to ©2 = 399°K, an increase of the susceptibility
transitions. Perfectly analogous conclusions can be is observed with increasing temperature, and only above
drawn also on the basis of the theory of Bean and Rod- 399°K does MnAs become paramagnetic and its suscep-
bell. Indeed, Eq. (2.5), obtained in the theory of Bean tibility obeys the Curie-Weiss law. These magnetic
and Rodbell, can be written in the form of (2.11), if the transformations are accompanied by changes of the
terms that do not depend on a are denoted by Ao: crystal structure. In the ferromagnetic (T < ®i) and
paramagnetic (T > ®2) states, MnAs has a lattice with
-g^-^-Ao + Atf + B^-Ho. (2.11) hexagonal symmetry D4jj. The magnetic transformation
©i causes an orthorhombic distortion of the crystal lat-
This equation corresponds to Eq. (2.9), and the condi- tice, which vanishes at the temperature ®2 :18] . In
tion A = 0, B > 0, which defines a second-order phase Guillaud's opinion, change of the magnetic properties of
transition in the theory of Bean and Rodbell, is perfectly MnAs can be explained by assuming that at ©i = 313°K
analogous to the condition for second-order phase tran- a transition takes place from the ferromagnetic into the
sitions in the theory of Landau and Iifshitz (a = 0, b >0). antiferromagnetic state, and at ©2 = 399° K the inverse
C14]
On this basis, Banerjee proposes that the theory of transition takes place. This assumption was confirmed
Landau and Iifshitz can be generalized also to include by Meyer and Taglang ci9] , who investigated the magneto-
the case of first-order phase transitions, the condition caloric effect of MnAs and observed a reversal of the
for which, in analogy with the theory of Bean and Rod- sign of the effect at 310 and 390° K. From thermody-
bell, should be taken to be a = 0 and b < 0. For an ex- namic considerations, it could be assumed that the ob-
perimental determination of the character of a magnetic served increase of the temperature under adiabatic de-
phase transition, Banerjee proposes to use the method magnetization of the sample is connected with antiferro-
t4]
of "thermodynamic coefficients," developed by Belov magnetic ordering of the spin magnetic moments of
and widely used at present for a precision determina- MnAs in the indicated temperature interval. However,
tion of the Curie point and of the spontaneous magnetiza- a neutron diffraction investigated by Bacon and Street [2C] ,
tion. In this case, according to Eq. (2.10), the quantities and also investigations of the antiferromagnetic reson-
a and b can be obtained from the isotherm H/CT = f(cr2), ance, did not confirm the presence of antiferromagnetic
the linear part of which is determined by the coefficient ordering in this compound. This contradiction in the ex-
b. The sign of this coefficient, as already noted, should perimental data has led to various interpretations of the
be negative for first-order phase transitions. nature of the magnetic transformations in MnAs. In
Thus, two independent criteria have been obtained Kittel's theory t l : , the transition at ©i = 313° is regarded
for determining the character of the magnetic phase as a magnetic transformation with inversion of the ex-
transition. One of them is the absolute magnitude of the change interaction, resulting from simultaneous realiza-
parameter r], which is calculated from the experimental tion of exchanges of different types, differing from each
data on the compressibility and on the temperature shift other in magnitude and having different dependences on
of the magnetic transformation with changing pressure, the interatomic distances. In the opinion of Bean and
RodbellC2], the transformation in MnAs at 313°K is a
*As is well known, at the present time the question of the singularities transition from the ferromagnetic to the paramagnetic
of the thermodynamic potential and of its derivatives at a second-order state, and the maximum of the susceptibility at 399°K
phase transition point cannot be regarded as solved, so that the problem is due to the Jahn- Teller effect and is connected with
of the validity of expanding the thermodynamic potential at the second-
order phase transition point should be regarded as open. the structural changes of the lattice. As already indica-
MAGNETIC FIRST ORDER PHASE TRANSITIONS 731

ted above, in this model one considers an exchange Table I


interaction that has one sign and varies isotropically
with the volume. E, dyne/cm2 G, dyne/cm2 V- x, cm2/dyne
Inasmuch as the main premise of the theoretical pa-
pers C1>2] was the strong dependence of the exchange- 3.22- ion 1.29-ion 0 27 4.5•10-12
interaction integrals on the interatomic distances and
the associated change of the elastic energy, the experi-
mental investigations of the elastic properties and of qualitative agreement with the results of an x-ray dif-
the effects of the temperature shift of magnetic trans- fraction investigation of the temperature variation of the
formations with changing pressure are of fundamental parameters of the MnAs crystal lattice, obtained earlier
importance for the verification of the validity of the by Willis and Rooksby t23] .
conclusions of these theories. Let us consider these ex- A comparison of the measured values of the elastic
perimental results in greater detail. moduli with the theoretically calculated ones' 1 ' 21 points
The elastic properties of MnAs were investigated by to a good agreement between the experiment and the
Grazhdankina and Burkhanov [21] . Their purpose was to theory of Bean and Rodbell' 21 . The compressibility
measure the elastic constants in a wide temperature calculated from formula (2.4) equals 2.2
interval that includes both the transition temperatures x 10" u cm2dyne"1, which differs only by a factor of 2
®i and ®2- The measurements were made by a resonance from the measured value K = 4.5 x 10"11 cm^yne" 1 , and
method using a composite vibrator and piezo-quartz ex- the additional allowance for the thermal expansion of
citation l221 at frequencies on the order of 60 kHz. Meas- the lattice' 151 in the theory of Bean and Rodbell leads to
urements of the Young's modulus E and of the shear a striking agreement between the theoretical value
modulus G on the same samples have made it possible K = 4.6 x 10~12 cmMyne"1 with the experimental data. An
to determine the Poisson coefficient and the compressi- estimate of E made by Kittel' 11 using formula (3.8)
bility K = 3 ( 1 - 2 fx)/E. In the indicated temperature in- gives a value of 4.2 x 1012 dyne/cm 2 for the Young's
terval, the thermal expansion and the magnetization of modulus, which is larger by one order of magnitude than
MnAs were also investigated. the measured value of E. The fact that there is no quan-
titative agreement with the predictions of Kittel's theory
Figure 3 shows the temperature dependences of the
is apparently not due only to the approximate character
magnetization a and of the reciprocal molar susceptibil-
of the calculation. Serious difficulties arise also in the
ity 1/X]yp measured 1213 at H = 10 kOe. In accordance
interpretation of the experimental data on the shift of
with data by others, a sharp drop of the magnetization
the transition temperature ®i with changing pressure.
takes place at T = 313° K, and at 392° K there is a peak
The sign of the effect d® i/dP does not correspond to the
of the function l/xjy[(T). Figure 4 shows plots of the
change of Aa (the crystal-lattice parameter of MnAs)
Young's modulus E, the shear modulus G, the Poisson
in the magnetic transformation. According to Kittel's
coefficient y., and the compressibility K as functions of
theory, the ferromagnetism-antiferromagnetism
the temperature. An anomalous growth of the moduli
(F —• AF) transition should be accompanied by an ex-
is observed only in the region of 400°K, i.e., in the
pansion of the lattice, determined by Eq. (3.4), but ac-
temperature region ®2 where a peak of the temperature
tually a strong contraction of the parameter a takes
dependence of the susceptibility is observed. However,
place.
in the temperature region ®i = 311°K, where a sharp
decrease of the magnetization takes place, no anomalous An investigation of the influence of hydrostatic com-
changes of the moduli were observed. The magnitudes pression on the temperature of the magnetic transforma-
of the moduli at room temperatures are given in Table I.
Figure 5 shows the temperature dependences of the
coefficient of thermal expansion a, according to which
the magnetic transformation a t ® 1 = 311°Kis accom-
panied by a jumplike change of the length of the sample,
and the transformation in the region ®2 ~ 400° K is ac-
companied by an abrupt change of a. These data are in

ISO
125
100
75 75
SO 50
25 - 25
0 JO i i 270Z30 310 330.750 370 330 410 430 450 470
70 120 170 ZZO Z7O 3Z0 370 4Z0 470
FIG. 4. Temperature dependence of Youngs modulus E, the shear
FIG. 3. Temperature dependences of the magnetization of MnAs (1) modulus G, the Poisson coefficient n, and the compressibility K of MnAs
and of the reciprocal molar susceptibility (2), measured at H = 10 kOe.
732 N. P . GRAZHDANKINA

seen from the plots that at atmospheric pressure ©x


= 314°K and decreases appreciably with increasing
pressure. Measurements made with increasing and de-
creasing P do not give the same results: a hysteresis is
observed, the width of which increases with increasing
pressure. In all cases the inverse transition, i.e., the
transformation observed when the sample is cooled, oc-
curs at a lower temperature and has a more diffuse
character. It is important that at low pressures (curves
1—3) hydrostatic compression of the sample leads only
to a shift of the temperature dependence of the induction
-Z5 - emf V(T) towards lower temperatures, without a change
in the magnitude of the emf and in the form of its tem-
FIG. 5. Temperature dependence of the coefficient of thermal perature dependence. At pressures above 2700 atm
expansion of MnAs [21]. (curves 5 and 6), there is not only a shift of ®u but also
a sharp decrease of V, this being connected with the
tion of MnAs has made it possible to verify further the change of the character of the magnetic transformation
theory of Bean and Rodbell. The shift of the magnetic- itself, due to the transformation of a non-magnetic modi-
transition temperature ®i with changing pressure was fication of MnAs. The occurrence under pressure of a
first determined by Rodbell and Wilson (see1241) with the new modification of MnAs at temperatures Tc is con-
aid of magnetic measurements performed at room tem- firmed also by investigations of electric resistance.
perature and in a relatively narrow pressure interval Figures 7 and 8 show R(P) plots obtained in the region
(up to 1000 atm). The value of d©i/dP was determined of the temperatures ®i and TQ. The different character
later on the basis of electric measurements up to of the variation of R(P) in the region of the indicated
5000 atm C2S] . The results obtained in these investiga- transitions indicates the presence of two different types
tions had to be refined, since the values of d®i/dP dif- of transformations. At low pressures (up to 2500 atm)
fered from each other. Further investigations at high
pressures were made by Grozhdankina and Bersenev'28-1,
who determined the temperature shift of the magnetic
transformation of MnAs under compression with the aid
of both magnetic and electric measurements in the tem- 273,5°K
perature range from 77 to 400° K and at pressures up to
11 000 atm. These measurements have shown that the
effect of the shift of ®i with changing pressure should be
determined only in the region of low pressures, up to
2500 atm. Higher pressures lead to a different phase ZSZ'K
transition, connected with the formation of a new modi-
fication of MnAs, which apparently has antiferromagnetic
ordering at low temperatures. On the basis of these r e -
sults, a P-T diagram of the magnetic transformations
in MnAs was constructed and agreed in general outline
with the theoretical predictions t l 2 ] .
Figure 6 shows the temperatures determined by
Grashdankina and Bersenev at various pressures with P, kg/cm5
the aid of magnetic induction measurements C26] . It is FIG. 7. Dependence of the electric resistance of MnAs on the pres-
sure. Measurements in the region of the magnetic-transition temperature

to
/?,ohm
35 \
30 II
1\ \
25 - \
h \
\\ \\
\\ \F
V \\
}
\ \
zn zoo zso 300 s/o szo sso I \
T,'K
FIG. 6. Temperature dependence of the induction emf of the meas-
uring coil (relative units)
2
at various pressures.2 1 — Atmospheric pressure,
2 P, kg/cm2
2 - P = 1400 kg/cm , 3 - P = 2350 kg/cm , 4 - P = 2780 kg/cm , 5 - FIG. 8. Dependence of the electric resistance of MnAs on the pres-
P = 2970 kg/cm2. The solid curves were obtained 26
while heating the sure. Measurements in the region of the transition temperature Tc [26].
sample, and the dashed ones while cooling [ ].
MAGNETIC F I R S T ORDER PHASE TRANSITIONS 733

the main effect of hydrostatic compression is the shift


of the temperature of the magnetic transformation
d®i/dP =-(16.0 ± 0.3) x 10"3 deg/atm (Fig. 9, curve 1).
It can be assumed that in this region a strong decrease
of the exchange interaction is not accompanied by a
noticeable change of the spontaneous magnetization of
MnAs. The transformation in the interval
2500—2700 atm is connected with the formation, under
pressure, of a nonmagnetic modification of MnAs, and
the temperature of this transformation decreases very
strongly with increasing P, namely dT^/dP = -(34 ± 7) Z30
x 10~3 deg/atm (Fig. 9, curve 2). On the basis of mag- Bom woaa
2
netic measurements made at low temperatures and high FIG. 9. Phase diagram of MnAs [ ]. 1 - kg/cm
26 f, Pressure dependence of the
pressures (from 4700 to 11 000 atm), a magnetic trans- temperature of the magnetic transformation 6,, 2 - changes of the trans-
formation was observed of the high-pressure phase of ition temperature Tc, 3 — change of the Neel point of the high-pressure
MnAs, due apparently to the occurrence of antiferro- phase; the dashed curve was obtained while cooling the sample, O -
magnetic ordering; the change of the Neel temperature obtained from magnetic measurements, • - from electric measurements.
under pressure is in this case equal to dTjj/dP
= (2.22 ± 0.7) x 10"3 deg/atm (Fig. 9, curve 3).
'/To
In Rodbell's opinion1-273, the above-described P-T
diagram of magnetic phase transitions in MnAs (Fig. 9)
is in qualitative agreement with the theoretical predic-
tion c l 2 ] . In1123 Bean and Rodbell considered, on the basis
of their model, the conditions for the equilibrium be- 0,*
tween the different magnetic states, for the four possi-
ble combinations of the signs of the parameters 0 and O Oft OS 12 W Pkp
T/Ta
To. It was shown, as a particular case, that for ferro- FIG. 10. Phase diagrams of magnetic transitions, a) Bean-Rodbell
magnetic compounds consisting of two magnetic sub- theory [12], b, c) experiment [26].
lattices, transitions from the ferromagnetic to the anti-
ferromagnetic and then to the paramagnetic state are Table II. Experimental data on the compressibility K and
possible when the pressure and temperature are d®/dP effects in certain ferro- and antiferromagnetic
changed. Figure 10a shows the equilibrium diagram
compounds (the parameter r\ was calculated from
taken from [12] . The diagram was constructed for the
formulas (2.1) and (2.4) of the Bean and Rodbell theory)
case rj = 2 in coordinates P*k /3 and T/T o , where P*
Cora- X.101!
= P — aT//c. The transition from P* to P changes the dSidP,
pound e, °K 103 deg/atm cm /dyne a 1 Type of transition Liter-
ature
initial slope of the phase-transition line, and rotation
of the coordinate axes through 90° leads to a phase dia- MnAs 313 -16.0±0,3 4.5 12 1.82 F - P lstotd. 21, 26
gram that corresponds to that obtained experimentally MnTe 310 2.3±0.3 -3.2 0.-15 AF - p 2dotd. 30. 31
MnP 291 -1.2±0.05 u 4 0.35 F - P 2d Old. 32
(Figs. 10 b—c). However, this agreement should be r e - CrTo 343 -6.2±0.4
2.2
8 0.3i F - P 2d ord. 32, 34
garded only as qualitative, since the model calculated by
Bean and Rodbell is a rather crude approximation when
applied to MnAs. The main defect of the calculation is qualitative description of magnetic phase transforma-
that the anisotropic character of the exchange interac- tions, since the conditions for realization of first or
tions existing in real crystals are ignored.* Attempts second-order transitions agree with ^he theoretical cri-
at a quantitative calculation of the concrete expressions terion T; defined by (2.4). This can be seen from Table
for the functional dependence of the temperature of the n, which gives the values of the parameters /3 and r]
transition on the external magnetic field intensity in the calculated from formulas (2.1) and (2.4) on the basis of
region of strong fields are likewise unsatisfactory 1153 . the experimental data on d®/dP and the compressibility
The causes of the discrepancy between the values of for certain ferro- and antiferromagnetic compounds.
d®i/dH calculated on the basis of this model and the
measured values in pulsed magnetic fields on the order It is seen from the presented data that r\ > 1 for
of 120 kOe are not clear, but the different determina- first-order phase transitions and r/ < 1 for second-order
tions of d®i/dH in constant fields (up to 30 kOe) agree transitions, in full agreement with the conclusions of the
with one another C28 ' 293 . Nonetheless, it can be stated on theory (see the summary Table III at the end of Sec. 8).
the whole that the Bean and Rodbell theory is a correct
3. MAGNETIC FIRST-ORDER TRANSITIONS OF THE
ORDER-ORDER TYPE
*In a recent paper, Goodenough and Kafalas [12S] present a new inter-
pretation of the magnetic phase transition in manganese arsenide. It is 4) Kittel's Theory of Magnetic Transformations with
proposed that besides the strong volume dependence of the exchange Inversion of the Exchange Interaction
parameters, there is also a change of the magnetic moment of the Mn
atoms with the changing volume of the MnAs. The magnetic transforma- Investigation of the magnetic transformations with
tion at the temperature @i is connected with electronic transitions due inversion of the exchange interaction is an interesting
to the change of the width of the 3d-band and to the considerable ex- theoretical problem, since ordinary thermodynamic
change striction in the basal planes. theories of phase transitions, using the model of the
734 N. P . GRAZHDANKINA

molecular field with a temperature-independent exchange take into account the dependence of the sublattice mag-
interaction, are unable to explain these transformations. netization on the temperature and on the interatomic
Smart'35-1 has shown that for a relative magnetization distances, i.e., he assumes |M^I = |MJJ| = M, and also
ffj in the sublattice j there exist (N/n)(l + <7j) spins ori- that the exchange interactions vary linearly with a, i.e.,
ented in one direction and (N/n)(l — a A spins oriented only the first term is taken into account in the Taylor
in the opposite direction; the entropy of such a system expansion of the exchange energy
is

The equilibrium value of the lattice parameter is ob-


tained by minimizing F with respect to a at the given
It is seen from this expressions that the entropy of the temperature:
system does not change if the directions of all the spins
of the given sublattices are reversed, since S is an even
function of Oj. Thus, if the directions of the magnetiza-
tion in some of the sublattices are reversed during the from which we determine the change of the parameter a
transformation without a change of cr, such a transforma- when the substance goes from one magnetic state to the
tion occurs without a change of entropy, and consequently other:
is a second-order phase transition. Experiments, how- Aa = o,i>-aAF = ^-M 2 . (3.4)
ever, refute this conclusion. It has been established ex-
perimentally that magnetic transformations connected It is seen from this expression that the sign of Aa is
with the change of the type of spin ordering are not determined by the sign of p, i.e., by the sign of the effect
second-order transitions but first-order transitions. of the change of the transition temperature T s with
Smart was the first to point out that allowance for the changing pressure. Substituting (3.3) in (3.1) yields
dependence of the coefficients of the Weiss molecular
field on the temperature makes it possible to explain F / p2 \ ,,» -r . , , ,, ,, ,n e\
~y~ — — I 2 ^ I (MAM B ) — p \aT — ac) MAMB. ("•"/
magnetic transformations with inversion of the exchange,
which is essentially equivalent to assuming that the ex- W h e n t h e s i g n of the e x c h a n g e i n t e r a c t i o n c h a n g e s , t h e
change integrals depend on the interatomic distances. s i g n of t h e f i r s t t e r m i n ( 3 . 5 ) r e m a i n s u n c h a n g e d ; i t i s
Neel C36] , Pratt C37] , and LiuC38] also attempted to explain p e r f e c t l y o b v i o u s t h a t t h e e f f e c t i v e s i g n of t h e e x c h a n g e
these transitions. Among the shortcomings of the cited i n t e r a c t i o n i s d e t e r m i n e d b y t h e s i g n of t h e s e c o n d t e r m
papers is that they did not reveal the cause of the change i n ( 3 . 5 ) . O n t h e o t h e r h a n d , if t h e m a g n e t i c s u b l a t t i c e s
in the magnitudes of the exchange interactions with a r e o r i e n t e d a t a n a n g l e t o e a c h o t h e r , a n d <p i s t h e a n g l e
changing temperature. b e t w e e n M/^ a n d M g , t h e n i t i s n e c e s s a r y t o i n t r o d u c e
in (3.5) a t e r m containing cos2q>, which a c t s like a p o -
Kittel [1] gives a physical justification for the change
tential barrier separating the two magnetic states. This
of the exchange integrals, and explains the main features
c a n l e a d t o a h y s t e r e s i s of t h e t r a n s i t i o n t e m p e r a t u r e
of the phase transitions connected with the change of the
T s , d e t e r m i n e d by the r e l a t i o n A T S = 2 a p M 2 / y c L , w h e r e
type of magnetic ordering. The transition from the anti-
C L is the lattice specific heat, and y is the Gruneisen
ferromagnetic into the ferromagnetic state is considered
c o n s t a n t . T h e i n f l u e n c e of t h e p r e s s u r e a n d of t h e m a g -
by Kittel on the basis of a two-sublattice model. It is
netic field on the transition t e m p e r a t u r e w a s determined
assumed that the change of the character of the magnetic
f r o m the Clausius-Clapeyron equation; this yielded
ordering at the transition point is due to the fact that
there exist in the substance simultaneously different
types of exchange interactions, differing from one dP ) B
(3.6)
another in magnitude, sign, and in the different depen- dT )
dence on the interatomic distances. If, owing to the DT, = J _ ( daT\
Aa 2MV (3.7)
thermal expansion, the lattice parameter assumes a dH
dT pM { dT )
certain critical value a c , at which the positive exchange-
interaction integral reverses sign, then a transition Kittel's attempt to use the theory developed by him
takes place from one magnetic state to another. The to explain the magnetic transformation in MnAs required
free energy at atmospheric pressure is written in the an additional calculation, connected with the considera-
form tion of the possible exchange interactions realized in
hexagonal crystals of the NiAs type. A formula was ob-
F = -^-RV(a — aTf — p (a — ac) FMAMB, (3.1) tained for Young's modulus in terms of the pressure
shift of the transformation temperature dT s /dP and the
where the first term determines the change of the elas- thermal coefficient of linear expansion a:
tic energy, and the second the change of the exchange
energy with changing lattice parameter a. Here M^ and
M3 are the sublattice magnetizations, V is the volume,
p = da/3a. is the rate of change of the exchange interac- T h e v a l u e o f E c a l c u l a t e d b y m e a n s of t h i s f o r m u l a f o r
tion as a function of the interatomic distance, R is the M n A s d i f f e r e d by o n e o r d e r of m a g n i t u d e f r o m t h e m e a s -
elastic constant divided by a2, and a-p is the value of the u r e d Y o u n g ' s m o d u l u s , a n d t h e s i g n of t h e e f f e c t d T s / d P
lattice parameter when M^. J- M Q , corresponding to zero a l s o d i s a g r e e d w i t h t h e c o n c l u s i o n s of K i t t e l ' s t h e o r y .
exchange energy, since the second term of Eq. (3.1) T h e s e q u e s t i o n s w e r e a l r e a d y c o n s i d e r e d by u s in s u b -
vanishes in this case. For simplicity, Kittel does not s e c t i o n 3 , i n t h e a n a l y s i s of t h e e x p e r i m e n t a l d a t a o n
M A G N E T I C F I R S T O R D E R PHASE TRANSITIONS 735

MnAs, which indicate that the magnetic transformation G-cm*


in MnAs is apparently not connected with inversion of
the exchange interaction, and is more likely due to 40 \
destruction of the spin disorder. \\
Recently, much experimental material was obtained 30
on magnetic transformations for which it has been FIG. 11. Temperature dependences
demonstrated experimentally and without any doubt that of the magnetization of the alloys
they are magnetic transitions with exchange inversion. Mn2.xCrxSb.l -x = 0.13, 2 - x = 0.05, ZO
3 - x = 0 (Mn2 Sb).
Such transitions were observed in substances with dif- - 1
ferent (crystal and magnetic) structures (subsection 8, 10
table III). We shall consider below only the most charac- r •
teristic cases of such transitions (Mn 2 _ x Cr x Sb, FeRh, l \1
and Mn3GaC), investigated most fully and most system- 0 WO
atically. T.'/f
change takes place in the lattice parameters: the c axis
5) Comparison of Theory with Experiment. Magnetic increases jumpwise, and the a axis decreases at the
Transformations in Mn 2 . x Cr x Sb, FeRh, and Mn3GaC point T s . It is seen from Fig. 12 that the transition
a) When Mn2Sb is alloyed with chromium, solid solu- temperature changes with the quantity x, but in different
tions of the type Mn 2 _ x Cr x Sb are produced. On the basis alloys the AF — F always occurs at the same value of
of neutron-diffraction141'42-1 and magnetic measure- the parameter 6.53 A. On this basis, the hypothesis was
ments [ 4 3 '" ] it was established that, depending on the advanced that the main cause of the AF — F transition
value of x, which determines the chromium concentra- in these compounds is the reversal of the sign of one of
tion, these alloys have different spin ordering at low the exchange-interaction constants, owing to the strong
temperatures: ferromagnetic (x = 0), helical (x < 0.022), dependence of the exchange integrals on the interatomic
and antiferromagnetic (0.025 < x < 0.20). When the tem- distances. In this connection, when the lattice parameter
perature is increased, a change takes place in the type reaches as a result of thermal expansion a certain criti-
of magnetic structure from antiferromagnetic (AF) or cal value, the AF — F transition takes place. It is as-
helical (H) ordering to a ferrimagnetic structure (F) of sumed that addition of chromium to the Mn2Sb leads only
the Mn2Sb type. In the concentration region 0.025 < x to a decrease of the lattice parameters, and causes no
< 0.035 the magnetic structure changes twice: AF — H significant changes in the exchange or magnetic interac-
— F t 4 5 ' 4 6 1 . All the indicated transformations, which tions, since the spontaneous magnetization, the Curie
occur with increasing temperature, are first-order temperature, and the magnetic crystallographic aniso-
phase transitions, as is evidenced by the release of heat, tropy t47] of the Mn 2 . x Cr x Sb alloys change only insignifi-
by the jumplike change of the volume and magnetization, cantly with changing x. In this connection, one might ex-
and also by the thermal hysteresis in the vicinity of the pect that the introduction into Mn2Sb of not only chrom-
transition temperature T s . The AF — F transitions are ium, but also of other elements with atomic radii
accompanied by a sharp change of kinetic effects such smaller than those of Mn and Sb would lead to a com-
as electric conductivity, thermal emf, Hall effect, and pression of the lattice and cause its parameters to ap-
change of the electric resistivity in the magnetic proach the critical dimensions, thereby resulting ulti-
field C5i: . Investigations of Mn 2 _ x Cr x Sb alloys in strong mately in an AF —- F transition with increasing tem-
pulsed magnetic fields150'52-1 have shown that the anti- perature. Indeed, addition of Cu, Va, Co, Zn, Ge, and
ferromagnetic state can be destroyed by a sufficiently As to the Mn2Sb also led to AF — F transformations' 433 ,
strong magnetic field exceeding a certain threshold thus demonstrating the validity of the indicated assump-
value H s . The transition temperature T s can also be tions concerning the nature of the magnetic phase tran-
altered by the action of high pressures. sitions in these alloys.
The first to observe first-order magnetic phase tran- Neutron-diffraction investigations by Claud and
sitions in Mn 2 _ x Cr x Sb were Swoboda et al. t 4 3 ] , who ini- Jarrett [ 4 1 ] , and later by Austin et al. C42] have shown that
tiated an extensive cycle of research on magnetic trans- the magnetic structure of Mn 2 . x Cr x Sb (in the ferrimag-
formations with inversion of the exchange interaction. netic state) is the same as that of Mn2Sb (Fig. 13).
X-ray diffraction and magnetic measurements of
Mn 2 . x Cr x Sb were made mlt31 in the range of concentra-
tions 0.025 £ x * 0.20. Figure 11 shows the temperature see
dependences of the magnetization for some of the inves-
tigated alloys. The plots reveal a sharp increase of a at
the point T s , connected with the occurrence of ferro-
G55
magnetic order, which in turn is destroyed with further
increase of temperature at the Curie point. X-ray dif- FIG. 12. Temperature dependences ^t
fraction studies have shown that the AF — F transition of the parameter c of the crystal lat-
tice of the alloys Mn2_xCrxSb.
leads to a change of the magnetic structure only, and is
not accompanied by a change of the crystallographic x-O/H
symmetry of the lattice. Both above and below T s , the
lattice has a tetragonal symmetry of the type P4/nmm. ff.W
In the case of the magnetic transformation, a sharp ZOO 300 400 SOO Sf
T,°K
736 N. P . GRAZHDANKINA

ferromagnetism is produced in the alloy at the transition


point T s with increasing temperature. Later, on the
basis of magnetic C54)55] , x-ray diffraction tS6 \ and
neutron-diffraction:57~60:I investigations it was estab-
lished that this transformation is connected with a
change in the type of the magnetic structure and is a
transition from the antiferromagnetic into the ferromag-
netic state. The alloys have an ordered body-centered
cubic structure of the CsCl type. In the antiferromag-
netic state, each atom of iron is surrounded by six other
atoms of iron with opposite spins; the magnetic moment
of the iron is equal to 3.3 /xg, and the magnetic moment
of the rhodium atoms is zero. At temperatures above
FIG. 13. Magnetic structures of the alloys Mn2-xCrxSb in the ferro- T s , the orientation and magnitude of the magnetic mo-
magnetic (a) and antiferromagnetic (b) states. The circles denote Mn I, ments change, i.e., the magnetic moments of the iron
the squares Mn II. In the antiferromagnetic state, the magnetic moments and rhodium in the ferromagnetic state are respectively
of all the neighboring layers are antiparallel; in the ferromagnetic state,
the Mn I layers are antiparallel to the Mn II layers. The arrows label the 2.8 ixg and 0.8 /xg and are parallel to each other. The
three-layer goups Mn II-Mn I-Mn II. transition temperature is very sensitive to the chemical
composition of this alloy [ " ' 5 5 ] . Hofer and CuckaC81J and
also Kouvel1623 have shown that a deviation of the com-
Substitution of Cr for Mn leads to a change in the values position of the FeRh from stoichiometric or the intro-
of the magnetic moments of Mn I and Mn II, which oc- duction of Pt, Pd, and Ir impurities leads to a strong
cupy different crystallographic positions. In the anti- change of T s . This apparently explains the disparity
ferromagnetic state, the magnetic moments of Mn I and between the values of the latent heat of the transition
Mn II are respectively 1.4 jug and 2.8 /xg. Upon transi- and the absolute values of T g and dT s /dH obtained by
tion into the ferrimagnetic state (T > T s ), the magnetic various authors [63 " 68] . It is perfectly obvious that for a
moment of Mn I increases by 0.4 /xg, and the magnetic reliable comparison of the experimental data with the
moment of Mn II decreases by 0.5 /xg- m order to ex- theory comprehensive measurement of several physical
plain the resultant value of the magnetic moment, namely quantities on the same samples are essential. This was
~0.9 HQ, it was proposed in :48 ' 49] to regard the magnetic done by Zakharov, Kadomtseva, et al. [66] who investiga-
structures of these compounds as three-layer groups ted the temperature dependences of the magnetization
Mn II-Mn I-Mn II, formed by the planes (001), having (Fig. 14), Young's modulus, and the crystal lattice
ferrimagnetic ordering above T s and antiferromagnetic parameter of the alloy FeRh; simultaneously they per-
ordering at low temperatures (Figs. 13a and b). Within formed dilatometric investigations of this alloy (Fig. 15).
the limits of each of the indicated planes, there is a They measured also the values of dT s /dP and dT s /dH.
strong exchange interaction of the spin magnetic mo- Analyzing the experimental data, the authors of the cited
ments, and the exchange between different planes is paper reached the conclusion that the AF — F interac-
much weaker. Starting from such a magnetic model, tion in FeRh can be explained on the basis of Kittel's
one could assume that these alloys are a very convenient theory. If we assume that the exchange interaction in a
system for an experimental verification of Kittel's
theory. If it is assumed that the interplanar exchange
interaction reverses sign at a certain critical value of
the parameter c, then according to the Kittel theory this
change should cause a first order AF — F phase transi-
tion accompanied by a jumplike change in the parameter
Ac = 2pM2/R- This conclusion of the theory, and also
the relations (3.6) and (3.7), were confirmed experimen-
tally [48~503. It should be noted that this agreement must 700
be regarded only as a qualitative verification of Kittel's T,'H
theory. This is connected with the fact that it is im- FIG. 14. Temperature dependence of the magnetization of the alloys
possible to measure the elastic modulus of the alloys FeRh at H = 1770 Oe (• - heating, O - cooling) and H = 14,500 Oe
Mn 2 . x Cr x Sb. For the same reason, Jarrett's (X - heating) (from [<*]).
attempts l3a: to estimate the different parameters of the
AF — F transitions on the basis of the more exact
theoretical relations derived by him are also of qualita-
tive character. Nonetheless, as shown by numerous
experimental investigations, the theory reflects cor-
rectly the main and most characteristic features of
transitions with inversion of the exchange, and can be
applied not only to layered uniaxial ferro- and antiferro-
magnets, but also in a number of cases to other magnetic
structures, such as for example FeRh and Mn3GaC.
b) Magnetic first-order phase transitions in FeRh FIG. 15. Temperature dependence of the thermal expansion of the
alloys were first observed by Fallot[53:l who noted that alloy FeRh. Solid — experimental data, dashed - calculated curve I66].
MAGNETIC FIRST ORDER PHASE TRANSITIONS 737

cubic crystal has an isotropic character and depends of the fact that the signs of the dT s /dP and dT s /dH
only on the volume, then expression (3.3) of Kittel's effects agree with the general thermodynamic relations
theory takes the form V / V T = 1 ± pfC-pM2. The tempera- given above, a difficulty arises in the interpretation of
ture dependence of the thermal expansion, calculated in these effects on the basis of Kittel's theory.
accordance with this formula, agrees qualitatively with The magnetic transformation in Mn3GaC differs from
the experimental data (see Fig. 15). the AF — F transitions in Mn 2 . x Cr x Sb and FeRh in the
c) As shown by Bouchaud and Fruchart [ 6 7 ' 6 8 : , the fact that the change of the spin ordering with increasing
compound Mn3GaC experiences an AF — F transition at temperature at the point T s in Mn3GaC is not accom-
the point T s = 150°K with increasing temperature. panied by an expansion of the lattice (as would be expec-
X-ray diffraction and magnetic measurements have ted from Kittel's theory), but by a strong contraction.
shown that this transformation, as well as the majority In this connection, the experimentally observed negative
of transitions with exchange inversion, is not accompan- values of the dT s /dP and dT s /dH effects do not agree
ied by a change in the crystallographic symmetry with Kittel's formulas (3.6) and (3.7). Pauthenet and
(perovskite-type lattice). The AF —* F transition is a Guillot t69] made an attempt to eliminate these contra-
typical magnetic first-order phase transition, as can be dictions. They also analyzed the AF — F transition on
clearly seen from the temperature and field dependences the basis of the exchange-inversion mechanism of mag-
of the magnetization (Figs. 16 and 17). It was established netic transitions, but took additional account of the
that the transition temperature changes under the influ- ferromagnetic and antiferromagnetic states of the spin
ence of an external magnetic field: dT s /dH = 0.33 entropy in the expression for the thermodynamic poten-
x 10"3 deg-Oe"1 and under the influence of high pressure: tials, and also of the exchange interactions, both inside
dT x /dP = - 2 . 5 1 x 10~3 deg/atm. From these data, on the the sublattices and between the sublattices; this led to
basis of the Clausius-Clapeyron equations a qualitative agreement between theory and experiment.

AV
<?! = (llT,)(dT,ldP)
4. MODELS OF MAGNETIC FIRST-ORDER TRANSI-
and TIONS NOT CONNECTED WITH THE EXCHANGE-
Ao STRICTION MECHANISM
S) (dHsldT)
So far we have considered investigations in which the
they calculated the heat of transition Q in the AF — F magnetic first-order phase transitions were explained
transformation. These two expressions give the closest on the basis of the exchange-striction mechanism.
values Qi = 0.93 and Q2 = 1.02 cal/g. However, in spite There are, however, theoretical papers in which these
transitions are explained with the aid of other model
representations, not connected with the assumption of a
G-cm' strong dependence of the exchange-interaction integral
on the volume of the crystal lattice. Thus, for example,
Yosida and MiwaC70] explain magnetic transformations
of the order-order type, i.e., transitions with inversion
of the volume, occurring in rare-earth elements with
helicoidal magnetic structure, on the basis of the com-
plicated temperature dependence of the constants of the
magnetic crystallographic anisotropy. According to
Dzyaloshinskii l l u , the magnetic structure of rare-earth
metals is determined to a considerable degree by the
shape of the Fermi surface. Transitions between differ-
ent structures are connected with changes of the elec-
tron energy spectrum, which become strong when the
wave vector of the magnetic structure coincides with
Iff - the extremal diameter of the Fermi surface. The source
of all these changes is the temperature dependence of
the parameter of the exchange interaction. In the case
when the exchange parameter has smooth values at which
FIG. 16. Temperature dependences of the magnetizations of the alloy conditions for the existence of the minimum of the
Mn3GaC [68].
thermodynamic potential are violated, jumplike changes
can occur in the magnetic structure, i.e., a first-order
G-cm phase transition.
Karpenko and Berdyshev [72] associate the inversion
of the exchange interaction in antiferromagnetic semi-
conductors of the Lio.iMno.9Se type with the change of the
electric carrier density. In this case it is assumed that
there are two competing types of exchange interactions:
H,iOe an antiferromagnetic interaction of the Kramers type
FIG. 17. Isotherms of magnetization of the alloy Mn3GaC [68]: 1 - (low temperatures) and a ferromagnetic interaction via
T = 146°K, 2 - T = 138°K, 3 - T = 134°K, 4 - T = 121°K, 5 - T = 111% the conduction electrons, the role of which increases
6 - T = 96°K. with increasing temperature as the result of the strong
738 N. P. GRAZHDANKINA

increase of the carrier density. Thus, the effective sign ies in which account is taken of the volume dependence
of the exchange is determined by the relative contribu- of the exchange interactions did not yield positive r e -
tions of interactions with opposite signs. According to sults. It was therefore assumed that the magnetic first-
BlumC73], magnetic phase transitions of the order-dis- order phase transition in the indicated compounds is due
order type can arise in many cases in connection with not to the exchange-striction mechanism, but to some
the special character of the splitting of the electronic other causes.
energy levels, due to the crystal and the magnetic fields. Rodbell, Jacobs, et a.l.l7B1 have shown that if account
The possibility of a first-order transition from the is taken of the biquadratic term in the isotropic indirect
ferromagnetic into the paramagnetic state was indicated exchange during the course of calculation of the tem-
by Shimizu190-1 on the basis of an analysis of the criterion perature dependence of the magnetization in the approxi-
of ferromagnetism in accordance with the band model. mation of the molecular-field theory, i.e., if it is as-
In the opinion of Mitsudo et al. t 7 5 : l , and also of Perm and sumed that the energy of the exchange interaction is of
CohenC76:, first-order transitions in the Neel point of the form
antiferromagnetic metals (e.g., the magnetic trans-
formation in CrC77:i) can be attributed to singularities of /(sV)-/(sV)», (6.1)
the electronic energy spectrum of these metals. The then it is possible to obtain satisfactory agreement be-
hypothesis were advanced that a large influence is exer- tween the calculatedCT(T)curve and the experimental
ted on the character of the magnetic phase transition by data. Anderson :80] first pointed out the possible pres-
the biquadratic indirect exchange in certain antiferro- ence of a term biquadratic in the spins, - j ( s a s b ) 2 , in the
magnetic compounds. [78] expression of the exchange energy, and estimated ap-
In a recently published paper, Bulaevskif and Khom- proximately the magnitude of this term. Later, on the
skii t7 * :i considered, on the basis of the self-consistent basis of an experimental investigation of the paramag-
field method, magnetic transitions in antiferromagnetic netic resonance t81] and a calculation of the interaction
semiconductors (such as V2O3 and Ti2O3[l20~122]) with of the paramagnetic ions in antiferromagnets C82] it was
allowance for the possible excitation of localized elec- also concluded that a biquadratic term must be taken
trons in the conduction band. In view of the strong de- into account in the isotropic indirect exchange. It was
pendence of the energy gap of the electronic spectrum shown there that the ratio j / I is of the order of
of the crystal on its magnetic structure, the magnetic 0.01—0.05. In™ , the temperature dependences of the
transformation is accompanied by an appreciable change magnetic moments of the sublattices of the antiferro-
of the electric properties, namely, when the magnetic magnetic compounds NiO and MnO were calculated on
order vanishes the substance is transformed from a the basis of the theory of the molecular field; no account
semiconductor into a metal. These transformations are was taken of the change in the elastic properties during
first-order phase transitions if the gap in the paramag- the magnetic transformation, and only the change of the
netic state is sufficiently small. exchange energy and of the entropy of the spin system
The influence of the magnetic crystallographic aniso- were taken into account. Allowance for the biquadratic
tropy on the magnetic transitions with exchange inver- term in the exchange energy leads to a change in the
sion were considered in many papers, and were dis- form of the temperature dependence of the magnetiza-
cussed in detail in the monograph of Belov, Belyanchi- tion in accordance with the formula
kova, et al. I 7 9 ] , so that we shall not concern ourselves
with these questions here. We shall dwell only on model (6.2)
representations of first-order magnetic transitions of
the order-disorder type, which are not connected with where
SNksa
the exchange-striction mechanism. F(o) = - OS*
"*" ' da
here S* is the entropy of the spin system, ji and j 2 are
6) Biquadratic Indirect Exchange and Magnetic Trans-
the parameters of the biquadratic exchange with the
formations in Certain Antiferromagnetic Compounds first and second coordination spheres, respectively, and
Neutron-diffraction investigations of the antiferro- I2 is the exchange parameter with the second coordina-
magnetic compounds NiO and MnO have shown that the tion sphere in the bilinear term of the Hamiltonian
change of the magnetization with temperature in the I(s a s b ). Figure 18 shows the dependence of a on T / T J J
region of the Neel point does not follow the Brillouin for NiO, calculated in accordance with formula (6.2)
function: the a(T) curve is steeper, and the magnetiza- under the assumption that Ii = I2 = I and ji = j 2 = j l 7 9 2 .
tion jumps down to zero in the immediate vicinity of TJJ. Agreement with the experimental data is obtained when
In this connection, it was proposed that the transition j / I = 0.015 for NiO and j / I = 0.02 from MnO. From the
from the antiferromagnetic into the paramagnetic state plots of Fig. 18 we see that the magnetization in the
in these compounds is a first-order phase transition. A region of the Neel point changes sharply when the tem-
characteristic and very significant feature of the des- perature changes, and the character of this change
cribed phase transitions is the sharp change of the elas- corresponds to a first-order phase transition. In the
tic constants of the crystal in the absence of any volume case when j = 0, i.e., when no account is taken of the
changes in the region of the antiferromagnetic trans- biquadratic indirect exchange, calculation in accordance
formation. The effects of the shift of the transition tem- with (6.2) leads to the usual dependence of the magne-
perature with changing pressure are relatively small in tization on the temperature, corresponding to the Brill-
these compounds. An attempt to calculate the jump of ouin function, and consequently to a second-order mag-
the magnetization at the Neel point on the basis of theor- netic phase transition (dashed curve of Fig. 18).
M A G N E T I C FIRST ORDER PHASE TRANSITIONS 739

use of a statistical method of investigating phase tran-


FIG. 18. Temperature dependence sitions. In this case, as is well known, it is necessary
0,8 of the magnetization of the antiferro- to know the distribution function Z. If the picture of the
\ \ \\ magnetic sublattices of NiO. The dashed splitting of the energy levels corresponds to the case
US \\ curves were calculated on the basis shown in Fig. 19, then the distribution function is of the
of the classical molecular-field theory;
Oft NiO the solid curves are calculated with form
allowances for the biquadratic exchange
0,2 \ [78]; the points are the experimental
i data of neutron-diffraction investiga- The magnetization of the system is given by
o,s T/TH tions.
(7-2)
7) Blum's Theory of the Magnetic Transformation in
Uranium Dioxide
If only the internal magnetic field H = XM is taken into
According to Blum1731, magnetic first-order transi-
account, the equation for the magnetization takes the
tions can arise in many cases in connection with the
special character of the splitting of the electronic en- form
ergy levels, due to the crystal and magnetic fields. * — e~x) (7.3)
Figure 19 shows by way of an example the schematic eA/kT-
picture of such a splitting. Under the influence of the XM kTx
electric field of the crystal, the degenerate level is first (7.4)
split into two sublevels, a singlet and a triplet. The
energy difference between the low-lying nonmagnetic The presence of the term e^/kt in (7.3) complicates
singlet and the higher triplet is A. The magnetic field somewhat the solution of this equation. Here, as in the
causes an additional splitting of the triplet into two sub- usual case, when the magnetization is determined by the
levels. The magnetic properties of the system, and also Brillouin function, the solution for M is obtained graph-
the character of the magnetic phase transition, are de- ically by plotting M against x and determining their
termined by the ratio of the splitting energies A and intersection points. For convenience, new variables
g/j.BH, due to the action of the electric and magnetic M' = M/g/^3 and A' = A/(gMB)2^ a r e introduced, and
fields. The calculation is performed in the approxima- the equation for the magnetization is written in the form
tion of the molecular-field theory, i.e., the magnetic (7 5)
field that splits the triplet is produced by the internal *'= ^ T ^ + o • '
molecular field due to the presence of exchange interac- By solving this equation for various values of A' , it was
tion between the magnetoactive ions. The magnitude of shown that for large A ' the magnetic-order state is im-
this field is proportional to the magnetization H = A.M, possible, since the exchange interaction does not suffice
where X is a constant determined by the integral of the to overlap the splitting due to the crystal field. For
exchange interaction. If a magnetic field H is weak, i.e., small values of A' , the solutions for M' are analogous
weaker than A/gjig (Fig. 19a), then the magnetization to those of the ordinary case, when the dependence of
of the system in the ground state will be zero, since the the magnetization of T and H is determined by the
lowest level is the nonmagnetic singlet, and the influ- Brillouin function (Fig. 20) and the character of the
ence of all the higher energy states of the system can variation of M' with T corresponds to a second-order
be neglected. On the other hand, if the magnetic field H phase transition. First-order phase transitions are
is sufficiently large, then one of the components of the possible only for definite values of A ' lying in the inter-
triplet will lie below the singlet and will make a contri- val 0.3662 < A' < 0.5. In this case the dependence of M'
bution to the magnetization, so that at T = 0°K the sys- on x has the form shown in Fig. 21, from which it is
tem has a resultant magnetic moment (see Fig. 19b). seen that only in the region of high temperatures T > T^
The decrease of the magnetization with increasing tem- (curve 3) is there one solution for M', namely M' = 0.
perature, connected with the population of the excited At low temperatures (curve 1) there are two solutions,
states, leads to a decrease of the splitting of the triplet and at values of T close to Tc (curve 2) there are three
and consequently to a change of the energy position of solutions for M'. The determination of the stable solu-
the magnetic component of the triplet. Under certain tions was based on an analysis of the change of the
conditions, it is a first-order phase transition. To de- thermodynamic Gibbs potential G as a function of the
termine the equilibrium properties of the system, and magnetization and the temperature:
also the conditions that lead to the particular character
of the magnetic transformations, Blum's theory makes

FIG. 19. Diagram of energy levels, a)


Singlet-triplet splitting A is smaller than the
splitting of the triplet due to the magnetic
field; b) splitting of the triplet by the mag-
netic field is larger than the splitting A of (0 Zfl 3fl 1J1 5fl
the crystal lattice by the electric field.
FIG. 20. Graphic solutions of Eq. (7.2) of Blum's theory [73] for the
case when the magnetic transformation is a second-order phase transition.
740 N. P . GRAZHDANKINA

analysis. In spite of the fact that Blum's calculations


give a qualitative correspondence between the theory and
experiment, recently doubts were advanced concerning
the possible interpretation of the magnetic transition in
UO2 on the basis of the foregoing model representations.
This is connected with the fact that recently published
results C84] of dilatometric measurements (Fig. 23) and
measurements of other properties of UO2 (Fig. 24) indi-
iff cate quite distinctly that the volume of the crystal lattice
FIG. 21. Graphic solutions of Eq. (7.2) of Blum's theory [73] for the changes during the transformation, and the elastic con-
case 0.366 < A' < 0.4. The magnetic transformation is a first-order phase stants decrease abruptly. These data indicate that it is
transition. necessary to review the mechanism of the magnetic
phase transition in UO2; this again raises the question
of the exchange-striction nature of the magnetic trans-
T)-G(O, T)} formation. For a final solution of this question further
experimental investigations are needed, primarily meas-
= - T l n [ 1 + '"' /T( urement of the effect of the shift of the Neel temperature
(7.6) of UO2 when the pressure is varied.
Here x = kT/A. Figure 22 shows the plots of the thermo- It must be noted that the antiferromagnetic trans-
dynamic potential against the magnetization, determined formation in Eu is also a first-order phase transi-
C8s]
from this formula, for various temperatures. In the tion . If account is taken of the fact that the picture
region T > T c (curve for x = 0.790) there is a single of the splitting of the initially degenerate levels in Eu
extremum for the states with M' = 0; this corresponds into "nonmagnetic" and "magnetic" sublevels is the
to the curve 3 of Fig. 21. At low temperatures, as seen same as in UO2, and the magnitude of the dTN/dP effect
from the curve for x = 0.770 on Fig. 22, the stable state is small C88] , then it would be of interest to verify the
is the one with the finite value of M'. At temperatures possibility of explaining this transition on the basis of
close to the magnetic transformation point (T = 0.780 the foregoing model.
and x = 0.786) the plots of 1 / A { G ( M ' , X) - G(0, x)}
against M' have two minima and one maximum, corre- 8) Criterion of Ferromagnetism in Accordance with the
sponding to the intersection points of the straight line 2 Band Model, and Magnetic First-order Phase Tran-
in Fig. 21. If both minima are at the same level for a sitions
certain temperature (the curve for x = 0.786), then a Theoretical investigations of magnetic phenom-
first-order phase transition takes place, for in this case ena [87 - 89] , based on the assumption of collectivization of
the magnetization drops suddenly from the finite value "magnetic" 3d- and "valence" 4s- and 4p-electrons
M' to zero. and their consideration as a single system of an elec-
The foregoing considerations were used by Blum as a tron gas have led to the creation of the band theory of
basis for explaining the nature of the magnetic trans- magnetism. According to this theory, the occurrence of
formation in UO2. Neutron-diffraction investigations' 833 ferromagnetic order of the spin magnetic moments is
have shown that the character of the temperature depen- possible only at definite relations between the kinetic
dence of the magnetization UO2 in the region of the Neel and exchange energies; it is necessary here that the
point (TJJ = 30.8° K) corresponds to a first-order phase level density at the Fermi surface y(e), and also the ex-
transition. The absence of volume changes of the crys- change energy, be sufficiently large. A calculation by
tal lattice during the transition, has led Blum to the as- StonerC89:l for the "normal" band, i.e., when the change
sumption that the transformation in UO2 is due to a of the density of states with energy has a parabolic form
change of the spin system only, and calls for a special v(i) ~ VT, had shown that the ferromagnetism criterion
for this case is a > a s , where a is the constant of the
Weiss molecular field and as is a constant characteriz-
0P05 ing the exchange interaction in the absence of a mole-

FIG. 23. Temperature dependences


of AL/L and of the coefficient of
thermal expansion of UO2.
•0005

F I G . 2 2 . D e p e n d e n c e of t h e free e n e r g y o n t h e m a g n e t i z a t i o n in ac-
c o r d a n c e w i t h B l u m ' s calculation [73]. First-order phase transition for
T = 0.786.
0 IB ZO 30 W 50
T,°K
MAGNETIC FIRST ORDER PHASE TRANSITIONS 741

minimizing E with respect to the magnetization:


4M2
dE li—l* , SEex (8.5)
dM -=0,

• 0. (8.6)
dM-
FIG. 24. Temperature dependence Following the Weiss hypothesis, it can be assumed that
of the elastic constants of UO2 in the the dependence of the exchange energy on M is given by
region of the antiferromagnetic trans- £ex -,-|ttJtf ! ; (8.7)
formation.
and then
HL ^ - aM and tfl-E
dM
taking (8.4) into account, we can conclude that E = 0 when
M = 0. Inasmuch as the total energy should be negative
0 W SO ISO WO ZOO ZW following ferromagnetic ordering, it is necessary to de-
T,°K termine the condition when E < 0 for M * 0. To analyze
the relations between E, E^, E e x , a, and M leading to a
cular field, and equal to the reciprocal of the paramag- stable ferro- or paramagnetic state, Shimizu used a
netic susceptibility of the electron gas at 0°K graphic method. Solving (8.4) by integrating by parts:
(a" 1 = Xo(0)). Attempts were made subsequently to de- El £0
termine the conditions for the occurrence of ferromag- E = G,-t,t)n- U1(e)dE-W2(e)d8 + £ e x , (8.8)
netism in a system of collectivized electrons with arbi- to £2
trary dispersion law, i.e., for an arbitrary band, when and plotting n against AS to correspond to the investiga-
the v{t) dependence can have an arbitrarily complicated ted density of states curve, he determined the change of
form. In an extensive investigation, Shimizu has the total energy, and also of the individual contributions
shown [90'9i:l that the ferromagnetism criterion in the Efc and E e x , as functions of the magnetization of the sys-
band model depends on the general topology of the den- tem. Figure 26 shows the change of n (number of singly
sity of states, the position of the Fermi level, and the occupied levels) for various values of the widths of the
magnitude of the molecular-field coefficient. It was es- energy band near the Fermi level; it is seen from this
tablished there that in a number of cases, when the figure that for the case when n = n' and AS = £1 - £2, the
Fermi level lies near the minimum of the density of area OG3G2F2n'O, between the curve and the ordinate
state curve, the transition from the ferromagnetic into axis is equal to the change of the kinetic energy deter-
the paramagnetic state can be a first-order phase tran- mined by the sum of the first three terms of (8.8). Tak-
sitions. ing into account (8.3), (8.5), and the functional depen-
In Shimizu's cited papers C9O'91], he investigated the dence of the exchange energy on the magnetization (8.7),
dependence of the kinetic and exchange energies on the we can determine the relation between A£ = 9E ex (n)/Sn
spontaneous magnetization of the system of electrons. and n (the straight lines OF 1; OF 2 , OF 3 , OF 4 , and OF 5 on
Figure 25 shows the change of the density of states as a Fig. 26), and also the absolute value of the exchange en-
function of the energy at 0°K. Here £0 is the Fermi ergy E e x , which equals the area between the indicated
level for the paramagnetic state and Vi(e) and v2(t) are straight lines and the ordinate axes. For the case
the density of states for e > £0 and e < £0, respectively. A£ = £1 - £2, this is the area OF 3 n'O. If the change of
The number of singly-occupied levels n(£), both above the kinetic energy is smaller than the contribution due to
and below the Fermi level, is the exchange energy E e x (the area OG3FiF2F3n'O is
smaller than the area OF3n'O), then the total energy is
) = I v, (e) de, n2 (y = § v2 (e) de. (8.1) negative and the system is in the ferromagnetic state.
If the change of Ej^ is exactly equal to the change of E e x ,
Depending on the form of the v(e) curve, the values of then the indicated two areas are equal and the state of
ni(£i) and n2(£2) differ. When spontaneous magnetization the system can be characterized as critical. For the
M occurs, some of the electrons occupying states below paramagnetic state, the total energy should be positive,
£o go over into states above £0, and then and consequently E^ must be larger than E e x . The de-
pendence of E on M for all the cases indicated above is
(8.2) shown in Fig. 27. An analysis of the change of the coeffi-

M = 2pBn. (8.3)

When ferromagnetic order sets in, a change takes place


in the kinetic energy E^ and in the exchange energy E e x ;
the total change of the electron energy is then
FIG. 25. Plot of the density
(8.4) of states vs. the energy.
- \
The equilibrium value of the energy is determined by (, („
742 N. P. GRAZHDANKINA

/N Ft MnAs with increasing pressure is explained on the basis


of the foregoing model in the following manner. With
increasing pressure, the degree of overlap of the wave
functions of the "magnetic" electrons increases, and
the curve of the dependence of the state density on the
energy becomes less steep. This effect leads to an in-
crease of the kinetic energy of the electrons, and by the
same token weakens the ferromagnetism criterion. In
this connection, the decrease of the temperature of the
magnetic transition with increasing pressure can occur
AC also in the case when the exchange energy and the co-
FIG. 26. Graphic solution of Eq. (8.8) in accordance with the Shimizu efficient of the molecular field do not change with in-
theory [91]. creasing pressure.
Systems in which magnetic first-order phase transi-
tions are observed are listed in Table III.
FIG. 27. Dependence of the energy
on the magnetization M. Curves 1 and 2 — 5. CONCLUSION
paramagnetic state, 5 and 6 - ferromag- On the basis of the foregoing review we can conclude
netic state with the usual second-order
phase transitions at the Curie point, 3 — that the occurrence of magnetic first-order phase tran-
critical state. 4 - the transition from the sitions is possible in a great variety of magnetic sys-
ferromagnetic into the paramagnetic
state is a first-order magnetic phase transit-
Table HI. Substances in which magnetic first-order
phase transitions were observed
Temper-
cients of the molecular field a has shown that Stoner's ature of
Magnetic magnetic
Substance Crystal structure transition transition, Literature
ferromagnetism criterion a > as is satisfied here, too
(curves 5 and 6 of Fig. 27). There may be cases, how-
ever, (curve 4 of Fig. 27), when the ferromagnetic state MnAs Hexagonal (type NiAs) F- P 313 98, 99, 16, 17, 19,
is possible even when a < a s , and the ferromagnetism orthorhombic313- 20, 21, 26, 15, 96
400°K 400
criterion is ac < a < as, while that for paramagnetism KMnF3 AF- p 100, 101
Monoclinic 88.3
is a < a c (a c is the coefficient of the molecular field
for the critical case). Cr3As2 Tetragonal F- p 213 102
It is seen from Fig. 27 that the dependences of E on TiCl3 Rhombohedral AF- p 210 103
M corresponding to the ferromagnetic state may differ MnBr2 Hexagonal layered AF-P 2.16 104, 105
in character (curve 4, on the one hand, and curves 5
and 6, on the other). This is of great significance, since NiO T < Tx, Rhombohedral AF - p 647 106, 107
it evidences that the character of the magnetic phase T>TN, Cubic 523
transitions is different for these cases. It can be as- MnO T < Ts, Rhombohedral AF-P 122 106
sumed that the magnetic transformation is a first-order T>TK, Cubic 115.9
phase transition when E(M) is determined by curve 4
(Fig. 27), and an ordinary second-order phase transition uo2 Cubic f. c. AF-P 30.8 83, 84
of the Curie point type in the case of curves 5 and 6. Eu Cubic, b. c. AF-P 90 85, 86
The conditions for the occurrence of first-order mag- Cr > AF ->. P 311.5 77
netic phase transitions within the framework of the M112 xCr^Sb Tetragonal AF - F 43, 41, 42, 40,
model of collectivized electrons are discussed in detail (0,025<x<0,20) = 200— 50, 52, 46, 51
350
in 1913 . It is noted there that for a band with an arbitrary » AF - H r.M = 46, 45, 42
Mn^Cr.Sb
dispersion law there can be realized conditions (depend-
ing on the position of the Fermi level) in which the 150 44
(0,025<a;< 0,035) H -* F
ferromagnetic state has two temperature boundaries, =50—150
i.e., the system is paramagnetic at high and low tem- Mn3Ge2 AF- F 148 92, 93, 94, 63, 95
peratures, and ferromagnetism arises in some inter- F* P 283
mediate temperature interval. It is important here that MnP Rhombohedral AF - F 50 108
both the occurrence and the vanishing of the ferromagne- FeRh Cubic (type CsCl) AF - F 350 53-55, 57, 58, 56,
59 61 62 63 64
tism, which occur on the boundaries of this temperature 65, 66
interval, are first-order phase transitions. A possible Mn3GaC AF - F 150 67, 68, 69
example of such transitions are the magnetic trans- Cubic (perovskite type)
formations in Mn3Ge2, observed by Fakidov, Grazhdan- a-Fe2O3 Rhombohedral AF- F 250 97, 109
kina, and Novogrudskii[92] and investigated in193"983. A CrSM7 Hexagonal (type NiAs) AF - F 158 110, 111, 63, 112, 113
general qualitative verification of the theory was made
by Shimizu by analyzing the experimental data obtained MnSn2 Tetragonal AF- F 73 114
for MnAs. The experimentally observed strong decrease Mno,9Lio,!Se Cubic (NaCl type) AF- F 71 115, 116
of the temperature of the magnetic transformation of
MAGNETIC FIRST ORDER PHASE TRANSITIONS 743

validity of t h e i d e a s of Y o s i d a a n d Miwa170-1. Since at


Table HI (Cont'd.)
present we have far from complete data on the struc-
Temper-
ature of
Magnetic magnetic
ture of t h e e l e c t r o n i c energy spectrum of transition
Substance Crystal structure Transition transition, Literature metals and their alloys or c o m p o u n d s , it i s impossible

°K to draw any concrete conclusions concerning the influ-


ence of t h i s f a c t o r on the character of t h e magnetic
uy Hexagonal F-> H 85 (see review on rare- transformations. However the theories of Dzyaloshin-
earth ferro- and
anti-fercomagnets skii [ 7 1 ]
, Shimizu[9O>91], and Blum [ 7 3 ]
are sufficiently
["•"'I)
convincing examples demonstrating the great importance
of t h e t o p o l o g y o f t h e F e r m i s u r f a c e , of t h e d e n s i t y o f
Mn3Pt AF - AF 383 119
s t a t e s , a n d of t h e s p l i t t i n g of t h e e n e r g y l e v e l s i n t h e
v2o3 » AF -P 161 120, 121
crystalline magnetic fields to the magnetic transitions.
Ti,O3 AF ^P 473 122
In t h i s c o n n e c t i o n , it c a n b e a s s u m e d t h a t in s u b s t a n c e s
NiS » AF -P 263 124
w i t h a s m a l l v o l u m e d e p e n d e n c e of t h e e x c h a n g e i n t e g -
In the case when two transition temperatures are given, one of them (upper) cor-r a l s , t h e m a g n e t i c t r a n s i t i o n s a r e d e t e r m i n e d t o a c o n -
responds to magnetic measurements and the other (lower) is established with the aid ofs i d e r a b l e d e g r e e b y t h e i n d i c a t e d c h a r a c t e r i s t i c s . T h u s ,
thermal measurements. j
t h e e s s e n t i a l c o n d i t i o n f o r f u r t h e r d e v e l o p m e n t of o u r
Symbols: F — ferromagnetism or ferrimagnetism, AF — antiferromagnetism, P —
paramagnetism, H — helical magnetic structure, WF — weak ferromagnetism of the c o n c e p t s c o n c e r n i n g t h e n a t u r e o f m a g n e t i c t r a n s f o r m a -
Dzyaloshinskii-Morin type. tions is comprehensive and purposeful r e s e a r c h on the
physical properties of v a r i o u s systems with spin order-

tems, and that the physical causes of t h i s phenomenon ing.

a r e different. Most frequently magnetic first-order


transitions are observed in s u b s t a n c e s in which effects
1
of t h e s h i f t of t h e magnetic-transformation temperature C. Kittel, Phys. Rev. 120, 335 (1960).
2
with increasing pressure are sufficiently large. In this C. P. B e a n a n d D . S. R o d b e l l , Phys. Rev. 126, 104
connection, principal attention was paid in this review (1963).
3
to the exchange-striction mechanism of m a g n e t i c trans- Ising, Zs. Phys. 31, 253 (1925).
formations, which is based on the assumption that the "K. P. Belov, Magnitnye prevrashcheniya (Magnetic
exchange interaction depends strongly on the inter- Transformations), Fizmatgiz, 1959.
atomic distances. This mechanism was used in the 5
L. D. Landau and E. M. Lifshitz, Statisticheskaya
theory of B e a n a n d R o d b e l l , and also in Kittel's theory fizika (Statistical Physics), Gostekhizdat, 1940 [Addison-
for a quantitative treatment of m a g n e t i c transformations Wesley, 1958].
of t w o d i f f e r e n t classes: order-disorder type, connec- 6
S. V. Vonsovskrf, Izv. A N SSSR, s e r . fiz. 11, 485
ted with the destruction of t h e s p i n o r d e r i n g , and order- (1947).
order type, due to inversion of t h e e x c h a n g e interaction. 7
V. L. Ginzburg, Zh. Eksp. Teor. Fiz. 17, 833 (1947).
The agreement between the experimental results and 8
V. K. S e m e n c h e n k o , ZhFKh ( J . of P h y s . Chem.) 25,
the main conclusions of t h e i n d i c a t e d t h e o r i e s indicate 121 (1951); Izbrannye glavy teoreticheskoi fiz. (Selected
that this mechanism is realized in a l a r g e n u m b e r of Chapters of T h e o r e t i c a l Physics), Prosveshenie, 1966.
magnetic systems that differ from one another in their 9
L. Tiza, Phase Transformations in Solids, N. Y.,
crystal a n d m a g n e t i c s t r u c t u r e s , i n t h e c h a r a c t e r of t h e ex- 1951.
change interactions, etc. Experimental investigations of 1 0
V. L. Ginzburg, Fiz. Tverd. Tela 11, 2031 (1960)
the elastic properties and effects of t h e s h i f t of t h e m a g -
[sic!].
netic-transition temperature with increasing pressure U
O . K. R i c e , J. Chem. Phys. 22, 1535 (1954).
are of p r i n c i p a l significance for a verification of the 1 2
D . S. R o d b e l l and C. P. Bean, J. Appl. P h y s . Suppl.
initial p r e m i s e s of t h e c o n s i d e r e d theories. Besides
33, 1037 (1962).
neutron-diffraction investigations, great importance 13
R. A. Farrell and P. H. E. Meijer, Physica 31, 725
attaches also to m e a s u r e m e n t s of t h e m a g n e t i c proper-
(1965).
ties in s t r o n g magnetic fields, particularly for order- 1 4
S. K. Banerjee, Phys. Lett. 1 2 , 17 (1964).
order transitions. These investigations yield important 1 5
R. W. DeBlois a n d D . S. R o d b e l l , Phys. Rev. 130,
information concerning the magnitudes of t h e internal
1347 (1963).
effective fields due to the magnetic and exchange inter- 1 6
C. Guillaud, J. de phys. 12, 223 (1951).
actions. A study of t h e c h a r a c t e r of t h e f o r c e s causing 1 7
A. S e r r e s , J. p h y s . et r a d i u m . 5, 146 (1947).
the magnetic ordering helps explain the causes of the 1 8
R. O. Kornelsen, Canad. J. Phys. 39, 1728 (1961).
change of t h i s order, and determine the true nature of 1 9
A. J. P. Meyer and P. Taglang, Compt. Rend. 246,
the magnetic transformation.
1820 (1958).
Magnetic transition models170"78'90'913 not connected 2 0
G. E. B a c o n and R. Street, Nature 175, 518 (1955).
2 1
with the exchange-striction mechanism are of undoubted N. P. Grazhdankina and A. M. Burkhanov, Zh. Eksp.
interest, and in m a n y cases the factors considered in Teor. Fiz. 50, 1519 (1966) [Sov. P h y s . - J E T P 23, 1013
these models apparently exert a definite influence on the (1966)].
2 2
character of t h e m a g n e t i c t r a n s f o r m a t i o n s . Thus, for J. Zacharias, Phys. Rev. 44, 116 (1933).
23
example, the experimentally established influence of the B. T. M. Willis and H. P . Rooksby, Proc. Phys. Soc.
magnetic crystallographic anisotropy on the magnetic- B67, 290 (1954).
t 9 6 ] 2 4
transition in t e m p e r a t u r e MnAs and in Mn 2 _ x Cr x Sb D . S. Rodbell, Proc. Inter. C o n f . of V e r y High
t 4 4 ]
alloys s e r v e to a certain degree as a proof of the P r e s s u r e Res., Lake George, N. Y., 1960.
744 N. P . GRAZHDANKINA
S 59
T . Kamigaichi, K. Masumoto and T. Hihara, J. Sci. G. Shirane, C. Cheen and P. Flinn, J. Appl. Phys.
Hiroshima Univ. A2, 29, 53 (1965). 34, 1044 (1963).
26 60
N. P. Grazhdankina and Yu. S. Bersenev, Zh. Eksp. L. Muldawer and F. de Bergevin, J. Chem. Phys. 35,
Teor. Fiz. 51, 1052 (1966) [Sov. Phys.-JETP 24, 702 1904 (1961).
61
(1967)]. E. M. Hofer and P. Cucka, J. Phys. Chem. Sol. 27,
27
D. S. Rodbell (private communication). 1552 (1966).
28 62
R. W. DeBlois and D. S. Rodbell, J. Appl. Phys. 34, J. S. Kouvel, J. Appl. Phys. 37, 625 (1966).
63
1101 (1963). R. B. Flippen and F. J. Darnell, J. Appl. Phys. 34,
29
A. J. P. Meyer and P. Taglang, J. phys. et radium 1094 (1963).
64
14, 82 (1953). D. Bloch, Ann. de Phys. 1, 93 (1966).
30 65
N. P. Grazhdankina, Zh. Eksp. Teor. Fiz. 33, 1524 E. A. Zavadskii and I. G. Fakidov, Fiz. Tverd. Tela
(1957) [Sov. Phys.-JETP 6, 1178 (1958)]. 9, 139 (1967) [Sov. Phys.-Solid State 9, 103 (1967)].
31 66
N. P. Grazhdankina and D. I. Gurfel', Zh. Eksp. A. I. Zakharov, A. M. Kadomtseva, R. Z. Levitin,
Teor. Fiz. 35, 907 (1958) [Sov. Phys.-JETP 8, 631 and E. G. Ponyatovskii, Zh. Eksp. Teor. Fiz. 46, 2003
(1959)]. (1964) [Sov. Phys.-JETP 19, 1348 (1964)].
32 67
N. P. Grazhdankina and Yu. S. Bersenev, Zh. Eksp. J. P. Bouchaud and R. Fruchart, Compt. Rend. 261,
Teor. Fiz. (1968) [sic!]. 458 (1965).
33 68
N. P. Grazhdankina, L. G. Gafdukov and K. P. J. P. Bouchaud, R. Fruchart, R. Pauthenet, M.
Rodionov, Zh. Eksp. Teor. Fiz. 40, 434 (1961) [Sov. Guillot, H. Bartholin and F. Chaisse, J. Appl. Phys. 37,
Phys.-JETP 13, 297 (1961)]. 971 (1966).
34 69
N. P. Grazhdankina, AN SSSR, ser. fiz. 30, 990 M. Guillot and R. Pauthenet, Compt. Rend. 263,
(1966). B-527 (1966).
35 70
J. S. Smart, Phys. Rev. 90, 55 (1953). K. Yosida and H. Miwa, J. Appl. Phys. 32, 85 (1961).
36 71
L. Neel, Compt. Rend. 242, 1824 (1956). 1. E. Dzyaloshinskii, Zh. Eksp. Teor. Fiz. 47, 336
37
G. W. Pratt, Phys. Rev. 108, 1233 (1957). (1964) [Sov. Phys.-JETP 20, 223 (1965)].
38 72
S. H. Iiu, D. Behrendt and S. Legyoldt, Phys. Rev. B. V. Karpenko and A. A. Berdyshev, Fiz. Tverd.
116, 1464 (1959). Tela 5, 3397 (1963) [Sov. Phys.-Solid State 5, 2494
39
H. S. Jarrett, Phys. Rev. A134, 942 (1964). (1964)].
40 73
W. A. Doerner and R. B. Flippen, Phys. Rev. 137, M. Blum, Phys. Rev. 141, 517 (1966).
74
A927 (1965). L. N. Bulaevskii and D. I. Khomskii, Fiz. Tverd.
41
W. H. Cloud and H. S. Jarrett, Phys. Rev. 120, 1969 Tela 9, 3070 (1967) [Sov. Phys.-Solid State 9, 2422
(1960). (1968)].
42 75
A. E. Austin, E. Adelson and W. H. Cloud, Phys. R. Mitsudo, K. Motizuki and T. Nagamiya, J. Phys.
Rev. 131, 1511 (1963). Soc. Japan 20, 710 (1965).
43 76
T. J. Swoboda, W. H. Cloud, T. A. Bither, M. S. D. Perm and M. Cohen, Phys. Rev. 155, 468 (1967).
Sadler and H. S. Jarrett, Phys. Rev. Lett. 4, 509 (1960). " A . Arrott, S. A. Werner and H. Kendrick, Phys.
44
F. J. Darnell, W. H. Cloud and H. S. Jarrett, Phys. Rev. Lett. 14, 1022 (1965).
ra
Rev. 130, 647 (1963). D. S. Rodbell, I. S. Jacobs, J. Owen and E. A.
4S
P. E. Bierstedt, F. J. Darnell, W. H. Cloud, R. B. Harris, Phys. Rev. Lett. 11, 10 (1963).
79
Flippen and H. S. Jarret, Phys. Rev.'Lett. 8, 15 (1962). K. P. Belov, M. A. Belyanchikova, R. Z. Levitin
48
P. E. Bierstedt, Phys. Rev. 132, 669 (1963). and S. A. Nikitin, Redkozemel'nye ferro-i antiferro-
47
H. S. Jarrett, P. E. Bierstedt, F. J. Darnell and M. magnetiki (Rare-earth Ferro- and Antiferromagnets),
Sparks, J. Appl. Phys. Suppl. 32, 87 (1961). Nauka, 1965.
48 80
H. S. Jarrett, W. H. Cloud, F. J. Darnell, P. E. P. W. Anderson, Solid State Phys. 14, 99 (1966).
81
Bierstedt, T. A. Bither, P. H. Walter and T. J. Swoboda, E. A. Harris and J. Owen, Phys. Rev. Lett. 11, 9
J. Phys. Soc. Japan 17, Bl (1962); Proc. Inter. Conf. of (1963).
82
Magnetism and Crystallography, 1961. Nai Li Huang and R. Orbach, Phys. Rev. Lett. 12,
49
W. H. Cloud, T. A. Bither and T. J. Swoboda, J. 275 (1964).
83
Appl. Phys. Suppl. 32, 55 (1961). B. C. Frazer, G. Shirane, D. E. Cox and C. E.
50
R. B. Flippen and F. J. Darnell, J. Appl. Phys. 34, Olsen, Phys. Rev. 140, A1448 (1965).
84
1094 (1963). O. G. Brand and C. T. Walker, Phys. Rev. Lett. 18,
51
N. P. Grazhdankina, Zh. Eksp. Teor. Fiz. 47, 2027 11 (1967).
85
(1964) [Sov. Phys.-JETP 20, 1362 (1965)]. A. M. Burkhanov, N. P. Grazhdankina and I. G.
52
R. B. Flippen, J. Appl. Phys. 34, 2026 (1963). Fakidov, Fiz. Tverd. Tela 9, 748 (1967) [Sov. Phys.-
63
M. Fallot, Compt. Rend. 205, 558 (1937). Solid State 9, 586 (1967)].
54 86
M. Fallot, Ann. de Phys. 10, 231 (1938). N. P. Grazhdankina, Zh. Eksp. Teor. Fiz. 52, 397
55
M. Fallot and R. Hocart, Rev. Sci. 8, 498 (1938). (1967) [Sov. Phys.-JETP 25, 258 (1967)].
56 87
F. de Bergevin and L. Muldawer, Compt. Rend. 252, Ya. I. Frenkel', Z. Physik 49, 31 (1928).
88
1347 (1962). J. Friedel, G. Leman and S. Olszewski, J. Appl.
57
F. Bertaut, F. de Bergevin and G. Roult, Compt. Phys. 32, 325 (1961).
B9
Rend. 256, 1698 (1963). E. C. Stoner, Proc. Roy. Soc. A154, 656 (1936);
58
F. Bertaut, A. Delapalm, F. Forrat, G. Roult, F. Proc. Roy. Soc. A165, 372 (1938).
de Bergevin and R. Pauthenet, J. Appl. Phys. Suppl. 33, " M . Shimizu, Proc. Phys. Soc. 84, 397 (1964).
91
1123 (1962). M. Shimizu, Proc. Phys. Soc. 86, 147 (1965).
MAGNETIC FIRST ORDER PHASE TRANSITIONS 745
92 112
1. G. Fakidov, N. P. Grazhdankina, and V. N. Novo- K. Dwicht, R. W. Germann, N. Menyuk and A. Wold,
grudskii, AN SSSR, ser. fiz. 20, 1509 (1956). J. Appl. Phys. Suppl. 33, 1341 (1962).
93 113
S. D. Margolin and I. G. Fakidov, Fiz. Metal. T. Kamigaichi, T. Okamoto, N. Iwata and E. Tatsu-
Metallov. 7, 157 (1957). moto, J. Phys. Soc. Japan 21, 2730 (1966).
94 114
R. Fontaine and R. Pauthenet, Compt. Rend. 254, J. S. Kouvel and C. C. Hartelius, Phys. Rev. 123,
650 (1962). 124 (1961).
95 115
E. A. Zavadskiiand I. G. Fakidov, Zh. Eksp. Teor. S. J. Pikart, R. Nathans and G. Shirane, Phys. Rev.
Fiz. 51, 1317 (1966) [Sov. Phys.-JETP 24, 887 (1967)]. 121, 707 (1961).
96 116
R. W. deBlois and D. S. Rodbell, J. Appl. Phys. 44, R. R. Heikes, T. R. Cuire and J. Happel, Phys. Rev.
1101 (1961). 121, 703 (1961).
97 117
R. A. Voskanyan, R. Z. Levitin and V. A. Shurov, D. R. Behrendt, S. Legvold and F. H. Spedding, Phys.
Zh. Eksp. Teor. Fiz. 53, 459 (1967) [Sov. Phys.-JETP Rev. 109, 1544 (1959).
118
26, 302 (1968)]. M. K. Wilkinson, W. C. Kochler, E. O. Wollan and
98
F. Heusler, Z. angew. Chem. 1, 260 (1904). J. Cable, J. Appl. Phys. 32, 485 (1961).
99 U9
L. F. Bates, Phil. Mag. 8, 714 (1929). E . Kren, G. Kadar, L. Pal, J. Solyon and P. Szabo,
100
A. J. Heeger, O. Becman and A. M. Portis, Phys. Phys. Rev. Lett. 20, 331 (1966).
120
Rev. 123, 1652 (1961). T. Shinjo, K. Kosuge, S. Kachi, H. Takaki, M. Shiga
101
A. Okazaki and Y. Suemune, J. Phys. Soc. Japan 16, and Y. Nakamura, J. Phys. Soc. Japan 21, 193 (1966).
m
671 (1961). M . Foex and M. P. Lebeau, Compt. Rend. 223, 1126
102
M. Yuzuri, J. Phys. Soc. Japan 15, 2007 (1960). (1946).
103 122
S. Ogawa, J. Phys. Soc. Japan 15, 1901 (1960). M. Foex and J. Loriers, Compt. Rend. 226, 901
104
F. O. Wollan, W. C. Kochler and M. K. Wilkinson, (1948).
123
Phys. Rev. 110, 638 (1958). 1. E. Dzyaloshinskii, Zh. Eksp. Teor. Fiz. 32, 1547
105
W. C. Kochler, J. Phys. et radium 20, 180 (1959). (1957) [Sov. Phys.-JETP 5, 1259 (1957)].
106 124
P. L. Richards, J. Appl. Phys. 34, 1237 (1963). J. T. Sparks and T. Komoto, J. Appl. Phys. 34, 1191
107
M. Trombe, J. de phys. 12, 170 (1951). (1963), Intern Congress on Magnetism, USA, Boston
108
E. E. Huber and D. H. Ridglay, J. Appl. Phys. 34, (1967).
125
1099 (1963). J. B. Goodenough and J. A. Kafalas, Phys. Rev. 157,
109
T. Kaneko and S. Abe, J. Phys. Soc. Japan 21, 451 389 (1967).
(1966).
110
H. Haraldsen and A. Neuber, Zs. anorg. allg. Chem.
234, 337 (1937).
111
N. P. Grazhdankina and I. G. Fakidov, Izv. AN SSSR,
ser. fiz. 21, 1116 (1957)]. Translated by J. G. Adashko

Potrebbero piacerti anche