Invisibility Physics Past Present and Future 2013 Progress in Optics

CHAPTER TWO
Invisibility Physics: Past, Present,

and Future
Greg Gbur
Department of Physics and Optical Science, University of North Carolina at Charlotte, Charlotte,
NC 28223, USA
Contents
1. Introduction 65
2. Early Concepts of Invisibility 67
3. Nonscattering Objects and Inverse Problems 75
4. Optical Cloaking 82
5. Other Cloaking Designs 90
6. Experiments and Material Properties 94
7. Related Applications 97
8. Invisibility in Science Fiction 100
9. Concluding Remarks 106
Acknowledgments 107
References 107
1. INTRODUCTION
The idea of invisibility has captured the imagination of humanity since
ancient times. Though the earliest visions of invisibility required magical
powers and artifacts bestowed by the gods, by the mid-19th century authors
of science fiction began to devise physical,if fanciful,explanations of how one
might “hide from light.”Around nearly the same time, physicists speculating
about the nature of the atom began unknowingly laying the groundwork
for invisibility physics. These early investigations led directly into invisibility
in the context of inverse scattering and eventually to the modern concepts
of optical cloaking.
In this chapter, we will review the physics of invisibility, and consider
developments from the past, present, and possible future. Far from being
a scientific novelty, the physics and mathematics of being unseen suggests
many possible applications in a variety of fields, including acoustics, imaging,
seismology, and quantum mechanics.
Progress in Optics, Volume 58 © 2013 Elsevier B.V.
ISSN 0079-6638, http://dx.doi.org/10.1016/B978-0-444-62644-8.00002-9 All rights reserved. 65
66 Greg Gbur
Figure 1 A demonstration of the “optical camouflage” system. (Photographs courtesy of

Susumu Tachi, The University of Tokyo).
Before we begin, however, we must be a little clear about the word

“invisibility,” whose colloquial usage can be much different than its scien-
tific one. A good illustration of this is provided in Figure 1, which shows a
demonstration of “optical camouflage” that was developed and introduced
by researchers at Tokyo University in 2003 (Tachi, 2003). The system con-
sists of a retroreflective coat worn by the target to be hidden, combined
with a camera recording the scene behind the target and transmitting it to a
projector in front. The projector displays the background scene on the coat,
giving the illusion of transparency. A similar technique was recently used in
2012 in a commercial for Mercedes-Benz: an “invisible car” was created by
recording the scene on one side of the car with a camera and displaying it
on the other side with an array of LEDs (Mercedes-Benz, 2012).
Though both of these systems were hyped as “invisible” in the media,
the transparency of the targets is not perfect, and only looks realistic from a
narrow range of viewing angles—the targets are far from unseen. This is a
recurring theme in the various invisibility schemes to be considered: some
work only for a limited range of frequencies, some for a limited range of
viewing angles, some possess unavoidable reflections, and most are only an
approximate form of the phenomenon as seen in popular science fiction.
With such a variety of schemes for invisibility (and a variety of associated
limitations),we take in this article the broadest definition of the term possible:
an object is referred to as invisible if it does not radiate or scatter waves under
circumstances in which it would be conventionally expected to do so. This
excludes things that are technically invisible to the naked eye (many gases
and individual atoms, for instance) but includes a lot of unusual physical and
optical phenomena.
In spite of serious restrictions, the invisibility schemes to be considered
have found many surprising applications.TheTokyo camouflage,for instance,
Invisibility Physics: Past, Present, and Future 67
is being applied to improving the field-of-view of automobile drivers and

airplane pilots,by making the interior of the transport effectively see-through
(Hornyak, 2012).
We begin this article by considering some of the earliest concepts of
invisibility physics in Section 2, primarily so-called “nonradiating sources.”
In Section 3, we look at the related concept of “nonscattering scatterers”
and their relation to inverse scattering problems. In Section 4, we discuss
the foundational cloaking papers of 2006 and the principles of transforma-
tion optics used to design them. In Section 5, we look at other varieties
of electromagnetic cloaking devices that have been introduced. Section 6
looks at the first experimental demonstrations of cloaking and discusses the
optical materials needed for such devices. The application of cloaking con-
cepts to fields outside of electromagnetics is considered in Section 7. Science
fiction has curiously anticipated many ideas now being introduced in invisi-
bility physics; in Section 8 we compare the early imaginings of authors with
the technology being developed today. Section 9 presents some concluding
remarks.
2. EARLY CONCEPTS OF INVISIBILITY

The first scientific discussion of the invisibility of a macroscopic object
seems to be due to Rayleigh (1884). In a curious footnote in an Ency-
clopaedia Britannica article on geometrical optics, Rayleigh suggested the
following:
Even when the optical differences are not small it is well to remember that trans-
parent bodies are only visible in virtue of a variable illumination. If the light falls
equally in all directions, as it might approximately do for an observer on a high
monument during a thick fog, the edge of (for example) a perfectly transparent
prism would be absolutely invisible. If a spherical cloud, composed of absolutely
transparent material, surround symmetrically a source of light, the illumination at
a distance would not be diminished by its presence.
Without proof, Lord Rayleigh introduced an unusual proposition: a non-

absorbing object is only visible by virtue of the nonuniformity of illumi-
nation upon it. If the object is equally illuminated from all sides, it should
effectively disappear from view.
Rayleigh provides no proof of this conjecture, which is likely true only
under special circumstances, if at all. His footnoted comment went unre-
marked for a number of years, until the physicist R.W. Wood endeavored
to test Rayleigh’s hypothesis with a simple experiment (Wood, 1902). Wood
68 Greg Gbur
+ −
+ − + − − −
Dynamid model Plum pudding model −
+ − − −
P. Lenard (1903) + − + − J.J. Thomson (1904)
−
+ − + − − −
+8
−
+
− − −
+
−
+
−
−
+
Saturnian model Archion model +
− +8 −
H. Nagaoka (1904) J. Stark (1910) −
− −
−
+
− −
+
−
+
Figure 2 Illustration of early incorrect models of the atom.
coated the inside of a pair of hemispheres with a luminous paint and drilled
a small hole in one of the hemispheres for observation. Wood found that
a “crystal ball or the cut glass stopper of a decanter” appeared to disappear
from view when sealed between the hemispheres.
Wood’s experiment, though intriguing, also failed to attract much atten-
tion. At about the same time, however, many physicists were occupied by
a very important and unsolved problem, namely the structure of the atom.
By 1903, there were only a few tantalizing clues available to researchers: the
presence of electrons, the structure of the periodic table, radioactivity, and
the Balmer and Rydberg formulas for atomic spectra. Some of the most
preeminent scientists of the time published their speculations on atomic
structure, most notably J.J. Thomson with his “plum pudding” model of
the atom (Thomson, 1904). Other examples include P. Lenard’s “dynamid”
model, which postulated that electrons in the atom are bound with positive
charges in pairs called dynamids (Lenard, 1903), H. Nagaoka’s “Saturnian”
model (Nagaoka, 1904), and J. Stark’s “Archion” model, which imagined that
atoms consist of magnetic dipoles arranged in rings, with electrons balancing
the positive charges (Stark, 1910). These models are illustrated in Figure 2.
All of these ideas possessed inherent difficulties. James Jeans performed
a dimensional analysis encompassing most hypotheses and astutely noted
that one could not derive a quantity with units of frequency from them,
suggesting that new physics was needed to explain the atom (Jeans, 1906).
GeorgeAdolphus Schott attempted to introduce this frequency by suggesting
that electrons were extended objects constantly trying to grow in size, and
only held back by the pressure of the electromagnetic “aether.” Curiously,

Schott’s assumptions resulted in an unexpected attraction between electrons
that he interpreted as the gravitational force (Schott, 1906).
The most obvious problem with these early atomic models, however,
was that they all included electric charges in accelerated motion (electrons
and unnamed positive charges). It was conventional wisdom in that era that
accelerating charges necessarily produce radiation; simple calculations indi-
cated that all of the classical models described would collapse quickly in a
burst of released energy.
In 1910, Paul Ehrenfest proposed one possible solution to this conun-
drum, by pointing out that there exist extended distributions of charge that
can accelerate without radiation under the right circumstances (Ehrenfest,
1910). He first noted that, by symmetry, an infinite, uniformly charged plane
that oscillates normal to its surface can only have an electric field and radi-
ation in the same normal direction; however, since the electric field must
be perpendicular to the direction of radiation, this system cannot radiate.
Similarly, a uniform sphere of charge that pulsates in the radial direction can
only have a radial electric field and radially outgoing radiation, and therefore
also cannot radiate. In both cases, the magnetic field must also point outward;
because ∇ · H = 0, this implies that H = 0 identically.
Ehrenfest then went further and provided a prescription for construct-
ing a general nonradiating distribution of charges and currents, which we
introduce as a theorem.
Theorem 2.1. Consider a charge distribution localized to a volume A
which possesses both a time-varying scalar potential φ(r, t) within A such
that ∇ 2 φ is nonsingular, and a time-independent scalar potential φ(r) in
the region B outside of A. Further assume that the potential φ(r, t) and its
gradient ∇φ(r, t) are continuous across the boundary between A and B, and
that H = 0 everywhere. Maxwell’s equations are then automatically satisfied,
and the charge distribution does not produce any radiation.The field, charge
and current may be found from the potential by
∂
E(r, t) = −∇φ(r, t), ρ(r, t) = −0 ∇ 2 φ(r, t), J(r, t) = 0 [∇φ(r, t)].
∂t
(2.1)
Ehrenfest’s intriguing observations seem to have been quickly forgot-

ten after Bohr developed his quantum model of the atom (Bohr, 1913),
which indirectly resolved the radiation problem. However, the aforemen-
tioned G.A. Schott remained convinced that the curious properties of atoms
70 Greg Gbur
could be explained by the use of a purely classical formalism. In 1933, Schott

published a remarkable result demonstrating that a uniform spherical shell
of charge can move in a periodic orbit without producing radiation (Schott,
1933).
The elegance of Schott’s work merits a description, though we follow
closely a later derivation by Goedecke (1964). We consider a rigid three-
dimensional distribution of total charge e and translational motion charac-
terized by the vector ξ (t). The charge density may be written in the form
ρ(r, t) = ef [r − ξ (t)],
where f [r] is the distribution of charge within the object, with volume
integral equal to unity.The current density of the distribution may be readily
written as
J(r, t) = e ξ̇ (t) f [r − ξ (t)]. (2.2a)
In the far zone of the moving charge distribution, the scalar and vector
potentials can be expressed in the form

μ0 r − s · r 3
A(rs, t) ≈ J r ,t − dr, (2.2b)
4πr c

1 r − s · r 3
φ(rs, t) ≈ ρ r ,t − dr, (2.2c)
4π0 r c
where s is the unit vector pointing in the direction of observation at distance
r and c is the speed of light.
If we assume that the distribution is moving periodically with period T
and is of finite support and thus restricted to a finite volume V , we may
expand the current and charge densities in a mixed Fourier integral/series
representation,
∞
J(r, t) = Jn (K)e i(K·r−ωn t) d3 K , (2.3a)
n=−∞
∞
1
ρ(r, t) = K · jn (K)e i(K·r−ωn t) d3 K , (2.3b)
n=−∞
ωn
with ωn = 2πn/T . On substitution from these expressions into
Equations (2.2), we find that the latter may be rewritten as
∞
(2π)3 μ0 e ikn r
A(rs, t) ≈ Jn (kn s) e −iωn t , (2.4a)
4π n=−∞ r
∞
(2π)3 e ikn r
φ(rs, t) ≈ s · Jn (kn s) e −iωn t , (2.4b)
4π0 c n=−∞ r
where kn ≡ ωn /c. The coefficients Jn (kn s) may be evaluated by the Fourier

relation,
T
1
Jn (K) = e −iK·r J(r, t)e iωn t d3 r dt, (2.5)
(2π)3 T V 0
which may be simplified via Equation (2.2a) to the form
T
e ˜
Jn (K) = f (K) e −iK·ξ (t) ξ̇ (t)e iωn t dt, (2.6)
T 0
where
1
f˜ (K) = f (r)e −iK·r d3 r. (2.7)
(2π)3 V
Using the standard relationship between the potentials and the electric field,
we finally arrive at the result
∞ T
(2π)3 e e ikn r

E(rs, t) = − ikn f˜ (kn s) e −iω n t
e −ikn s·ξ (t )
4π0 cT n=−∞ r 0

×(s × [s × ξ̇ (t )])e iωn t dt . (2.8)
The terms of this series depend upon the time-independent Fourier trans-
form f˜ (kn s) of the charge distribution. This leads to the following theorem,
due to Schott,
Theorem 2.2. A rigid charge distribution ef (r) undergoing periodic trans-
lational motion will not produce any radiation if the condition
f˜ (kn s) = 0 (2.9)
is satisfied for all nonzero integers n and for all directions of observation s,
independent of the precise path of the distribution.
Schott considered the special case of a uniformly charged sphere of radius
a, and demonstrated that this sphere will produce no radiation for a radius
a = mcT /2, where m is an integer. It is immediately apparent that the period
of oscillation is an integer multiple of the amount of time it takes for light to
cross the sphere’s diameter. This in turn indicates that the sphere’s range of
motion is limited by relativity to be much less than its diameter: the sphere’s
motion is more a “wobble” than an “orbit.”
72 Greg Gbur
Schott continued working on radiationless motions until his death, pro-

ducing a series of papers on orbiting and spinning charged spheres (Schott,
1936a, 1936b, 1937a, 1937b). The most significant development to come
from this later work was the observation that a charged sphere might also
self-oscillate without external force, according to the classical theory of radi-
ation reaction. Bohm and Weinstein continued this line of reasoning, not-
ing additional classes of approximately nonradiating distributions (Bohm
& Weinstein, 1948). Goedecke (1964) developed a more general theory
of radiationless motions, and derived examples of asymmetric spinning
distributions.
All of these works read, in a sense, like solutions in search of a problem.
Schott suggested that his radiationless orbits might provide a stable model of
the neutron. Bohm and Weinstein speculated that the muon might be a self-
oscillating excited state of the electron. Goedecke went even further,suggest-
ing that radiationless orbits might lead to a new “theory of nature” in which
all stable particles or aggregates are described as radiationless excitations.
Radiationless motion generated renewed interest in the context of the
so-called inverse source problem: the possible determination of the structure of
a source from its radiation pattern. If sources exist which do not produce any
radiation, then no information can be deduced about their structure and the
inverse problem is nonunique.The history of such nonradiating sources is much
too involved to be described here; we refer the reader to Gbur (2003) for a
detailed discussion and simply note the results and properties most relevant
to later discussion. Some of the earliest papers were those by Friedlander
(1973) and Bleistein and Cohen (1977) who looked at the properties of
nonradiating sources and their connection to the inverse problem; at about
the same time, Devaney and Wolf (1973) demonstrated the existence of
monochromatic nonradiating classical current distributions.
The problem is commonly formulated in the context of scalar monochro-
matic fields u(r) and sources q(r) oscillating at frequency ω, which satisfy an
inhomogeneous Helmholtz equation,
∇ 2 u(r) + k2 u(r) = −4πq(r), (2.10)
where k = ω/c. The source is assumed to be restricted to a domain D; the
notation is illustrated in Figure 3.
The solution to this equation may be found in terms of the Green’s
function of the Helmholtz equation in the form

e ik|r−r |
u(r) = q(r ) . (2.11)
D |r − r |
u(rs)
rs
q(r)
Figure 3 Illustration of the notation related to a monochromatic radiation source

and field.
In the far zone of the source (kr 1),this expression may be approximated as
e ikr
u(rs) ≈ (2π)3 q̃(ks), (2.12)
r
where
1
q̃(K) = q(r )e −iK·r d3 r . (2.13)
(2π)3 D
The function q̃(rs) is often referred to as the radiation pattern of the source.
Equation (2.12) leads to the following theorem.
Theorem 2.3. A source will be nonradiating, i.e., it will not produce any
power flow in the far zone of the source, if and only if
q̃(ks) = 0 for all directions s. (2.14)
It is to be noted that this theorem is essentially a monochromatic ver-
sion of Theorem 2.2 shown earlier. A simple example of such a source is a
homogeneous sphere of radius a. It was shown by Kim and Wolf (1986) that
such a sphere will not radiate if it satisfies the condition j1 (ka) = 0, where
j1 (x) is the first-order spherical Bessel function.
Another important theorem, first shown in its present form by
Friedlander (1973), is as follows:
Theorem 2.4. The field of a nonradiating source vanishes everywhere
outside the domain of the source.
If a source is nonradiating, it therefore produces no radiation anywhere
outside of the source domain. Finally, we note the following theorem by
Gamliel, Kim, Nachman, and Wolf (1989), useful in the mathematical con-
struction of nonradiating distributions.
74 Greg Gbur
Theorem 2.5. A bounded nonradiating source distribution q(r) of finite

support and the field u(r) that it generates are related by the inhomogeneous
Helmholtz equation, Equation (2.10), subject to the boundary conditions
u(r)|r∈S = 0, (2.15a)

∂u(r)
= 0, (2.15b)
∂n r∈S
where S is the boundary of the source domain and ∂/∂n denotes differen-
tiation along the outward normal.
This theorem is strikingly similar to Ehrenfest’s early electromagnetic
construction of radiationless sources, discussed earlier, and is essentially a
scalar version of it. Both theorems reach the same conclusion:a source will be
nonradiating if its time-varying fields go smoothly to zero on the boundary
of the source domain.
The nonradiating condition may be interpreted as a unique wave phe-
nomenon in which the field everywhere outside of a three-dimensional
source cancels via complete destructive interference; this is discussed in Kim
and Wolf (1986) as well as Gbur (2003).
In spite of the many years they have been studied, nonradiating sources
continue to reveal new surprising effects. It was already noted in very early
investigations that even partially coherent sources can be nonradiating; see,
for instance, Baltes and Hoenders (1978), Hoenders and Baltes (1979), and
Devaney and Wolf (1984). The existence of partially coherent nonradiat-
ing sources has in turn led to a number of unusual phenomena, includ-
ing highly incoherent sources that produce fully coherent fields outside the
source (Gbur & Wolf, 1997) and unpolarized sources that produce almost
completely polarized fields (Gbur & James, 2000).
Nonradiating phenomena are not restricted to three-dimensional prob-
lems. Berry, Foley, Gbur, and Wolf (1998) demonstrated that localized exci-
tations, now referred to as nonpropagating excitations, can be produced on a
vibrating string. A simple explanation of the effect was suggested soon after
by Denardo (1998). It has been shown that such excitations exist even in
the presence of more realistic conditions such as damping and finite band-
width driving forces (Gbur, Foley, & Wolf, 1999) and weak nonuniformity
(Denardo & Miller, 2003). The related existence of directional excitations
in one-dimensional wave problems was demonstrated by Essl (2004), and
general guidelines for construction of nonpropagating and directional exci-
tations was considered by Moses, Gan, and Gbur (2009). Analogous two-
dimensional nonpropagating waves were considered by Gbur (2001).
It was shown some time ago that primary planar sources in three-
dimensional space must necessarily be radiating (Friberg, 1978). More
recently, however, Devaney (2004) suggested that nonradiating surface
sources are possible, if one considers a broader class of singlet and doublet
surface elements. This latter observation has some relevance in the context
of cloaking, to be discussed soon.
3. NONSCATTERING OBJECTS AND INVERSE PROBLEMS

Perhaps the earliest attempt to make invisible objects, at least in a lim-
ited sense, was the attempt during the Second World War to make airplanes
undetectable by radar using a thin surface layer of high absorption and vac-
uum impedance. As Sommerfeld (1964) recalls,
During the war the problem arose to find, as a counter measure against allied
radar, a largely nonreflecting (“black”) surface layer of small thickness. This layer
was to be particularly nonreflecting for perpendicular or almost perpendicular
incidence of the radar wave.
The solution was analogous to the perfectly matched layers now used in
finite-difference time-domain electromagnetic simulations to absorb outgo-
ing waves and prevent spurious reflections (see, for instance, Sullivan, 2000.)
The impedance Z of a medium is given by the ratio of the permeability μ
to the permittivity , i.e.,
μ
Z≡ . (2.16)

The development of a minimally reflecting surface therefore requires match-
ing the impedance of the material to that of vacuum. In an interesting antic-
ipation of future cloaking schemes, this requires a material with a nontrivial
magnetic permeability.
A different solution was found by Kay and Moses (1956),who constructed
a class of gradient dielectric index layers that could perfectly transmit waves of
any direction of incidence, albeit only at a single frequency and polarization.
These results can also be applied to quantum mechanical potential scatter-
ing, and some time later Toyama, Nogami, and Zhao (1993) and Nogami
and Toyama (1998) looked at reflectionless potentials in the Dirac equation.
It was found that reflectionless scalar and pseudoscalar potentials could be
constructed, but not reflectionless vector potentials. Hellberg, Karlsson, and
Thärning (1992) have looked at the challenging problem of making a poten-
tial that is nonreflecting for an incident pulse or, equivalently, for a range of
frequencies.
76 Greg Gbur
The first three-dimensional “invisible bodies” were introduced by Kerker

(1975b), who noted that subwavelength-size compound dielectric ellipsoids
can be constructed which produce no scattered field.1 In Kerker’s scheme, an
outer shell of high permittivity surrounds an inner sphere of low permittivity.
The dipoles induced by an illuminating field in the shell and the sphere
are in opposing directions and, with the right choice of permittivity and
radii, the dipole moment can be made zero. Chew (1976) considered the
implications of this result in the Rayleigh–Gans approximation, and Chew
and Kerker (1976) noted that Kerker’s original condition leads to abnormally
low scattering cross-sections in larger spheres.
With the development of medical imaging techniques such as computed
tomography (CT) and magnetic resonance imaging (MRI),an understanding
of invisible objects acquired practical implications. The existence or nonex-
istence of such objects is of fundamental importance to the uniqueness of
inverse problems, which we now discuss in more detail.
Typical problems studied in physics represent the determination of an
“effect” produced by a given “cause,” and can be referred to as a direct problem.
For instance,a standard problem in antenna theory is to calculate the radiation
pattern (effect) produced by an oscillating system of currents (cause). An
inverse problem represents a reversal of this process: the determination of a
“cause” from a given “effect.” In the example of antenna theory, the inverse
problem would be deriving the distribution of currents in the antenna from
measurements of the radiation pattern it produces. Other examples of inverse
problems involve the determination of crystal structure from X-ray scattering
experiments (X-ray crystallography) and the determination of the internal
structure of the human body from X-ray absorption experiments (CT).
Even a crime scene investigation may be broadly imagined as an inverse
problem: discovering the cause of a crime (the criminal and his motive)
from measurements of the effect (the crime scene).
A large number of inverse problems have been found to be ill-posed, in a
sense introduced by Hadamard (1902). A well-posed problem has a solution
that exists, is unique, and is continuous, where ill-posed problems lack one or
more of these characteristics. Difficulties in ill-posed problems are illustrated
in Figure 4. An operator A converts a point in the object space into a point
in the image space; the inverse problem is represented by the inverse operator
A−1 that returns an image point to an object point. For a nonunique inverse
1 An erratum appeared in Kerker (1975a). Kerker’s original paper notes that the research “was
supported in part by the Paint Research Institute in connection with a numerical study of
the effect of microvoids upon the hiding power of paints.”
object space A image space

physical solutions nonuniqueness
(prior knowledge)
noncontinuity
noise result nonexistence
Figure 4 Illustration of various types of ill-posedness.
problem, a single image can represent multiple objects. Closely related is a

noncontinuous inverse problem, in which very similar images can represent
very different objects: any noise in the system can result in a wildly incorrect
object reconstruction. In an inverse problem suffering from nonexistence,
noise in the imaging process can push the resultant image outside of the
image space, making reconstruction impossible.
Strategies exist for coping with and minimizing problems associated with
ill-posedness; often prior knowledge of the physics of the problem is used to
restrict to class of possible objects and enforce uniqueness and continuity. A
detailed discussion of inverse problems is outside the scope of this article; we
refer the reader to Bertero and Boccacci (1998) or Devaney (2012) more
information. Important for our purposes is the observation that nonunique-
ness in an inverse problem implies the existence of invisible objects in the
direct problem, and vice versa. This is immediately evident in the inverse
source problem, in which the source structure q(r) is determined from the
radiated field u(r). If qNR (r) represents a nonradiating source, it is clear from
Equation (2.11) that the source given by the linear combination q(r)+qNR (r)
also results in the same radiated field u(r), and the problem is nonunique.
The possible existence of objects that do not scatter incident waves (orig-
inally referred to as nonscattering scatterers) is therefore of fundamental impor-
tance in the uniqueness of inverse scattering/absorption techniques such
as computed tomography (Kak & Slaney, 1988), magnetic resonance imag-
ing (Vlaardingerbroek & den Boer, 2003) or diffraction tomography (Wolf,
1996). Furthermore, the existence of nonscattering scatterers has a close
connection to the existence of nonradiating sources, as we now show.
We consider a scattering object of inhomogeneous refractive index n(r, ω)
localized to a domain D, and consider a monochromatic illuminating field
ui (r, ω) incident upon it; the object results in a scattered field us (r, ω). The
notation is illustrated in Figure 5. The total field u(r, ω) = ui (r, ω) + us (r, ω)
78 Greg Gbur
us (r)
ui (r)
D
n(r,ω)
Figure 5 Illustrating the scattering of a monochromatic field from an inhomogeneous

refractive index.
satisfies the generalized Helmholtz equation,
[∇ 2 + n2 (r, ω)k2 ]u(r, ω) = 0. (2.17)
Due to the nonconstant refractive index n(r, ω), this equation does not typ-
ically possess simple analytic solutions. However, we can write it in a more
useful form by adding k2 u to both sides of the equation, and bringing the
refractive index term to the right-hand side,
[∇ 2 + k2 ]u(r, ω) = k2 [1 − n2 (r, ω)]u(r, ω). (2.18)
Defining the scattering potential as
k2 [n2 (r, ω) − 1]
F(r, ω) ≡ , (2.19)
4π
we may then write
[∇ 2 + k2 ]u(r, ω) = −4πF(r, ω)u(r, ω). (2.20)
The incident field is taken to be the field that would exist throughout space
in the absence of the scattering object, and it therefore satisfies the homo-
geneous Helmholtz equation,
∇ 2 ui + k2 ui = 0. (2.21)
Equation (2.20) then becomes an inhomogeneous equation for the scattered

field,
∇ 2 us + k2 us = −4πFu. (2.22)
On comparison of this equation to those for radiation, Equations (2.10) and

(2.11), we find that the scattered field may be written as

e ik|r−r |
us (r, ω) = F(r , ω)u(r , ω) . (2.23)
D |r − r |
It is to be noted that this equation cannot be easily solved, as the scattered
field appears implicitly in the integral, in the total field u(r, ω). However,
the expression is formally the same as the radiated field for a primary source,
Equation (2.11), and we may make the following statement.
Theorem 3.1. For a given incident field ui (r, ω), a scattering object acts as
a primary radiation source with a source function q(r, ω) = F(r, ω)u(r, ω).
This may be understood physically by noting that a field incident upon a
scatterer excites oscillating dipoles within it, and this collection of dipoles act
as a primary source. If we further continue the analogy between sources and
scatterers,we obtain the following theorem by a straightforward modification
of Theorem 2.14.
Theorem 3.2. An object with a scattering potential F(r, ω) will be non-
scattering for a given incident field ui (r, ω) if

F(r , ω)u(r , ω)e iks·r d 3 r = 0 for all s, (2.24)
D
where u = us + ui .
Because of the presence of us on both sides of Equation (2.23), it is
difficult to assess the existence or nonexistence of nonscattering scatterers.
In the special case that the scattering potential is extremely weak, however,
we may then approximate u(r, ω) in Equation (2.23) by ui (r, ω), in what is
known as the first Born approximation (Born & Wolf, 1999, Section 13.1.2).
A weakly scattering object will then be nonscattering if

F(r , ω)ui (r , ω)e ik r̂·r d 3 r = 0 for all r̂. (2.25)
D
It is therefore in principle possible to construct an object which is non-

scattering for a single direction of incidence. This raised the question of
whether it is possible to construct a perfectly nonscattering object. Devaney
(1978) demonstrated that potentials can be constructed which are nonscat-
tering for a finite number of incident directions within the context of the
first Born approximation; this result was later extended to problems of diffu-
sion and radiative transport by Hoenders (1997). However, it was also shown
80 Greg Gbur
in two independent articles that an object that is perfectly nonscattering for

all directions of incidence.We will explicitly prove the simpler result byWolf
and Habashy (1993).
Theorem 3.3. Within the accuracy of the first Born approximation, there
is no medium occupying a finite region of space which is a nonscattering
scatterer for all directions of incidence.
We will consider incident fields in the form of plane waves,
ui (r, ω) = A0 e iks0 ·r , (2.26)
for all possible directions s0 of illumination. On substitution of this result

into Equation (2.25), we find that the scattered field in the far zone is of the
form
e ikr
us (rs, ω) = (2π)3 F̃[k(s − s0 ), ω], (2.27)
r
where F̃(K, ω) is the three-dimensional spatial Fourier transform of the
scattering potential. For a fixed direction of incidence s0 , measurements
of the field in the far zone of the scatterer for all possible directions of
scattering s can be used to determine those components of F̃(K, ω) that
lie on a sphere of radius k, centered on K = ks0 , as shown in Figure 6.
This sphere, originally used in the theory of X-ray scattering by crystals, is
known as the Ewald sphere of reflection. For the scatterer to be nonscattering for
all directions of incidence, the Fourier transform of F(r, ω) must therefore
vanish on Ewald spheres of reflection for every direction of incidence and
ks
—ks0
ΣL
Figure 6 Illustrating the Ewald spheres of reflection (dashed) and the Ewald limiting
sphere, L .
must consequently vanish within a sphere of radius |K| ≤ 2k, known as the
Ewald limiting sphere. However, because F̃(K, ω) is the Fourier transform of
a function of finite support, it is the boundary value of an entire analytic
function in three complex variables (see Fuks, 1963, p. 352). It follows then
that F̃(K, ω) cannot vanish over any three-dimensional region of K-space
unless it vanishes identically—in which case F(r, ω) is itself identically zero.
This theorem was also proven in an exact form by Nachman (1988), who
demonstrated the general uniqueness of the inverse scattering problem. We
state his result as a theorem for future reference.
Theorem 3.4. We consider the class of inhomogeneous, isotropic scatter-
ing objects of finite conductivity and finite extent. Perfectly nonscattering
objects of this class do not exist.
The nonexistence of nonscattering scatterers implies the uniqueness of
inverse scattering problems: by probing an object with illumination from
a sufficiently large number of incident directions and measuring the field
for a sufficiently large number of scattering directions, an accurate image of
the object can be constructed. However, the theorem by Nachman, though
correct and of broad significance, does not encompass all possible scatter-
ing materials, leaving the possibility of “loopholes” in the theory allowing
invisibility.
One such loophole was introduced relatively early in scattering theory,
when Alexopoulos and Uzunoglu (1978) observed that active spherical scat-
terers (ones with a negative imaginary part of the refractive index, implying
gain) can have a vanishing extinction cross-section, implying that the inci-
dent field is unaffected. However, Kerker (1978) pointed out that, although
the extinction cross-section may be small for such objects, the scattering
cross-section will still be quite large. In such cases, the loss of the incident
field due to scattering is perfectly balanced by its amplification in the gain
medium. Recently, a number of authors (Arslanagic & Ziolkowski, 2010,
2011) have demonstrated that a lossy particle surrounded by a shell of gain
medium can possess significant transparency.
Such “invisible” gain particles are strikingly similar to the properties
of some recent optical systems exhibiting parity-time (PT) symmetry. First
discussed by Bender and Boettcher (1998) in the context of quantum mechan-
ics, a PT-symmetric system in optics is one which possesses alternating
regions of gain and loss with particular values; Berry (2008) and Makris,
El-Ganainy, Christodoulides, and Musslimani (2008) produced some early
work on the optical case. In recent years, Lin et al. (2011) demonstrated
82 Greg Gbur
computationally that certain PT-symmetric photonic lattices can exhibit uni-

directional invisibility, with perfect transmission but strong scattering for the
opposite direction of illumination. Longhi (2011) pointed out that this invis-
ibility can break down for sufficiently thick lattices. Soon after,Regensburger
et al. (2012) experimentally demonstrated the unidirectional invisibility of
such a lattice.
4. OPTICAL CLOAKING
The study of invisibility physics was reinvented and reinvigorated with
the publication of two back-to-back papers in Science, by Leonhardt (2006b)
and Pendry, Schurig, and Smith (2006). These articles theoretically demon-
strated that it is not only possible to construct an invisible object, but that it
is possible to create an invisibility cloak that hides an arbitrary object within
an interior cloaked region. Simulations of the effect of these cloaks on light
rays are shown in Figure 7.
Both of these cloak designs are based on the observation that light prop-
agating in a vacuum region of curved space is mathematically analogous to
light propagating in a medium with gradient material properties in Cartesian
Figure 7 The behavior of light rays in (a) the Leonhardt cloak, for two different directions
of incidence and (b) the Pendry–Schurig–Smith cloak. (From Leonhardt (2006b) and
Pendry et al. (2006), respectively) Reprinted with permission from AAAS.
space. The tools of differential geometry may therefore be used to design

regions of space with a desired effect on light (such as cloaking) which
can then be reverse-engineered into a material structure with the necessary
optical properties. This strategy of optical design is now referred to as trans-
formation optics, and has been key in the development of a variety of different
cloaking schemes.
For a detailed discussion of transformation optics, we refer the reader
to the review article by Leonhardt and Philbin (2009) or the book by the
same authors, Leonhardt and Philbin (2010). Here we briefly introduce the
subject by considering the mathematics used in the groundbreaking cloak
designs.
Leonhardt (2006b) considered the propagation of monochromatic scalar
waves in two dimensions satisfying the Helmholtz equation,
[∂x2 + ∂y2 + n2 (r)k2 ]u(r) = 0, (2.28)
where ∂x represents the partial derivative with respect to x, and so forth.

Initially restricting our attention to geometrical optics, we assume that the
refractive index varies slowly over the scale of a wavelength λ = 2π/k. Using
the complex variable z = x+iy,the Helmholtz equation may be rewritten as
[4∂z∗ ∂z + n2 (r)k2 ]u(z, z∗ ) = 0. (2.29)
Let us now introduce a complex coordinate w = w(z) which is an analytic

function of z. A transformation from z to w is a conformal mapping, and
the Helmholtz equation can be written in w-space in the form
[4∂w∗ ∂w + n2 (r)k2 ]u(w, w ∗ ) = 0, (2.30)
where n satisfies the expression
n2
n2 = . (2.31)
|dw/dz|2
The form of the Helmholtz equation is therefore preserved under the map-
ping w(z), and the effect of the map can be absorbed into a new effective
refractive index n . To design a cloak, we begin by choosing a mapping such
that w(z) = z for large values of w, and such that n = 1. Rays will form
straight lines in w-space, propagating as in vacuum. It therefore seems that,
regardless of their distortion in z-space, rays will return to their original
trajectories far from the distortion region.
84 Greg Gbur
Not every mapping will form a cloaking device, however, and will not
even necessarily result in a device that exhibits invisibility; even straightfor-
ward transformations necessarily result in a multi-sheeted Riemann surface
within which rays can be “lost.” For instance, the simple map
w(z) = z + a2 /z, (2.32)
with a a positive real number, results in the refractive index profile
n(x, y) = |1 − a2 /z2 |. (2.33)
The inverse transform, however, is of the form

1
z= w ± w 2 − 4a2 , (2.34)
2
which is a two-sheeted Riemann surface. There are a pair of branch points
at w = ±2a, and the branch cut in z-space is the unit circle.
In z-space, rays that do not cross the branch cut will be deflected from
the interior region near the cloak and return to their original trajectory far
away, satisfying our concept of a cloaking device. However, rays that do cross
the branch cut in w-space (or cross the unit circle in z-space) will transfer to
the other Riemann sheet and, in z-space, will asymptotically approach the
origin. The unit circle acts as an “event horizon” of sorts, and any rays that
pass within will never emerge again, resulting in the cloaking device casting
a significant shadow.2
This difficulty can be fixed by replacing the functional behavior on the
second Riemann sheet with a function that guides incoming rays around
a circular or elliptical path to cross the branch cut a second time and with
the same trajectory and position by which they entered. A possible refractive
index which does this is the Kepler potential,
r0
n2 = − 1, (2.35)
|w − w1 |
where w1 is taken to be one of the branch points, e.g., w1 = +2a. Values
of w for which this potential becomes negative are inaccessible to light rays
and are cloaked from observation, because the refractive index becomes pure
imaginary.
2 This relationship between black holes and cloaking devices will be further discussed in
Section 7.
There are a few important observations to be made about this cloak.

First, the use of conformal mapping restricts this particular method to two-
dimensional cloaks. Second,The asymmetric behavior of the Kepler potential
about w = 0 results in a cloak whose behavior is asymmetric:though cloaked
from all directions, the local ray behavior is quite different for different direc-
tions of incidence, as seen in Figure 7a. Third, the use of the Kepler potential
within the unit circle introduces a refractive index discontinuity that nat-
urally results in reflections. This means that, from a wave perspective, the
invisibility of this cloak design is not perfect; this agrees with Theorem 3.4.
However, as noted by Leonhardt (2006b), there is nothing in that theorem
precluding arbitrarily good (albeit not perfect) cloaking, and reflections can
be reduced by an appropriate antireflection coating. Finally, it is to be noted
that the boundary of the inaccessible region, as described by Equation (2.35),
has n = 0, implying an infinite phase velocity of light. This represents a fun-
damental challenge in the practical construction of the cloak, but it is not
unique to this particular design, as we shall see.
The work on conformal invisibility was extended by Leonhardt (2006a),
who investigated a number of such devices in detail and noted that the inte-
rior Riemann surface introduces a uniform time delay to those rays passing
through it. Hendi, Henn, and Leonhardt (2006) considered the relationship
between conformal invisibility and the inverse scattering problem. A more
sophisticated conformal cloak (Ochiai, Leonhardt, & Nacher, 2008) using
negative index materials3 was shown to eliminate the time delay, minimize
the reflections, and negate the need for infinite phase velocities on the inte-
rior surface of the cloak. We have noted that the original cloak was designed
only within the limit of geometrical optics; Chen, Leonhardt, andTyc (2011)
used simulations to show that it performs almost perfectly for waves as well,at
least for certain quantized frequencies. Even more recently, Xu, Liu, Zhang,
Ong, and Ma (2012) demonstrated that the reflections at the inner Riemann
sheet could be eliminated entirely by an appropriate choice of the inner
function.
The other fundamental cloaking design was proposed by Pendry et al.
(2006), and it applies transformation optics to produce a fully electromag-
netic cloak in three-dimensional space. In terms of Maxwell’s equations,
transformation optics is built on the observation that the free space Maxwell’s
equations in an arbitrary curved space are equivalent to the macroscopic
Maxwell’s equations in a Cartesian coordinate system. This can be done most
3 These materials will be discussed in Section 6.

86 Greg Gbur
elegantly by using tensor analysis,as we now briefly demonstrate. A complete

derivation can be found in Leonhardt and Philbin (2009); we summarize the
main observations of that article.
Maxwell’s equations in a free space in Cartesian coordinates may be
written in the form
ρ
∇ · E = , ∇ · B = 0,
0
∂B 1 ∂E
∇ ×E=− , ∇ ×B= 2 + μ0 J. (2.36)
∂t c ∂t
In an arbitrary coordinate system, the equations may be written as
1 √ i ρ 1 √ i
√ gE ,i = , √ gB ,i = 0,
g 0 g
∂B i 1 ∂E i
ijk Ek, j = − , ijk Bk, j = 2 + μ0 J i .
∂t c ∂t
Here g is the determinant of the metric tensor gij of the space or coordinate
system and ijk is the Levi–Civita tensor, which in an arbitrary right-handed
coordinate system may be expressed as
1
ijk = √ [ijk], (2.37)
g
where [ijk] is the familiar permutation symbol, defined by
⎧
⎨ +1, ijk even permutation,
[ijk] = −1, ijk odd permutation, (2.38)
⎩
0 otherwise.
The notation (F),j indicates the partial derivative of the quantity with respect
to the jth component, i.e.,
∂F
(F),j ≡ . (2.39)
∂xj
By use of the inverse metric tensor, Equation (2.37) may be expressed in
the suggestive form,
√
√ ij gρ √ ij
gg Ej ,i = , gg Bj ,i = 0,
0
√ √
∂( gg ij Bj ) 1 ∂( gg ij Ej ) √
[ijk]Ek,j = − , [ijk]Bk,j = 2 + μ0 g J i .
∂t c ∂t
(2.40)
Because these equations are in free space, the B-field can also be written in
terms of the H in the usual manner,
Bj = μ0 Hj . (2.41)
This revised set of equations can be compared to the macroscopic
Maxwell’s equations in a Cartesian coordinate system but in a material
medium,
D,ii = ρ, B,ii = 0,
∂B i ∂D i
[ijk]Ek,j = − , [ijk]Hk,j = + J i. (2.42)
∂t ∂t
The two sets of equations can be reconciled if we take the following consti-
tutive relations,
D i = 0 ij Ej , B i = μ0 μij Hj , (2.43)
with the following identifications of the permittivity and permeability,
√
ij = μij = gg ij . (2.44)
These relations provide a prescription for constructing an electromagnetic
invisibility device. One designs a “virtual” warping of space for the free-
space Maxwell’s equations, and then determines directly the type of material
properties that will produce the effect in real Cartesian space.
There are several important observations to be made immediately from
this construction. Because the permittivity and permeability are equal, the
material is of unit impedance, and in fact produces no reflections. Further-
more, these quantities are in general of tensor form, which means that the
material is anisotropic. Moreover, the presence of a nontrivial permeability
indicates that the material is in general magnetic. Because most materials in
nature do not possess a significant magnetic response at optical frequencies,
this cloaking design needs to be fabricated out of novel substances known
as metamaterials, to be discussed in Section 6.
The electromagnetic cloak described by Pendry et al. (2006) is a spherical
shell of inner radius R1 and an outer radius R2 .A spatial transformation is used
that compresses all fields within the sphere r < R2 into the shell R1 < r < R2 ,
of the form
r = R1 + r(R2 − R1 )/R2 , θ = θ, φ = φ. (2.45)
The warping of a Cartesian region of space under this coordinate transfor-
mation is illustrated in Figure 8, with the path of a ray shown for clarity. This
88 Greg Gbur
Figure 8 The virtual distortion of space created by the Pendry–Schurig–Smith cloak.

A diagonal ray is shown for clarity.
transformation results in the following diagonal values of the permittivity,
R2 (r − R1 )2
r = , (2.46a)
R2 − R1 r 2
R2
θ = , (2.46b)
R2 − R1
R2
φ = , (2.46c)
R2 − R1
and with μij = ij .
It should be noted immediately, that r → 0 as r → R1 . This indicates
that n → 0 on the inner surface, or that the phase velocity of light becomes
infinite on that surface, as it did for the conformal cloak. The reason for this
is clear from a consideration of Figure 7b. In order to provide an undistorted
wavefront at the output of the cloak, all rays must travel the same optical path
within the cloak. However,those rays that pass close to the inner surface must
travel a longer spatial distance, and must have an increased phase velocity to
make up the difference. Though it is certainly possible to produce a material

with a large or infinite phase velocity, this can only be done exactly at a
single frequency, which necessarily makes the cloaking effect produced here
a narrowband one. Much of the research that has been done since this
original concept was introduced has consisted of attempts to understand and
remove this restriction.
This cloaking device, in principle perfect, does not violate Theorem 3.4,
which was only proven for isotropic materials. The use of anisotropy there-
fore provides yet another “loophole” for invisibility. This observation seems
to have been anticipated by Greenleaf, Lassas, and Uhlmann (2003), who
noted that there exist anisotropic conductivities that are undetectable in the
problem of electrical impedance tomography.
The perfect invisibility of an ideal cloak was soon demonstrated by Chen,
Wu, Zhang, and Kong (2007) using an exact Mie scattering method, though
it was noted that a cloak with loss will only be perfect in the backscattering
direction; a simulation of a perfect cloak is shown in Figure 9. Even a cloak
which fails to meet exactly the stringent material requirements of Equation
(4.19) can still achieve a significant degree of invisibility, as was shown by
Cummer, Popa, Schurig, and Smith (2006) using a full-wave analysis. Qiu,
Hu, Xu, and Feng (2009) have shown that an anisotropic spherical cloak can
be well approximated by many thin layers of homogeneous material.
A number of extensions to the original Pendry–Schurig–Smith cloak
(PSSC) have been investigated.Yaghjian and Maci (2008) introduced a bound-
ary condition technique for deriving electromagnetic cloaks and derived a
whole family of spherical and cylindrical cloaks, with the original PSSC
Figure 9 E-field distribution and Poynting vector of a plane wave interacting with an
ideal cloak. (From Chen et al. (2007)) Copyright American Physical Society.
90 Greg Gbur
as a special case. The authors further noted that causality prohibits a per-
fect cloak of this form that functions over a finite bandwidth. Chen et al.
(2007) and Kildishev, Cai, Chettiar, and Shalaev (2008) considered extend-
ing the bandwidth of operation by tailoring the dispersion properties of the
cloak material, but also observed that this approach was subject to signifi-
cant causal limitations. These limitations were quantified by Hashemi, Qiu,
McCauley, Joannopoulos, and Johnson (2012), who derived an uncertainty
relation of sorts between the cloaking bandwidth and the diameter of the
cloak.
5. OTHER CLOAKING DESIGNS

The introduction of cloaking physics, and the efforts to overcome
the limitations of the first cloaks, have spawned a number of variations on
the initial concepts. Right before the current cloaking “craze” began, Alù
and Engheta (2005) demonstrated that the cross-section of a spherical or
cylindrical object can be dramatically reduced by coating it with a low-loss
and low permittivity shell that has a plasmon resonance. The principle of
the device is analogous to that of Kerker (1975b), mentioned earlier: the
dipole moments of a high refractive index core and low index shell can
be made to cancel almost entirely. Though the cloaking is not perfect, and
the cloak design is dependent on the nature of the object to be cloaked,
the plasmonic device possesses several advantages. Alù and Engheta (2008)
showed that multiple frequencies can be cloaked simultaneously with multi-
layered plasmonic shells. Because the electromagnetic field is not identically
zero within the cloak, it is possible to design a plasmonic cloak that can hide
a functioning sensor, as Alù and Engheta (2009) demonstrated. Using this
latter idea, Alù and Engheta (2010) have proposed noninvasive near-field
imaging using a cloaked near-field scanning optical microscope. Though
the initial plasmonic cloak design was for spherical and cylindrical objects,
Tricarico, Bilotti, Alù, and Vegni (2010) have applied the same technique to
cloak irregular objects with anisotropic scattering properties.
Two of the biggest challenges in optical cloaking are the bandwidth
limitations and the complex anisotropic materials required. Both of these
challenges were alleviated by Li and Pendry (2008), who introduced a new
strategy called “hiding under the carpet.”The principle of operation is illus-
trated in Figure 10. Using transformation optics, space is bent away from a
planar surface, leaving a gap underneath within which objects may be hid-
den. Light waves incident upon the carpet cloak are reflected as if from the
light rays
cloaking
“carpet”
Figure 10 The principle of the “carpet cloak,” which bends incident rays and waves
away from an excluded volume, making it appear that the surface is flat.
planar surface below. Unlike the PSSC, this device can be constructed with
nearly isotropic materials; furthermore, it does not require infinite phase
velocities and therefore can be constructed to work over a larger frequency
range. These simplifications have resulted in a number of experimental real-
izations, which will be discussed in the next section. However, it has been
shown by Zhang, Chan, and Wu (2010) that a carpet cloak designed without
any anisotropy results in a lateral shift of the reflected wave that in principle
makes the cloaked region detectable.
To simplify the design of a cylindrical cloak, Cai, Chettiar, Kildishev, and
Shalaev (2007) introduced a reduced set of cloaking material parameters
that are nonmagnetic. This cloak, however, is not impedance matched and
will result in some scattered fields. By using a higher-order transformation,
Cai, Chettiar, Kildishev, Shalaev, and Milton (2007) reduced the scattering
considerably.
Another strategy for simplifying the design of cloaks is to consider devices
that only work for a single direction of illumination (somewhat reminiscent
of the early nonscattering scatterer theory of Devaney (1978)). Xi,Chen,Wu,
and Kong (2009) introduced a unidirectional cloak that can be fashioned only
out of isotropic materials. A similar strategy was independently arrived at by
Urzhumov and Smith (2012) some time later.
From its very name, one naturally assumes that an invisibility cloak must
wrap around the object to be hidden. Remarkably, however, a number of
researchers have pointed out that a cloak may be designed that hides objects
outside of itself. Vasquez, Milton, and Onofrei (2009) demonstrated that an
active cloak,4 consisting of a finite system of sensors coupled to wave sources,
can conceal an object lying outside of the source domain. This idea was
4 In contrast to the use of the word “active” in Section 3, here it refers to a system that senses
an incoming field and produces an active response.
92 Greg Gbur
(a) (b) (c)

ε = −1 ε=1 ε = −1 ε=1
= −1 =1 = −1 =1
ε'', µ''
antiobject object ε', µ'
Figure 11 External cloaking using complementary media. (a) A negative index slab
essentially “negates” propagation through an adjacent vacuum region. (b) The scatter-
ing of an external object may be canceled by a complementary object within the slab.
(c) A spherical object can also be used for a complementary medium.
inspired by an active cloaking design by Miller (2006) that could hide an

arbitrary object within its domain; a detailed investigation of active cloaking
was also performed by Greenleaf, Kurylev, Lassas, and Uhlmann (2007b).
Even more surprising, Lai, Chen, Zhang, and Chan (2009) have shown
that it is possible to perform passive external cloaking by exploiting the
properties of negative index materials. The idea is illustrated in Figure 11. A
negative refractive index slab with n = −1 essentially cancels the propagation
effects of the slab and a region of equivalent thickness beyond it: in terms
of transformation optics, the slab may be said to compress that space to zero
width. If an object is put in this compressed region,scattering associated with
it can be negated by the introduction of a negative index “mirror image”
within the slab; a similar strategy can even be applied to a spherical external
cloak.
In Section 3 it was noted that the existence of invisible objects implies
nonuniqueness of the inverse scattering problem.This nonuniqueness in turn
implies that it is in principle possible to use anisotropic metamaterials to make
one object look like another.This idea of illusion optics was first introduced by
Lai et al. (2009), who used transformation optics to demonstrate an illusion
scheme to make a dielectric spoon look like a metal cup; the results of their
simulation is shown in Figure 12. Jiang, Ma, and Cheng (2010) proposed a
modified illusion strategy which requires only positive values of permittivity
and permeability. Just as objects can be externally hidden by active cloaks,
they can also be externally disguised, as shown by Zheng, Xiao, Lai, and
Chan (2010).
The preceding illusion cloaks require either active sources or must be
tailored specifically to the object to be disguised. Recently, Luo, He,Wang,
Figure 12 A numerical demonstration of illusion optics, showing how the scattering

pattern of a dielectric spoon can be converted into that of a metal cup through an
appropriate illusion cloak. (Figure from Lai et al. (2009)) Copyright American Physical
Society.
Chan, and Zhu (2011) combined the principles of cloaking and illusion to
create a passive device that works independently of the shape of the concealed
object.
One of the more surprising results of the illusion paper by Lai et al.
(2009) is the argument that it is possible to make an illusion of a hole in an
opaque screen, in essence frustrating the action of the screen. Along similar
lines, it has been suggested by Chen, Luo, Ma, and Chan (2008) that it is
possible to embed an anisotropic “anti-cloak” within a cloaked region to
make the cloak visible. The idea of anticloaking has been further developed
by Castaldi, Gallina, Galdi,Alù, and Engheta (2009).
In what might be considered another consequence of inverse scattering
nonuniqueness, Yang, Chen, Luo, and Ma (2008) have proposed the idea
of a “superscatterer,” an object with a scattering cross-section greatly larger
than the geometric cross-section. Luo,Yang, Gu, Chen, and Ma (2009) have
proposed that such superscatterers could be used to conceal an otherwise
open entryway.
A number of other results in the subject of cloaking are worth mentioning
that do not fit into the broader categories discussed above. Alù (2009) has
introduced the idea of a mantle cloak, which achieves invisibility by the use
of a thin patterned surface region on the object to be hidden. It has been
shown by Baumeier, Leskkova, and Maradudin (2009) that regions of a metal
surface may be cloaked from surface plasmons by means of an array of point
scatterers. Hakansson (2007) has approached the problem of cloaking by
the use of inverse design of scattering optical elements. Setälä, Hakkarainen,
Friberg, and Hoenders (2010) have looked at the problem of cloaking of
a slab object within the context of the first Born approximation, making a
connection between modern cloaking theory and the earlier researches.
94 Greg Gbur
6. EXPERIMENTS AND MATERIAL PROPERTIES

Many of the cloaking designs considered share the need for complex
material structures that meet one or more of the following conditions:
• a magnetic response,
• anisotropy,
• a high-gradient refractive index.
These properties are difficult to achieve in general, and especially at opti-
cal frequencies where most natural materials are nonmagnetic, i.e., μ = μ0 .
Furthermore, the optical properties of natural materials are dictated by their
atomic and molecular behavior and have traditionally been considered inher-
ent and immutable.
However, it has long been appreciated that the structure of materials can
play a huge role in their interactions with light, and this structure can be
engineered to produce effects that are rare or nonexistent in nature. An
early example of such engineering is the development of photonic crystals,
materials with periodic structures comparable in scale to the wavelength
of light, which can be used to create photonic bandgaps (see, for instance,
Joannopoulos, Meade, & Winn, 1995).
More recently, it has been realized that the engineering of material
structures on the subwavelength scale can produce even more dramatic
effects. The first demonstration of this principle came from Pendry, Holden,
Robbins, and Stewart (1999), who showed that an array of subwavelength
nonmagnetic conducting sheets can produce an effective magnetic perme-
ability tunable to values not found in nature.5 One fundamental structure
introduced in this paper was the split ring resonator, a schematic of which is
shown in Figure 13. The inherent capacitance of the split ring design even
allows the permeability to be negative, an important result in what follows.
It had been shown many years earlier by Vesalago (1968) that a material
with simultaneously negative values of and μ will have a refractive index
less than zero; this in turn suggested the possibility of many unusual and
novel optical phenomena, such as negative refraction, lensless imaging, and
a reverse Cerenkov effect. It was unclear in Vesalago’s time how one could
achieve such material properties, and his results spent many years in neglect.
5 In curious synchronicity with earlier work of Sommerfeld, Pendry first began work on
metamaterials when he was called to consult for a company that had developed a radar-
absorbing carbon coating for battleships, the physics of which they did not understand
(Grant & Hapgood, 2009).
(a)
(b)
Figure 13 An illustration of a cylinder with a split ring structure, from (a) side and
(b) top.
n=1 n = −1 n=1
Figure 14 An illustration of imaging with a perfect lens.
Negative refraction drew new interest (and controversy) when Pendry

(2000) suggested that Vesalago’s negative index lens could in fact be “per-
fect,” providing resolution beyond the ordinary Rayleigh criterion. This is
illustrated schematically in Figure 14. Geometric rays will be focused by the
lens and the evanescent waves will be amplified within the material. Pendry’s
hypothesis sparked intense research into such unusual materials,soon dubbed
metamaterials. An experimental demonstration of negative refraction using
microwaves was published soon after by Shelby, Smith, and Schultz (2001),
using split ring resonators to achieve the required magnetic properties of the
negative index.
Broadly speaking,a metamaterial is a material engineered with a subwave-
length structure in order to achieve optical properties not found in nature. It
is outside the scope of this article to discuss the numerous experimental and
theoretical results relating to metamaterials; we refer the interested reader to
Smith,Pendry,andWiltshire (2004), Shalaev (2007) and Litchinitser,Gabitov,
Maimistov, and Shalaev (2008) for reviews. We instead consider only those
results relevant particular to the subject of invisibility and cloaking.
The first experimental demonstration of a cloaking device was performed
by Schurig et al. (2006) at microwave wavelengths. This two-dimensional
96 Greg Gbur
Figure 15 The first experimental cloaking device, developed for microwaves. (Figure
courtesy of David Smith of Duke University).
device used concentric circles of split ring resonators to produce the appro-
priate electric and magnetic material responses; a photograph of the device is
shown in Figure 15. Placed within a planar waveguide, the cloak was shown
to effectively hide a copper cylinder within it. However, it is to be noted that
even for this planar cloaking device, a simplified set of material parameters
were chosen for fabrication, with nontrivial reflections as a result.
As was noted in the previous section,the material requirements for cloak-
ing are much easier in the “hiding under the carpet” configuration of Li and
Pendry (2008), which requires much less extreme variations in refractive
index and can use approximately isotropic materials. The first realization of
such a “ground-plane cloak” was done in the microwave regime by Lie et al.
(2009), and they demonstrated effective cloaking over a wavelength range
from 1.9 to 2.3 cm. Perhaps as important as the demonstration itself was the
development of an algorithm that converts a desired index profile into the
appropriate configuration of metamaterial elements. The cloak was assumed
to be embedded in a background medium with refractive index n = 1.331,
which resulted in cloak indices ranging from n = 1.08 to n = 1.67, all
achievable with nonresonant metamaterials.
Other designs followed,in general increasing the complexity of the device
and reducing the wavelength of operation.Valentine, Li, Zentgraf, Bartal, and
Zhang (2009) fabricated a carpet cloak that functions in a wavelength range
of 1.4–1.8 mm. The device consisted of an array of subwavelength holes of
varying density etched in a silicon slab, cloaking a region 3.8 by 0.4 µm.
At nearly the same time, Lee et al. (2009) and Gabrielli, Cardenas, Poitras, and
Lipson (2009) independently produced carpet cloaks fashioned from silicon
nanorods, operating over a range of wavelengths around 1.5 µm.
All of these carpet designs shared certain features in common: a high
background refractive index, a cloaked region with dimensions comparable
to or smaller than the wavelength, and two-dimensional geometry. The lim-
itation on geometry was overcome by Ergin, Stenger, Brenner, Pendry, and
Wegener (2010), who introduced a cloak that is effective for a broad range of
viewing angles and for unpolarized light at wavelengths from 1.4 to 2.7 µm.
This design was later miniaturized by Fischer, Ergin, and Wegener (2011)
to develop a cloak that functions at wavelengths between 650 to 900 nm,
including part of the visible spectrum. By fashioning a cloak out of optical
calcite, Zhang, Luo, Liu, and Barbastathis (2011) and Chen et al. (2011) were
able to demonstrate the cloaking of a macroscopic region of space using
visible light.
Other types of cloaks have also been demonstrated experimentally. The
reduced cloak of Cai, Chettiar, Kildishev, and Shalaev (2007) was successfully
tested by Smolyaninov, Hung, and Davis (2008). Scattering-cancelation via
plasmonic cloaking was shown by Edwards, Alù, Silveirinha, and Engheta
(2009). A unidirectional two-dimensional cloak was demonstrated by Landy
and Smith (2013); this is perhaps the first cloak that applies transformation
optics without any simplifying approximations in the design. In a curious
synthesis, Xu et al. (2012) experimentally tested a cloak that combines strate-
gies of transformation optics, conformal mapping, and plasmonic cloaking.
One of the major bottlenecks in future experimental realizations of opti-
cal cloaks is the development of robust,lossless,three-dimensional metamate-
rials at optical frequencies.Theoretically,Silveirinha,Alù,and Engheta (2007)
have introduced metamaterial composite structures that could be used for
cloak designs, and Xie et al. (2011) have proposed an invisibility cloak design
using only silver nanowires. Though the experimental side is difficult, bulk
metamaterials that function in at visible and near-infrared wavelengths have
been achieved on a limited scale.Yao et al. (2008) has demonstrated negative
refraction in collections of silver nanowires, whileValentine et al. (2008) has
produced a fishnet structure of silver to similar effect.
7. RELATED APPLICATIONS
The insights gained from research on cloaking have not been restricted
to optical invisibility. Cloaking concepts have been introduced in a variety of
98 Greg Gbur
other fields where shielding from waves is important; furthermore, transfor-

mation optics has been used to develop a number of other surprising devices
and effects. In this section, we summarize some interesting results of both
aspects.
Not long after the introduction of optical cloaking, Greenleaf, Kurylev,
Lassas, and Uhlmann (2008a) suggested that approximate cloaking is possible
for quantum mechanical matter waves. For certain special energies, however,
the cloak can function as a novel type of ion trap. These ideas, and possi-
ble applications, were further developed by Greenleaf, Kurylev, Lassas, and
Uhlmann (2008b), and similar concepts were considered by Zhang, Genov,
Sun, and Zhang (2008).
Other applications abound.Wood and Pendry (2007) introduced a super-
conducting cube lattice to act as a shield from static magnetic fields; this pos-
sibility was experimentally tested by Magnus et al. (2008). It has been shown
by Sanchez, Navau, Prat-Camps, and Chen (2011) and Gömöry, Solovyov,
Šouc,Navau,Prat-Camps,and Sanchez (2012) that a combination of a ferrous
shell, which attracts fields, and a superconducting shell, which repels them,
can create a cloak for static magnetic fields. Similarly, a DC electric cloak
was experimentally implemented by Yang, Mei, Jin, and Cui (2012). Trans-
formation optics has even been applied to the cloaking and concentrating
of heat flux (Guenneau,Amra, & Veynante, 2012).
Cloaking in acoustics has also been extensively investigated, though it is
complicated by the fact that,unlike Maxwell’s equations,acoustic wave equa-
tions are not in general form-invariant under coordinate transformations.
The first work was done by Cummer and Schurig (2007), who introduced a
design of a two-dimensional acoustic cloak. A structure to achieve this design
was proposed byTorrent and Sánchez-Dehesa (2008). Chen and Chan (2007)
and Cummer et al. (2008) formulated plans for three-dimensional cloaks.
A broadband acoustic cloak has been constructed for surface waves on a
fluid (Farhat, Enoch, Guenneau, & Movchan, 2008), and one has also been
fabricated for underwater ultrasound (Zhang, Xia, & Fang, 2011).
A novel type of water wave cloak was recently proposed by Alam (2012)
to protect offshore platforms from dangerous high-amplitude waves. Noting
that waves propagate not only on the surface of the ocean but also underwater
at interfaces between layers of different temperatures,Alam has suggested that
an appropriate set of seafloor features could couple surface waves into these
interfacial waves and back again, leaving a region of relative calm. This is
illustrated in Figure 16.
Closely related to acoustic cloaks are elastic wave cloaks, which have also
received significant attention. The first study of such cloaks seems to have
Figure 16 Illustration of a seafloor-based cloak for water waves. (Figure from Alam
(2012)) Copyright American Physical Society.
been done by Milton, Briane, and Willis (2006). A particular form of cylin-
drical cloak has been introduced by Brun, Guenneau, and Movchan (2009).
Recently, experiments have been done to demonstrate elastic cloaking in
thin plates (Stenger, Wilhelm, & Wegener, 2012). One possible application
of such elastic cloaking is the protection of buildings from seismic waves.
Parnell (2012) has considered using pre-stressed elements to guide waves
around a protected target, while Kim and Das (2012) has suggested that seis-
mic waves could be converted to evanescent ones and dissipated in the cloak,
preventing collateral damage to nearby structures.
A two-dimensional cloak, in essence,“stretches” space in the x − y plane
away from the origin x = 0, y = 0. If one conceptually replaces y by the
time variable t, the result is a spacetime cloak, as first envisioned by McCall,
Favaro, Kinsler, and Boardman (2011). The effect of such a cloak on a pulse
is illustrated in Figure 17a. The spacetime cloak was experimentally realized
by Fridman, Farsi, Okawachi, and Gaeta (2012), by using a “split-time lens”
to produce chirped red-shifted and blue-shifted sections of a CW beam.
A dispersive fiber was used to separate the frequency-shifted regions, pro-
ducing a 50 picosecond gap. Beyond this, a dispersion-compensated fiber
brought the regions together again, and the chirp was undone by a second
split-time lens. This configuration is illustrated in Figure 17b. A nonlinear
optical phenomenon was used as the “event” to hide from the CW beam.
We conclude this section by noting that a number of other unusual devices
and phenomena have been created by the use of transformation optics. An
early discussion of the possibilities was presented by Leonhardt and Philbin
100 Greg Gbur
(a) (b)
time
optical
fiber
“split-time
laser lens”
event
region
“split-time
lens” dispersion-
detecort compensating
fiber
position
Figure 17 Illustrating (a) the action of a temporal cloak on a pulse, in terms of space
and time coordinates and (b) the experimental setup used to realize such cloak.
(2006). It was shown by Leonhardt (2009) that a Maxwell’s fish eye lens
is a transformation medium and can produce a perfect image, albeit in the
unusual case when both source and image lie within the lens. Transforma-
tion optics has been used to design reflectionless beam shifters and split-
ters (Rahm, Cummer, Schurig, Pendry, & Smith, 2008), as well as lossless
fiber bends (Roberts, Rahm, Pendry, & Smith, 2008), fiber-to-chip cou-
plers (Markov,Valentine, & Weiss, 2012) and light squeezers and expanders
(García-Meca et al., 2011).
Transformation optics also has the potential to model the behavior of
light around celestial bodies, both real and hypothetical (Genov, Zhang, &
Zhang, 2009). For instance, Chen, Miao, and Li (2010) introduced a trans-
formation that mimics the behavior of light outside the event horizon of a
Schwarzschild black hole. A detailed physical model of a black hole, however,
must also include warpings of time as well as space; Philbin et al. (2008) have
experimentally implemented a fiber-optic analog of a black hole’s event hori-
zon. Perhaps even more unusual, Greenleaf, Kurylev, Lassas, and Uhlmann
(2007a) have introduced the concept of an electromagnetic wormhole.
8. INVISIBILITY IN SCIENCE FICTION

Long before physicists seriously considered the idea of invisibility,
authors of science fiction and horror were coming up with their own “sci-
entific” visions of how invisibility could work. A number of these ideas were
surprisingly close to modern developments; in this section we take a brief
look at the history of invisibility in fiction,6 and wonder whether authors

will inspire the next generation of optical devices—or vice versa.
The first “physical” description of an invisible object can be traced back
to the Irish-born American writer Fitz James O’Brien (1828–1862), who
in 1859 penned the horror story, “What Was It?” The tale concerns a pair
of men who spend the night in a seemingly haunted house, only to find
that the residence is occupied by a very living—but invisible—humanoid
monster. The character Hammond calms his friend Harry with a scientific
explanation:
“Let us reason a little, Harry. Here is a solid body which we touch, but which we
cannot see. The fact is so unusual that it strikes us with terror. Is there no parallel,
though, for such a phenomenon? Take a piece of pure glass. It is tangible and
transparent. A certain chemical coarseness is all that prevents its being so entirely
transparent as to be totally invisible. It is not theoretically impossible, mind you,
to make a glass which shall not reflect a single ray of light—a glass so pure and
homogeneous in its atoms that the rays from the sun shall pass through it as they
do through the air, refracted but not reflected. We do not see the air, and yet we
feel it.”
This notion of a transparent body that refracts light, but does not reflect
it, sounds very much like the “invisible bodies” of Kerker and the “non-
scattering scatterers” of Devaney and Wolf. It could be said that Wolf and
Habashy (1993) disproved O’Brien’s hypothesis by disproving the existence
of perfectly nonscattering scatterers, at least for isotropic materials!
The most famous story about an unseen being is the early science fiction
novel “The Invisible Man” by H.G. Wells (1866–1946), first published in
1897. (See Figure 18a) In the book, a chemist named Griffin invents an
elaborate process by which he makes his own body invisible,without realizing
the dire difficulties that come with such a power. Seeking shelter in the house
of his old colleague Kemp, Griffin explains how invisibility is possible:
“If a sheet of glass is smashed, Kemp, and beaten into a powder, it becomes much
more visible while it is in the air; it becomes at last an opaque white powder. This
is because the powdering multiplies the surfaces of the glass at which refraction
and reflection occur... But if the white powdered glass is put into water, it forthwith
vanishes...
You make the glass invisible by putting it into a liquid of nearly the same refractive
index; a transparent thing becomes invisible if it is put in any medium of almost
the same refractive index. And if you will consider only a second, you will see also
that the powder of glass might be made to vanish in air, if its refractive index could
6 Most of the stories in this section have been reprinted multiple times. We therefore only
mention the title, author, and publication year, and do not cite the original source.
102 Greg Gbur
Figure 18 (a) Original 1897 advertisement for “The Invisible Man.” (b) Illustration from
a 1906 publication of “The Shadow and the Flash” in The Windsor Magazine.
be made the same as that of air; for then there would be no refraction or reflection
as the light passed from glass to air.”
Wells, who acquired a scientific education in biology under Thomas

Henry Huxley,knew that index matching is required in order to avoid reflec-
tions at the surface of a would be invisible dielectric object. If he had been
aware of magnetic materials and the more general concept of impedance
matching, used in modern cloaks, his description of the science might have
been quite different.
Griffin’s above description only explains how invisibility works, but it
does not explain the process by which it is achieved. Griffin explains this to
Kemp as well:
“But the essential phase was to place the transparent object whose refractive index
was to be lowered between two radiating centers of a sort of ethereal vibration,
of which I will tell you more fully later. No, not those Roentgen vibrations—I don’t
know that these others of mine have been described. Yet they are obvious enough.”
Clearly Wells was inspired by the discovery of X-rays (Roentgen rays)

only two years earlier, in 1895, and he would not be the only writer to be
influenced by “invisible rays.” In 1894, the famed author and satirist Ambrose
Bierce (1842–1913) wrote “The Damned Thing,” a four-part story centered
around the mysterious and savage murder of a man named Hugh Morgan.
By the end of the tale, it is revealed in Morgan’s own diary that he was being
stalked by a creature unseen:
“It is known to seamen that a school of whales basking or sporting on the surface
of the ocean, miles apart, with the convexity of the earth between them, will
sometimes dive at the same instant—all gone out of sight in a moment. The
signal has been sounded—too grave for the ear of the sailor at the masthead and
his comrades on the deck—who nevertheless feel its vibrations in the ship as the
stones of a cathedral are stirred by the bass of the organ.
As with sounds, so with colors. At each end of the solar spectrum the chemist can
detect the presence of what are known as ‘actinic’ rays. They represent colors–
integral colors in the composition of light—which we are unable to discern. The
human eye is an imperfect instrument; its range is but a few octaves of the real
‘chromatic scale’ I am not mad; there are colors that we can not see.
And, God help me! the Damned Thing is of such a color!”
The “actinic” rays are what are now known as ultraviolet (UV) light. The
discovery of UV,infrared,and X-rays drove home in the public consciousness
that there are significant and powerful things unseen by the naked eye.
This idea of an invisible creature, or of making a living creature invisible,
sounds outlandish. However, it should be noted that in 2011 a new reagent
called Scale was introduced that can make organic tissue highly transparent
(Hama et al., 2011).This reagent was only used on dead tissue, but one of the
authors suggested more ambitious possibilities in a news article (Gonzalez,
2011):
We are currently investigating another, milder candidate reagent which would
allow us to study live tissue in the same way, at somewhat lower levels of trans-
parency. This would open the door to experiments that have simply never been
possible before.
Scale is not the only optical research anticipated by a science fiction

author. In the 1903 story “The Shadow and the Flash” by Jack London
(1876–1916), two lifelong rivals race to achieve invisibility before the other,
and each pursues a different plan. Paul Tichlorne attempts to achieve perfect
transparency, but finds that he is still dogged by the “flash” of refraction and
reflection. Lloyd Inwood, however, tries to hide himself by the development
of perfect absorption (see Figure 18b):
“Colour is a sensation,” he was saying. “It has no objective reality. Without light, we
can see neither colors nor objects themselves. All objects are black in the dark, and
in the dark it is impossible to see them. If no light strikes upon them, then no light is
flung back from them to the eye, and so we have no vision-evidence of their being.”
104 Greg Gbur
“But we see black objects in daylight,” I objected.

“Very true,” he went on warmly. “And that is because they are not perfectly black.
Were they perfectly black, absolutely black, as it were, we could not see them—ay,
not in the blaze of a thousand suns could we see them! And so I say, with the right
pigments, properly compounded, an absolutely black paint could be produced
which would render invisible whatever it was applied to.”
Lloyd’s scheme also has trouble, because a truly black object still casts a
shadow that can be observed.7 However, surfaces that are almost perfectly
absorbing are of importance in the development of solar energy collectors
and thermal detectors; Mizuno et al. (2009) have designed such a surface
fashioned from a vertical forest of carbon nanotubes.
In a quite remarkable feat of imagination, the author A. Merritt (1884–
1943) even anticipated the development of the modern cloaking device. In
Merritt’s 1923 novel “The Face in the Abyss,” the hero Graydon comes into
contact with a series of dragon-like Messengers that can disappear at will.
Like Hammond and Harry long before, Graydon uses science to reassure
himself and explain the effect:
Science knew the thing was possible, and scientists the world over were trying to
find the secret to use in the next war.
Yes, the invisible Messengers were easily explained. Conceive something that nei-
ther absorbs light nor throws it back. In such case the light rays stream over that
something as water in a swift brook streams over a submerged boulder. The boul-
der is not visible. Nor would be the thing over which the light rays streamed. The
light rays would curve over it, bringing to the eyes of the observer whatever image
they carried from behind. The intervening object would be invisible. Because it
neither absorbed nor threw back light, it could be nothing else.
There is a traveler in the desert. Suddenly he sees before him a rivulet and green
palms. They are not there. They are far behind the mountain at whose base they
seem to be. The rays of light carrying their images have struck upward, angled over
the mountain, struck down, and have been reflected in denser hot air. It is a mirage.
The example was not entirely analogous, but the basic principle was the same.
The truly amazing thing about this quote is that Merritt has not only made
the proper connection between the gradient refractive index in mirages and
cloaking,but anticipated the exact analogy of water flowing around a boulder
that authors of the first cloaking papers would later use! From an article by
Ball (2006), describing the release of the first cloaking concepts:
[Ulf Leonhardt] and John Pendry of Imperial College London, UK, and their co-
workers have independently described similar ways to create an invisible “hole” in
7 And also generates blackbody radiation.
space, inside which objects can be hidden. They say it is possible to guide light
around the hole, rather like water flowing around a rock in a river, so that the
object inside it cannot be seen.
One curious question remains: what inspired the first scientific explana-
tions of invisibility given by F.J. O’Brien and H.G.Wells? No direct evidence
survives, as O’Brien’s private correspondence was lost and Wells notoriously
did not discuss the inspiration for his writing. However, from the above
descriptions, we can make a reasonable guess.
Wells’ discussion of making powdered glass invisible by submerging it in
water is a very good clue. In a paper by Coblentz (1904), written at about
the same time as Wells’ book, we find the following remark:
Occasionally there appear notices of methods for making objects invisible by
selecting combinations of media having about the same index of refraction as
the immersed solid. This subject was first investigated by Christiansen in 1884. His
object was to show that white powders are transparent.
The discovery by Christiansen (1884) being referenced is now known as
a Christiansen filter. A powder placed in an index-matched liquid will become
quite transparent; however, because the liquid and the powder typically have
different dispersive properties, the mixture will in general transmit light over
a narrow range of wavelengths and strongly scatter at all others.This research
was quite popular at the time of Wells’ education, and as it is known that he
read textbooks on physics and optics, he was likely aware of it.
It should be noted that Coblentz mentions another piece of work on
Christiansen filters, the aforementioned article by Wood (1902):
The most recent notice on this subject is by Wood, who dissolved chloralhydrate
in glycerine. This solution has almost the same dispersion as glass, and finely
powdered glass, in it, is transparent without showing a trace of Christiansen’s
colors.
We discussed another aspect of this paper at the beginning of this review,
bringing us in a sense full circle. Wood introduced his article on his crude
invisibility device by demonstrating how transparent objects can be made
invisible by index-matching.
But where did Christiansen get the inspiration for his own experiments?
He references the pioneering and influential optics work of Newton (1721)
himself. Isaac Newton described in detail how transparent objects may be
made unseen by the use of index-matching (page 221):
But the truth of this Proposition will farther appear by observing, that in the
Superficies interceding two transparent Mediums, (such as are Air, Water, oil, com-
mon Glass, Crystal, metalline Glasses, Island Glasses, white transparent Arsenick,
106 Greg Gbur
Diamonds, etc.) the Reflexion is stronger or weaker accordingly, as the Superficies

hath a greater or less refracting Power. For in the Confine of Air and Salgem ’tis
stronger than in the Confine of Air and Water, and still stronger in the Confine of Air
and common Glass or Crystal, and stronger in the Confine of Air and a Diamond. If
any of these, and such like transparent Solids, be immerged in Water, its Reflexion
becomes much weaker than before; and still weaker if they be immerged in the
more strongly refracting Liquors of well rectified Oil of Vitriol or Spirit of Turpentine.
Perhaps unsurprisingly, it seems therefore that the inspiration for H.G.

Wells’invisible man may be traced right back to the work of Newton himself!
Additionally, Newton’s research also leads us to understand the muse for F.J.
O’Brien (Newton, page 224):
But farther, that this discontinuity of parts is the principal Cause of the opacity of
Bodies, will appear by considering, that opake Substances become transparent by
filling their Pores with any Substance of equal or almost equal density with their
parts... so, on the contrary, the most transparent Substances, may, by evacuating
their Pores, or separating their parts, be render’d sufficiently opake; as Salts or wet
Paper, or the Oculus Mundi Stone by being dried, Horn by being scraped, Glass by
being reduced to Powder, or otherwise flawed...
Newton’s experiments led him to conclude, in essence, that objects are

only opaque due to multiple scattering between the constituent parts. A
piece of paper, for example, may be made transparent by soaking it in oil;
in reverse, a piece of clear glass may be made opaque by grinding it into
powder. But this brings us back to O’Brien’s description of visible and
invisible objects:“A certain chemical coarseness is all that prevents its being
so entirely transparent as to be totally invisible.” It seems quite reasonable
that O’Brien’s first physical explanation of invisibility is based on Newton’s
views on the nature of light/matter interactions!
9. CONCLUDING REMARKS
We have seen in this article that invisibility has a long and varied history
in both science and science fiction. The observations of early science fiction
visionaries highlight how far the science of cloaking, and more generally
transformation optics, have come. Though science fiction has anticipated
many of the discoveries in optical physics, it is fair to say that the science has
now surpassed it. With concepts such as optical black holes, temporal cloaks,
external cloaks, and perfect lenses, the science has gone beyond the wildest
dreams of many a science fiction author.
It would be dangerous to try and predict where the study of transfor-
mation optics and cloaking will go even within the next few years. It seems
clear that many challenges remain in the fabrication of suitable metamaterials

for cloaking devices, and it is to be expected that much effort will be put
into the development of bulk three-dimensional cloaking materials. On the
theoretical side, fundamental causal limitations on cloaks remain an obstacle
to progress; however, alternative cloaking strategies exist which can avoid
this barrier, and it seems likely that other schemes will come to light.
At the very least, it seems likely that the ideas and lessons learned from
the optics of cloaking will inspire a new generation of optical devices with
unexpected and useful applications.
ACKNOWLEDGMENTS
The author would like to thank Elisa Hurwitz, Charlotte Stahl and Dr.Yalong Gu for a careful
reading of the manuscript, and to Susumu Tachi of Tokyo University and David Smith of
Duke University for the use of their photographs. This article is dedicated to Milo.
REFERENCES
Alam,M.-R. (2012). Broadband cloaking in stratified seas. Physical Review Letters 108,084502.
Alexopoulos, N., & Uzunoglu, N. (1978). Electromagnetic scattering from active objects:
Invisible scatterers. Applied Optics 17, 235–239.
Alù, A. (2009). Mantle cloak: Invisibility induced by a surface. Physical Review B 80, 245115.
Alù, A., & Engheta, N. (2005). Achieving transparency with plasmonic and metamaterial
coatings. Physical Review E 72, 016623.
Alù,A.,& Engheta,N. (2008). Multifrequency optical invisibility cloak with layered plasmonic
shells. Physical Review Letters 100, 113901.
Alù,A., & Engheta, N. (2009). Cloaking a sensor. Physical Review Letters 102, 233901.
Alù, A., & Engheta, N. (2010). Cloaked near-field scanning optical microscope tip for non-
invasive near-field imaging. Physical Review Letters 105, 263906.
Arslanagic, S., & Ziolkowski, R. (2010). Active coated nano-particle excited by an arbitrarily
located electric Hertzian dipoleresonance and transparency effects. Journal of the Optical 12,
024014.
Arslanagic, S., & Ziolkowski, R. (2011). Active coated nanoparticles: Impact of plasmonic
material choice. Applied Physics A 103, 795–798.
Ball, P. (2006), Invisibility cloaks are in sight. Nature News. <http://bit.ly/YTlMSE>.
Baltes, H., & Hoenders, B. (1978). A nonradiating linear combination of quasihomogeneous
current correlations. Physics Letters A 69, 249–250.
Baumeier, B., Leskkova,T., & Maradudin, A. (2009). Cloaking from surface plasmon polari-
tons by a circular array of point scatterers. Physical Review Letters 103, 246803.
Bender, C., & Boettcher, S. (1998). Real spectra in non-Hermitian Hamiltonians having PT
symmetry. Physical Review Letters 80, 5243–5246.
Berry, M. (2008). Optical lattices with PT symmetry are not transparent. Journal of Physics A
41, 244007.
Berry, M., Foley, J., Gbur, G., & Wolf, E. (1998). Nonpropagating string excitations. The
American Journal of Physics 66, 121–123.
Bertero, M., & Boccacci, P. (1998). Introduction to inverse problems in imaging. Bristol and
Philadelphia: Institute of Physics Publishing.
Bleistein, N., & Cohen, J. (1977). Nonuniqueness in the inverse source problem in acoustics
and electromagnetics. Journal of Mathematical Physics 18, 194–201.
108 Greg Gbur
Bohm, D., & Weinstein, M. (1948). The self-oscillations of a charged particle. Physical Review
74, 1789–1798.
Bohr, N. (1913). On the constitution of atoms and molecules. Philosophical Magazine 26,
1–25.
Born, M., & Wolf, E. (1999). Principles of optics (7th ed.). Cambridge: Cambridge University
Press.
Brun, M., Guenneau, S., & Movchan, A. (2009). Achieving control of in-plane elastic
waves. Applied Physics Letters 94, 061903.
Cai, W., Chettiar, U., Kildishev, A., & Shalaev, V. (2007). Optical cloaking with metamate-
rials. Nature Photonics 1, 224–227.
Cai, W., Chettiar, U., Kildishev, A., Shalaev, V., & Milton, G. (2007). Nonmagnetic cloak
with minimized scattering. Applied Physics Letters 91, 111105.
Castaldi, G., Gallina, I., Galdi, V., Alù, A., & Engheta, N. (2009). Cloak/anti-cloak interac-
tions. Optics Express 17, 3101–3114.
Chen, H., & Chan, C. (2007). Acoustic cloaking in three dimensions using acoustic meta-
materials. Applied Physics Letters 91, 183518.
Chen, H., Leonhardt, U., & Tyc, T. (2011). Conformal cloak for waves. Physical Review A 83,
055801.
Chen, H., Liang, Z.,Yao, P., Jiang, X., Ma, H., & Chan, C. (2007). Extending the bandwidth
of electromagnetic cloaks. Physical Review B 76, 241104.
Chen, H., Luo, X., Ma, H., & Chan, C. (2008). The anti-cloak. Optics Express 16, 14603–
14608.
Chen, H., Miao, R.-X., & Li, M. (2010). Transformation optics that mimics the system
outside a Schwarzschild black hole. Optics Express 18, 15183–15188.
Chen, H.,Wu, B.-I., Zhang, B., & Kong, J. (2007). Electromagnetic wave interactions with
a metamaterial cloak. Physical Review Letters 99, 063903.
Chen, X., Luo,Y., Zhang, J., Jiang, K., Pendry, J., & Zhang, S. (2011). Macroscopic invisibi-
lity cloaking of visible light. Nature Communications 2, 1–6.
Chew, H. (1976). Anomalous scattering and invisible bodies in the Rayleigh–Gans approxi-
mation. Journal of the Optical Society of America 66, 493–494.
Chew, H., & Kerker, M. (1976). Abnormally low electromagnetic scattering cross sections.
Journal of the Optical Society of America 66, 445–449.
Christiansen, C. (1884). Untersuchungen über die optischen Eigenschaften von fein ver-
theilten Körpern. Annalen der Physik 23, 298–306.
Coblentz,W. (1904). Optical notes. Physical Review 19, 89–97.
Cummer, S., Popa, B.-I., Schurig, D., & Smith, D. (2006). Full-wave simulations of electro-
magnetic cloaking structures. Physical Review E 74, 036621.
Cummer, S., Popa, B.-I., Schurig, D., Smith, D., Pendry, J., Rahm, M. et al. (2008). Scattering
theory derivation of a 3D acoustic cloaking shell. Physical Review Letters 100, 024301.
Cummer, S., & Schurig, D. (2007). One path to acoustic cloaking, New Journal of Physics
9, 45.
Denardo, B. (1998). A simple explanation of simple nonradiating sources in one dimension—
Comment on nonpropagating string excitations. The American Journal of Physics 66, 1020–
1021.
Denardo, B., & Miller, G. (2003). Quasi-nonpropagating wave sources in one dimension.
The American Journal of Physics 71, 778–782.
Devaney, A. (1978). Nonuniqueness in the inverse scattering problem. Journal of Mathema-
tical Physics 19, 1526–1531.
Devaney, A. (2004). Nonradiating surface sources. Journal of the Optical Society of America A
21, 2216–2222.
Devaney, A. (2012). Mathematical foundations of imaging. Tomography and wavefield inversion.
Cambridge: Cambridge University Press.
Devaney, A., & Wolf, E. (1973). Radiating and nonradiating classical current distributions
and the fields they generate. Physical Review D 8, 1044–1047.
Devaney, A., & Wolf, E. (1984). Non-radiating stochastic scalar sources. In L. Mandel, &
E. Wolf (Eds.), Coherence and quantum optics V (pp. 417–421). New York: Plenum Press.
Edwards, B., Alù, A., Silveirinha, M., & Engheta, N. (2009). Experimental verification of
plasmonic cloaking at microwave frequencies with metamaterials. Physical Review Letters
103, 153901.
Ehrenfest, P. (1910). Ungleichförmige Elektrizitätsbewegungen ohne Magnet- und
Strahlungsfeld. Physik in unserer Zeit 11, 708–709.
Ergin, T., Stenger, N., Brenner, P., Pendry, J., & Wegener, M. (2010). Three-dimensional
invisibility cloak at optical wavelengths. Science 328, 337–339.
Essl, G. (2004). Trapping and steering on lattice strings:Virtual slow waves and directional
and nonpropagating excitations. Physical Review E 69, 066601.
Farhat, M., Enoch, S., Guenneau, S., & Movchan, A. (2008). Broadband cylindrical acoustic
cloak for linear surface waves in a fluid. Physical Review Letters 101, 134501.
Fischer, J., Ergin, T., & Wegener, M. (2011). Three-dimensional polarization-independent
visible-frequency carpet invisibility cloak. Optics Letters 36, 2059–2061.
Friberg, A.T. (1978). On the question of the existence of nonradiating primary planar sources
of finite extent. Journal of the Optical Society of America 68, 1281–1283.
Fridman, M., Farsi, A., Okawachi, Y., & Gaeta, A. (2012). Demonstration of temporal
cloaking. Nature 481, 62–65.
Friedlander, F. (1973). An inverse problem for radiation fields. Proceedings of the London Math-
ematical Society 2, 551–576.
Fuks, B. (1963). Introduction to the theory of analytic functions of several complex variables. Provi-
dence, R.I.: American Mathematical Society.
Gabrielli, L., Cardenas, J., Poitras, C., & Lipson, M. (2009). Silicon nanostructure cloak oper-
ating at optical frequencies. Nature Photonics 3, 461–463.
Gamliel, A., Kim, K., Nachman, A., & Wolf, E. (1989). A new method for specifying
nonradiating monochromatic sources and their fields. Journal of the Optical Society of America
A 6, 1388–1393.
García-Meca, C., Tung, M., Galán, J., Ortuño, R., Rodríguez-Fortuno, F., Martí, J. et al.
(2011). Squeezing and expanding light without reflections via transformation optics. Optics
Express 19, 3562–3575.
Gbur, G. (2001). Nonradiating sources and the inverse source problem. PhD thesis. University of
Rochester.
Gbur, G. (2003). Nonradiating sources and other invisible objects. In E. Wolf (Ed.), Progress
in optics (Vol. 45, pp. 273–315). Amsterdam: Elsevier.
Gbur, G., Foley, J., & Wolf, E. (1999). Nonpropagating string excitations—finite length and
damped strings. Wave Motion 30, 125–134.
Gbur, G., & James, D. (2000). Unpolarized sources that generate highly polarized fields
outside the source. Journal of Modern Optics 47, 1171–1177.
Gbur, G., & Wolf, E. (1997). Sources of arbitrary states of coherence that generate comple-
tely coherent fields outside the source. Optics Letters 22, 943–945.
Genov, D., Zhang, S., & Zhang, X. (2009). Mimicking celestial mechanics in metamaterials.
Nature Physics 5, 687–692.
Goedecke, G. (1964). Classically radiationless motions and possible implications for quantum
theory. Physical Review 135, B281–B288.
Gömöry, F., Solovyov, M., Šouc, J., Navau, C., Prat-Camps, J., & Sanchez,A. (2012). Experi-
mental realization of a magnetic cloak. Science 335, 1466–1468.
Gonzalez, R. (2011), A chemical that can turn your organs transparent, io9.
<http://bit.ly/qnfwtj>.
110 Greg Gbur
Grant, A., & Hapgood, F. (2009), Metamaterial revolution: The new science of making
anything disappear, Discover Magazine <http://bit.ly/2pkrp>.
Greenleaf, A., Kurylev,Y., Lassas, M., & Uhlmann, G. (2007a). Electromagnetic wormholes
and virtual magnetic monopoles from metamaterials. Physical Review Letters 99, 183901.
Greenleaf, A., Kurylev, Y., Lassas, M., & Uhlmann, G. (2007b). Full-wave invisibility of
active devices at all frequencies. Communications in Mathematical Physics 275, 749–789.
Greenleaf, A., Kurylev, Y., Lassas, M., & Uhlmann, G. (2008a). Approximate quantum
cloaking and almost-trapped states. Physical Review Letters 101, 220404.
Greenleaf, A., Kurylev, Y., Lassas, M., & Uhlmann, G. (2008b). Isotropic transformation
optics:Approximate acoustic and quantum cloaking. New Journal of Physics 10, 115024.
Greenleaf, A., Lassas, M., & Uhlmann, G. (2003). Anisotropic conductivities that cannot be
detected by EIT. Physiological Measurement 24, 413–419.
Guenneau, S., Amra, C., & Veynante, D. (2012). Transformation thermodynamics: Cloaking
and concentrating heat flux. Optics Express 20, 8207–8218.
Hadamard, J. (1902). Sur les problèmes aux dérivées partielles et leur signification physique.
Bull University Princeton 13, 49.
Hakansson, A. (2007). Cloaking of objects from electromagnetic fields by inverse design of
scattering optical elements. Optics Express 15, 4328–4334.
Hama, H., Kurokawa, H., Kawano, H., Ando, R., Shimogori, T., Noda, H. et al. (2011).
Scale: A chemical approach for fluorescence imaging and reconstruction of transparent
mouse brain. Nature Neuroscience 14, 1481–1488.
Hashemi, H., Qiu, C.-W., McCauley, A., Joannopoulos, J., & Johnson, S. (2012). Diameter-
bandwidth product limitation of isolated-object cloaking. Physical Review A 86, 013804.
Hellberg, R., Karlsson, A., & Thärning, P. (1992). Non-reflecting dispersive media. Smart
Materials and Structures 1, 341–346.
Hendi, A., Henn, J., & Leonhardt, U. (2006). Ambiguities in the scattering tomography
for central potentials. Physical Review Letters 97, 073902.
Hoenders, B., & Baltes, H. (1979). The scalar theory of nonradiating partially coherent
sources. Lettere al Nuovo Cimento 23, 206–208.
Hoenders, B. J. (1997), Existence of invisible nonscattering objects and nonradiating sources.
Journal of the Optical Society of America A 14, 262–266.
Hornyak, T. (2012), Could a see-through toyota prius prevent accidents? CNET.
<http://cnet.co/t1ensb>.
Jeans, J. (1906), On the constitution of the atom. Philosophical Magazine 11, 604–607.
Jiang, W., Ma, H., Cheng, Q. and Cui, T. (2010), Illusion media: Generating virtual objects
using realizable metamaterials. Applied Physics Letters 96, 121910.
Joannopoulos, J., Meade, R., & Winn, J. (1995). Photonic crystals. Princeton, NJ: Princeton
University Press.
Kak, A., & Slaney, M. (1988). Principles of computerized tomographic imaging. New York: IEEE
Press.
Kay, I., & Moses, H. (1956). Reflectionless transmission through dielectrics and scattering
potentials. Journal of Applied Physics 27, 1503–1508.
Kerker, M. (1975a). Erratum. Journal of the Optical Society of America 65, 1085.
Kerker, M. (1975b). Invisible bodies. Journal of the Optical Society of America 65, 376–379.
Kerker, M. (1978), Electromagnetic scattering from active objects. Applied Optics 17, 3337–
3339.
Kildishev, A., Cai, W., Chettiar, U., & Shalaev, V. (2008). Transformation optics: approac-
hing broadband electromagnetic cloaking. New Journal of Physics 10, 115029.
Kim, K., & Wolf, E. (1986). Non-radiating monochromatic sources and their fields. Optics
Communications 59, 1–6.
Kim, S.-H., & Das, M. (2012). Seismic waveguide of metamaterials. Modern Physics Letters B
26, 1250105.
Lai, Y., Chen, H., Zhang, Z.-Q., & Chan, C. (2009). Complementary media invisibility
cloak that cloaks objects at a distance outside the cloaking shell. Physical Review Letters 102,
093901.
Lai,Y., Ng, J., Chen, H., Han, D., Xiao, J., Zhang, Z.-Q. et al. (2009). Illusion optics: The
optical transformation of an object into another object. Physical Review Letters 102, 253902.
Landy, N., & Smith, D. (2013). A full-parameter unidirectional metamaterial cloak for
microwaves. Nature Materials 12, 25–28.
Lee, J., Blair, J., Tamma,V., Wu, Q., Rhee, S., Summer, C. et al. (2009). Direct visualization
of optical frequency invisibility cloak based on silicon nanorod array. Optics Express 17,
12922–12928.
Lenard, P. (1903), Über die Absorption der Kathodenstrahlen verschiedener Geschwind-
igkeit. Annals of Physics 12, 714–744.
Leonhardt, U. (2006a). Notes on conformal invisibility devices. New Journal of Physics 8, 118.
Leonhardt, U. (2006b). Optical conformal mapping. Science 312, 1777–1780.
Leonhardt, U. (2009), Perfect imaging without negative refraction. New Journal of Physics 11,
093040.
Leonhardt, U., & Philbin,T. (2006). General relativity in electrical engineering. New Journal
of Physics 8, 247.
Leonhardt, U., & Philbin, T. (2009). Transformation optics and the geometry of light. In
E. Wolf (Ed.), Progress in optics (Vol. 53, pp. 69–152). Amsterdam: Elsevier.
Leonhardt, U., & Philbin, T. (2010). Geometry and light:The science of invisibility. New York:
Dover Publications.
Li, J., & Pendry, J. (2008). Hiding under the carpet: A new strategy for cloaking. Physical
Review Letters 101, 203901.
Lie, R., Ji, C., Mock, J., Chin, J., Cui, T., & Smith, D. (2009). Broadband ground-plane cloak.
Science 323, 366–369.
Lin, Z., Ramezani, H., Eichelkraut, T., Kottos, T., Cao, H., & Christodoulides, D. (2011).
Unidirectional invisibility induced by PT-symmetric periodic structures. Physical Review
Letters 106, 213901.
Litchinitser, N., Gabitov, I., Maimistov, A., & Shalaev, V. (2008). Negative refractive index
metamaterials in optics. In E. Wolf (Ed.), Progress in optics (Vol. 51, pp. 1–67). Amsterdam:
Elsevier.
Longhi, S. (2011), Invisibility in PT-symmetric complex crystals. Journal of Physics A 44,
485302.
Luo, X., Yang, T., Gu, Y., Chen, H., & Ma, H. (2009). Conceal an entrance by means of
superscatterer. Applied Physics Letters 94, 223513.
Luo, Y., He, L.-X., Wang, Y., Chan, H., & Zhu, S.-Z. (2011). Changing the scattering of
sheltered targets. Physical Review A 83, 043809.
Magnus, F., Wood, B., Moore, J., Morrison, K., Perkins, G., Fyson, J. et al. (2008). A d.c.
magnetic material. Nature Materials 7, 295–297.
Makris, K., El-Ganainy, R., Christodoulides, D., & Musslimani, Z. (2008). Beam dynamics
in PT symmetric optical lattices. Physical Review Letters 100, 103904.
Markov, P., Valentine, J., & Weiss, S. (2012). Fiber-to-chip coupler designed using an optical
transformation. Optics Express 20, 14705–14713.
McCall, M., Favaro, A., Kinsler, P., & Boardman, A. (2011). A spacetime cloak, or a history
editor. Journal of Optics 13, 024003.
Mercedes-Benz (2012). Mercedes-benz invisible car campaign. <http://bit.ly/yv0hka>.
Miller, D. (2006). On perfect cloaking. Optics Express 14, 12457–12466.
Milton, G., Briane, M., & Willis, J. (2006). On cloaking for elasticity and physical equations
with a transformation invariant form. New Journal of Physics 8, 248.
Mizuno, K., Ishii, J., Kishida, H., Hayamizu, Y., Yasuda, S., Futaba, D. et al. (2009). A black
body absorber from vertically aligned single-walled carbon nanotubes. Proceedings of the
National Academy of Sciences 106, 6044–6047.
112 Greg Gbur
Moses, D., Gan, C., & Gbur, G. (2009). Directional, nonpropagating, and polychromatic
excitations in one-dimensional wave systems. Physical Review E 79, 026606.
Nachman, A. (1988). Reconstructions from boundary measurements. Annals of Mathematics
128, 531–576.
Nagaoka, H. (1904). Kinetics of a system of particles illustrating the line and the band
spectrum and the phenomena of radioactivity. Philosophical Magazine 7, 445–455.
Newton, I. (1721). Opticks. London:William and John Innys.
Nogami, Y., & Toyama, F. (1998). Reflectionless potentials for the one-dimensional Dirac
equation: Pseudoscalar potentials. Physical Review A 57, 93–97.
Ochiai, T., Leonhardt, U., & Nacher, J. (2008). A novel design of dielectric perfect invisibility
devices. Journal of Mathematical Physics 49, 032903.
Parnell, W. (2012). Nonlinear pre-stress for cloaking from antiplane elastic waves. Proceedings
of the Royal Society of London 468, 563–580.
Pendry, J. (2000). Negative refraction makes a perfect lens. Physical Review Letters 85, 3966–
3969.
Pendry, J., Holden, A., Robbins, D., & Stewart,W. (1999). Magnetism from conductors and
enhanced nonlinear phenomena. IEEE Transactions on Microwave Theory and Techniques 47,
2075–2084.
Pendry, J., Schurig, D., & Smith, D. (2006). Controlling electromagnetic fields. Science 312,
1780–1782.
Philbin,T., Kuklewicz, C., Robertson, S., Hill, S., König, F., & Leonhardt, U. (2008). Fiber-
optical analog of the event horizon. Science 319, 1367–1370.
Qiu,C.-W.,Hu,L.,Xu,X.,& Feng,Y. (2009). Spherical cloaking with homogeneous isotropic
multilayered structures. Physical Review E 79, 047602.
Rahm, M., Cummer, S., Schurig, D., Pendry, J., & Smith, D. (2008). Optical design of reflec-
tionless complex media by finite embedded coordinate transformations. Physical Review
Letters 100, 063903.
Rayleigh, L. (1884). Geometrical optics. In Encyclopaedia Britannica (Vol. 17, pp. 798–807).
Regensburger, A., Bersch, C., Miri, M.-A., Onishchukov, G., Christodoulides, D., &
Peschel, U. (2012). Parity-time synthetic photonic lattices. Nature 488, 167–171.
Roberts, D., Rahm, M., Pendry, J., & Smith, D. (2008). Transformtion-optical design of sharp
waveguide bends and corners. Applied Physics Letters 93, 251111.
Sanchez, A., Navau, C., Prat-Camps, J., & Chen, D.-X. (2011). Antimagnets: Controlling
magnetic fields with superconductor-metamaterial hybrids. New Journal of Physics 13,
093034.
Schott, G. (1906). On the electron theory of matter and the explanation of fine spectrum
lines and of gravitation. Philosophical Magazine 12, 21–29.
Schott, G. (1933). The electromagnetic field of a moving uniformly and rigidly electrified
sphere and its radiationless orbits. Philosophical Magazine 15, 752–761.
Schott,G. (1936a).The electromagnetic field due to a uniformly and rigidly electrified sphere
in spinless accelerated motion and its mechanical reaction on the sphere, I. Proceedings of
the Royal Society of London 156, 471–486.
Schott, G. (1936b). On the spinless rectilinear motion of a uniformly and rigidly electrified
sphere, II. Proceedings of the Royal Society of London 156, 487–503.
Schott, G. (1937a).The general motion of a spinning uniformly and rigidly electrified sphere,
III. Proceedings of the Royal Society of London 159, 548–570.
Schott, G. (1937b). The uniform circular motion with invariable normal spin of a rigidly and
uniformly electrified sphere, IV. Proceedings of the Royal Society of London 159, 570–591.
Schurig, D., Mock, J. J., Justice, B. J., Cummer, S. A., Pendry, J. B., Starr, A. F. et al. (2006).
Metamaterial electromagnetic cloak at microwave frequencies. Science 314, 977–980.
Setälä, T., Hakkarainen, T., Friberg, A., & Hoenders, B. (2010). Object-dependent cloaking
in the first-order Born approximation. Physical Review A 82, 013814.
Shalaev, V. (2007). Optical negative-index metamaterials. Nature Photonics 1, 41–47.

Shelby, R., Smith, D., & Schultz, S. (2001). Experimental verification of a negative index of
refraction. Science 292, 77–79.
Silveirinha, M., Alù, A., & Engheta, N. (2007). Parallel-plate metamaterials for cloaking
structures. Physical Review E 75, 036603.
Smith, D., Pendry, J., & Wiltshire, M. (2004). Metamaterials and negative refractive index.
Science 305, 788–792.
Smolyaninov, I., Hung, Y., & Davis, C. (2008). Two-dimensional metamaterial structure
exhibiting reduced visibility at 500 nm. Optics Letters 33, 1342–1344.
Sommerfeld, A. (1964), Optics. New York and London:Academic Press.
Stark, J. (1910). Prinzipien der Atomdynamik: Die Elektrischen quanten. Hirsel, Leipzig:Verlag
von S.
Stenger, N., Wilhelm, M., & Wegener, M. (2012). Experiments on elastic cloaking in thin
plates. Physical Review Letters 108, 014301.
Sullivan, D. (2000), Electromagnetic simulation using the FDTD method. New Jersey:Wiley-IEEE
Press.
Tachi, S. (2003). Telexistence and retro-reflective projection technology (RPT). In S. Richir,
P. Richard, B. Taravel, (Eds.), Proceedings of the 5th virtual reality international conference
(pp. 69/1–69/9). Angers: ISTIA Innovation.
Thomson, J. (1904). On the structure of the atom: an investigation of the stability and periods
of oscillation of a number of corpuscles arranged at equal intervals around the circumference
of a circle; with application of the results to the theory of atomic structure. Philosophical
Magazine 7, 237–265.
Torrent, D., & Sánchez-Dehesa, J. (2008). Acoustic cloaking in two dimensions: A feasible
approach. New Journal of Physics 10, 063015.
Toyama, F., Nogami,Y., & Zhao, Z. (1993). Relativistic extension of the Kay-Moses method
for constructing transparent potentials in quantum mechanics. Physical Review A 47, 897–
902.
Tricarico, S., Bilotti, F., Alù, A., & Vegni, L. (2010). Plasmonic cloaking for irregular objects
with anisotropic scattering properties. Physical Review E 81, 026602.
Urzhumov,Y., & Smith, D. (2012). Low-loss-directional cloaks without superluminal velocity
or magnetic response. Optics Letters 37, 4471–4473.
Valentine, J., Li, J., Zentgraf, T., Bartal, G., & Zhang, X. (2009). An optical cloak made of
dielectrics. Nature Materials 8, 568–571.
Valentine, J., Zhang, S., Zentgraf,T., Ulin-Avila, E., Genov, D., Bartal, G. et al. (2008). Three-
dimensional optical metamaterial with a negative refractive index. Nature 455, 376–379.
Vasquez, F., Milton, G., & Onofrei, D. (2009). Active exterior cloaking for the 2D Laplace and
Helmholtz equations. Physical Review Letters 103, 073901.
Vesalago, V. (1968). The electrodynamics of substances with simultaneously negative values
of and μ. Soviet Physics Uspekhi 10, 509–514.
Vlaardingerbroek, M., & den Boer, J. (2003). Magnetic resonance imaging: Theory and practice (3rd
ed.). Berlin: Springer.
Wolf, E. (1996). Principles and development of diffraction tomography. In: A. Consortini
(Ed.),Trends in optics (pp. 83–110). San Diego:Academic Press.
Wolf, E., & Habashy, T. (1993). Invisible bodies and uniqueness of the inverse scattering
problem. Journal of Modern Optics 40, 785–792.
Wood, B., & Pendry, J. (2007). Metamaterials at zero frequency. Journal of Physics Condensed
Matter 19, 076208.
Wood, R. (1902). The invisibility of transparent objects. Physical Review 15, 123–124.
Xi, S., Chen, H., Wu, B.-I., & Kong, J. (2009). One-directional perfect cloak created with
homogenous material. IEEE Microwave and Wireless Components Letters, 131–133.
114 Greg Gbur
Xie,Y., Chen, H., Xu,Y., Zhu, L., Ma, H., & Dong, J.-W. (2011). An invisibility cloak using
silver nanowires. Plasmonics 6, 477–481.
Xu, S., Cheng, X., Xi, S., Zhang, R., Moser, H., Shen, Z. et al. (2012). Experimental demon-
stration of a free-space cylindrical cloak without superluminal propagation. Physical Review
Letters 109, 223903.
Xu, T., Liu, Y., Zhang, Y., Ong, C., & Ma, Y. (2012). Perfect invisibility cloaking by
isotropic media. Physical Review A 86, 043827.
Yaghjian, A., & Maci, S. (2008). Alternative derivation of electromagnetic cloaks and con-
centrators. New Journal of Physics 10, 115022.
Yang, F., Mei, Z., Jin, T., & Cui,T. (2012). Dc electric invisibility cloak. Physical Review Letters
109, 053902.
Yang, T., Chen, H., Luo, X., & Ma, H. (2008). Superscatterer: enhancement of scattering with
complementary media. Optics Express 16, 18545–18550.
Yao, J., Liu, Z., Liu,Y.,Wang,Y., Sun, C., Bartal, G. et al. (2008). Optical negative refraction in
bulk metamaterials of nanowires. Science 321, 930.
Zhang, B., Chan,T., & Wu, B.-I. (2010). Lateral shift makes a ground-plane cloak detectable.
Physical Review Letters 104, 233903.
Zhang, B., Luo,Y., Liu, X., & Barbastathis, G. (2011). Macroscopic invisibility cloak for visible
light. Physical Review Letters 106, 033901.
Zhang, S., Genov, D., Sun, C., & Zhang, X. (2008). Cloaking of matter waves. Physical Review
Letters 100, 123002.
Zhang, S., Xia, C., & Fang, N. (2011). Broadband acoustic cloak for ultrasound waves. Physical
Review Letters 106, 024301.
Zheng, H., Xiao, J., Lai,Y., & Chan, C. (2010). Exterior optical cloaking and illusions by using
active sources:A boundary element perspective. Physical Review B 81, 195116.

Invisibility Physics Past Present and Future 2013 Progress in Optics

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Invisibility Physics Past Present and Future 2013 Progress in Optics

Caricato da

Copyright:

Formati disponibili

CHAPTER TWO

Invisibility Physics: Past, Present,

Figure 1 A demonstration of the “optical camouﬂage” system. (Photographs courtesy of

Before we begin, however, we must be a little clear about the word

is being applied to improving the field-of-view of automobile drivers and

2. EARLY CONCEPTS OF INVISIBILITY

Without proof, Lord Rayleigh introduced an unusual proposition: a non-

only held back by the pressure of the electromagnetic “aether.” Curiously,

Ehrenfest’s intriguing observations seem to have been quickly forgot-

could be explained by the use of a purely classical formalism. In 1933, Schott

where kn ≡ ωn /c. The coefficients Jn (kn s) may be evaluated by the Fourier

Schott continued working on radiationless motions until his death, pro-

Figure 3 Illustration of the notation related to a monochromatic radiation source

Theorem 2.5. A bounded nonradiating source distribution q(r) of finite

3. NONSCATTERING OBJECTS AND INVERSE PROBLEMS

The first three-dimensional “invisible bodies” were introduced by Kerker

object space A image space

noise result nonexistence

Figure 4 Illustration of various types of ill-posedness.

problem, a single image can represent multiple objects. Closely related is a

Figure 5 Illustrating the scattering of a monochromatic ﬁeld from an inhomogeneous

satisfies the generalized Helmholtz equation,

[∇ 2 + n2 (r, ω)k2 ]u(r, ω) = 0. (2.17)

[∇ 2 + k2 ]u(r, ω) = k2 [1 − n2 (r, ω)]u(r, ω). (2.18)

Defining the scattering potential as

[∇ 2 + k2 ]u(r, ω) = −4πF(r, ω)u(r, ω). (2.20)

Equation (2.20) then becomes an inhomogeneous equation for the scattered

On comparison of this equation to those for radiation, Equations (2.10) and

It is therefore in principle possible to construct an object which is non-

in two independent articles that an object that is perfectly nonscattering for

ui (r, ω) = A0 e iks0 ·r , (2.26)

for all possible directions s0 of illumination. On substitution of this result

computationally that certain PT-symmetric photonic lattices can exhibit uni-

space. The tools of differential geometry may therefore be used to design

[∂x2 + ∂y2 + n2 (r)k2 ]u(r) = 0, (2.28)

where ∂x represents the partial derivative with respect to x, and so forth.

[4∂z∗ ∂z + n2 (r)k2 ]u(z, z∗ ) = 0. (2.29)

Let us now introduce a complex coordinate w = w(z) which is an analytic

[4∂w∗ ∂w + n2 (r)k2 ]u(w, w ∗ ) = 0, (2.30)

where n satisfies the expression

w(z) = z + a2 /z, (2.32)

with a a positive real number, results in the refractive index profile

n(x, y) = |1 − a2 /z2 |. (2.33)

The inverse transform, however, is of the form

There are a few important observations to be made about this cloak.

3 These materials will be discussed in Section 6.

elegantly by using tensor analysis,as we now briefly demonstrate. A complete

Figure 8 The virtual distortion of space created by the Pendry–Schurig–Smith cloak.

transformation results in the following diagonal values of the permittivity,

make up the difference. Though it is certainly possible to produce a material

5. OTHER CLOAKING DESIGNS

(a) (b) (c)

antiobject object ε', µ'

inspired by an active cloaking design by Miller (2006) that could hide an

Figure 12 A numerical demonstration of illusion optics, showing how the scattering

6. EXPERIMENTS AND MATERIAL PROPERTIES

Figure 14 An illustration of imaging with a perfect lens.

Negative refraction drew new interest (and controversy) when Pendry

other fields where shielding from waves is important; furthermore, transfor-

8. INVISIBILITY IN SCIENCE FICTION

look at the history of invisibility in fiction,6 and wonder whether authors

Wells, who acquired a scientific education in biology under Thomas

Clearly Wells was inspired by the discovery of X-rays (Roentgen rays)

Scale is not the only optical research anticipated by a science fiction