Reversing Light With Negative Refraction

Materials engineered to have negative permittivity and permeability demonstrate exotic behavior,
from a negative refractive index to subwavelength focusing.
John B. Pendry; David R. Smith

Physics Today 57 (6), 37–43 (2004);

https://doi.org/10.1063/1.1784272

Selectable Content List

No matches found for configured query.

CrossMark

 
View Export

Downloaded from http://pubs.aip.org/physicstoday/article-pdf/57/6/37/11083469/37_1_online.pdf

Online Citation

Related Content

Nanowire laser measures the refractive index of its surroundings

Novel Composite Medium Exhibits Reversed Electromagnetic Properties

Physics Today (May 2000)

The solid state

Physics Today (January 1952)
Reversing Light With Negative
Refraction
ments, then a negative magnetic re-
Materials engineered to have negative permittivity and sponse would exist.
permeability demonstrate exotic behavior, from a negative Although somewhat less common
refractive index to subwavelength focusing. than positive materials, negative ma-
terials are nevertheless easy to find.
Materials with negative e include met-
John B. Pendry and David R. Smith als (such as silver, gold, and aluminum)
at optical frequencies; materials with
ictor Veselago, in a paper1 published in 1968, pondered negative m include resonant ferromagnetic or antiferro-
V

Downloaded from http://pubs.aip.org/physicstoday/article-pdf/57/6/37/11083469/37_1_online.pdf

the consequences for electromagnetic waves interacting magnetic systems.
with a hypothetical material for which both the electric per- That negative material parameters occur near a reso-
mittivity e and the magnetic permeability m were simulta- nance has two important consequences. First, negative ma-
neously negative. Because no naturally occurring material terial parameters will exhibit frequency dispersion: They
or compound has ever been demonstrated with negative e will vary as a function of frequency. Second, the usable
and m, Veselago wondered whether this apparent asymme- bandwidth of negative materials will be relatively narrow
try in material properties was just happenstance or perhaps compared with positive materials. These consequences can
had a more fundamental origin. He concluded that not only help answer our initial question as to why materials with
should such materials be possible, but if ever found, they e and m both negative are not readily found. In existing ma-
would exhibit remarkable properties unlike those of any terials, the resonances that give rise to electric polariza-
known materials and would give a twist to virtually all elec- tions typically occur at very high frequencies—in the opti-
tromagnetic phenomena. Foremost among these properties cal for metals, and at least in the terahertz-to-IR region for
is a negative index of refraction. semiconductors and insulators. On the other hand, reso-
Veselago always referred to the materials as “left nances in magnetic systems typically occur at much lower
handed,” because the wave vector is antiparallel to the frequencies and usually tail off toward the THz and IR re-
usual right-handed cross product of the electric and mag- gion. In short, the fundamental electronic and magnetic
netic fields. We prefer the negative-index description. The processes that give rise to resonant phenomena in materi-
names mean the same thing, but our description appeals als simply do not occur at the same frequencies, although
more to everyday intuition and is less likely to be confused no physical law would preclude such overlap.
with chirality, an entirely different phenomenon.
Why are there no materials with negative e and m? Metamaterials extend material response
One first needs to understand what it means to have a neg- Because of the seeming separation in frequency between
ative e or m and how negative values occur in materials. electric and magnetic resonant phenomena, Veselago’s
The Drude–Lorentz model of a material is a good starting analysis of materials with e and m both negative might
point, because it conceptually replaces the atoms and mol- have remained a curious exercise in electromagnetic the-
ecules of a real material by a set of harmonically bound ory. However, in the mid-1990s, researchers began looking
electron oscillators resonant at some frequency w0. At fre- into the possibility of engineering artificial materials to
quencies far below w0, an applied electric field displaces have a tailored electromagnetic response. Although the
the electrons from the positive cores and induces a polar- field of artificial materials dates back to the 1940s, ad-
ization in the same direction as the applied field. At fre- vances in fabrication and computation—coupled with the
quencies near resonance, the induced polarization be- emerging awareness of the importance of negative mate-
comes very large, as is typical in resonance phenomena; rials—led to a resurgence of effort in developing new struc-
the large response represents accumulation of energy over tures with novel material properties.
many cycles, such that a considerable amount of energy is To form an artificial material, we start with a collec-
stored in the resonator (in this case, the medium) relative tion of repeated elements designed to have a strong re-
to the driving field. So large is this stored energy that even sponse to applied electromagnetic fields. As long as the size
changing the sign of the applied electric field has little ef- and spacing of the elements are much smaller than the elec-
fect on the polarization near resonance! That is, as the fre- tromagnetic wavelengths of interest, incident radiation
quency of the driving electric field is swept through the cannot distinguish the collection of elements from a homo-
resonance, the polarization flips from in-phase to out-of- geneous material. We can thus conceptually replace the in-
phase with the driving field, and the material exhibits a homogeneous composite by a continuous material de-
negative response. If instead of electrons the material re- scribed by material parameters e and m. At lower
sponse were due to harmonically bound magnetic mo- frequencies, conductors are excellent candidates from
which to form artificial materials, because their response
John Pendry is a professor of physics at Imperial College to electromagnetic fields is large.
London. David Smith is an adjunct professor of physics at the A metamaterial mimicking the Drude–Lorentz model
University of California, San Diego. can be straightforwardly achieved with an array of wire

© 2004 American Institute of Physics, S-0031-9228-0406-010-7 June 2004 Physics Today 37

 a
b

e
m m<0
+1
e<0

–1

–10

TRANSMITTED POWER (dB)

Figure 1. Metamaterials can be designed to create negative –20
refraction. (a) In this example of a metamaterial used in
microwave experiments, unit cells consist of a split-ring

Downloaded from http://pubs.aip.org/physicstoday/article-pdf/57/6/37/11083469/37_1_online.pdf

resonator and a wire spanning the cell, just visible on the
reverse of the supporting sheets. (b) Schematic variation of –30
e (green) and m (blue) with frequency. The shaded green
and blue bands denote negative regions for e and m, re-
spectively. (c) The transmitted power spectra16 for a meta-
material of cut wires (green), a metamaterial of split ring –40
resonators (blue), and a metamaterial combining wires and
split-ring resonators (red). The yellow band, corresponding
to the red curve’s transmission window, indicates the region –50
of negative refractive index. 8 9 10 11 12
FREQUENCY (GHz)

elements into which cuts are periodically introduced. The quency set by the geometry of the element. Such a split-
effective permittivity for the cut-wire medium has the form ring resonator (SRR), in its various forms, can be viewed
as the metamaterial equivalent of a magnetic atom. The
(1) SRR medium could be described by the resonant form
Fw2
where the plasma frequency wp and the resonance fre- m(w) ⊂ 1 ⊗
. (2)
quency w0 are determined only by the geometry of the lat- w ⊗ w20 ⊕ iwG 2

tice rather than by the charge, effective mass, and density The wire medium and the SRR medium represent two
of electrons, as is the case in naturally occurring materi- basic building blocks—one electric, the other magnetic—
als. For w0 < w < wp, the permittivity is negative and, be- for a large range of metamaterial response, including Vese-
cause the resonant frequency can be set to virtually any lago’s hypothesized material (see figure 1).
value in a metamaterial, phenomena—including negative
e—usually found at optical frequencies can be reproduced Negative refraction
at frequencies as low as a few megahertz. Structures are Maxwell’s equations determine how electromagnetic
often designed with continuous wires so that w0 ⊂ 0. waves propagate within a medium and can be solved to ar-
The path to achieving magnetic response from con- rive at a wave equation of the form
ductors is slightly different. From the basic definition of a
] 2E(x,t) ] 2E(x,t)
magnetic dipole moment, m ⊂ 1/2*r × j d3r for current den- ⊂em . (3)
]x2 ]t2
sity j, one can see that a magnetic response can be obtained
if local currents can be induced to circulate in closed loops. In this equation, e and m enter as a product, so it would
Introducing a resonance into the element should enable a not appear to matter whether their signs were both posi-
very strong magnetic response, potentially one that can tive or both negative. Indeed, solutions of the wave equa-
lead to a negative m. tion have the form exp[i(nkd – wt)], where n ⊂ √em is the
In 1999, one of us (Pendry) and colleagues proposed a refractive index. Propagating solutions exist in the mate-
variety of structures that, they predicted, would form mag- rial whether e and m are both positive or both negative. So
netic metamaterials.2 Those structures consisted of loops what, if anything, is the difference between positive and
or tubes of conductor with a gap inserted. One can view negative materials?
such structures as miniature circuits: A time-varying mag- It turns out that one needs to be more careful in tak-
netic field induces an electromotive force in the plane of ing the square root, because e and m are analytic functions
the element, driving currents within the conductor. A gap whose values are generally complex. There is an ambigu-
in the plane of the structure introduces capacitance into ity in the sign of the square root that is resolved by a proper
the planar circuit and gives rise to a resonance at a fre- analysis. For example, if instead of writing e ⊂ –1 and

38 June 2004 Physics Today http://www.physicstoday.org

a b Figure 2. A negative-index material
will refract light through a negative
angle. (a) In this simulation17 of a
Snell’s law experiment, a negative-
index wedge with e ⊂ –1 and m ⊂ –1
deflects an electromagnetic beam by a
negative angle relative to the surface
normal: The beam emerges on the
same side of the surface normal as the
incident beam. Color represents inten-
sity: red, highest; blue, lowest. (b) A
positive-index wedge, in contrast, will
positively refract the same beam. Red
lines trace the path of the beams, and
the surface normals are shown in
black. Experiments confirm this behav-
ior. (c) The deflection angle (horizontal
axis) observed for a beam traversing a
negative wedge as a function of fre-
quency (vertical axis). (d) The deflec-
c d tion angle observed for a positive-
14 index Teflon® wedge as a function of
frequency. In the negative wedge there

Downloaded from http://pubs.aip.org/physicstoday/article-pdf/57/6/37/11083469/37_1_online.pdf

is strong dispersion with frequency:
The condition e ⊂ –1, m ⊂ –1 is real-
ized only over a narrow bandwidth
FREQUENCY (GHz)

around 12 GHz.

12

sample and strikes the second inter-

10 face at an angle. Because of the differ-
–90° –60° –30° 0° 30° 60° 90° ence in refractive index between the
DEFLECTION ANGLE material and free space, the beam exits
the wedge deflected by some angle
from the direction of incidence.
m ⊂ –1 we write e ⊂ exp(ip) and m ⊂ exp(ip), then n ⊂ One might imagine that an experimental determina-
√em ⊂ exp(ip/2)exp(ip/2) ⊂ exp(ip) ⊂ –1. The important tion of Snell’s law would be a simple matter. The peculi-
step is that the square root of either e or m alone must arities of metamaterials, however, add a layer of complex-
have a positive imaginary part—a necessity for a passive ity that renders the experimental confirmation somewhat
material. more difficult. Present samples, based on SRRs and wires,
This briefly stated argument shows why the material are frequency dispersive with fairly narrow bandwidths
Veselago pondered years ago is so unique: The index of re- and exhibit considerable loss. The first experiment show-
fraction is negative. A negative refractive index implies ing negative refraction was performed in 2001 by one of us
that the phase of a wave decreases rather than advances (Smith) and colleagues at the University of California, San
with passage through the medium. As Veselago pointed Diego.3 In an experiment similar to that depicted in figure
2, they measured the power refracted from a two-dimen-
out, this fundamental reversal of wave propagation con-
sional wedge-shaped metamaterial sample as a function of
tains important implications for nearly all electromagnetic
angle, confirming the expected properties.
phenomena. Many of the exotic effects of negative index While the UCSD data were compelling, the concept of
have been or are currently being pursued by researchers. negative index proved counterintuitive enough that many
But perhaps the most immediately accessible phenomenon other researchers needed further convincing. In 2003, An-
from an experimental or computational point of view is the drew Houck and colleagues at MIT repeated the negative-
reversal of wave refraction, illustrated in figure 2. refraction experiment on the same sort of negative-index
Snell’s law, which describes quantitatively the bending metamaterial and confirmed the original findings.4 Look-
of a wave as it enters a medium, is perhaps one of the old- ing at wedges with different angles, the MIT group showed
est and most well known of electromagnetic phenomena. In that the observed angle of refraction was consistent with
the form of a wedge refraction experiment, as depicted in Snell’s law for the metamaterial. In the same year, Clau-
figure 2, Snell’s law is also the basis for a direct measure- dio Parazzoli and coworkers at Boeing Phantom Works
ment of a material’s refractive index. In this type of exper- also confirmed the negative-refraction results in a sepa-
iment, a wave is incident normal to a wedge-shaped sam- rately designed sample.5 In their measurements, the de-
ple. The wave is transmitted through the transparent tector’s distance from the sample was significantly larger

http://www.physicstoday.org June 2004 Physics Today 39

 Figure 3. Lightweight, compact a
lenses can be designed from metama-
terials to be relatively free of aberra-
x
tion. (a) A positive-index lens with an
z
index of refraction n of 2.3. On the
right is its focusing pattern: Light trav- 1
y 0.5
eling in the +z-direction is focused at 0
the red peak. Color represents inten- –5 30
sity: red, highest; blue, lowest. (b) A
metamaterial lens with n ⊂ –1 and 0

y
20

)
(c

m
with the same radius of curvature as

(c
)
in (a), and its focusing pattern. The

z
5 10
meta-lens is much lighter than the
positive-index lens, a significant ad-
vantage for aerospace applications. b
Although both lenses have the same x
radius of curvature, the negative lens z
has a much shorter focal distance.
(c) A concave metamaterial lens, de- y
1
signed by Claudio Parazzoli and his
colleagues at Boeing Phantom 0.5
Works,18 with n  –1 at microwave 0
frequencies near 15 GHz. On the left

Downloaded from http://pubs.aip.org/physicstoday/article-pdf/57/6/37/11083469/37_1_online.pdf

–5 40
is its unit cell; on the right,
y( 20 )
a picture of the lens. cm 0
cm
)
5 0 z(
c

than for previous demonstrations. z

Although it has proven to be a x
valuable concept, a rigorously de- Lens
fined negative index of refraction y
may not necessarily be a prerequisite
for negative-refraction phenomena. z
An alternate approach to attaining
negative refraction uses the proper- y
ties of photonic crystals,6,7 materials
that lie on the transition between a
metamaterial and an ordinary structured dielectric. Pho- into a topic of materials—or metamaterials—physics. As
tonic crystals derive their properties from Bragg reflection metamaterials are being designed and improved, we are
in a periodic structure engineered in the body of a dielec- now free to consider the ramifications associated with a
tric, typically by drilling or etching holes. The periodicity negative index of refraction. That material property, per-
in photonic crystals is on the order of the wavelength, so haps because it is so simply stated, has enabled the rapid
that the distinction between refraction and diffraction is design of new electromagnetic structures—some with very
blurred. Nevertheless, with photonic crystals many novel unusual and exotic properties.
dispersion relationships can be realized, including ranges
in which the frequency disperses negatively with wave vec- A better focus
tor as required for a negative refraction. Using photonic Refraction is the phenomenon responsible for lenses and
crystals,8,9 researchers have observed focusing, as pre- similar devices that focus or shape radiation. Although
dicted for negative-index materials. usually thought of in the context of visible light, lenses are
The concept of negative refraction has also been gen- utilized throughout the electromagnetic spectrum; they
eralized to transmission-line structures, which are com- thus represent a good starting point to implement nega-
mon in electrical engineering applications. By pursuing tive index materials.
the analogy between lumped circuit elements and mate- In his early paper, Veselago noted that a negative-
rial parameters, George Eleftheriades and coworkers at index focusing lens would need to be concave rather than
the University of Toronto have demonstrated negative- convex—a seemingly trivial matter, but there is more to
refraction phenomena in microwave circuits.10 The trans- the story. For thin lenses, geometrical optics—valid for ei-
mission-line model has proven exceptionally valuable for ther positive or negative index—gives the result that the
the development of microwave devices: Tatsuo Itoh and focal length f is related to the lens’s radius of curvature,
Christophe Caloz at UCLA have applied the model to de- R, by f ⊂ R/(n – 1). The denominator in the focal-length
velop novel microwave components, including antennas, formula implies an inherent distinction between positive-
couplers, and resonators.11 and negative-index lenses: A material with n ⊂ ⊕1 does
Those experiments and applications have shown that not refract electromagnetic fields, whereas a material with
the material Veselago hypothesized more than 35 years n ⊂ –1 does. The result is that negative-index lenses can
ago can now be realized using artificially constructed be more compact, with a host of other benefits, as shown
metamaterials; the discussion of negative refractive index in figure 3.
is thus more than a theoretical curiosity. The question of Making a conventional lens with the best possible res-
whether such a material can exist has been answered, and olution requires a wide aperture. Each ray emanating from
the development of negative-index structures has turned an object, as shown in figure 4a, has wave vector compo-

40 June 2004 Physics Today http://www.physicstoday.org

 a c Figure 4. Resolution limitations. (a) For
e O –1
m O –1 good resolution, conventional lenses
n O –1 need a wide aperture to refract rays at
large angles q, but even so, they are
q z z limited in resolution by the wavelength
used. (b) The missing Fourier compo-
nents of the image are contained in the
near field, which decays exponentially
(blue curve) and makes negligible con-
tribution to the image. (c) A lens made
from a planar slab of negative-index
b d material not only brings rays to a focus
but has the capacity (d) to amplify the
near field so that it contributes to the
image. Such a negative lens thus re-
moves the wavelength limitation. How-
ever, the resonant nature of the amplifi-
cation places severe demands on
materials: They must be very low loss.

Downloaded from http://pubs.aip.org/physicstoday/article-pdf/57/6/37/11083469/37_1_online.pdf

nents along the axis of the lens, kz ⊂ k0cosq, and perpen- If by some magic we could amplify the near fields, we
dicular to the axis, kx ⊂ k0sinq, where k0 is the wavenum- could in principle recoup their contribution, but the am-
ber and q is the angle of the ray with respect to the axis. plification would have to be of just the right amount and
The axial projection kz is responsible for transporting the possibly very strong for the most localized components.
light from object to image; kx represents a Fourier compo- That is a tall order, but by a remarkable chance, a planar
nent of the image. For good resolution, the larger one can slab of negative material achieves this feat.12
make kx, the better. The best that can be achieved is kx ⊂ Figure 4c shows rays contributing to the image formed
k0, and hence the resolution limit is D  p/k0 ⊂ l/2, where by a negative slab. Just as for a conventional lens, the rays
l is the wavelength. This restriction is a huge problem in only contribute details greater than about half a wave-
many areas of optics. The feature size achieved in com- length in diameter. In contrast, the behavior of the near
puter chips and the storage capacity of DVDs, for exam- field is remarkably different, as shown in figure 4d. The
ple, are wavelength-limited. Even a modest relaxation of near field has the capacity to excite short-wavelength res-
the wavelength limitation would be of great value. onances of the negative-index surface that are akin to the
In contrast to the image, the object has no limit to its surface plasmons on the surfaces of metals such as silver.
electromagnetic details, but unfortunately not all of that Interaction with the plasmonlike excitation kicks the de-
information makes it across the lens to the image. The caying wave into a growing wave. The negative medium
problem lies with the wave vector’s z-component, which thus amplifies the wave and compensates for the decay
we can write as kz ⊂ √k02 – kx2. Evidently, for large values that occurred in an equal thickness of vacuum. The reso-
of kx, corresponding to fine details in the object, kz is nances have a finite width and the requirement of e ⊂ –1
imaginary and the waves decay exponentially as and m ⊂ –1 can be met only at one frequency because of
exp(–√kx2 – k02 z), as shown in figure 4b. By the time these the inherent dispersion of negative media. Therefore, this
so-called evanescent waves reach the image, they have super lensing effect is a narrow-band phenomenon.
negligible amplitude. For that reason, they are commonly In the case of evanescent waves, amplification does
referred to as the “near field” and the propagating rays not imply a sustained input of power. Evanescent waves
as the “far field.” carry no power and hence, in the absence of loss, a large-

a c d d
d d
Figure 5. Generalizing the perfect lens.
(a) A slab with refractive index n ⊂ –1
(black region) draws light to a perfect
focus. (b) The negative slab achieves
this focus by “unwinding” or negating
the phase acquired in passing through
free space. The phase velocity (green
arrow) advances in the positive medium
(white region) but retreats in the nega-
d tive medium. (c) Focusing can occur
through two objects that are more com-
b Energy Energy plex, provided that one is the inverse
mirror image of the other. (d) A graphi-
cal equation of optical cancellation:
Mirror-antisymmetric regions of space
optically annihilate one another. A
Phase Phase
negative medium is, in effect, a piece
of optical antimatter.

http://www.physicstoday.org June 2004 Physics Today 41

a b c

1
e O –1 e = +1
m O –1 m = +1

Figure 6. An optical paradox. (a) Ray tracing predicts that some rays, such as ray 2, will be rejected from this system, even
though the mirror theorem illustrated in figure 5 predicts that all waves should be transmitted. (b) The solution of Maxwell’s
equations for a single negative-index cylinder shows the expected diffraction pattern. (c) When the complementary layer is
added, all scattering is removed, within the accuracy of the calculation. Color represents intensity: red, highest; blue, lowest.

Downloaded from http://pubs.aip.org/physicstoday/article-pdf/57/6/37/11083469/37_1_online.pdf

amplitude evanescent wave can be sustained indefinitely stored by passing through the slab. These effects suggest
in a purely passive medium. another view of the focusing action, that of the slab anni-
For a conventional lens, resolution is limited by the hilating an equal thickness of vacuum. One can see why
aperture. This new lens based on negative materials will materials with negative n are so special: Negative media
also, in practice, have limitations, chiefly due to losses. behave like optical antimatter.
Any real material will always have small, positive imagi- In fact, the result is more general.14 Two slabs of ma-
nary components to e and m that represent resistive losses terial optically annihilate if one is the negative mirror
in the system and damp the resonances responsible for am- image of the other—that is, if they meet in the plane z ⊂ 0
plifying the near fields. Because of their high quality fac- and, at equal and opposite distances from that plane,
tors Q, the sharpest resonances give the greatest amplifi- e(x, y, z) ⊂ –e(x, y, –z) and m(x, y, z) ⊂ –m(x, y, –z).
cation, but they are also the most susceptible to energy Figures 5c and d illustrate the cancellation. The two
dissipation: They are the first to be killed by the losses, so media have varying refractive indices, and in general, light
with increasing loss the resolution is rapidly degraded. does not follow a straight line. Nevertheless, complemen-
Nick Fang and colleagues13 have explored near-field tary paths in each medium are traced such that the over-
amplification by exploiting the fact that for very small sys- all phase acquired in the first medium is canceled by the
tems—much smaller than the free-space wavelength—the contribution from the second. Likewise, if the waves have
electric and magnetic fields are independent of one an- a decaying nature, decay in one half would be followed by
other and can be controlled separately. Therefore if one is amplification in the other.
only concerned with the electric fields, m is irrelevant and This result may seem straightforward, but some con-
one need only ensure that e is negative. figurations have surprises. The two halves of figure 6a are
Silver has a negative real part to e and therefore a inverse mirror images as required by the cancellation the-
thin film should behave like a negative-index slab and am- orem, and therefore we expect that incident waves are
plify the near field. Fang and coworkers experimented on transmitted without attenuation and without reflection.
several silver films of different thickness, but each time se- Yet a ray-tracing exercise holds a surprise. Ray 2 in the
lected the same kx. The film clearly amplified waves up to figure hits the negative cylinder and is twice refracted to
a critical film thickness of about 50 nm, above which losses be ejected from the system rather than transmitted. The
intervened and the amplification process collapsed. Nev- rays don’t support our theorem!
ertheless, considerable amplification is possible: Amplifi- Further investigation shows that the cylinder is ca-
cation factors of around 30 were achieved before collapse. pable of trapping rays in closed orbits, shown by dotted
Thus we can be optimistic that some limited subwave- lines in the center of the figure. Such closed paths are the
length focusing can be achieved with silver films. signature of a resonance and a clue as to how the paradox
With a microwave transmission-line lens, Eleftheri- is resolved. A full solution of Maxwell’s equations shows
ades’s group in Toronto10 has recently realized conditions for that when the incident light is first switched on, the ray
subwavelength focusing and produced images significantly predictions are initially obeyed. With time, some of the in-
enhanced by evanescent-wave amplification. The image res- cident energy will feed into the resonant state in the mid-
olution of about l/5 was consistent with losses in the system; dle of the system, which in turn will leak energy into a
reducing the loss would improve resolution even further. transmitted wave and into a contribution to the reflected
wave that cancels with the original reflection. As always
Negative refraction as negative space in negative media, resonant states play a central role.
A slab of negative material with e ⊂ m ⊂ –1 acts like a lens: Figures 6b and c show the equilibrium solutions. With
Objects on one side are brought to a focus on the other side. only the negative sphere in figure 6b, there is strong scat-
As shown in figures 5a and b, as a wave progresses through tering. Figure 6c includes the mirror-antisymmetric layer
the negative medium, its phase is wound backward. Over- that, within the accuracy of the calculations, removes the
all, the slab undoes the effect of an equal thickness of vac- reflected contributions and the spurious forward scatter-
uum. Similarly, decaying waves have their amplitude re- ing to leave transmission unhindered, as predicted.

42 June 2004 Physics Today http://www.physicstoday.org

 An interesting question arises if there is absorption—
represented by positive imaginary parts of either or both
of e and m—in the system. Conditions for the theorem may
still be satisfied but require that for every instance of a
positive part to e or m, there is a mirror-antisymmetric neg-
ative part somewhere else in the system. In other words,
parts of the system must exhibit gain: Loss can only be
compensated by active amplification with a sustained
input of power.
Building on the foundations
Negative refraction is a subject with constant capacity for
surprise: Innocent assumptions lead to unexpected and
sometimes profound consequences. This new field has gen-
erated great enthusiasm but also controversy, yet even the
controversies have had the positive effect that key concepts
have been critically scrutinized in the past 18 months. In the
past year, experimental data have been produced that vali-
date the concepts. As a result, we have a firm foundation on
which to build.15 Many groups are already moving forward
with applications. The microwave area has naturally been
most productive, because the metamaterials required are

Downloaded from http://pubs.aip.org/physicstoday/article-pdf/57/6/37/11083469/37_1_online.pdf

easier to fabricate. In addition to microwave lenses, novel
waveguides and other devices are under consideration.
One of the most exciting possibilities is imaging be-
yond the wavelength limit. Practical applications will re-
quire low-loss materials, which are a great challenge to the
designers of new metamaterials. Proposals to employ thin
silver films as lenses are being explored in several labora-
tories. And the challenges are not purely experimental: We
are not yet done with theory, because the assumption of
negative refraction has many ramifications that are still
being explored and are sure to cast more light on this
strange but fascinating subject. Not surprisingly, many re-
searchers are joining the field: 2003 saw more than 200
papers published on negative refraction. We expect even
more in 2004!
References
1. V. G. Veselago, Sov. Phys. Usp. 10, 509 (1968).
2. J. B. Pendry, A. J. Holden, D. J. Robbins, W. J. Stewart, IEEE
Trans. Microwave Theory Tech. 47, 2075 (1999).
3. R. A. Shelby, D. R. Smith, S. Schultz, Science 292, 77 (2001).
4. A. A. Houck, J. B. Brock, I. L. Chuang, Phys. Rev. Lett. 90,
137401 (2003).
5. C. G. Parazzoli, R. B. Greegor, K. Li, B. E. C. Koltenbah, M.
Tanielian, Phys. Rev Lett. 90, 107401 (2003).
6. M. Notomi, Phys. Rev. B 62, 10696 (2000).
7. C. Luo, S. G. Johnson, J. D. Joannopoulos, J. B. Pendry, Opt.
Express 11, 746 (2003).
8. P. V. Parimi, W. T. Lu, P. Vodo, S. Sridhar, Nature 426, 404
(2003).
9. E. Cubukcu, K. Aydin, E. Ozbay, S. Foteinopolou, C. M. Souk-
oulis, Phys. Rev. Lett. 91, 207401 (2003).
10. A. Grbic, G. V. Eleftheriades, Phys. Rev. Lett. 92, 117403
(2004).
11. L. Liu, C. Caloz, T. Itoh, Electron. Lett. 38, 1414 (2002).
12. J. B. Pendry, Phys. Rev. Lett. 85, 3966 (2000).
13. N. Fang, Z. Liu, T. J. Yen, X. Zhang, Opt. Express 11, 682
(2003).
14. J. B. Pendry, S. A. Ramakrishna, J. Phys.: Condens. Matter
14, 6345 (2003).
15. Further reading can be found in a special edition of Opt. Ex-
press, “Focus Issue: Negative Refraction and Metamaterials,”
11, 639–755 (April 2003), and in M. W. McCall, A. Lakhtakia,
W. S. Weiglhofer, Eur. J. Phys. 23, 353 (2002).
16. R. A. Shelby, D. R. Smith, S. C. Nemat-Nasser, S. Schultz,
Appl. Phys. Lett. 78, 4 (2001).
17. P. Kolinko, D. R. Smith, Opt. Express 11, 640 (2003).
18. C. G. Parazzoli, R. B. Greegor, J. A. Nielson, M. A. Thompson,
K. Li, A. M. Vetter, M. H. Tanielian, Appl. Phys. Lett. 34,
3232 (2004). 䊏

June 2004 Physics Today 43 Circle number 25 on Reader Service Card