Early View publication on wileyonlinelibrary.

com
(issue and page numbers not yet assigned;
citable using Digital Object Identifier – DOI )

LASER & PHOTONICS

Laser Photonics Rev., 1–21 (2011) / DOI 10.1002/lpor.201000043 REVIEWS

Abstract During the past decade electromagnetic metamateri-

als have realized many exotic phenomena that are difficult or im-
possible using naturally occurring materials. It is their resonantly
enhanced interaction with electromagnetic waves that under-
pins their attractive qualities, which are increasingly important
in the terahertz frequency range. Passive and active terahertz
metamaterials and devices have enabled novel functionality and
unprecedented terahertz device performance. These demon-
strations prove their potential to address the so-called terahertz
gap, a technology vacuum associated with the deficiency of
natural materials with a desirable terahertz response.

Manipulation of terahertz radiation using metamaterials

Hou-Tong Chen * , John F. O’Hara, Abul K. Azad, and Antoinette J. Taylor

1. Introduction decades of the electromagnetic spectrum, and discoveries

of new phenomena including prominent proof-of-concept
Metamaterials have been developed over the past ten years demonstrations such as super-resolution in optical imaging,
in the form of composites of ordered subwavelength con- perfect metamaterial absorbers, electromagnetic invisibility
stituents. They are man-made effective media that exhibit (cloaking) and transformation optics. One may refer to a
unusual properties not normally found in nature. The idea few review articles or books for these topics [7–10].
was inspired by the pioneering works of Pendry in reduc- A designed and controllable metamaterial resonant re-
ing electrical plasma frequency in metal wires [1, 2] and sponse, either passive or active and linear or nonlinear,
creating an artificial magnetic response via metal split-ring would enhance the interaction between electromagnetic
resonators (SRRs) [3]. The first demonstration of composite waves and materials. Novel functionality and possibly im-
metamaterials was at microwave frequencies by Smith and proved performance could be expected using metamaterials.
colleagues [4, 5], and used a combination of plasmonic-type This is of particular interest and importance in the tera-
metal wires and an SRR array to create negative electric hertz (1 THz 1012 Hz) frequency range, which is roughly
permittivity εeff and negative magnetic permeability μeff , re- defined between 0.1 and 10 THz. It is the least developed
spectively. It was demonstrated that electromagnetic waves regime in the electromagnetic spectrum, though it is very
(light) are able to propagate in such composite metamate- attractive for numerous applications [11] including chemi-
rials with simultaneously negative values of constitutive cal and biological spectroscopy and sensing, material char-
parameters (ε and μ), and the light refraction is opposite to acterization, nondestructive concealed weapons detection,
normal materials, i. e., a negative index of refraction. These imaging, and secure short-range wireless communication,
are the kind of hypothetical “substances” that Veselago [6] in addition to fundamental spectroscopy studies of com-
had speculated over 40 years ago but were never found be- plex materials [12]. This THz gap is essentially due to the
fore in naturally existing materials. In his paper, Veselago lack of materials having a suitable THz electronic response
called these substances left-handed materials (LHM), and which is present for microwave and lower frequencies, or
predicted several fundamental phenomena occurring in or photonic response which is present for infrared and higher
in association with such substances [6], including the char- frequencies. The deficiency of THz materials has largely pre-
acteristic frequency dispersion, negative index of refraction, vented the development of THz technology for real-world
reversal of Snell’s law, focusing with a flat slab, and rever- applications, as compared to its neighboring microwave and
sal of Doppler effect and Čerenkov radiation, all of which infrared regimes. By leveraging the tunable electromagnetic
have been experimentally observed using metamaterials. response, metamaterials could play a key role in solving this
The growing worldwide interest in these exotic properties THz materials issue, a new opportunity that would not be
has resulted in metamaterial structures operating over many possible with naturally existing materials. This has already

Center for Integrated Nanotechnologies, Los Alamos National Laboratory, MS K771, Los Alamos, NM 87545, USA
*
Corresponding author: e-mail: [email protected]

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

2 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

proven true with a few demonstrations of THz devices, for SRR in Fig. 1a, the wire pair can realize both an electric and
instance, THz modulators using hybrid THz metamateri- magnetic material response, yet has the advantage of much
als. In this article we comprehensively review the recent simpler planar fabrication, particularly at THz and higher
progress in THz metamaterials and devices, including de- frequencies. It was demonstrated first and has been heavily
sign, fabrication, and characterization of THz metamaterials, used in the optical frequencies to realize optical magnetic
with a focus on their passive and active properties towards resonance and negative index of refraction [24–27]. Such
THz device applications. Although photonic crystals and pe- an idea was then used in the THz frequency range [28–30].
riodic metal hole arrays (extraordinary optical transmission) Similar to fishnet structures in the near-infrared [25, 26] and
have been also considered as metamaterials in a broadened stacked periodic hole arrays in the millimeter waves [31],
sense, they are not included in this review article. negative refraction in the THz frequency range was also
recently demonstrated in stacked periodic metal hole arrays
under normal incidence [32–34]. These alternative meta-
2. Terahertz metamaterials: design, material designs, demonstrated first in optical frequencies,
fabrication, and characterization could be very useful where magnetic resonance is important
in realizing particular functionality, for example, a recently
The design, fabrication, and characterization of metamate- proposed deep subwavelength THz waveguide [35]. Ad-
rials operating in the microwave, THz, infrared and visible ditional proposals of composite right/left-handed (CRLH)
regimes all have different requirements. Generally, SRR- transmission lines used metamaterial principles to achieve
based designs are suitable for microwave [4,5], THz [13–16] a uniform spatial mode in THz quantum cascade lasers,
and infrared [17] metamaterials, though further increase of making them immune to spatial hole burning [36].
their resonance frequency is prevented by increasing metal In SRR-based THz metamaterials, the unit cell has
losses and kinetic inductance [18]. In the THz frequency outer dimensions of a few tens of microns ( λ0 =10 where
regime, metal losses are not yet significant and the majority λ0 is the wavelength in vacuum), and the smallest fea-
of THz metamaterials have employed SRRs and their vari- ture size is on the micron scale. Therefore, planar THz
ants. In Fig. 1 we illustrate several unit cell designs typically metamaterials are naturally compatible with conventional
used in THz metamaterials. These are periodically patterned micro-fabrication processes, which include photolithogra-
in two dimensions (2D) on the planar surface of an appropri- phy, metal film deposition, lift-off process and wet/dry etch-
ate substrate. The heart of the metamaterial response is the ing. Additional fabrication techniques for planar THz meta-
SRR shape: the asymmetry introduced by the capacitive ring materials include ink-jet printing [37, 38] and microfluidic-
gap enables electric field coupling, while the magnetic field jetting [39], both of which have the advantage of creating
inductively couples to the ring loop. Together these create a large area metamaterials without clean room facilities. Typi-
fundamental resonant response analogous to a lumped LC cally insulators such as high resistivity semiconductors are
resonator. The single SRR in Fig. 1a enables coupling to used as THz metamaterial substrates. These not only provide
both the electric and magnetic fields of the incident elec- mechanical support, but can also introduce electromagnetic
tromagnetic waves, while the structure in Fig. 1b allows functionality. Recently, there has been increasing interest in
coupling to the electric field alone, therefore inducing a THz metamaterials on polymer and silk-based flexible films
purely electrical response [19–21]. Like cavity resonators, for biomedical applications [40–44].
SRRs exhibit higher-order oscillation modes, which occur Planar metamaterials are more precisely described as
at frequencies higher than the fundamental. Modes that are surface scatterers rather than effective media [45], the latter
closely spaced in frequency can couple each another, and requiring three-dimentional (3D) structures. The fabrication
unit cells may be designed to enhance or eliminate this of 3D THz metamaterials remains a hurdle that imposes
coupling by utilizing rectangular (or otherwise reshaped) difficult contraints on THz metamaterial design. Quasi-3D
SRRs [22, 23]. The unit cell illustrated in Fig. 1c is a pair of THz metamaterials can be obtained simply through layer-by-
metal wires separated by a dielectric spacer. Like the basic layer fabrication or stacking 2D metamaterials with appro-

Figure 1 (online color at: www.lpr-

journal.org) Unit cells typically used in
THz metamaterials. (a) Single split-ring
resonator; (b) electrical split-ring res-
onator; and (c) a pair of metal wires.
The arrows indicate the current flow.
The bottom insets illustrate the po-
larization of the incident electromag-
netic waves.

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 3

priate spacers [15,30,46–50]. But one key issue in achieving 3. Passive properties of
truly 3D THz metamaterials lies in the formation of out- terahertz metamaterials
of-plane resonators, which is not easily accomplished us-
ing standard micro-fabrication techniques. Microelectrome-
chanical systems (MEMS) fabrication approaches provide 3.1. Magnetic and electric response
a viable solution realizing reconfigurable THz metamate-
rials [51]. This form of fabrication uses a combination of Natural materials that exhibit a magnetic response are far
mechanically dissimilar materials to create an array of free- less common than those exhibit an electric response, particu-
standing, out-of-plane SRRs that are connected to the sup- larly at THz frequencies and beyond. The artificial magnetic
porting substrate by cantilever legs. response in SRRs [3], however, is scalable and is suitable
Recently, there has been some progress in a three-step for operating in the THz frequency range. In a metal SRR
hierarchical self-assembly approach that results in forma- shown in Fig. 1a, the time-varying magnetic field of the
tion of microarrays composed of metallic polyhedra [52]. incident electromagnetic wave can excite oscillating cur-
Conventional photolithographic methods are used to pattern rents in the loop through Faraday induction. The capacitive
2D templates with special metals (nickel for the panel and split gap and the inductive loop make the structure behave
lead-tin for the solder hinge). Templates are designed so like a subwavelength LC resonator, withpthe fundamental
that cubic and parallelepiped shapes could be assembled resonance frequency given by ω0 = 1= LC when damp-
spontaneously by heating through a minimization of surface ing is negligible. Obviously the capacitance, and therefore
energy. Another 3D metamaterial fabrication technique was the resonance frequency, will be affected by the substrate
demonstrated at mid-infrared frequencies [53, 54] but with dielectric constant [16], foreshadowing a method to tune
the potential to be employed in the THz. It is based on mem- the magnetic response by altering the substrate dielectric
brane projection lithography (MPL), a multi-step process constant. In order to excite a magnetic resonant response,
to create out-of-plane SRRs. The basic idea begins with a the incident magnetic field must have a component perpen-
developable and relatively thick polymer film, on which a dicular to the SRR plane, which requires an oblique incident
thin membrane material is deposited. By lithographic meth- angle for typical planar THz metamaterials.
ods the membrane is patterned with holes in the shape of THz metamaterials were first demonstrated by Yen et
SRRs, for example. These structured holes provide access of al. [13] and Moser et al. [14]. In the work by Yen et al.,
a developer to dissolve the polymer beneath the membrane a magnetic response at THz frequencies was observed in
and form cavities. Finally, angled electron beam evapora- planar metamaterials composed of periodic arrays of 3-μm-
tion projects the deposition of metals through the holes to thick, nonmagnetic copper SRRs fabricated on quartz sub-
create out-of-plane SRRs under the membrane and on the strates using a photoproliferated process [13]. The mea-
cavity walls. surements were performed using spectroscopic ellipsometry
The characterization of THz metamaterials typically adapted with FTIR for S-polarized and P-polarized light
employs two approaches: Fourier transform infrared spec- reflection measurements at an oblique incidence angle of
troscopy (FTIR) and terahertz time-domain spectroscopy 30 , as shown in Fig. 2a. The S-polarized light allows for a
(THz-TDS). FTIR is relatively easy to use and is able to perpendicular component of magnetic field to be coupled to
cover a wide spectral range. But it is an intensity mea- the SRR, resulting in a magnetic resonance. The frequency
surement that lacks phase information. The majority of dependent ellipsometry parameter tan 2 (Ψ) reveals a mag-
THz metamaterials have been characterized using THz- netic resonance peak, which is a function of the geometric
TDS, which may be performed in reflection or transmission dimensions of the resonant elements, as shown in Figs. 2b
mode. Sample absorption can be obtained by the relation and 2c, the experimental and simulation results, respectively.
A = 1 T R, where A, T and R are the absorbance, trans- Early THz metamaterials employed relatively thick SRR
mittance and reflectance, respectively. THz-TDS is a coher- structures [13–15], which proved unnecessary as long as
ent measurement technique, where linearly polarized ultra- the metal thickness remains larger than the skin depth [56],
fast broadband THz pulses are generated and detected using typically in the order of 100 nm at THz frequencies for good
femtosecond near-infrared lasers. It enables one to tempo- metals. Thicker metal results in marginal changes of reso-
rally map the THz waveform with femtosecond resolution, nance strength, but can affect the resonance frequency [57].
and it also allows time-windowing of the measured data, The dependence of the magnetic resonance on incidence
which is often useful in removing multiple reflection effects. angle in planar THz metamaterials was further investigated
Since it is a coherent technique, one can retrieve both the in several other works [42, 58–61]. In these experiments,
spectral amplitude and phase via numerical Fourier trans- the magnetic coupling to planar THz metamaterials seems
formation [55]. THz pulses usually contain usable spectral to be rather weak even with relatively large incidence an-
components ranging from 100 GHz to 3 THz. Normaliza- gles, as revealed by the shallow resonant dips. It should
tion of the spectra is often used to separate the metamaterial be noted that the conventional SRRs are bianisotropic in
response from a reference material, for example, a bare nature [62]. The incident magnetic field excites not only
substrate, free space, or a perfect mirror (in reflection mea- a magnetic moment but also an electric dipole at the split
surements). The spectral amplitude and phase information gap. In addition, if the incident wave polarization is parallel
can also be used to retrieve the effective constitutive param- to the gap-bearing side of the SRR, as shown by the right
eters (ε and μ). polarization under Fig. 1a, the incident electric field also

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

4 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

Figure 2 (online color at: www.lpr-

journal.org) (a) Illustration depicting the
orientation of the 30 ellipsometry exper-
iment. The inset to (a) shows a unit cell
image taken by focused ion-beam mi-
croscopy. (b) Experimental and (c) nu-
merical simulation results of the ratio
of the magnetic (S-polarized) to electric
(P-polarized) response for three differ-
ent structures. Figure adapted from Yen
et al. [13].

ing of high dielectric constant materials [65, 66]. A recent

review paper on Mie resonance-based dielectric metamate-
rials can be found in [67]. Currently, investigations of meta-
material resonances are still highly dependent on numerical
simulations. However, a THz near-field imaging technique
for measuring the profile of the electric and magnetic fields
would provide valuable information in determining the ori-
gin of the metamaterial resonances [68, 69].
By changing the design of the SRRs, the above men-
tioned bianisotropy can be eliminated. In the unit cell struc-
ture shown in Fig. 1b, when the magnetic field has a com-
ponent perpendicular to the SRR plane, the currents in the
sub-loops will cancel each other at the center arm, and effec-
tively create a closed outer ring, which is non-resonant. This
Figure 3 Transmitted THz amplitude spectra under normal inci- design is also capable of eliminating any magnetoelectric
dence through double SRR-based metamaterials fabricated on
response. Here, the counter-circulating currents excited by
silicon (solid curve) and quartz (dotted curve) substrates. Inset
the electric field create a pair of magnetic moments with
shows the incident THz field orientation.
opposite directions, which cancel each other and leave a
purely electrical resonant response [19]. A series of THz
excites the fundamental circulating mode at the same fre- planar metamaterial structures consisting of a variety of
quency as the magnetic excitation [15, 16, 63, 64]. This is electric SRRs is shown in Fig. 4a. These were all demon-
sometimes termed electric excitation of the magnetic reso-
nance [15, 63]. We avoid this terminology because in this
orientation the excited magnetic moments do not coherently
combine to produce scattering in any direction; it is the
electric field excited electric dipoles that produce the entire
observed response. Azad et al. [16] demonstrated such a
THz electrical excitation of LC resonance in double SRR-
based metamaterials under normal incidence. As depicted in
Fig. 3, the transmission amplitude spectra, measured using
THz-TDS, revealed strong resonances excited by the in-
plane electric field of THz radiation along the gap-bearing
arms of SRRs. Additionally, identical structures fabricated
on silicon and quartz substrates revealed shifting in the res-
onance frequency, clearly showing the effect of substrate
index on the resonance frequency via modifying the effec-
tive capacitance in the SRR gaps.
A magnetic resonance at THz frequencies was also re- Figure 4 (online color at: www.lpr-journal.org) (a) Illustration
alized using a pair of metal strips separated by a dielectric of metamaterial unit cells exhibiting purely electric resonant re-
spacer [Fig. 1c] [28–30]. Further, a magnetic response can sponse. (b) Inverse metamaterial structures termed complemen-
be obtained from Mie resonances in particles or rods consist- tary metamaterials.

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 5

Figure 5 (online color at: www.lpr-journal.org) (a) Transmission amplitude and (b) phase spectra for the planar metamaterials with
unit cell structure shown in Fig. 4. The red and blue curves are for original and complementary metamaterials, respectively.

strated [20, 21] to reveal strong purely electric resonances, both the original and complementary metamaterials in Fig. 4.
as evidenced by the sharp transmission dips and strong There were claims that a magnetic resonance will arise in
Lorentzian phase dispersion, shown in the red curves in the complementary metamaterial when its corresponding
Fig. 5. As already discussed, SRRs exhibit higher order res- original metamaterial has an electrical resonance, and vice
onances excited by the incident electric field, similar to versa [71, 73]. However, the mechanism of the claimed mag-
those in cut wire arrays. The resonance frequencies are de- netic resonance could not be clarified.
termined by the vertical side length of the SRR (along the The transmission amplitude and phase spectra in Fig. 5
direction of incident electric field) as well as the distance could be used to retrieve the effective permittivity of meta-
between SRRs [68, 70]. Higher order modes can strongly materials using the reversal of Fresnel equations and by
couple with the fundamental LC mode if the resonance fre- assuming some appropriate effective metamaterial thick-
quencies are quite close. This has the effect of reshaping the ness, for instance, the dimension of the unit cell. However,
transmission and reflection spectra, and can be enhanced or the effective permittivity and permeability are bulk prop-
eliminated using rectangular SRRs [22, 23]. erties, more appropriately designated for multiple layered
THz metamaterials. Single layered planar metamaterials are
not well defined in thickness. Therefore, regardless of the
3.2. Complementary planar metamaterials resonance origin (electrical or magnetic), the transmission
(and reflection) amplitude and phase spectra are more fun-
The unit cell structures, shown in Fig. 4b, constitute the damentally important than permittivity and permeability in
so-called complementary planar THz metamaterials [20] the design of advanced single layered THz metamaterial
satisfying Babinet’s principle. Complementary metamate- devices. This notion has led to exacting studies of properties
rials were first investigated in the microwave frequencies of meta-films or meta-surfaces [45].
by Falcone et al. [71]. According to Babinet’s principle,
when the polarization of the incident THz waves rotates
90 with respect to the corresponding original metamaterial 3.3. Terahertz metamaterial polarimetric devices
structures, or equivalently, the samples rotate 90 , the com-
plementary metamaterials should exhibit complementary As shown in the previous section, the resonant response
transmission properties, as illustrated in Fig. 5a. That is, in in THz metamaterials exhibits strong frequency dispersion
place of a transmission dip in the original metamaterials, the in reflection and transmission amplitude and phase. Phase
complementary metamaterials show a sharp transmission effects are the backbone of many polarization effects such
peak. Further transmission and reflection measurements re- as dichroism, birefringence and optical activity, or devices
veal that the absorption in complementary metamaterials such as wave plates and beam-splitters. Metamaterials there-
is rather small, as expected to satisfy Babinet’s principle, fore offer a unique approach to creating THz polarimetric
and mainly due to the metallic losses [72]. We emphasize devices. This takes advantage of the anisotropic existing in
that the observed resonant response is purely electrical in many metamaterial designs with reduced symmetry, such

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

6 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

as the resonators in Figs. 1, 2, and the first two structures circular polarization (i. e. a quarter wave plate) is achieved
in Fig. 4. The anisotropic resonators behave as two orthog- in one metamaterial layer. Based on the same principle,
onal oscillators with distinct resonance frequencies for the Strikwerda et al. fabricated and measured THz quarter-wave
polarizations along the two orthogonal axes; one is driven plates based on both a single layer anisotropic SRRs as well
above its resonance frequency, and the other is driven be- as a dual-layer of meanderlines [75], both of which achieved
low its resonance frequency. The result is that the com- > 99% circular polarization from linearly polarized THz
ponent along one axis experiences a negative phase shift waves with reasonable bandwidth. Other devices such as
and the other component experiences a positive phase shift, polarizing beam splitters and pulse shapers can also be cre-
a birefringent effect suitable for the construction of THz ated [74], all of which are based on the simple birefringent
wave plates [74, 75]. response of metamaterials.
The inset to Fig. 6b shows a unit cell studied by Peralta
et al. in creating THz wave plates and polarizing beamsplit-
ters [74]. The figure also shows the THz transmission am-
plitude and phase response of a single layer of that metama- 3.4. Chiral terahertz metamaterials
terial to linearly polarized incident waves. Separate curves
indicate different polarization orientation angles of the in-
cident wave. The vertical black lines show frequencies that Chirality is another material property that affects wave po-
are of particular interest, where the amplitude of the meta- larization, in the form of optical activity. Unlike birefrin-
material response is minimally affected by orientation angle, gence, chirality requires a non-zero interaction between the
yet the phase is greatly affected. At 1.06 THz the phase dif- electric (magnetic) fields and the magnetic (electric) flux
ference is approximately 45 . This point deserves emphasis: densities. That is, chiral media are bi-isotropic, whereas
half of the required phase shift necessary to change linear to the previously discussed birefringent metamaterials were
simply anisotropic [76]. Chiral materials not only enable
p 
optical activity and circular dichroism; they also change
the index of refraction to n = με κ, where the added
term, κ, is called the chirality parameter. The chirality pa-
rameter can be so large in metamaterials that it enables
a new route to negative refraction [77]. Multi-layer THz
metamaterials have been shown to exhibit optical activity
and verify that the observed effects are not from simple
anisotropy [78]. Engelbrecht et al. realized THz optical
activity to 600=λ [79]. Light-induced active control over
THz optical activity has also been demonstrated, though the
effect is very weak [80]. Zhang et al. reported negative re-
fractive index in THz metamaterials via chirality [81]. And
Singh et al. have demonstrated “extrinsic” chirality, where
the chiral effects of optical activity and circular dichroism
were measured in an intrinsically achiral metamaterial [82].
In the work by Zhang et al. [81], significant dichroism
is realized by using a vertical metallic chiral resonator as
shown in Fig. 7, in which electrical and magnetic responses
in the THz frequency range are simultaneously excited with
a small angle between the electrical dipole and the mag-
netic moment. Measured using THz-TDS, the transmission
spectra of right and left circularly polarized (RCP and LCP)
beams are significantly different in magnitude and disper-
sion, particularly near the resonance frequency. That is, for
the metamaterial shown in Fig. 7, the transmission of the
LCP beam exhibits much stronger resonant dispersion than
the RCP beam. The experimentally extracted refractive in-
dex for LCP shows negative values over 1.06 to 1.27 THz
with a minimum value below 5, while the refractive in-
dex for RCP shows only positive values. It is further shown
that, although the permittivity exhibits negative values, the
permeability is always positive due to the relatively weak
Figure 6 (online color at: www.lpr-journal.org) Response of magnetic response. Thus, the observed negative index of
metatmaterial with a polarization sensitive unit cell (a) transmis- refraction is due to the chiral design of the metamaterial,
sion amplitude and (b) phase at various incident polarization an- providing a new opportunity in exploring negative reflection
gles. and negative refraction [81].

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 7

to find appropriate coating materials that satisfy the index

matching requirement. Usually, it is also quite challenging
to fabricate high quality and relatively thick films required
for THz wavelengths.
One THz antireflection approach involves fabricating
surface relief structures, which achieve impedance matching
by creating an effective index satisfying the quarter-wave an-
tireflection constraints [88]. Such structures may also utilize
gradient index schemes [87, 89, 90] to possibly accomplish
broadband antireflection performance. Chen et al. fabricated
micropyramid structures on a silicon surface, which enables
enhancement of THz transmission over a broad bandwidth,
though the anti-reflection behavior is still narrowband [90].
Later, they accomplished broadband THz antireflection [87]
employing a three-dimensional inverted photonic grating
structure illustrated in Fig. 8a–c. This scheme produces a
measured reflectance of less than 5% spanning from 2.2 THz
to 5.5 THz, as shown in Fig. 8d. Further, the enhanced trans-
mittance 3-dB bandwidth spans from 1.4 to 6 THz, shown
in Fig. 8e. The main drawback of this approach is that the
substrate surface is no longer flat or smooth limiting its
further integration.
Chen et al. developed THz antireflection coatings based
on planar metamaterials consisting of an electric SRR array
and a metal mesh separated by a dielectric spacer [91], a unit
cell shown in Fig. 9a. In principle, this approach is suitable
for substrates with any dielectric constant value, and there
is very little requirement for the spacer dielectric materials.
As shown in Fig. 9b for normal incidence, the metamaterial
antireflection coating is capable of a dramatic reduction of
reflection and significant enhancement of transmission near
the designed frequency ( 1:2 THz), despite being much
thinner than a wavelength ( λ0 =20). These were further
demonstrated to operate over a wide range of incidence
angles for both transverse magnetic (TM) and transverse
electric (TE) polarization [91, 92], as shown in Figs. 9c
Figure 7 (a) The schematic of the chiral structure made of gold.
and 9d. In fact, the angular-dependence of antireflection
(b) The inductor-capacitor circuit model of the chiral structure.
performance for TE and TM polarizations could be close to
(c,d) SEM images of the chiral metamaterials at a tilted angle. identical when the spacer thickness is optimized to realize
(e,f) Linearly polarized THz-TDS measurements of the chiral meta- zero reflection at normal incidence [92].
materials. t1 and t2 are the transmission coefficients when polar- This work also identified that destructive and construc-
izer and analyzer are parallel and perpendicular with each other, tive interference, due to multiple reflections and transmis-
respectively. (g,h) Calculated circular polarization transmission sions at the boundaries of the metamaterial coating, schemat-
= = +
coefficients tL t1 it2 and tR t1 t2 . Figure adapted from ically shown in Fig. 10, are responsible for the reduction
Zhang et al. [81]. of reflection and enhancement of transmission [91]. The
SRR array at the air-spacer interface and the mesh at the
spacer-substrate interface can be treated as meta-surfaces,
3.5. Antireflection coatings with metamaterials which impart strong magnitude and phase shifts to the reflec-
tion and transmission coefficients. With appropriate spacer
Many THz devices and components, including the metama- thickness, the metamaterial coating accomplishes a situa-
terial devices described in this review, suffer from insertion tion similar to quarter-wave antireflection but with a much
losses due to Fresnel reflections of the substrates. Antireflec- smaller overall coating thickness and relaxed requirements
tion coatings are of significant importance when considering for the properties of the coating materials. This mechanism
the generally low power in many far infrared or THz sys- was proved by modeling calculations assuming that the top
tems. Single layered quarter-wave antireflection coatings (SRRs) and bottom (mesh) meta-surfaces are completely
require a particular refractive index and a quarter wavelength decoupled. The complex transmission and reflection coef-
thickness [83–86]. In the far infrared and microwave bands, ficients at the SRR and mesh interfaces are then obtained,
substrate materials such as semiconductors and ceramics which are used to calculate the overall reflection and trans-
often have large refractive indices, which make it difficult mission taking into account the multi-reflection between the

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

8 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

Figure 8 (online color

at: www.lpr-journal.org) (a–
c) Schematic and scanning
electron microscope images
of a inverted photonic grating.
(d) Reflectance and (e) trans-
mittance spectra of 20 μm
(blue curve) and 15 μm (red
curve) period devices. The
corresponding dashed curves
are simulation results. Figure
adapted from Chen et al. [87].

two interfaces. The results reproduce the most important sion dips are due to reflection rather than absorption [72].
features observed in experiments and numerical simulations The first perfect metamaterial absorber was developed by
when the antireflection metamaterial coatings are treated as Landy et al. in the microwave frequency range using a bi-
a whole. This indicates that the assumed magnetic resonant layered metamaterial structure [94]. This approach was used
response plays a less important role in realizing metama- shortly thereafter to demonstrate metamaterial absorbers in
terial antireflection coatings, since in the above modeling the THz frequency range [95]. The unit cell of the THz
calculations the decoupled SRRs and mesh indicate no mag- metamaterial absorber is shown in Fig. 11a consisting of
netic resonance. The coupling between the top and bottom two distinct metallic elements: an electric SRR and a split
metallic structures slightly shifts the metamaterial resonance wire separated by a dielectric spacer (polyimide). In simula-
frequency and therefore can be considered as a perturbation. tions, nearly 100% absorbance was achieved, while in exper-
It is worth pointing out that the metamaterial antireflection iments, due to the fabrication tolerance, a resonant absorp-
operates at an off-resonance frequency, which dramatically tivity of 70% was demonstrated at 1.3 THz [95]. This corre-
reduces the loss and enhances the transmission. This antire- sponds to a power absorption coefficient of α = 2000 cm 1 ,
flection mechanism further suggests a general rule in de- which is significant at THz frequencies. The absorbance
signing metamaterial antireflection coatings. That is, many was further experimentally improved to 97% in a flexible
other resonant/nonresonant structures, including frequency THz metamaterial absorber, where the bottom split wires
selective surfaces or FSSs, can be employed for antireflec- are replaced by a gold ground plane with no underlying
tion and absorber applications. The mechanism enables the substrate [96]. This free standing THz metamaterial ab-
correct interpretation of the perfect metamaterial absorbers sorber is also capable of operating over a wide range of
(see next section) and electromagnetic wave tunneling [93]. incident angles. Similar concepts of using a ground plane
and wide angle operation were numerically investigated by
Diem et al. [97]. The above THz metamaterial absorbers
3.6. Perfect metamaterial absorbers are polarization dependent, but polarization insensitive THz
metamaterial absorbers were accomplished by using more
In single layered planar metamaterials, the sharp transmis- symmetric unit cell designs [98, 99].
sion dips are caused by the large surface impedance mis- The principle under which the first metamaterial ab-
match near the resonance frequency. That is, the transmis- sorbers were designed is the ability to tailor the electric and

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 9

Figure 9 (online color at: www.lpr-journal.org) (a) Schematic design illustrating one unit cell of the metamaterial antireflection coating.
The structures should be chosen in such a way that the SRR and mesh (or other structures) provide a capacitive and inductive response,
respectively. (b) Experimentally measured reflectance and transmittance under normal incidence. The horizontal solid and dashed
gray lines are the expected THz reflectance and transmittance, respectively, of a bare substrate surface. The angular dependence
of experimental reflectance spectra from the metamaterial-coated surface are shown for (c) TM and (d) TE polarized incident THz
radiation. Inset: Angular dependent reflectance at the antireflection frequency.

will facilitate a design in which the impedance Z = μ =ε

p
matches that of the environment (free space) resulting in no
reflection. If they are also designed to be sufficiently lossy,
light entering into the metamaterials may be completely
absorbed. In this treatment, the magnetic resonance plays a
key role in metamaterial absorbers.
However, there are a few issues arising from this effec-
tive media treatment. All experimental and simulation re-
sults have shown dramatically different behaviors when the
electromagnetic waves propagate from the two opposite di-
rections even for free-standing metamaterial absorbers [100].
For instance, when a ground plane is used, the metamaterial
behaves like a perfect absorber when light is incident from
the front to the resonators, while it is a perfect mirror (re-
Figure 10 (online color at: www.lpr-journal.org) Schematic flector) when light is incident from the back to the ground
of the multiple reflection and transmission in the metamaterial plane. Further, the metamaterial absorbers consist of only
antireflection coatings and perfect absorbers. two layers of metallic structures, and are strongly inhomoge-
neous in the wave propagating direction. This is in contrast
magnetic responses independently [94–96, 98]. The eSRR to the requirement of multiple, periodic and homogeneous
couples strongly to the electric field and negligibly to the subwavelength structures to treat metamaterials as effective
magnetic field [20, 21], while the split wire and the center media. If metamaterial absorbers are effective media with
section of the eSRR form a wire pair enabling a magnetic re- effective index and impedance, then the system should be-
sponse [24, 25]. Both can be tuned independently by chang- have equally, i. e., exhibiting minimized reflection and high
ing the geometry of the SRR, split wire and the thickness absorbance from either propagation direction. However, this
of the spacer. In multi-layered or bulk metamaterials where was not observed in experiments or simulations even when
the metamaterials can be treated as effective media, this the medium surrounding the absorber was air, i. e., a free-

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

10 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

Figure 11 (online color at: www.lpr-journal.org)

(a) Schematic of the unit cell of the THz meta-
material absorber. (b) The simulated reflectance
(green), transmittance (red) and absorbance (blue).
Figure adapted from Tao et al. [95].

standing absorber without a substrate. Even when cut-wires enable real devices, particularly for applications in the THz
replace the ground plane, the metamaterial absorber reflects frequency range. Part of the metamaterial response is due
almost 100% when light is incident from the back to the to the conductivity of metals, which is required to be as
cut-wires – the cut-wires are equivalent to the ground plane. high as possible to facilitate the strongest resonant response.
With the above concerns, we performed theoretical cal- There is very little we can do to actively/dynamically mod-
culations using the interference model with decoupled meta- ify metal conductivity, even though this may be an avenue
material interfaces, as described in the previous section. to dynamically alter the resonant response. One exception
The calculation results are in excellent agreement with the is based on high temperature superconducting films. Below
experimental measurements and numerical simulations in the transition temperature, they have a conductivity higher
literature where the metamaterial absorbers were treated than metals at low frequencies. Even at THz frequencies the
as a whole. This, once again, suggests that the impedance conductivity is still rather high and can be tuned over a wide
matching by tuning magnetic permeability μ plays a neg- range by varying the temperature. While we will describe
ligible role. Instead, Fabry-Perot-like resonance is respon- the properties of superconducting THz metamaterials at the
sible for the absorption – with destructive interference in end of this section, for most tunable metamaterials it is ob-
reflection, light is trapped inside the metamaterial and even- viously preferential to alter the embedded environment of
tually absorbed. Similarly, Wen et al. developed a trans- the resonant elements. For example, changing the metamate-
mission line model to analyze the mechanism of metamate- rial substrate affects both resonance strength (via substrate
rial absorbers [101]. In this model, the top eSRRs and the losses) and the resonance frequency (via the substrate di-
bottom split wires are decoupled and modeled as surface electric constant). In contrast to the passive metamaterial
impedances, while the spacer is modeled as the transmission properties in the previous section, the functionality in the
line for wave propagation. This approach is essentially the active/dynamical metamaterials described in this section is
same as the interference approach we developed [91, 102]. essentially determined by modifying the properties of the
The model calculations also nicely reproduce the asymmet- metamaterial substrate, or some materials incorporated in
ric absorption as well as other absorption features. critical regions of the resonant elements. These dynamic
Despite the controversy in the mechanism, maximiz- changes occur through application of an external stimulus,
ing the absorbance in THz metamaterial absorbers will which may be thermal, optical, electronic or magnetic.
be useful for detecting THz or creating thermal radiation
detectors. The THz metamaterial absorbers could be de-
signed to have a small volume, small thermal mass, high
absorption coefficient, and spectral selectivity, which are 4.1. Resonance switching and frequency tuning
ideal for radiation detecting bolometers [103], e. g. for THz by photoexcitation
imaging applications. Additionally, broadening the absorp-
tion bandwidth [104] or achieving multi-frequency absorp- Semiconductors play an important role in resonance switch-
tion [105, 106] is also of importance in certain applications. able and frequency tunable THz metamaterials. They can
be dynamically tuned from insulating to highly metallic
through photoexcitation of free charge carriers. Padilla et al.
4. Active and dynamical demonstrated the first dynamically switchable THz metama-
terahertz metamaterials terial by photoexcitation [59], where an intrinsic GaAs wafer
is used as the substrate of a planar array of single SRRs.
In general, after the metamaterial structure is designed and Photocarriers are excited using near-infrared femtosecond
fabricated on an insulating substrate, its resonance strength laser pulses, which increase the surface conductivity and
and frequency are already determined and fixed. Dynami- therefore the loss of the GaAs substrate. The otherwise
cally and actively switchable and frequency tunable meta- strong resonant response is dramatically damped depending
materials are highly desired to enhance functionality and on the fluence of the photoexcitation, and consequently the

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 11

THz metamaterial resonances and consequently the THz

transmission through an electric SRR array and its comple-
mentary metamaterial [108]. This substrate was also used
in ultrafast switching of surface plasmon enhanced reso-
nant THz transmission in two-dimensional subwavelength
metallic hole arrays [109].
Accompanying the resonance switching, the results in
Fig. 12 also show a slight blue-shifting of the resonances
due to the decreasing real part of the complex dielectric
constant in the semiconductor substrate after photoexci-
tation [59, 110]. Chen et al. demonstrated more dramati-
cally frequency agile THz metamaterials by incorporating
semiconducting material as an integral part of the resonant
elements. The result is dynamical tuning of the metama-
terial resonance frequency through photoexcitation [111].
As shown in Figs. 13a and 13b, the electric SRR array was
fabricated on a silicon-on-sapphire substrate, where the 0.6-
μm-thick highly resistive silicon layer was etched away
Figure 12 (online color at: www.lpr-journal.org) Experimentally except for the two strips near the split gap. These formed
measured THz transmission amplitude spectra through an SRR capacitive plates when they were photoexcited and became
array fabricated on an intrinsic GaAs substrate under various highly conducting. The modification of capacitance at the
power levels of photoexcitation. split gap by photoexcitation results in a red-shifting of meta-
material resonance frequency by over 20% in experiments,
transmission of the THz radiation is also switched, most sig- shown in Fig. 13c, which agrees well with numerical simu-
nificantly near the resonance frequencies, and at a relatively lation results, shown in Fig. 13d, using appropriate values
low power as shown in Fig. 12. This all-optical dynamical of conductivity corresponding to the fluence levels of pho-
switching (modulation) of THz radiation can be very fast toexcitation.
if one reduces the lifetime of photocarriers, which is in the Two other metamaterial structures were also proposed
nanosecond range in GaAs and is responsible for the rel- to achieve a blue-shifting of the metamaterial resonance
atively slow recovery of the metamaterial resonance [59]. frequency upon photoexcitation, where instead the incorpo-
GaAs/ErAs nanoisland superlattices can also be used as rated silicon strips formed parts of the SRR current loops
the metamaterial substrate and have carrier lifetimes that when they were photoexcited, resulting in a reduction of the
are engineered to be anywhere from sub-picosecond to tens overall inductance [111]. In numerical simulations, however,
of picoseconds [107]. Through femtosecond laser excita- this blue-shifting of resonance frequency required one order
tion of GaAs/ErAs nanoisland superlattice substrate, Chen of magnitude higher in conductivity over the silicon strips
et al. demonstrated an ultrafast switching (20 ps) of the in the previous case, which could not be easily achieved by

Figure 13 (online color at: www.lpr-

journal.org) Scanning electron microscopy
images of (a) an individual unit cell and
(b) a square array of electric SRRs where
silicon strips were incorporated at the split
gap, resulting in the frequency agile THz
metamaterial upon photoexcitation. (c) Ex-
perimentally measured THz transmission
amplitude spectra at various photoexcita-
tion power levels. (d) Numerical simulation
results with the silicon conductivity esti-
mated from the photoexcitation fluence.

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

12 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

photoexcitation. Shen et al. revised the designs using an elec- constant resulted in a tunable resonance frequency and
tric SRR variant unit cell structure with multiple loops and thereby tunable permeability [65]. Another approach uses
split gaps, and incorporated silicon at the side gaps but not insulator-to-metal transition (IMT) materials, in which the
the central gap [112]. After photoexcitation, a blue-shifting dielectric constant and conductivity can experience orders
of the resonance frequency on the order of 40% was ob- of magnitude change when varying the temperature. Such
served in numerical simulations due to the short-circuiting materials can be also used as the metamaterial substrate or
of the side gaps. The tunability of a metamaterial resonance to compose the metamaterial resonant elements to achieve
frequency is of particular importance when considering the resonance tuning.
large frequency dispersion and narrow bandwidth in meta- Vanadium oxide (VO2 ) is a typical IMT material which
materials. The creation of frequency agile metamaterials undergoes a thermally triggered insulator-to-metal phase
would extend the spectral range over which devices function transition near 340 K. Driscoll et al. investigated the tem-
and also enable the manufacture of new devices such as perature dependent resonant properties of an SRR array
dynamically tunable notch filters. fabricated on a 90-nm-thick VO2 film atop a sapphire sub-
strate [114], as shown in Figs. 14a and 14b. The local electro-
magnetic fields of the SRR overlap the VO2 layer as shown
4.2. Thermally tunable terahertz metamaterials in Fig. 14c and interact with the VO2 material response. In-
creasing the temperature to its insulator-to-metal phase tran-
The dielectric constant of many natural materials, such as sition, the divergent permittivity of VO2 results in a modified
ferroelectrics, depends on temperature. They can serve as gap capacitance in SRRs and therefore decreases the reso-
the metamaterial substrate or compose part of the metama- nance frequency about 20% as shown in Fig. 14d. Thicker
terial resonant elements, so that resonance properties can VO2 film should increase the frequency tuning range [114].
be tuned by changing temperature. For example, when an Due to the increasing conductivity with temperature in VO2
electric SRR array is fabricated on single crystal strontium films, the resonance shifting is also accompanied by damp-
titanate (STO) substrate, the resonance frequency can be ing, similar to the resonance switching by photoexcitation of
continuously tuned, from 0.223 THz to 0.128 THz when tem- semiconducting substrates described in the previous section.
perature decreases from 409 K to 150 K, while the resonance One interesting and important phenomena observed in such
strength (or quality factor) has no significant change [113]. thermally tunable THz metamaterials is the memory effect
Němec et al. demonstrated a negative permeability THz due to the hysteretic behavior in the phase transition [115].
metamaterial composed of ferroelectric materials with high In this demonstration, although the phase transition in the
dielectric constant. The temperature dependent dielectric VO2 film was caused by the application of electrical pulses,

Figure 14 (online color at: www.lpr-journal.org)

Resonance tuning by temperature using SRRs fab-
ricated on VO2 films. (a) Close-up of the SRR gap
on top of a near-field image of the VO2 film during
phase transition. (b) Device layout and experimen-
tal setup. (c) A single SRR plotted with resonant
electric field amplitude from simulations. (d) Experi-
mentally measured transmission spectra plotted as
1 T at increasing sample temperatures. Figure
adapted from Driscoll et al. [114].

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 13

the tunability and memory are inherently thermal in nature

since the electrical pulses change the local temperature in
the VO2 film [115].
Wen et al. investigated the resonance in a cut-wire array
made from VO2 film on silica glass [116]. When tempera-
ture increases from 30 °C to 70 °C, the conductivity of VO2
film increases about 200-fold, changing from moderately
insulating to highly metallic. This resulted in the appearance
of a cut-wire resonance at 0.6 THz with a modification in
transmission amplitude over 65%. Seo et al. investigated
THz transmission through a gold slot antenna array on a
VO2 film to achieve a thermally tunable resonant THz re-
sponse [117]. The gold slot antenna array was fabricated
on top of a 100-nm-thick VO2 film. Using a narrow slot
antenna, they realized an unprecedented extinction ratio of
over 1000 when the underlying VO2 thin film experienced a
phase transition. To put this in perspective, for a VO2 film
without the nano slot antenna array, the THz transmission
in the high-temperature phase (metallic) remains as high as
30% compared to the low-temperature phase (insulating).
Ultra-broadband operation was accomplished by using an
array of so-called supercell containing nano slot antennas
with various lengths [117].

4.3. Electrically switchable metamaterials:

modulation of terahertz amplitude and phase

Metamaterials that are dynamically tunable by photoexci-

tation or temperature depend on powerful lasers or slow
thermal processes. Therefore it is quite attractive and practi-
cal to have an electrical approach to control the metamaterial
response. Room-temperature electrically switchable meta-
materials were first created by fabricating planar metallic
metamaterials on a thin n-doped GaAs layer, which was
epitaxially grown through MBE on an intrinsic GaAs sub-
strate [118], as shown in Fig. 15a. Due to the conducting sub-
strate, the resonance of SRRs is almost completely damped.
However, the gold electric SRRs and the n-GaAs form a
Schottky diode structure so that the application of a reverse
voltage bias is able to deplete charge carriers away from
the split gaps, schematically shown in Fig. 15b for another
unit cell structure [119]. The modification of the substrate
conductivity, i. e. from highly conducting to insulating, re-
covers the metamaterial resonant response and consequently
tunes the THz transmission shown in Figs. 15c and 15d for
the amplitude and phase spectra, respectively, as a function
of the applied voltage bias. The first generation device was
capable of THz intensity modulation of 50% at the reso- Figure 15 (online color at: www.lpr-journal.org) (a) Design
nance frequency [118], which was increased to 80% with an schematic of the active THz metamaterial device capable of elec-
improved metamaterial design [119], a state-of-the-art THz trical control of the metamaterial resonance and THz transmission.
modulation performance unlike anything available using (b) An improved design of the device unit cell with the depletion
naturally existing materials. This concept was also used in region shown. The THz transmission amplitude (c) and phase
switching the extraordinary transmission due to surface plas- (d) spectra at various voltage biases reveal the voltage tunable
mon resonances in metal periodic hole arrays [120]. Due metamaterial resonance. The correlated transmission amplitude
to the relatively large dimensions of the resonant element, and phase shifting both contribute to (e) a broadband modulated
the electric field near the split gaps extends well beyond THz signal roughly covering the frequency range between the
the thickness of the doping layer. This leaves the resonance two resonances.

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

14 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

Figure 16 (online color at: www.lpr-

journal.org) (a) Design schematic il-
lustrating the terahertz spatial light

modulator consisting of a 4 4 pixel
array, and each pixel is independently
controlled by an external voltage be-
tween the Schottky electric pad and
the ohmic contact. (b) A transmission
image of the THz spatial light mod-
ulator at the resonance frequency
of 0.36 THz, with two pixels turned
off (biased), and the rest turned on
(zero bias).

frequency largely unchanged even though the change of car- THz-TDS, typically a bulky mechanical chopper is used and
rier density due to depletion results in a modified dielectric the modulation speed is limited to the kilohertz range. The
constant in the doped layer. In order to operate at higher electrically driven metamaterial THz modulator is compact
frequencies, it is necessary to scale down the dimensions and integrable, and has demonstrated a modulation speed
of the metamaterial resonant element and increase the sub- of 2 MHz [124]. Higher modulation speed is possible, by
strate doping, which makes the resonance frequency more reducing the overall capacitance of the structure as a whole.
sensitive to the modification of dielectric constant in the The active THz metamaterials were employed by Chan et
doped layer. Essentially this makes frequency tuning domi- al. to demonstrate a spatial light modulator for terahertz
nant over resonance switching [121]. This mechanism was beams [125]. The 4  4 pixels THz spatial modulator is
responsible for the metamaterial devices demonstrated to schematically shown in Fig. 16a, where each pixel consists
externally modulate THz quantum cascade lasers at 2.4 and of thousands of electric SRRs which were linked all to-
2.8 THz, with a maximum modulation depth of 60% and a gether. Voltage biases are applied to each pixel so that the
large degree of modulation linearity [122]. resonance and therefore the THz transmission is indepen-
In addition to switching the THz transmission ampli- dently controlled, resulting in a spatially modulated THz
tude, its phase also varies as a function of the applied volt- beam. Figure 16b shows a transmission image of the 4  4
age bias, as shown in Fig. 15d. Under a voltage bias of 16 pixel array at 0.36 THz, with two pixels turned off (biased),
volts, a phase shift of Δφ = π =6 was accomplished for such and the rest turned on (zero bias). The device is capable
a single planar metamaterial layer, which makes it an ef- of high modulation depth at the resonance frequency of
ficient solid state THz phase modulator [119]. It was also 0.36 THz with a negligible amount of crosstalk among pix-
found that, from the results shown in Figs. 15c and 15d, els. A modulation of THz wavefront was also accomplished
the amplitude switching and phase shifting as a function by appropriately applying the voltage bias to the metama-
of voltage bias are correlated through the Kramers-Kronig terial pixel array to simulate double-slit diffraction. Such a
(KK) relations. Near frequencies where the amplitude is not spatial THz modulator would possibly find applications in
strongly dependent on the applied bias voltage, but its slope THz imaging and communication systems.
is, the phase experiences a maximum shift, and vice versa.
Thus, although the amplitude modulation and phase shifting
are inherently narrowband phenomena associated with the
resonant nature of metamaterials, both the correlated am- 4.4. Reconfigurable terahertz metamaterials
plitude and phase modulation contribute to the total THz
modulation signal, making the device capable of a broad- In planar THz metamaterials, it is the electrical field that can
band modulation. This broadband modulation is shown in be efficiently coupled, while the coupling of the magnetic
Fig. 15e, with the bandwidth roughly covering the frequen- field can only be realized with an oblique incidence angle
cies between the two resonances. The device was employed or out of plane SRRs, the latter creating significant fabrica-
in THz time-domain measurements to externally modulate tion challenges. Tao et al. developed a unique solution in
the broadband THz radiation, and satisfying results were accomplishing reconfigurable THz metamaterials based on
obtained in measuring the transmission spectrum through a MEMS technology [51]. In this approach, the free standing
second metamaterial sample [119]. gold SRRs are attached to two SiNx cantilever legs, which
While aforementioned devices are polarization sensitive are also coated with gold, on the supporting substrate. By
due to connecting wires among SRRs, a more symmetric changing the temperature, the difference in the thermal ex-
design of resonators and very thin connecting wires enables pansion coefficients of SiNx and gold results in bending of
polarization independent operation [123]. In many THz ap- the cantilever legs. The SRRs would therefore bend out of
plications, e. g., THz imaging, the modulation speed plays the plane of the substrate, as shown in Fig. 17a, which leads
an important role in improving a system performance. In to dramatic changes in the electromagnetic response.

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 15

Figure 17 (online color at: www.lpr-journal.org)

(a) SEM pictures showing the bending of the SRRs
as fabricated (first picture) and after rapid thermal an-
nealing at 350, 400, and 450 °C. Transmission spectra
for various orientations (temperatures) of the SRRs
are shown in (b) magnetic response and (c) electric
response, with the polarization of the incident THz ra-
diation illustrated on left and right in (d), respectively.
Additional transmission spectra show (e) bianisotropic
response and (f) nonresonant response, with the polar-
ization on the left and right in (g). Figure adapted from
Tao et al. [51].

Figures 17d and g illustrates the four possible configura- mation of an elastomeric substrate to controllably modify
tions of polarization for the normally incident THz radiation the dimensions of resonators and therefore the metamaterial
while the SRRs are popped out of the substrate plane. In the resonance frequency [126]. Elastic tuning was achievable
first configuration illustrated on the left in Fig. 17d, the mag- up to 10% of the deformation (i. e. the resonance frequency
netic field has a component perpendicular to the SRR plane, would be recovered after relaxation), while further increase
and therefore the magnetic resonance is observed as shown of the deformation would result in a permanent change of
in Fig. 17b for various SRR bending angles (temperatures). the dimensions and the resonance frequency.
This becomes stronger for increasing SRR bending angles
corresponding to increasing annealing temperature. In this
case, there is no fundamental resonance (LC resonance) ex- 4.5. Superconducting metamaterials
cited by the electric field. In the second configuration illus-
trated on the right in Fig. 17d, the electric field is efficiently The conductivity of metal depends on temperature, which
coupled to the fundamental electric resonance at low bend- results in a temperature-dependent metamaterial resonance
ing angles (temperatures), and the coupling decreases as the strength though it is quite marginal [127]. In superconduc-
bending angle (temperature) increases, as shown in Fig. 17c. tors, the temperature dependence of conductivity is much
There is no magnetic coupling because the magnetic field more significant, and therefore the metamaterial resonance
is always in the SRR plane. In the third configuration illus- tuning is remarkable. There have been only a few prelimi-
trated on the left in Fig. 17g, both the electric and magnetic nary recent metamaterial demonstrations incorporating su-
fields excite the fundamental LC resonance and the structure perconducting films, but the interest is growing. Part of the
exhibits a bianisotropic response. The transmission spectra goal of superconducting metamaterials, mainly focusing on
in Fig. 17e reveal the sharp resonance dip which does not the microwave frequency range, is to reduce the resonant
depend on the bending angle. Finally, with the configuration losses and improve the resonance quality factor [128–131],
illustrated on the right in Fig. 17g, neither the electric field since at low temperature superconducting materials possess
nor the magnetic field is able to excite the resonance at any superior conductivity than metals at frequencies up to low
bending angle, as is obvious in the transmission spectra in THz. It was found that the metamaterial resonance, includ-
Fig. 17f, which shows no resonance dip. The variation in ing resonance strength and frequency, could be tuned by
transmission amplitude with bending angle is due to the applying external dc or rf magnetic fields [132, 133] in addi-
high order resonances. tion to changing the temperature. It has also been extended
The above reconfigurable THz metamaterials were to the THz frequency range using high temperature super-
achieved using a thermal approach to control the orientation conducting (HTS) films [134, 135], where THz radiation
of the SRRs in the unit cells, and in principle they could would not easily break Cooper pairs and the low tempera-
be also realized by other MEMS-based approaches such as ture THz conductivity is still rather high enabling a strong
electrostatic or piezoelectric actuation [51]. Another type of THz metamaterial resonant response. In these works, planar
mechanical tuning of metamaterial resonance was demon- superconducting THz metamaterials were made from HTS
strated in the near-infrared range but with the potential at cuprate films of YBa2 Cu3 O7 δ (YBCO) with the supercon-
THz frequencies. It employed high-strain mechanical defor- ducting transition temperature Tc  90 K. Square arrays of

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

16 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

Figure 18 (online color at: www.lpr-journal.org) Microscopy image of (a) a split-ring resonator array and (b) a single unit cell
composed of YBCO films. The resonant THz transmission spectra at various temperatures were shown for 180 nm (c) and 50 nm
(d) thick YBCO metamaterials, clearly revealing the temperature tunable metamaterial resonant response. For the 180 nm thick YBCO
metamaterial sample, the corresponding resonant transmission minimum (e) and the resonance frequency (f) were plotted as a
function of temperature, and they were compared with results obtained from numerical simulations and theoretic modeling, using the
experimentally measured temperature-dependent complex conductivity shown in the inset to (e).

SRRs were fabricated using conventional photolithographic temperature and then a blue-shifting when the temperature
methods and wet chemical etching of the YBCO films. THz further increases, while the 50-nm-thick YBCO metamate-
time domain spectroscopy incorporated with a cryostat was rial sample only shows the red-shifting; the blue-shifting
used to characterize the metamaterials transmission proper- is not observable since the resonance is already completely
ties under normal incidence. switched off. The observed resonance switching and fre-
Gu et al. observed a significant enhancement of reso- quency tuning are more complicated than what had been
nance strength when temperature decreases below Tc with expected. Results of numerical simulations and theoretical
negligible frequency tuning [134], possibly due to the rela- modeling, shown in Figs. 18e and 18f for the 180-nm-thick
tively thick films. Using epitaxial YBCO films with smaller metamaterial sample, reveal that both the real and imagi-
thicknesses, however, Chen et al. observed interesting tuning nary parts of the temperature-dependent YBCO complex
features that were not previously observed [135]. The micro- conductivity have to be consistently taken into account to
scopic images of the metamaterials are shown in Figs. 18a correctly interpret the experimental results, particularly the
and 18b, where the thickness of the YBCO film is either blue-shifting in the resonance frequency at higher tempera-
180 nm or 50 nm. The THz transmission spectra of the fun- tures and the thickness dependent resonance tuning [135].
damental resonance (RLC resonance) are shown in Figs. 18c It should be emphasized that the resonance tuning in
and 18d, respectively, for these two metamaterial samples. superconducting metamaterials is mainly due to the large
Similar with the results of Gu et al., at temperatures above Tc , variation in the imaginary part of the complex conductivity
the resonance is rather weak (180-nm-thick sample) or not in superconducting films. It is expected that such a con-
observable (50-nm-thick sample). By decreasing the temper- ductivity tuning would also be achieved by application of
ature below Tc , the resonance is dramatically increased, and optical excitation and electrical and magnetic fields. No-
is accompanied by a significant resonance frequency tuning. tably, the tuning mechanism in superconducting materials
However, the differences are also obvious: 1) The resonance is a property of the conductors, in contrast to other tun-
for 50-nm-thick YBCO metamaterial sample shows a lower able metamaterials which rely on changing properties of the
resonance frequency, though the structure dimensions are conductors’ environment.
identical with the 180-nm-thick YBCO metamaterial sam-
ple; 2) The resonance strength is switched more efficiently
in the 50-nm-thick YBCO metamaterial sample, and the 5. Sensing with terahertz metamaterials
resonance frequency tuning range is also larger; 3) In the
180-nm-thick YBCO metamaterial sample, the resonance After the development of tabletop time-domain THz spec-
frequency first experiences a red-shifting with increasing troscopy (THz-TDS), one of the first proposed applications

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 17

was non-invasive sensing. As with any other frequency als. This is a result of our inability to decrease their radiation
range, materials exhibit unique THz resonances and sig- resistance [148]. Multiple resonant modes of the SRRs can
natures that may prove useful in label-free biological and be coupled to favorably reshape the response [22], an effect
chemical sensing. In addition, the non-ionizing nature of that was utilized in the work by Debus et al. [142]. The
THz waves makes them well-suited to delicate chemical and third issue is analyte composition. Most THz analytes are,
biological environments. In spite of the slow development of course, non-magnetic. Therefore we rely on the alter-
of practical THz sources and detectors, research continues ation of electric flux, or equivalently the ring capacitance,
toward the goal of improved THz sensing. Non-invasive to induce measurable changes in the SRR resonance. High
detection of explosives and sensing of cancer, DNA or other sensitivity requires the analytes to substantially alter the
biological species have been the subject of many feasibility electric flux, either through absorption or modification of the
studies. Such studies are often performed with quantities flux density. Fortunately, small losses within the SRR gap
or concentrations much higher than would be available in a can significantly dampen resonances [118], which supports
practical application. Therefore, the key to forward progress high-sensitivity. However, THz metamaterials are often fab-
is improving sensitivity. Waveguide-based THz sensors can ricated on silicon or gallium arsenide, both of which have
improve the sensitivity of thin films by dramatically in- very large dielectric constants in the THz regime. As such,
creasing the wave-material interaction length [136]. THz the substrates contain most of the electric flux surrounding a
micro-resonators and filters [137–139] improve sensing by resonating SRR, rendering that flux inaccessible to analyte
using analytes to shift their natural resonance. For example, interception. This will be a major detriment to sensing unless
using the micro-resonator method, THz sensitivity to the suitable THz metamaterials can be made free-standing or
binding state of DNA was improved about 1,000 times over on thin substrates with low dielectric constants [41–44, 49].
conventional TDS [137]. This is an inherently sensitive pro- The final issue is resonance matching. Here, the idea is to
cess with high-Q resonators and shares some qualities with design metamaterial resonances that coincide with analyte
metamaterial sensing, as will be revealed. resonances. Resonant dielectric constants can be large and
Originally noted in terms of improving nonlinear phe- highly dispersive, which favorably enhances the analyte’s
nomena [3], the resonant behavior of SRR-based metama- effect on the SRR capacitance. But further enhancement
terials leads to a localization and concentration of electric may be gleaned through coupled-oscillator methods. Strong
and/or magnetic flux in the SRR structure as well as high coupling creates effects beyond the simple superposition of
dispersion in reflection/transmission spectra [140], which multiple resonance effects. These include resonance split-
can be exploited for the purpose of THz sensing. Very strong ting [149, 150], shifting, and changes to absorption [151],
resonances thereby enable high-sensitivity by amplifying all of which may be exploited. As already alluded, reso-
the effects of minor perturbations. Small shifts in resonance nance matching may also greatly improve discrimination by
frequency, induced via dielectric loading for example, can providing numerous simultaneous analyte signatures [145].
change the amplitude response by tens of decibels. Metama- The future of sensing with THz metamaterials seems
terials began to be explored as THz sensors in a few demon- optimistic. As a path forward, research will likely focus on
strations utilizing thin-film overlayers deposited on single creating new SRRs that feature higher Q-factors and greater
layer metamaterials [141–144]. In one of these cases [142], field localization, both of which would enhance every aspect
asymmetric double-gap SRRs were designed to steepen of sensing.
a resonance flank, thereby enhancing sensitivity. Another
demonstrations used metamaterials in which several prim-
itive unit cells are tiled together to create a multi-resonant 6. Conclusions
response [145, 146]. This can be tuned to align with the nat-
ural modes of a particular analyte in order to provide better Metamaterials have provided researchers with a fresh per-
selectivity or discrimination in sensing. Sensing with THz spective on controlling electromagnetic radiation, a fact
metamaterials is still quite immature, but its motivation and that has fundamentally and practically important implica-
technical approach are becoming clear. The combination of tions, especially in the THz frequency regime. With the
strong fields, high dispersion, and localization underpin the ongoing discovery of electromagnetic phenomena and the
substantial motivation for the development of THz metama- unceasing progress in fabrication techniques, new active
terial sensors. and passive THz devices such as wave modulators, sen-
O’Hara et al. summarized a few general issues that gov- sors, absorbers, and polarization-control components, are
ern the effectiveness of sensing with metamaterials [144]. rapidly being developed. Leveraging the strong interaction
First, no sensing modality will perform well without ensur- of tuned metamaterials with electromagnetic waves, it is not
ing sufficient interaction between the electromagnetic fields uncommon for these devices to exhibit orders of magnitude
and the analyte. In this sense, SRRs that perform better will improvements in performance over the previous state-of-the-
more effectively localize electromagnetic fields within their art. Furthermore, existing THz techniques have been used
gap where analytes are placed [147]. However, the distri- for high-quality metamaterial demonstrations including me-
bution of THz fields near the gap still occupies a volume chanically reconfigurable and superconducting metamate-
much larger than the analyte volumes of interest in high- rials, often revealing formerly unknown physics. Indeed,
sensitivity sensing systems. The second issue is resonator the marriage of metamaterials and THz has proven very
Q-factor, which is limited to  10 in most THz metamateri- symbiotic and fruitful, as evidenced by the ever-increasing

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

18 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

number of publications. In the near term, we expect this

trend to continue. In the long term, it is likely that these John F. O’Hara received his B. Sc. in
scientific pursuits will lead to more dedicated technology Electrical Engineering in 1998 from the
development, hopefully culminating in a variety of prac- University of Michigan and his Ph. D.
tical and real-world THz devices, and finally closing the in electrical engineering from Oklahoma
State University in 2003. From there, he
THz gap.
began studies in terahertz surface waves
and scattering as an Intelligence Commu-
nity Postdoctoral Fellow in the Condensed
Matter and Thermal Physics group at
Acknowledgements. We acknowledge support from the Los
LANL. In 2006, this work transitioned into
Alamos National Laboratory LDRD Program. This work was per-
studies of THz metamaterials, which he has continued as a
formed, in part, at the Center for Integrated Nanotechnologies,
Staff Scientist in CINT/LANL. His research interests include
a US Department of Energy, Office of Basic Energy Sciences
synthetic materials from RF to infrared, THz technologies,
Nanoscale Science Research Center operated jointly by Los
electromagnetic modeling, and nonlinear phenomena.
Alamos and Sandia National Laboratories. Los Alamos National
Laboratory, an affirmative action/equal opportunity employer, is
operated by Los Alamos National Security, LLC, for the National Antoinette (Toni) Taylor is the Leader of
Nuclear Security Administration of the US Department of Energy the Materials Physics and Applications
under contract DE-AC52-06NA25396. Division at LANL. Prior to this position
she was Director of the Center for In-
Received: 20 December 2010, Revised: 14 March 2011, tegrated Nanotechnologies, a joint San-
Accepted: 22 March 2011 dia/LANL Nanoscience Research Center
Published online: 3 May 2011 funded through BES. Her research inter-
ests include the investigation of ultrafast
Key words: Metamaterials, terahertz, split ring resonators, dynamical processes in materials and the
inductive-capacitive resonance, solid state device, active device, development of novel ultrafast measure-
passive properties. ment techniques for the understanding of new phenomena.
Currently, she is a Member-at Large, Division of Laser Science
of the American Physical Society and OSA’s representative to
the Joint Council of Quantum Electronics. She is a LANL Lab-
oratory Fellow and a Fellow of the American Physical Society,
Hou-Tong Chen received his B. S. and the Optical Society of America and American Association for
M. S. from University of Science and Tech- the Advancement of Science.
nology of China (USTC) in 1997 and
2000, respectively, and a Ph. D. from
Rensselaer Polytechnic Institute (RPI), References
USA, in 2004, all in Physics. After a short
period of Postdoc Fellow at RPI, from [1] J. B. Pendry, A. J. Holden, W. J. Stewart, and I. Youngs,
2005 to 2008 he was a Postdoctoral Re- Phys. Rev. Lett. 76, 4773–4776 (1996).
search Associate at Los Alamos National [2] J. B. Pendry, A. J. Holden, D. J. Robbins, and W. J. Stewart,
Laboratory (LANL). He holds a Technical J. Phys. Condens. Matter 10, 4785–4809 (1998).
Staff Member position since 2008 in the Center for Integrated [3] J. B. Pendry, A. J. Holden, D. J. Robbins, and W. J. Stew-
Nanotechnologies (CINT) at LANL, and is a CINT Scientist art, IEEE Trans. Microwave Theory Tech. 47, 2075–2084
since 2010. His research areas include metamaterials and (1999).
devices, nanophotonics and plasmonics, ultrafast optics and [4] D. R. Smith, W. J. Padilla, D. C. Vier, S. C. Nemat-Nasser,
terahertz time domain spectroscopy, and terahertz scanning and S. Schultz, Phys. Rev. Lett. 84, 4184–4187 (2000).
near-field microscopy. [5] R. A. Shelby, D. R. Smith, and S. Schultz, Science 292,
77–79 (2001).
Abul Azad received his M. S. and Ph. D. [6] V. G. Veselago, Sov. Phys. Usp. 10, 509–514 (1968).
degree in Electrical and Computer En- [7] V. M. Shalaev, Nature Photon. 1, 41–48 (2007).
gineering from Oklahoma State Univer- [8] B. Wood, Laser Photon. Rev. 1, 249–259 (2007).
sity in 2003 and 2006, respectively. From [9] G. Goussetis, A. P. Feresidis, and A. R. Harvey, J. Mod. Opt.
June 2006 to January 2007 he was work- 57, 1–16 (2010).
ing as a Postdoctoral Research Asso- [10] W. Cai and V. Shalaev, Optical Metamaterials: Fundamen-
ciate in the Department of Physics in RPI. tals and Applications, 1st edn. (Springer, New York, 2010).
He joined LANL in January 2007 as a [11] P. U. Jepsen, D. G. Cooke, and M. Koch, Laser Photon. Rev.
postdoc. Currently he is a Technical Staff 5, 124–166 (2011).
Member in CINT/LANL. His research in- [12] R. D. Averitt and A. J. Taylor, J. Phys. Condens. Matter 14,
terests are in the area of metamaterials, plasmonics, ultrafast R1357–R1390 (2002).
[13] T. J. Yen, W. J. Padilla, N. Fang, D. C. Vier, D. R. Smith,
terahertz spectroscopy, ultrafast properties of materials, and
J. B. Pendry, D. N. Basov, and X. Zhang, Science 303, 1494–
microwave antennas.
1496 (2004).

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 19

[14] H. O. Moser, B. D. F. Casse, O. Wilhelmi, and B. T. Saw, [39] Y. L. Hor, Z. Szabo, H. C. Lim, J. F. Federici, and E. P. Li,
Phys. Rev. Lett. 94, 063901 (2005). Appl. Opt. 49, 1179–1184 (2010).
[15] N. Katsarakis, G. Konstantinidis, A. Kostopoulos, R. S. Pen- [40] M. Aznabet, M. Navarro-Cia, S. A. Kuznetsov,
ciu, T. F. Gundogdu, M. Kafesaki, E. N. Economou, A. V. Gelfand, N. I. Fedorinina, Y. G. Goncharov,
T. Koschny, and C. M. Soukoulis, Opt. Lett. 30, 1348–1350 M. Beruete, O. El Mrabet, and M. Sorolla, Opt. Express 16,
(2005). 18312–18319 (2008).
[16] A. K. Azad, J. M. Dai, and W. L. Zhang, Opt. Lett. 31, 634– [41] H. Tao, A. C. Strikwerda, K. Fan, C. M. Bingham,
636 (2006). W. J. Padilla, X. Zhang, and R. D. Averitt, J. Phys. D: Appl.
[17] S. Linden, C. Enkrich, M. Wegener, J. F. Zhou, T. Koschny, Phys. 41, 232004 (2008).
and C. M. Soukoulis, Science 306, 1351–1353 (2004). [42] F. Miyamaru, M. W. Takeda, and K. Taima, Appl. Phys.
[18] J. Zhou, T. Koschny, M. Kafesaki, E. N. Economou, Express 2, 042001 (2009).
J. B. Pendry, and C. M. Soukoulis, Phys. Rev. Lett. 95, [43] X. Liu, S. MacNaughton, D. B. Shrekenhamer, H. Tao,
223902 (2005). S. Selvarasah, A. Totachawattana, R. D. Averitt, M. R. Dok-
[19] D. Schurig, J. J. Mock, and D. R. Smith, Appl. Phys. Lett. meci, S. Sonkusale, and W. J. Padilla, Appl. Phys. Lett. 96,
88, 041109 (2006). 011906 (2010).
[20] H. T. Chen, J. F. O’Hara, A. J. Taylor, R. D. Averitt, C. High- [44] H. Tao, J. J. Amsden, A. C. Strikwerda, K. Fan, D. L. Kaplan,
strete, M. Lee, and W. J. Padilla, Opt. Express 15, 1084– X. Zhang, R. D. Averitt, and F. G. Omenetto, Adv. Mater.
1095 (2007). 22, 3527–3531 (2010).
[21] W. J. Padilla, M. T. Aronsson, C. Highstrete, M. Lee, [45] C. L. Holloway, A. Dienstfrey, E. F. Kuester, J. F. O’Hara,
A. J. Taylor, and R. D. Averitt, Phys. Rev. B 75, 041102(R) A. K. Azad, and A. J. Taylor, Metamaterials 3, 100–112
(2007). (2009).
[22] J. F. O’Hara, E. Smirnova, A. K. Azad, H. T. Chen, and [46] M. Gokkavas, K. Guven, I. Bulu, K. Aydin, R. S. Penciu,
A. J. Taylor, Active and Passive Electronic Components M. Kafesaki, C. M. Soukoulis, and E. Ozbay, Phys. Rev. B
2007, 1–10 (2007). 73, 193103 (2006).
[23] A. K. Azad, A. J. Taylor, E. Smirnova, and J. F. O’Hara, [47] B. D. F. Casse, H. O. Moser, J. W. Lee, M. Bahou, S. Inglis,
Appl. Phys. Lett. 92, 011119 (2008). and L. K. Jian, Appl. Phys. Lett. 90, 254106 (2007).
[24] V. M. Shalaev, W. S. Cai, U. K. Chettiar, H. K. Yuan, [48] A. K. Azad, H. T. Chen, X. Lu, J. Gu, N. R. Weisse-
A. K. Sarychev, V. P. Drachev, and A. V. Kildishev, Opt. Bernstein, E. Akhadov, A. J. Taylor, W. Zhang, and
Lett. 30, 3356–3358 (2005). J. F. O’Hara, THz Sci. Technol. 2, 15–22 (2009).
[25] S. Zhang, W. J. Fan, N. C. Panoiu, K. J. Malloy, R. M. Os- [49] X. G. Peralta, M. C. Wanke, C. L. Arrington, J. D. Williams,
good, and S. R. J. Brueck, Phys. Rev. Lett. 95, 137404 I. Brener, A. Strikwerda, R. D. Averitt, W. J. Padilla,
(2005). E. Smirnova, A. J. Taylor, and J. F. O’Hara, Appl. Phys.
[26] G. Dolling, C. Enkrich, M. Wegener, C. Soukoulis, and Lett. 94, 161113 (2009).
S. Linden, Science 312, 892–894 (2006). [50] F. Miyamaru, S. Kuboda, K. Taima, K. Takano, M. Hangyo,
[27] G. Dolling, C. Enkrich, M. Wegener, J. F. Zhou, and and M. W. Takeda, Appl. Phys. Lett. 96, 081105 (2010).
C. M. Soukoulis, Opt. Lett. 30, 3198–3200 (2005). [51] H. Tao, A. C. Strikwerda, K. Fan, W. J. Padilla, X. Zhang,
[28] T. F. Gundogdu, N. Katsarakis, M. Kafesaki, R. S. Penciu, and R. D. Averitt, Phys. Rev. Lett. 103, 147401 (2009).
G. Konstantinidis, A. Kostopoulos, E. N. Economou, and [52] J. S. Randhawa, S. S. Gurbani, M. D. Keung, D. P. Demers,
C. M. Soukoulis, Opt. Express 16, 9173–9180 (2008). M. R. Leahy-Hoppa, and D. H. Gracias, Appl. Phys. Lett.
[29] M. Awad, M. Nagel, and H. Kurz, Opt. Lett. 33, 2683–2685 96, 191108 (2010).
(2008). [53] D. B. Burckel, J. R. Wendt, G. A. Ten Eyck, A. R. Ellis,
[30] O. Paul, C. Imhof, B. Reinhard, R. Zengerle, and I. Brener, and M. B. Sinclair, Adv. Mater. 22, 3171–3175
R. Beigang, Opt. Express 16, 6736–6744 (2008). (2010).
[31] M. Beruete, M. Sorolla, and I. Campillo, Opt. Express 14, [54] J. R. Wendt, D. B. Burckel, G. A. Ten Eyck, A. R. Ellis,
5445–5455 (2006). I. Brener, and M. B. Sinclair, J. Vac. Sci. Technol. B 28,
[32] C. Croenne, F. Garet, E. Lheurette, J. L. Coutaz, and D. Lip- C6O30 (2010).
pens, Appl. Phys. Lett. 94, 133112 (2009). [55] D. Grischkowsky, S. Keiding, M. van Exter, and C. Fattinger,
[33] S. X. Wang, F. Garet, K. Blary, E. Lheurette, J. L. Coutaz, J. Opt. Soc. Am. B 7, 2006–2015 (1990).
and D. Lippens, Appl. Phys. Lett. 97, 181902 (2010). [56] R. Singh, E. Smirnova, A. J. Taylor, J. F. O’Hara, and
[34] S. X. Wang, F. Garet, K. Blary, C. Croenne, E. Lheurette, W. Zhang, Opt. Express 16, 6537–6543 (2008).
J. L. Coutaz, and D. Lippens, J. Appl. Phys. 107, 074510 [57] S. Y. Chiam, R. Singh, J. Gu, J. Han, W. Zhang, and
(2010). A. A. Bettiol, Appl. Phys. Lett. 94, 064102 (2009).
[35] A. Ishikawa, S. Zhang, D. A. Genov, G. Bartal, and [58] T. F. Gundogdu, I. Tsiapa, A. Kostopoulos, G. Kon-
X. Zhang, Phys. Rev. Lett. 102, 043904 (2009). stantinidis, N. Katsarakis, R. S. Penciu, M. Kafesaki,
[36] A. A. Tavallaee, P. C. Hon, K. Mehta, T. Itoh, and E. N. Economou, T. Koschny, and C. M. Soukoulis, Appl.
B. S. Williams, IEEE J. Quantum Electron. 46, 1091–1098 Phys. Lett. 89, 084103 (2006).
(2010). [59] W. J. Padilla, A. J. Taylor, C. Highstrete, M. Lee, and
[37] M. Walther, A. Ortner, H. Meier, U. Loeffelmann, R. D. Averitt, Phys. Rev. Lett. 96, 107401 (2006).
P. J. Smith, and J. G. Korvink, Appl. Phys. Lett. 95, 251107 [60] T. Driscoll, G. O. Andreev, D. N. Basov, S. Palit, T. Ren,
(2009). J. Mock, S. Y. Cho, N. M. Jokerst, and D. R. Smith, Appl.
[38] K. Takano, T. Kawabata, C. F. Hsieh, K. Akiyama, F. Miya- Phys. Lett. 90, 092508 (2007).
maru, Y. Abe, Y. Tokuda, R. P. Pan, C. L. Pan, and [61] C. Menzel, R. Singh, C. Rockstuhl, W. Zhang, and F. Led-
M. Hangyo, Appl. Phys. Express 3, 016701 (2010). erer, J. Opt. Soc. Am. B 26, B143–B147 (2009).

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

 LASER & PHOTONICS
REVIEWS

20 H.-T. Chen, J. F. O’Hara, et al.: Manipulation of terahertz radiation using metamaterials

[62] X. L. Xu, B. G. Quan, C. Z. Gu, and L. Wang, J. Opt. Soc. [90] Y. W. Chen, P. Y. Han, and X. C. Zhang, Appl. Phys. Lett.
Am. B 23, 1174–1180 (2006). 94, 041106 (2009).
[63] N. Katsarakis, T. Koschny, M. Kafesaki, E. N. Economou, [91] H. T. Chen, J. Zhou, J. F. O’Hara, F. Chen, A. K. Azad, and
and C. M. Soukoulis, Appl. Phys. Lett. 84, 2943–2945 A. J. Taylor, Phys. Rev. Lett. 105, 073901 (2010).
(2004). [92] H. T. Chen, J. Zhou, J. F. O’Hara, and A. J. Taylor, Terahertz
[64] B. G. Quan, X. L. Xu, H. F. Yang, X. X. Xia, Q. Wang, Science and Technology 3, 66–73 (2010).
L. Wang, C. Z. Gu, C. Li, and F. Li, Appl. Phys. Lett. 89, [93] L. Zhou, W. J. Wen, C. T. Chan, and P. Sheng, Phys. Rev.
041101 (2006). Lett. 94, 243905 (2005).
[65] H. Němec, P. Kužel, F. Kadlec, C. Kadlec, R. Yahiaoui, and [94] N. I. Landy, S. Sajuyigbe, J. J. Mock, D. R. Smith, and
P. Mounaix, Phys. Rev. B 79, 241108(R) (2009). W. J. Padilla, Phys. Rev. Lett. 100, 207402 (2008).
[66] R. Yahiaoui, H. Němec, P. Kužel, F. Kadlec, C. Kadlec, and [95] H. Tao, N. I. Landy, C. M. Bingham, X. Zhang, R. D. Averitt,
P. Mounaix, Opt. Lett. 34, 3541–3543 (2009). and W. J. Padilla, Opt. Express 16, 7181–7188 (2008).
[67] Q. Zhao, J. Zhou, F. Zhang, and D. Lippens, Materials Today [96] H. Tao, C. M. Bingham, A. C. Strikwerda, D. Pilon,
12(12), 60–69 (2009). D. Shrekenhamer, N. I. Landy, K. Fan, X. Zhang,
[68] G. Acuna, S. F. Heucke, F. Kuchler, H. T. Chen, A. J. Taylor, W. J. Padilla, and R. D. Averitt, Phys. Rev. B 78, 241103
and R. Kersting, Opt. Express 16, 18745–18751 (2008). (2008).
[69] A. Bitzer, H. Merbold, A. Thoman, T. Feurer, H. Helm, and [97] M. Diem, T. Koschny, and C. M. Soukoulis, Phys. Rev. B
M. Walther, Opt. Express 17, 3826–3834 (2009). 79, 033101 (2009).
[70] A. Bitzer, J. Wallauer, H. Helm, H. Merbold, T. Feurer, and [98] N. I. Landy, C. M. Bingham, T. Tyler, N. Jokerst,
M. Walther, Opt. Express 17, 22108–22113 (2009). D. R. Smith, and W. J. Padilla, Phys. Rev. B 79, 125104
[71] F. Falcone, T. Lopetegi, M. A. G. Laso, J. D. Baena, (2009).
J. Bonache, M. Beruete, R. Marques, F. Martin, and [99] D. Y. Shchegolkov, A. K. Azad, J. F. O’Hara, and
M. Sorolla, Phys. Rev. Lett. 93, 197401 (2004). E. I. Simakov, Phys. Rev. B 82, 205117 (2010).
[72] J. F. O’Hara, E. Smirnova, H. T. Chen, A. J. Taylor, [100] Y. X. Li, Y. S. Xie, H. W. Zhang, Y. L. Liu, Q. Y. Wen, and
R. D. Averitt, C. Highstrete, M. Lee, and W. J. Padilla, J. W. W. Ling, J. Phys. D.: Appl. Phys. 42, 095408 (2009).
Nanoelectron. Optoelectron 2, 90–95 (2007). [101] Q. Y. Wen, Y. S. Xie, H. W. Zhang, Q. H. Yang, Y. X. Li, and
[73] T. H. Hand, J. Gollub, S. Sajuyigbe, D. R. Smith, and Y. L. Liu, Opt. Express 17, 20256–20265 (2009).
S. A. Cummer, Appl. Phys. Lett. 93, 212504 (2008). [102] J. O’Hara, A. K. Azad, and A. J. Taylor, Eur. Phys. J. D 58,
[74] X. G. Peralta, E. I. Smirnova, A. K. Azad, H. T. Chen, 243–247 (2010).
A. J. Taylor, I. Brener, and J. F. O’Hara, Opt. Express 17, [103] P. L. Richards, J. Appl. Phys. 76, 1–24 (1994).
773–783 (2009). [104] Y. Q. Ye, Y. Jin, and S. L. He, J. Opt. Soc. Am. B 27, 498–
[75] A. C. Strikwerda, K. Fan, H. Tao, D. V. Pilon, X. Zhang, 504 (2010).
and R. D. Averitt, Opt. Express 17, 136–149 (2009). [105] Q. Y. Wen, H. W. Zhang, Y. S. Xie, Q. H. Yang, and Y. L. Liu,
[76] J. A. Kong, Electromagnetic Wave Theory, 2nd ed. (John
Appl. Phys. Lett. 95, 241111 (2009).
Wiley Sons, New York, 1990). [106] H. Tao, C. M. Bingham, D. Pilon, K. Fan, A. C. Strikwerda,
[77] J. B. Pendry, Science 306, 1353–1355 (2004).
D. Shrekenhamer, W. J. Padilla, X. Zhang, and R. D. Averitt,
[78] N. Kanda, K. Konishi, and M. Kuwata-Gonokami, Opt.
J. Phys. D: Appl. Phys. 43, 225102 (2010).
Express 15, 11117–11125 (2007).
[79] S. Engelbrecht, M. Wunderlich, A. M. Shuvaev, and A. Pi- [107] R. Prasankumar, A. Scopatz, D. Hilton, A. Taylor, R. Averitt,
menov, Appl. Phys. Lett. 97, 081116 (2010). J. Zide, and A. Gossard, Appl. Phys. Lett. 86, 201107
[80] N. Kanda, K. Konishi, and M. Kuwata-Gonokami, Opt. Lett. (2005).
34, 3000–3002 (2009). [108] H. T. Chen, W. J. Padilla, J. M. O. Zide, S. R. Bank,
[81] S. Zhang, Y. S. Park, J. Li, X. Lu, W. Zhang, and X. Zhang, A. C. Gossard, A. J. Taylor, and R. D. Averitt, Opt. Lett.
Phys. Rev. Lett. 102, 023901 (2009). 32, 1620–1622 (2007).
[82] R. Singh, E. Plum, W. Zhang, and N. I. Zheludev, Opt. Ex- [109] A. K. Azad, H. T. Chen, S. R. Kasarla, A. J. Taylor, Z. Tian,
press 18, 13425–13430 (2010). X. C. Lu, W. Zhang, H. Lu, A. C. Gossard, and J. F. O’Hara,
[83] K. Kawase and N. Hiromoto, Appl. Opt. 37, 1862–1866 Appl. Phys. Lett. 95, 011105 (2009).
(1998). [110] J. M. Manceau, N. H. Shen, M. Kafesaki, C. M. Soukoulis,
[84] C. R. Englert, M. Birk, and H. Maurer, IEEE Trans. Geosci. and S. Tzortzakis, Appl. Phys. Lett. 96, 021111 (2010).
Remote Sens. 37, 1997–2003 (1999). [111] H. T. Chen, J. F. O’Hara, A. K. Azad, A. J. Taylor,
[85] A. J. Gatesman, J. Waldman, M. Ji, C. Musante, and S. Yn- R. D. Averitt, D. B. Shrekenhamer, and W. J. Padilla, Nature
gvesson, IEEE Microw. Guided Wave Lett. 10, 264–266 Photon. 2, 295–298 (2008).
(2000). [112] N. H. Shen, M. Kafesaki, T. Koschny, L. Zhang,
[86] J. Lau, J. Fowler, T. Marriage, L. Page, J. Leong, E. Wish- E. N. Economou, and C. M. Soukoulis, Phys. Rev. B 79,
now, R. Henry, E. Wollack, M. Halpern, D. Marsden, and 161102(R) (2009).
G. Marsden, Appl. Opt. 45, 3746–3751 (2006). [113] R. Singh, A. K. Azad, Q. X. Jia, A. J. Taylor, and H. T. Chen,
[87] Y. W. Chen, P. Han, X. C. Zhang, M. L. Kuo, and S. Y. Lin, Opt. Lett. 36, 1230–1232 (2011).
Opt. Lett. 35, 3159–3161 (2010). [114] T. Driscoll, S. Palit, M. M. Qazilbash, M. Brehm, F. Keil-
[88] A. Wagner-Gentner, U. U. Graf, D. Rabanus, and K. Jacobs, mann, B. G. Chae, S. J. Yun, H. T. Kim, S. Y. Cho, N. M. Jok-
Infrared Phys. Technol. 48, 249–253 (2006). erst, D. R. Smith, and D. N. Basov, Appl. Phys. Lett. 93,
[89] Y. F. Huang, S. Chattopadhyay, Y. J. Jen, C. Y. Peng, 024101 (2008).
T. A. Liu, Y. K. Hsu, C. L. Pan, H. C. Lo, C. H. Hsu, [115] T. Driscoll, H. T. Kim, B. G. Chae, B. J. Kim, Y. W. Lee,
Y. H. Chang, C. S. Lee, K. H. Chen, and L. C. Chen, Na- N. M. Jokerst, S. Palit, D. R. Smith, M. Di Ventra, and
ture Nanotech. 2, 770–774 (2007). D. N. Basov, Science 325, 1518–1521 (2009).

© 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.lpr-journal.org

 REVIEW
ARTICLE

Laser Photonics Rev. (2011) 21

[116] Q. Y. Wen, H. W. Zhang, Q. H. Yang, Y. S. Xie, K. Chen, [133] B. B. Jin, C. H. Zhang, S. Engelbrecht, A. Pimenov, J. B. Wu,
and Y. L. Liu, Appl. Phys. Lett. 97, 021111 (2010). Q. Y. Xu, C. H. Cao, J. Chen, W. W. Xu, L. Kang, and
[117] M. Seo, J. Kyoung, H. Park, S. Koo, H. S. Kim, H. Bernien, P. H. Wu, Opt. Express 18, 17504–17509 (2010).
B. J. Kim, J. H. Choe, Y. H. Ahn, H. T. Kim, N. Park, [134] J. Gu, R. Singh, Z. Tian, W. Cao, Q. Xing, M. He,
Q. H. Park, K. Ahn, and D. S. Kim, Nano Lett. 10, 2064– J. W. Zhang, J. Han, H. T. Chen, and W. Zhang, Appl. Phys.
2068 (2010). Lett. 97, 071102 (2010).
[118] H. T. Chen, W. J. Padilla, J. M. O. Zide, A. C. Gossard, [135] H. T. Chen, H. Yang, R. Singh, J. F. O’Hara, A. K. Azad,
A. J. Taylor, and R. D. Averitt, Nature 444, 597–600 (2006). S. A. Trugman, Q. X. Jia, and A. J. Taylor, Phys. Rev. Lett.
[119] H. T. Chen, W. J. Padilla, M. J. Cich, A. K. Azad, 105, 247402 (2010).
R. D. Averitt, and A. J. Taylor, Nature Photon. 3, 148–151 [136] J. Q. Zhang and D. Grischkowsky, Opt. Lett. 29, 1617–1619
(2009). (2004).
[120] H. T. Chen, H. Lu, A. K. Azad, R. D. Averitt, A. C. Gossard, [137] M. Nagel, P. H. Bolivar, M. Brucherseifer, H. Kurz,
S. A. Trugman, J. F. O’Hara, and A. J. Taylor, Opt. Express A. Bosserhoff, and R. Buttner, Appl. Opt. 41, 2074–2078
16, 7641–7648 (2008). (2002).
[121] X. Y. Miao, B. Passmore, A. Gin, W. Langston, S. Vangala, [138] M. Nagel, F. Richter, P. Haring-Bolivar, and H. Kurz, Phys.
W. Goodhue, E. Shaner, and I. Brener, Appl. Phys. Lett. 96, Med. Biol 48, 3625–3636 (2003).
101111 (2010). [139] C. K. Tiang, J. Cunningham, C. Wood, I. C. Hunter, and
[122] X. Peralta, I. Brener, W. J. Padilla, E. Young, A. Hoffman, A. G. Davies, J. Appl. Phys. 100, 066105 (2006).
M. Cich, R. Averitt, M. Wanke, J. Wright, H. T. Chen, [140] G. Sisó, M. Gil, F. Aznar, J. Bonache, and F. Martin, Laser
J. O’Hara, A. Taylor, J. Waldman, W. Goodhue, J. Li, and Photon. Rev. 3, 12–29 (2009).
J. Reno, Metamaterials 4, 83–88 (2010). [141] T. Driscoll, G. O. Andreev, D. N. Basov, S. Palit, S. Y. Cho,
[123] O. Paul, C. Imhof, B. Laegel, S. Wolff, J. Heinrich, S. Hoe- N. M. Jokerst, and D. R. Smith, Appl. Phys. Lett. 91, 062511
fling, A. Forchel, R. Zengerle, R. Beigang, and M. Rahm, (2007).
Opt. Express 17, 819–827 (2009). [142] C. Debus and P. H. Bolivar, Appl. Phys. Lett. 91, 184102
[124] H. T. Chen, S. Palit, T. Tyler, C. M. Bingham, J. M. O. Zide, (2007).
J. F. O’Hara, D. R. Smith, A. C. Gossard, R. D. Averitt, [143] Y. M. Sun, X. X. Xia, H. Feng, H. F. Yang, C. Z. Gu, and
W. J. Padilla, N. M. Jokerst, and A. J. Taylor, Appl. Phys. L. Wang, Appl. Phys. Lett. 92, 221101 (2008).
Lett. 93, 091117 (2008). [144] J. F. O’Hara, R. Singh, I. Brener, E. Smirnova, J. Han,
[125] W. L. Chan, H. T. Chen, A. J. Taylor, I. Brener, M. J. Cich, A. J. Taylor, and W. Zhang, Opt. Express 16, 1786–1795
and D. M. Mittleman, Appl. Phys. Lett. 94, 213511 (2009). (2008).
[126] I. M. Pryce, K. Aydin, Y. A. Kelaita, R. M. Briggs, and [145] C. M. Bingham, H. Tao, X. Liu, R. D. Averitt, X. Zhang,
H. A. Atwater, Nano Lett. 10, 4222–4227 (2010). and W. J. Padilla, Opt. Express 16, 18565–18575 (2008).
[127] R. Singh, Z. Tian, J. G. Han, C. Rockstuhl, J. Q. Gu, and [146] Y. Yuan, C. Bingham, T. Tyler, S. Palit, T. H. Hand,
W. L. Zhang, Appl. Phys. Lett. 96, 071114 (2010). W. J. Padilla, N. M. Jokerst, and S. A. Cummer, Appl. Phys.
[128] M. Ricci, N. Orloff, and S. M. Anlage, Appl. Phys. Lett. 87, Lett. 93, 191110 (2008).
034102 (2005). [147] C. Debus and P. H. Bolivar, Proc. SPIE 6987, 69870U/1–
[129] M. C. Ricci and S. M. Anlage, Appl. Phys. Lett. 88, 264102 69870U/8 (2008).
(2006). [148] R. Singh, A. K. Azad, J. F. O’Hara, A. J. Taylor, and
[130] C. Kurter, J. Abrahams, and S. M. Anlage, Appl. Phys. Lett. W. Zhang, Opt. Lett. 33, 1506–1508 (2008).
96, 253504 (2010). [149] R. Singh, C. Rockstuhl, F. Lederer, and W. Zhang, Phys.
[131] V. A. Fedotov, A. Tsiatmas, J. H. Shi, R. Buckingham, Rev. B 79, 085111 (2009).
P. de Groot, Y. Chen, S. Wang, and N. I. Zheludev, Opt. [150] N. Liu, S. Kaiser, and H. Giessen, Adv. Mater. 20, 4521–
Express 18, 9015–9019 (2010). 4525 (2008).
[132] M. C. Ricci, H. Xu, R. Prozorov, A. P. Zhuravel, A. V. Usti- [151] S. Zhang, D. A. Genov, Y. Wang, M. Liu, and X. Zhang,
nov, and S. M. Anlage, IEEE Trans. Appl. Supercond. 17(2, Phys. Rev. Lett. 101, 047401 (2008).
Part 1), 918–921 (2007),

www.lpr-journal.org © 2011 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim