www.advopticalmat.de
www.MaterialsViews.com
COMMUNICATION
Four-Color Metamaterial Absorber THz Spatial
Light Modulator
David Shrekenhamer, John Montoya, Sanjay Krishna, and Willie J. Padilla*
The ability to dynamically control the response of a photonic
device to electromagnetic radiation is a very powerful concept
that has long been a goal of scientists and engineers. Of particular interest is altering the surface reflection coefficient—as
is done in the pixel of a spatial light modulator (SLM)—without
any moving parts or gratings, opening up a new paradigm for
sensing. Metamaterials are engineered electromagnetic materials which enable the precise control over light and are optimistic candidates to achieve this degree of freedom.[1–5] Here
we report a metamaterial absorber (MMA) 8 × 8 THz SLM in
which the reflection and absorption in each pixel is dynamically controlled by all-electronic means. Four central frequencies in THz were chosen for the MMA-SLM and were arranged
in a square array pattern to form a multi-color superpixel. By
altering the carrier density between the frequency selective
patterned surface and a ground plane, an average modulation
index of 62% was achieved across the pixel array with switching
speeds of up to 12 MHz. Our demonstration of a multicolor
THz MMA-SLM can be leveraged for greater control over
light, leading to more compact, efficient and versatile photonic
components.
The capability to control light dynamically, spatially, and/
or spectrally has shown numerous potential advantages;
specifically with regard to spectroscopic and imaging applications.[6–8] Commercially available digital micro-mirror
(DMD)[9] and liquid-crystal (LC)[10] devices have shown great
promise from the near-infrared to the ultraviolet, but do
not operate at longer wavelengths. These devices, however,
suffer other significant limitations. DMDs are mechanical
and can only operate up to a few kilohertz. Most LCs perform optimally near 1 kHz and are thus also relatively
slow.[6,11] Obtaining spectral information with these devices
can be challenging requiring complex, high cost equipment
and instrumentation with significantly limited acquisition
speeds.[7] If SLMs, however, had a built-in ability to discriminate between different spectral bands at the physical layer,
this would yield significant improvements while enabling a
host of applications.
Dr. D. Shrekenhamer, Prof. W. J. Padilla
Department of Physics
Boston College, 140 Commonwealth Avenue
Chestnut Hill, Massachusetts, 02467, USA
E-mail: willie.padilla@bc.edu
J. Montoya, Prof. S. Krishna
Department of Electrical and Computer Engineering
Center for High Technology Materials
University of New Mexico
Albuquerque, New Mexico, 87106, USA
DOI: 10.1002/adom.201300265
Adv. Optical Mater. 2013,
DOI: 10.1002/adom.201300265
Metamaterials can be designed to yield highly efficient
absorption by careful tailoring of their electric and magnetic
response functions. Metamaterial absorbers have sparked significant interest since their first demonstration[12,13] and hold
great potential for use in applications ranging from thermal
emitters[14] and energy harvesting,[15] to sensors and narrow
band absorbers.[16] Switching between near perfect absorption and mirror-like reflection is possible by functionalizing
metamaterial absorbers where both the effective permittivity
and permeability can be controlled within the same unit cell.
Integrating tunable absorbers as a reflection-based SLM would
offer the intensity contrast and low insertion losses found
within commerical SLMs, that are currently unavailable at THz
frequencies[17–19] due to fundamental limitations of a metasurface to generate a sufficient magnet response to match the
impedance of free space.[20] The electromagnetic properties of
metamaterials can be dynamically tuned through a variety of
different mechanisms,[21] and made to be both frequency and
spatially selective.[22] When implemented as room-temperature
THz modulators,[23,24] metamaterial based devices have demonstrated significant advantages in modulation depth, switching
speed and spectral sensitivity over alternative architectures
such as quantum well structures, which require cryogenic temperatures,[25] and LCs which possess relatively slow switching
speeds.[26] They are thus attractive candidates to bridge the technological limitations of current state-of-the-art SLMs, especially
at THz frequencies where components necessary to efficiently
manipulate THz radiation remain in need of substantial development in order to provide realistic solutions.[27]
Here we demonstrate a doped semiconducting metamaterial
SLM with multi-color super-pixels composed of arrays of electronically controlled THz metamaterial absorbers. In Figure 1a
we show a photograph of an 8 × 8 pixel implementation of the
MMA-SLM. The pixels are tiled in a square array and we modulate each at a unique frequency in the THz range. The overall
SLM system architecture is shown schematically in Figure 1b
and consists of metamaterial absorber pixels flip chip bonded
to a Silicon chip carrier with routing to bond pads which are
wire-bonded to a leadless chip carrier (LCC). Individual pixels
are biased through a commercial 64 channel output controller
and each THz color pixel consists of approximately 400 individual metamaterial absorber unit cells.
Our metamaterial absorber consists of two metallic layers
with a dielectric spacer lying in-between. The top metal layer
is patterned in order to respond resonantly to the electric
component of an incident electromagnetic wave. A bottom
ground plane layer is spaced relatively close to the top layer,
thus allowing the external magnetic field to couple, as shown
in Figure 1b. Altering the geometry of the metallic pattern and
dielectric thickness allows tuning of the impedance and loss,
© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
www.advopticalmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 1. Design and structure detail of the electronically controlled THz metamaterial absorber based spatial light modulator (MMA-SLM). (a) Image
of MMA-SLM as assembled in chip carrier package. (b) Cross-sectional schematic view of a single pixel. (c,d) Optical microscope images for each
color pixel with dimensions as listed in Table 1.
thus enabling resonant absorption at nearly any frequency. In
our implementation we utilize a 2 μm (<λ/35) thick n-doped
(2.0 × 1016 cm−3) epitaxial layer of GaAs as the dielectric spacer.
A junction is formed from the electric inductive capacitive
(ELC) resonator and serves as the Schottky contact, while the
bottom ground plane is used as an Ohmic contact. This configuration enables depletion of carriers in the epi-layer underneath the ELCs as a function of applied reverse bias and results
in tuning of the absorption peak and modulation of the electromagnetic spectrum.
Optical microscope images of the MMA-SLM are shown in
Figure 1c,d and the dimensions of each color pixel is listed
in Table 1. Each pixel has a separate ground plane which provides electrical isolation to enable independent modulation,
and necessitates fabrication of mesas extending through the
epi-layer to define a single pixel. The metamaterial unit cells
Table 1. Dimensions for each frequency (THz) color pixel as relating to
Figure 1c,d. The line width w is 3 μm and the pixel pitch p is 600 μm are
held constant for each pixel design.
dimension
[micrometers]
2
2.72 THz
3.27 THz
3.81 THz
4.34 THz
a
25
27
28
30
l
19.2
15.2
12.5
10.6
f
500
486
476
480
wileyonlinelibrary.com
are connected to their neighbors via horizontal metallic wires.
The arrays have a pitch, p, of 600 μm and the isolated Ohmic
contact ground planes are centered about each pixel with a side
length of 570 μm. The mesas lie in-between the ground planes
and have a width of 10 μm (see Supporting Information for fabrication details).
In Figure 2a–d we show the measured frequency dependent
polarized reflectance R(ω) (bottom panels) for zero applied bias
and for a reverse bias of Vbias = −26.5 V. The modulation index,
defined as M(w ) ≡ 100% × R−26.5V (w ) − R0V (w ) / Rmax (w ), is
also shown in the top panels of Figure 2a–d. When a metamaterial pixel is unbiased, the GaAs epi-layer has a maximum
carrier concentration (nGaAs). As Vbias is increased, depletion
of nGaAs results in a monotonic redshift of the reflectance
minimum until saturation, as shown for each color pixel in
Figure 2e. A maximum Vbias = −26.5 V results in sufficient
depletion of nGaAs to shift the frequency of the reflectance
minimum by nearly 5%, in the case of the lowest frequency
color pixel. As the resonance redshifts the minimum reflectance value for each pixel decreases, with the largest decrease
occurring at lower frequencies where the minimum value
decreases by 0.12. Values for the modulation index near ω0
average about 54%, but are larger on the low frequency side
of ω0, for each resonance, and we find an average value of
M = 62%.
In many applications, amplitude modulation over a narrow
band is desired. As an example we consider operating at a fixed
© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Optical Mater. 2013,
DOI: 10.1002/adom.201300265
www.advopticalmat.de
www.MaterialsViews.com
Frequency (THz)
COMMUNICATION
Frequency (THz)
simulations (see the Supporting Information
for computational details). Recent computational studies have looked at the relation
0.5
0.5
0.0
0.0
between the semiconductor physics and elec-0.5
-0.5
tromagnetic response of metamaterials.[28]
1.0
1.0
We model the frequency dependent dielec0.8
0.8
tric permittivity of the epi-layer ε˜Ga As(ω)
2.72 THz
0.6
0.6
using Drude theory where, at THz frequenV (Volts) n
(cm )
0.4
0.4
3.27 THz
cies, n-doped GaAs has been well character0
2e16
0.2
0.2
-26.5
1e14
ized for similar doping levels.[29,30] We model
0.0
0.0
the depletion to be directly underneath the
2.8
3.0
3.2
3.4
3.6
2.4
2.6
2.8
3.0
3.2
Frequency (THz)
Frequency (THz)
Schottky junction formed at the Au:GaAs
Frequency (THz)
Frequency (THz)
interface and, at max bias, to be completely
3.4
3.6
3.8
4.0
4.2
4.0
4.2
4.4
4.6
4.8
c
d
depleted across the 2 μm epi-layer thickness.
0.5
0.5
Results from our computational model are
0.0
0.0
plotted as the dashed curves in Figure 2a–d
-0.5
-0.5
and show excellent agreement with experi1.0
1.0
ment. Notably, as nGaAs is changed from
0.8
0.8
2 × 1016 cm−3 to 1 × 1014 cm−3 the tunability
4.34
THz
0.6
0.6
3.81 THz
in the reflectance saturates as shown in
0.4
0.4
Figure 3a, in agreement with experiment.
0.2
0.2
The pixels perform collectively as a SLM
0.0
0.0
and we individually characterize the tunable
3.4
3.6
3.8
4.0
4.2
4.0
4.2
4.4
4.6
4.8
Frequency (THz)
Frequency (THz)
reflectance for each pixel. In Figure 3a a spatial map of the maximum absolute change in
e 4.35
f
reflectance | R(ω )| = | R−26.5V(ω ) − R0V(ω )| is
2.72 THz
4.30
0.70
3.27 THz
4.25
shown. Our THz MMA-SLM demonstrates
3.81 THz
3.80
4.34 THz
tunable reflectance that spans from 0.25 to
3.75
0.55
over 0.50 change in total reflectance. We note
3.70
3.28
that, since each of our pixels has a filling
3.20
0.40
fraction f of 1/4, that R(ω) and ΔR(ω) values
3.12
2.72
shown should be multiplied by f when using
2.64
0.25
the area of the entire SLM, but specified
2.56
modulation index values will be unchanged.
-25
-20
-15
-10
-5
0
-25
-20
-15
-10
-5
0
Our fabricated device had two dead pixels,
VBias (Volts)
VBias (Volts)
displayed as black in Figure 3a, which is
therefore unmodulated.
Figure 2. Experimentally measured and computationally simulated frequency dependent reflecThe frequency at which the unbiased
tance and modulation index (defined in the text) of the metamaterial absorber shown for each
minimum in reflectance occurs is displayed
color sub-pixel. (a–d) (Bottom panels) The experimental curves are shown for applied bias
voltage Vbias equal to both 0 V (blue) and −26.5 V (gold), with the simulated curves shown as a false color plot in Figure 3b. The 2 × 2
multi-color subarray, outlined in white, is
for carrier concentration of the n-doped GaAs epitaxial layer nGaAs equal to 2.0 × 1016 cm−3
(blue dashed) and 1.0 × 1014 cm−3 (gold dashed). (Top panels) Simulated modulation index replicated across the entire 8 × 8 spatial array
are shown as the dashed gray curves in the top panels of (a–d). The colored vertical dashed
with each color appearing distinct from one
lines indicate frequency of the unbiased reflectance minimum, Rmin (Vbias = 0) = ω0, for their another. The operational frequency of each of
respective sub-pixels. (e) Frequency location of Rmin as a function of Vbias for each color subthe four color pixels was chosen to minimize
pixel, color shading on inset image indicates location within the super-pixel. (f) The reflectance
values R(ω0) for ω0 as indicated by the vertically dashed color lines in Figure 2a–d, shown as a spectral overlap, which allows for both spatial
and spectral information to be gathered in
function of Vbias for each color sub-pixel.
parallel by creating an effective 4 × 4 spatially
modulated array with four uniquely controlled frequencies.
frequency ω0 which we take to be the reflectance minimum of
In order to quantify advantages our semiconductor based
the unbiased case for each pixel. In Figure 2f we plot R(ω0) as a
MMA-SLM has over current state-of-the-art devices, we charfunction of Vbias for each color pixel. We observe that the reflecacterized the modulation speed and the results are shown in
tance level changes monotonically as a function of Vbias and
Figure 3c. We measured the device RC time constant[31] by
appears to saturate upon approaching the breakdown threshold.
loading with an external series resistor Rs, which allowed for
The doped semiconductor layer thus provides an all electronic
extraction of the device capacitance. We find a 3 dB roll-off fremeans of both frequency and amplitude tuning of the metaquency fc of 12 MHz (see Supporting Information for details).
material reflectance and here we realize an average change in
The improved modulation speed, compared to previous metaR(ω0) of 0.35 for all pixels at THz frequencies.
material designs,[31] is a result of having the Ohmic ground
In order to clarify the mechanism underlying operation of
plane directly underneath the Schottky layer which minimizes
the tunable absorber, we perform full wave 3D electromagnetic
2.8
3.0
3.2
3.2
3.4
3.6
Reflectance
Rmin (THz)
Adv. Optical Mater. 2013,
DOI: 10.1002/adom.201300265
GaAs
-3
Reflectance
M
3.0
M
Bias
M
Reflectance
2.8
b
Reflectance
2.6
Reflectance (ω0)
2.4
M
a
© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
3
www.advopticalmat.de
COMMUNICATION
www.MaterialsViews.com
Figure 3. Spatial light modulator performance. (a) Magnitude of the maximum change in reflectance | R (ω )| ≡ | R − 26.5V (ω) − R 0V (ω )| shown for
each pixel. (b) The corresponding frequency for each pixels respective |R max (ω)|. (c) The measured RC time constant by loading with an external
series resistor Rs, where the solid curves represent the fits used to calculate the device roll-off.
the device capacitance. Another added benefit of our Ohmic
ground plane configuration is that THz radiation is shielded
from the carrier wire routing thus allowing for higher filling
factors which further accommodates larger array sizes.
In conclusion, we have demonstrated an all-electronic multicolor spatial light modulator with super-pixels composed of
arrays of terahertz metamaterial absorbers. Our device is
capable of modulation of THz radiation at frequencies up to
12 MHz and an average modulation depth of 62% across the
pixel array. By utilizing a flip-chip bonded n-doped gallium arsenide epitaxial layer in the metamaterial absorber, we are able
to achieve high pixel density and minimal device capacitance
thus allowing for high speed modulation. Our design is scalable and may accommodate large active areas for much larger
pixel count as required for high resolution applications. This
work demonstrates a new path for construction of high speed
terahertz electronic devices and has implications in numerous
scientific and technological areas rich in applications, particularly in sensing, imaging, and dynamic scene projectors.
Experimental Section
The frequency dependent reflectance [R(ω)] was characterized at an
incident angle of 20° from 2 to 10 THz using a Hyperion-2000 infrared
microscope connected to a Fourier-transform infrared spectrometer,
Hg-arc lamp source, liquid helium-cooled Si bolometer detector, and a
germanium coated 6 μm mylar beamsplitter. The measured reflectance
spectra are normalized with respect to a gold mirror with square metal
apertures set to 450 μm on each side. Measurements were performed
with the THz electric field perpendicular to the metal connecting wires,
as depicted in Figure 1d. The DC potential between the Schottky metal
layer and the Ohmic ground plane is a reverse bias voltage (Vbias) as
shown in Figure 1b. The transmission was simulated and found to be
zero across the range characterized – not shown.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
D. Shrekenhamer and J. Montoya contributed equally to this work.
The research presented in this work performed at Boston College was
4
wileyonlinelibrary.com
supported through funding from the Office of Naval Research under US
Navy contracts No. N00014–07–1–0819 (semiconductor modeling) and
No. N00014–11–1–0864 (metamaterial spatial light modulator design
and fabrication). The University of New Mexico also acknowledges
additional funding from AFOSR through contract FA9550–10–1–0113.
Received: June 24, 2013
Revised: August 7, 2013
Published online:
[1] J. B. Pendry, Phys. Rev. Lett. 2000, 85, 3966–3969.
[2] D. R. Smith, W. J. Padilla, D. Vier, S. C. Nemat-Nasser, S. Schultz,
Phys. Rev. Lett. 2000, 84, 4184–4187.
[3] R. Shelby, D. Smith, S. Schultz, Science 2001, 292, 77–79.
[4] N. Fang, H. Lee, C. Sun, X. Zhang, Science 2005, 308, 534–537.
[5] D. Schurig, J. Mock, B. Justice, S. A. Cummer, J. Pendry, A. Starr,
D. Smith, Science 2006, 314, 977–980.
[6] U. Efron, Spatial Light Modulator Technology: Materials, Devices, and
Applications, Vol. 47, Marcel Dekker, New York 1995.
[7] R. A. DeVerse, R. R. Coifman, A. C. Coppi, W. G. Fateley,
F. Geshwind, R. M. Hammaker, S. Valenti, F. J. Warner, G. L. Davis,
Proc. SPIE 2003, 4959, 12–22.
[8] M. F. Duarte, M. A. Davenport, D. Takhar, J. N. Laska, T. Sun,
K. F. Kelly, R. G. Baraniuk, IEEE Signal Process. Mag. 2008, 25, 83–91.
[9] D. Dudley, W. Duncan, J. Slaughter, Proc. SPIE 2003, 4985, 14–25.
[10] K. M. Johnson, D. J. McKnight, I. Underwood, IEEE J. Quantum Electron. 1993, 29, 699–714.
[11] J. P. Rice, J. E. Neira, M. Kehoe, R. Swanson, Proc. SPIE 2009, 7210,
72100D.
[12] N. Landy, S. Sajuyigbe, J. Mock, D. Smith, W. Padilla, Phys. Rev. Lett.
2008, 100, 207402.
[13] C. M. Watts, X. Liu, W. J. Padilla, Adv. Mater. 2012, 24, 98–120.
[14] X. Liu, T. Tyler, T. Starr, A. F. Starr, N. M. Jokerst, W. J. Padilla, Phys.
Rev. Lett. 2011, 107, 45901.
[15] K. Aydin, V. E. Ferry, R. M. Briggs, H. A. Atwater, Nat. Commun.
2011, 2, 517.
[16] X. Liu, T. Starr, A. F. Starr, W. J. Padilla, Phys. Rev. Lett. 2010, 104,
207403.
[17] W. L. Chan, H. T. Chen, A. J. Taylor, I. Brener, M. J. Cich,
D. M. Mittleman, Appl. Phys. Lett. 2009, 94, 213511.
[18] B. Sensale-Rodriguez, R. Yan, M. M. Kelly, T. Fang, K. Tahy,
W. S. Hwang, D. Jena, L. Liu, H. G. Xing, Nat. Commun. 2012, 3, 780.
[19] B. Sensale-Rodriguez, S. Rafique, R. Yan, M. Zhu, V. Protasenko,
D. Jena, L. Liu, H. G. Xing, Opt. Express 2013, 21, 2324–2330.
[20] Y. Ra’di, V. S. Asadchy, S. A. Tretyakov, arXiv preprint arXiv:1211.3549.
2012.
© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Optical Mater. 2013,
DOI: 10.1002/adom.201300265
www.advopticalmat.de
www.MaterialsViews.com
Adv. Optical Mater. 2013,
DOI: 10.1002/adom.201300265
COMMUNICATION
[21] N. I. Zheludev, Y. S. Kivshar, Nat. Mater. 2012, 11, 917–924.
[22] B. Walther, C. Helgert, C. Rockstuhl, F. Setzpfandt, F. Eilenberger,
E. B. Kley, F. Lederer, A. Tünnermann, T. Pertsch, Adv. Mater. 2012,
24, 6300–6304.
[23] M. Rahm, J. Li, W. J. Padilla, J. Infrared Millim. Terahz. Waves 2012,
34, 1–27.
[24] S. H. Lee, M. Choi, T. Kim, S. Lee, M. Liu, X. Yin, H. K.
Choi, S. S. Lee, C. Choi, S. Choi, Nat. Mater. 2012, 11, 936–
941.
[25] R. Kersting, G. Strasser, K. Unterrainer, Electron. Lett. 2000, 36,
1156–1158.
[26] T. R. Tsai, C. Y. Chen, R. P. Pan, C. L. Pan, X. C. Zhang, IEEE Microwave Wireless Comp. Lett. 2004, 14, 77–79.
[27] M. Tonouchi, Nat. Photonics 2007, 1, 97–105.
[28] Y. Urzhumov, J. S. Lee, T. Tyler, S. Dhar, V. Nguyen, N. M. Jokerst,
P. Schmalenberg, D. R. Smith, Phys. Rev. B 2012, 86, 075112.
[29] T. I. Jeon, D. Grischkowsky, Appl. Phys. Lett. 1998, 72, 3032–3034.
[30] P. Huggard, J. Cluff, G. Moore, C. Shaw, S. Andrews, S. Keiding,
E. Linfield, D. Ritchie, J. Appl. Phys. 2000, 87, 2382–2385.
[31] H. T. Chen, S. Palit, T. Tyler, C. M. Bingham, J. M. O. Zide,
J. F. O'Hara, D. R. Smith, A. C. Gossard, R. D. Averitt, W. J. Padilla,
Appl. Phys. Lett. 2008, 93, 091117.
© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
5