Frequency-dependent transmission-type

digital coding metasurface controlled by

light intensity
Cite as: Appl. Phys. Lett. 113, 091601 (2018); https://doi.org/10.1063/1.5045718
Submitted: 23 June 2018 . Accepted: 07 August 2018 . Published Online: 28 August 2018

Xin Ge Zhang, Wei Xiang Jiang, and Tie Jun Cui

ARTICLES YOU MAY BE INTERESTED IN

Design of digital coding metasurfaces with independent controls of phase and amplitude
responses
Applied Physics Letters 113, 063502 (2018); https://doi.org/10.1063/1.5043520

Resonant metasurface with tunable asymmetric reflection

Applied Physics Letters 113, 094103 (2018); https://doi.org/10.1063/1.5046948

Reconfigurable conversions of reflection, transmission, and polarization states using active

metasurface
Applied Physics Letters 110, 121901 (2017); https://doi.org/10.1063/1.4979033

Appl. Phys. Lett. 113, 091601 (2018); https://doi.org/10.1063/1.5045718 113, 091601

© 2018 Author(s).
 APPLIED PHYSICS LETTERS 113, 091601 (2018)

Frequency-dependent transmission-type digital coding metasurface

controlled by light intensity
Xin Ge Zhang, Wei Xiang Jiang,a) and Tie Jun Cuib)
State Key Laboratory of Millimeter Waves, School of Information Science and Engineering,
Southeast University, Nanjing 210096, China
(Received 23 June 2018; accepted 7 August 2018; published online 28 August 2018)
A great challenge with metasurfaces is tunability for tailoring electromagnetic waves dynamically.
Metasurfaces designed digitally, also known as digital coding metasurfaces, provide a low-overhead
strategy for constructing controllable or even programmable metasurfaces. However, most reported
digital metasurfaces have been single-band ones of the reflection type that have been realized by
controlling the phase response of the digital elements. This study reports on a light-controllable and
frequency-dependent digital coding metasurface that allows wave transmission to be manipulated
more freely and flexibly. By remotely tuning the illumination intensity or shifting the frequency of
the incident waves, the transmission response of the designed digital elements can be tuned dynami-
cally. The proposed device is assessed experimentally at microwave frequencies. The presented
transmissive digital metasurface offers unprecedented opportunities to produce reconfigurable devi-
ces that are controlled in multiple ways with a single design. Published by AIP Publishing.
https://doi.org/10.1063/1.5045718

An electromagnetic (EM) metasurface is an artificial type and whose encoding was based on the phase response
structure that comprises subwavelength elements arranged of the EM wave. However, as well as the phase of an EM
periodically or nonperiodically on a two-dimensional sur- wave, its amplitude is also an important property; many use-
face. Because of the unusual EM response at its surface,1–4 a ful structures and devices can be realized by manipulating
metasurface can exert powerful control over the phase and the amplitude, such as frequency-selective surfaces,35–38
amplitude of the EM waves. Also, because a metasurface absorbers,39–41 radomes,42,43 and low-sidelobe antennas.44,45
is extremely thin, its inherent loss is low and it avoids the Herein, we present a tunable digital coding metasurface
fabrication and assembly problems associated with three- for dynamic control of the transmission state of EM waves
dimensional metamaterials. Because of the aforementioned by tuning the external illumination. By remotely tuning the
properties, metasurfaces have attracted much attention in illumination intensity or shifting the frequency of the inci-
recent eight years.5–9 Some functional devices and systems dent wave, the digital metasurface can achieve free switches
have been realized based on metasurfaces, such as flat of the transmission status of the EM waves. We design a dig-
lenses,10–13 focusing and scattering screens,14,15 polar- ital element and verify it with full-wave simulation, and we
izers,16–18 imaging systems,19–21 and other exciting applica- also fabricate and measure a digital metasurface comprising
tions.22–28 However, most previously reported metasurfaces 15  15 designed elements.
cannot be controlled or tuned freely; their functions are fixed To realize the aforementioned transmissive digital coding
once they have been fabricated. Therefore, a challenge metasurface that can be tuned by illumination or the frequency
remains to realize multi-functional or reconfigurable EM of the incident wave, we propose and design the digital ele-
devices. ment shown in Fig. 1(a). A circular-ring gap with inner radius
Digital codings “0” and “1” have been adopted innova- r is excavated on the top copper layer, and a varactor diode is
tively to denote the corresponding unit cells with different loaded inside the gap to connect the inner copper patch and
phase responses of EM waves, and these have been used to outer copper layer. In the center of the element, a metallic via
design digital coding and programmable metasurfaces.29 An connects the inner circular copper patch to the bottom copper
external signal can be used to control the states of the struc- ground line, which is arranged in the x direction. To control
tural elements on a digital metasurface. Therefore, the func- the wave transmission of the element dynamically with exter-
tion of a digital metasurface can be altered in real time by nal illumination, we use a silicon PIN photodiode series array
changing its control signal dynamically.30–33 With some to provide a direct-current (DC) reverse voltage for tuning the
complicated systems, there is a need for remote-control non- capacitance (CT) of the embedded varactor in real time, as
contact tuning, and this has led to another class of tunable shown in Fig. 1(a).
metasurfaces being proposed, namely, the frequency coding To investigate the transmission characteristics of the
metasurface.34 By changing the EM-wave frequency, the light-controllable element and further to realize its expected
phase responses of the coding elements can be tuned freely. transmission performance, we use the CST Microwave Studio
However, most digital metasurfaces that have been dem- software package to simulate and optimize the element. In the
onstrated to date have been single-band ones of the reflective simulations, the “unit cell” boundary condition is used. We
select the dielectric material F4B with a dielectric constant of
a)
E-mail: wxjiang81@seu.edu.cn 2.65 and loss tangent of 0.001 as the substrate of the element,
b)
E-mail: tjcui@seu.edu.cn the thickness h of which is 2.0 mm. The sidelength a of the

0003-6951/2018/113(9)/091601/5/$30.00 113, 091601-1 Published by AIP Publishing.

091601-2 Zhang, Jiang, and Cui Appl. Phys. Lett. 113, 091601 (2018)

In the numerical simulations, the transmission peaks

appear at the resonance frequencies of the designed digital ele-
ment,47–49 and some details are analyzed in the supplementary
material. Next, we investigate the resonance performance of
the element. With CT ¼ 2.67 pF, resonance modes (i) and (ii)
are excited near the ring gap, as shown in Figs. 2(a) and 2(b).
The surface currents on the sides of the ring gap are in opposite
directions, indicating that strong magnetic resonance occurs at
each resonance frequency. Meanwhile, the paths of the surface
currents differ considerably at the resonance frequencies of
FIG. 1. (a) Proposed transmissive digital element that is controlled by light
2.33 and 5.72 GHz. For the fundamental resonance mode (i),
illumination. (b) Simulated amplitudes of transmission coefficient of two virtual current loops are formed along the ring gap,50,51 as
designed digital element for CT ¼ 2.67 pF and 0.63 pF. The two orange areas denoted by the blue curves in Fig. 2(a). For the high-order res-
represent two working frequency bands in which the state of the digital ele- onance mode (ii) that differs from the fundamental resonance
ment can be switched dynamically between “0” and “1.”
mode (i), four virtual current loops are induced on the struc-
ture, as denoted by the blue curves in Fig. 2(b). The distribu-
element is 18.0 mm, which is around 1/5 of the wavelength at tions of surface current in resonance modes (i) and (ii) for
a working frequency of 3.12 GHz. The diameter of the metallic CT ¼ 0.63 pF are shown in Figs. 2(c) and 2(d). The resonance
via is 0.3 mm, the width w of the ground line is 0.4 mm, and phenomena for CT ¼ 0.63 pF is clearly similar to that for
the thickness of the copper patches and ground line is CT ¼ 2.67 pF. Two virtual current loops and four virtual cur-
0.018 mm. Because there is no existing model of a varactor rent loops also form for the fundamental resonance mode (i)
diode in CST, in the simulations, we use an RLC series circuit and high-order resonance mode (ii), respectively. By changing
comprising a loss resistance Rs, a parasitic inductance Ls, and CT, the two resonance frequencies shift to 3.12 and 6.64 GHz.
a variable capacitance CT to simulate the varactor diode. From We select the Skyworks SMV1405-079LF varactor
the handbook of varactor diodes,46 Rs ¼ 0.8 X, Ls ¼ 0.7 nH, diode to satisfy the dynamic capacitance needed by the digi-
and the adjustable value of CT depends on the reverse DC bias. tal element. The capacitance of this varactor can be tuned as
To obtain good transmission characteristics at two different 0.63–2.67 pF, and its maximum reverse bias voltage is 30 V;
bands, we perform many numerical simulations with different we use the BPW-34-S photodiode to provide the required
values of g, r, and CT. Finally, by observing the transmission bias voltage. In general, each photodiode can generate a volt-
amplitude of the element, we arrive at g ¼ 0.4 mm, r ¼ 8.4 mm, age of 0–0.5 V when the light intensity changes. Here, we
and CT ¼ 2.67 or 0.63 pF. To reduce the effect of the ground connect 70 photodiodes in series to provide sufficient voltage
line, the incident wave is y-polarized and the varactor diode is under normal light intensity.
loaded at y end of the ring gap, as shown in Fig. 1(a). The To verify the proposed digital metasurface experimen-
simulated amplitudes of the transmission coefficient S21 of the tally, we fabricated and measured one sample. The fabricated
designed element with frequencies of 1.5–7.5 GHz are shown
in Fig. 1(b). The element clearly has two resonance modes (i)
and (ii). With CT ¼ 2.67 pF, the two resonance frequencies are
f1 ¼ 2.33 GHz and f3 ¼ 5.72 GHz; with CT ¼ 0.63 pF, the reso-
nance frequencies shift to f2 ¼ 3.12 GHz and f4 ¼ 6.64 GHz. At
resonance frequency f2, the amplitude of the transmission coef-
ficient is 18.72 dB with CT ¼ 2.67 pF and 0.176 dB with
CT ¼ 0.63 pF. At resonance frequency f3, the amplitude of the
transmission coefficient is 0.46 dB with CT ¼ 2.67 pF and
23.95 dB with CT ¼ 0.63 pF.
Here, we encode an element whose transmission-
coefficient amplitude is below 13 dB as a “0” element and
one whose transmission-coefficient amplitude exceeds
1 dB as a “1” element. Therefore, at 3.12 GHz, the digital
element acts as a “0” element with CT ¼ 2.67 pF and as a “1”
element with CT ¼ 0.63 pF. By contrast, at 5.72 GHz, the dig-
ital element acts as a “1” element with CT ¼ 2.67 pF and as a
“0” element with CT ¼ 0.63 pF. Therefore, for CT ¼ 2.67 pF
or CT ¼ 0.63 pF, the digital element clearly has different dig-
ital states at the two resonance frequencies of 3.12 and
5.72 GHz; that is, the digital states also depend on the work- FIG. 2. Distributions of induced surface currents at resonance frequencies
ing frequency bands. According to the two aforementioned (a) 2.33 GHz and (b) 5.72 GHz with CT ¼ 2.67 pF and (c) 3.12 GHz and (d)
6.64 GHz with CT ¼ 0.63 pF. The blue loops numbered 1–4 represent the vir-
coding states, we note that the element has two working fre-
tual current loops induced in different resonance modes, and the blue arrows
quency bands of 2.91–3.32 GHz and 5.51–5.94 GHz instead indicate the flow directions of the currents on both edges of the metallic
of two resonance frequencies of 3.12 and 5.72 GHz. patches inside and outside the ring gap.
091601-3 Zhang, Jiang, and Cui Appl. Phys. Lett. 113, 091601 (2018)

metasurface comprises (i) a 15  15 array of digital elements diode)-based fluorescent tube is roughly 80 mW. With the light
and (ii) 70 photodiodes connected in series and located by source turned off, the light intensity is zero and the generated
the side of the metasurface, as shown in Fig. 3(a). The out- DC bias is zero. With the light source turned on, the illumina-
side copper layers of all the elements of the digital metasur- tion intensity is 6000 lx and the generated DC bias is 30 V.
face are connected together and then connected to the For light intensities EV ¼ 0 lx and 6000 lx, the measured
positive electrode of the photodiode array. The metasurface amplitudes of the transmission coefficient of the metasurface
is divided into 5  5 super unit cells (SUCs), each of which for y-polarized incident waves are shown in Fig. 3(d). With
comprises 3  3 elements. For each SUC, the feeding lines zero light intensity, there are two resonance frequencies of
are connected together, meaning that all elements in one 2.51 and 5.96 GHz; with a light intensity of 6000 lx, the two
SUC are controlled together. In the ideal case, each SUC is resonance frequencies are 3.28 and 6.71 GHz. The two work-
connected to one series of photodiodes and thus can be con- ing frequency bands shift slightly to higher frequencies and
trolled independently. Here, to simplify the fabrication, we change to the bands of 3.10–3.56 GHz and 5.84–6.24 GHz.
divide the metasurface into five columns of five SUCs each The measured results differ slightly from the simulated ones,
and then connect the ground lines of the columns together. and the reasons for this are analyzed in the supplementary
In the first experiment, the ground lines of all the columns material. Note that although there is a small difference
are connected to the negative electrode of the photodiode between the tested transmission performance and the simula-
array by a bias line. The fabricated sample covers a total tions, the measured dual resonance characteristics of the dig-
area of 300  280 mm2. ital metasurface agree very well with the simulation ones,
We use the free-space method to measure the transmission which validates the feasibility of light-controllable and
coefficients of the fabricated digital metasurface,52 and the frequency-dependent coding metasurfaces.
experimental environment is shown schematically in Fig. 3(b). Thus far, we have studied the tunability of the transmis-
In the experiment, large-area wave-absorbing materials are sion state of a tunable digital coding metasurface in the case
placed around the fabricated sample to avoid wave diffraction. of normal incidence. Next, to further verify the excellent tun-
The fabricated sample is placed on a supporting board, and the ing performance of the digital metasurface, we also investi-
distance from the sample to the transmitting antenna and that gate the case of EM-wave oblique incidence. The simulation
from the sample to the receiving antenna are both 1.5 m. We results are shown in Figs. 4(a) and 4(b) for incidence angles
first test the voltage generated by the 70 photodiodes when the of 30 and 60 , respectively. From the transmission curves,
illumination intensity (EV) changes; the measured results are the metasurface also works well with waves of larger incident
shown in Fig. 3(c). When the intensity of the light projected angle. The measured transmission curves agree very well
onto the photodiodes is zero and 6000 lx, the required voltages with the simulated ones, as shown in Figs. 4(c) and 4(d).
of zero and 30 V, respectively, are achieved. To provide light However, as the angle of oblique incidence increases, the
of the required intensity, we use a white-light fluorescent tube transmissivity at the resonance frequencies decreases, thereby
as the experimental light source and fix it above the photodiode narrowing the working bandwidth of the metasurface.
array, as shown in the inset of Fig. 3(b). In the measurements, To better understand the working mechanism of the digi-
the illumination beam power of the LED (light-emitting tal metasurface when controlling the transmission amplitude,
we plot the simulated surface-induced current distributions
for different coding patterns in Fig. 5. The whole metasurface
comprises 5  5 SUCs, each of which has the same coding

FIG. 3. (a) Photograph of the fabricated sample. (b) Schematic of experimental

setup, including the sample, light source, and their position relationship. Here,
the inset photograph shows part of the real environment. (c) Measured relation-
ship between voltage generated by silicon photodiode array and illumination FIG. 4. Transmission amplitude of designed digital element with oblique
intensity EV. (d) Measured amplitudes of transmission coefficients of fabri- incidence. The incidence angles of the EM waves are 30 in the left column
cated digital metasurface for EV ¼ 0 and EV ¼ 6000 lx. The two gray areas and 60 in right column. (a), (b) Simulated transmission curves. (c), (d)
denote the two tested working frequency bands. Measured transmission curves.
091601-4 Zhang, Jiang, and Cui Appl. Phys. Lett. 113, 091601 (2018)

In conclusion, we have proposed and verified a light-

tunable coding metasurface that offers great flexibility for
controlling EM wave transmission in real time. The digital
state of the proposed digital element can be controlled
dynamically by remote illumination. As an example, a light-
controllable and frequency-dependent digital metasurface
was fabricated and measured. The experimental results agree
well with the simulation ones. Unlike previous work in which
the phase of the EM wave was quantized by means of digital
“0”s and “1”s, here the EM-wave amplitude was quantized
by “0”s and “1”s. The proposed digital coding metasurface
could be used to realize various high-performance configura-
ble devices and to achieve low-sidelobe antennas and shaped
antennas.

See supplementary material for details about (i) the

surface-induced current distributions of the element, (ii) the
analysis of measurement errors, (iii) the methods used to
measure the transmission coefficients, and (iv) the theoretical
basis for near field measurement.

This work was supported in part by the National Key

Research and Development Program of China (Grant
Nos. 2017YFA0700201, 2017YFA0700202, and
2017YFA0700203), the National Natural Science
Foundation of China (Grant Nos. 61522106, 61571117,
FIG. 5. Simulated induced surface-current distributions of the digital coding 61501112, 61401089, 61501117, and 61631007), the
metasurface with different coding patterns: (a) 01010; (b) 10101; (c) 01010/ Foundation of National Excellent Doctoral Dissertations of
01010; and (d) 10101/10101. The working frequency in (a) and (c) is China (Grant No. 201444), and the 111 Project (111-2-05).
3.12 GHz, and that in (b) and (d) is 5.72 GHz. (e), (f) Measured intensity dis-
tributions of electric field on the xy-plane of digital coding metasurface with 1
coding patterns (e) 01010 and (f) 10101. N. F. Yu, P. Genevet, M. A. Kats, F. Aieta, J. P. Tetienne, F. Capasso, and
Z. Gaburro, Science 334, 333 (2011).
2
N. F. Yu and F. Capasso, Nat. Mater. 13, 139 (2014).
3
state, either “0” or “1.” As shown in Figs. 5(a) and 5(c), the L. X. Liu, X. Q. Zhang, M. Kenney, X. Q. Su, N. N. Su, C. M. Ouyang, Y.
surface current distributions of the metasurfaces at 3.12 GHz L. Shi, J. G. Han, W. L. Zhang, and S. Zhang, Adv. Mater. 26, 5031
(2014).
differ distinctly for the coding patterns of 01010 and 01010/ 4
S. L. Jia, X. Wan, P. Su, Y. J. Zhao, and T. J. Cui, AIP Adv. 6, 045024
01010. The “1” element has strong resonance whereas the (2016).
5
“0” element has almost no resonance, meaning that the digital H. T. Chen, A. J. Taylor, and N. F. Yu, Rep. Prog. Phys. 79, 076401
metasurface can control the transmission amplitude by simply (2016).
6
C. L. Holloway, E. F. Kuester, J. A. Gordon, J. O’Hara, J. Booth, and D.
changing the spatial coding patterns. When the frequency R. Smith, IEEE Antennas Propag. Mag. 54, 10 (2012).
shifts to 5.72 GHz, the transmission states of the binary ele- 7
H. H. Hsiao, C. H. Chu, and D. P. Tsai, Small Methods 1, 1600064 (2017).
8
ments are all switched to opposite ones; that is, the digital F. Ding, A. Pors, and S. I. Bozhevolnyi, Rep. Prog. Phys. 81, 026401
metasurface has excellent frequency tunability of its coding (2018).
9
T. J. Cui, Natl. Sci. Rev. 5, 134 (2018).
patterns, as shown in Figs. 5(b) and 5(d). 10
X. Chen, L. Huang, H. M€ uhlenbernd, G. Li, B. Bai, Q. Tan, G. Jin, C. W.
To validate experimentally the performance shown in Qiu, S. Zhang, and T. Zentgraf, Nat. Commun. 3, 1198 (2012).
11
Figs. 5(a)–5(d), we fabricated another sample. In this experi- F. Aieta, Nano Lett. 12, 4932 (2012).
12
M. Khorasaninejad, F. Aieta, P. Kanhaiya, M. A. Kats, P. Genevet, D.
ment, we connect the ground lines of only the second and Rousso, and F. Capasso, Nano Lett. 15, 5358 (2015).
fourth columns of SUCs to the photodiode series. Thus, with 13
K. Chen, Y. J. Feng, F. Monticone, J. M. Zhao, B. Zhu, T. Jiang, L.
a light intensity of 6000 lx, the sample can generate the cod- Zhang, Y. Kim, X. M. Ding, S. Zhang, A. Al u, and C. W. Qiu, Adv.
ing patterns 01010 at 3.28 GHz and 10101 at 5.96 GHz. We Mater. 29, 1606422 (2017).
14
T. Cai, G. M. Wang, S. W. Tang, H. X. Xu, J. W. Duan, H. J. Guo, F. X.
measure the near field distributions of electric fields of the Guan, S. L. Sun, Q. He, and L. Zhou, Phys. Rev. Appl. 8, 034033 (2017).
metasurface with the coding pattern configurations shown in 15
L. H. Gao, Q. Cheng, J. Yang, S. J. Ma, J. Zhao, S. Liu, H. B. Chen, Q.
Figs. 5(a) and 5(b), and some explanations are given in the He, W. X. Jiang, H. F. Ma, Q. Y. Wen, L. J. Liang, B. B. Jin, W. W. Liu,
supplementary material. The scanning step is 5 mm, around L. Zhou, J. Q. Yao, P. H. Wu, and T. J. Cui, Light: Sci. Appl. 4, e324
(2015).
1/10 of the wavelength at 5.96 GHz. As shown in Fig. 5(e), 16
B. Ratni, A. de Lustrac, G. P. Piau, and S. N. Burokur, Appl. Phys. Lett.
the measured intensity distribution of the electric field on the 111, 214101 (2017).
17
xy-plane at 3.28 GHz indeed satisfies the coding pattern K. Grady, J. E. Heyes, D. R. Chowdhury, Y. Zeng, M. T. Reiten, A. K.
Azad, A. J. Taylor, D. A. R. Dalvit, and H. T. Chen, Science 340, 1304
01010. At 5.96 GHz, the metasurface exhibits the coding pat-
(2013).
tern 10101, and the measured intensity distribution of the 18
Z. Tao, X. Wan, B. C. Pan, and T. J. Cui, Appl. Phys. Lett. 110, 121901
electric field is shown in Fig. 5(f). (2017).
091601-5 Zhang, Jiang, and Cui Appl. Phys. Lett. 113, 091601 (2018)

19 35
L. Huang, X. Chen, H. M€ uhlenbernd, H. Zhang, S. Chen, B. Bai, Q. Tan, K. Sarabandi and N. Behdad, IEEE Trans. Antennas Propag. 55, 1239
G. Jin, K. W. Cheah, C. W. Qiu, J. Li, T. Zentgraf, and S. Zhang, Nat. (2007).
36
Commun. 4, 2808 (2013). C. Debus and P. H. Bolivar, Appl. Phys. Lett. 91, 184102 (2007).
20 37
X. Ni, A. V. Kildishev, and V. M. Shalaev, Nat. Commun. 4, 2807 (2013). M. N. Hussein, J. F. Zhou, Y. Huang, M. Kod, and A. P. Sohrab, IEEE
21
J. Lin, P. Genevent, M. A. Kats, N. Antoniou, and F. Capasso, Nano Lett. Trans. Microwave Theory Tech. 65, 4851 (2017).
38
13, 4269 (2013). S. Ghosh and K. V. Srivastava, IEEE Trans. Electromagn. Compat. 60,
22
C. Huang, C. L. Zhang, J. N. Yang, B. Sun, B. Zhao, and X. G. Luo, Adv. 166 (2018).
39
Opt. Mater. 5, 1700485 (2017). N. I. Landy, S. Sajuyigbe, J. J. Mock, D. R. Smith, and W. J. Padilla, Phys.
23
L. Q. Cong, P. Pitchappa, Y. Wu, L. Ke, C. Lee, N. Singh, H. Yang, and Rev. Lett. 100, 207402 (2008).
40
R. Singh, Adv. Opt. Mater. 5, 1600716 (2017). A. B. Li, S. Kim, Y. Luo, Y. B. Li, J. Long, and D. F. Sievenpiper, IEEE
24
D. R. Smith, O. Yurduseven, L. P. Mancera, P. Bowen, and N. B. Kundtz, Trans. Microwave Theory Tech. 65, 2810 (2017).
41
Phys. Rev. Appl. 8, 054048 (2017). G. M. Akselrod, J. N. Huang, T. B. Hoang, P. T. Bowen, L. Su, D. R.
25
L. Q. Cong, P. Pitchappa, C. Lee, and R. Singh, Adv. Mater. 29, 1700733 Smith, and M. H. Mikkelsen, Adv. Mater. 27, 8028 (2015).
42
(2017). F. Costa and A. Monorchio, IEEE Trans. Antennas Propag. 60, 2740 (2012).
26 43
Y. J. Zheng, X. Y. Cao, J. Gao, H. H. Yang, Y. L. Zhou, and T. Liu, Opt. P. Mei, X. Q. Lin, J. W. Yu, P. C. Zhang, and A. Boukarkar, IEEE Trans.
Express 25, 30001 (2017). Antennas Propag. 66, 409 (2018).
27 44
S. Liu, Q. Cheng, Q. Xu, T. Q. Wang, L. L. Du, K. Luan, Y. H. Xu, D. P. Kumar, A. Kedar, and A. K. Singh, IEEE Trans. Antennas Propag. 63,
Bao, X. J. Fu, J. G. Han, W. L. Zhang, and T. J. Cui, Adv. Opt. Mater. 4, 4723 (2015).
45
384 (2016). W. P. M. N. Keizer, IEEE Trans. Antennas Propag. 55, 715 (2007).
28 46
W. X. Jiang, C. Y. Luo, S. Ge, C. W. Qiu, and T. J. Cui, Adv. Mater. 27, See http://www.skyworksinc.com/uploads/documents/SMV1405_1430_
4628 (2015). Series_200068V.pdf for the data sheet of SMV1405 Plastic Packaged
29
T. J. Cui, M. Q. Qi, X. Wan, J. Zhao, and Q. Cheng, Light: Sci. Appl. 3, Abrupt Junction Tuning Varactor.
47
e218 (2014). B. A. Munk, Frequency Selective Surfaces: Theory and Design (Wiley,
30
S. Liu and T. J. Cui, Adv. Opt. Mater. 5, 1700624 (2017). New York, 2000).
31 48
L. L. Li, T. J. Cui, W. Li, S. Liu, J. Ding, X. Wan, Y. B. Li, M. H. Jiang, Z. Bai, Q. Zhang, Y. Ju, G. Tao, X. Jiang, N. Kang, C. Liu, and L. Zhang,
C. W. Qiu, and S. Zhang, Nat. Commun. 8, 197 (2017). J. Phys. D: Appl. Phys. 49, 065002 (2016).
32 49
H. H. Yang, F. Yang, S. H. Xu, Y. L. Mao, M. K. Li, X. Y. Cao, and J. N. Xu, J. Gao, J. Zhao, and X. Feng, AIP Adv. 5, 077157 (2015).
50
Gao, IEEE Trans. Antennas Propag. 64, 2246 (2016). C.-Y. Chen, S.-C. Wu, and T.-J. Yen, Appl. Phys. Lett. 93, 034110 (2008).
33 51
P. Pitchappa, C. P. Ho, L. Q. Cong, R. Singh, N. Singh, and C. Lee, Adv. F. J. Rodrıguez-Fortu~no, R. Ortu~no, C. Garcıa-Meca, J. Martı, and A.
Opt. Mater. 4, 391 (2016). Martınez, J. Appl. Phys. 112, 103104 (2012).
34 52
H. T. Wu, S. Liu, X. Wan, L. Zhang, D. Wang, L. L. Li, and T. J. Cui, D. K. Ghodgaonkar, V. V. Varadan, and V. K. Varadan, IEEE Trans.
Adv. Sci. 4, 1700098 (2017). Instrum. Meas. 38, 789 (1989).