A birefringent reconfigurable metasurface at visible wavelengths was obtained by combining an Au nanograting metasurface, which shows very high birefringence characteristics in visible light, and a microelectromechanical system actuator. The system was based on the electrostatic out-of-plane motion of the nanograting and it was developed by microfabrication. The modulation of retardation was achieved at a wavelength of 633 nm by up to 25.3° (from 21.5° to 46.8°) by applying a voltage in the range of 0–200 V.

Metasurfaces are planar metamaterials that have attracted a significant amount of attention because by periodically arranging fine patterns on the surface.1–3 They can exhibit unique optical characteristics, which cannot be seen in natural materials. Metasurfaces have high affinity for microfabrication technologies including lithography, film deposition and etching. Due to recent advances in their fabrication method, metasurfaces for visible light applications can now be produced. In addition, research on novel display applications including holography4,5 has been actively conducted. Since the characteristics of the metasurfaces depend on their structure, after their formation, the optical properties are fixed. Therefore, to perform dynamic modulation, some modifications are required.

Most of the methods for the dynamic control of the characteristics of metasurfaces are based on changing the refractive indices of the media surrounding the metasurfaces by applying external stimuli. Various media have been used including photoactive polymers,6 phase-transition materials,7,8 and liquid crystals.9 In these years, as an approach alternative to these methods, several studies on the mechanical deformation of metasurfaces have been conducted: they combined the metasurfaces with microelectromechanical system (MEMS) actuators to achieve their mechanical modulation.10–20 The advantage of mechanical modulation is its potential wide dynamic range because the near-field optical interaction between meta-atoms is exponentially dependent on the inter-atomic distance. For examples of mechanical reconfigurable metasurfaces, transmittance modulators,11–15 polarization modulators,16 color filters17–19 and active lenses20 have been reported in the visible range.

This study aims to demonstrate birefringence modulation in the visible region; hence, an Au nanograting, which is a metasurface showing very high birefringence characteristics in the visible light,21 was combined with a MEMS actuator. Au nanograting has been studied, with possible applications such as micro-optical retarders including a half-wave plate, a quarter-wave plate, and a radial polarizer.22,23 This structural birefringence can be controlled by dimensional parameters such as the thickness of the Au nanograting.

A schematic of the birefringent reconfigurable metasurface (BRM) device realized for this study is shown in Fig. 1. The device was fabricated on a glass substrate and the movable beams and fixed bars were designed by selectively etching the sacrificial Si layer under the grating area. The bars were fixed on the glass substrate, while the movable beams were connected to suspended Au membranes, which were moved by the electrostatic force generated by the above transparent electrode [indium tin oxide (ITO)] [Fig. 1(b)]. As the movable beams were deformed toward the electrode, the geometry of the grating cross-section was changed, so that the birefringence property of the Au nanograting could be modulated.

FIG. 1.

(a) Overview of the BRM device. The movable beams and the fixed bars were alternately arranged by selectively etching the underlying Si sacrificial layer. (b) A transparent electrode was placed on top of the device to pull up via electrostatic forces by applying voltage, changing the geometry of the grating cross-section and, consequently, modulating the characteristics of the metasurface.

FIG. 1.

(a) Overview of the BRM device. The movable beams and the fixed bars were alternately arranged by selectively etching the underlying Si sacrificial layer. (b) A transparent electrode was placed on top of the device to pull up via electrostatic forces by applying voltage, changing the geometry of the grating cross-section and, consequently, modulating the characteristics of the metasurface.

Close modal

The retardation Δ due to the birefringence of a grating slit with width w and thickness t can be described as23,24

Δ(w,t)=2πtλ0ΔN(w)=2πtλ0|Re[ky(w)]k0Re[kx(w)]k0|,
(1)

where λ0 is the wavelength of the incident light, kx and ky are the wave-numbers in the slow-axis direction [transverse magnetic (TM) mode: electric field perpendicular to the grating] and in the fast-axis direction [transverse electric (TE) mode: electric field parallel to the grating] of the birefringent material, respectively, and k0 is the wave-number in a vacuum. kx and ky are functions of w, and as w decreases, the difference between these values increases.23 In the proposed BRM device, the birefringence can be modulated by changing the effective thickness of the Au layer by driving beams. As shown in Fig. 1(b), the cross-section of the BRM device in the ON state can be approximated to be a three-layer grating stack having slit widths of w1 and w2. The total retardation of the stack Δtotal can roughly be estimated as

Δtotal=Δ1(w1,td)+2Δ2(w2,d),
(2)

where d is the displacement of the movable beams. The first term indicates bar-beam interaction and the second term indicates inter-bar or inter-beam interactions, respectively. The device was designed to have the grating period p =800 nm, slit widths w1 = 500 nm and w2 = 1300 nm, the grating length lg = 50 μm, the thickness of Au t =300 nm, and the whole device length l =500 μm. Based on the design parameters and Eqs. (1) and (2), the initial retardation was calculated to be 45°. The BRM device was fabricated by a collective process of surface micromachining as shown in Fig. 2.

FIG. 2.

Fabrication process of the BRM device by electron beam lithography (EBL) and plasma etching. By sputtering, Si as an etching mask, Au and Cr as adhesion layers, and Si as a sacrificial layer, with thicknesses of 150, 300, 5, and 500 nm, respectively, were deposited on a quartz-glass substrate. A positive electron beam (EB) resist (gL2000, Gluon Lab.) was patterned by EBL (JBX-6300FS, JEOL) (a) and the top Si layer was also patterned by reactive ion etching (RIE) with SF6 gas (b). Then, the Au layer was etched by ion milling with Ar plasma, followed by etching of the lower Si layer by SF6 RIE (c). A second EBL was performed with an EB resist to distinguish fixed bars and movable beams (d). The EB resist was diluted to facilitate its penetration into the slits of grating. The exposed parts of Si serving as movable beams were etched with XeF2 gas (e). Finally, the resist was removed with O2 plasma (f).

FIG. 2.

Fabrication process of the BRM device by electron beam lithography (EBL) and plasma etching. By sputtering, Si as an etching mask, Au and Cr as adhesion layers, and Si as a sacrificial layer, with thicknesses of 150, 300, 5, and 500 nm, respectively, were deposited on a quartz-glass substrate. A positive electron beam (EB) resist (gL2000, Gluon Lab.) was patterned by EBL (JBX-6300FS, JEOL) (a) and the top Si layer was also patterned by reactive ion etching (RIE) with SF6 gas (b). Then, the Au layer was etched by ion milling with Ar plasma, followed by etching of the lower Si layer by SF6 RIE (c). A second EBL was performed with an EB resist to distinguish fixed bars and movable beams (d). The EB resist was diluted to facilitate its penetration into the slits of grating. The exposed parts of Si serving as movable beams were etched with XeF2 gas (e). Finally, the resist was removed with O2 plasma (f).

Close modal

The images of the fabricated BRM device are shown in Fig. 3. As designed, the movable beams connected to the suspended Au membrane and fixed bars isolated from the pattern were alternately arranged. Figure 3(d) shows the profile of the A–A′ cross-section of the grating part shown in Fig. 3(b) as measured using a laser microscope (VK-X 250, KEYENCE) after its release. It shows that the suspended Au nanograting has deformed into an upward convex shape along the direction perpendicular to the slit [Fig. 3(d)]. This could to be due to the residual stress of the deposited films. This stress-induced deformation caused an initial displacement of about 600 nm at the center of the element.

FIG. 3.

(a) SEM image of the entire BRM device. (b) and (c) Laser microscopy images of the nanograting of the BRM device (b) and of the root of its beams (c): the modulation was measured by focusing on the part indicated by the white circle in (b). (d) A two-dimensional profile across the device [red line in (b)] measured with a laser microscope.

FIG. 3.

(a) SEM image of the entire BRM device. (b) and (c) Laser microscopy images of the nanograting of the BRM device (b) and of the root of its beams (c): the modulation was measured by focusing on the part indicated by the white circle in (b). (d) A two-dimensional profile across the device [red line in (b)] measured with a laser microscope.

Close modal

To apply electrostatic forces to the BRM device in the vertical direction, a transparent indium tin oxide (ITO) glass electrode was placed above it. An insulating polyimide tape with a thickness of 50 μm was attached to the sample substrate to create a gap between the grating and the transparent electrode. A wiring was connected to the ITO electrode and the Au layer of the device, and a voltage was applied to elevate the self-supporting Au membrane toward the electrode. A SourceMeter (2410, KEITHLEY) was used to apply and monitor this driving voltage and perform retardation measurements. Since the retardation of the Au nanograting occurs mainly at the slit between adjacent Au beams, a larger modulation amount should be obtained by measuring at a position as small initial displacement as possible. Therefore, the characteristics were evaluated at the edge region of the grating indicated by the white circle in Fig. 3(b). The measurement spot diameter was 2.5 μm, which is four times that of the grating period, and it was obtained by using a fiber with a core diameter of 50 μm and a 20X objective lens. Figure 4 shows the measurement setup. The retardation was measured based on the rotating-analyzer method with a microspectroscopic system (DF-1037, Techno Synergy, Inc.) and a miniature fiber optic spectrometer (USB2000, OceanOptics).25 To calculate the optical retardation, the transmitted light intensity for each rotation angle of the analyzer was acquired by a spectroscope and fitted to the following Stokes vector of the emitted light S

S=LPθ·X45°·LP90°·Sunp,
(3)

where LP and X are the Mueller matrices of the linear polarizers and of the sample, respectively, and Sunp is the Stokes vector of the non-polarized light source. X denotes the BRM sample modeled as a combination of a partial polarizer and an optical retarder. The intensity of the transmitted light, the S0 component of S, is given by

S0=(p12+p22)2p1p2cosΔcos2θ+(p12p22)sin2θ8,
(4)

where Δ,p1,p2 are the retardation and amplitude transmittances of the TE and TM polarization modes, respectively.

FIG. 4.

Set-up for measuring the retardation of the BRM device with a microspectroscopic system.

FIG. 4.

Set-up for measuring the retardation of the BRM device with a microspectroscopic system.

Close modal

The change in the retardation and transmittances of TM and TE waves at the wavelength of 633 nm, when applying voltages from 0 to 700 V, is shown in Fig. 5. In our BRM device, a maximum modulation of 25.3° (from 21.5° to 46.8°) was achieved at the voltage range of 0–200 V. In the 0–200 V voltage range, the retardation increased, whereas transmittances hardly changed. As shown in the two-dimensional profile in Fig. 3(d), the initial displacement of the movable beams in the measurement spot is negative. Therefore, it is assumed that the effective thickness of the grating increased in the voltage range of 0–200 V. The retardation decreased between 200 and 500 V. It is considered to be due to the decrease in the effective thickness in the positive-displacement region. However, the retardation increased again at voltages exceeding 500 V. The possible interpretation of this phenomenon is the following. The nanograting deformed into an upward convex shape upon release. When a voltage was applied to this convex structure, the beams not only lifted up but were also bent because of non-uniformity of the electrostatic force. Consequently, the beams at the edge region were first lifted up (0–500 V). Then, the horizontal slit between the edge beams and bars narrowed (500–700 V): as the retardation generated in the Au nanograting is strongly affected by the slit width,23 the retardation increases again in this voltage range (500–700 V). However, it should be noted that the device burned out and was rendered inoperable while reproducing the experiment with an applied voltage of 700 V.

FIG. 5.

Measured modulation of the retardation and the transmittances of TM and TE waves of the BRM device for an incident light with 633 nm wavelength when applying a voltage from 0 to 700 V.

FIG. 5.

Measured modulation of the retardation and the transmittances of TM and TE waves of the BRM device for an incident light with 633 nm wavelength when applying a voltage from 0 to 700 V.

Close modal

Using another device fabricated on the same chip with identical design parameters, the relationship between the deformation and the retardation was investigated by measuring the retardation at different points on the convex structure. In this device [shown in Fig. 6(a)], a voltage of 0–700 V was applied together with the first device; the device was able to withstand the voltage but, as in the first device, stiction occurred at the grating edge [Fig. 6(b)] at 500–700 V, resulting in a more convex shape [Fig. 6(c)].

FIG. 6.

(a) SEM image of the second BRM device from a diagonal view, fabricated by the same manufacturing process. The grating area is clearly curved upward. The measurement positions of the grating section with initial displacement are indicated by A–G. (b) SEM image of the slits of the second device. White circle indicates measurement point A. (c) A two-dimensional profile across the second device measured with a laser microscope. (d) Retardation measured by focusing on each point in (a) at the wavelength of 633 nm as a function of the initial displacement measured by a laser microscope.

FIG. 6.

(a) SEM image of the second BRM device from a diagonal view, fabricated by the same manufacturing process. The grating area is clearly curved upward. The measurement positions of the grating section with initial displacement are indicated by A–G. (b) SEM image of the slits of the second device. White circle indicates measurement point A. (c) A two-dimensional profile across the second device measured with a laser microscope. (d) Retardation measured by focusing on each point in (a) at the wavelength of 633 nm as a function of the initial displacement measured by a laser microscope.

Close modal

The retardation was measured at points A–G, as shown in Fig. 6(a), using the same method but without applying voltage. Figure 6(d) shows the measurement results of retardation as a function of the displacement: the retardation decreased as the displacement increased at points A–C. The high retardation at point A is considered to be attributed to the slit narrowing due to the stiction. In addition, the retardation hardly changed at the points C–F, where the displacement exceeded the initial thickness. It is considered that inter-bar and inter-beam interactions Δ2 become dominant instead of bar-beam interaction Δ1 when the displacement exceeded the thickness t of the grating. The curvature of the BRM device can be avoided by proper stress control during film deposition.26,27

In this study, the birefringence modulation in the visible light range by metasurfaces was obtained using an Au nanograting combined with a MEMS actuator. By the electrostatic out-of-plane motion of the nanograting, a maximum modulation of 25.3° (from 21.5° to 46.8°) was achieved at the edge of the grating for a wavelength of 633 nm. However, due to the stress-induced convex shape, modulation could not be achieved from the entire grating. This issue can be improved in future studies by conducting proper stress control during the deposition of thick Au and Si sacrificial layers. In addition, the relationship between retardation and deformation was investigated as a function of the initial deformation in an upward convex shape at the time at which the nanograting structure was released. In future studies, the driving voltage can be significantly decreased by using thinner insulators. It may also improve the stiction problem. The BRM device presented in this study could be expected to be incorporated into small and multi-function display devices and non-invasive medical devices. This research aims to contribute toward the expansion of the applications of metasurfaces in the visible light region.

This work was supported by Grant-in-aid Nos. 16J08512 and 17H02754 for scientific research from the Japan Society for the Promotions of Science (JSPS). This work was also supported by the VLSI Design and Education Center (VDEC) at the University of Tokyo.

1.
N. I.
Zheludev
and
Y. S.
Kivshar
,
Nat. Mater.
11
,
917
(
2012
).
2.
S. B.
Glybovski
,
S. A.
Tretyakov
,
P. A.
Belov
,
Y. S.
Kivshar
, and
C. R.
Simovski
,
Phys. Rep.
634
,
1
(
2016
).
3.
N.
Yu
and
F.
Capasso
,
Nat. Mater.
13
,
139
(
2014
).
4.
L.
Huang
,
X.
Chen
,
H.
Mühlenbernd
,
H.
Zhang
,
S.
Chen
,
B.
Bai
,
Q.
Tan
,
G.
Jin
,
K. W.
Cheah
,
C. W.
Qiu
,
J.
Li
,
T.
Zentgraf
, and
S.
Zhang
,
Nat. Commun.
4
,
2808
(
2013
).
5.
D.
Wen
,
F.
Yue
,
G.
Li
,
G.
Zheng
,
K.
Chan
,
S.
Chen
,
M.
Chen
,
K. F.
Li
,
P. W. H.
Wong
,
K. W.
Cheah
,
E.
Yue Bun Pun
,
S.
Zhang
, and
X.
Chen
,
Nat. Commun.
6
,
8241
(
2015
).
6.
M. X.
Ren
,
W.
Wu
,
W.
Cai
,
B.
Pi
,
X. Z.
Zhang
, and
J. J.
Xu
,
Light: Sci. Appl.
6
,
e16254
(
2017
).
7.
D.
Wang
,
L.
Zhang
,
Y.
Gu
,
M. Q.
Mehmood
,
Y.
Gong
,
A.
Srivastava
,
L.
Jian
,
T.
Venkatesan
,
C. W.
Qiu
, and
M.
Hong
,
Sci. Rep.
5
,
15020
(
2015
).
8.
J.
Rensberg
,
S.
Zhang
,
Y.
Zhou
,
A. S.
McLeod
,
C.
Schwarz
,
M.
Goldflam
,
M.
Liu
,
J.
Kerbusch
,
R.
Nawrodt
,
S.
Ramanathan
,
D. N.
Basov
,
F.
Capasso
,
C.
Ronning
, and
M. A.
Kats
,
Nano Lett.
16
,
1050
(
2016
).
9.
D.
Franklin
,
R.
Frank
,
S.-T.
Wu
, and
D.
Chanda
,
Nat. Commun.
8
,
15209
(
2017
).
10.
N. I.
Zheludev
and
E.
Plum
,
Nat. Nanotechnol.
11
,
16
(
2016
).
11.
J. Y.
Ou
,
E.
Plum
,
L.
Jiang
, and
N. I.
Zheludev
,
Nano Lett.
11
,
2142
(
2011
).
12.
J. Y.
Ou
,
E.
Plum
,
J.
Zhang
, and
N. I.
Zheludev
,
Nat. Nanotechnol.
8
,
252
(
2013
).
13.
J.
Valente
,
J. Y.
Ou
,
E.
Plum
,
I. J.
Youngs
, and
N. I.
Zheludev
,
Appl. Phys. Lett.
106
,
111905
(
2015
).
14.
O.
Buchnev
,
N.
Podoliak
,
T.
Frank
,
M.
Kaczmarek
,
L.
Jiang
, and
V. A.
Fedotov
,
ACS Nano
10
,
11519
(
2016
).
15.
T.
Shimura
,
K.
Iwami
, and
N.
Umeda
,
IEEJ Trans. Sens. Micromachines
137
,
400
(
2017
).
16.
T.
Kan
,
A.
Isozaki
,
N.
Kanda
,
N.
Nemoto
,
K.
Konishi
,
H.
Takahashi
,
M.
Kuwata-Gonokami
,
K.
Matsumoto
, and
I.
Shimoyama
,
Nat. Commun.
6
,
8422
(
2015
).
17.
K.
Yamaguchi
,
M.
Fujii
,
T.
Okamoto
, and
M.
Haraguchi
,
Appl. Phys. Express
7
,
012201
(
2014
).
18.
H.
Honma
,
K.
Takahashi
,
M.
Ishida
, and
K.
Sawada
,
Appl. Phys. Express
9
,
027201
(
2016
).
19.
H.
Honma
,
M.
Mitsudome
,
M.
Ishida
,
K.
Sawada
, and
K.
Takahashi
,
J. Micromech. Microeng.
27
,
034001
(
2017
).
20.
K.
Yamaguchi
,
H.
Yamanaka
,
T.
Ohtsu
, and
S.
Ishii
,
Appl. Phys. Lett.
108
,
111903
(
2016
).
21.
S.-Y.
Hsu
,
K-l
Lee
,
E.-H.
Lin
,
M.-C.
Lee
, and
P.-K.
Wei
,
Appl. Phys. Lett.
95
,
013105
(
2009
).
22.
K.
Iwami
,
M.
Ishii
,
Y.
Kuramochi
,
K.
Ida
, and
N.
Umeda
,
Appl. Phys. Lett.
101
,
161119
(
2012
).
23.
M.
Ishii
,
K.
Iwami
, and
N.
Umeda
,
Appl. Phys. Lett.
106
,
021115
(
2015
).
24.
M.
Born
and
E.
Wolf
,
Principles of Optics
, 5th ed. (
Pergamon Press
,
1975
), pp.
691
692
.
25.
M.
Ishii
,
K.
Iwami
, and
N.
Umeda
,
Opt. Express
24
,
7966
(
2016
).
26.
R.
Liu
,
B.
Jiao
,
Y.
Kong
,
Z.
Li
,
H.
Shang
,
D.
Lu
,
C.
Gao
, and
D.
Chen
,
J. Micromech. Microeng.
23
,
095019
(
2013
).
27.
K.
Dong
,
S.
Lou
,
H. S.
Choe
,
K.
Liu
,
Z.
You
,
J.
Yao
, and
J.
Wu
,
Appl. Phys. Lett.
109
,
023504
(
2016
).