of 18
1
Supporting Information
Electro
-
Optically Tunable Multifunctional Metasurfaces
Ghazaleh Kafaie Shirmanesh,
Ruzan Sokhoyan,
Pin Chieh Wu,
,
and Harry A. Atwater
†,
,*
Thomas J. Watson Laboratory of Applied Physics and
Kavli Nanoscience Institut
e, California Institute
of Technology, Pasadena, California 91125, USA
Department of Photonics, National Cheng Kung University, Tainan 70101, Taiwan
*
E
-
mail
:
haa@caltech.edu
2
1. Comparison to the previously p
roposed design
In our previous paper
,
1
a dual
-
gated Al
-
based metasurface was used, while, in the new design, a
single
-
gated Au
-
based metasurface is presented. In the design presented in our previous paper, all
metasurface elements were connected together a
nd the same bias voltage was applied to all of
them. However, the new metasurface operates by individually controlling each metasurface pixel.
This requires the metasurface pixels to be electronically isolated. Th
e
requirement
of individually
-
addressable m
etasurface elements makes
the single
-
gated
metasurface
easier to be fabricated
compared to
its dual
-
gated counterpart
.
In the proposed single
-
gated metasurface, the lower dielectric layer acts as an optical
dielectric, removing the limitation of using HAOL
as the lower dielectric layer. Replacing the
HAOL
layer
by a lower index dielectric (Al
2
O
3
) resulted in a higher reflectance minimum.
In addition, the proposed Au
-
based metasurface can provide a smaller amplitude
modulation accompanied by a phase shift co
mparable to the single
-
gated Al
-
based metasurface.
This supports the operating principle of the
multifunctional
metasurface that mainly relies on the
phase modulation. Fig
ure S
1 illustrates the amplitude and phase shift of the Al
-
based and Au
-
based metasur
faces as a function of applied biases. As can be seen, the proposed metasurface can
provide a more modest amplitude modulation with an elevated reflectance minimum level
compared to the
previous
metasurface design.
Figure S1. Reflectance an
d
phase shift
as a function of applied bias for the new design compared with
the previous dual
-
gated metasurface design.
1
2
. Calculating electrostatic properties of ITO
In order to calculate the optical response of metasurfaces under applied bias, we first need to
extr
act the properties of the ITO layer. To this end, we obtained the spatial distribution of charge
carriers in the ITO layer by coupling finite difference time domain optical simulations (FDTD
Lumerical) with device physics simulations (Device Lumerical). In
our device physics
calculations, the assumed parameters for ITO were obtained from our previous
work
.
1
Once the
spatial distribution
s
of charge carriers under different applied biases w
ere
identified, we calculated
the complex dielectric permittivity of I
TO
"#$
by using the Drude
model
.
1
Fig
ure
S
2
a
-
c
show the
simulated spatial distribution of charge carrier concentration, real and imaginary part of the
permittivity of ITO for different applied biases. As can be seen in Fig
ure
S
2
b
, by changing the
appl
ied bias, an
epsilon near zero (ENZ) region is observed within the ITO layer, at a region close to
the interface of the ITO and HAOL film.
Voltage (V)
-
4
-
2
0
2
4
16
12
8
4
0
Reflect
ance (%)
300
250
200
150
100
50
0
Phase shift (
)
Al
-
based
Au
-
based
Al
-
based
Au
-
based
3
Figure S2.
Calculated spatial distribution of
(a) charge carrier density, (b) real and (c) imaginary
part of the pe
rmittivity of ITO for different applied biases.
3
. Full
-
wave simulation of reconfigurable metasurface
Figure S3. (a) Simulated phase shift spectrum for different applied biases. The phase shift was
calculated with respect to the phase of the reflection f
rom the metasurface without applied voltage.
(b) Maximum achievable reflectance change (square) and phase modulation (diamond) at different
wavelengths.
To gain further insight, we analyze
d
the
electromagnetic
fields
in the metasurface at the
resonance wav
elength.
The spatial distribution of the
electric field intensity
, under
an
applied bias
of
-
6V, 0V, and +6V are presented in
Fig
ure
S
4
a
-
c
. Zoomed
-
in images of
the
electric field
inside
the dielectric spacer of the metasurface, consisting of
the Al
2
O
3
/ITO/
HAOL
heterostructures
are
shown in the insets. As can be seen, at the applied bias of +6V, there will be a strong field
enhancement in the accumulation region of the ITO layer which occurs subsequent to the existence
of ENZ condition in the mentioned layer
. By leveraging this strong field enhancement in the
active
region of the ITO layer, the complex permittivity change can be drastically emphasized.
Our electromagnetic calculations show
that the metasurface phase and amplitude tunability
is based on an int
erplay between magnetic plasmon resonance and the ENZ region of ITO. To
show this, the amplitude of the magnetic field inside the metasurface is shown in Figure S5. The
field distributions are presented at different wavelengths and under different applied
biases. As
can be seen in Figure S5d
-
f, at the resonance wavelength
l
=
1510 nm, there exists
a
large
-
magnitude magnetic field that is localized in the gap region between gold antenna and gold back
plane.
Figure S6 shows the spatial distribution of the z
-
c
omponent of the electric field
within the
metasurface under different applied biases. As can be seen, the z components of the electric field
around the right and left edges of the antenna are antiparallel to each other. This antiparallel field,
accompanied
by the curl of the current density is consistent with large magnetic field shown in
N
ITO
Distance from ITO/HAOL
interface (nm)
Voltage (V)
x 10
20
Distance from ITO/HAOL
interface (nm)
Voltage (V)
a
b
Re
{
푰푻푶
}
Distance from I
TO/HAOL
interface (nm)
Voltage (V)
Im
{
푰푻푶
}
c
Phase shift (
°
)
(R
max
-
R
min
)/R
min
Wavelength (nm)
a
b
Phase shift (
°
)
4
Figure S5d
-
f
,
indicat
ing
that the considered resonance is a magnetic plasmon resonance. We note
that the strength of the magnetic dipole is strongly modified by altering t
he applied bias.
F
igure S4.
Spatial distribution of the amplitude of the electric field at the operating wavelength of
l
=
1510
nm for different voltages of (a) V
=
-
6 V, (b) V
=
0, and (c) V
=
6 V. The insets show the
zoomed
-
in image of the Al
2
O
3
/ITO/HA
OL nano
-
sandwich.
Figure S5.
Spatial distribution of the amplitude of the magnetic field at the operating wavelength of
(a
-
c)
l
=
1100 nm, (d
-
f)
l
=
1510 nm, and (g
-
i)
l
=
1510 nm for different voltages of V
=
-
6 V (a, d,
and g), V
=
0 (b, e, and h), and
V
=
6 V (c, f, and i).
HAOL
Al
2
O
3
ITO
Anten
na
Back
reflector
|E| (V/m)
a
c
b
|H| (A/m)
a
c
b
d
f
e
g
i
h
5
Figure S6. Spatial distribution of the z
-
component of the electric field at the wavelength of
l
=
1510
nm, under applied bias (a) V
=
-
6 V, (b) V
=
0, and (c) V
=
6 V.
4. Increasing metasurface reflectance level
Due to the
stro
ng field confinement in the ITO layer of the metasurface at ENZ condition, which
is indeed the operation principle of the modulation provided by the metasurface, these plasmonic
active metasurfaces show small reflectance values. It should be noted that in
the proposed tunable
metasurface, one can always see a tradeoff between the amplitude of the reflection from the
metasurface and the maximum achievable phase shift.
Figure S
7
shows how changing the thickness of different layers could increase the reflectan
ce
value at the cost of decreasing the phase shift. Figure S
7
a shows the reflectance spectrum for
different antenna thicknesses for the unbiased case (see Figure S
7
b for a zoomed in presentation
at the resonant vicinity).
Figure S
7
c presents t
he spectrum o
f the maximum achievable phase shift
defined by
Max
.
phase
shift
=
Acquired
phase
(
V
=
6V
)
Acquired
phase
(
V
=
6V
)
(S1)
As can be seen, increasing the thickness of the antennas could increase the unbiased reflectance
level to some extent. While
further increasing the antenna thickness will result in a higher
reflectance, it will drastically decrease the maximum achievable phase shift. The same tradeoff
can be observed when increasing the thickness of the lower dielectric layer and the ITO layer
for
which
reflectance (maximum achievable phase shift) spectra are depicted in Figure S
7
d,e (Figure
S
7
f) and Figure S
7
g,h (Figure S
7
i), respectively.
Another method to alter the reflectance from the metasurface is to cover the antennas by a dielectric
laye
r. Figure S
7
j,k show the unbiased reflectance from the metasurface for different thicknesses
of a SiO
2
layer covering the metasurface
area
. The maximum achievable phase shift spectrum is
depicted in Figure S
7
l. As can be seen, increasing the thickness of t
he top SiO
2
layer up to 80 nm,
would
increase the reflectance from the metasurface. However, a 120 nm
-
thick SiO
2
layer will
drastically decrease the maximum achievable phase shift. It should be noted that the dimensions
of the antenna and the electrode are
adjusted to
l
a
=
210 nm,
w
a
=
115 nm, and
w
e
=
130 nm
such
that the resonance wavelength of the metasurface when covered by an 80 nm
-
thick SiO
2
layer is
the same as
that of
the original metasurface.
In addition to changing the metasurface physical dimensi
ons,
replacing the ITO layer by
transparent conducting oxides with higher electron mobilities such as CdO,
is expected to
increase
the reflectance level of the metasurfac
e
.
However,
investigating this
would be
beyond the scope
of this paper.
E
z
(V/
m)
a
c
b
6
Figure S7.
E
ffect of different layers of the tunable metasurface on reflectance and maximum
achievable phase shift of the metasurface. (a) Reflectance spectrum, (b) zoomed
-
in image of the
reflectance spectrum, and (c) maximum achievable phase shift for different anten
na thickness values.
(d) Reflectance spectrum, (e) zoomed
-
in image of the reflectance spectrum, and (f) maximum
achievable phase shift for different lower Al
2
O
3
thickness values. (g) Reflectance spectrum, (h)
zoomed
-
in image of the reflectance spectrum, an
d (i) maximum achievable phase shift for different
ITO thickness values. (j) Reflectance spectrum, (k) zoomed
-
in image of the reflectance spectrum, and
(l) maximum achievable phase shift for different top SiO
2
thickness values.
1100
1300
1500
1700
1900
Wavelength (nm)
1100
1300
1500
1700
1900
Wavelength (nm)
1100
1300
1500
1700
1900
Wavelength (nm)
Reflectance (%)
0
20
40
60
80
100
Max. phase shift (
)
-
400
-
300
-
200
-
100
0
100
1100
1
300
1500
1700
1900
Wavelength (nm)
1450
1500
1550
1600
Wavelength (nm)
Reflectance (%)
0
10
20
1100
1300
1500
1700
1900
Wavelength (nm)
1100
1300
1500
1700
1900
Wavelength (nm)
1400
1450
1500
1600
Wavelength (nm)
1550
1400
1450
1500
1600
Wavelength (nm)
1550
1100
1300
1500
1700
1900
Wavelength (nm)
1450
1500
1600
Wavelength (nm)
1550
1100
1300
1500
1700
1900
Wavelength (nm)
a
d
g
j
b
e
h
k
c
f
i
l
7
5
.
Multifunctiona
l
metasurface
fabrication
Figure S
8
.
Sample fabrication steps. (a) Patterning the out
er
most connecting pads; (b) patterning the
back reflector; (c) deposing the Al
2
O
3
layer; (d) patterning the ITO layer; (e) patterning the
connecting pads of the ITO lay
er; (f) depositing the HAOL gate dielectric; and (g) patterning the
antenna array and the inner connecting pads.
6
.
Universal
measurements setup
In order to measure all the functions p
erform
ed by our
multifunctional
metasurface, we designed
and built a
cu
stom
optical measurement setup. This
measurement
setup is capable of measuring
reflectance spectrum, phase shift, beam steering and reconfigurable focusing. Fig
ure
S9
shows our
custom
-
built setup. Each measurement is performed through a part of our
univers
al
setup that are
discussed in detail in the following sections. For all measurements, we utilized an uncollimated
white light source from a halogen lamp (LA) and a visible
CMOS image sensor
camera (V
-
C) to
visualize the sample surface.
b
e
c
f
a
d
g