2104 Vol. 9, No.

10 / October 2021 / Photonics Research Research Article

Reconfigurable metasurface-based 1 × 2
waveguide switch
AMGED ALQULIAH,1,2 MOHAMED ELKABBASH,3,4,5 JINLUO CHENG,1,2,6 GOPAL VERMA,1,2
CHAUDRY SAJED SARAJ,1,2 WEI LI,1,2,7 AND CHUNLEI GUO3,8
1
GPL, Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, China
2
University of Chinese Academy of Sciences, Beijing 100049, China
3
The Institute of Optics, University of Rochester, Rochester, New York 14627, USA
4
Current address: The Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA
5
e-mail: [email protected]
6
e-mail: [email protected]
7
e-mail: [email protected]
8
e-mail: [email protected]

Received 22 April 2021; revised 13 August 2021; accepted 20 August 2021; posted 24 August 2021 (Doc. ID 428577);
published 30 September 2021

Reconfigurable nanophotonic components are essential elements in realizing complex and highly integrated pho-
tonic circuits. Here we report a novel concept for devices with functionality to dynamically control guided light in
the near-visible spectral range, which is illustrated by a reconfigurable and non-volatile (1 × 2) switch using an
ultracompact active metasurface. The switch is made of two sets of nanorod arrays of TiO2 and antimony tri-
sulfide (Sb2 S3 ), a low-loss phase-change material (PCM), patterned on a silicon nitride waveguide. The metasur-
face creates an effective multimode interferometer that forms an image of the input mode at the end of the stem
waveguide and routes this image toward one of the output ports depending on the phase of PCM nanorods.
Remarkably, our metasurface-based 1 × 2 switch enjoys an ultracompact coupling length of 5.5 μm and a record
high bandwidth (22.6 THz) compared to other PCM-based switches. Furthermore, our device exhibits low losses
in the near-visible region (∼1 dB) and low cross talk (−11.24 dB) over a wide bandwidth (22.6 THz). Our pro-
posed device paves the way toward realizing compact and efficient waveguide routers and switches for applications
in quantum computing, neuromorphic photonic networking, and biomedical sensing and optogenetics.
Published by The Optical Society under the terms of the Creative Commons Attribution 4.0 License. Further distribution of this work
must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI.
https://doi.org/10.1364/PRJ.428577

1. INTRODUCTION via two phase shifters that are independently tuned through
The advancement of nanofabrication, coupled with the attain- volatile and weak modulation of the waveguide refractive index
ment of a high level of complexity in photonic integrated cir- commonly using electro-optic or thermo-optic effects [4], lead-
cuits (PICs), triggered tremendous interest toward realizing ing to devices with a limited tunability, high energy consump-
miniaturized all-optical interconnects that are superior to elec- tion [several milliwatts (mW)], and large footprints [hundreds
tronic circuits in terms of bandwidth density, speed, and energy of micrometers (μm)] [14]. On the other hand, micro-ring res-
efficiency as well as mitigating the von Neumann data trans- onators (MRRs) [15–20] and micro-electromechanical systems
mission bottleneck [1–3]. Particularly, the reconfigurable con- (MEMS) [21,22] offer high modulation depth and a relatively
trol of light propagating in PICs is crucially important for many small footprint. Still, they suffer from a narrow operational
emerging applications such as programmable PIC [4], neuro- bandwidth (less than 3 dB) [20], low tolerance to temperature
inspired computing [5,6], quantum information processing variations and fabrication errors, as well as a large actuation
[7,8], optical communication [9,10], microwave photonics voltage (>40 V) [23]. Notably, systems based on exciting sur-
[11], and sensor applications [12,13]. face plasmon resonances (SPRs) enjoy the highest switching
Reconfigurable photonic computing cores are convention- rates and smallest footprints. However, they suffer from high
ally implemented using waveguide meshes of Mach–Zehnder insertion and propagation losses as they require coupling
interferometers (MZIs) where the interference is controlled from/to a photonic waveguide [24,25], and plasmonic metals

2327-9125/21/102104-12 Journal © 2021 Chinese Laser Press

 Research Article Vol. 9, No. 10 / October 2021 / Photonics Research 2105

are highly lossy [25], which hinder their widespread usage. The application specific because the optical properties of the constitu-
common volatility of these schemes necessitates an “always-on” tional meta-atoms are permanent once fabricated. To actively
power supply to retain the switching state, rendering them en- tune the optical properties of metasurfaces, several approaches
ergy inefficient [14,26]. have been introduced [47,48] with considerable interest in using
To circumvent these hurdles, phase-change materials PCMs [28,49] due to their advantages as discussed above. So far,
(PCMs) emerged as candidates to demonstrate photonic recon- such investigations on reconfigurable metasurfaces are focusing
figurability owing to their unique tunable properties [27,28]. on controlling light propagating in free space, e.g., for bichro-
PCMs possess high contrasts in the electrical resistivity and matic and multifocus Fresnel zone plates [50], beam steering
refractive index between the resonant-bonded crystalline and [31], tunable color generators [51], dynamic spectrum control-
covalent-bonded amorphous phase states over a wide spectral lers [52], switchable spectral filters [53], information processing
range. They are non-volatile, reversible, and they provide fast [54], communication [55], imaging [56], hologram and aug-
and low energy actuation (<10 aJ∕nm3 [29]) by ultrashort mented reality [57–59], vortex beam generators, and illusion
electrical or optical pulses (up to subnanosecond) [30] and and cloaking [60].
stable switching ability of more than 1015 cycles [31]. In ad- In this paper, we extend the concept of reconfigurable meta-
dition, the scalability of PCMs makes their nanofabrication rel- surfaces to PICs by reporting a novel dynamic near-visible nano-
atively approachable and compatible with other substrates, as photonic (1 × 2) switch enabled by combining two nanorod
their amorphous state is used during deposition [23]. Several arrays of TiO2 and a novel ultralow-loss PCM (antimony trisul-
PCM-based integrated photonic devices were recently demon- fide Sb2 S3 ) superimposed on silicon nitride waveguide. Our de-
strated, e.g., photonic memories [32,33], optical modulators signed device provides unprecedented functionality as it
[34,35], optical switches [14–20], and optical computing facilitates the non-volatile dynamic routing of light inside a mul-
[36,37]. In these applications, a top-cladding layer of timode waveguide toward predefined outputs. The device shows
Ge2 Sb2 Te5 (GST) and, recently, Ge2 Sb2 Se4 Te1 (GSST) was a wide bandwidth of 22.6 THz, low loss (∼1 dB), and low cross
deposited onto a waveguide. However, the high absorption talk (< − 11.29 dB) with a compact active length (5.5 μm). To
losses in one of the phases of such materials fundamentally re- the extent of our knowledge, our design enjoys a record footprint
strict their use in phase modulation schemes for large-area PICs compared to PCM-based optical switches, which would over-
[4] and deep-neural networks [5,6], where light would propa- come the challenges associated with large-area PCM actuation.
gate through several PCM-based interconnects. Although losses Table 1 compares the design principles, the type of PCM de-
can be alleviated in devices working at the telecom wavelengths, signs, and performance of the reported (1 × 2) PCM-based
this comes with the cost of sacrificing the device footprint. In switches and of our metasurface PCM-based switch.
addition, such approach is impractical at the visible and near- Appendix B discusses the need for visible/near-visible PICs.
visible wavelengths due to the large intrinsic losses. Finally, the
use of large-area PCMs creates a considerable barrier to the ac- 2. MATERIALS AND METHODS
tuation mechanism (see Appendix A for more details). Figure 1(a) schematically illustrates the structure and function-
On the other hand, optical metasurfaces enable unprec- ality of our device, which is a Y-branch waveguide with a meta-
edented flexibility in controlling the propagation of light through surface located on the surface of its stem. From bottom to top,
a spatially dependent and abrupt phase change at an interface the waveguide is formed by a silicon substrate, a 3 μm thick
that is imposed by ultrathin artificial arrays of engineered sub- SiO2 layer, and a ridge SiN waveguide. The metasurface con-
wavelength nanoantennas [38]. Metasurfaces have realized a sists of two sets of nanorods. Each set includes 28 equally
plethora of ultracompact, broadband, and efficient on-chip pho- spaced nanorods with a length L and a center-to-center distance
tonic devices, such as mode converters [39], polarization rotators between neighbor nanorods Λ. These two sets are aligned to
[39], mode-pass polarizers [40], power splitters [41], asymmetric form two adjacent rectangles separated by a gap g. The total
power transmitters [39], second-harmonic generators [42], re- footprint of the metasurface is Lx .
mote near-field controllers [43], and guided waves to free space The materials are chosen to realize a device appropriate for
wave couplers [44–46]. However, these devices are passive and near-visible integrated photonics. The SiN waveguide has the

Table 1. Comparison of Previously Reported PCM-Based (1 × 2) Switches with Our Metasurface-Based Switch
Design Principle PCM Design Bandwidth (THz) Cross Talk (dB) Operating Wavelength Range Reference
Directional coupler GST layer 3.74 −10 IR [14]
Directional coupler GSST layer 4.4 −32 IR [61]
Contra-directional coupler GST-Si grating 0.275 −30 IR [62]
Micro-ring resonator GST layer 0.125 −6 IR [16]
Micro-ring resonator GST layer 0.125 −33 IR [17]
Micro-ring resonator GST layer 0.125 −42 IR [18]
Micro-ring resonator GST layer 0.125 −5 IR [15]
Micro-ring resonator GST nanodisk 0.125 −5 IR [20]
Micro-ring resonator GST layer 0.125 −14.1 IR [19]
Micro-ring resonator Sb2 S3 layer 0.125 NA Visible [63]
Y branch Sb2S3/TiO2 metasurface 22.6 −11.24 Near visible This work
 2106 Vol. 9, No. 10 / October 2021 / Photonics Research Research Article

pulses with long time duration (100 ms) [67] to reach the
Sb2 S3 glass transition temperature at 270°C [67]. In contrast,
the re-amorphization process requires pulses with shorter time
duration (400 ns) [67] to reach the Sb2 S3 melting temperature
at 527.85°C [67] and then quickly cool it at a rate >20°C∕ns
[53,67,69].
In our designed metasurface, the nanorod antennas cause a
spatial linear phase shift of the optical mode in the waveguide in
addition to the propagation phase along the propagation direc-
tion (x axis). The accumulative local phase shifts from the two
nanorod sets result in a constructive interference that focuses
the input fundamental mode in the stem multimode wave-
guide. Following that, the focused input mode partially con-
verts the input fundamental mode to higher-order modes,
creating an effective asymmetric multimode interferometer
(MMI). Consequently, an image of the input mode is formed
at the end of the stem multimode waveguide due to the self-
imaging principle of the MMI [70]. The produced image is
then routed in one of the predefined output ports depending
on the phase of the PCM nanorods (specifically depending on
the effective refractive index of the nanorod arrays). Therefore,
Fig. 1. Design of the proposed metasurface-based reconfigurable when Sb2 S3 is in the crystalline state, it has a significantly
(1 × 2) integrated switch working around λ
800 nm. (a) 3D illus- higher effective index, causing a larger phase delay at the side
trations of the device structure and its functionality when the Sb2 S3 where the PCM nanorods are located. In contrast, when the
metasurface structure is in a crystalline or amorphous state. The top- Sb2 S3 is in the amorphous state, the TiO2 causes a larger phase
left inset shows the top view of the device with dimensions of the meta- delay at the other side of the waveguide, leading to the field
surface consisting of a set of passive TiO2 nanorods and another set of constructively interfering at the opposite side.
PCM (Sb2 S3 ) nanorods. The top middle inset shows the conditions
The optical properties of the device were simulated using
required for the Sb2 S3 to undergo a reversible structural transition
from amorphous to crystalline states. (b) Cross section of the device. the finite-difference time-domain method (FDTD, Mode
(c) and (d) Wavelength dependence of the complex optical constants Solutions, Lumerical Ansys Inc.). The simulation domain
(n, k) of amorphous Sb2 S3 , crystalline Sb2 S3 , amorphous titanium di- was enclosed by eight standard perfectly matched layers as
oxide, and silicon nitride. Highlighted parts in red indicate the spectral boundary conditions. To minimize the numerical dispersion
region of interest where Sb2 S3 exhibits low loss and high switching and to enhance the interfaces’ resolution, we chose the nonuni-
contrast. form auto mesh setting with a minimum size of conformal
mesh cell of 2.5 nm. The fundamental TE mode was launched
into the stem SiN waveguide using a broadband mode source
advantages of thermodynamic stability, wide spectral range of and enabling multifrequency calculations. The frequency-
transparency (λ
0.25–8 μm), higher fabrication flexibility, domain time power monitors were used to record the profile
and lower propagation losses compared to silicon waveguides of the normalized electric field intensity, transmission at output
[64]. The materials of the two sets of nanorods are chosen ports, reflection, and scattering. For calculating the effective
as PCM antimony trisulfide (Sb2 S3 ) for one set and amorphous indices as well as the dispersion in the SiN waveguide, the
titanium dioxide (a-TiO2 ) for the other. Compared to the finite-difference eigenmodes solver (FDE, Mode Solutions,
prototypical PCM (GST), the new class of PCM (Sb2 S3 ) Lumerical Ansys Inc.) was used. To estimate the effective in-
has a bandgap tunable from 2.0 eV (crystalline) to 1.7 eV dices of metasurface nanorod arrays, the effective medium
(amorphous), and thus it works in the near-visible spectrum theory was considered by using Rytov’s approximations.
with a low absorption coefficient and a relatively higher con- To optimize the device performance, the ridge SiN multi-
trast in the change of the real part of its refractive index Δn mode waveguide is chosen to have a rectangular shape with a
[65–67]. A-TiO2 is chosen because it is lossless in the near- height h
220 nm, a thickness of under-etched layer
visible range and enjoys a high refractive index to provide s
200 nm, and a sufficiently large width d
3.5 μm to sup-
strong guided light–nanorod interaction [68]. The top middle port the multimode interference and to reduce the cross talk
inset in Fig. 1(a) shows the reversible optical switching of the between the output ports [70] [see Fig. 1(b)]. Note that the
Sb2 S3 from the crystalline to amorphous phases. The optical two output waveguides are narrower to match the spot size
switching [65–67] can be done using a 630 nm laser source of the produced single self-images and optimize the device cross
with a power of up to 90 mW. This laser source provides stable talk. The mode of the output waveguides can be coupled with
performance and cycling durability of Sb2 S3 during the experi- other devices with arbitrary widths using techniques such as
ment [67]. The reversible switching between both structural waveguide tapers [71–74] or integrated mode size converters
states of Sb2 S3 is realized by controlling the width and focus [75–77]. Appendix C shows the dependence of effective refrac-
of the laser pulse [67]. The crystallization process requires tive indices and dispersion of modes on the width of the SiN
 Research Article Vol. 9, No. 10 / October 2021 / Photonics Research 2107

waveguide. The two output ports have a width d out
footprint (Fig. 2, inset). The suggested fabrication method
1.65 μm and are attached to the end of the stem waveguide and working setup of the proposed device are provided in
at a position where the self-image is formed. During the sim- Appendix E.
ulation, the complex optical constants of SiN, SiO2 , Si, and
TiO2 are obtained from the database of Palik [78], while
the experimental values of Sb2 S3 are taken from 3. RESULTS AND DISCUSSION
Refs. [67,79]. Figures 1(c) and 1(d) list the optical constants Before presenting the functionality of the metasurface, it is in-
of a-TiO2 , SiN, and two phases of Sb2 S3 . To compromise be- structive to characterize how the single metasurface nanoan-
tween the absorption coefficient (k) and the refractive index tenna confines the light inside the waveguide. Figures 3(a)
contrast (Δn), the operating bandwidth is set around and 3(b) show the calculated field distribution (E z compo-
800 nm as shown in the red bars in Figs. 1(c) and 1(d). nents) and the electric field intensity jEj2 distribution in the
The performance metrics of a waveguide optical switch are nanorods and for both phases of Sb2 S3 . In our design, the di-
the cross talk and the transmission, where the former is defined mensions of the TiO2 nanorods are larger than those of the
as the contrast transmission ratio between the two output ports Sb2 S3 nanorods in order to increase the magnitude of phase
[14]. Although it is not easy to define a qualitative quantity for delay (i.e., effective refractive index) and compensate for its
the performance metric, an efficient optical switch should lower refractive index compared to a-Sb2 S3 . Therefore, for
exhibit low cross talk and high transmission. The metasurface a-Sb2 S3 , the field is mostly confined in the TiO2 nanorod.
was engineered adopting the direct design approach [80]. To When the PCM changes its state to c-Sb2 S3 , the field becomes
obtain the target performance, first, the width (W Sb2 S3 ) and strongly confined in the Sb2 S3 rods. Figure 3(c) shows the
height (hs ) of Sb2 S3 nanorods were fixed at (W Sb2 S3
length dependence of the calculated effective refractive index
55 nm, hs
50 nm) to support Mie resonant modes in the for different nanorod antennas positioned on the SiN substrate
nanoantennas [39,42,81,82] and to meet the current state at λ
800 nm. The FDE method was implemented to obtain
of manufacturing capacity [65,66]. The period (Λ) was set the effective refractive indices caused by metasurface nanoan-
at 200 nm to avoid diffraction effects [83,84]. Following that, tennas. A significant neff difference is obtained between a-Sb2 S3
the dimensions of TiO2 -nanorods (width and height) were op- and c-Sb2 S3 antennas due to their high refractive index con-
timized to realize the desired optical routing response (see trast. The neff of TiO2 lies between the neff of a-Sb2 S3 and
Appendix D for further details). The gap width (g) between c-Sb2 S3 , which in turn partly explains the switching function-
the two nanorod sets was kept at 180 nm to bring about degrees ality of our device. Appendix F discusses the method used for
of freedom to the metasurface fabrication. Finally, Lx , L, g, and calculating the equivalent effective indices of the Sb2 S3 and
Λ were parametrically swept in sequential order as shown in TiO2 nanorod arrays. These results show that employing
Figs. 2(a)–2(d). The final dimensions of the metasurface are high-index dielectric/PCM nanoantennas on a waveguide offers
as follows: the TiO2 nanorods have a width of 72 nm and a a unique ability to independently and dynamically tune the
thickness of 250 nm; Λ
200 nm, L
1060 nm, localization of guided light by engineering the relative effective
g
180 nm, and Lx
5.5 μm, indicating an ultracompact indices induced by the metasurface nanoantennas.
Figure 4(a) shows the normalized electric field intensity pro-
file jEj2 in the switch for a-Sb2 S3 at a wavelength 800 nm in

Fig. 3. Characterization of nanoantenna structure. (a) Profile (zy

plane) of the normalized near field of the E z component showing scat-
tering Mie modes in the TiO2 and Sb2 S3 nanoantennas for a-Sb2 S3
(left panel) and c-Sb2 S3 (right panel) at λ
800 nm. (b) Normalized
electric field intensity jEj2 distribution (zy plane) in the same nanorods
for a-Sb2 S3 and c-Sb2 S3 . The boundaries of the SiN waveguide and
Fig. 2. Metasurface parametric sweep. (a)–(d) Device cross talk as a nanoantennas are indicated in solid black lines. (c) Effective refractive
function of metasurface footprint (Lx ), nanorod length (L), separated index of the TE00 mode as a function of TiO2 nanoantenna length and
gap width (g), and the nanorods’ center-to-center distance (Λ), respec- Sb2 S3 nanoantenna length for the a-Sb2 S3 and c-Sb2 S3 phases. The
tively, for a-Sb2 S3 and c-Sb2 S3 . The dotted gray lines show the dimen- dashed gray line indicates the length of the nanorods (L) used in our
sions used in our metasurface. simulation. The inset figure shows the FDE simulation setup.
 2108 Vol. 9, No. 10 / October 2021 / Photonics Research Research Article

mode and do not experience any polarization rotation (see

Fig. 12 in Appendix G).

4. CONCLUSION
In conclusion, we have proposed a conceptually novel approach
for devices that dynamically control guided light using a
reconfigurable metasurface. We have demonstrated a broad-
band compact and low-loss (1 × 2) switch for near-visible wave-
lengths by integrating a metasurface consisting of two nanorod
arrays of TiO2 and Sb2 S3 on a silicon nitride waveguide.
Our demonstrated device enjoys a record high bandwidth
(22.6 THz) compared to other phase-change-material-based
switches while having low loss (∼1 dB), low cross talk
Fig. 4. Simulated device performance for a-Sb2 S3 (upper panel) and (−11.24 dB), and ultracompact active length (5.5 μm). The
c-Sb2 S3 (lower panel). (a) and (d) Full-wave simulation showing the proposed device could be a reliable component in the meshes
optical field intensity jEj2 in the switch for the fundamental TE mode of future energy-efficient large-scale PICs. Moreover, we believe
in the xy plane at λ
800 nm. The boundaries of the SiN wave- that integrating active metasurfaces with photonic waveguides,
guide and metasurface structure are indicated by dashed lines and rec- as demonstrated in our example, may provide a step change
tangles, respectively. Inset: enlarged view of the field profile in the toward realizing several tunable, efficient, and non-volatile
metasurface. (b) and (e) Transmission spectra at two output ports chip-scale devices for applications in neuromorphic computing
Port2 and Port3 . (c) and (f ) Total transmission at output ports
[5,6], quantum information processing [7,8], optical commu-
(Port2  Port3 ), reflection, and scattering of the device.
nication [9,10], microwave photonics [11], and biomedical
sensing [12,13].

the xy plane, when the fundamental TE mode is launched from APPENDIX A: THE DRAWBACKS OF USING
the stem multimode waveguide to the Y branches along the LARGE-AREA PCMS FOR DEVICES’ ACTUATION
propagation direction (x axis). In this case, the field is localized MECHANISMS
in the TiO2 nanoantennas as shown in the enlarged view of the The use of large-area PCMs creates a considerable barrier to the
inset of Fig. 4(a); the result is consistent with Fig. 3(a), where actuation mechanism because of the following reasons: first, the
TiO2 nanorods show stronger confinement than a-Sb2S3 nano- lack of optimization for additional scaling and integration be-
rods. Figure 4(b) shows the calculated transmission spectra cause of the inaccurate, slow, and diffraction-limited alignment
from Port1 to Port2 and from Port1 to Port3 . The average process [85]; second, the difficulty in attaining similar levels of
simulated insertion loss in the target port is <0.706 dB. crystallization and amorphization using a thermodynamic
The insertion losses are caused mainly by the reflected optical mechanism over a large-area PCM [67,86]; third, large-area
power resulting from the impedance mismatch with the meta- PCMs prevent achieving the sufficient cooling rate [31,53] that
surface nanoantennas and the scattering losses are due to the is necessary for the re-amorphization process of PCMs (because
strong light–antenna interactions at subwavelength intervals. PCMs suffer from low thermal conductivity [31,87,88]). These
The average value of the cross talk is found to be −11.24 dB reasons lead to weak and inconsistent switching behavior of
throughout the operating bandwidth. large-area PCMs [23,67,86]. In addition, another significant
To further characterize the device performance, Fig. 4(c) limitation is the filamentation phenomenon associated with
shows the spectra of total transmission (T, Port2  Port3 ), re- electrical switching. Such filamentation causes a nonuniform
flection (R), and scattering (S) over the simulated wavelength crystallization throughout large-area PCMs. These limitations
region. eventually lead to devices with weak and inconsistent switching
Figures 4(d)–4(f ) show the functionality of the switch for behavior [86,88,89].
c-Sb2 S3 . In this case, the field is localized in the c-Sb2 S3 nano- A practical solution is reducing the PCMs’ volume to the
rods because of its stronger light confinement than TiO2 nano- subwavelength scale [31,53,86,88,89]. This solution is adopted
rods [see Fig. 3(b)]. From the calculated transmission spectra in from the current well-established phase-change random access
Fig. 4(e), the average value of cross talk is −10.01 dB through- memory (PRAM), where the PCM volumes are deeply scaled
out the operating bandwidth. The average calculated insertion and embedded in thermally optimized units to achieve the
loss is ∼1 dB, which is lower than the efficiency when the required cooling rate [30,90]. As a result, PRAMs enjoy a high
switch is in the a-Sb2 S3 . The higher insertion losses are due to switching cyclability (above 1 × 106 ) [69] and do not suffer
increased reflection and scattering losses are because of the from the filamentation issue [86]. The introduction of PCM-
higher refractive index of c-Sb2 S3 compared to a-Sb2 S3 . based metasurfaces [27,28,31,86–89,91] represents a viable
Figure 4(f ) shows the spectra of total transmission (T, technological solution for many potential applications. In such
Port2  Port3 ), reflection (R), and scattering (S) over the simu- a platform (PCM-based metasurfaces), each meta-atom is ther-
lated wavelength region. We note that the produced modes at mally modulated independently rather than modulating the
Port2 and Port3 maintain the polarization of the input TE00 whole large-area PCM. Hence, the speed and reliability increase
 Research Article Vol. 9, No. 10 / October 2021 / Photonics Research 2109

without suffering cooling rate inaccessibility, switching nonuni- of stem waveguide width (W ) on the effective refractive index
formity, or crystallization filamentation [27,28,31,86–89,91,92]. (neff ) and D of different modes as shown in Figs. 5(a) and
5(b). A wide range of neff (from 1.72 to 1.878) can be obtained,
which corresponds to the propagation constant of supported
APPENDIX B: THE NEED FOR VISIBLE/NEAR-
modes in the waveguide. As it can be seen in Fig. 5(a), the
VISIBLE PHOTONIC INTEGRATED DEVICES
neff increases rapidly with the increasing W and reaches a maxi-
The PCM-based integrated devices reported thus far are mal constant value [maxneff 
1.878] for the TE00 mode.
restricted to the infrared region, where the silicon waveguide Also, the ∇neff between the modes decreased with the increas-
exhibits optical transparency and where the PCM (GST and ing W . Note that the height (h) of the SiN waveguide is kept
GSST) shows low absorption and high Δn [28]. It is therefore constant at 220 nm during the simulation, which is below the
crucially important to realize visible and near-visible integrated maximum value (800 nm) for a crack-free (low pressure chemi-
photonics that display more compact size, as the footprint of cal vapor deposition, LPCVD) SiN layer [39,95,96]. As we are
the device relies on the wavelength of its source [93], and to interested in the TE00 mode that is injected into our device,
harness the exotic optical phenomena at this spectral region Fig. 5(c) shows the simulated neff and D curves of the fundamen-
[94], which have been elusive because of the incompatibility tal TE mode in a SiN waveguide with a width 3.5 μm over the
with silicon-based PICs due to the high absorption of Si. In simulated wavelengths. In Figs. 5(b) and 5(c) we can see that the
addition, visible/near-visible photonic integrated devices are waveguide width 3.5 μm features a flat anomalous dispersion for
crucial to many applications, particularly in integrated quan- TE00 mode over the simulated wavelength. The dependence of
tum photonics, as quantum light sources, e.g., color centers waveguide dispersion on the neff can be given by [97]
and quantum dots, emit light in this wavelength range.
λ d2 neff
Also, it is important for biomedical sensing and optogenetics. D
− , (C1)
c dλ2
where c is the speed of light in a vacuum.
APPENDIX C: THE DESIGN OF A STEM
MULTIMODE WAVEGUIDE
APPENDIX D: THE ROLE OF OPTIMIZING THE
To study the MMI effect that takes place in our SiN waveguide,
HEIGHT AND WIDTH OF TiO2 IN THE DEVICE
it is vital to check the properties of the supported modes in the
PERFORMANCE
waveguide. Therefore, in Mode Solutions (Lumerical Ansys
Inc), the FDE method was used to calculate the dependence To justify the selected dimensions for TiO2, in Fig. 6, we show
the device performance (i.e., cross talk and transmission at the
desired output) for c-Sb2 S3 and a-Sb2 S3 , considering variations
in the height and width of TiO2 nanorods. It is to be noted that
the sole reason for increasing the dimensions of TiO2 nanorods
is to optimize the device performance in the amorphous state of
Sb2 S3 . As it can be seen in Fig. 6(b), the cross talk decreases
notably with the increasing dimensions of TiO2 nanorods and
reaches a global minimum at height 250 nm and width 72 nm.
Moreover, the transmission at the desired output in the amor-
phous state reaches global maximum at the same TiO2 nanorod
dimensions [see Fig. 6(d)].
To further clarify the reason for this behavior, we conducted
full-wave simulations to show the propagation of electric field
intensity in the device in the crystalline and amorphous states
in three different scenarios: (1) the dimensions of the TiO2
nanorods are exactly the same as those of the Sb2 S3 nanorods
(width
55 nm, height
50 nm), (2) the selected TiO2
nanorods’ dimensions (width
72 nm, height
250 nm),
and (3) the larger TiO2 nanorods’ dimensions
(width
88 nm, height
300 nm). As it can be seen from
the first scenario [Figs. 7(a) and 7(d)], the light remains local-
Fig. 5. Multimode waveguide characterization. The dependence of ized in the Sb2 S3 nanoantennas (enlarged inset figures) due to
waveguide width on the (a) neff and (b) dispersion D of different SiN its higher refractive index in both phases. This biased localiza-
waveguide modes at λ
800 nm. The shaded region indicates the tion in the Sb2 S3 nanorods results in unacceptable performance
waveguide widths that support asymmetric multimode. The dashed
in the amorphous state because the light is routed to the un-
black lines show the maximal modal index (neff ) in the waveguide
and the waveguide width that support a single mode. The gray dotted desired output. After optimizing the dimensions of the TiO2
line indicates the width considered in the stem SiN waveguide. (c) neff nanorods [Figs. 7(b) and 7(e)], the localization of light changes
and D of the fundamental TE00 mode as a function of simulated wave- significantly between the Sb2 S3 ∕TiO2 nanorod arrays depend-
lengths. The inset figure shows the E y distribution of the TE00 mode ing on the phase of Sb2 S3 . Moreover, the produced self-imaged
at λ
800 nm. mode takes relatively confined paths in the output ports with
 2110 Vol. 9, No. 10 / October 2021 / Photonics Research Research Article

APPENDIX E: THE SUGGESTED FABRICATION

METHOD AND EXPERIMENTAL SETUP OF
RECONFIGURABLE METASURFACE-BASED
(1 × 2) WAVEGUIDE SWITCH
1. Suggested Fabrication Method
The fabrication of nanorod arrays on top of nanophotonic
waveguides was previously described by Yu et al. [39] and
Wang et al. [42]. However, the demonstration of the two nano-
rod arrays of two different materials patterned on a waveguide
has not yet been reported. Therefore, we believe that it is useful
to present the fabrication method for this particular design.
Figure 8 shows the suggested fabrication process, which can
be described as follows. A 3 μm film of silica (SiO2) is trans-
ferred onto a silicon substrate by plasma enhanced chemical
vapor deposition (PECVD). A 0.42 μm layer of stoichiometric
silicon nitride (SiN) is deposited onto the (SiO2 ∕Si) wafer us-
ing LPCVD. Energy dispersive X-ray spectroscopy (EDX) is
used to determine the thickness and composition of the SiN
film. The first step of electron beam lithography (EBL; positive
resist) is then conducted to pattern the nanorod arrays as well as
the alignment marks using JEOL 6300-FX at 250 pA [39].
After the development, a 55 nm film of amorphous Sb2S3 is
deposited using electron-beam deposition (EBD) and followed
by an overnight liftoff using a Microposit Remover 1165 at a
temperature of 80°C. The second round of positive resist is
Fig. 6. Parametric sweep of TiO2 nanorods for the TE00 mode at conducted to deposit a 250 nm film of TiO2. To define the SiN
λ
800 nm. (a) and (b) Simulated device cross talk considering var- waveguide, a second step of EBL (negative resist) is carried out.
iations of the height and width of TiO2 nanorods when Sb2 S3 is in the To pattern the output Y-branch waveguides, an aligned expo-
crystalline and amorphous state, respectively. (c) and (d) Calculated sure of a beam current (4 nA [39]) is used. After the develop-
transmission at the desired output as a function of height and ment, a chromium photomask is deposited onto the wafer via
width of TiO2 nanorods for the crystalline and amorphous state, re-
electron-beam evaporation and followed by an overnight liftoff
spectively. The black dashed lines indicate the dimensions used in our
metasurface.

high transmittance. Note that the further increase in the di-

mensions of TiO2 nanorods [Figs. 7(c) and 7(f )] leads to in-
creased scattering and reflection losses, coupled with altering
the position where the self-image is produced.

Fig. 7. Full-wave simulation showing the optical field intensity jEj2

in the switch for the fundamental TE mode in the xy plane at Fig. 8. Suggested fabrication method employing three steps of elec-
λ
800 nm and for different TiO2 dimensions for (a)–(c) c-Sb2 S3 tron beam lithography: (a) two positive resists followed by LPCVD to
and (d)–(f ) a-Sb2 S3 , The boundaries of the SiN waveguide and meta- transfer the desired patterns of the nanorod arrays onto the developed
surface structure are indicated by dashed lines and rectangles, respec- gaps and (b) a negative resist followed by reactive ion etching (RIE) to
tively. Inset: enlarged view of the field profile in the metasurface. define the SiN waveguide.
 Research Article Vol. 9, No. 10 / October 2021 / Photonics Research 2111

Fig. 9. Device fabrication tolerance. (a) and (b) Simulated device

cross talk for the fundamental mode operating at λ
800 nm, con-
sidering possible parallel misalignment (offset) and axial misalignment Fig. 10. Schematic illustration of the experimental setup suggested
(gap width), respectively. The gray dotted lines in the figures indicate for characterizing the performance of the proposed reconfigurable
the dimensions used in our device. switch. Here, PPG is a programmable pulse generator, PC is a power
controller, and M is a mirror.

using a Microposit Remover 1165 at a temperature of 75°C.

To eliminate the residual resist, an oxygen plasma is used at used to couple light from the fiber to the stem SiN waveguide
20°C for 15 s. The SiN waveguides (i.e., stem and Y branches) [14]). Convex lenses are used to collect the light that decouples
are etched through the SiN film via plasma reactive ion etching. from the output ports of the switch. The polarization controller
The residual mask is removed by wet etching [98]. is employed to ensure matching the fundamental quasi-TE
To estimate the device performance against fabrication mis- mode of the stem SiN waveguides [14,39,63,85]. Finally, a
alignment between the nanorod arrays in the first and second camera is used for imaging the light spots at the outputs of
lithography processes, we simulated the cross-talk response be- the switch. Note that if the switch is characterized by an
tween the output ports considering possible parallel misalign- off-chip optical fiber setup using focusing grating couplers at
ment (offset) [Fig. 9(a)] and axial misalignment (gap) the input and output ports of the switch, then the need for
[Fig. 9(b)]. Within the entire calculated dimensions consider- the convex lens is eliminated. And the camera is replaced by
a low-noise power meter to measure the power from the output
ing both parallel and axial misalignment, the device showed
ports [14,19].
high tolerance to possible nanorod-array misalignments, as
the simulated device cross talk showed minor changes in both
a-Sb2 S3 and c-Sb2 S3. Specifically, for the parallel misalignment, APPENDIX F: THE EFFECTIVE INDICES OF A
the simulated cross-talk value remained below −10.0 dB, and NANOROD-ARRAY-LOADED SiN WAVEGUIDE
for the axial misalignment, the cross talk remained below The effective refractive indices of each nanorod array in our
−8.4 dB, even for the worst case, when the nanorod arrays were metasurface were calculated using Rytov’s approximation
attached or highly separated from each other. These results con- [99] as follows.
firm the reliability and robustness of our device. The effective indices of Sb2 S3 nanorod arrays in the amor-
phous (nas∥ ) and crystalline (ncs∥ ) states can be given by
2. Suggested Experimental Setup  
Figure 10 illustrates the suggested experimental setup for real- W Sb2 S3 2 W Sb2 S3 2
n2as∥
na  1 − nbare , (F2)
izing the reversible optical switching of the Sb2 S3 and charac- Λ Λ
terizing the switch performance. For optically switching the  
phase Sb2 S3 , the pump source requires a pulsed diode laser with W Sb2 S3 2 W Sb2 S3 2
ncs∥

2
nc  1 − nbare , (F3)
power of 90 mW working at λ
630 nm. The laser pulses are Λ Λ
modulated electrically using a programmable pulse generator to where na and nc are the effective indices of a waveguide cross
produce pulses with varying durations of time. For the crystal- section with a-Sb2 S3 or c-Sb2 S3 nanorods, respectively, and
lization process, long pulses (100 ms) are used, whereas, for the nbare is the effective index of a bare SiN waveguide (without
amorphization process, shorter pulses (400 ns) are used [67]. A nanorods).
power controller (PC) composed of a polarizing beam splitter The effective index of a TiO2 nanorod array can be
and liquid-crystal retarder is employed to control the power of described as
each pulse [67]. The PC enables increasing the pulses’ repro-  
W TiO2 2 W TiO2 2
ducibility by keeping the diode at a constant current [67]. To n2t∥
n2  1 − nbare , (F4)
obtain the high fluence μm spot size required for the amorph- Λ Λ
ization process, an objective lens is used [67]. In contrast, the where n2 is the effective index of waveguide cross section with a
spot size is required to slightly be defocused to avoid damage TiO2 nanorod.
effects in the spot center during the crystallization process [67]. Figures 11(a)–11(c) schematically illustrate x-periodic nano-
The sample is fixed on a six-axis stage and monitored using antenna arrays patterned on a SiN waveguide. We started by
alignment microscopy to control the focus of the laser spot pre- calculating the effective indices of the bare SiN waveguide as
cisely [14,39,67]. For the probe source, a tunable near-visible well as the nanorod-loaded waveguide using the FDE solver
laser injects light to the input port of the switch by butt- (Lumerical Inc.) as shown in Fig. 11(d). Following that, we
coupling [39,42] (alternatively, grating couplers can also be employed Rytov’s formulas to calculate the effective indices
 2112 Vol. 9, No. 10 / October 2021 / Photonics Research Research Article

mode in both Sb2S3 phases. Figure 12(a–c) shows the electric

field component Ey for TE00 mode at the input and output
ports of the device. The arrows illustrate the corresponding vec-
tor diagrams of the electric fields of the modes at the input and
output ports.

Funding. CAS-TWAS Presidents Fellowship Program;

National Natural Science Foundation of China (62134009,
62121005); Innovation Grant of Changchun Institute of
Optics, Fine Mechanics and Physics (CIOMP); Jilin
Provincial Science and Technology Development Project
(YDZJ202102CXJD002); Development Program of the
Science and Technology of Jilin Province
(20200802001GH); Scientific Research Project of the
Chinese Academy of Sciences (QYZDB-SSW-SYS038).

Acknowledgment. A. A. acknowledges the CAS-TWAS

Presidents Fellowship Program. M. E. initiated the project and
conceived the approach. C. G., W. L., M. E., and J. C. super-
vised the project. A. A. designed the metasurface and performed
simulations. A. A., M. E., W. L., J. C., and G. V. analyzed the
data. C. S. contributed to the discussions. A. A. wrote the paper
with input from all the authors. All authors discussed the
Fig. 11. Sketches showing x-periodic Sb2 S3 and TiO2 nanoanten-
research.
nas patterned on a SiN waveguide in (a) xy plane and (b), (c) zy
plane, where na∕c are the effective indices of a-Sb2 S3 and c-Sb2 S3
nanorods, n2 is the effective index of the TiO2 nanorods, and nbare Disclosures. The authors declare no conflicts of interest.
is the effective index of bare SiN waveguide. (d) Effective refractive
indices for the waveguide cross section calculated using FDE. The in- Data Availability. The data that support the findings of
set shows the FDE simulation setup. (e) Effective indices for nanorod this study are available from the corresponding author upon
arrays calculated by Rytov’s approximation. reasonable request.

REFERENCES
of the periodic nanorod arrays. As shown in Fig. 11(e), the
1. B. J. Shastri, A. N. Tait, T. Ferreira de Lima, W. H. P. Pernice, H.
magnitude of the effective index of a TiO2 nanorod array is Bhaskaran, C. D. Wright, and P. R. Prucnal, “Photonics for artificial
larger than that of a-Sb2 S3 after optimizing their dimensions intelligence and neuromorphic computing,” Nat. Photonics 15,
as discussed in the previous section. 102–114 (2021).
2. Z. Chen and M. Segev, “Highlighting photonics: looking into the next
decade,” eLight 1, 2 (2021).
APPENDIX G: MODAL PROFILES AT THE 3. G. Wetzstein, A. Ozcan, S. Gigan, S. Fan, D. Englund, M. Soljačić, C.
DEVICE INPUT AND OUTPUT PORTS Denz, D. A. B. Miller, and D. Psaltis, “Inference in artificial intelligence
with deep optics and photonics,” Nature 588, 39–47 (2020).
It is important to note that the produced modes at the output 4. W. Bogaerts, D. Pérez, J. Capmany, D. A. B. Miller, J. Poon, D.
ports of the device maintain the polarization of the input TE00 Englund, F. Morichetti, and A. Melloni, “Programmable photonic cir-
cuits,” Nature 586, 207–216 (2020).
5. J. Feldmann, N. Youngblood, C. D. Wright, H. Bhaskaran, and
W. H. P. Pernice, “All-optical spiking neurosynaptic networks with
self-learning capabilities,” Nature 569, 208–214 (2019).
6. J. Torrejon, M. Riou, F. A. Araujo, S. Tsunegi, G. Khalsa, D. Querlioz,
P. Bortolotti, V. Cros, K. Yakushiji, A. Fukushima, H. Kubota, S.
Yuasa, M. D. Stiles, and J. Grollier, “Neuromorphic computing with
nanoscale spintronic oscillators,” Nature 547, 428–431 (2017).
7. J. Wang, F. Sciarrino, A. Laing, and M. G. Thompson, “Integrated pho-
tonic quantum technologies,” Nat. Photonics 14, 273–284 (2020).
8. A. Blais, S. M. Girvin, and W. D. Oliver, “Quantum information
processing and quantum optics with circuit quantum electrodynam-
Fig. 12. Modes in the input and output ports of SiN waveguides. ics,” Nat. Phys. 16, 247–256 (2020).
9. S. Leedumrongwatthanakun, L. Innocenti, H. Defienne, T. Juffmann,
Simulated Ey component at λ
800 nm for the TE00 mode in the
A. Ferraro, M. Paternostro, and S. Gigan, “Programmable linear quan-
(a) input stem SiN waveguide before interacting with the metasurface, tum networks with a multimode fibre,” Nat. Photonics 14, 139–142
(b) output Port3 for a-Sb2 S3 and (c) output Port2 for c-Sb2 S3. The (2020).
arrows indicate the vector diagrams of the electric field component 10. Q. Ma, L. Chen, H. B. Jing, Q. R. Hong, H. Y. Cui, Y. Liu, L. Li, and T. J.
of the modes. The dashed black lines show the boundaries of the Cui, “Controllable and programmable nonreciprocity based on detach-
SiN waveguides. able digital coding metasurface,” Adv. Opt. Mater. 7, 1901285 (2019).
 Research Article Vol. 9, No. 10 / October 2021 / Photonics Research 2113

11. D. Marpaung, J. Yao, and J. Capmany, “Integrated microwave pho- 30. D. Loke, T. H. Lee, W. J. Wang, L. P. Shi, R. Zhao, Y. C. Yeo, T. C.
tonics,” Nat. Photonics 13, 80–90 (2019). Chong, and S. R. Elliott, “Breaking the speed limits of phase-change
12. Y. Wang, W. Li, M. Li, S. Zhao, F. De Ferrari, M. Liscidini, and F. G. memory,” Science 336, 1566 (2012).
Omenetto, “Biomaterial-based ‘structured opals’ with programmable 31. C. R. de Galarreta, A. M. Alexeev, Y.-Y. Au, M. Lopez-Garcia, M.
combination of diffractive optical elements and photonic bandgap ef- Klemm, M. Cryan, J. Bertolotti, and C. D. Wright, “Nonvolatile recon-
fects,” Adv. Mater. 31, 1805312 (2019). figurable phase-change metadevices for beam steering in the near
13. Z. Li, J. Zou, H. Zhu, B. T. T. Nguyen, Y. Shi, P. Y. Liu, R. C. Bailey, J. infrared,” Adv. Funct. Mater. 28, 1704993 (2018).
Zhou, H. Wang, Z. Yang, Y. Jin, P. H. Yap, H. Cai, Y. Hao, and A. Q. 32. C. Rios, P. Hosseini, C. D. Wright, H. Bhaskaran, and W. H. P.
Liu, “Biotoxoid photonic sensors with temperature insensitivity using a Pernice, “On-chip photonic memory elements employing phase-
cascade of ring resonator and Mach–Zehnder interferometer,” ACS change materials,” Adv. Mater. 26, 1372–1377 (2014).
Sens. 5, 2448–2456 (2020). 33. Z. Cheng, C. Ríos, N. Youngblood, C. D. Wright, W. H. P. Pernice, and
14. P. Xu, J. Zheng, J. K. Doylend, and A. Majumdar, “Low-loss and H. Bhaskaran, “Device-level photonic memories and logic applica-
broadband nonvolatile phase-change directional coupler switches,” tions using phase-change materials,” Adv. Mater. 30, 1802435 (2018).
ACS Photon. 6, 553–557 (2019). 34. M. A. Kats, D. Sharma, J. Lin, P. Genevet, R. Blanchard, Z. Yang,
15. M. Rudé, J. Pello, R. E. Simpson, J. Osmond, G. Roelkens, J. J. G. M. M. M. Qazilbash, D. N. Basov, S. Ramanathan, and F. Capasso,
van der Tol, and V. Pruneri, “Optical switching at 1.55 μm in silicon “Ultra-thin perfect absorber employing a tunable phase change
racetrack resonators using phase change materials,” Appl. Phys. material,” Appl. Phys. Lett. 101, 221101 (2012).
Lett. 103, 141119 (2013). 35. H. Liang, R. Soref, J. Mu, A. Majumdar, X. Li, and W.-P. Huang,
16. M. Stegmaier, C. Ríos, H. Bhaskaran, C. D. Wright, and W. H. P. “Simulations of silicon-on-insulator channel-waveguide electrooptical
Pernice, “Nonvolatile all-optical 1 × 2 switch for chipscale photonic 2 × 2 switches and 1 × 1 modulators using a Ge2Sb2Te5 self-holding
networks,” Adv. Opt. Mater. 5, 1600346 (2017). layer,” J. Lightwave Technol. 33, 1805–1813 (2015).
17. J. Zheng, A. Khanolkar, P. Xu, S. Colburn, S. Deshmukh, J. Myers, J. 36. J. Feldmann, M. Stegmaier, N. Gruhler, C. Ríos, H. Bhaskaran, C. D.
Frantz, E. Pop, J. Hendrickson, J. Doylend, N. Boechler, and A. Wright, and W. H. P. Pernice, “Calculating with light using a chip-scale
Majumdar, “GST-on-silicon hybrid nanophotonic integrated circuits: all-optical abacus,” Nat. Commun. 8, 1256 (2017).
a non-volatile quasi-continuously reprogrammable platform,” Opt. 37. M. Xu, X. Mai, J. Lin, W. Zhang, Y. Li, Y. He, H. Tong, X. Hou, P. Zhou,
Mater. Express 8, 1551–1561 (2018). and X. Miao, “Recent advances on neuromorphic devices based on
18. Y. Zhang, J. B. Chou, J. Li, H. Li, Q. Du, A. Yadav, S. Zhou, M. Y. chalcogenide phase-change materials,” Adv. Funct. Mater. 30,
Shalaginov, Z. Fang, H. Zhong, C. Roberts, P. Robinson, B. 2003419 (2020).
Bohlin, C. Ríos, H. Lin, M. Kang, T. Gu, J. Warner, V. Liberman, 38. A. H. Dorrah, N. A. Rubin, A. Zaidi, M. Tamagnone, and F. Capasso,
K. Richardson, and J. Hu, “Broadband transparent optical phase “Metasurface optics for on-demand polarization transformations along
change materials for high-performance nonvolatile photonics,” Nat. the optical path,” Nat. Photonics 15, 287–296 (2021).
Commun. 10, 4279 (2019). 39. Z. Li, M.-H. Kim, C. Wang, Z. Han, S. Shrestha, A. C. Overvig, M. Lu,
19. C. Wu, H. Yu, H. Li, X. Zhang, I. Takeuchi, and M. Li, “Low-loss in- A. Stein, A. M. Agarwal, M. Lončar, and N. Yu, “Controlling propaga-
tegrated photonic switch using subwavelength patterned phase tion and coupling of waveguide modes using phase-gradient metasur-
change material,” ACS Photon. 6, 87–92 (2019). faces,” Nat. Nanotechnol. 12, 675–683 (2017).
20. C. Zhang, M. Zhang, Y. Xie, Y. Shi, R. Kumar, R. R. Panepucci, and D. 40. B. Wang, S. Blaize, and R. Salas-Montiel, “Nanoscale plasmonic TM-
Dai, “Wavelength-selective 2 × 2 optical switch based on a pass polarizer integrated on silicon photonics,” Nanoscale 11, 20685–
Ge2Sb2Te5-assisted microring,” Photon. Res. 8, 1171–1176 20692 (2019).
(2020). 41. A. Alquliah, M. Elkabbash, J. Zhang, J. Cheng, and C. Guo,
21. T. J. Seok, N. Quack, S. Han, R. S. Muller, and M. C. Wu, “Large-scale “Ultrabroadband, compact, polarization independent and efficient
broadband digital silicon photonic switches with vertical adiabatic cou- metasurface-based power splitter on lithium niobate waveguides,”
plers,” Optica 3, 64–70 (2016). Opt. Express 29, 8160–8170 (2021).
22. T. J. Seok, J. Luo, Z. Huang, K. Kwon, J. Henriksson, J. Jacobs, L. 42. C. Wang, Z. Li, M.-H. Kim, X. Xiong, X.-F. Ren, G.-C. Guo, N. Yu, and
Ochikubo, R. S. Muller, and M. C. Wu, “Silicon photonic wavelength M. Lončar, “Metasurface-assisted phase-matching-free second har-
cross-connect with integrated MEMS switching,” APL Photon. 4, monic generation in lithium niobate waveguides,” Nat. Commun. 8,
100803 (2019). 2098 (2017).
23. J. Zheng, S. Zhu, P. Xu, S. Dunham, and A. Majumdar, “Modeling 43. V. Ginis, M. Piccardo, M. Tamagnone, J. Lu, M. Qiu, S. Kheifets, and
electrical switching of nonvolatile phase-change integrated nanopho- F. Capasso, “Remote structuring of near-field landscapes,” Science
tonic structures with graphene heaters,” ACS Appl. Mater. Interfaces 369, 436–440 (2020).
12, 21827–21836 (2020). 44. X. Guo, Y. Ding, X. Chen, Y. Duan, and X. Ni, “Molding free-space
24. M. Ono, M. Hata, M. Tsunekawa, K. Nozaki, H. Sumikura, H. Chiba, light with guided wave–driven metasurfaces,” Sci. Adv. 6, eabb4142
and M. Notomi, “Ultrafast and energy-efficient all-optical switching (2020).
with graphene-loaded deep-subwavelength plasmonic waveguides,” 45. R. Guo, M. Decker, F. Setzpfandt, X. Gai, D.-Y. Choi, R. Kiselev, A.
Nat. Photonics 14, 37–43 (2020). Chipouline, I. Staude, T. Pertsch, D. N. Neshev, and Y. S. Kivshar,
25. M. Thomaschewski, V. A. Zenin, C. Wolff, and S. I. Bozhevolnyi, “High–bit rate ultra-compact light routing with mode-selective on-chip
“Plasmonic monolithic lithium niobate directional coupler switches,” nanoantennas,” Sci. Adv. 3, e1700007 (2017).
Nat. Commun. 11, 748 (2020). 46. S. Sun, Q. He, S. Xiao, Q. Xu, X. Li, and L. Zhou, “Gradient-index
26. D. Pérez, I. Gasulla, P. Das Mahapatra, and J. Capmany, “Principles, meta-surfaces as a bridge linking propagating waves and surface
fundamentals, and applications of programmable integrated photon- waves,” Nat. Mater. 11, 426–431 (2012).
ics,” Adv. Opt. Photon. 12, 709–786 (2020). 47. A. M. Shaltout, V. M. Shalaev, and M. L. Brongersma, “Spatiotemporal
27. M. Wuttig, H. Bhaskaran, and T. Taubner, “Phase-change materials light control with active metasurfaces,” Science 364, eaat3100 (2019).
for non-volatile photonic applications,” Nat. Photonics 11, 465–476 48. O. Tsilipakos, A. C. Tasolamprou, A. Pitilakis, F. Liu, X. Wang, M. S.
(2017). Mirmoosa, D. C. Tzarouchis, S. Abadal, H. Taghvaee, C. Liaskos, A.
28. S. Abdollahramezani, O. Hemmatyar, H. Taghinejad, A. Krasnok, Y. Tsioliaridou, J. Georgiou, A. Cabellos-Aparicio, E. Alarcón, S.
Kiarashinejad, M. Zandehshahvar, A. Alù, and A. Adibi, “Tunable Ioannidis, A. Pitsillides, I. F. Akyildiz, N. V. Kantartzis, E. N.
nanophotonics enabled by chalcogenide phase-change materials,” Economou, C. M. Soukoulis, M. Kafesaki, and S. Tretyakov,
Nanophotonics 9, 1189–1241 (2020). “Toward intelligent metasurfaces: the progress from globally tunable
29. N. Farmakidis, N. Youngblood, X. Li, J. Tan, J. L. Swett, Z. Cheng, metasurfaces to software-defined metasurfaces with an embedded
C. D. Wright, W. H. P. Pernice, and H. Bhaskaran, “Plasmonic network of controllers,” Adv. Opt. Mater. 8, 2000783 (2020).
nanogap enhanced phase-change devices with dual electrical-optical 49. F. Ding, Y. Yang, and S. I. Bozhevolnyi, “Dynamic metasurfaces using
functionality,” Sci. Adv. 5, eaaw2687 (2019). phase-change chalcogenides,” Adv. Opt. Mater. 7, 1801709 (2019).
 2114 Vol. 9, No. 10 / October 2021 / Photonics Research Research Article

50. Q. Wang, E. T. F. Rogers, B. Gholipour, C.-M. Wang, G. Yuan, J. 70. K. R. Safronov, D. N. Gulkin, I. M. Antropov, K. A. Abrashitova, V. O.
Teng, and N. I. Zheludev, “Optically reconfigurable metasurfaces Bessonov, and A. A. Fedyanin, “Multimode interference of Bloch sur-
and photonic devices based on phase change materials,” Nat. face electromagnetic waves,” ACS Nano 14, 10428–10437 (2020).
Photonics 10, 60–65 (2016). 71. P. Sethi, A. Haldar, and S. K. Selvaraja, “Ultra-compact low-loss
51. S. G.-C. Carrillo, L. Trimby, Y.-Y. Au, V. K. Nagareddy, G. Rodriguez- broadband waveguide taper in silicon-on-insulator,” Opt. Express
Hernandez, P. Hosseini, C. Ríos, H. Bhaskaran, and C. D. Wright, 25, 10196–10203 (2017).
“A nonvolatile phase-change metamaterial color display,” Adv. Opt. 72. Y. Fu, T. Ye, W. Tang, and T. Chu, “Efficient adiabatic silicon-on-
Mater. 7, 1801782 (2019). insulator waveguide taper,” Photon. Res. 2, A41–A44 (2014).
52. Z. Zhu, P. G. Evans, R. F. Haglund, and J. G. Valentine, “Dynamically 73. J. Zhang, J. Yang, H. Xin, J. Huang, D. Chen, and Z. Zhaojian,
reconfigurable metadevice employing nanostructured phase-change “Ultrashort and efficient adiabatic waveguide taper based on thin flat
materials,” Nano Lett. 17, 4881–4885 (2017). focusing lenses,” Opt. Express 25, 19894–19903 (2017).
53. C. Ruiz de Galarreta, I. Sinev, A. M. Alexeev, P. Trofimov, K. 74. C. Sun, Y. Yu, and X. Zhang, “Ultra-compact waveguide crossing for a
Ladutenko, S. G.-C. Carrillo, E. Gemo, A. Baldycheva, J. Bertolotti, mode-division multiplexing optical network,” Opt. Lett. 42, 4913–4916
and C. D. Wright, “Reconfigurable multilevel control of hybrid all- (2017).
dielectric phase-change metasurfaces,” Optica 7, 476–484 (2020). 75. Y. Zhang, Y. He, H. Wang, L. Sun, and Y. Su, “Ultra-broadband mode
54. T. J. Cui, M. Q. Qi, X. Wan, J. Zhao, and Q. Cheng, “Coding meta- size converter using on-chip metamaterial-based Luneburg lens,”
materials, digital metamaterials and programmable metamaterials,” ACS Photon. 8, 202–208 (2021).
Light Sci. Appl. 3, e218 (2014). 76. J. M. Luque-González, R. Halir, J. G. Wangüemert-Pérez, J. de-Oliva-
55. L. Zhang, M. Z. Chen, W. Tang, J. Y. Dai, L. Miao, X. Y. Zhou, S. Jin, Rubio, J. H. Schmid, P. Cheben, Í. Molina-Fernández, and A. Ortega-
Q. Cheng, and T. J. Cui, “A wireless communication scheme based on Moñux, “An ultracompact GRIN-lens-based spot size converter using
space- and frequency-division multiplexing using digital metasurfa- subwavelength grating metamaterials,” Laser Photon. Rev. 13,
ces,” Nat. Electron. 4, 218–227 (2021). 1900172 (2019).
56. L. Li, Y. Shuang, Q. Ma, H. Li, H. Zhao, M. Wei, C. Liu, C. Hao, C.-W. 77. C. Yao, S. C. Singh, M. ElKabbash, J. Zhang, H. Lu, and C. Guo,
Qiu, and T. J. Cui, “Intelligent metasurface imager and recognizer,” “Quasi-rhombus metasurfaces as multimode interference couplers
Light Sci. Appl. 8, 97 (2019). for controlling the propagation of modes in dielectric-loaded wave-
57. R.-B. Hwang, “Binary meta-hologram for a reconfigurable holographic guides,” Opt. Lett. 44, 1654–1657 (2019).
metamaterial antenna,” Sci. Rep. 10, 8586 (2020). 78. E. D. Palik, Handbook of Optical Constants of Solids (Academic,
58. C. Liu, W. M. Yu, Q. Ma, L. Li, and T. J. Cui, “Intelligent coding meta- 2012).
surface holograms by physics-assisted unsupervised generative ad- 79. M. Delaney, I. Zeimpekis, D. Lawson, D. Hewak, and O. Muskens,
versarial network,” Photon. Res. 9, B159–B167 (2021). “A new family of ultra-low loss reversible phase change materials for
59. J. Xiong and S.-T. Wu, “Planar liquid crystal polarization optics for photonic integrated circuits: Sb2S3 and Sb2Se3,” Adv. Funct. Mater. 30,
augmented reality and virtual reality: from fundamentals to applica- 2002447 (2020).
tions,” eLight 1, 3 (2021). 80. M. M. R. Elsawy, S. Lanteri, R. Duvigneau, J. A. Fan, and P. Genevet,
60. X. G. Zhang, W. X. Jiang, H. L. Jiang, Q. Wang, H. W. Tian, L. Bai, “Numerical optimization methods for metasurfaces,” Laser Photon.
Z. J. Luo, S. Sun, Y. Luo, C.-W. Qiu, and T. J. Cui, “An optically driven Rev. 14, 1900445 (2020).
digital metasurface for programming electromagnetic functions,” Nat. 81. K. Koshelev and Y. Kivshar, “Dielectric resonant metaphotonics,”
Electron. 3, 165–171 (2020). ACS Photon. 8, 102–112 (2021).
61. Q. Zhang, Y. Zhang, J. Li, R. Soref, T. Gu, and J. Hu, “Broadband 82. I. Staude, T. Pertsch, and Y. S. Kivshar, “All-dielectric resonant meta-
nonvolatile photonic switching based on optical phase change mate- optics lightens up,” ACS Photon. 6, 802–814 (2019).
rials: beyond the classical figure-of-merit,” Opt. Lett. 43, 94–97 (2018). 83. R. Halir, P. J. Bock, P. Cheben, A. Ortega-Moñux, C. Alonso-Ramos,
62. H. Hu, H. Zhang, L. Zhou, J. Xu, L. Lu, J. Chen, and B. M. A. Rahman, J. H. Schmid, J. Lapointe, D.-X. Xu, J. G. Wangüemert-Pérez, Í.
“Contra-directional switching enabled by Si-GST grating,” Opt. Molina-Fernández, and S. Janz, “Waveguide sub-wavelength struc-
Express 28, 1574–1584 (2020). tures: a review of principles and applications,” Laser Photon. Rev.
63. Z. Fang, J. Zheng, A. Saxena, J. Whitehead, Y. Chen, and A. 9, 25–49 (2015).
Majumdar, “Non-volatile reconfigurable integrated photonics enabled 84. P. Cheben, R. Halir, J. H. Schmid, H. A. Atwater, and D. R. Smith,
by broadband low-loss phase change material,” Adv. Opt. Mater. 9, “Subwavelength integrated photonics,” Nature 560, 565–572
2002049 (2021). (2018).
64. X. Li, N. Youngblood, Z. Cheng, S. G.-C. Carrillo, E. Gemo, W. H. P. 85. J. Zheng, Z. Fang, C. Wu, S. Zhu, P. Xu, J. K. Doylend, S. Deshmukh,
Pernice, C. D. Wright, and H. Bhaskaran, “Experimental investigation E. Pop, S. Dunham, M. Li, and A. Majumdar, “Nonvolatile electrically
of silicon and silicon nitride platforms for phase-change photonic in- reconfigurable integrated photonic switch enabled by a silicon PIN di-
memory computing,” Optica 7, 218–225 (2020). ode heater,” Adv. Mater. 32, 2001218 (2020).
65. W. Dong, H. Liu, J. K. Behera, L. Lu, R. J. H. Ng, K. V. Sreekanth, X. 86. Y. Zhang, C. Ríos, M. Y. Shalaginov, M. Li, A. Majumdar, T. Gu, and J.
Zhou, J. K. W. Yang, and R. E. Simpson, “Wide bandgap phase Hu, “Myths and truths about optical phase change materials: a per-
change material tuned visible photonics,” Adv. Funct. Mater. 29, spective,” Appl. Phys. Lett. 118, 210501 (2021).
1806181 (2019). 87. K. Shportko, S. Kremers, M. Woda, D. Lencer, J. Robertson, and M.
66. K. V. Sreekanth, Q. Ouyang, S. Sreejith, S. Zeng, W. Lishu, E. Ilker, Wuttig, “Resonant bonding in crystalline phase-change materials,”
W. Dong, M. ElKabbash, Y. Ting, C. T. Lim, M. Hinczewski, G. Strangi, Nat. Mater. 7, 653–658 (2008).
K.-T. Yong, R. E. Simpson, and R. Singh, “Phase-change-material- 88. Y. Wang, P. Landreman, D. Schoen, K. Okabe, A. Marshall, U.
based low-loss visible-frequency hyperbolic metamaterials for Celano, H. S. P. Wong, J. Park, and M. L. Brongersma, “Electrical
ultrasensitive label-free biosensing,” Adv. Opt. Mater. 7, 1900081 tuning of phase-change antennas and metasurfaces,” Nat.
(2019). Nanotechnol. 16, 667–672 (2021).
67. M. Delaney, I. Zeimpekis, D. Lawson, D. W. Hewak, and O. L. 89. Y. Zhang, C. Fowler, J. Liang, B. Azhar, M. Y. Shalaginov, S. Deckoff-
Muskens, “A new family of ultralow loss reversible phase-change ma- Jones, S. An, J. B. Chou, C. M. Roberts, V. Liberman, M. Kang, C.
terials for photonic integrated circuits: Sb2S3 and Sb2Se3,” Adv. Funct. Ríos, K. A. Richardson, C. Rivero-Baleine, T. Gu, H. Zhang, and J.
Mater. 30, 2002447 (2020). Hu, “Electrically reconfigurable non-volatile metasurface using low-
68. W. Zhu, R. Yang, G. Geng, Y. Fan, X. Guo, P. Li, Q. Fu, F. Zhang, C. loss optical phase-change material,” Nat. Nanotechnol. 16, 661–666
Gu, and J. Li, “Titanium dioxide metasurface manipulating high- (2021).
efficiency and broadband photonic spin Hall effect in visible regime,” 90. T. Akiyama, M. Uno, H. Kitaura, K. Narumi, R. Kojima, K. Nishiuchi,
Nanophotonics 9, 4327–4335 (2020). and N. Yamada, “Rewritable dual-layer phase-change optical disk
69. M. Wuttig and N. Yamada, “Phase-change materials for rewriteable utilizing a blue-violet laser,” Jpn. J. Appl. Phys. 40, 1598–1603
data storage,” Nat. Mater. 6, 824–832 (2007). (2001).
 Research Article Vol. 9, No. 10 / October 2021 / Photonics Research 2115

91. W. Zhang, R. Mazzarello, M. Wuttig, and E. Ma, “Designing crystal- 95. H. M. Mbonde, H. C. Frankis, and J. D. B. Bradley, “Enhanced non-
lization in phase-change materials for universal memory and neuro- linearity and engineered anomalous dispersion in TeO2-coated Si3N4
inspired computing,” Nat. Rev. Mater. 4, 150–168 (2019). waveguides,” IEEE Photon. J. 12, 2200210 (2020).
92. C. R. De Galarreta Fanjul, “Reconfigurable phase-change optical 96. H. El Dirani, L. Youssef, C. Petit-Etienne, S. Kerdiles, P. Grosse, C.
metasurfaces: novel design concepts to practicable devices,” Ph.D. Monat, E. Pargon, and C. Sciancalepore, “Ultralow-loss tightly
thesis (University of Exeter, 2020). confining Si3N4 waveguides and high-Q microresonators,” Opt.
93. P. Trofimov, A. P. Pushkarev, I. S. Sinev, V. V. Fedorov, S. Bruyère, Express 27, 30726–30740 (2019).
A. Bolshakov, I. S. Mukhin, and S. V. Makarov, “Perovskite–gallium 97. R. R. Grote and L. C. Bassett, “Single-mode optical waveguides on na-
phosphide platform for reconfigurable visible-light nanophotonic chip,” tive high-refractive-index substrates,” APL Photon. 1, 071302 (2016).
ACS Nano 14, 8126–8134 (2020). 98. K. Bi, Q. Wang, J. Xu, L. Chen, C. Lan, and M. Lei, “All-dielectric meta-
94. B. Desiatov, A. Shams-Ansari, M. Zhang, C. Wang, and M. Lončar, material fabrication techniques,” Adv. Opt. Mater. 9, 2001474 (2021).
“Ultra-low-loss integrated visible photonics using thin-film lithium nio- 99. S. Rytov, “Electromagnetic properties of a finely stratified medium,”
bate,” Optica 6, 380–384 (2019). J. Exp. Theor. Phys. 2, 466–475 (1956).