In the NOISE-Lab, we are building ultra low-power nanoscale optoelectronic devices by engineering the light-matter interaction. To that end, we design and fabricate various nano-photonic structures, including photonic crystal resnators, ring resonators and dielectric metasurfaces. On top of these nanophotonic structures we integrate new materials, including 2D materials, solution processed quantum materials (including quantum dots and nanoplatelets), and phase change materials. With this hybrid platform, in our lab we are currently pursuing three different project.

Few of the talks on our research by Arka might be useful to see to get a better idea:
Silicon Photonics for Optical Computing (2014)
Dielectric Metasurface (2019)
Metaphotonic Computational Image Sensor (March, 2020)
Hybrid Integrated Photonics (2020)
Metaphotonic Computational Image Sensor (June, 2020)
Software Defined Optics (November, 2021)
Here are talks given by group members:
Design and Optimization of Dielectric Metasurfaces (by Alan Zhan, MSR, 2016)


MCIS: Metaphotonic Computational Image Sensors

With increase in wearable technology, implantable bio-sensors, Internet of things, and in this effort to make everything smart, one needs a lot of sensors, which needs to be compact, low power, and also intelligent to reduce the subsequent data processing. In our research, we are looking into this problem, by using nanophotonics. The compactness of the sensor demands to have integrated photonics to be used. Unfortunately, with small size, the performance of the sensor goes down. Hence we are researching to supplement the sensors with computing, to gain back the performance. We are mainly interetsed in revolutinizing the field of cameras, spectrometers and microscopes.

NEURON: NEuromorphic Universally Reconfigurable Optical Network

Photons can pass each other without interacting with each other. That provides an attractive way to create large-scale optical network exploiting free-space geometry. For example a simple display with a million pixels can be placed in front of a CMOS camera with a million detectors, and just by introducing a lens in between them we can create few billion channels. This inherent parallellism of light, which made optics an attractive candidate at the first place for optical information processing, is lost when we go for integrated photonics, which relies on waveguides, or photonic wires. Using metasurface as lenses, and DBR as mirrors, we can create a monolithic geomtery, where we can take advantage of both free space (inherent parallellism) and integrated photonics (small size, robustness to misalignment). Using these devices, along with emerging nonlinear and tunable materials, we can create an optical resonator, which can process images, and can also potentially implement a monolithic optical neuron.

PHOENIQS: PHOtonics-Enabled Noisy Intermediate-Scale Quantum System

The project aims to realize an nanophotonic array of coupled nonlinear cavities, where the quantum materials enable nonlinearity at the few-photon level. The resulting quantum photonic platform is impossible to simulate in any classical computer and will allow synthesis of novel quantum states of light. The materials we are exploring for this work are 2D materials and their heterostructures, colloidal quantum dots, nanoplatelets and atomic vapor.

The previous research direction of the group (before 2020) are below:

Hybrid silicon (compatible) photonics(HySiP)

To improve the transceivers in current silicon photonics (SiP), we are looking into new materials, cavities and new modulation techniques. The current SiP devices are limited either by the large size of the devices, and hence large power and low speed (in MZI); or by high Q-resonators (thermal stabilization necessitates large power consumption; and photon lifetime reduces speed). We are exploring nanophotonic innovation to solve this three dimensional optimization problem (speed, power and size). Our approach is to explore a hybrid silicon photonic platform, where the underlying photonic devices are made of silicon, on top of which we will integrate new materials (like electro-optic oxides, polymers etc.). We, however, want to go beyond signal communication, and want to explore the avenues of optical computing. For that we are actively working on new nonlinear optical materials. We want to push the energy of these devices to few photon levels, where we can also study quantum optical effects. These devices can be thought of as precursors to future quantum information processing devices.

Self electro-optic devices for optical computing

Using a photo-absorbing material in a silicon ring resonator, we have proposed a platform, where an optoelectronic feedback could be easily implemented. This device provides a way to have optical bistability, without explicitly relying on any optical nonlinearity. Moreover, this device is shown to satisfy all th criteria, an optical swicth should have to build a scalable digital optical computing system. In our current research, we are looking into new materials, that can be integrated on top of silicon photonics, and can absorb light. Then we will build the optoelectronic feedback.
Funding Sources: AFOSR (YIP-Program); AFOSR (SBIR-CFDRC)

2D-material nanophotonics

We are actively collaborating with leading researchers in the 2D material community to build photonic devices using 2D materials. 2D materials are a newly discovered materials, which are monolayer and single-atom thick. Due to such low volume, the energy required to change this material can be very low. Moreover, these materials can be easily transferred to other materials. In our reseach, we are looking into building new light source, electro-optic modulator as well as strongly nonlinear optical devices using the 2D materials. These materials will be integrated on a large-scale silicon nitride photonic integrated circuits to build active devices. The choice of silicon nitride is motivated by their low loss, and lack of two-photon absorption. Unfortunately, they are essentially dielectric, and hence no active devices can be fabricated. This is where 2D materials will play a very important role.
Funding Sources: NSF-EFRI-2DARE; AFOSR (YIP-Program); NSF-EPMD; NSF-MRSEC

Single photon nonlinear optics

We are exploring ways to realize single photon nonlinear optics in a scalable fashion. For this we are optimizing various multi-modal cavities to enhance the effective nonlinerity, as well as exploring strongly nonlinear material, including 2D materials, and organic materials. Another research direction is to explore polariton-polariton interaction to reach few photon nonlinear devices.
Funding Sources: NSF-EFRI-ACQUIRE; AFOSR (YIP-Program); NSF-MRSEC

Tunable photonics

As part of the core HySiP interest, we are exploring various different materials, including complex oxides, polymer and phase change materials to build tunable and reconfigurable optical devices. This work also connects the other research interest of iCOS (tunable dielectric metasurface).
Funding Sources: NSF-EFRI-ACQUIRE; Intel-SRC

Intelligent compact optical sensor (iCOS)

With increase in wearable technology, Internet of things, and in this effort to make everything smart, one needs a lot of sensors, which needs to be compact, low power, and also intelligent to reduce the subsequent data processing. In our research, we are looking into this problem, by using nanophotonics. The compactness of the sensor demands to have integrated photonics to be used. Unfortunately, with small size, the performance of the sensor goes down. Hence we are researching to supplement the sensors with computing, to gain back the performance. We are mainly interetsed in two type of sensors: image sensors and spectrometer.

Tunable dielectric metasurface

Metasurfaces are two-dimensional quasi-periodic array of subwavelength features. Dielectric metasurfaces allow wavefront shaping of the incident light. However, the true potential of such metasurface can be realized, if one can tune them. We are looking into new materials with tunable refractive index to achieve this goal, or using flexible substrates to mechanically tune the metasurfaces. The goal is to build fast (~10's of MHz) sub-wavelength spatial light modulators.
Funding Sources: UW-RRF; Samsung-GRO

Cell-size Optical Microscope

Using the metasurface technology, we plan to build untra-compact, implantable optical microscopes. Using such microscopes,we can implant several of them inside a mouse brain, and image several parts of the brain simultaneously, thus truly enabling large scale brain imaging. Details of the project can be found here (PDF).

Freeform optics

Using metasurfaces, we are designing and building freeform optics, where we realize higher order polynominal surfaces, such as cubic surfaces. Details of the project can be found here. These freeform metasurfaces can be useful for various applications, including tunable eyewears, augmented reality visors, laser beam shaper and retroreflectors. Going beyond a single metasurface, we are exploring stacked meteasurfaces to create a simple volume optics.
Funding Sources: Tunoptix Inc.

Monolithic photonics for nonlinear image processing and optical neural network

Photons can pass each other without interacting with each other. That provides an attractive way to create large-scale optical network exploiting free-space geometry. For example a simple display with a million pixels can be placed in front of a CMOS camera with a million detectors, and just by introducing a lens in between them we can create few billion channels. This inherent parallellism of light, which made optics an attractive candidate at the first place for optical information processing, is lost when we go for integrated photonics, which relies on waveguides, or photonic wires. Using metasurface as lenses, and DBR as mirrors, we can create a monolithic geomtery, where we can take advantage of both free space (inherent parallellism) and integrated photonics (small size, robustness to misalignment). Using these devices, along with emerging nonlinear and tunable materials, we can create an optical resonator, which can process images, and can also potentially implement a monolithic optical neuron.

Research Facilities

  • Optics: We have optical characterization facilities in the wavelength range 400-1600nm, by using Fianium supercontinuum source, Priceton Instruments spectrometers (visible+IR), Santec CW laser (1425-1565nm) and spatial light modulators. We also have the capability to characterize photonic devices using grating-coupler in a fiber-in fiber-out setup. Addtionally, we have numeorus light sources in the visible frequency range to characterize metasurfaces and imaging optics.
  • Fabrication: We have access to state-of-the-art electron beam lithography, etching and deposition tools. We also have a 2D materials and cavity trasnfer station in our lab.
  • Computation: We have access to many electromagnetic solvers including, Lumerical FDTD, and have a high-end server for faster computation.