The growing maturity of nanofabrication has ushered massive sophisticated optical structures available on a photonic chip. The integration of subwavelength-structured metasurfaces and metamaterials on the canonical building block of optical waveguides is gradually reshaping the landscape of photonic integrated circuits, giving rise to numerous meta-waveguides with unprecedented strength in controlling guided electromagnetic waves. Here, we review recent advances in meta-structured waveguides that synergize various functional subwavelength photonic architectures with diverse waveguide platforms, such as dielectric or plasmonic waveguides and optical fibers. Foundational results and representative applications are comprehensively summarized. Brief physical models with explicit design tutorials, either physical intuition-based design methods or computer algorithms-based inverse designs, are cataloged as well. We highlight how meta-optics can infuse new degrees of freedom to waveguide-based devices and systems, by enhancing light-matter interaction strength to drastically boost device performance, or offering a versatile designer media for manipulating light in nanoscale to enable novel functionalities. We further discuss current challenges and outline emerging opportunities of this vibrant field for various applications in photonic integrated circuits, biomedical sensing, artificial intelligence and beyond.

Optical meta-waveguides for integrated photonics and beyond.

Research on two-dimensional designer optical structures, or metasurfaces, has mainly focused on controlling the wavefronts of light propagating in free space. Here, we show that gradient metasurface structures consisting of phased arrays of plasmonic or dielectric nanoantennas can be used to control guided waves via strong optical scattering at subwavelength intervals. Based on this design principle, we experimentally demonstrate waveguide mode converters, polarization rotators and waveguide devices supporting asymmetric optical power transmission. We also demonstrate all-dielectric on-chip polarization rotators based on phased arrays of Mie resonators with negligible insertion losses. Our gradient metasurfaces can enable small-footprint, broadband and low-loss photonic integrated devices.

Controlling propagation and coupling of waveguide modes using phase-gradient metasurfaces

The rapid growth of brain-inspired computing coupled with the inefficiencies in the CMOS implementations of neuromrphic systems has led to intense exploration of efficient hardware implementations of the functional units of the brain, namely, neurons and synapses. However, efforts have largely been invested in implementations in the electrical domain with potential limitations of switching speed, packing density of large integrated systems and interconnect losses. As an alternative, neuromorphic engineering in the photonic domain has recently gained attention. In this work, we demonstrate a purely photonic operation of an Integrate-and-Fire Spiking neuron, based on the phase change dynamics of Ge$_2$Sb$_2$Te$_5$ (GST) embedded on top of a microring resonator, which alleviates the energy constraints of PCMs in electrical domain. We also show that such a neuron can be potentially integrated with on-chip synapses into an all-Photonic Spiking Neural network inferencing framework which promises to be ultrafast and can potentially offer a large operating bandwidth.

/pdf/toward-fast-neural-computing-using-all-photonic-phase-change-4prfl6kc5g.pdf

Toward Fast Neural Computing using All-Photonic Phase Change Spiking Neurons

Low propagation loss in high confinement waveguides is critical for chip-based nonlinear photonics applications. Sophisticated fabrication processes which yield sub-nm roughness are generally needed to reduce scattering points at the waveguide interfaces in order to achieve ultralow propagation loss. Here, we show ultralow propagation loss by shaping the mode using a highly multimode structure to reduce its overlap with the waveguide interfaces, thus relaxing the fabrication processing requirements. Microresonators with intrinsic quality factors (Q) of 31.8 $\pm$ 4.4 million are experimentally demonstrated. Although the microresonators support 10 transverse modes only the fundamental mode is excited and no higher order modes are observed when using nonlinear adiabatic bends. A record-low threshold pump power of 73 $\mu$W for parametric oscillation is measured and a broadband, almost octave spanning single-soliton frequency comb without any signatures of higher order modes in the spectrum spanning from 1097 nm to 2040 nm (126 THz) is generated in the multimode microresonator. This work provides a design method that could be applied to different material platforms to achieve and use ultrahigh-Q multimode microresonators.

Exploiting Ultralow Loss Multimode Waveguides for Broadband Frequency Combs.

Deep learning has led to unprecedented successes in solving some very difficult problems in domains such as computer vision, natural language processing, and general pattern recognition. These achievements are the culmination of decades-long research into better training techniques and deeper neural network models, as well as improvements in hardware platforms that are used to train and execute the deep neural network models. Many application-specific integrated circuit (ASIC) hardware accelerators for deep learning have garnered interest in recent years due to their improved performance and energy-efficiency over conventional CPU and GPU architectures. However, these accelerators are constrained by fundamental bottlenecks due to 1) the slowdown in CMOS scaling, which has limited computational and performance-per-watt capabilities of emerging electronic processors, and 2) the use of metallic interconnects for data movement, which do not scale well and are a major cause of bandwidth, latency, and energy inefficiencies in almost every contemporary processor. Silicon photonics has emerged as a promising CMOS-compatible alternative to realize a new generation of deep learning accelerators that can use light for both communication and computation. This article surveys the landscape of silicon photonics to accelerate deep learning, with a coverage of developments across design abstractions in a bottom-up manner, to convey both the capabilities and limitations of the silicon photonics paradigm in the context of deep learning acceleration.

A Survey on Silicon Photonics for Deep Learning

https://onlinelibrary.wiley.com/doi/pdf/10.1002/adpr.202000212

Integrated Neuromorphic Photonics: Synapses, Neurons, and Neural Networks

Ultra-compact mode-order converters with dielectric slots are demonstrated on a silicon-on-insulator platform. We propose a mode converter that converts the TE0 mode into the TE1 mode with an ultra-small footprint of only 0.8×1.2µm2. The measured insertion loss is less than 1.2 dB from 1520 nm to 1570 nm. To reduce the insertion loss, we further optimize the structure and design two mode converters that convert the TE0 mode into the TE1 mode and the TE2 mode with footprints of 0.88×2.3µm2 and 1.4×2.4µm2, respectively. Their measured insertion losses are both less than 0.5 dB. Additionally, the proposed devices are cascadable and scalable for high-order mode conversion.

Ultra-compact silicon mode-order converters based on dielectric slots

Neuromorphic photonics that aims to process and store information simultaneously like human brains has emerged as a promising alternative for next generation intelligent computing systems. The implementation of hardware emulating the basic functionality of neurons and synapses is the fundamental work in this field. However, previously proposed optical neurons implemented with SOA-MZIs, modulators, lasers or phase change materials are all dependent on active devices and pose challenges for integration. Meanwhile, although the nonlinearity in nanocavities has long been of interest, the previous theories are intended for specific situations, e.g., self-pulsation in microrings, and there is still a lack of systematic studies in the excitability behavior of the nanocavities including the silicon photonic crystal cavities. Here, we report for the first time a universal coupled mode theory model for all side-coupled passive microresonators. The excitability of microresonators resulting from the subcritical Andronov-Hopf bifurcation is categorized as Class 2 neural excitability and can be used to mimic the Hodgkin-Huxley neuron model. We demonstrate that the microresonator-based neuron can exhibit the typical characteristics of spiking neurons: excitability threshold, leaky integrating dynamics, refractory period, cascadability and inhibitory spiking behavior, paving the way to realize all-optical spiking neural networks.

All-Optical Spiking Neuron Based on Passive Microresonator

On-chip spatial mode operation, represented as mode-division multiplexing (MDM), can support high-capacity data communications and promise superior performance in various systems and numerous applications from optical sensing to nonlinear and quantum optics. However, the scalability of state-of-the-art mode manipulation techniques is significantly hindered not only by the particular mode-order-oriented design strategy but also by the inherent limitations of possibly achievable mode orders. Recently, metamaterials capable of providing subwavelength-scale control of optical wavefronts have emerged as an attractive alternative to manipulate guided modes with compact footprints and broadband functionalities. Herein, we propose a universal yet efficient design framework based on the topological metamaterial building block (BB), enabling the excitation of arbitrary high-order spatial modes in silicon waveguides. By simply programming the layout of multiple fully etched dielectric metamaterial perturbations with predefined mathematical formulas, arbitrary high-order mode conversion and mode exchange can be simultaneously realized with uniform and competitive performance. The extraordinary scalability of the metamaterial BB frame is experimentally benchmarked by a record high-order mode operator up to the twentieth. As a proof of conceptual application, an 8-mode MDM data transmission of 28-GBaud 16-QAM optical signals is also verified with an aggregate data rate of 813 Gb/s (7% FEC). This user-friendly metamaterial BB concept marks a quintessential breakthrough for comprehensive manipulation of spatial light on-chip by breaking the long-standing shackles on the scalability, which may open up fascinating opportunities for complex photonic functionalities previously inaccessible. 

/pdf/metamaterial-enabled-arbitrary-on-chip-spatial-mode-1g0bqawm.pdf

Metamaterial-enabled arbitrary on-chip spatial mode manipulation

/pdf/metamaterial-enabled-arbitrary-on-chip-spatial-mode-31ad1omi.pdf

Jinlong Xiang

Papers

Integrated Neuromorphic Photonics: Synapses, Neurons, and Neural Networks

Ultra-compact silicon mode-order converters based on dielectric slots

All-Optical Spiking Neuron Based on Passive Microresonator

Metamaterial-enabled arbitrary on-chip spatial mode manipulation

Metamaterial-enabled arbitrary on-chip spatial mode manipulation