This PDF file contains the front matter associated with SPIE Proceedings Volume 12891, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Multiple application segments from data transmission to sensing drive the demand for high-performance photonic integrated circuits. We review advancements in silicon photonics manufacturing platform for datacom and multi-Tb/s optical interconnects.

In this work, we explore the manifestation of optical nonlinearities in silicon, given illumination by radiation with wavelengths in the optical communication (C-band) spectrum, near 1550 nm, and extreme intensities, spanning 100-1000 GW/cm². We photoexcite a silicon photodiode with femtosecond-duration 1550-nm laser pulses and observe the resulting optical autocorrelations as a function of the peak pulse intensity. Such measurements in silicon reveal (i) negligible single-photon absorption, suggesting that there are few defect (trap) states in the bandgap that can assist below-bandgap photoexcitation, (ii) significant two-photon absorption at intensities above 100 GW/cm², (iii) growing three-photon absorption at intensities rising above a threshold of 300 GW/cm², and (iv) increasing saturation at intensities rising above a threshold of 650 GW/cm². We attribute this saturation to the extremely high density of charge carriers brought about by three-photon absorption—as this depletes the available electrons in the valence band and the available states in the conduction band. We hope that this work will be a foundation for the future integration of telecom (C-band) technologies and silicon nanostructures.

Binarized neural networks offer substantial reductions in memory and computational requirements compared to full precision networks. However, conventional CMOS-based hardware implementations still face challenges with resilience for deployment in harsh environments like space. This paper proposes an optical XOR-based accelerator for binarized neural networks to enable low power and resilient operation. The optical logic gates rely on wavelength-specific intensity propagation rather than absolute intensity levels. This provides inherent robustness against fabrication process variations and high energy particle strikes. Simulations of an optical hardware prototype for XNOR-Net show the accelerator achieves 1.2 μs latency and 3.2 mW power. The binarized network maintained 2-4% accuracy degradation compared to the full precision baseline on MNIST and CIFAR-10. The proposed optical accelerator enables efficient and resilient deployment of binarized neural networks for harsh environment applications like spacecraft and satellites.

Single-mode and low-loss operation of optical waveguides is typically limited to a 200-500 nm wide wavelength range. The lower limit is the boundary between single and multi-mode operation, and the upper limit comes from the decreasing confinement of the fundamental mode inside the core, which eventually leads to too large bending radii, waveguide cross-talk and poor integration density. Many interferometric waveguide components, such as grating couplers and multi-mode interference (MMI) couplers, have even narrower wavelength range. This paper demonstrates photonic integrated circuits (PICs) with ultra-broadband operation from 1.2 to 2.4 μm wavelength based on 3 μm thick silicon-on-insulator (SOI) waveguides. Such thick waveguides maintain ultra-high mode confinement for over 1 μm bandwidth, which supports dense integration with low-loss crossings, Euler bends and total internal reflection (TIR) mirrors. While some parts of the PICs are based on multi-moded strip waveguides, mode filters with rib-waveguides allow to keep the PICs effectively single-moded. The focus of the paper is on passive PICs, although the platform also enables active components. Ultra-broadband test results are provided for long waveguide spirals and waveguide-fiber coupling, as well as for echelle gratings, arrayed waveguide gratings (AWGs) and different types of 2x2 couplers. Low-loss operation is demonstrated with continuous transmission spectra measured from 1.25 μm up to 2.4 μm wavelength, i.e. up to 1.15 μm bandwidth. The measured bandwidths are limited by the available measurement setup, rather than the PIC components themselves. Remaining challenges for ultra-broadband operation, such as anti-reflection coatings, are discussed. Applications for broadband operation in communication, imaging and sensing are also presented.

We present a 1×16 optical phased array (OPA) in state-of-the-art 45-nm SOI CMOS technology for free-space optical communication and imaging. Width and grating period of the antenna is optimized to have longitudinal radiation angle of 10˚ and radiation efficiency of 35.3 % at an operating wavelength of 1310 nm. To reduce insertion loss of the OPA, a thermo-optic phase shifter based on n-i-n junction is used. The proposed OPA has been fabricated in 45-nm electronic-photonic process from GlobalFoundries, which is intended to fabricate both integrated photonics and electronics. Far-field measurement result shows 1-D two-sided beam-steering angle of 34.3˚ in transversal direction and full-width half-maximum beam-width of 3.7˚.

Color centers are promising candidates for quantum technologies due to their long coherence times and high-quality spin-photon interfaces. Silicon has recently emerged as a host for color centers operating in the telecommunication bands, in a technological platform featuring the world’s most advanced manufacturing, electronics, and photonics. In this talk, I will present our recent work on the fabrication and isolation of individual G-centers in silicon photonic waveguides, their spectral reconfiguration, and the enhancement of their light-matter interaction via coupling to photonic crystal cavities.

A high-quality heralded single photon source is realized on silicon-on-insulator (SOI) platform. With the help of specially designed ultra-low loss fiber-chip edge couplers, the heralding efficiency of the single photon source system is 56%, after calibrating for a 38% detector efficiency. Compared with the state of the art, this measured heralding efficiency marks a new milestone for integrated, on-chip silicon sources.

Integrating quantum optics with silicon would substantially benefit from the extensive advancements made in manufacturing classical electronic and optical components. We present the first all-silicon quantum light source by embedding a single silicon-based defect within a silicon nanophotonic cavity. We have observed a 30-fold enhancement of luminescence, achieving near-unity atom-cavity coupling efficiency and an 8-fold acceleration of emission from the all-silicon quantum emissive center. These findings pave the way for large-scale integrated cavity quantum electrodynamics and quantum light-matter interfaces.

Quantum photonic computing may extend Moore's Law beyond sub-1nm CMOS technology nodes. There are additional performance advantages with Millimeter wave processors. Unlike lights, short wavelength microwaves are not so easily blocked by features in the integrated circuit, due to diffraction. 3D or more dimensional microwave communications between photonic microwave CMOS transistors can be achieved even without waveguides. In this paper, we will present techniques for novel photonic waveguides, which not only improve the photon transmission efficiency, but also increase the photonic CMOS switching speed. Photonic SRAMs are fabricated with local interconnected optical waveguides. Significant SRAM speed improvement can be accomplished. Similar to Phase Shift Masks, nonlinear holographic interference produces better optical signals. Optoelectronic Microwave Tunnel-Junction CMOS, Optoelectronic IMPATT CMOS, Photonic BARITT CMOS, and Transferred electron Gunn Microwave CMOS will be illustrated with photon-accelerated avalanche breakdown and frequency responses.

Photonic integrated circuits enable the miniaturization of photonic systems by integrating key optical functions on a chip. While CMOS compatible silicon and silicon nitride are very efficient platforms for passive circuits, they lack active key functionalities for the realization of a full system on a chip. A versatile solution is to use micro-transfer printing for heterogeneous integration of active devices on such platforms. Here we present the recent advances of micro-transfer printing on silicon nitride and discuss the remaining challenges.

We demonstrate a III-V-on-silicon-nitride mode-locked laser through the heterogeneous integration of a semiconductor optical amplifier on a passive silicon nitride cavity using the technique of micro-transfer printing. Specifically, we explore the impact of the gain voltage and saturable absorber current on the locking stability of a tunable mode-locked laser. By manipulating these parameters, we demonstrate the control of the optical spectrum across a wide range of wavelengths spanning from 1530 nm to 1580 nm. Furthermore, we implement an optimization approach based on a Monte Carlo analysis aimed at enhancing the mode overlap within the gain region. This adjustment enables the achievement of a laser emitting a 23 nm wide spectrum while maintaining a defined 10 dB bandwidth for a pulse repetition rate of 3 GHz.

This paper presents the design of two different control approaches for stabilizing the operating point of a microring resonator modulator (MRM) when fast disturbances occur. These arise in particular when the properties of the high-frequency modulation signal change due to the rapid electrical and optical self-heating effects. The challenge lies in the relatively slow dynamics of the heater used as an actuator and the significant limitation of the input. The first approach involves the design of a Model Predictive Control (MPC) suitable for a system like MRM. MPC enables a wide operating range by considering the non-linearities of the system. It can operate in both the stable and unstable regions while accounting for constraints on the control input and states of the system. A drawback for real-time implementation, especially in the case of a fast system like the MRM, is that the computational effort is relatively high in each time step due to the included optimization of the control input over a prediction horizon. On one hand, MPC can function as a benchmark design to showcase achievable control quality in simulation, despite its relatively high computational cost. On the other hand, employing a low-order approximation of the dynamic MRM model allows for offline pre-computation of the MPC internal optimization, thereby significantly reducing the online computational effort. The second approach proposes a computationally efficient PID control with two modes. The first mode is designed for normal operation of the PID controller in a region close to the operating point, while the second mode facilitates returning the system back to the operating point (where the PID controller can be restarted) from outside this region, achieved by applying a feedforward control. The system's state variable, ring temperature, determines the operating regions and the corresponding mode selection. To address the challenge of unknown temperature, an Unscented Kalman Filter (UKF) is designed to estimate the temperature.

In this paper, we discuss 3 examples in which microlenses can be a useful tool to address challenges in coupling between a fiber array and photonic integrated circuit (PIC). The (arrays of) microlenses used in this work were realized by the photoresist reflow method and can either be monolithically integrated on the back side of a PIC or separately fabricated microlens blocks can be mounted on the device side of a PIC. The first example involves the implementation of silicon microlenses etched at the back side of a sensing PIC (operating in C-band) aiming at relaxed alignment tolerances and keeping the device side free of interfacing fibers. The second example involves the implementation of a 4-mm long working distance expanded beam (30 μm mode field diameter, C-band) interface for telecom/datacom applications which also greatly relaxes lateral and longitudinal alignment tolerances between grating couplers on PIC and a fiber array. The final example involves the integration of an isolator in this long working distance expanded beam interface. The isolator stack consisted of a polarizer (0.55 mm thick), a non-reciprocal Faraday Rotator (485 μm thick film latching Faraday Rotator) and half-wave plate (HWP, 91 μm quartz) glued on top of each other. We obtained broadband operation exhibiting a very low (between 1 and 1.5 dB) insertion loss and good extinction ratio (between 17 and 20 dB) in C-band (around 1550 nm).

We propose a temperature-insensitive silicon waveguide using a polymer on the top surface of the cladding. The temperature insensitivity of the waveguide is achieved by combining materials with positive and negative thermo-optic coefficients. The material used for the athermal silicon optical circuit proposed in this study is silicone resin. The length of the temperature compensation waveguide was determined based on the phase shifts in the light. This is because, if TO effect is nullified, there should be no phase variation. We input light into the waveguide and simulated the phase after passing through the temperature compensation structure. The amount of phase shift varied with the length of the waveguide when silicone resin was used for the top surface of the clad. The phase shift of the athermal waveguide was approximately 1% to 12% compared to the phase shift of the conventional waveguide.

Power-efficient thermo-optic phase shifters have been demonstrated using 3 μm thick silicon on insulator (SOI) waveguides fabricated on cavity-SOI wafers. In cavity-SOI the cavities are premade in the SOI wafer which simplifies the processing of the waveguides with thermally insulated heater structures. Measurement results of asymmetric Mach-Zehnder interferometric TO switches show a 10-fold decrease in required power for α π phase shift in devices fabricated on cavity-SOI when compared to devices fabricated on plain SOI. With the cavities the required heating power for the π phase shift is only 2.1 mW. Numerical simulations support the experimental results.

Using a highly-scalable and physics-informed design platform with custom Mach-Zehnder interferometers (MZIs), we design and experimentally demonstrate a 1 × 2 wideband duplexer on silicon operating within 1450-1630 nm. The device is constructed from six layers of cascaded MZIs whose geometries are optimized using an equivalent artificial neural network, in a total timeframe of 75 seconds. Experimental results show below 0.72 dB deviation from the arbitrarily-specified target response, and less than 0.66 dB insertion loss. Demonstrated capabilities and the computational efficiency of our design framework pave the way towards the scalable deployment of custom MZI networks in communications, sensing, and computation applications.

We demonstrate a vertical-junction, carrier-injection, micro-ring modulator that is fabricated using AIM Photonics’ 300 mm Quantum FLEX Platform which shows results with high modulation efficiency and a large ON-OFF ratio. The modulator device includes a ring and a single-bus, straight waveguide. The ring has a radius of 7 μm and a 220 nm silicon-on-insulator (SOI) waveguide is used both for the ring and the straight waveguides with a rib structure of 110-nm slab thickness. The width of the core waveguide is 550 nm for both the ring and the straight waveguides. The slab width between the full-height silicon core and contact area is kept at 1 μm on both sides from the 550-nm core. The coupling gap between the ring and the bus waveguide is designed to be 150 nm. To make the waveguide core vertical junction, the upper half of the core is n-doped and the lower half is p-doped. To have a smooth electrical connectivity between the core and the contact area, three-level doping is applied where the core is doped with the minimum concentration and the contact silicon area is doped with the highest concentration. The modulator is tested with a tunable laser over a 100-nm window extending from 1485 nm to 1585 nm. The light is coupled to the modulator using grating couplers which are used to couple input and output light. The vertical junction shows excellent direct current (DC) I-V characteristics and the modulator performs at high modulation efficiency of about 1.14 nm and a large ON-OFF ratio of about 21 dB at 1.0 V.

This talk will highlight opportunities for terahertz science and technology from nonlinear integrated photonic circuits by exploring waveguides, resonators and terahertz antennas. Their present and future applications in metrology, emission and waveform control are discussed.

Due to their tunable bandgap, compound semiconductors have become the primary materials for synthesizing infrared (IR) photodetectors. However, the manufacturing processes for most compound semiconductors are costly and energyintensive. Therefore, developing a low-cost, low-energy, and silicon-compatible IR photodetector is worthwhile. Here, we create an inverted pyramid structure (IPS) for Ag/n-Si photodetector and utilize its structure to enhance localized surface plasmon resonance (LSPR). A thick metal surface makes the hot carriers take a long time to pass through, while the short lifetime of the hot carriers leads to their decay before reaching the destination. Thus thick metal could result in degraded signals. However, examination of the response from the mid-infrared light at 4.26 μm, the 10nm-Ag/n-Si Schottky photodetectors show that the device with a thickness of 14 nm has a 2.6 times improvement in response compared to devices with a thickness of 10 nm, and the signal-to-noise ratio (SNR) increased by 3.6 times. Based on the results of scanning electron microscopy (SEM), the hot carrier effect and the variation of optical response intensity are found to depend highly on the space and size of nanoparticles (NPs). As a result, LSPR enhances light absorption and improves the optical response.

We propose a novel waveguide-coupled germanium lateral uni-traveling carrier photodetector (L-UTC PD) structure, which offers tremendous potential for coherent communication applications at 130 Gbaud and beyond. This PD structure is very simple and can be easily fabricated in a commercial silicon photonics (SiPh) foundry. The fabricated L-UTC PD has a very low dark current (~4.2 nA), a high responsivity (~0.66 A/W@1550 nm) and a small capacitance (<14.6 fF). Moreover, the L-UTC PD exhibits excellent tolerance to high optical input power, with a 3-dB bandwidth exceeding 67 GHz, even at a high photocurrent of 3.0 mA.

This paper discusses the performance of a Silicon based avalanche photodiode, SAP500, that is able to operate in Geiger and linear mode. This detector is based on a ‘reach through’ structure for excellent quantum efficiency, extremely low bulk dark current and noise and can achieve high gains as they are designed for ultra-low light level applications.

We propose a novel control scheme to induce transparency through thermal tuning in multi-stage cascaded ring structures. Higher-order MicroRing Resonator structures are fundamental blocks for high Q-factor filtering as well as wavelength-sensitive applications. Through the proposed thermal control scheme, which does not require additional heating elements with respect to the traditional calibration and uncertainty-compensation setup, we can manage both the central wavelength of the add-drop filter, as well as toggle the transparent state. Through this technique the add-drop filter can be switched off, allowing the signal to pass through without frequency-dependent effects and just introducing flat-band crosstalk.

Micro-ring resonators (MRRs) have emerged as vital components in photonic applications, offering precise control of light at the nanoscale. Achieving optimal MRR design parameters is crucial for maximizing their performance in high-speed applications. This study aims to employ feature engineering and supervised machine learning (ML) techniques to comprehensively analyze MRRs. This includes impact of change in MRR design geometries, such as radius, coupling geometry, waveguide properties to key MRR output parameters, including the quality factor, full width at half maximum (FWHM), rise/fall time, and free spectral range (FSR). By utilizing results of over 1000 simulations in Lumerical, as well as incorporating the theoretical knowledge of MRRs, the study seeks to build highly accurate predictive model.

To achieve more compact integrated photonic devices, reducing size of elements is crucial. A factor that limits sizereduction is electro-optic components that require large optical interaction length. In this work, we designed and fabricated an optical modulator where a photonic crystal structure is used to create large phase difference in short distance. Our design is a 2x2 Mach-Zehnder interferometer on the platform of silicon-on-insulator. A left-handed photonic crystal structure that is designed to operate at 1.55 um is placed on one arm of the interferometer to add phase to light. The phase difference between two arms yields amplitude modulation at the output of the interferometer. The photonic crystal is hexagonal air hole lattice and used to switch between negative and positive effective refractive indices. This change is triggered by applying voltage which decreases the refractive index of silicon from 3.480 to 3.477 due to plasma dispersion effect, and causes photonic band-to-band transition. By this way, effective refractive index of the structure jumps from negative to positive values. To be able to realize this, photonic crystal region is sandwiched between n-doped and p-doped materials, which creates a p-i-n diode. By taking the advantage of band-to-band transition at left-handed photonic crystal, we experimentally demonstrated that interaction length is reduced from 255 um to 4.4 um. This reduction leads to low optical insertion loss as well as more compact devices.

In this work, we investigate add-drop and all-pass micro-ring resonators to optimize their required parameters to achieve an improved hysteresis loop for silicon-based memory applications. Thermo-optic effect is applied to tune their working wavelength. By evaluating and comparing these two device configurations, we aim to determine the optimal choice for achieving better optical bistability. The provided theoretical analysis describes the impact of the thermo-optic coefficients of the micro-ring resonator on the bistable behavior. By varying these coefficients, we examine how the thermal response affects the switching characteristics and hysteresis effects in both micro-ring resonators. We also run numerical simulations where thermo-optic effect is applied by placing a micro-heater on top of each micro-ring resonator. The research involves varying several key parameters, such as the coupling coefficients, ring radius, and waveguide configurations, to optimize the hysteresis loop behavior. Through careful analysis and comparison, we determine the most suitable device configuration for achieving optimal performance and stability in memory systems.

Inverse design approaches with topology optimization can yield in highly efficient devices; however designing fabrication-compatible, broadband, yet simultaneously fabrication-tolerant devices still widely remains a challenge. Here, we design a broadband and fabrication-tolerant 10% silicon-based power tap using 3D-FDTD simulations and topology optimization, and demonstrate its experimental performance. The power tap has a compact footprint of 7.0 μm×3.1 μm, and achieves a broadband and spectrally flat operation from 1500 nm to 1600 nm. The device was specifically built to be fabrication-tolerant using an approach that maintains high performance under over-etch and under-etch scenarios by maximizing the contiguous area of the silicon layer in the final device. This tolerance was verified with 3D-FDTD simulations with 15 nm over-etch and under-etch modifications, demonstrating a change of less than 0.64 dB at either output port compared to the original device response at 1550 nm. The designed power tap was fabricated using a standard 220 nm thick silicon-on-insulator platform. The experimental measurements match closely with the design target and 3D-FDTD results, achieving state-of-the-art performance with excess losses as low as 0.23 dB and broadband operation. The output ports of the device also exhibit extremely flat spectra, where the transmission remains between 0.86 and 0.92 for the through port, and between 0.06 and 0.14 for the tap port throughout the 1500-1600 nm spectral range. These results represent the state-of-the-art experimental performance in compact power taps, and prove the effectiveness of fabrication-tolerant optimization.

We propose a deep photonic interferometer network architecture for designing fabrication-tolerant photonic devices. Our framework incorporates layers of variation-aware, custom-designed Mach-Zehnder interferometers and virtual wafer maps to optimize broadband power splitters under fabrication variations. Specifically, we demonstrate 50/50 splitters with below 1% deviation from the desired 50/50 ratio, even with up to 15 nm over-etch and under-etch variations. The significantly improved device performance under fabrication-induced changes demonstrates the effectiveness of the deep photonic network architecture in designing fabrication-tolerant photonic devices, and showcases the potential for improving circuit performance by optimizing for expected variations in waveguide width.

As silicon photonics-based circuit designs transition from lab to fab, an end-to-end automated measurement flow is required to address a unique combination of high flexibility in test conditions and high volume. This paper describes such a flow for process design kit (PDK) development in the state-of-the-art 300 mm CMOS-compatible silicon photonics foundry at the Albany NanoTech Complex in Albany, NY. Presenting details of this measurement flow will offer considerable cost and time savings to new users in this area. The measurement flow begins at the layout stage, where users can instantiate various combinations of pre-characterized padsets that contain DC/RF pads and optical couplers, which are compatible with the automated electro-optic setup used for measurements. These padsets are offered via two options: (1) a script-based layout builder tool or (2) a parametric cell in a “Measurement Design Kit” offering in a design automation platform, which is an analog to a PDK. Special marker layers are added to the padsets, whose coordinates are extracted after the layout is complete. The coordinates are then passed to fiber positioners on the semi-automated prober while performing measurements. Electro-optic measurements are performed across the wafer using vertical coupling, which is well-suited for large-scale measurements. The wafer is placed on a 300 mm prober with automated fiber positioners that can optimize optical coupling across six degrees of freedom. The electro-optic measurement setup is based on the Keysight Photonic Application Suite. It includes a tunable laser, polarization synthesizer, and multi-channel detectors that measure transmission in both TE and TM polarizations. A lowloss optical switch matrix is programmed to switch connections between lasers and detectors to 16 grating couplers in the padset. The entire measurement setup, including the prober and instruments, is driven using the Python-based SweepMe! automation framework, which is modular and allows for the easy creation of test plans.

Surface Plasmon Polariton resonant sensors (SPPs) have found wide-ranging applications, particularly for nanoscale sensing. The SPP dispersion is determined by the properties of a limited number of suitable metals and cannot be arbitrarily tuned. The proposed gas sensor is based on surface plasmon polariton manipulated in the metamaterial surface. The plasmonic sensor utilizes a metal-air interface to detect and analyze various substances and phenomena in the mid-infrared range. The utilization of the Mid-infrared (MIR) wavelength range offers numerous benefits across a wide range of applications, including chemical and biological detection. This paper introduces a metasurface composed of highly doped silicon that demonstrates a plasmonic effect in the mid-infrared wavelength range. Silicon has several benefits, including compatibility with CMOS technology and easy manufacturing utilizing traditional silicon fabrication techniques. In addition, operating in the mid-infrared (mid-IR) range and using doped silicon material enables the development of integrated plasmonic devices at the microscale. In this paper, a meta surface plasmonic grating is proposed for sensing and detecting the refractive index changes. In order to test the performance of the plasmonic sensor, a commercial Lumerical software based on finite difference time domain (FDTD) has been used. The suggested design shows high sensitivity at λ = 13.9807 μm and it can be used for gas sensing applications.