This PDF file contains the front matter associated with SPIE Proceedings Volume 11691, including the Title Page, Copyright Information, and Table of Contents.

Silicon photonics has emerged as a game-changing technology for data communications in recent years. However, the technology is also applicable to an increasing number of other applications. Even in the data-comm application, the relentless demand for more data requires the performance of silicon photonics to continue to improve. In this presentation I will discuss our work in three aspects of Silicon Photonics technology. The first is high-speed optical transmitters. Our approach of co-design of the photonic and electronic components of systems has enabled us to demonstrate 100Gb/s OOK from a single silicon modulator without any equalisation. Secondly, I will discuss a technology that we have developed that allows comprehensive wafer scale testing of silicon photonics circuits, as well as trimming of individual devices, and a non-volatile method of programming silicon photonics circuits without the need for large power consumption to maintain the state of the programmed circuit. Finally, I will discuss our work with Pointcloud Inc., on 3D imaging via an integrated LIDAR system, which has demonstrated millimeter accuracy for measurements made at distances beyond 70m.

We present polarization splitters for micron-scale silicon photonics based on Mach-Zehnder interferometers (MZI). We have designed the polarization splitters based on an advanced transfer matrix model, then confirmed by numerical simulations. We have fabricated the devices on our 3 µm-thick silicon on insulator (SOI) platform and successfully measured polarization splitting over a wide wavelength range (1450-1650 nm). We have achieved an extinction ratio (ER) >14dB for both polarizations on a bandwidth of 35 nm. However, high ER upto 20 dB is observed when considering one polarization.

We demonstrate a polarization rotator-splitter (PRS) design on standard 220 nm silicon-on-insulator (SOI) wafers with all rib waveguides and 2 μm silicon dioxide (SiO₂) claddings. The design is fully compatible with the imec iSiPP50G silicon photonics platform. We show TM0-TE0 converting loss < 0.5 dB and all polarization crosstalk < -10 dB in the wavelength range of 1500- 1600 nm.

Optical power splitters are widely used in many applications and different typologies have been developed for devices dedicated to this function. Among them, the multimode interference design is especially attractive for its simplicity and performance making it a strong candidate for low-cost applications, such as photonics lab-on-chips for biomedical point of care systems. Within this context, splitting the optical beam equally into multiple channels is of fundamental importance to provide reference arms, parallel sensing of different biomarkers and allowing multiplexed reading schemes. From a theoretical point of view, the multimode structure allows implementation of the power splitting function for an arbitrary number of channels, but in practice its performance is limited by lithographic mask imperfections and waveguide width. In this work we analyze multimode waveguide structures, based on amorphous silicon (a-Si:H) over insulator (SiO₂), which can be produced by the PECVD deposition technique. The study compares the performance of several 1 to N designs optimized to provide division of the fundamental quasi-TM mode as a function of input polarization and lithographic roughness. The performance is analyzed in terms of output power uniformity and attenuation and is based on numerical simulations using the Beam Propagation Method and Eigenmode Expansion Propagation Methods.

We present theoretical designs of high performance optical filters in integrated silicon photonic nanowire resonators. We use mode interference in formed by zig-zag waveguide coupled Sagnac loop reflectors (ZWC-SLRs), tailored to achieve diverse filtering functions with good performance. These include compact bandpass filters with improved roll-off, optical analogues of Fano resonances with ultrahigh spectral extinction ratios (ERs) and slope rates, and resonance mode splitting with high ERs and low free spectral ranges. The analysis verifies the feasibility of multi-functional integrated photonic filters based on ZWC-SLR resonators for flexible spectral engineering in diverse applications.

Periodic silicon waveguides with a pitch that is below half the effective wavelength of light support diffraction-less Bloch modes. These modes propagate as through a homogeneous, artificial-core metamaterial waveguide whose optical characteristics can be engineered by lithographic patterning. Subwavelength gratings (SWGs) provide designers with unique tools to control the refractive index, dispersion and birefringence of the equivalent metamaterial, yielding improved device performance. Based on this approach many high-performance optical devices have been designed and experimentally demonstrated in the last years. In this paper we will review the fundamentals of SWG engineering and present some of our latest findings.

Data center IP traffic is doubling every 2.5 years, driving the need to scale connectivity bandwidth on a similar cadence. At higher data rates the electrical link reach shrinks but energy efficiency does not improve significantly. Co-packaging optics close to ASICs enables data throughput scaling by reducing the SERDES power and hence overall power due to shorter electrical channels. Despite the advantages, co-packaging optics next to electronics can be challenging. This paper reviews Intel’s advancements in demonstrating industry’s first fully operational Silicon Photonic integrated circuit co-packaged with switch ASICs, describing in detail component level, integrated photonic integrated circuit (PIC) level, and transceiver module level design and performance.

This Conference Presentation, “Compact modeling and parametric extraction of phase shifters in carrier-depletion Mach-Zehnder silicon modulators,” was recorded for the Photonics West 2021 Digital Forum.

The goal of SiEPICfab is to conduct research in the fabrication of silicon photonic devices and photonic integrated circuits, and to make leading-edge silicon photonic manufacturing accessible to Canadian and international academics and industry. SiEPICfab builds on the success of the Silicon Electronic Photonic Integrated Circuits (SiEPIC) program, which has been offering research training workshops since 2008, by adding a fabrication facility “fab”. We have developed a rapid prototyping facility to support a complete ecosystem of companies involved in silicon photonics product development, including modelling, design, library development, fabrication, test, and packaging of silicon photonics. SiEPICfab allows designers to rapidly complete design-fabricate-test cycles, with technologies such as sub-wavelength sensors, PN junction ring modulators, silicon defect-based detectors, single photon detectors, single photon sources, and photonic wire bond integration of lasers and optical fibres.

Silicon (Si) photonic micro-electro-mechanical systems (MEMS), with its low-power phase shifters and tunable couplers, is emerging as a promising technology for large-scale reconfigurable photonics with potential applications for example in photonic accelerators for artificial intelligence (AI) workloads. For silicon photonic MEMS devices, hermetic/vacuum packaging is crucial to the performance and longevity, and to protect the photonic devices from contamination. Here, we demonstrate a wafer-level vacuum packaging approach to hermetically seal Si photonic MEMS wafers produced in the iSiPP50G Si photonics foundry platform of IMEC. The packaging approach consists of transfer bonding and sealing the silicon photonic MEMS devices with 30 μm-thick Si caps, which were prefabricated on a 100 mm-diameter silicon-on-insulator (SOI) wafer. The packaging process achieved successful wafer-scale vacuum sealing of various photonic devices. The functionality of photonic MEMS after the hermetic/vacuum packaging was confirmed. Thus, the demonstrated thin Si cap packaging shows the possibility of a novel vacuum sealing method for MEMS integrated in standard Si photonics platforms.

Silicon photonics can make complex circuits. We now understand that meshes of interferometers can perform arbitrary linear operations, beyond previous optical approaches. But configuring and stabilizing large meshes could be challenging. Fortunately, a class of architectures can both implement any linear transform and can also be self-configured and self-stabilized, even without any calculations. They can also adapt in real time to changing problems and environments. Applications include adaptive mode multiplexing and demultiplexing, automatic beam coupling, complex linear transforms for classical and quantum processing, and full measurement of amplitude and phase of multimode fields. Architectures, algorithms, and applications will be discussed.

Accurate 3D imaging is essential for machines to map and interact with the physical world1,2. While numerous 3D imaging technologies exist, each addressing niche applications with varying degrees of success, none have achieved the breadth of applicability and impact that digital image sensors have achieved in the 2D imaging world3-10. A large-scale twodimensional array of coherent detector pixels operating as a light detection and ranging (LIDAR) system could serve as a universal 3D imaging platform. Such a system would other high depth accuracy and immunity to interference from sunlight, as well as the ability to directly measure the velocity of moving objects11. However, due to difficulties in providing electrical and photonic connections to every pixel, previous systems have been restricted to fewer than 20 pixels12-15. Here, we demonstrate the first large-scale coherent detector array consisting of 512 (32×16) pixels, and its operation in a 3D imaging system. Leveraging recent advances in the monolithic integration of photonic and electronic circuits, a dense array of optical heterodyne detectors is combined with an integrated electronic readout architecture, enabling straightforward scaling to arbitrarily large arrays. Meanwhile, two-axis solid-state beam steering eliminates any tradeoff between field of view and range. Operating at the quantum noise limit16,17, our system achieves an accuracy of 3.1 mm at a distance of 75 meters using only 4 mW of light, an order of magnitude more accurate than existing solid-state systems at such ranges. Future reductions of pixel size using state-of-the-art components could yield resolutions in excess of 20 megapixels for arrays the size of a consumer camera sensor. This result paves the way for the development and proliferation of low cost, compact, and high-performance 3D imaging cameras, enabling new applications from robotics and autonomous navigation to augmented reality and healthcare.

Future applications of ultrasound and photoacoustic imaging require a matrix of small and sensitive ultrasound sensors with read-out through a flexible cable. Silicon photonic ultrasound sensors have good prospects: small and sensitive sensors, wafer-scale fabrication, and matrix read-out via single optical fiber using photonic multiplexing. Here, we discuss different types of silicon photonic ultrasound sensors and their applications. This includes our optomechanical ultrasound sensor with extreme sensitivity that is achieved with an innovative optomechanical silicon photonic waveguide in an acoustical membrane. We discuss limitations of state-of-the-art piezoelectric sensors, how silicon photonic sensors overcome these, and applications in medical imaging.

LIDAR on a silicon chip holds strong potentials for LIDAR system solutions featuring low cost, small size, and high robustness. In line with this effort, on-chip circulators are of great interest as they bring significant benefit for system complexity reduction and SNR improvement by enabling the LIDAR transmitter and receiver to share a single common aperture. Here, we present our recent study on passive silicon photonics nonlinear switches as conditional circulators for LIDAR applications. We propose a device implementation to address the nonlinear switch working principle by controlling waveguide nonlinear coefficient using sub-wavelength gratings. This implementation is foundry-compatible using only regular passive silicon waveguide components and are fully demonstrated in the experiment. In addition, we propose a sub-splitting coupler-based switch potentially can achieve a better fabrication tolerance than sub-wavelength grating-based switch. This work builds up signal processing functions in silicon photonics technology for optical communication and sensing applications. In particular, for LIDAR applications, this work contributes to the critical components of important use, and the easy integration with other existing functions such as optical phased arrays and spectral filters pronounces the potential for LIDAR on a silicon chip.

In the last decade, silicon photonics has undergone an impressive development driven by an increasing number of technological applications. Plasmonics has not yet made its way to the microelectronic industry, mostly because of the lack of compatibility of typical plasmonic materials with foundry processes. In this framework, we have developed a plasmonic platform based on heavily n-doped Ge grown on silicon substrates. We developed growth protocols to reach n-doping levels exceeding 10²⁰ cm^-3, allowing us to tune the plasma wavelength of Ge in the 3-15 μm range. The plasmonic resonances of Ge-on-Si nanoantennas have been predicted by simulations, confirmed by experimental spectra and exploited for molecular sensing. Our work represents a benchmark for group-IV mid-IR plasmonics.

The sensitivity of a hybrid distributed feedback semiconductor (DFB) laser heterogeneously integrated onto silicon (Si) is extensively characterized in the presence of external optical feedback at different bias and temperature conditions. The unique modal engineering approach of the device allows the light generated in the III-V material to be stored in the low-loss Si region to significantly enhance the quality (Q) factor of the cavity resonator. This design leads to an increased temperature tolerance of the laser without impacting the transmission efficiency even under the most severe feedback conditions. At a temperature of T = 35◦C, the laser continuous to unveil optimal performance and exhibits feedback insensitivity when externally modulated at 10 Gbps transmission over a 10 km fiber coil. The study presented here demonstrates the ability of a high-Q laser to achieve floor-free transmission at different operating conditions with a power penalty degradation no greater than 1.5 dB. The prolonged transition to the coherence collapse regime at a much higher reflection level evidenced by this device when compared to its III-V counterparts in addition to its ability to withstand perturbations associated with temperature variations and unintentional back-reflections delivers a step forward towards isolator-free applications. This work suggest that this type of semiconductor lasers can serve as a promising solution for the development of compact and reliable photonic integrated circuits (PICs).

Quantum dot lasers directly grown on silicon are excellent candidates to achieve energy and cost-efficient optical transceivers thanks to their outstanding properties such as high temperature stability, low threshold lasing operation, and high feedback tolerance. In order to reach even better performance, p-type doping is used to eliminate gain saturation, gain broadening due to hole thermalization and to further reduce the linewidth enhancement factor. Optical transceivers with low relative intensity noise are also highly desired to carry broadband data with low bit-error rate. Indeed, the intensity noise stemming from intrinsic optical phase and frequency fluctuations caused by spontaneous emission and carrier noise degrades the signal-to-noise ratio and the bit-error rate hence setting a limit of a highspeed communication system. This paper constitutes a comprehensive study of the intensity noise properties of epitaxial quantum dot lasers on silicon. Results show minimal values between - 140 dB/Hz and - 150 dB/Hz for doping level between 0 and 20 holes/dot in the active region. In particular, the intensity noise is insensitive to temperature for p-doped QD laser. Modulation properties such as damping, carrier lifetime, and K-factor are also extracted from the noise characteristics and analyzed with respect to the doping level. We also provide numerical insights based on an excitonic model illustrating the effects of the Shockley-Read-Hall recombination on the intensity noise features. These new findings are meaningful for designing high speed and low noise quantum dot devices to be integrated in future photonic integrated circuits.

Silicon photonic modulators are a key component for electro-optic transmitter within data centers. Electro-refractive modulators relying on free carrier plasma dispersion in Mach-Zehnder interferometer have become the most popular solution. Accumulation–based capacitive modulators are an efficient approach, which can reduce the modulation power consumption. In this work we study the behavior of capacitive modulators with polycrystalline silicon to form the capacitance. The modulators are made within the standard fabrication flow with only few add-ons. In this work we show that furnace annealing conditions and excimer laser annealing conditions during the polycrystalline silicon formation enhance the modulator bandwidths.

Silicon modulators paved the way for silicon photonics to take control of optical interconnects. Since its popularization, most works use the 1-D diode model approximation to design the horizontal PN junction, which estimates the modulator bandwidth and efficiency. Some works do not even consider the effects of fringe capacitance, alleging that the junction’s dimensions are large. The 1-D model is suitable for vertically uniform PN junctions. However, there are essential deviations for the typical rib waveguide used in most horizontal-junction silicon modulators. Our work aims to quantify such deviations incorporating details from 2D model simulations and offer a corrected 1-D model for estimating modulation bandwidth. This study was carried out as follows: firstly, we incorporated an improved scheme for phase shifting and loss for different junction locations and widely used doping concentrations. Next, we analyzed the generation-recombination effects and their impact on the depletion width at the top and bottom of the waveguide. We calculated the depletion width via the 1-D model and the two-dimensional Poisson’s equation finite-element calculation for the rib and identified an important mismatch. Lastly, we propose and demonstrate an accurate equivalent circuit with our 1-D model corrections. Our model considers the total depletion capacitance, the fringe capacitance, the capacitance due to the wider depletion widths at the top and bottom surfaces of the diode, and other capacitive effects at the border of the rib as a result of high reverse bias. We found that although the 1-D model is well-suited for small reverse biases, higher voltages and extreme junction locations affect the bandwidth’s estimation dramatically.

Silicon photoconductive heaters-detectors have been demonstrated to be useful for their ability to simultaneously act as thermo-optic phase shifters and in-waveguide photodetectors, as well for their ease of integration with silicon photonic fabrication processes. This functionality allows for the automated control of circuit elements through detect-and-tune control loops, which enable the efficient scaling of large integrated optoelectronic circuits. We have developed a compact model for the optoelectronic properties of silicon photoconductive devices in Lumerical INTERCONNECT based on measured results from fabricated devices, allowing designers to estimate the performance of such devices in circuits before fabrication. We demonstrate relative device performance compared to germanium detectors, and highlight target applications for such devices through simulation and fabricated devices, including a compact and widely reconfigurable notch filter.

In this study, we found giant photothermal nonlinearity with 𝑛₂ = 10^-1𝜇𝑚²/𝑚𝑊 in ~100𝑛𝑚 silicon nanoblocks, based on Mie-resonance enhanced absorption and efficient temperature increase via thermal insulation. Through a continuouswave pump-probe setup, we demonstrated an ultrasmall high-contrast all-optical switch with 90% modulation depth. Due to the 0.001𝜇𝑚³ small geometrical size, thermal dissipation is as fast as nanosecond, leading to modulation speed at GHz, which is much faster than other thermal optic switches. The large and fast all-optical switching could open the possibility toward high-density integrated photonic nanocircuits based entirely on silicon.

This paper describes the design of the model of the digital half-adder based on silicon photonics using the micro-ring resonator (MRR) structures. Silicon photonics is a promising field of study that could replace or modify modern electronics. The development of this field leads to the discovery of new solutions for existing electronic circuits. One of the examples is a digital half-adder that is proposed to be designed using MRR structures. In this work, we designed a half-adder based on MRR structures with two inputs and two outputs. The static response of the digital circuit was also investigated. Finally, the performance of the model was demonstrated under different bitrates - 100, 400, 700 and 1000 kbps.

The Genetic Algorithm (GA) is one of the most popular heuristic methods due to its natural and fast implementation. However, at the same time, it has the disadvantage of poor optimization. To improve performance, it’s necessary avoid stuck in local maximums throught choosing proper methods and parameters that vary for each application. In photonic devices, although the GA has been recently used to optimize passive silicon Y-branches, its performance is still trailing behind other optimization algorithms based on swarms, for instance. In this work, we present a new three-part heuristic method for optimizing Y-branches. We used the Finite-difference Time-domain (FDTD) method and the Particle Swarm Optimization (PSO) to generate an optimal data set as initial population for the GA. Considering an adequate population model, we demonstrate improvement in the performance for the design of a Y-branch through the GA. Next, we used a variation of a gradient-based search method to fine-tune the final parameters to find the absolute maximum. As a result, we produced new non-intuitive Y-branch devices with on-chip areas smaller than 2µm² and excess loss down to 0.05 dB @1550 nm for the TE mode. A complete study of fabrication feasibility and uv-lithography typical fabrication errors and its effects on the bandwidth will be shown at the time of the conference. Our method will be compared against other widely-used heuristic methods in photonic device design in terms of number of iterations.

We propose a general procedure for the automatic design of N x N integrated optical switches using the Beneš configuration based on ring filters, suitable for multi-hop routing in meshed optical networks. The basic switch components (waveguides, couplers, optical crossings) are selected from the process development kit (PDK) of the target platform, and can be integrated with user-defined custom components. The described procedure is able to automatically select the optimal number of rings and their proper routing, minimizing the number of optical crossings and reducing the system transmission losses. The generated topology can be directly simulated in the Synopsys OptSim© environment at a system level. There, the impact of the design options on transmission performances is evaluated for the chosen signal modulation scheme and numerical simulations are automatically executed to determine, among the 2^M possible optical paths across the electrically tuned M rings, the only N! configurations required to generate all the permutations of the inputs signals to the output ports minimizing the transmission BER. Finally, the generated Beneš switch can be converted to a GDSII mask targeting the foundry of choice and manual adjustment to the layout can be made. The automatic procedure takes advantage of the vertical integration of the Synopsys environment, which allows to perform a comprehensive investigation of the whole photonic device, from the component design to the system level analysis, thanks to the integration of Optsim- optical communication system, Optsim schematic-driven photonic circuit, Rsoft component design tools, and OptoDesigner© mask layout.

We report on the statistical assessment of the properties of directional couplers based on silicon waveguides, growth by IMEC fab in Belgium in the framework of the Europractice partnership. We characterized 25 chips from a multi-project wafer, each one containing several passive add-drop ring resonators with different coupling strengths. The analysis was repeated for chip temperatures ranging from 25°C to 55°C. The measurements we performed confirmed the reliability of the iSiPP50G platform used to growth the considered components.

Coupling between monomode silicon waveguides and Ge photodetectors is challenging due to the size mismatch between both components. This work proposes a non-linear tapered coupling device to couple light from a silicon on insulator waveguide to an SOI Ge photodetector. For our 32.1 µm non-linear spot-size converter, the simulated transmission coefficients in the SCL band for 1550 nm, 1460 nm, and 1625 nm wavelength are -0.0113, -0.059, -0.0092 dB, respectively; while state of the art linear spot-size converters of the same transmission coefficients are around 211% bigger.

The nanowire transistors on silicon-on-insulator (SOI) substrate embedded with ferroelectric hafnium-zirconium-oxide (HfZrO₂) are elaborately probed when the devices are illuminated with the ultraviolet (UV) laser. The basic functionality of the ferroelectric nanowire transistor can be verified by monitoring the drain current hysteresis during the bidirectional gate voltage scan. Therefore, this study mainly analyzes and summarizes the electrical response of the device to ultraviolet (UV) irradiation; the main emphases will be placed on the rotational direction of the hysteresis window and the width of the hysteresis window when components of different dimensions are compared with one another. To administer the comparisons impartially, the pertinent surface-to-volume ratios of these nanowire transistors are used as the gauging parameters. As the device measurements would demonstrate, Hysteresis rotating in a clockwise direction is attributed to the oxide layer defects, while the counterclockwise direction is induced by the ferroelectric effect. Needless to say, the quality of the device itself is still contingent upon the gate oxide robustness and the quality of its adjacent interfaces. And last but not least, the threshold voltage shift is also used as an indicator to illuminate the impact of changing polarization effect on the nanoscale devices. Through the effective modulation of the hysteretic window by irradiating the nanowire FETs with a UV laser, we believe many unique applications involving the optical modulation and photodetection that are commonly found in silicon photonics can be realized.

Photonic crystals demonstrate photonic band-gaps in their optical spectra, due to the periodic contrast of refractive index. The periodic nature characteristic of photonic crystals exhibits the Bragg scattering phenomena, which only captures the targeted wavelengths to create resonance effect within the cavity structure. Using this phenomena, we present the model of High-Q 1D photonic crystal waveguide resonator for wavelengths centered at 2.4 µm in silicon-on-insulator (SOI). Moreover, silicon proves to be a low-loss material for mid-IR range of up to 4 µm. The modelling and optimization of the silicon photonic crystal is achieved by performing 3D FDTD simulations using a commercial software. A well-known L3 cavity is designed in 400 nm thick SOI layer by removing three holes from the center. Due to fabrication tolerance circular holes are used to achieve photonic crystal effect. The cavity is optimized in two stages by placing a magnetic dipole source at the center. In first stage, a two-step optimization was performed by optimizing the air hole period followed by the radius. In second stage, the period and radius optimization of the inner most air holes of the cavity were carried out to achieve the maximum High-Q factor of 77,538 at wavelength of 2.4 µm. Our proposed design is consistent with the available lithography for achieving the mid–IR sensors in SOI. Also, Multi-Project-Wafer (MPW) monolithic integration can be achieved using single etch process reducing the effective cost. These sensors can widely be used for gas sensing applications for example environmental monitoring of public areas to monitor the concentrations of carbon dioxide gas.

We recently achieved n_2 ~10-1 μm^2/mW on a single silicon Mie resonator, i.e. five orders-of-magnitude improvement of silicon nonlinearity. Here we present the direct evidence quantitatively linking the nonlinearity to temperature rise with 10K precision, and unravel that the huge optical nonlinearity is due to nonlinear temperature rise, coupled with nonlinear absorption, resulting in 1000K increase with ~10 mW/μm^2 excitation. We developed corresponding numerical simulation tools that confirm our observations and can be adapted to explain general nanostructure heating.