Photonic Integrated Circuits Research Articles

On-chip topological beamformer for multi-link terahertz 6G to XG wireless.

Terahertz (THz) wireless communication holds immense potential to revolutionize future 6G to XG networks with high capacity, low latency and extensive connectivity. Efficient THz beamformers are essential for energy-efficient connections, compensating path loss, optimizing resource usage and enhancing spectral efficiency. However, current beamformers face several challenges, including notable loss, limited bandwidth, constrained spatial coverage and poor integration with on-chip THz circuits. Here we present an on-chip broadband THz topological beamformer using valley vortices for waveguiding, splitting and perfect isolation in waveguide phased arrays, featuring 184 densely packed valley-locked waveguides, 54 power splitters and 136 sharp bends. Leveraging neural-network-assisted inverse design, the beamformer achieves complete 360° azimuthal beamforming with gains of up to 20 dBi, radiating THz signals into free space with customizable user-defined beams. Photoexciting the all-silicon beamformer enables reconfigurable control of THz beams. The low-loss and broadband beamformer enables a 72-Gbps chip-to-chip wireless link over 300 mm and eight simultaneous 40-Gbps wireless links. Using four of these links, we demonstrate point-to-4-point real-time HD video streaming. Our work provides a complementary metal-oxide-semiconductor-compatible THz topological photonic integrated circuit for efficient large-scale beamforming, enabling massive single-input multiple-output and multiple-input and multiple-output systems for terabit-per-second 6G to XG wireless communications.

Free-standing millimeter-range 3D waveguides for on-chip optical interconnects

Next-generation energy-efficient photonic integrated systems, such as neuromorphic computational chips require efficient heterogeneous integration of ultracompact light sources and photodetectors through highly dense waveguide circuits. However, interconnecting these devices in emitter-receiver communication circuits remains a challenge and an obstacle towards upscaling heterogeneous photonic chips. Here we report on versatile air-cladded free-standing 3D polymer waveguides (OrmoCore, n≈1.5) spanning up to 900 µm in length without intermediate mechanical support structures, microprinted via two-photon polymerization. The presented waveguides are suitable for on-chip out-of-plane light coupling as well as non-connected 3D crossings, needed for high density optical circuits. The waveguides show optical transmission losses of 1.93 dB mm−1 at λ = 635 nm, and of 3.71 dB mm−1 at λ = 830 nm in the wavelength range of GaAs-based microLEDs spectral emission. On-chip imaging for high-precision alignment and TPP microfabrication are performed seamlessly by utilizing the same laser source for both steps, allowing accurate 3D printing on microstructured substrates. As proof of concept, we interconnect two GaAs-based microLEDs via an on-chip microprinted 3D waveguide. Such combined systems can serve as building blocks of future complex integrated heterogeneous photonic networks.

An Integrated Photonic Biosensing Platform for Pathogen Detection in Aquaculture.

Aquaculture is expected to play a vital role in solving the challenge of sustainably providing the growing world population with healthy and nutritious food. Pathogen outbreaks are a major risk for the sector, so early detection and a timely response are crucial. This can be enabled by monitoring the pathogen levels in aquaculture facilities. This paper describes a photonic biosensing platform based on silicon nitride waveguide technology with integrated active components, which could be used for such applications. Compared to the state of the art, the current system presents improvements in terms of miniaturization of the Photonic Integrated Circuit (PIC) and the development of wafer-level processes for hybrid integration of active components and for material-selective chemical and biological surface modification. Furthermore, scalable processes for integrating the PIC in a microfluidic cartridge were developed, as well as a prototype desktop readout instrument. Three bacterial aquaculture pathogens (Aeromonas salmonicida, Vagococcus salmoninarum, and Yersinia ruckeri) were selected for assay development. DNA biomarkers were identified, corresponding primer-probe sets designed, and qPCR assays developed. The biomarker for Aeromonas was also detected using the hybrid PIC platform. This is the first successful demonstration of biosensing on the hybrid PIC platform.

Research on optical interferometric imaging with flexible control using optical fibers and PIC chip

We propose a prototype called a flexible integrated resolution and efficient light-imaging-expanded synthetic system (FIREFLIES). This paper describes the design, manufacturing, and experimental demonstration of the proposed system. FIREFLIES enables interferometric imaging at approximately 1550 nm using a variable baseline sampling technique, in which the baseline-collected light field forms interference fringes that are captured by an on-chip photodetector. This innovation extends the limited sampling distance imposed by the processing size restrictions of traditional photonic integrated circuit (PIC) links. Furthermore, we introduce a unique method for achieving super-resolution sampling by flexibly controlling the baseline. An experimental platform is constructed to test the FIREFLIES against a one-dimensional grating target. The experimental curves closely align with the theoretical predictions, confirming the efficacy of the system in super-resolution sampling and imaging performance.

CMOS photonic integrated source of broadband polarization-entangled photons

We showcase a fully on-chip CMOS-fabricated silicon photonic integrated circuit employing a bidirectionally pumped microring and polarization splitter-rotators tailored for the generation of broadband (>9 THz), high-fidelity (90–98%) polarization-entangled photons. Spanning the optical C+L-band and producing over 116 frequency-bin pairs on a 38.4-GHz-spaced grid, this source is ideal for flex-grid wavelength-multiplexed entanglement distribution in multiuser networks.

(Invited) Advanced III-V-on-Si Heterogeneously Integrated Platforms for Next Generation Silicon Photonics Integrated Circuits

The field of Silicon Photonics has experienced a solid and continuous progress over the last few years, gaining in technological maturity, design tools, and new methods [1]. The deployment of low-cost, compact, and power-efficient photonic circuits with a high wafer yield and robustness stands as one of the fundamental pillars that sustain such progress [2]. Presently, photonic circuit technology has diversified its number of available platforms. Despite the fact that indium phosphide (InP) [3] and silicon-on-insulator (SOI) platforms are still considered as the workhorses of integrated photonics in terms of maturity and deployment of active (InP) and passive (SOI) components, other alternatives such as germanium-on-silicon [4], silicon nitride-on-insulator [2] or hybrid solutions combining different functional materials with Si are gaining momentum [5]. A representative example is the heterogeneous III-V/Si platform [6], which has been used to develop compact photonic circuits with on-chip gain. Contrarily to the hybrid integration, the III-V-on-Si heterogeneous integration avoids the constraints of chip-to-chip alignment while enabling the simultaneous integration of hundreds of III-V gain chips in a scalable fashion. Still, the integration methods of such III-V materials on silicon need to be improved to attain the maturity level of the monolithic III-V platform, which benefits from a complete palette of technological solutions not yet available in the III-V-on-Si heterogeneous integration. More recently, an advanced heterogeneous scheme based on wafer-seed-bonding and epitaxial regrowth has emerged [7]–[9]. The ambition is to create a generic integration scheme combining the best offered by the III-V and the Si-photonics platforms. The regrowth capability gives access to the large epitaxial toolkit available in the conventional InP monolithic platform, where several epitaxial steps are often implemented [10][11]. To cite some of them, the epitaxial regrowth of III-V materials to bury III-V lasers bonded onto silicon are object of intense research nowadays to overcome the thermally inefficient buried oxide [12].In this paper, we will review the advances on III-V-on-Si heterogeneous integration through the implementation of several key demonstrators and building blocks for silicon photonics, including on-chip semiconductor optical amplifiers, lasers and electro-absorption modulators. We will discuss the progress and benefits of the direct seed bonding and regrowth as well as new device designs to improve the performance.

AC‐Driven Plasmon Waveguide Integrated Electroluminescent Device

AbstractPhotonic elements have greater information‐carrying capabilities compared to electronic components, making the development of high‐speed, chip‐integrated optical communication devices crucial for addressing interconnect bottlenecks in high‐speed computing systems. Significant progress has been made in studying light sources and their transmission. Despite these advancements, achieving a satisfactory integration of photonic circuits with both light source and optical waveguide functions has remained a challenge. Here, a silver nanowire (AgNW) waveguide‐integrated light source is demonstrated driven by alternating current (AC) voltage. AgNW functions as both the electrode and waveguide to simplify the device structure, and the 2D transition metal dichalcogenides (TMDs) monolayer acts as the gain media in device. The electroluminescence (EL) can be tunably modulated via adjusting the frequency and voltage of the AC in this device, and it is also successfully coupled into the waveguide with its transmission distance up to 25 µm. The experiments of electroluminescence and waveguide transmission of different materials and their interlayer excitons have made a preliminary exploration of multiple wavelength transmission. It provides a new research platform forward in the development of chip‐integrated optical communication devices, thereby hopefully addressing the interconnect bottleneck and unlocking the potential for enhanced performance and efficiency in high‐speed computing systems.

Power stabilization of a terahertz-frequency quantum-cascade laser using a photonic-integrated modulator

We demonstrate a technique to stabilize the emission from a 3.4-THz quantum-cascade laser against power drifts, using a recently developed photonic integrated circuit (PIC) structure formed by coupling a racetrack resonator with a ridge waveguide. This structure enables a dynamic power-control range of ±15% and locking over >600 s using a proportional–integral control loop. The resonator yields a 50% weaker perturbation to the laser emission frequency when compared with direct laser modulation, and hence offers the prospect of simultaneous, quasi-independent control of power and frequency. Progress towards integrating the PIC with a precision micromachined rectangular metallic waveguide module has also been demonstrated, with power modulation of the principal laser emission line being observed during pulsed operation.

Integrating deep convolutional surrogate solvers and particle swarm optimization for efficient inverse design of plasmonic patch nanoantennas

Abstract Plasmonic nanoantennas with suitable far-field characteristics are of huge interest for utilization in optical wireless links, inter-/intrachip communications, LiDARs, and photonic integrated circuits due to their exceptional modal confinement. Despite its success in shaping robust antenna design theories in radio frequency and millimeter-wave regimes, conventional transmission line theory finds its validity diminished in the optical frequencies, leading to a noticeable void in a generalized theory for antenna design in the optical domain. By utilizing neural networks, and through a one-time training of the network, one can transform the plasmonic nanoantennas design into an automated, data-driven task. In this work, we have developed a multi-head deep convolutional neural network serving as an efficient inverse-design framework for plasmonic patch nanoantennas. Our framework is designed with the main goal of determining the optimal geometries of nanoantennas to achieve the desired (inquired by the designer) S 11 and radiation pattern simultaneously. The proposed approach preserves the one-to-many mappings, enabling us to generate diverse designs. In addition, apart from the primary fabrication limitations that were considered while generating the dataset, further design and fabrication constraints can also be applied after the training process. In addition to possessing an exceptionally rapid surrogate solver capable of predicting S 11 and radiation patterns throughout the entire design frequency spectrum, we are introducing what we believe to be the pioneering inverse design network. This network enables the creation of efficient plasmonic antennas while concurrently accommodating customizable queries for both S 11 and radiation patterns, achieving remarkable accuracy within a single network framework. Our framework is capable of designing a wide range of devices, including single band, dual band, and broadband antennas, with directivities and radiation efficiencies reaching 11.07 dBi and 75 %, respectively, for a single patch. The proposed approach has been developed as a transformative shift in the inverse design of photonics components, with its impact extending beyond antenna design, opening a new paradigm toward real-time design of application-specific nanophotonic devices.

Design and numerical analysis of high contrast ratio and ultra-compact all-optical 2-bit reversible comparator

In this paper, a novel interference-based nanostructure was designed and simulated to realize an all-optical 2-bit reversible comparator by employing a novel technique. The plane wave expansion (PWE) method was adopted to analyze the encoder design and frequency modes. Aside from downsizing, the finite-difference time-domain (FDTD) method was utilized for the simulation and numerical analysis of the design proposed herein. An ultra-compact nanostructure with a 129.8 μm2 footprint was utilized for the all-optical 2-bit reversible comparator. One of the noteworthy characteristics of the proposed nanostructure was its excellent contrast ratio (i.e., 13.8 dB) in comparison to other nanostructures. The bitrate and delay time in this nanostructure were 3.33 Tb/s and 300 fs, respectively. Based on the findings of the simulations conducted at a central wavelength of 1.55 μm, it is recommended to employ the nanostructure proposed herein during the third telecom window. A photonic crystal nano-resonator was utilized to design the high-performance all-optical 2-bit reversible comparator, which may also be employed in integrated optical circuits (IOCs).

Hazelnut proteins detection in cookies using an immersible label-free photonic chip sensor

Rapid and sensitive methods to detect allergenic proteins in foods at the point-of-need could help to protect allergic consumers from life-threatening accidental exposure. In this context, the development of a label-free optical immunosensor for detecting hazelnut proteins in cookies is presented. The sensor is based on silicon photonic chips containing two integrated Mach-Zehnder interferometers (MZI) with light input and output on one side of the chip, and sensing window openings on the other side. Coupling with a white-light LED and a spectrometer for signal recording is achieved via a bifurcated fiber. The sensing window of one of the MZIs is modified with a mouse monoclonal antibody against hazelnut proteins, while the other sensing window is modified with bovine serum albumin to serve as blank. The assay follows a two-step sandwich immunoassay format, involving a 10-min reaction with the hazelnut proteins calibrator or cookie sample, followed by a 2-min reaction with a mixture of monoclonal antibodies. All reactions were performed by immersing the chip side where the MZIs windows are located into the solutions. The recorded spectrum is processed in real-time to transform the spectrum shifts due to immunoreactions taking place on the sensing windows of the two MZIs into phase shifts. The developed assay detects hazelnut proteins at concentrations as low as 25 ng/mL in calibrators prepared in buffer or in aqueous extract of hazelnut-free cookies. Additionally, cookies containing hazelnuts were analyzed demonstrating the sensor ability for fast and sensitive detection of the hazelnut proteins in commercial products.

Subwavelength Grating Cascaded Microring Resonator Biochemical Sensors with Record-High Sensitivity.

Photonic integrated circuit biochemical and biomedical sensors show promising applications in medical diagnosis, food security, healthcare, and environmental monitoring. Silicon-on-insulator subwavelength grating waveguides and cascaded microring resonator structures enhance photon-analyte interaction, offering superior sensing performance (higher sensitivity with lower limit of detection and larger free spectral range) compared to traditional strip and slot waveguide microring resonator structures. In this study, we design, simulate, and experimentally demonstrate a novel and compact biochemical sensor integrating subwavelength grating cascaded microring resonators and multibox subwavelength grating straight waveguides on a silicon-on-insulator platform. We achieve a record-high refractive index sensitivity of 810 nm/RIU with a limit of detection value of 2.04 × 10-5 RIU. The measured concentration sensitivity for sodium chloride solutions is 1430 pm/% with a limit of detection of 0.04%. The free spectral range is 35.8 nm, and the measured Q factor is 1.9 × 103. By combining the advantages of cascaded microring resonators with those subwavelength gratings, our sensor offers unprecedented sensitivity for biochemical sensing applications, promising significant enhancements in healthcare diagnostics and environmental monitoring systems.

Variable optical attenuator using a multi-path interferometer embedded in an optical cavity

Variable optical attenuators (VOAs) used in photonic integrated circuits suffer from trade-off between dynamic range of attenuation and large footprint and introduce inadvertent changes in optical phase while attenuating the input signal. In this work, we propose a VOA architecture using micro-ring resonators which provides ≥60dB optical attenuation of selected wavelengths, while preserving the input optical phase, in a compact footprint. It enables tailoring of the dynamic range and insertion loss for different circuit applications in photonic signal processing, programmable photonics, high performance computing, and datacenter communication.

Broadband and easily fabricated double-tip edge coupler based on thin-film lithium niobate platform

Due to the significant difference in spot size between lithium-niobate-on-insulator (LNOI) waveguide and fiber, there is a substantial coupling loss between them, which limits the interconnection and application of LN photonic chips. Although some LNOI edge couplers have been designed and fabricated with low coupling loss between chip and fiber, they require high etching accuracy. In this paper, a low-loss, broadband and easily fabricated edge coupler is designed and fabricated on the LNOI platform. To reduce the etching requirement for tip width, a design has been adopted with symmetrical double tips at the edge. Additionally, a multimode interferometer (MMI) is introduced to ensure large fabrication tolerance. The proposed design only requires a single full etching step, significantly reducing stringent requirements for process with high precise. Experimental results demonstrate that the device exhibits a coupling loss less than 3.5 dB/facet in the wavelength range of 1500–1640 nm, while achieving a low coupling loss of 2.71 dB/facet at 1550 nm.

Plasmonic-based electro-absorption modulator integrated with different dielectric materials at 1.55 µm wavelength

In this paper an electro-absorption modulator (EAM) based on plasmonic material along with different dielectric materials were investigated. The study focuses on the performance matrices of the device based on four different dielectric materials such as hafnium dioxide (HfO2), silicon nitride (Si3N4), boron nitride (hBN), and silicon dioxide (SiO2). A finite element method (FEM) is used to carry out the investigated results like insertion loss (IL), extinction ratio (ER), and figure of merit (FOM). The device yielded a very high ER of 35.42 dB/µm and an IL of 0.02 dB/µm at 1.55 µm wavelength. A modulation frequency (f3dB) of 0.46 THz, 1.50 THz, 2.39 THz, and 4.24 THz was achieved for the EAMs with HfO2, Si3N4, hBN, and SiO2as the dielectric material, respectively. Also, the investigated EAM with different dielectric materials has the potential to be integrated in Photonic Integrated Circuits (PICs).

Glass interposer for heterogeneous integration of flip-chipped photonic and electronic integrated circuits

Glass interposer for heterogeneous integration of flip-chipped photonic and electronic integrated circuits

Fast topological pumps via quantum metric engineering on photonic chips.

Topological pumps have garnered substantial attention in physics. However, the requirement for slow evolution speed to satisfy adiabaticity greatly restricts their application in on-chip devices. Here, we discover a direct link between adiabaticity and quantum metric, the real part of quantum geometry that has been relatively less explored compared to its imaginary counterpart, the Berry curvature. We demonstrate that the evolution speed of topological pumps between nontrivial edge states can be increased by reducing the quantum metric via introduction of long-range coupling to the celebrated Rice-Mele model. This fast topological pump can occur without affecting the bulk state evolution, which challenges the common understanding. We experimentally confirm our findings by using a platform consisting of bilayer integrated silicon waveguides operating at telecommunication wavelengths. Our work provides possibilities for lifting topological pumps from the constraints of slow evolution and paves the way toward compact photonic integration.

Symmetric silicon microring resonator optical crossbar array for accelerated inference and training in deep learning

Photonic integrated circuits are emerging as a promising platform for accelerating matrix multiplications in deep learning, leveraging the inherent parallel nature of light. Although various schemes have been proposed and demonstrated to realize such photonic matrix accelerators, the in situ training of artificial neural networks using photonic accelerators remains challenging due to the difficulty of direct on-chip backpropagation on a photonic chip. In this work, we propose a silicon microring resonator (MRR) optical crossbar array with a symmetric structure that allows for simple on-chip backpropagation, potentially enabling the acceleration of both the inference and training phases of deep learning. We demonstrate a 4×4 circuit on a Si-on-insulator platform and use it to perform inference tasks of a simple neural network for classifying iris flowers, achieving a classification accuracy of 93.3%. Subsequently, we train the neural network using simulated on-chip backpropagation and achieve an accuracy of 91.1% in the same inference task after training. Furthermore, we simulate a convolutional neural network for handwritten digit recognition, using a 9×9 MRR crossbar array to perform the convolution operations. This work contributes to the realization of compact and energy-efficient photonic accelerators for deep learning.

Piezoelectrically tunable, narrow linewidth photonic integrated extended-DBR lasers

Recent advancements in ultra-low-loss silicon nitride (Si3N4)-based photonic integrated circuits have surpassed fiber lasers in coherence and frequency agility. However, high manufacturing costs of DFB and precise control requirements, as required for self-injection locking, hinder widespread adoption. Reflective semiconductor optical amplifiers (RSOAs) provide a cost-effective alternative solution but have not yet achieved similar performance in coherence or frequency agility, as required for frequency modulated continuous wave (FMCW) LiDAR, laser locking in frequency metrology, or wavelength modulation spectroscopy for gas sensing. Here, we overcome this challenge and demonstrate an RSOA-based and frequency-agile fully hybrid integrated extended distributed Bragg reflector (E-DBR) laser with high-speed tuning, good linearity, high optical output power, and turn-key operability. It outperforms Vernier and self-injection locked lasers, which require up to five precise operating parameters and have limitations in continuous tuning and actuation bandwidth. We maintain a small footprint by utilizing an ultra-low-loss 200 nm thin Si3N4 platform with monolithically integrated piezoelectric actuators. We co-integrate the DBR with a compact ultra-low-loss spiral resonator to further reduce the intrinsic optical linewidth of the laser to the Hertz-level—on par with the noise of a fiber laser—via self-injection locking. The photonic integrated E-DBR lasers operate at 1550 nm and feature up to 25 mW fiber-coupled output power in the free-running and up to 10.5 mW output power in the self-injection locked state. The intrinsic linewidth is 2.5 kHz in the free-running state and as low as 3.8 Hz in the self-injection locked state. In addition, we demonstrate the suitability for FMCW LiDAR by showing laser frequency tuning over 1.0 GHz at up to 100 kHz triangular chirp rate with a nonlinearity of less than 0.6% without linearization by modulating a Bragg grating using monolithically integrated aluminum nitride (AlN) piezoactuators.

Narrow nano-beam generation inside of silicon photonic chip with a hemi-cylinder

A photonic nanojet (PNJ) can be shrunk using a high index dielectric micro-particle to generate a narrow nano-beam. Considering silicon has a high refractive index and low absorption in near-infrared waves, here, we present a narrow nano-beam generation inside of a silicon photonic chip with a hemi-cylinder. The proposed design consists of a silicon hemi-cylinder on a silicon substrate with a micro-hole. At first, we investigate the narrow PNJ inside of the silicon substrate produced by a silicon hemi-cylinder on a silicon substrate. In order to make the PNJ inside of the silicon substrate usable, the properties of the PNJ influenced by the position of the added micro-hole filled with water are studied. A narrow nano-beam can be obtained when the position of the added micro-hole is well-chosen. Then, the narrow nano-beam affected by the silicon hemi-cylinder immersed different mediums is presented. Finally, we discuss the influence of the perturbation of illumination angle on the properties of the nano-beam. The proposed structure may have potential applications in enhanced Raman spectroscopy and optofluidic chips.