Neuromorphic photonics combines the advantages of photonics with the computational power of neural networks to create novel reconfigurable processing devices. Doing so enables new applications which are difficult or impossible for conventional digital electronic or RF processors to handle. This talk will highlight recent progress in neuromorphic photonic integrated circuits (PICs), beginning with photonic neurons which integrate both the linear and nonlinear functionality. It will then cover recent demonstrations utilizing PICs, including model predictive control (MPC), RF blind source separation (BSS), nonlinearity compensation in long-haul communications, and RF fingerprinting.

Subwavelength materials have become a fundamental tool for silicon photonic design, enabling devices with unique performance characteristics. We will briefly review some fundamentals here and will then discuss some of the latest advances in the field, with a particular focus on polarization handling. Furthermore, we will discuss advances in integrated optical sensing, addressing both fundamental issues such as the optimization of detection limits, as well as state-of-the-art results with novel sensing architectures. We will also discuss which benefits subwavelength structures can provide in such sensors.

In this invited presentation, we will show our most recent results on the use of subwavelength engineering of photons and phonons in suspended and non-suspended silicon optomechanical devices. These results open exciting new perspectives to exploit optomechanical effects with scalable silicon photonics technology, having great potential for applications in optical and wireless communications, radar, sensing, metrology and quantum technologies.

The recent progress of artificial intelligence (AI) has boosted the computational possibilities in fields where standard computers are not able to perform. The AI paradigm is to emulate human intelligence and therefore breaks the familiar architecture on which digital computers are based. In particular, neuromorphic computing, artificial neural networks (ANN) and deep learning models mimic how the brain computes. With this respect, photonics is a suitable platform to implement ANN hardware thanks to its speed, low power dissipation and multi-wavelength opportunities. One photonic device candidate to perform as an optical neuron is the optical microring resonator. Indeed microring resonators show both a nonlinear response and a capability of optical energy storage, which can be interpreted as a fading memory. Here, we describe the physics of silicon microring resonators and of arrays of microring resonators for application in neuromorphic computing. We describe different types of ANNs from feed-forward networks to photonics extreme learning machines and reservoir computing. In addition, we discuss also hybrid systems where silicon microresonators are coupled to other active material

Photonic neural networks based on semiconductor lasers hold a big promise for future, high performance computing at ultra-high bandwidths. We integrate such networks in a variety of semiconductor laser structures, demonstrate their scalable photonic integration. Finally, we apply our fully implemented neural networks to technologically and commercially relevant tasks in optical telecommunication, demonstrating their realization in realtime and without active involvement of classical digital electronic computers.

As part of the European Defense Agency-funded PICTURE project, we have investigated the potential of integrated photonics for radar systems, targeting a photonics-based architecture for electronically scanned active array antenna systems, including multifunctional signal generation and detection. In this context, we have developed integrated photonic transmit and receive modules with up- and down-conversion capability, powered by a mode-locked laser for frequency reference. Key elements such as narrowband optical bandpass filters are evaluated for different technology platforms (silicon-on-insulator and silicon nitride). Perspectives and lessons learned are presented, based on the evaluation of system performance with IF-to-RF and RF-to-IF conversion efficiencies.

SiC has emerged as a highly promising material for harnessing nonlinear processes, attributed to its inherent χ(2) and χ(3) nonlinearities. The recent introduction of SiC-on-insulator nanowaveguides with high-index contrast has enabled efficient dispersion engineering and strong light-matter nonlinear interaction. This advancement has consequently led to various nonlinear phenomena, including the generation of Kerr combs, supercontinuum, and second harmonic signals. Notably, none of the existing demonstrations have combined χ(2) and χ(3) nonlinear processes within a single device. In this study, we present a dispersion-engineered 4H-SiC-on-insulator nanowaveguide, which not only allows for the generation of octave-spanning supercontinuum through dispersive wave generation but also enables efficient frequency doubling of this dispersive wave. Our demonstration shows that the 4H-SiC-on-insulator is a versatile integrated platform for compact, multifunctional nonlinear devices.

Photonic integrated quantum circuits could fully unlock quantum systems' potential. However,monololithic platforms face several challenges. We explore hybrid quantum photonic integration, combining different photonic materials to mitigate monolithic approaches disadvantages . Our work introduces hybrid circuits integrating different quantum sources and materials. We present a technique integrating nanowire single photon sources with silicon nitride waveguides, enabling circuit reconfiguration and emission wavelength control using heat, and strain tuning which is compatible with superconducting single photon detectors . We realize a full quantum transceiver on chip using on demand single photon sources and SNSPDS, in additio to dispersion engineering of superconducting transmission lines for SNSPDs multi-pixel integration. We discuss progress towards room-temperature quantum integrated photonics using hexagonal boron nitride quantum emitters. Finally, we delve into scalable generation and quantum state reconstruction of W-states and construction using phase retrieval algorithms, and ongoing research on topological photonics, and photonic lattices.

One of the strong potentials of photonic quantum computing consist in the possibility to leverage matured technologies, currently implemented in several industrial domains. Integrated photonic, one of the most prominent, has already been implemented in proof of principle quantum computing platforms demonstrating its viability for high precision control of large number of qubits. Nevertheless, the tools required to generate, manipulate and detect single and entangled photons are not part of standard fabrication processes; such integrated elements still require extensive research work before integrated photonic circuit could be implemented in large scale systems. We present quantum computing protocols and a dedicated hardware architecture that takes advantage of photonic modularity and integrated photonic circuit with simplified architectures.

Quantum photonics has emerged as a key driver for advancing applications, such as quantum communication and sensing, that harness the potential of quantum effects beyond classical capabilities. Photonic integrated circuits favor such implementations by providing low loss and interferometric stability. Nonetheless, to fully leverage these advantages, two major challenges need to be addressed: highly efficient fiber-to-chip light coupling and waveguide-integrated single-photon detection. In this work, we utilize subwavelength grating (SWG) metamaterials to design (i) novel off-chip couplers enabling sub-decibel coupling efficiency and (ii) a new superconducting nanowire single-photon detector (SNSPD) concept featuring enhanced photon absorption and fast detection.

We demonstrate the co-integration of SiN waveguide-based building block components in an established 200 mm Silicon photonics platform, extending its capabilities to the 1 µm wavelength range, and addressing bioimaging applications based on Optical Coherence Tomography (OCT). Measured coupling and propagation loss, at wafer-scale, are low for edge couplers (< 2 dB), and PECVD SiN waveguides (< 0.4 dB/cm) respectively. SiN waveguide-coupled Ge-on-Si photodetectors with 0.6 A/W responsivity at 1060 nm and dark currents lower than 15 nA at -1 V operating voltage were achieved at wafer-scale, with a smaller footprint compared to standard Si-based photodetectors. This platform has the potential to be used for a variety of applications, such as optical communications, imaging, and sensing. The achieved results at 1060 nm pave the way for the deployment of miniaturized OCT-based devices to enable large scale patient diagnosis.

Optics and photonics already occupy a prominent position in the instrumentation of life sciences. For molecular diagnostics, DNA sequencing, cellular imaging often sensitive optical equipment is deployed. These instruments, however, are often bulky and expensive. In addition, the classical optical schemes do not lend themselves to parallelized detection schemes, as required in high throughput applications. Silicon photonics has seen a remarkable rise in the last decades, allowing indeed to integrate complex optical functionality on a small footprint, including on-chip routing and manipulation of free-space radiation. Life science applications, however, usually require visible wavelengths where silicon is not transparent. We have been developing a back-end CMOS compatible integrated photonics platform for visible applications based on silicon nitride. We will review applications of the platform in molecular sensing, cytometry, spectroscopy and imaging.

In this work, a 4D microscopic imaging system was developed for simultaneously 3D topography and hyperspectral measurement. The 4D imaging system exhibited excellent performance, with lateral and axial resolutions of 10 μm in 3D space, and a spectral resolution of 3 nm. Sample experiments were performed on a corn seed. A 2D true color image was fused onto the 3D topography, presenting the seed surface in 4D manner. Additionally, fluorescent hyperspectral imaging was conducted, revealing the different activities of germ and endosperm of the corn seed.

In this talk I will review our recent results in the field of optical parametric oscillation. In particular, the demonstration of Degenerate Optical Parametric Oscillation (DOPO) using a phase-coherent dual-tone pump scheme will be discussed. This work highlights the potential of photonic molecules as a platform for achieving high-quality DOPO, which could have important applications in areas such as quantum communication, quantum cryptography, and quantum computing. The phase-matching conditions are carefully controlled to achieve maximum DOPO efficiency, and the results show that the DOPO signal can be efficiently generated and optimized. This work opens new avenues for further investigation in the field of DOPO and its applications.

Innovative photonic devices and systems aim at achieving simultaneously a large scale of integration and high performance, leading to complex designs based on metamaterials and non-trivial geometries characterized by a large number of geometrical and material parameters. At the same time, modern cutting-edge designs usually involve multiple deterministic and stochastic figures of merit that account for both performance metrics and fabrication requirements, thus complicating the selection of the final design candidates. In this invited talk, we will discuss the potentiality of combining multi-objective analysis and optimization tools with machine learning techniques for the design of highly performing photonic devices and systems.

The advent of the multimode photonics era requires new devices for on-chip modal control. One of the most important applications is the simultaneous and uniform excitation of pairs of distinct modes. This represents a fundamental key-step for nonlinear multimode operations, where couple of different spatial modes should be excited in-phase and should preserve the temporal synchronisation. In this work we design, fabricate and test integrated silicon nitride asymmetric tapers, whose central longitudinal axis is shifted with respect to the axis of the corresponding waveguide. By controlling the centre-to-centre shift S between taper and waveguide, we achieve simultaneous and in-phase excitation of both the TE00 (TM00) and the TE01 (TM01) mode, with a power ratio that depends on S.

The integration of III-V compound semiconductors on a silicon platform has emerged as a transformative approach to enhance the performance and functionality of photonic and optoelectronic devices. This paper presents recent achievements, challenges, and future prospects of GaAs monolithic integration on silicon with a specific focus in the development of near-infrared (NIR) emitters and photodetectors. We address the challenges associated with III-V monolithic integration on silicon and its compatibility with CMOS processes. These challenges include lattice mismatches, thermal management, and process scalability. We present our latest results obtained on near infrared resonant cavity enhanced photodetector and light emitting devices integrated onto a nominal Si(001) substrate. The devices structures have been optimized by incorporating active regions based on InGaAs/GaAsP strained-layer superlattices and GaAs/AlGaAs distributed Bragg Reflectors. We also show an alternative solution to fabricate low-threshold emitters based on III-As membranes and lateral injection devices.

We report on the design, fabrication, and characterization of the first photonic crystal (PhC)-based red multiple-quantum-well (MQW) color converters fully optimized for augmented reality (AR) microdisplays through a quasi-3D light harnessing principle. This principle leverages an aluminum (Al) bottom reflector and a silicon dioxide (SiO2) gap to harness the bottom-emitted light, along with copper (Cu) lateral mirrors and a silicon nitride (SiN) phase-matcher for Bloch-mode replication. These structures were designed using 3D-FDTD simulations. As a proof-of-principle, we fabricated corresponding devices that exhibit promising characteristics, including record light extraction efficiencies over 40% for 4µm pixels and emission patterns with high directionality. Time-resolved photoluminescence (TRPL) analyses, along with a four-wave intensity model developed in this work, indicate that there is still room for improvement. We believe that the guidelines established in this study could pave the way for the use of MQW color converters in the next generation of very bright, high-resolution RGB microdisplays for AR glasses and beyond.

Tunable lasers and photonic integrated circuits are a promising technology to provide compact and high performance solutions for coherent remote sensing applications such as Lidar, and distributed acoustic fiber sensing (DAS). A hybrid tunable laser was fabricated within the EU funded INSPIRE project, based on the micro-transfer printing of a pre-fabricated InP gain section on the IMEC low-loss silicon nitride platform. By simultaneously modulating the laser SOA current and Vernier ring resonators, we demonstrate a 20 GHz chirp amplitude, while maintaining a <5 kHz linewidth. DAS measurement with this laser are presented.

Integrated widely tunable lasers are commonly used in communications. However, applications in, e.g., Lidar, fiber sensing or gas spectroscopy, have different requirements, such as continuous sweeping or low-noise operation. In this paper, I will discuss novel ideas for continuous sweeping, potentially eliminating mode-hopping, and options for low-frequency-noise operation.

We review our recent results on integration of lasers, detectors and microfluidic channels on silicon and germanium based mid-infrared photonic circuits, as well as initial sensing experiments for medical and gas sensing applications. Wideband devices, such as waveguides, couplers, splitters and detectors are required for mid-IR absorption spectroscopy. We show that such Si and Ge devices can operate over large spectral bands.

We present a framework for designing high-performance Time-Stepped Optical Frequency Comb (TSOFC) lasers comprising: (i) Thin Film Lithium Niobate (TFLN) Integrated Circuit (IC) for high-bandwidth photonic devices; (ii) Semiconductor Optical Amplifier for optical gain; (iii) silicon CMOS IC for drive circuits; and (iv) three-dimensional integration for dense connectivity between CMOS and TFLN. We target high-resolution Optical Coherence Tomography (OCT) as an example application; our TSOFC design switches between 128 optical wavelengths at 3 GHz switching rate (to meet lateral, axial, and temporal resolution requirements) with <2 Watts total system power consumption (using a 180 nm technology node for CMOS IC).

Due to explosive growth of internet traffic during past decades, there is an imminent need for scalable technologies that can enable both high-speed and low-power consumption requirements of today’s optical networks. Silicon photonics (SiPh) has emerged as the best solution to address these challenges. SiPh-based products are solid state devices that are built via standard CMOS fabrication methods that convert between optical and electrical transmission states. These products need to be reliable for long-term use in datacenter and communication networks as well as for novel applications such as AI, sensing and edge computing. Passive SiPh components such as waveguides, couplers, polarization splitters/rotators and multiplexers/de-multiplexers have different qualification and reliability test methods and requirements versus active SiPh components such as integrated lasers, modulators, photodiodes, and thermal tuners. To better understand and predict the potential failure mechanism and the micro-optic lifetime of the various devices that are monolithically integrated into SiPh chips, there is a need for well-defined qualification and reliability protocols across the industry.

Metamaterials are a new, emerging class of high-performance materials that derive their unique, physical properties from the way they are structured. In this presentation, I will focus on the creation of 2-dimensional metamaterials (i.e. metasurfaces) by nanopatterning glass, semiconductor and metal films. I will argue that these metasurfaces are ideal building blocks for the next generation of optical elements and optoelectronic devices. I will highlight how metasurface functionalities can start to impact a variety of optical sensing, imaging and display technologies. For example, I will show how one can create transparent optical sensors on glass substrates that can extract valuable information from an optical scene. I will also discuss the use of integrated metasurfaces for new imaging modalities. The proposed optical elements can be fabricated by scalable fabrication technologies, opening the door to many commercial applications.

Light scattered by subwavelength dielectric nanoantennas can be manipulated by engineering the antennas' Mie resonances. Spectral overlapping of the dipole-like electric and magnetic Mie resonances results in suppression of backward scattering. This phenomenon is known as Forward Kerker effect which enables the individual nanoantennas to act as Huygens' scatterers. A periodic chain of such Huygens' scatterers may effectively act as a so-called Huygens' metawaveguide with interesting new optical properties compared to non-resonant subwavelength grating waveguide counterparts. Inspiring by these recent findings, in our current research we are investigating more complex integrated photonic components, e.g., micro-ring resonators. In this talk, first I will explain the Forward Kerker effect of the resonant dielectric nanoantennas followed by the functionalities of a straight Huygens' metawaveguide. Next, I will demonstrate our high-Q Huygens' meta-ring-resonators and its different aspects. In conclusion, I will brief on the potential future applications of our demonstrated meta-ring-resonator in nonlinear and quantum integrated photonic systems.

We demonstrate a novel optical phased array (OPA) technology of all-solid state LiDAR module in autonomous vehicles. The OPA chip is designed a reflective structure with electro-optical modulation phase array. A key feature of this design is to apply the reflective multi-dielectric layer and bipolar NPN electronic control structure in each nano-resonator. The result enables to exhibit wide FOV of 30° × 30° and high-resolution of 0.2° × 0.2° at estimation rate of 1GHz to adjust independently the amplitude. The newly developed OPA chip is beneficial for using in 3D sensing, optical communication, and biomedical imaging.

With the blooming of artificial intelligence technology, the distant sensing systems with an agile deployment, a high refreshing rate and a high precision are demanded to map the fast-changing environment in terms of 3D point cloud for autonomous vehicles and smart home products. Frequency modulated continuous wave (FMCW), as a coherent detection method, shows a high precision, a high selectivity against ambient light, the intrinsic capability in speed sensing and eye-safety. However, FMCW usually needs a discrete Fourier transform to resolve the distance from the beating between the transmitted and received signals, leading to limitation on the acquisition rate and higher requirement in the frequency chirping span, speed, and linearity of the FMCW laser. Here we demonstrate an electro-optic comb-enabled absolute distance ranging technology with an acquisition rate up to 10 kHz and a precision better than 100 mm. We show that real-time distance resolving can be achieved with only two adjacent interference peaks among many peaks generated over one modulation frequency chirping period, boosting the acquisition rate of LiDAR system for ambiguity-free distance measurement.

Photonic chips are becoming increasingly complex, combining even more optical building blocks on the same chip. With this growing complexity we also see an expanding need for, and use of electrical tuning. This imposes opportunities, as photonic circuits can now become reconfigurable at run time, even to the point of creating arbitrary connectivity between functional building blocks, serving as a general-purpose optical processor. But at the same time, large-scale configurability comes with some tremendous challenges in terms of power consumption, electrical and optical packaging, driver electronics and control algorithms. We will discuss our recent progress in these domains in our path to building general-purpose programmable photonic chips. Expanding silicon photonics with high-efficiency electro-optic tuners, high-density packaging solutions, and electronics and software layers to govern the behavior of these photonic circuits that can be used for both photonic and microwave analog signal processing.

In this talk, I will discuss different PCMs that can be used in conjunction with silicon and silicon nitride photonics, to create reconfigurable optical switches for visible and infrared wavelengths. I will also talk about different heaters that are needed to actuate the phase transitions on-chip. Specifically, I will show how using ultrathin graphene as a heater element can provide very high energy-efficiency, close to the fundamental limit set by thermodynamics.

Phase change materials (PCMs) are currently revolutionizing nanophotonics by providing ways to tune and reconfigure optical functionalities without any moving parts. Building on this phenomenon, the last decade has witnessed many exciting reports of novel devices exploiting PCMs such as for example beam-steering, tunable light emission, reflection and absorption, programmable metasurfaces and reconfigurable neural networks. A large majority of the first studies were using standard PCMs such as GeSbTe as simple binary on-off switches, in which the ON state is the amorphous phase and the OFF state the fully crystalline phase. However, PCMs present another degree of freedom for tunability: the possibility to encode multilevel non-volatile states via partial crystallization. Furthermore, recently a new class of low-loss PCMs emerged (e.g. Sb2S3 and Sb2Se3), with negligible optical absorption in the near-infrared.

In this paper, we will present recent results on methods to program PCMs into various multilevel states of crystallization. We will then present nanophotonic devices leveraging this multilevel programming and conclude on the perspectives for this technology.

Phase change materials (PCMs) are commonly used in rewritable optical disks and memory devices. Recently, there have been efforts to incorporate PCMs into optical components and photonic circuits for developing reconfigurable optics, which necessitates the use of larger-scale PCMs. However, enlarging PCM-based devices has proven challenging due to difficulties in the switching mechanism, which demands significant thermal energy density. In this study, through computational models and experimental observations, we explore the effective parameters in phase transformation of Ge2Sb2Te5 and Ge2Sb2Se4Te, with areal length scale on the order of tens of micrometers. Our findings offer insight regarding the development of next-generation adaptive optics such as filters, zoom lenses, and beam steering devices.

We present fully-reconfigurable photonic integrated circuits (PICs) based on liquid-crystal-cladded slab waveguides. Our design utilizes a single-mode slab waveguide with a thin nematic liquid crystal (LC) cladding layer. We achieve local effective index modulation by controlling the LC’s director’s direction through electrodes integrated beneath the slab. As a proof of concept, we demonstrate a straight waveguide at 1550nm with voltage-controlled activation. The reconfiguration time and the excess loss due to LC were measured as 0.1 ms and 7.3 dB/cm, respectively. The low-temperature fabrication process enables cost-effective PICs on CMOS chips for 3D sensing, telecommunications, and data processing applications.

Selective control of light is essential for optical science and technology with numerous applications. I will review nanophotonic waveguides and integrated couplers used for selective coupling and spatial control of multiple degrees of freedom of light. I will highlight our recent development of hybrid nanophotonic circuits for optical logic operations and on-chip information processing. I will present the design and application of an OAM-controlled hybrid nanowire circuit for optical logic operations including AND and OR gates, as well as of an active hybrid nanophotonic circuit that enables nonlinear optical selectivity based on the spin and OAM of light.

Single solid-state quantum emitters have demonstrated considerable potential for the implementation of important quantum photonic devices such as on-demand single-photon sources or deterministic quantum logic gates. Converting a bare quantum emitter into a device with sufficient performance for use in quantum photonic systems requires an efficient, high cooperativity interface to accessing optical fields. Such interfaces may in principle be implemented through nanophotonic geometries that leverage cavity quantum electrodynamics effects with strongly confined optical modes. A number of critical factors, however, such as preservation of the emitter coherence post-fabrication, and fabrication yield and scalability, must be simultaneously addressed, imposing significant challenges to device development. In this talk I will describe our past and current efforts to target such requirements through nanophotonic design, towards the development of scaled chip-integrated quantum photonic systems with functionality enabled by quantum emitters.