Freeform Optics enables the design of more compact or more complex optical systems. Here, we will demystify designing with freeform surfaces, yet demonstrate their power in optical design. Designing for manufacture is important for affordability, and this is even more critical with freeform optics due to their complexity. We will seize the opportunity to also discuss the importance of concurrent engineering in designing freeform optical systems. Finally, we will briefly introduce a novel optical component, the metaform.

Adaptive illumination systems are capable of changing their emission pattern in a dynamic and flexible manner. Such systems can be realized with tunable optical components. We analyze the possibilities and limitations of phase-only spatial light modulators, implemented as a kind of programmable freeform optics, to realize adaptive illumination systems. First, the calculation of the required phase shift patterns to generate specific target irradiance distributions from arbitrary incident wavefronts, is elaborated. Second, the practical limitations of generating prescribed target patterns are experimentally tested and critically discussed.

Freeform metrology is an enabling technology for today’s research and advanced manufacturing. The Tilted Wave Interferometer is a full field measurement system for fast and flexible measurements. It is based on an off-axis illumination scheme based on a microlens array. In this contribution, we present a novel illumination system for the tilted wave interferometer, that allows to reduce the measurement time by a factor of four using parallelization based on wavelength multiplexing. Here we present a design solution that utilizes the flexibility of 3D-printing. The microlenses are realized as multi-order diffractive optical elements, providing a high efficiency compared to colorfilter based realizations. To boost the light efficiency of the novel illumination system further, a field lens functionality is added to the system by adding individual micro-prisms to each microlens. The system is manufactured by the use of grayscale two-photon polymerisation.

Since 1990, adaptive optics are used in astronomy to remove the effects of atmospheric turbulence, and then retrieve diffraction-limited images, even in bad seeing conditions. Thanks to its strong knowledge in Shack-Hartmann wavefront sensing and deformable mirror, Imagine Optic has developed a simple and affordable adaptive optics system for astronomers. We present the current prototype as well as first experimental results on both natural stars and extended sources, with the main goal of allowing an effective correction in all sky conditions regardless of the object.

Modern bio-imaging techniques such as light-sheet, multiphoton and PALM/STORM are now aiming to image more complex biological samples at larger depth and therefore face larger-amplitude and more complex aberrations. We provide an analysis of key requirements driving optimal implementation of adaptive optics (AO) in microscopy, with a focus on wavefront modulators. We show that some specifications of wavefront modulators such as linearity, hysteresis or actuators performance & layout can end up to better AO performance in microscopy systems, when specifically optimized for such use. We then provide design details and characterization results of a newly developed deformable mirror, and report on experimental images obtained from AO-enhanced microscopes based on the device, for several modalities such as light-sheet, multiphoton or super-resolution single molecule localization systems. Finally, we provide recommendations on how to define the right set of AO components, algorithms and overall method depending on modality, instrument and sample constraints.

Conventional approaches to microscopy record essentially two-dimensional images with a trade between transverse resolution and depth of field. Advances in computational imaging, using engineered point-spread functions have enabled an increase in depth of field, but generally with poor image quality arising from axial variations in the point-spread function. We report how the axial variations in Airy beams can be exploited to enable diffraction-limited, aberration-free 3D microscopy in a single snapshot for imaging of both 3D surfaces and 3D volumes. Localisation microscopy of point emitters enables microscopy with nanoscale nanoscale resolution and when implemented with various exotic point-spread functions this can be extended to 3D imaging. We will show how 3D locational microscopy based on Airy beams can enable much higher emitter densities, which enables the essential high-speed 3D measurement required in applications ranging through fluid dynamics, cardio-vascular monitoring and single-molecule imaging.

Extreme ultraviolet (XUV) light applications are still a very promising field that was heavily enlivened by the definition of the new wavelength for semiconductor lithography within the XUV range. But the detection of XUV light is also important for the exploration in the field of space science (i.e., monitoring the formation and evolution of solar storms) and high-energy physics (i.e., dark matter detection). The advancement of this technology mainly depends on the performance optimization of XUV sources, optical systems and related photodetectors. In this work, the optical design of a high flux XUV setup was simulated and defined to optimise the beam path which is the backbone of the initial evaluation process for the characterisation of luminescent materials under XUV irradiation. Additionally, the paper focused on the conceptualisation and realisation of the experimental setup as well as the alignment of the optical components and the detector calibration.

Concatenated backward ray mapping is an alternative for ray tracing in 2D. It is based on the phase space description of an optical system. Phase space is the set of position and direction coordinates of rays intersecting an optical line. The original algorithm is limited to optical systems consisting of only straight line segments; we extend it to accommodate curved segments. The algorithm is applied to the compound parabolic concentrator, a standard optical system that collects parallel light and reshapes it to a focused beam. We compare the accuracy and speed of the extended algorithm to the original algorithm and Monte Carlo ray tracing. The results show that the extended algorithm outperforms both methods.

We present a novel approach to computing reflectors with a scattering surface in illumination optics. A scattering model governed by a Fredholm integral equation is derived. Solving this integral relation yields a virtual specular target distribution, which we insert into a Monge-Ampère least-squares numerical solver to get a scattering reflector that yields the desired illumination.

We present a novel approach to minimize aberrations in imaging systems. The energy distributions at the source and target of an optical system play a crucial role in designing freeform surfaces through illumination optics methodologies. We quantify the on-axis and off-axis aberrations using a merit function that depends on the energy distributions. The minimization of the merit function yields optimal energy distributions, which subsequently enable us to design freeform reflector surfaces that cause the least aberrations. We validate our method by testing it for two configurations, a single-reflector system with a parallel source to a near-field target, and a double-reflector system with a parallel source to a point target.

-

Project iFLOWS aims to develop a novel framework for the effective and uninterrupted screening of postal/courier flows involving all actors across the transport chain. The main concept of iFLOWS is based on a multi-tiered approach to screening of letters and parcels, enhancing cross-organisation collaboration and intelligence and upgrading the threat, illicit material and dangerous substances detection process.

Our project IPN-Bio aims to foster and develop long-term international, interdisciplinary, and inter-sectoral collaboration between Europe, USA, Latin America, and China. IPN-Bio consortium consists of 13 world-leading organizations (4 EU/UK universities, 3 EU/UK companies, and 6 third country partner organizations) from four continents and eight countries working at the frontier of the field with the complementary expertise in the multidiscipline of Photonics, Nanotechnology and Biotechnology.

Transforming imaging with simultaneous light modulation

Many products and devices depend on imaging technology, from projection displays to remote sensors. The EU-funded DYNAMO project hopes to achieve a new paradigm in imaging techniques by creating spatial light modulators which can operate simultaneously. Conventional spatial light modulators operate sequentially: a beam of light is shaped into different patterns, and the time interval between patterns is governed by the refresh rate of the device. Instead, researchers propose sending all patterns in one short nanosecond pulse, creating a dynamic spatiotemporal light modulation device. This will result in ultra-fast imaging with a refresh rate for dynamic pixels equivalent to that of the GHz range.

Objective

Imaging technologies form the basis of a vast range of products and devices and improvements would have a huge impact both scientifically and commercially. We have identified a key bottleneck, how light is modulated in the imaging system, that we can unlock to achieve a new paradigm in imaging technologies. Spatial light modulators, and similar components, operate sequentially: the light beam is shaped in different patterns but the time interval between patterns is limited by the refresh rate of the device. We will remove this limitation, thereby creating a technological breakthrough; our advance will be to send all possible patterns of the device simultaneously, and encoded in a short nanosecond pulse, creating the concept of parallel beam shaping or dynamic spatio-temporal light modulation device. In DYNAMO, we will shape optical beams in two spatial dimensions plus the temporal one. The equivalent refresh rate of the dynamic pixel will start at GHz, although we are confident it will become much higher by the end of the project. To give an idea of our ambition, we compare this improvement in the time to process images with the improvement in the clock frequency of computers: the first general-purpose electronic computer, the ENIAC, had a clock frequency of 100kHz in 1945. It was not until 2000 where AMD reached 1 GHz in their computers. Processing images is broadly similar to processing data so this is indicative of the fifty-year acceleration in the realm of imaging that we will achieve. DYNAMO is an ambitious and integrated project that begins by studying the fundamentals of acoustic wave scattering and ends by developing ultra-fast imaging applications in optics. The success of this pathway requires the synergy of the disciplines of physical acoustics, photonics and imaging. The outcomes from this project offer to accelerate imaging technologies and place European science and industry at the forefront of the inventions and advances that will follow.

V4F aims to show proof-of-principle of a new technology capable of unprecedented control over interactions with specially synthesised targets to significantly improve the energy balance of aneutronic fusion reactions. New concepts and advanced simulations of inertial confinement of aneutronic fusion reactions and particle acceleration will inform pioneering experiments in high-energy matter-interactions. Results could offer the prospect of breakthrough increases in alpha-particle yields from fusion reactions and mitigate the instabilities found in conventional fusion reactions. This work offers the tantalising possibility of aneutronic fusion as a waste-free nuclear energy source and radical new configurations of particle accelerators, leading to an efficient positron beam acceleration. The results will benefit society with game-changing new approaches to clean, safe energy production and significant downscaling of positron accelerators with dramatic impacts in medicine, industry and fundamental science.

We present the experimental demonstration of the integration of group IV and III/V quantum well and quantum dot structures to silicon nitride, the co-integration of phase change materials on silicon nitride and post fabrication trimming capability of silicon nitride structures.

In this paper, we present controllable and efficient supercontinuum generation with multiple dispersive waves exploiting the quasi-phase-matching (QPM) condition in Si3N4 waveguide. The frequency component needed by QPM condition is introduced by varying the width of the waveguide through propagation. We demonstrated that integrated photonics is an ideal platform to apply QPM strategy, offering new opportunities to tailor the dispersive wave.

Inductively coupled plasma chemical vapor deposition was used to obtain thin films of silicon-rich silicon nitride with a refractive index of 2.44 at optical telecommunications wavelength. The resulting layer was patterned into a 1.6 μm wide waveguide and tested for its nonlinear behavior using a 90-fs all-fiber laser centered at 1630 nm. A significant spectral broadening is demonstrated with a supercontinuum generation from 1300 nm to 1985 nm. Simulations are in fair agreement with the experiments, assuming a nonlinear index of 2 x 10-18 m2/W.

Microresonator solitons enable high repetition-rate optical frequency combs. Their spectral span

scales inversely with the strength of the resonator’s anomalous group velocity dispersion. In a thick (800 nm) silicon nitride platform, wide resonator waveguides (>2 µm) with weak anomalous dispersion are especially promising for the generation of broadband spectra. As wider waveguides have a weaker evanescent field, the coupling strength to their bus waveguide is reduced. To address this challenge, alternative coupler designs, such as pulley couplers are required. Here, we investigate pulley couplers for wide waveguide specifically targeted at broadband soliton generation. We observe significant improvement of the coupling ideality compared to conventional coupler geometries and broadband four-wave mixing spectra are observed in 2.8 µm wide

microresonators.

We present designs and experiments of single-etched amorphous silicon (α-Si) surface grating couplers on a silicon nitride (SiN) waveguide, operating at the telecom (C-band) and datacom (O-band) wavebands. SiN-only grating couplers demonstrate experimental coupling loss of -3.9 dB at 1.55 μm wavelength with a 1-dB bandwidth of 37 nm. Utilizing the hybrid α-Si/SiN platform with subwavelength grating structure, we obtain improved coupling performances, with the optimized fiber-chip coupling loss of -2.0 dB and a 1-dB spectral bandwidth of 42 nm centered at 1.31 μm.

Artificial intelligence (AI) is a potentially disruptive tool for physics and science in general. One crucial question is how this technology can contribute at a conceptual level to help acquire new scientific understanding or inspire new surprising ideas. I will talk about how AI can be used as an artificial muse in quantum physics, which suggests surprising and unconventional ideas and techniques that the human scientist can interpret, understand and generalize to its fullest potential.

[1] Krenn, Kottmann, Tischler, Aspuru-Guzik, Conceptual understanding through efficient automated design of quantum optical experiments. Physical Review X 11(3), 031044 (2021).

[2] Krenn, Pollice, Guo, Aldeghi, Cervera-Lierta, Friederich, Gomes, Häse, Jinich, Nigam, Yao, Aspuru-Guzik, On scientific understanding with artificial intelligence. Nature Reviews Physics 4, 761–769 (2022).

[3] Krenn, Zeilinger, Predicting research trends with semantic and neural networks with an application in quantum physics. PNAS 117(4), 1910-1916 (2020).

We report the first application of the Machine Learning technique of data-driven dominant balance to optical fiber noise-driven Modulation Instability, with the aim to automatically identify local regions of dispersive and nonlinear interactions governing the dynamics. We first consider the analytical solutions of Nonlinear Schrödinger Equation – solitons on finite background – where it is shown that dominant balance distinguishes two particularly different dynamical regimes: one where the nonlinear process is dominating the dispersive propagation, and one where nonlinearity and second order dispersion act together driving the localization of breathers. By means of numerical simulations, we then analyse the spatio-temporal dynamics of noise-driven Modulation Instability and demonstrate that data-driven dominant balance can successfully identify the associated dominating physical regimes even within the turbulent dynamics.

Long-term forecasting of extreme events such as oceanic rogue waves, heat waves, floods, earthquakes, has always been a challenge due to their highly complex dynamics. Recently, machine learning methods have been used for model-free forecasting of physical systems. In this work, we investigated the ability of these methods to forecast the emergence of extreme events in a spatiotemporal chaotic passive ring cavity by detecting the precursors of high intensity pulses. To this end, we have implemented supervised sequence (precursors) to sequence (pulses) machine learning algorithms, corresponding to a local forecasting of when and where extreme events will appear.

The aim of kW-flexiburst is to develop a high-power Ultra-short Pulse (USP) laser generating bursts that can be arbitrarily adjusted in terms of burst repetition rate, intra-burst repetition rate, number of pulses per burst, relative intensities in the burst while maintaining 1 kW average power. This will be enabled by a radically new concept of seed oscillator, which offers the opportunity to work at GHz repetition rates.

This high power USP laser will be adaptable to efficiently process any material (metals, dielectrics, semiconductors) using a variety of laser parameters that can be continuously tuned from a few high energy pulses to a large number of pulses in a high repetition rate burst.

The flexible laser performance will be demonstrated in relevant industrial applications, which require high throughput/ high quality laser processing methods and therefore will benefit significantly from the high mean power and the tunable pulses provided by the kW-flexiburst system.

The selected applications span a wide range of industrial fields from micro-structuring of metals, ceramics and other dielectrics, drilling of hard substrates and cutting of transparent materials. Each of them carries the potential for significant or even disruptive improvements of the related industrial production process by employing the kW-flexiburst laser technology in combination with the beam delivery concepts and process methods proposed by the project.

Photoelectrochemical cells (PECs) that mimic photosynthesis belong to the group of direct systems for converting sunlight to stored chemical energy. Common to those is the potential to become more efficient and cost effective because, unlike indirect ones, they do not involve unnecessary steps such as the sunlight to electricity conversion. Despite their greater potential, there is yet no direct conversion device that works on any technological scale. Indeed, there seems to be a large barrier linked to a poor PEC efficiency in absorbing sunlight and driving the catalysis for water oxidation (WO) and selective CO2 reduction (CO2R) to carbon-based compounds to store chemical energy. In addition, most PEC designs incorporate non-abundant or highly toxic elements precluding their future use at a larger scale.

In LICROX we will implement a new PEC type incorporating three complementary light absorbing elements driving WO and CO2R. The latter consists of a tandem assembly that combines Cu nanocatalysts with molecular catalysts made of only abundant elements. BiVO4 photoanodes have been fabricated and incorporated in tandem structures with organic photovoltaics (OPV) providing sufficient photovoltage and photocurrent to drive the bias free CO2R reaction to C2 products, targeting ethylene. Several light trapping mechanisms have been incorporated, which have been proven to be very effective in boosting the light harvesting efficiency in thin film solar cells.

To accelerate the endeavor of converting the triple junction PEC proposed into a working technology for transforming light and CO2 into compounds capable of storing chemical energy, LICROX brings together an interdisciplinary team of scientists with a comprehensive expertise in materials chemistry, semiconductor physics, electrochemistry, and photonics from EPFL, TUM, ICIQ and ICFO. Designing a strategy by DBT to overcome societal resistance, LICROX will set the route for a new scalable renewable energy technology to be initially pushed towards an industrial implementation and commercialization by AVA, HST and a newly developed spin-off from ICFO.

In this talk, the overall LICROX project will be exposed, and the light management will be specifically targeted by ICFO's presenter.

With the total amount of worldwide data skyrocketing, the global data storage demand is predicted to grow to 1.75 × 10^14 GB by 2025. Traditional storage methods have difficulties keeping pace given that current storage media have a maximum density of 10^3 GB/mm3. As such, data production will far exceed the capacity of currently available storage methods. The costs of maintaining and transferring data, as well as the limited lifespans and significant data losses associated with current technologies also demand advanced solutions for information storage. Nature offers a powerful alternative through the storage of information that defines living organisms in unique orders of four bases (A, T, C, G) located in molecules called deoxyribonucleic acid (DNA). DNA molecules as information carriers have many advantages over traditional storage media. Their high storage density, potentially low maintenance cost, ease of synthesis, and chemical modification make them an ideal alternative for information storage. To this end, rapid progress has been made over the past decade by exploiting user-defined DNA materials to encode information. In our project we explored data storage relying on DNA nanostructures (as opposed to DNA sequence) as well as on other combinations of nanomaterials.

It is estimated that 10 % of global greenhouse gas emissions are related to cooling buildings and environments. With demand for cooling expected to grow tenfold by 2050, and the increasing frequency of extreme heat waves, improving the efficiency of cooling systems plays a critical role in addressing the global climate challenge. Passive Radiative Cooling (PRC) materials – an emerging technology that can cool to sub-ambient temperatures, even in direct sunlight, without using electricity – could be an efficient alternative to conventional systems to save energy and reduce heat gains. However, the lack of standardisation and guidance for testing PRC materials and their properties, along with no standardised methods for testing their real-world performance, are limiting their uptake.

The PaRaMetriC project aims at developing a metrological framework to classify and compare PRC materials, assessing and validating appropriate benchmark materials and laboratory testing methods. It will also focus on characterising the properties of PRC materials and develop modelling methods, setting standards for quality control and allowing long-term effectiveness to be evaluated. The project will also create protocols and best-practice guides for in-field testing and set up long-term tests across several sites to assess material performance under a variety of real-world conditions. This project will help drive innovation in PRC technology, producing more energy-efficient cooling to meet rising needs.

ENSEMBLE

Neuromorphic models are proving capable of performing complex machine learning tasks, overcoming the structural limitations imposed by software systems and electronic neuromorphic models. Unlike computers, the brain uses a unified geometry whereby memory and computation occur in the same physical location. The neuromorphic approach tries to reproduce the functional blocks of biological neural networks. In the photonics field, one possible and efficient way is to use integrated circuits based on soliton waveguides, ie channels self-written by light. Thanks to the nonlinearity of some crystals, propagating light can write waveguides and then can modulate them according to the information it carries. Thus, the created structures are not static but they can self-modify by varying the input information pattern. These hardware systems show a neuroplasticity which is very close to the one which characterize the brain functioning. The solitonic neuromorphic paradigm this work introduces is based on X-junction solitonic neurons as the fundamental elements for complex neural networks. These solitonic units are able to learn information both in supervised and unsupervised ways by unbalancing the X-junction. The storage of information coincides with the evolution of structure that changes plastically. Thus, complex solitonic networks can store information as propagation trajectories and use them for reasoning.

Machine learning has opened a new realm of possibilities in photonic circuit design and manufacturing. First, we describe our approach of using deep learning to optimize the multi-dimensional parameter space for hundreds of optical chips on a mask, resulting in homogeneity of performance in high volume applications. Second, we present our approach of using a support vector machine to predict the performance of optical devices by wafer probing. This approach eliminates the expensive and labour-intensive process of optical chip testing, and allows unprecedented control over the fabrication process, including in-situ monitoring of wafer fabrication and real-time process adjustments. The combination of these two approaches paves the way for accelerated adoption of photonics in high volume applications.

We employ one- and two-photon polymerizatin, i.e. flash-TPP printing, which is compatible with metal-oxid-semiconductor (CMOS) technology, to fabricate polymer-cladded and single-mode 3D photonic waveguides and adiabatic splitters. Our 3D technology is a major step forward towards the highly-interconnection required in optical neural networks, which removes the high energy dissipation of electronics and where 3D integration enables scalability that is challenging to realize in 2D.

Photonic neural networks are a highly sought-after area of research due to their potential for high-performance complex computing. Unlike artificial neural networks, which use simple nonlinear maps, biological neurons transmit information and perform computations through spikes that depend on spike time and/or rate. Through comprehensive studies and experiments, a strong foundation has been laid for the development of photonic neural networks. We have recently developed a large-scale spiking neural network, consisting of more than 30.000 neurons, which serves as a proof-of-concept experiment for novel bio-inspired learning concepts. This breakthrough is significant because it demonstrates the potential of using photonic neural networks for advanced computing and highlights the importance of incorporating biological principles into artificial intelligence research.

Photonic computing has attracted much attention due to its great potential to accelerate artificial neural network operations. However, the processing of a large amount of data, such as image data, basically requires large-scale photonic circuits and is still challenging due to its low scalability of the photonic integration. Here, we propose a scalable image processing approach, which uses a temporal degree of freedom of photons. In the proposed approach, the spatial information of a target object is compressively transformed to a time-domain signal using a gigahertz-rate random pattern projection technique. The time-domain signal is optically acquired at a single-input channel and processed with a microcavity-based photonic reservoir computer. We experimentally demonstrate that this photonic approach is capable of image recognition at gigahertz rates.

We experimentally demonstrate an autonomous, fully tuneable and scalable neural network of 350+ parallel nodes based on a large area, multimode semiconductor laser. We implement online learning strategies based on reinforcement learning. Our system achieves high performance and a high classification bandwidth of 15KHz for the MNIST dataset. Our approach is highly scalable both in terms of classification bandwidth and neural network size.

A high-performance photonic reservoir, which utilizes the injection locking effect in a

highly multimodal semiconductor laser, has been developed. This innovative design allows for fully parallel and high-bandwidth operation. The output of this system is projected in space and imaged onto a digital micromirror device, which provides a readout and facilitates the hardware integration of programmable output weights. By using a highly multimodal semiconductor laser, the injection locking effect enables a large number of modes to be simultaneously locked to the injected signal, resulting in high dimensionality of the reservoir, reducing the computational time and complexity. The use of a digital micromirror device provides a flexible readout, allowing the output to be programmed to suit a range of applications. The hardware integration of programmable output weights enables the system to be optimized for specific tasks, improving performance and reducing power consumption.

In this paper, we experimentally demonstrated a novel all-optical reservoir computer with an all optical input mask. The combination of the binary random masks and the time-stretched ultrashort pulses has increased the system's classification performance. Compared with the traditional digital masks, this method shows superior classification performance in spoken-digit classification tasks and eliminates the need for high-speed modulation for digital masks.

Deep neural networks are increasingly applied in many branches of applied science such as computer vision and image processing by increasing performances of instruments. Different deep architectures such as convolutional neural networks or Vision Transformers can be used in advanced coherent imaging techniques such as digital holography to extract various metrics such as autofocusing reconstruction distance or 3D position determination in order to target automated microscopy or real-time phase image restitution. Deep neural networks can be trained with both datasets simulated and experimental holograms, by transfer learning. Overall, the application of deep neural networks in digital holographic microscopy and 3D computer micro-vision has the potential to significantly improve the robustness and processing speed of holograms to infer and control a 3D position for applications in micro-robotics.

Modern optical microscopes, from super-resolved fluorescence to label-free mechanisms of contrast, are powerful instruments able to produce images that are rich sources of molecular information towards an unprecedented insight into the morphological and functional properties of biological cells at the nanoscale. Today we are in the realm of multimodal optical microscopy boosted by artificial intelligence that makes intelligent the microscope. Super-resolved fluorescence microscopy, incorporating photochemical parameters from brightness to lifetime, and non-linear approaches, like those associated with multi-photon excitation able to exploit intrinsic fluorescence and SHG/THG, is coupled to label-free polarisation methods like Mueller matrix microscopy, expanding the available data set. Such a data set is the core for developing an artificial microscope aiming to transform a label-free interrogation of the sample into a molecular-rich fluorescence-based image. The intelligent microscope is AI-guided through a computational core based on three modules based on a convolutional neural network (CNN) and a tensor independent component analysis (tICA) un-supervised machine learning within a supervised deep learning strategy having the ambitious target to create a robust virtual environment "to see "what we could not perceive before". An interesting case study is related to understanding the visualisation of chromatin organisation.

Optical approaches have made great strides enabling high-speed, scalable computing necessary for modern deep learning and AI applications. In this study, we introduce a multilayer optoelectronic computing framework that alternates between optical and optoelectronic layers to implement matrix-vector multiplications and rectified linear functions, respectively. The system is designed to be real-time and parallelized, utilizing arrays of light emitters and detectors connected with independent analog electronics. We experimentally demonstrate the operation of our system and compare its performance to a single-layer analog through simulations.

Constraint Bayesian optimization approach is used to optimize the beam pointing and spatial filtering of a laser beam using the capillary transmission and the output beam profile, as the optimization criteria. We have demonstrated that the developed method was able to robustly find the optimal laser parameters and it will be implemented in the SwissFEL UV photocathode laser in the future.

Encoded-diffraction hybrid systems—optical encoding and simple electronic decoding—offers feature distillation via model training. Additionally, the most faithfully reconstructed images are not the ones that are best classified. We parametrize our results with singular value decomposition (SVD) entropy, a proxy for image complexity.

Despite the breadth of scientific literature on micro- and nanoplastics (MNPs), a standardized procedure for detecting MNPs is still lacking so far, leading to incomparable results between published studies. This work innovatively proposed the combination of machine learning with advanced optical photothermal infrared (O-PTIR) spectroscopy to develop an efficient and reliable detection framework for MNPs. Spectra of MPs and non-MPs were first collected and inputted to build a classification model, based on which four important wavenumbers were selected. A simplified support vector machine (SVM) model was subsequently developed using the selected four wavenumbers. Good predictive ability was evidenced by a high accuracy of 0.9133. The developed method can improve speed as well as the reliability of results, having a great potential for routine analysis of MNPs, ultimately leading to the standardization of detection methods.

Silicon photonics has been largely developed as a platform to address the future challenges in several applications including datacom, sensing or optical communications, among others. However, the properties of silicon itself is not enough to overcome all limitations in terms of speed, power consumption and scalability. Especially, silicon exhibits strong two photon absorption (TPA) and is centrosymmetric limiting the use of its nonlinear optical properties including Kerr and Pockels effects. New strategies have then been encouraged based on the heterogeneous integration of new materials in the silicon photonics platform. In this perspective, the recent trends in the development of new materials with optical properties complementary to the silicon and silicon nitride properties will be presented. In particular, the presentation will focus in a new integration approach of doped crystalline-oxides on silicon and silicon nitride photonics platform. Especially, the integration of doped-zirconia (ZrO2), compatible with CMOS technology has been developed. Indeed, such an oxide exhibits linear and nonlinear optical properties suitable to address the challenges of silicon photonics: low propagation loss, no two-photon absorption (TPA) due to its large bandgap energy, a large transparency window from the ultraviolet to the mid-infrared, Kerr and Pockels effect, and light emission.

We study the antimony trisulfide’s (Sb2S3) linear optical properties for potential applications in reconfigurable chip-based supercontinuum mid-IR sources. We experimentally demonstrate that Sb2S3 cladding on SiGe-on-Si waveguides induces relatively low extra propagation loss below 1 dB/cm between 3.3 and 3.9 μm wavelength.

Since their first demonstration more than 15 years ago, subwavelength metamaterials in silicon photonic devices have attracted increasing research interest while also breaking into commercial applications. We will discuss recent advances in this research field, in particular novel components and circuits for beam steering applications, on-chip filtering and quantum optics. The use of Mie resonant particle chains as on-chip waveguides has only recently been demonstrated and is opening the door to a new and exciting branch of integrated metamaterials research. We will review the early work in this area.

Recently, we successfully realized amorphous silicon carbide (a-SiC) integrated photonics with

optical losses as low as 0.78 dB/cm. Moreover, the deposition of a-SiC was done at 150 ℃, which enables

successful lift of a-SiC as an additive step to existing photonics circuits. In this work, we present an adiabatic

taper coupler which provides bidirectional lossless connection between two integrated photonics platforms:

thin-film silicon nitride (Si3N4) and a-SiC. Normalized power transmission of 96.61% is presented, and the

coupler enables strong confinement when coupling from weakly confined thin-film device to normal thickness

device. By utilizing such a coupler as bridge, switching back and forth between Si3N4 and a-SiC platforms

can be easily realized. This allow us to carry out applications including quantum interference and digital

Fourier spectroscopy, in which long optical delay lines are constructed on Si3N4 and highly integrated circuits

are built on a-SiC.