To improve data security and authentication, we designed and fabricated a cryptographic key generator based on a photonic integrated circuit that converts a digital input key into a digital output key by means of a physical unclonable function. The physical unclonable function is realised by an imperfect multi-mode interferometer controlled by low-voltage MEMS phase shifters with very low power dissipation. The MEMS phase shifters are fabricated using our recently developed MEMS-on-PIC technology. A cross-platform approach applicable to all common material platforms used in integrated photonics, enabled by a combination of a protective interfacial layer and a sacrificial layer technology.

Photonic integrated circuits are a promising technology for quantum applications, which are known to impose stringent requirements on the performance characteristics of utilized components. Besides achieving high efficiency of active photonic components, low optical losses of waveguiding and coupling structures are of the same importance. In this contribution we focus on the analysis of optical losses related to waveguiding of single photons generated by InAs quantum dots in GaAs strip waveguides. We perform a simulation study of the effects of GaAs waveguide nanoscale surface roughness on the waveguide propagation losses. This study is also supported by experimental data on line edge roughness and surface roughness of fabricated GaAs waveguides determined from SEM and AFM analyses. The roughness applied in simulation is based on the statistical properties of this data. The results of our analysis strengthen our understanding of scattering losses and their individual contributing factors. We also conclude that for the investigated GaAs waveguides the contribution of scattering on the waveguide top surface roughness to the propagation losses is very small compared to the contribution of sidewall scattering.

We developed and optimized a double-layer edge coupler design and fabrication process to ensure the compliance of low-loss photonic wire bonding on the lithium niobate platform under the collaborative effort between Vanguard Automation (VA) and Swiss Center for Electronics and Microtechnology (CSEM).

This paper presents a novel design of grating couplers in the SiPh platform, as part of the development of a gas sensing system, that measures specific gas concentration through on-chip absorption based spectroscopy in the mid-infrared. The grating is based on 400 nm SOI rib waveguide platform, targeting gaussian emission profile. We investigate both uniform and non-unifοrm grating designs. We show maximum coupling efficiency optimized by 65% at the center wavelength of 4.3 μm, and broadband response. A tolerance study investigating the Si height and etch depth variations indicate very good tolerance to fabrication imperfections across the investigated spectral range of 4.25-4.35 μm. The proposed methodology can be used to engineer grating coupler emission profiles for other applications and platforms.

Devices engineered for slowing light, utilizing one-dimensional grating waveguides and fabricated from silicon nitride, often necessitate large footprints to secure the required delay, a consequence of the material’s inherently low refractive index. Our approach employs a genetic algorithm to optimize 100×100nm^{^}2 etchings on a predetermined grating waveguide topology, allowing for either the selective guidance of peak pulse intensity of the output or the augmentation of true time delay within the identical unit length. Within the chosen predetermined topology, the optimal configuration was identified based on the properties of the signal excitation in the time domain. This approach significantly facilitates the application-specific selection of peak intensity decay rate and time delay behavior within a 1D grating waveguide system.

Second-order nonlinearity can be demonstrated in popular CMOS materials such as silicon and silicon nitride by breaking the centrosymmetry of their crystalline structure. We present a detailed theoretical investigation of the electric field-induced second harmonic generation (EFISH) in silicon nitride waveguide. Up-and-down frequency-conversion operations are achieved through the periodic spatial distribution of metal electrodes around the waveguides, which induces an effective second-order nonlinearity and high-efficiency SHG. We use a computational model to numerically simulate the EFISH process inside the waveguide and calculate the charge carriers responsible for the electric field creation. The SHG efficiency is massively dependent on both the waveguide dimensions and the electric field creation. We obtain second-order electric susceptibility (χ(2)) up to 0.1735 pm/V with spectral bandwidth of around 4 nm.

Within the Interreg North-West Europe project OIP4NWE, we have set up an open innovation next-generation pilot line for InP-based photonic integrated circuits (PICs) to improve PIC manufacturing and make packaging more reliable and affordable. We present the service offering within the different components of the pilot line, consisting of InP PIC manufacturing in the front-end, and external interfacing optics and packaging in the back-end. We also report on how we supported 6 European SMEs with an innovation support voucher with the OIP4NWE pilot line.

In this work, we propose and implement a bi-junction depletion-type silicon electro-optic phase shifter. The phase shifter has a lateral profile of implants, that closely resembles that of a common bipolar junction transistor, and thus, has two polarities. These are acceptor-donor-acceptor (PNP) as well as donor-acceptor-donor (NPN). We realize both variants in IMEC ISIPP50G open-access silicon photonic technology and compare them to lateral and interleaved phase shifters. Both PNP and NPN phase shifters exhibit a V_πL_π figure that is at least 14.47% and up to 45.1% lower than that of the lateral and interleaved phase shifters realized in the same technology. Bi-junction phase shifters can be implemented in any planar silicon photonic technology that offers bipolar implantations within silicon photonic waveguides.

We present the design and simulation of an integrated graphene-on-silicon nitride (GOSiN) dual-mode electroabsorption modulator at λ = 1550 nm. We started from designing GOSiN TE₀ and TE₁ mode filters with minimum length and insertion losses. Then, we designed a dual-mode modulator that combines both mode filters, tuning the absorption by applying a gate voltage. We show that the (0,0), (0,1), (1,0) and (1,1) logical values can be generated, with modulation depths up to 608-670 dB/cm for the TE0 mode and 228-284 dB/cm for the TE₁ mode. Our results improve, to the best of our knowledge, the state of the art in integrated graphene dual-mode modulators, helping to develop efficient mode division multiplexing (MDM) on-chip interconnections.

Lithium niobate photonics provides a low-loss platform with great properties for high-speed modulation, wavelength conversion and quantum optics. Micro-transfer printing allows scalable integration with CMOS compatible silicon photonics technologies.

The emergence of co-packaged optics and of parallel optics pluggable form factors with an increased number of fibers poses new challenges for optical packaging technologies. We describe micro-optical interposers manufactured with isothermal glass molding that enable the parallelized connection of large fiber arrays with photonic integrated circuits in a low-rise form factor as well as the management of scrambled light polarization from a remote laser source coupled by single mode fiber. A resonantly assisted but temperature tolerant silicon photonics Mach-Zehnder modulator operated in lumped element configuration is co-developed for that purpose and a concept for a 6.4 Tb/s bi-directional light engine described.

An ultra-compact 1.31μm-emission photonic crystal (PC) nano-ridge laser directly grown on a silicon substrate without thick buffer layers achieves lasing with a cavity length as small as 50 μm at a remarkably low pumping threshold of 4.42 kW/cm2. This laser exhibits a lasing peak with side-mode suppression ratio of over 17 dB and a linewidth as narrow as 1.47 nm under 22.91 kW/cm2 pulsed pumping.

All-optical wavelength conversion is a key functionality for large WDM networks with dynamic traffic, e.g., to enable flexible wavelength allocation within the subnetworks. All-optical techniques mostly rely on exploiting the nonlinearities in semiconductor optical amplifiers, such as gain-saturation, FWM, and difference-frequency generation. These methods, however, require additional probe laser, or phase matching. Here we describe an alternative approach based on a feedback-controlled integrated DBR-based multi-wavelength laser (MWL). Our MWL is designed to emit at multiple and controllable modes, thus removing the need for an additional external probe light. Injecting an optical signal around one of the modes of the MWL leads to the spectral multiplication of the signal to the other modes of the MWL. By varying the phase and amplitude of the feedback we show frequency conversion of a 1 GBd ASK signal at offsets ranging from tens of GHz to 1.2 THz. The emission of the MWL can be controlled at nanosecond time scales by changing the feedback phase from a monolithically-integrated feedback cavity. Our approach is, in principle, only limited by the gain bandwidth of the active medium which can reach up to 10 THz in InP.

We introduce a monolithic disk laser design where a whispering-gallery [WG] mode hosts the laser, while the pump is perpendicularly enhanced by exploiting a vertical Fabry–Pérot [FP] resonance. Pump-laser orthogonality here permits the WG laser roundtrip to be 1000 times longer than the FP-pump mode, which accordingly enhances coherency. In this manner, a 1 cm coherency laser diode can pump a 10-meter coherency erbium-based WG laser mode to enable an integrated ultrahigh Q laser. Orthogonal on-chip lasers, electrically pumped using a semiconductor diode, might impact high-Q photonics. Applications include communication, navigation, metrology, and sensing, where foundry-compatible mass-produced ultracoherent emitters are needed.

Lasers capable of emitting two or more distinct wavelengths with a control over their power balance appear as promising and versatile key devices in the context of THz signal generation, telecommunication or wavelength conversion using optical injection. For such applications, fast wavelength switching is required but devices implying tunability through thermal or mechanical actions can be slow or bulky. Photonic integration has the potential to overcome these obstacles. While merging the beam of distinct lasers on the same chip seems to be a straightforward solution, the absence of modal competition and phase noise correlation limits their use in some applications.

Here, we consider DBR-based multi-cavity lasers with multiple wavelengths sharing a common broadband gain medium. The wavelength switching principle relies on controlling the phase in a monolithically integrated optical feedback cavity. In this contribution, by monitoring the emitted time-resolved power signal, we characterize the dynamics of the wavelength switching mechanism and report a measured response time below 4 ns. Additionally, from numerical investigations using a multi-mode extension of the Lang-Kobayashi rate equations, we identify key parameters influencing the switching time.

The emergent thin-film lithium niobate on insulator (TFLN or LNOI) photonic integrated circuits (PICs) offer significant advantages in various applications due to their unique properties. This paper briefly explores recent advancements in TFLN PIC developments and their broad applications, emphasizing transformative capabilities in telecommunications and beyond. We highlight CSEM's pioneering initiative in establishing the first open-access foundry for this technology, addressing challenges associated with limited access to manufacturing facilities and process design kits (PDK).

A new low-cost BiCMOS Heterojunction Bipolar Phototransistor (HPT) is fabricated for the first time in an industrial BiCMOS technology from STMicroelectronics with a “no change in process” approach. The static responsivity as a function of the biasing is determined from measurements at 850 nm for various HPT designs. Devices with a static responsivity level up to 40 A/W at 2.5V collector-emitter voltage biasing can be achieved when the device operates near its breakdown voltage. In its active region, at V_CE = 1V, a static responsivity level up to 12.2 A/W was obtained selecting the appropriate base biasing.

Integrated photonic circuits represent a significant advancement in optical technology, providing notable benefits in speed, efficiency, and miniaturization. By consolidating various optical components onto a single chip, these circuits enable precise manipulation and control of light. As a result, they are playing a pivotal role in technologies such as high-speed data interconnects, sensors, artificial intelligence accelerators, and quantum computing. However, one of the key challenges that hinders the widespread adoption of integrated optics is fabrication imperfections. This work¹ highlights a scalable and non-volatile technique for post-fabrication tuning of photonic computational memories by automated silicon ion implantation to precisely align high-quality resonant devices to targeted wavelengths. Spectral shifts ranging from less than 10 pm to several nanometers are obtained showing long-term stability while adding negligible loss.

The realization of integrated active optical systems is crucial for the use of components such as amplifiers, lasers or photodetectors on a chip. Alumina (Al₂O₃) doped with erbium is a promising material for the realization of various active functions. Indeed, the low propagation losses of Al₂O₃, the rare earth compatibility and its wide transmission band makes Al₂O₃ suitable for a wide range of applications. Nevertheless, current methods for producing such waveguides are often costly and difficult, requiring complex and potentially loss-making processing steps like etching. In this context, Pulsed Laser Deposited (PLD) combined with lift-off is a relevant method for avoiding etchings. The process is composed of three main steps: photolithography, PLD and lift-off. In this work, we present how the different steps have to be optimized to make suitable waveguides for light propagation. Notably, photolithography needs a precise cross-section profile to obtain smooth sidewall, to ease lift-off and get high resolution patterning. For PLD, SEM images showed the importance of plume directivity and orientation in the PLD chamber to achieve a good control of the waveguide shapes. Finally, we also have shown that the Erbium photoluminescence is dependent on the annealing temperature. These results highlight the essential parameters which need to be precisely controlled to achieve accurate microstructures by liftoff processing performed in PLD layers, paving the way for the demonstration of low-loss waveguides and fully integrated erbium laser without etching.

We demonstrate a monostatic LiDAR based on an InP photonic integrated optical phased array (OPA). The system utilizes an OPA with on-chip amplification which transmits and receives light simultaneously through an array of eight end-fire waveguide antennas. The OPA is capable of a 4.6° angular resolution and a 41° field of view. The on-chip amplifiers provide up to 21.5dB gain in a 1465-1600nm wavelength range. We show proof-of-principle FMCW (frequency modulated continuous wave) sensing through the monostatic OPA. The system relies on the frequency modulation with up to 10GHz frequency excursion of an external optically isolated DFB laser, which allows the simultaneous detection of range and velocity. The measurements were performed with a reflective target located ~2m away from the OPA, by varying the target position and velocity of 30 cm and ±5cm/s respectively. To the best of our knowledge, we demonstrated the first monostatic FMCW LiDAR implementation on an integrated InP OPA.

Achieving high degree of tunability in photonic devices has been a focal point in the field of integrated photonics for several decades, enabling a wide range of applications from telecommunication and biochemical sensing to fundamental quantum photonic experiments. We introduce a novel technique to engineer the thermal response of photonic devices resulting in large and deterministic wavelength shifts across various photonic platforms, such as amorphous Silicon Carbide (a-SiC), Silicon Nitride (SiN) and Silicon-On-Insulator (SOI). In this paper, we demonstrate bi-directional thermal tuning of photonic devices fabricated on a single chip. Our method can be used to design high-sensitivity photonic temperature sensors, low-power Mach-Zehnder interferometers and more complex photonics circuits.

This study provides an in-depth evaluation of two fundamental techniques for fabricating sensing windows on silicon nitride platforms: a traditional etching strategy using reactive ion etching (RIE) combined with wet etching, and a lift-off-based process in which the top cladding material is deposited onto a suitable resist which is subsequently stripped of the distinct sensing waveguides. The analysis, based on a side-by-side comparison, meticulously examines the effectiveness of these methods. Key evaluation metrics include propagation and bending loss in the sensing windows, process robustness, and uniformity of critical dimensions and heights across the wafer. This will provide a comprehensive understanding of the strengths, weaknesses, and potential application limitations of each technique. An integral part of the study is the careful revision of the waveguide material stack to address specific challenges and applications. This precise tuning and adaptation of the material stack serve as a proxy for the demands likely to be encountered in real-world applications. The conservative etching technique has the advantage that it can be easily combined with subsequent facet etching processes for edge coupling approaches. Conversely, the lift-off resist based approach, despite its relative complexity and sensitivity to high-temperature deposition on the resist, reduces the negative impact of the process on surface roughness and sidewall angles. The knowledge gained from this research provides valuable guidance in the selection of appropriate fabrication techniques for specific silicon nitride sensor applications to increase the robustness of the processing steps for potential mass production stability.

The field of integrated photonics has expanded since the last century due to the need for even smaller devices, finding numerous applications in areas such as sensing, communications, and information technology. Particularly, the use of optical sensors has increased in recent years due to advantages over classical sensors, including versatility, minimal sample quantities, and label-free quantitative detection of chemical and biological samples. One technique employed for creating integrated photonic structures is ultrashort laser inscription, enabling the fabrication of optical waveguides in transparent materials without the need for masks or chemical processes. This work presents significant advancements in the design, fabrication, and characterization of Mach-Zehnder interferometers (MZIs) based on optical waveguides, utilizing the femtosecond direct laser writing (FDLW) technique. These interferometers have enabled the creation of integrated systems and their application in detecting physical variables such as temperature changes and variations in the refractive index of solutions with varying concentrations, including urea. We manufactured an embedded Mach-Zehnder interferometer in soda-lime glass, exhibiting sensitivity comparable to Silicon-on-Insulator (SOI) devices. A substantial enhancement in sensitivity (~54 pm/°C) was achieved, thanks to the unique three-dimensional (3D) capabilities provided by FDLW, surpassing the typically low thermo-optic coefficient of soda-lime glass. As a proof of concept, we also applied the first FDLW-fabricated MZI for concentration changes detection through evanescent field interaction in fused silica, demonstrating a sensitivity of ~1.22 nm/mM. Such miniaturized structures will significantly impact the development of compact and highly sensitive integrated photonic devices.

We present a new concept for generating wideband signals of higher spectral efficiency from signals of low-bandwidth and lower efficiency. With pure electronics, we are able to generate broad-bandwidth signals with low-modulation format from low-bandwidth sub-DACs. This is based on electrical orthogonal sampling with sinc-pulse sequences in N parallel branches. In photonics, a higher spectral efficiency can be achieved from M branches at different optical powers. The proposed method can be integrated into any silicon platform and might be of great interest for bandwidth, and data hungry applications.

Advanced flat optical and integrated photonic systems emerge from research and highly specialized solutions to more common use cases. This requires a concurrently developed infrastructure in the manufacturing. The necessary technologies are already available in semiconductor manufacturing. These technologies are mature and provide the accuracy and process stability required. However, the quality parameters which are typically used in semiconductor manufacturing do not map optical quality parameters in a straightforward manner. In consequence, the chosen process conditions often do not fit well to the specific optical application.

We will show on the example of electron-beam lithography, that while quality parameters describing optical applications and semiconductor applications differ, they are linked. Further we show, that tuning of the lithography process is a powerful lever to maximize overall optical quality. Since curved or any-angle structures are prevalent in many optical applications the data preparation and writing strategy are key-factors to improve optical performance and to decrease write time.

By means of artificial intelligence (AI), the amount of data processed has experienced an unprecedented surge, necessitating advanced computational techniques and technologies. Simultaneously, the hardware responsible for processing this data must demonstrate energy efficiency while maintaining a compact design. Phase change material based photonic neuromorphic computing addresses these challenges by enabling energy efficient and fast in-memory computation with a high degree of parallelization. The input data, encoded optically, is channeled into an on-chip matrix multiplication setup capable of executing parallel Multiply-Accumulate (MAC) operations using multiple wavelengths. Central to this computation are the PCM cells, which alter their refractive index according to their crystalline state[1]. These PCM cells function as non-volatile on-chip memory units. Typically, the state of these PCM units is established optically, with additional in-plane inputs required for each matrix cell. However, our approach seeks to set the state of the PCM cell using a out-of-plane vertical-cavity surface-emitting laser (VCSEL) array positioned atop the photonic-logic matrix. This approach significantly enhances the design flexibility for large scale matrix-vector multiplications and therefore opens up new possibilities for efficient computation.

This work reports on the fabrication and optimization of silicon microring resonators on SOI platforms with a focus on rapid device prototyping. Such resonators are instrumental in expanding the functionality of photonic circuits, be it by leveraging the inherent nonlinearities of silicon or by improving frequency filtering. Central to this investigation is the detailed fabrication techniques developed for SOI waveguides, specifically tailored to minimize losses. Devices are created using Electron Beam Lithography and are etched using Reactive Ion Etching. The performance of microrings with single-mode waveguides is compared with that of multimode variants, and it is shown that the latter mitigate the impact of sidewall roughness, thereby reducing scattering losses. Through optimization of the patterning parameters, etching recipes and thermal treatment, developed devices exhibit propagation losses as low as 0.28dB/cm and Q factors in the vicinity of 2×10⁵.

Integrated diamond photonics presents one of the most promising platforms for solid-state quantum computing due to diamond’s high refractive index, large electronic bandgap and its large number of defects with highly stable fluorescence of single photons and coherent electron spin control like the silicon vacancy. This work presents the design, simulation, fabrication and measurement of free-standing diamond photonic crystal cavities from bulk single-crystal diamond substrate for emission enhancement of the SiV^-. Simulation yields Q-factors of up to 2.4 million and normalised mode volumes down to 0.52. Confocal measurements of the fabricated devices yield Q-factors of up to 1800 near the SiV- zero phonon line. These diamond resonators offer a promising approach to realizing large-scale diamond quantum registers.

On-chip positional control of quantum dot single photon source (QDSPS) is an important problem for quantum photonic integrated circuits (PICs) due to the dependence of node and antinode locations inside an optical cavity and polarization states of QDSPS. To couple emitted photons at certain polarizations and wavelengths that are matched with allowed waveguide mode requires QDSPS deposition at specific locations with low fabrication tolerances. In this research, we simulated QDSPS for investigating the effects of position dependence on emitted single photons coupling efficiency and spontaneous emission rate by observing Purcell enhancement factor for both fundamental TE and TM modes.

We experimentally demonstrate a new electro-optic SRAM element fully CMOS compatible. Inspired by the Esaki diode, presenting negative differential resistance (NDR), we designed a new type of NDR diode based on a horizontal PN junction and a region with higher acceptor concentration, P⁺, in silicon. We embedded the new NDR into a photonic micro-ring resonator to enable a bistable device with electrical and optical readout capabilities. Our device is remarkable for its simplicity, CMOS compatibility and its low power consumption around the nanowatt, but it’s also an important steppingstone on the way to new nonlinear electro-optic and neuromorphic computing structures.

AR and VR technologies are advancing rapidly, offering immersive experiences in the digital world. Researchers are exploring new ways to improve visual quality and user immersion. One promising solution is combining Metalens Arrays and Color Filters, which can enhance AR/VR experiences by manipulating light at a tiny scale. These technologies, integrated into AR/VR glasses, promise to revolutionize various fields by improving image resolution, color accuracy, and brightness. Users can expect more lifelike virtual environments, allowing deeper exploration and engagement in simulations and applications. Overall, the integration of Metalens Arrays and Color Filters represents a significant advancement in immersive experiences, opening up new possibilities in entertainment, education, and professional fields. In our research, we create a reflective Color Filter (CF) using a metasurface for RGB color filtering intended for AR/VR displays. This CF reflects light of specific wavelengths for desired colors while absorbing the rest. We assess color purity and accuracy using linear plots in Ansys Lumerical FDTD and chromaticity diagrams. Additionally, to focus this filtered light at various spots on the same plane, we design a metalens array in Ansys Lumerical FDTD and analyze its focusing profile.