A Si optical phase shifter is essential for high-speed optical interconnect and efficient computing. However, a Si optical phase shifter based on the free-carrier effect is facing the fundamental difficulty to achieve its modulation bandwidth beyond 100 GHz because of the tradeoff relationship between the modulation efficiency and optical loss. The huge power consumption and slow switching speed of a thermo-optic phase shifter are also issues for programmable photonic integrated circuits for computing. To overcome this obstacle, we have investigated the hybrid integration of III-V semiconductors on Si photonics platform. In particular, we have proposed a III-V/Si hybrid metal-oxide-semiconductor (MOS) capacitor for phase modulation, where an n-type III-V membrane is bonded on a p-type Si waveguide with a gate oxide. A III-V/Si hybrid MOS capacitor enables electrons accumulation at the III-V MOS interface, contributing to efficient and low-loss optical phase modulation owing to the superior electron-induced refractive index change in III-V to in Si. We will review the recent progress of III-V/Si hybrid optical phase shifter based on MOS capacitor for optical communication as well as emerging computing applications.

This paper highlights the optimization of Deep Rib High Speed Phase Modulators for 400G applications thanks to the optimal choice of structure and implants through a Design Of Experiment analysis, including the proposal of a new Deep Rib HSPM with Z-implants. Results show 1,6dB gain in OMA (Optical Modulation of Amplitude) at the same cutoff frequency (fc=[2Pi*RC]^-1) compared to the previous generation [1] with optimized vertical implants, and up to 2,15dB gain vs [1] with the new Z-implant Deep Rib device.

Mach-Zehnder (MZ) modulators in Silicon-on-Insulator (SOI) are key components for integrated silicon photonic devices. Reducing their energy dissipation is a crucial step for applications of silicon photonics, especially in large data centers. In this work, we combine band-edge slow light structures consisting of silicon waveguide gratings with a periodic (interleaved) p-n junction. The slow-light structures consist of a waveguide grating with wide/narrow sections realized in a 300-nm thick silicon layer, on top of an unetched silicon layer of 50 to 150 nm thickness, fully embedded in SiO2. The grating gives rise to a photonic stop band and to a slow-light region close to the lowest band edge. The profile of the p-n junction varies periodically along the waveguide with interleaved n and p regions. This structure maximizes the spatial overlap between the optical mode and the depletion regions, yielding a further improvement of modulator efficiency beyond the slow-light effect. --- Under an applied reverse bias, the silicon refractive index is modified by the plasma dispersion effect, thus the waveguide grating acts as a phase shifter. The modulator efficiency VpiLpi is strongly improved in comparison with modulators without slow light or with a lateral p-n junction. Thanks to the optimized overlap between electric field and depletion regions, this improvement takes place over a spectral interval that is much larger than the slow-light bandwidth. Insertion losses due to free carriers are also lower than in conventional modulators. The advantage of combining slow-light grating waveguides with an interleaved p-n junction is especially pronounced at low driving voltage (of the order of 1V), where the dissipated energy can be as low as 0.4 pJ/bit over an optical bandwidth of several 10 nm. Thus, the present modulator structure is promising in view of realizing integrated MZ modulators with low power dissipation.

Optical frequency combs (OFCs) have played an important role over the past years for optical frequency metrology and synthesis, for astronomy, telecommunications, and spectroscopy. Among the different methods that have been studied for OFC generation, electro-optic frequency combs (EOFCs) using electro-optical modulators show a large flexibility in comb repetition rate, making it a solution of choice for absorption spectroscopy where a fine sampling in the frequency domain is usually required. Silicon photonics is a promising platform for EOFC generation, thanks to its high volume production and strong light confinement allowing to achieve small footprint photonic integrated circuits (PICs). Additionally, silicon PICs benefit from a direct compatibility with complementary metal-oxide semiconductor (CMOS) fabrication process. Carrier depletion-based modulators have already proved to be efficient and to achieve high bandwidth operation, which makes them suitable for EOFC generation. In this work we will show the first dual comb spectroscopy experiment using silicon optical modulators. As a proof of concept for spectroscopy applications, beating of two silicon EOFCs with slightly different repetition rates is observed in the RF domain using a multi-heterodyne detection technique. Each EOFC is generated from a silicon push-pull Mach-Zehnder modulator and shows typically 12 equally separated lines. The comb repetition rate is swept from 500 MHz to 12.5 GHz, thanks to the inherent flexibility of EOFCs, while their relative offset is kept steady (4 MHz). This technique is used to recover the transfer function of an optical band-pass filter without any tunable laser.

Silicon photonic modulators are a key component for electro-optic transmitters within data centers. Electro-refractive modulators relying on free carrier plasma dispersion in a Mach-Zehnder interferometer (MZI) have become the most popular solution. Among the different electrical configurations, PN silicon modulators show a high bandwidth but at the price of a low efficiency. Accumulation–based capacitive modulators are an alternative, allowing to reduce the modulator power consumption. Additionally, strained SiGe exhibits a stronger plasma dispersion effect than silicon for holes. In this work we study the behavior of capacitive modulators with a thin layer of strained SiGe. The modulator fabrication process is based on the standard process flow with only few add-ons. In the first demonstration we show that the thin SiGe layer improved the modulator efficiency by 25%. In addition, further improvement is possible by optimization of the SiGe deposition condition to maximize the SiGe layer stress.

Enabled by technological improvements, photonic devices and circuits are becoming increasingly more complex. Non-trivial geometries are designed to reduce device footprint, improve performance, and introduce novel functionalities. However, the number of design variables required to properly represent these geometries quickly grows, limiting the effectiveness of classical design approaches. Moreover, parameters are often strongly interdependent, restricting the use of sequential optimizations or independent parameter sweeps. Although several optimization techniques can be effective for multi-parameter design, they commonly allow to optimize for a single or a handful designs and the optimization process needs to be repeated if new performance criteria are introduced. In contrast to classical design approaches, the in uences of the design parameters remain hidden as well as the general behavior of the design space. In this paper we present an extension of our recent work on the application of machine learning pattern recognition to the design of multi-parameter photonic devices. In particular, we propose using a combination of local optimization based on the adjoint method and the use of dimensionality reduction. Adjoint optimization is used multiple times to generate a small set of different designs with high performance. Dimensionality reduction is applied to analyze the relationship between these degenerate designs and identify a lower-dimensional design sub-space that includes all alternative good designs. This sub-space can be mapped for any performance criteria thus enabling informed decisions based on the relative priorities of all relevant performance specifications. As a proof of concept, we demonstrate a ten-parameter design of an integrated photonic power splitter using silicon-on-insulator technology. We identify a region of possible high performance design solutions and select two design candidates either maximizing the splitter efficiency or minimizing back-reflection.

Computing enabled by electronics has been improved so extensively as to make machine learning algorithms such as deep neural network so powerful than ever before. Data communications enabled by optics have been one of the cornerstones of the modern society built upon Internet. The rapidly increasing demands for communication bandwidth due to numerous emerging applications such as AI-based cloud computing have significantly increased the bisection bandwidth of intra-datacenter networks. This trend will necessitate the optics-electronics co-packaging on board, which can only be realized by the substantial development of integrated photonics such as silicon photonics. This opportunity, making optics and electronics so close to each other, will in turn offer a chance to reconsider building EO-hybrid computational and intelligent systems. Although optical computing and optical neural networks have been proposed since a while ago, recent demonstrations of deep neural networks implemented on silicon photonics reactivate the study of exploiting photonics for machine learning. The current program-based deep neural networks are powerful, but consume huge computational resources. Suppose that just optical propagation in compact photonic chips could realize similar functions, it would greatly decrease the power and latency. Even though the scalability and the integration of nonlinear activation functions on photonic chips are still challenging, for some applications such as classifier and digit recognition, photonics systems could be viable and beneficial from such aspects. This talk will introduce our efforts to develop the topology concepts, algorithms, and applications in order to implement silicon photonics for calculation and machine learning applications. First, a high-bit reconfigurable DAC based on generic photonic circuits will be presented. Second, without using any nonlinear activation functions or building deep neural networks, we demonstrate a photonic classifier based on only linear optical components. At last, we will show several ways of implementing silicon photonic circuit to recognize digits.

In almost all modern communication and sensing systems analog-to-digital and digital-to-analog converters (ADCs, DACs) are used and their performance parameters like bandwidth and resolution are critical for the overall system performance. Nowadays the resolution of the best broadband ADCs and DACs is fundamentally limited by the clock jitter of the electronic sampling clock. Mode-locked lasers have demonstrated exceptionally low timing jitter for the envelope of the optical pulse train which can be orders of magnitudes lower than the jitter of electronic clocks. This has motivated research into electronic-photonic ADC and DAC concepts. The talk will describe conventional and novel electronic-photonic ADC and DAC concepts as well as implementations of core circuit blocks in silicon photonics technology.

Over the past few decades, EO (electro-optic) polymers have been developed into modulator devices that have incredible speed and low power. The talk will explore how advanced EO polymer materials can fill in the gaps of the photonics roadmap through extremely high performance. As the appetite for increased traffic continues unabated, there are needs to look for faster and lower power consuming optoelectronic devices. One such device that merits potential that can be integrated on a photonics platform as a PIC (photonic integrated circuit) is the EO polymer modulator that is attractive not only in raw speed, but also in low power consumption for systems through direct driving without the use of dedicated (and often expensive) driver chips. Polymer modulator data will be presented that aligns with industry-based photonics roadmaps.

Silicon photonic technologies have undoubtedly revolutionized the optical communication and photonic industries during the last 10 years. Silicon on insulator (SOI) devices are now routinely used to build high performance transceivers and optical components, with a privileged route to the integration with electronic devices. However, in order to increase the density of integration and demonstrate advanced functionality, other materials showing complementary characteristics to those exhibited by silicon are needed. In this context, we recently developeds a tuneable index back end of line (BEOL)-compatible platform, based on silicon nitride, in which the content of silicon can be selectively varied to tailor the optical characteristics of the material. This granted the ability to change the material transparency window and its linear and nonlinear optical response in various wavelength regions. This advanced platform allowed us to demonstrate various photonic devices, exhibiting extremely low loss and capable of performing functionality that are not allowed in silicon. Moreover, the developed platform is fully CMOS and BEOL compatible and only low temperature processes (<350 °C) are utilized during the fabrication steps. In this work we describe the CMOS fabrication protocols developed for this platform as well as various recently realized applications/devices, such as MUX/DEMUX in the O-band, advanced nonlinear devices and novel photonic components operating in the 2 um window.

The rising production and consumption of data worldwide is posing challenges on how it can be efficiently analyzed and absorbed on a human scale. Emerging technologies in data visualization are rapidly expanding to try to address this problem. One technology in particular, Augmented Reality (AR) glasses, has the capacity to allow individuals to process live virtual information while keeping an eye on their surrounding environment. Our team has recently presented a unique retinal projection concept for augmented reality applications¹. The concept combines a photonic integrated circuit (PIC) and holography. The photonic integrated circuit is made of silicon nitride (Si₃N₄) for its ability to guide visible wavelengths² (λ = 532 nm in our case) and its compatibility with the CMOS fabrication process technology. In our concept, the role of this circuit is to distribute and extract light at specific locations on the surface of a glass. The light emissions pass through a holographic layer deposited on the surface of the photonic circuit. The holographic layer is made of a 2D array of small individual holograms. The role of the hologram is to modify the light properties (emission angle, phase…) of the light emissions from the PIC in order to generate a composite plane wavefront. The eye can focus this wavefront on the retina even at a small eye relief distance (few centimeters). The focused wavefront represents one pixel of the projected retinal image. Several emissive point distributions on the surface of the PIC will create a full 2D retinal image. The retinal projector concept is schematically described in fig.1. Until now, both parts of the AR glass – the PIC and the holographic layer – have been developed independently^3,4,5. In previous work⁶, the basic building blocks of the circuit were designed both analytically and with numerical simulations: single-mode waveguide, MMI (MultiMode Interference) coupler, diffraction gratings and more. This paper presents in section 2 the design and numerical simulations of a double-tip edge coupler at λ = 532 nm. This additional component will be useful to couple light in our future prototypes with a compact design. In section 3, it is combined with previously designed PIC building blocks to design a unique device dedicated to evaluate the interaction between a silicon nitride PIC and a hologram.

High-speed electronic integrated circuits are essential to the development of new fiber-optic communication systems. As a consequence of the increasing speeds and multi-channel operation, close integration and co-design of photonic and electronic devices have become a necessity to realize high-performance sub-systems. Such co-design on the other hand also enables the design of new electro-optic architectures to create and process multi-level optical signals. This presentation will illustrate a number of recent and ongoing developments in IDLab, an imec research group, from various H2020 projects with a focus on application-specific high-speed electronic transceiver circuits such as driver amplifiers and transimpedance amplifiers (TIAs).

Optical Random Access Memories (RAMs) have been conceived as high-bandwidth alternatives of their electronic counterparts, raising expectations for ultra-fast operation that can resolve the ns-long electronic RAM access bottleneck. In addition, with electronic Address Look-Up tables operating still at speeds of only up to 1 GHz, the constant increase in optical switch i/o data rates will yield severe latency and energy overhead during forwarding operations. In this invited paper, we present an overview of our recent research, introducing an all-optical RAM cell that performs both Write and Read functionalities at 10Gb/s, reporting on a 100% speed increase compared to state-of-the-art optical/electrical RAM demonstrations. Moreover, we present an all-optical Ternary-CAM cell that operates again at 10 Gb/s, doubling the speed of the fastest optical/electrical CAMs so far. To achieve this, we utilized a monolithically integrated InP optical Flip-Flop and a Semiconductor Optical Amplifier-Mach-Zehnder Interferometer (SOA-MZI) operating as an Access Gate to the RAM, and as an XOR gate to the T-CAM. These two demonstrations pave the way towards the vision of integrated photonic look-up memory architectures in order to relieve the memory bottlenecks.

Coherent optical frequency combs have revolutionized precision measurement with light. However, the complex setups needed to generate and stabilize these systems limit their relevance in field-testing scenarios. Combs generated in a microcavity (“micro-combs”) are evolving rapidly, and provide a pathway to miniaturize frequency comb systems. While early micro-combs were subject to instabilities and lacked the critical ability to form femtosecond pulses, an important advancement has been the realization of temporal soliton in optical micro-cavities. First observed in optical fibers, this form of ultrashort pulse formation has been demonstrated across many micro-cavity material platforms. Photonic chip-based soliton microcombs have shown rapid progress and have already been used in many system-level applications, including coherent communications, astrophysical spectrometer calibration, optical-frequency synthesis and ultrafast ranging. One challenge to realize a fully integrated soliton microcomb is to combine a high-Q integrated microresonator (made of e.g. silica or silicon nitride) and a high-power laser, in order to avoid the use of an optical amplifier or an external modulator to meet the power requirements for soliton generation. While there has been substantial progress in realizing soliton microcombs that rely on compact laser diodes, culminating in devices that only utilize a semiconductor amplifier or a self-injection-locked laser as a pump source, accessing and generating single-soliton states with electronically detectable line rates from a compact laser module has remained challenging. In this talk, the recent demonstration of a compact photonic package will be presented, where a high-Q microresonator is driven by an ultralow-noise diode laser and generates a single-soliton state with a repetition rate below 100 GHz.

Second order optical nonlinearities induced in silicon waveguides by the application of an electric field are evidenced by measuring second harmonic generation (SHG) in the mid infrared. The electric field is produced by lateral p-i-n junctions which are periodically disposed across the waveguide to reach a quasi-phase matching condition. Here, we report on the modeling of the experimental results by using stochastic variations of waveguide and junction geometries which are compatible with the fabrication technique. These variations lead to a broad band multiple peaked spectrum of the SHG efficiency around the nominal phase matched wavelength. Agreement between experiments and simulations is found.

Photonic chip based Kerr frequency combs are transforming diverse applications including spectroscopy, telecommunication, signal processing and metrology among others. Integrated silicon nitride (SiN) waveguides with anomalous dispersion have the potential to bring practical nonlinear optics to mainstream photonic integrated circuits; however, high stress and high processing temperatures for SiN deposited by low pressure chemical vapour deposition (LPCVD) remain an obstacle to mass adoption. We successfully demonstrate fully CMOS-compatible high confinement SiN microring resonators based on reactive sputtering thin-films at a maximum processing temperature of 400°C. We deposit 0.85 μm thick SiN thin-films with a low stress value of 41.5 MPa and bulk material losses of 0.3 dB/cm. Linear waveguides losses of 0.7 dB/cm (Q_int= 4.9 × 10⁵) and 0.5 dB/cm (Q_int= 6.6 × 10⁵) have been achieved at 1560 nm and 1580 nm, respectively. We characterised the nonlinear properties of the waveguides and measured a nonlinear coefficient of Υ = 2.1 W^-1 m^-1 and a nonlinear refractive index n₂ of 5.6 × 10^-19 m² W^-1. Modulation-instability (MI) optical frequency combs are observed by pumping a 120 μm radius microring resonator at 1560 nm with an estimated on-chip pump power of 850 mW, showing a native FSR spaced frequency comb covering a >250 nm wide spectral range.

Four years ago, the EU project PIX4life set out to mature an open access pilot line for silicon nitride integrated photonics, focused on life science applications. The synergies of industrial and academic project partners enabled the creation and validation of a unique pilot line using carefully selected demonstrator projects. Simultaneously, the software infrastructure (process design kits, design tools and building blocks) needed to enable early access of the pilot line through multi-project wafer (MPW) fabrication runs was created. After ten MPW fabrication runs in last three years at two foundries, and successful realization of dozens of designs from the project partners and the external customers, the pilot line is fully operational and ready for open access. In this presentation, we intend to share the experiences we have gained in setting up the pilot line, and to discuss the opportunities and challenges we can expect in future.

The consolidation of the silicon photonics ecosystem has gained a lot of momentum in recent years. This is attributed to various initiatives in Europe and elsewhere as well. Despite these efforts, there are still outstanding issues that need to be resolved to further empower fabless end-users through a supply chain that is predictable, low risk, on-time, and provides low-cost services. This paper discusses topics such as design IP, diversity in platforms, access for prototyping (in particular MPW), skilled personnel, and routes for manufacturing in silicon photonics. These issues are of key importance for the development of a high-performance silicon photonics ecosystem. Furthermore, the paper describes how ePIXfab – the European Silicon Photonics Alliance – is playing its role to develop a silicon photonics ecosystem in Europe that can meet end-users expectations consistently.

In high performance computer systems and large-scale data centers, data movement becomes a critical problem. To overcome this problem, co-packaging technology, which realizes high density integration of optical modules and switch ASICs (application specific integrated circuits) in one small package, is attracted much attention. Due to the multiple photonics dies and the ASICs are integrated in a small volume, low-latency, high-bandwidth-density and low-power data transfer can be realized. To provide the co-packaging technology compatible with standard large-scale integration (LSI) packaging process, we have proposed a new package substrate called active optical package (AOP) substrate. The AOP substrate is an organic package substrate where Si photonics dies are embedded. On the surface of the substrate, fan-out polymer waveguides, connecting high density Si-photonics I/O and low density single-mode fiber array, are integrated. The polymer waveguides and the SMF is connected using an optical connector at the edge of the AOP substrate. Thus, the AOP substrate will be used as same as conventional standard package substrates. Only a different point is to connect a pluggable optical fiber. In this paper, we introduce some key technologies for the AOP substrate.

As the market adoption of silicon photonics technologies continues to rise, and ever more fabless companies enter the market, there is a clear need for a flexible device prototyping foundry service that retains the ability for device level innovation, whilst also offering a clear route to market. The CORNERSTONE platform offers an affordable multi-project-wafer (MPW) service that allows a degree of customisation, which may not be accessible at other foundries. Through the use of DUV projection lithography, fabrication processes can be easily transferred to other foundries for mass production. Additionally, the ability to exploit high resolution e-beam lithography for certain layers mimics more advanced technology nodes, should this be deemed necessary. Several silicon-on-insulator platforms enable a plethora of applications including datacoms, LIDAR and mid-IR sensing. This talk gives an overview of the present status of the CORNERSTONE platforms, and an outlook for the future.

VTT's silicon photonics platform: present status and future outlook

The direct epitaxial growth of III-V materials on Si substrates is one of the most promising methods to provide Si-based lasers for monolithic integration. Direct epitaxy of III-V on Si encounters three major challenges, which are large lattice mismatch, polarity difference, and thermal expansion mismatch, leading to threading dislocations, antiphase boundaries, and microthermal cracks, respectively. In this talk, I will show that, by homoepitaxial growth on a U-shaped patterned Si (001) substrate, highly uniform (111)-faceted Si sawtooth with underlying hollow structures can be formed. The (111)-faceted sawtooth structures can effectively annihilate the antiphase boundaries and terminate mostly the lattice mismatch induced dislocations at the III-V/Si interface, while the hollow structures can effectively reduce the thermal stress. The GaAs layers on (111)-faceted hollow Si (001) show a threading dislocation density of ~106/cm2 via the electron channeling contrast image method. The O-band emission of InAs/GaAs QDs on Si (001) shows similar intensity to that on the GaAs substrate. With such high quality III-V materials on Si, we demonstrate the high performance microcavity, microdisk and electrically pumped lasers both on Si and SOI substrates.

We present our recent results on the use of two quite different approaches for photonic integration. First we shall describe how we used the concept of bound states in the continuum (BiC) to make channel guided devices without the need for any dry etching. The BiC channel waveguide employs a substrate that is completely flat. The completely flat structure is attractive for hybrid integration of 2D materials because it does not introduce sharp corners which can reduce the electrical mobility of the 2D material. Channel guiding of light can nonetheless be achieved by spin coating a lower-refractive-index polymer/photoresist on the 2D material and developing it to form a channel. This approach for integrating 2D materials also increases the optical overlap with the 2D material. We used this approach for the hybrid integration of graphene on lithium niobate for making 40-GHz-bandwidth channel-guided photodetectors and electroabsorption modulators on lithium niobate. The BiC concept facilitates the hybrid integration of 2D materials on different substrates and may also be used to increase the effective optical nonlinearity of the underlying substrate by hybrid integration of the appropriate 2D material. Second we shall discuss the InP membrane waveguide platform for nonlinear applications. InP has a third order nonlinearity that is over an order of magnitude larger than silicon, and is therefore of potential interest for spontaneous four wave mixing to produce entangled photons. The use of InP membranes can potentially facilitate the integration of active III-V lasers, and Geiger mode avalanche photodiodes for single photon detection and nonlinear devices on large scale silicon wafers which can integrate the large delay interferometers and filters needed for quantum information processing. We discuss the advantages and disadvantages of InP for SFWM and present recent results on the use of InP membranes for generating heralded single photons.

We report on the CMOS-compatible hybrid III-V/Silicon platform developed in CEA-LETI. In order to follow the large-scale integration capabilities of silicon photonics, already available worldwide in 200mm or 300mm through different foundries, the development of CMOS-compatible process for the III-V integration is of major interest. The technological developments involve not only the hybridization on top of a mature silicon photonic front-end wafer through direct molecular bonding but the patterning of the III-V epitaxy layer, low access resistance contacts, as well as planar multilevel BEOL must also be investigated and optimized. Test vehicles for the process validation based on either distributed feedback (DFB) or distributed Bragg reflector (DBR) laser cavities were designed. A modular approach is proposed in order to minimize the impact on the already qualified silicon photonics devices. Next, a collective III–V die bonding and processing have been successfully developed in this platform. The collective bonding, based on a flexible template holder, allows for large scale die to wafer transfer in both 200 and 300mm. After the III-V substrate removal and III-V patterning relying on optimized dry etching processes, CMOS compatible metallization’s are used to realize ohmic contact on n-InP and P-InGaAs leading to contact resistivity in the range of 10−6 Ω·cm². While first demonstrations have been obtained through wafer bonding, the fabrication process was subsequently validated on III-V dies bonding with a fabrication yield of Fabry-Perot lasers of 97% in 200mm. A planarized two-metal-level BEOL was used to connect the devices, leading to a drastic reduction of series resistance between 5.5 and 7 Ω. Finally, the functionality of DFB and DBR lasers is demonstrated with SMSR up to 50 dB and maximum output power of 3 mW in CW. The overall technological features are expected improve the efficiency, density, and cost of silicon photonics PICs.

Combining acoustic waves and optomechanical resonators could lead to a new generation of ultra-stable optomechanical oscillators with high-spectral purity and low phase noise signals directly at the frequency of interest. Moreover, this approach would allow an easy on-chip integration at μm scale, where the lack of good quality and miniaturized source is a severe issue. The resonator part of such a system will be made of a 1D photonic crystal nanobeam suspended over a silicon waveguide, sustaining optical modes around 1.55 μm and mechanical modes around a few GHz. On the other hand, Interdigital transducers (IDT) will provide acoustic waves in the GHz allowing a full mechanical control of the oscillating structure. In addition, phononic waveguides can be designed to engineer the acoustic velocity of the mechanical propagating modes in order to achieve the desired delay for a correct stabilization of the oscillations. The proposed system can be then exploited for potential on-board applications such as navigation and telecommunication systems, metrology or sensing. Design, simulation, fabrication and experimental results of resonant mechanical excitation in the GHz will be discussed, with particular emphasis on the time-domain studies of the propagating acoustic waves, where a specific time-gating technique is needed in order to properly extrapolate the delay information.

We present our results on the generation and manipulation of high dimensional quantum frequency states of light generated with AlGaAs chips working in the telecom band at room temperature. In the case of devices based on a collinear phase matching scheme we propose a method to generate and control the symmetry of broadband biphoton frequency combs, exploiting the interplay of cavity effects and relative temporal delay between the two photons of each pair. In the case of devices based on a transverse pump configuration we demonstrate that engineering the spatial properties of the pump beam allows to produce frequency-anticorrelated, correlated and separable states, and to control the parity of the biphoton wavefunction to induce either bosonic or fermionic particle statistics. These results open promising perspectives for communication and computation protocols exploiting high-dimensional quantum states, as well as for the quantum simulation of fermionic problems with photons on an integrated platform.

With the advent of silicon and silicon nitride (Si₃N₄) whispering gallery mode (WGM) resonators, a high quality factor Q (> 10⁶) can be achieved. These high Q cavities provide an ideal platform to study light-matter interactions. However, such high Q can only be achieved when the cavity radius is large compared to the operating wavelength. In small circular cavities, the bending losses, originating from the change in the direction of propagation of light, dominate. This radiation of energy from the cavity cannot be eliminated by improving the fabrication techniques and material quality. This is the biggest hurdle in reducing size of the WGM cavities, which is crucial for their efficient use as ultra-sensitive biosensors, sources of nonlinear effects and quantum optical sources. In all these application, a large Q/V ratio is essential to improve the performance. Here, we propose a new design in which a silicon nitride cavity of small radius (R ≤ 2μm) is encircled by one or more external shells. We show that by optimally choosing the thicknesses and the radii of the external shells, there can be dramatic reduction of the radiation of light from the inner cavity. In fact, Q as high as > 10⁶ (only found in bare cavities of larger radii) can be obtained. The external shells, being ultra-thin cannot guide light and hence the mode is trapped in the inner cavity. Thus, the quality factor can be significantly enhanced keeping the small mode volume. Owing to the improved Q/V, applications in ultra-sensitive biosensing and the study of nonlinear effects is viable such unusually small cavities. Additionally, the proposed design can be fabricated with the existing state-of-the-art technology.

In this paper we discuss the technology of SiO₂:TiO₂ waveguide films produced via sol-gel method and dip-coating technique, mastered at Silesian University of Technology. The technology allows fabrication of 150–300 nm thick waveguide films in single step deposition process and up to 500 nm thick with multiple step deposition process. The typically fabricated waveguide films have refractive index of ~1.8, however this value can be arbitrary altered from 1.45 to 1.94. The developed technology allows fabrication of waveguides having propagation losses below 0.2 dB/cm. The up-to-now fabricated waveguide films seem to be very promising. However biggest stumbling block in their application lays in elaboration of efficient patterning techniques. Currently the technology is developed towards an efficient way of waveguide layers patterning and introducing rare earth ions into the waveguide layers. The patterning technology will be developed in a twofold way, including mainly deep and shallow strip waveguide structures. The deeply patterned waveguide structures will be fabricated either with the direct nanoimprint of waveguide layer or in a more conventional way by using electron beam lithography and dry plasma etching. The fabrication of shallowly patterned waveguide structures will be carried out using also direct nanoimprint or alternatively photolithography combined with wet chemical etching.

A novel fabrication method for apodized diffractive structures is proposed and examined experimentally and theoretically. It combines the principles of Laser Interference Lithography and a moiré effect in order to create high-quality subwavelength gratings with 2D adiabatic variation of depth. Synthesized structures can be used in optical security systems, plasmonic research and high-effective light couplers.

Silicon photonics owes its success for a large part to the re-use of technologies originally developed for electronic integrated circuits. Therefore, not surprisingly, the silicon photonics world has also been inspired by open-access methodologies that are well established in the CMOS-world. This presentation will sketch a historical perspective of open-access in silicon photonics and will discuss the strengths but also the challenges in today’s state of the art. The presentation will then spell out a roadmap for the future.

We present an intuitive method of port analysis based on an S-matrix formalism for the design of a 2D metasurface coupler integrated with strip waveguide in SOI for 3.8 μm wavelength. The metasurface design has been achieved using Lumerical by systematically optimizing the radius of circularly shaped unit cells and the period between them. The effective coupling efficiency achieved for the designed 2D metasurface integrated with waveguide is ~98% in the in-plane waveguide for the out-of-plane surface illumination. The achieved bandwidth of the structure is 1 μm. The S-parameters of the optimized design are calculated and the 3-port analysis of the device is carried out using Interconnect by Lumerical. The suggested method is general and can be applied to any multi-port devices with some simple changes made in the layout of the design. Gain is the performance metric in our case and it is defined as the ratio of power transmitted at the output port to that of power in the input port. The gain of all the three ports are calculated. We believe our design is a good alternative for conventionally employed grating couplers and inverse tapers.

The Si transparency (1.1 μm – 8 μm wavelength) contains the strongest absorption features of a wide range of chemical and biological substances. However, the use of SOI in the mid-IR is hampered by the large absorption of the buried oxide (BOX) for wavelengths above 4 μm. Silicon membranes have garnered great interest for their unique capability to overcome the BOX limitation while leveraging the advantages of Si photonics. On the other hand, silicon is uniquely poised for the implementation of wideband mid-IR sources based on nonlinear frequency generation. Promising supercontinuum and frequency comb generation have already been demonstrated in Si. Still, current implementations have a limited flexibility in the engineering of phase-matching conditions and dispersion, which complicates the shaping of the nonlinear spectrum. Patterning Si with features smaller than half of the wavelength (well within the capabilities of standard large-volume fabrication processes) has proven to be a simple and powerful tool to implement metamaterials with optimally engineered properties. Here, we present the design of nanostructured silicon membrane waveguides with ultra-wideband flat anomalous dispersion in a wavelength span exceeding 5 µm. Our three-dimensional finite difference time domain (FDTD) calculations predict flat anomalous dispersion near 50 ps/km⋅nm between 2.5 µm and 8 µm wavelength. These results illustrate the potential of subwavelength metamaterial engineering to control chromatic dispersion in Si membrane waveguides. This is a promising step towards the implementation of wideband nonlinear sources in the mid-IR for silicon photonics.

Fluorescence microscopy is an indispensable tool in biology and medicine, fuelling many breakthrough discoveries in a wide set of sub-domains. Yet, the resolution is intrinsically restricted by the diffraction limit. The last two decades have witnessed the emergence of several super-resolution fluorescence microscopy techniques that are breaking this limit (cfr. Nobel Prize in Chemistry 2014). Structured illumination microscopy (SIM) is one of such techniques that has gained much popularity and is available in commercial systems. In SIM, a biological sample is illuminated by a spatially structured light field — a sinusoidal interference pattern — which causes normally inaccessible high-resolution information to be encoded into the observed image due to the Moiré effect. Typically, the illumination patterns are generated by a grating or SLM and focused onto the sample through an objective lens to excite the fluorophores. Next, the fluorescence is transmitted to the imager through the same lens. Multiple images are taken under certain illumination patterns and used to reconstruct a single super-resolved image. However, making use of free space optics, state-of-the-art SIM systems require multiple bulky and precision optical components that give these systems the drawbacks of high cost and cumbersome size. In this talk, a structured illumination microscopy system based on a photonic integrated circuit (PIC) is presented. The unique properties of photonic integrated circuits allow us to create truly innovative microscopy systems that have the potential to go well beyond the current state-of-the-art: higher resolution due to the use of high refractive index materials, easier alignment with on-chip illumination light path, large field of view, a compact form factor resulting from on-chip integration, and lower cost due to compatibility with CMOS chip fabrication.

We present the fabrication, characterization, and demonstration of high-efficiency ultrasound-powered micro- light emitting diodes (μLED) for use in optogenetic applications. InGaN based blue-emitting LED material wafers grown on a patterned sapphire substrate (PSS) were used to assist in the out-scattering of the light. The turn-on voltage of the LEDs is around 2.5 volts and the electrical ideality factor is 1.2 confirming high radiative recombination efficiency. A power density of more than 50 mW/mm² was obtained from a 130 x 300 μm² LED with a mesa of 100 μm diameter at 3 mA which is much more than is required to excite channelrhodopsin transfected neural cells. A high external quantum efficiency (EQE) of 33% is obtained at 3 mA measured in an integrating sphere. The peak wavelength of the μLED was measured at 483 nm at different current densities. The μLEDs are integrated directly onto a rectifier and Piezoelectric Transducer (PZT) harvester to realise a highly efficient ultrasound-powered light delivery unit capable to generate mWs of optical power. The concept was validated by powering the integrated device with ultrasound.

An integrated high-speed novel electro-optic power splitter based on interleaved PN junction in a silicon microring resonator has been proposed, which can be used as an ultra-fast, dynamically tunable optical power splitter to efficiently distribute power between various optical stations in Network-on-chip (NoC). The main working principle behind such a device is that we can change the refractive index of the PN junction waveguide under an application of an external reverse bias voltage, enabling the resonator to work as a power splitter. The interleaved PN junction waveguides utilizing the features of greater overlap of the optical waveguide mode with the depletion layer compared to normal PN junction, provide the opportunity to maximize the change in refractive index with minimum applied reverse bias voltage. The carriers (electrons and holes) are depleted near to the p region and n region interface on applying a reverse bias voltage. The depletion region increases with the applied reverse bias voltage, which enhances more interaction with the propagating light. To evaluate the performance of the splitter, we have calculated the change in the effective refractive index and the optical loss due to the free carrier concentrations. The proposed optical power splitter is capable of tuning the power splitting ratio from 0.15 to 7.82, which is a very wide range of tunability with low power consumption (35 mW) and low external voltage (0-3.5 V). We have proposed the use of a high-speed Digital to Analog converter (DAC) to apply an external reverse bias voltage to the interleaved PN junction waveguide.

In this work, a novel characterization technique taking advantage of the maturity and high sensitivity of Fourier domain optical coherence tomography (FD-OCT). The method has been experimentally verified on a silicon photonic chip fabricated using imec-ePIXfab passive technology with a waveguide height and width of 220 nm and 500 nm, respectively. The silicon is surrounded by air from top and by silicon-oxide from bottom and sides. The design assumes quasi Transverse Magnetic (TM) mode with an effective refractive index n_eff of 1.61 and group effective refractive index n_g of about 3.471 at 1550 nm. The method is applied on a simple 1.46 mm length waveguide ended by grating couplers from both sides. Using that length and by inspecting the spatial response obtained after inverse Fast Fourier Transformation (FFT), the extracted mean group effective refractive index n_g of the straight waveguide is 3.597. This value is attributed to the fabrication tolerance in the width of the structure leading to a width error of about 7.7 %.

In this paper, tri-state buffer and tri-state inverter circuits using the electro-optical effect have been proposed. Both the tri-state devices have three terminals as input, output and enable. An enable terminal is used to control the direction of the flow of the input towards the output. Mach-Zehnder interferometer (MZI) is used as the switching element for the designing of the tri-state buffer and tri-state inverter. The tri-state buffer circuit transfers the data, as it is, to the desired output port and the tri-state inverter sends the inverted data to the desired output port. Using the tri-state buffer and inverter, an optical integrated circuit is proposed. To analyse the operation of the device, the performance parameters like extension ratio, insertion loss, and contrast ratio are computed and found as 26.65, 19.41 and 0.07 dB respectively.

Data communication around the whole globe has seen an enormous increase in the exchange of data across the networks. This has urged the need of optical communication systems that can provide the desired bandwidth. This paper is a step to the path of building such optical processing elements. The paper presents optical designs of a 3-bit incrementer and a 3- bit decrementer logic function. The two optical circuits increment or decrement each 3-bit input binary number by unity. The optical designs have been implemented using the LiNbO₃ based Mach-Zehnder Interferometer (LNMZI) electrooptical switches. The incrementer circuit has employed just five LNMZIs and the decrementer circuit has only six LNMZIs. The optical circuits have been designed and tested using beam propagation method (BPM). A critical analysis of the two designs has been performed by calculation of the important performance metric parameters like extinction ratio (ER), contrast ratio (CR), insertion loss (IL). The incrementer/decrementer shows best values of ER, CR and IL as 31.43 dB, 35.71 dB and 0.95 dB respectively. The calculated parameters have been found within the acceptable limits which support the feasibility of the design.

Lithographic technology has been one of the main upholders to Moore's law in the semiconductor industry for the last decades. The underlying reason that enabled the evolution in semiconductor industry has been a steady silicon wafer printing cost, while being able to dramatically increase the number of nodes that can be printed per chip. Key developments in lithography such as wavelength decreasing, together with performance increase in lens and imaging technology, should be accounted for almost all the reduction of cost per function in integrated circuits technology. In this work, we will be presenting the simulation of two mitigation techniques for the impact of defects introduced by manufacturing processes. Namely, the lithographic mask limited resolution on the geometry of the representative device. These perturbations are a consequence of the lithographic mask limited resolution on the geometry of the representative device. For this purpose, the Beam Propagation and Finite Differences Time Domain methods will be used to simulate a multimode interference structure based on silicon nitride. The structure will be affected by previously mentioned perturbations and we expect results revealing a strong dependence between mask resolution, and imbalance and power loss. Two strategies will be followed concerning the mitigation of power loss and imbalance introduced by the limited resolution of lithographic mask: - Access waveguides tapering; - Adjustable power splitting ratios through the electro-optic effect. Through both strategies we aim to achieve an improvement on device’s performance but, in the latter are expected finer tuning capabilities, being enabled by dynamic compensation of power loss and imbalance when in a closed loop control architecture.

In this work, double ring resonator is analyzed with a microfluidic channel, the coupling of light is observed from the bus waveguide to ring waveguide and then to bus waveguide. The light couples inside the ring for some duration of time that is till the length required to circulate inside the ring. The design is simulated in the Lumerical mode solutions, the light source of 1550nm is allowed to pass through the input port of the top bus waveguide. And the same light will be coupled into the ring waveguide and then enters through the bottom bus waveguide at port 3. The designed double ring resonator is placed on a SiO2, the ring and the bus waveguide is of the material silicon. The light couples from the bus waveguide initially and then couples with the ring waveguide. In the double ring resonator structure, the effective refractive index 3.4254 is observed and its group index of 3.5226. The effective index and the group index is calculated at the mode 5, where the light coupling is observed at the two ring waveguide. Later by placing a micro fluidic channel on top of the double ring waveguide the effective refractive index was 3.4168 and its group index was 3.5328. From the length and the FSR, the effective index of 3.4168 and the group index of 3.5328 was calculated, which is desired for the bio sensing application. Hence it was observed an increase in the group index indicates that the proposed design will be useful in bio sensing application and it can be further fabricated for a point of care devices.

Application of photonics for smart farming could have a tremendous impact on the economy and the well-being of world population. Silicon Photonics should be a promising platform to apply the photonic technologies at reasonable cost in agricultural-based countries for their economic agriculture products. Among several group-IV materials, Silicon Nitride (SiN) is one of the most interesting Si-compatible materials suitable for on-chip spectroscopic sensing in the near- and mid-IR region as it has a relatively large transparent windows and a sufficiently-high refractive index of ~ 2. In this presentation, we will discuss the use of Si3N4 on SiO2 waveguides for a wideband spectroscopic sensing. The target design would include optical wavelength from 2.5 µm to 3.5 µm, in which several agriculturally-important gas molecules can be detected such as H2O (water vapor), CO2 (carbon dioxide), N2O (Nitrous oxide), NH3 (Ammonia), C2H4 (Ethylene), CH4 (Methane), and the spectral range is also within the transparent windows of Si3N4 on SiO2 waveguide. Significantly, we focus our effort on achieving a waveguide structure that can support single-mode optical propagation for a broad spectral range within 2.5-3.5 µm, and at the same time, can maintain a good sensing performance for multi-gas spectroscopic sensing despite the compromised evanescent field confinement values owing to the wideband design. We investigated the detection performance in term of compactness, polarization dependence, detection limit, and fabrication tolerance.

We have investigated the ability to monitor the dynamics transition phase of various substances by resonant probe light. Such a specific Micro-Total Analysis Systems (μTAS) can be used in food, cosmetic and biology applications. Such labon- chip sensors present the possibility of data treatment with an embedded system. The serial of transduced spectra are then acquired with an optical spectrum analyzer linked to a computer on which Matlab software record and process the data in real time. Then specific quantities can be linked to the intrinsic physico-chemical characteristics of the substances. As an example (not exhaustive) the development and the ability of an optical integrated polymeric resonator, acting as a surface light probe, for monitoring temperature-induced supramolecular phase transitions will be presented. The homogeneous detection of the transitions between different self-assembled structures in an aqueous solution of fatty acids (12-hydroxystearic acid, in association with amino-pentanol) was studied by investigating the coupling between the solution and the integrated photonic micro-cavity. Tuning the self-organized assemblies of surfactant is very attractive for many applications, such as cosmetic products, food, drug delivery and medical, and the development of alternative tools - especially those requiring minute amount of solution - to monitor their structural changes are essential. These original studies at temperatures ranging from 17 to 24 °C, based on a statistical treatment of optical resonance spectra, have evidenced the thermoresponsive nature of the optical features, and that different regimes occur with temperature. The optical results were corroborated with the measurement of the solution viscosity as a function of temperature, confirming that we can ascribe the optically-detected regimes to a surfactant assembly shifting reversibly from a tubular shape to a micellar one. The comparison between the optical and the rheological responses showed different accuracies: while the viscosity data exhibited a rather smooth and monotonous transition, the behavior changes were sharper and non-monotonous in terms of optical properties, allowing us to unambiguously identify in intermediate regime between 18.5 and 20°C. These morphological transition experiments represent a unique opportunity to extend the numbers of available techniques studying these systems through integrated optical techniques with potential opportunities of real time detection and working on low sampling volume. Other examples will be developed as the detecting of phase transition of sphingomyelin in biology and health corroborate by differential scanning calorimetry.

We present dual DFB lasers each integrated with one heater developed in a generic foundry platform. The thermal effects are experimentally investigated and exhibited a continuous wavelength difference tuning of 0-12.33 nm.