This PDF file contains the front matter associated with SPIE Proceedings Volume 9891, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and Conference Committee listing.

In the investigation we demonstrated technique of direct femtosecond laser writing of tracks with induced refractive index at record low depth under surface of lithium niobate (3-15 μm). It was shown that with the help of proposed technique one can be written claddings of near surface optical waveguides that plays a key role in fabrication of fast electro-optical modulators with low operating voltage. Fundamental problem resolved in the investigation consists in suppression of negative factors impeding femtosecond inscription of waveguides at low depths. To prevent optical breakdown of crystal surface we used high numerical aperture objectives for focusing of light. It was shown, that advanced heat accumulation regime of femtosecond inscription is inapplicable for writing of near-surface waveguides, and near the surface waveguides should be written in non-thermal regime in contrast to widespread femtosecond writing at depths of tens micrometers. Inscribed waveguides were examined for optical losses and polarization properties. It was experimentally shown, that femtosecond written near surface waveguides have such advantages over widely used proton exchanged and Ti-diffusion waveguides as lower optical losses (down to 0.3 dB/cm) and maintaining of all polarization states of propagation light, which is crucial for development of electro-optical modulators for broadband and ultrashort laser emission. Novelty of the results consists in technique of femtosecond inscription of waveguides at record low depths under the surface of crystals. As compared to previous investigations in the field (structures at depths near 50 um with buried electrodes), the obtained waveguides could be used with simple closely adjacent on-surface electrodes.

Large scale, planar optronic systems allowing spatially distributed functionalities can be well used in diverse sensor networks, such as for monitoring the environment by measuring various physical quantities in medicine or aeronautics. In these systems, mechanically flexible and optically transparent polymeric foils, e.g. polymethyl methacrylate (PMMA) and polyethylene terephthalate (PET), are employed as carrier materials. A benefit of using these materials is their low cost. The optical interconnections from light sources to light transmission structures in planar optronic systems occupy a pivotal position for the sensing functions. As light sources, we employ the optoelectronic components, such as edgeemitting laser diodes, in form of bare chips, since their extremely small structures facilitate a high integration compactness and ensure sufficient system flexibility. Flexographically printed polymer optical waveguides are deployed as light guiding structures for short-distance communication in planar optronic systems. Printing processes are utilized for this generation of waveguides to achieve a cost-efficient large scale and high-throughput production. In order to attain a high-functional optronic system for sensing applications, one of the most essential prerequisites is the high coupling efficiency between the light sources and the waveguides. Therefore, in this work, we focus on the multimode polymer waveguide with a parabolic cross-section and investigate its optical coupling with the bare laser diode. We establish the geometrical model of the alignment based on the previous works on the optodic bonding of bare laser diodes and the fabrication process of polymer waveguides with consideration of various parameters, such as the beam profile of the laser diode, the employed polymer properties of the waveguides as well as the carrier substrates etc. Accordingly, the optical coupling of the bare laser diodes and the polymer waveguides was simulated. Additionally, we demonstrate optical links by adopting the aforementioned processes used for defining the simulation. We verify the feasibility of the developed processes for planar optronic systems by using an active alignment and conduct discussions for further improvements of optical alignment.

In this paper we present a tandem UV/VIS/ NIR wavelength selector based on a multilayer a-SiC:H optical filter that requires appropriate ultraviolet steady states optical switches to select the desired wavelengths in the ultraviolet/visible/infrared spectral ranges. Five ultra-light communication channels are transmitted together, each one with a specific bit sequence. Results show that the background side and intensity works as a selector in the infrared, visible and ultra violet light regions, shifting the sensor sensitivity. This nonlinearity allows the identification and decoding of the different input channels. This concept is extended to implement a 1 by 5 wavelength division multiplexer with channel separation in the visible range and a transmission capability of 30 Kbps. The relationship between the optical inputs and the output signal is established and an algorithm to decode the MUX signal presented.

Structuring surface and bulk of crystalline silicon on different length scales can significantly alter its properties and possibly improve the performance of opto-electronic devices and sensors based on silicon. Different dominant feature scales are responsible for modification of some of electronic and optical properties of silicon. Several easily reproducible chemical methods for facile structuring of silicon on nano and micro-scales, based on both electroless and anodic etching of silicon in hydrofluoric acid based etchants, and chemical anisotropic etching of silicon in basic environments, are presented. We show how successive micro and nano structuring creates hierarchical silicon surfaces, which can be used to simultaneously exploit the advantages of both structuring feature length scales. Finally, we demonstrate a large increase in photocurrent obtained from a hybrid structured silicon/organic near-infrared photodetector. Improved silicon/6,6’-dibromoindigo hybrid photodiodes were prepared by nano- and micro-structuring the silicon part of the heterojunction by wet chemical etching methods. Photocurrent and spectral responsivity were improved in comparison to planar diodes by up to two orders of magnitude by optimization of the silicon structuring process. We show that the improvement in photocurrent is not due to the increase in surface area or light trapping.

Polarization management is a key functionality in many photonic applications, including optical communications, imaging or quantum information. Developing integrated devices capable of reliably controlling polarization state would result in compact and low cost circuits with improved stability compared with fiber or bulk optics solutions. However, stringent fabrication tolerances make the integration of polarization managing elements highly challenging. The main challenge in polarization controllers, composed by polarization rotators and polarization phase shifters, is to precisely control rotation angle in integrated polarization rotators. Proposed solutions typically require sophisticated fabrication processes or extremely tight fabrication tolerances, seriously hindering their practical application. Here we present a technology independent polarization controller scheme that relies on phase shifters to largely relax fabrication tolerances of polarization rotators. In addition, these phase shifters enable dynamic wavelength tuning. In our scheme, three polarization rotation elements are interconnected with two tunable phase shifters to adjust the polarization extinction ratio, while an output polarization phase shifter is used to select the relative phase. This way we can achieve any desired output state of polarization. We have implemented this scheme in the silicon-on-insulator platform, experimentally demonstrating a record polarization extinction range of 40 dB (± 20 dB) with a 98% coverage of the Poincaré sphere. Furthermore, the device is tunable in the complete C-band. These results constitute, to the best of our knowledge, the highest polarization extinction range achieved in a fully integrated device.

Implementation of optical components in microprocessors can increase their performance by orders of magnitude. However, the size of optical elements is fundamentally limited by diffraction, while miniaturization is one of the essential concepts in the development of high-speed and energy-efficient electronic chips. Surface plasmon polaritons (SPPs) are widely considered to be promising candidates for the next generation of chip-scale technology thanks to the ability to break down the fundamental diffraction limit and manipulate optical signals at the truly nometer scale. In the past years, a variety of deep-subwavelength plasmonic structures have been proposed and investigated, including dielectric-loaded SPP waveguides, V-groove waveguides, hybrid plasmonic waveguides and metal nanowires. At the same time, for practical application, such waveguide structures must be integrated on a silicon chip and be fabricated using CMOS fabrication process. However, to date, acceptable characteristics have been demonstrated only with noble metals (gold and silver), which are not compatible with industry-standard manufacturing technologies. On the other hand, alternative materials introduce enormous propagation losses due absorption in the metal. This prevents plasmonic components from implementation in on-chip nanophotonic circuits. In this work, we experimentally demonstrate for the first time that copper plasmonic waveguides fabricated in a CMOS compatible process can outperform gold waveguides showing the same level of mode confinement and lower propagation losses. At telecommunication wavelengths, the fabricated ultralow-loss deep-subwavelength hybrid plasmonic waveguides ensure a relatively long propagation length of more than 50 um along with strong mode confinement with the mode size down to lambda^2/70, which is confirmed by direct scanning near-field optical microscopy (SNOM) measurements. These results create the backbone for design and development of high-density nanophotonic circuits and their integration with electronic logic on a silicon chip.

Waveguide optics takes up a prominent role in the progressing miniaturization of optical devices. Chip integrated photonic waveguides especially allow for complex routing schemes of light across a chip. In/out-coupling diffraction gratings form an essential tool in waveguide systems, as they facilitate the interaction between the waveguide system and the near or far-field.[1,2] Ideally, these gratings would couple out all light in the waveguide into a beam with a predefined polarization and, phase and intensity profile. As such they should be able to produce any functional beam that is typically prepared by free space optics. Yet, in practice there is typically a design trade-off between beam quality and out-coupling efficiency.[2] Light in the waveguide has to travel laterally through the grating to be coupled out. The light therefore decays exponentially over the grating, causing much more light to be coupled out at the start of the grating than at the end. This asymmetry results in a warped out-coupling intensity that heavily influences the light beam’s intensity profile. Especially when the grating is addressing points in the near field, as is the case for focusing waveguide grating couplers, this effect can be highly disruptive.

In this work we present a grating constructed from a field of sub-wavelength scatterers, rather than full grating lines. By tuning the position and the density of the scatterers, the phase and the intensity of the out-coupled light can be set precisely over large grating areas. An iterative design algorithm is developed that carefully tunes the density so as to control the light intensity in the waveguide and the amount of out-coupled light. Using FDTD simulations we show that these gratings can efficiently couple out light into a nearly diffraction limited spot with an even angular intensity. We verify this experimentally by fabricating these gratings in the SiN/SiO2 system using e-beam lithography. In addition, we also show that these gratings can couple out more complex holographic patterns.

These density controlled out-coupling gratings let us efficiently address the near-field on optical chips, making them ideal waveguide components for on-chip optical trapping, holographic imaging or fluorescent excitation.[3]

The ever growing demands of bandwidth in optical communication systems are making traditional Wavelength Division Multiplexing (WDM) based systems to reach its limit. In order to cope with future bandwidth demand is necessary to use new levels of orthogonality, such as the waveguide mode or the polarization state. Mode Division Multiplexing (MDM) has recently attracted attention as a possible solution to increase aggregate bandwidth. In this work we discuss the proposition a of mode converter that can cover the whole C-Band of optical communications. The Mode Converter is based on two Multimode Interference (MMI) couplers and a phase shifter. Insertion loss (IL) below 0.2 dB and Extinction ratio (ER) higher than 20 dB in a broad bandwidth range of 1.5 μm to 1.6 μm have been estimated. The total length of the device is less than 30 μm.

Array waveguide gratings (AWGs) are a key component in WDM systems, allowing for de-multiplexing and routing of wavelength channels. A high-resolution AWG able to satisfy challenging requirements in terms of insertion loss and X-talk is what is needed to contribute to the paradigm change in the deployment of optical communication that is nowadays occurring within the ROADM architectures. In order to improve the performances and keep down the footprint, we modified the design at the star coupler (SC) and at the bending stages. We evaluated how the background noise is modified within a whiskered-shaped SC optimized to reduce the re ectivity of the SOI slab and keep down back-scattered optical signal. A dedicated heating circuit has also been designed, in order to allow for an overall tuning of the channel-output. A high-performance AWG has also to cope with possible thermal-induced environmental changes, especially in the case of integration within a Photonic Integrated Circuit (PIC). Therefore, we suggested a way to reduce the thermal-sensitivity.

In this paper, we report on time resolved electro-optic measurements in strained silicon resonators. Strain is induced by applying a mechanical deformation to the device. It is demonstrated that the linear electro-optic effect vanishes when the applied voltage modulation varies much faster than the free carrier lifetime, and that this occurs independently on the level of the applied stress. This demonstrates that, at frequencies which lie below the free carrier recombination rate, the electro-optic modulation is caused by plasma carrier dispersion. After normalizing out free carrier effects, it is found an upper limit of (8 ± 3) pm/V to the value of the strain induced χ⁽²⁾_{eff, zzz} tensor component. This is an order of magnitude lower than the previously reported values for static electro-optic measurements.

The electro-optic Pockels effect may be generated in silicon photonics structures by breaking the crystal symmetry by means of a highly stressing cladding layer (typically silicon nitride, SiN) deposited on top of the silicon waveguide. In this work, the influence of the waveguide parameters on the strain distribution and its overlap with the optical mode to enhance the Pockels effect has been analyzed. The optimum waveguide structure have been designed based on the definition and quantification of a figure of merit. The fabrication of highly stressing SiN layers by PECVD has also been optimized to characterize the designed structures. The residual stress has been controlled during the growth process by analyzing the influence of the main deposition parameters. Therefore, two identical samples with low and high stress conditions were fabricated and electro-optically characterized to test the induced Pockels effect and the influence of carrier effects. Electro-optical modulation was only measured in the sample with the high stressing SiN layer that could be attributed to the Pockels effect. Nevertheless, the influence of carriers were also observed thus making necessary additional experiments to decouple both effects.

Ge quantum wells are emerging as a relevant material system for enabling fast and power-efficient optical modulation in the framework of Si-photonics. The need for reliable designs of QW structures, matching given operating wavelengths and bias voltages, calls for the implementation of modelling tools capturing the optical properties of SiGe heterostructures. Here we report on the calculation of the quantum confined Stark effect based on an eight-band k×p model. The obtained spectra are analysed and compared with experimental data showing a good agreement between calculation and measurements.

Despite all the eminent advantages of silicon photonics, other materials need to be integrated to fulfill the functions that are difficult to realize with silicon alone. This is because silicon has a low light emission efficiency and a low electro-optic coefficient, limiting the use of silicon as a material for light sources and modulators. A strong two-photon absorption (TPA) at high intensities also limits the use of silicon in applications exploiting nonlinear effects. In addition, signal amplification is needed to compensate the insertion and propagation losses in silicon nanowaveguides. To address these issues we have demonstrated the integration of atomic layer deposited nanolaminates on silicon waveguides. Firstly we demonstrate slot waveguide ring resonators patterned on a silicon-on-insulator (SOI) wafer coated with an atomic layer deposited organic/inorganic nanolaminate structure, which consists of alternating layers of tantalum pentoxide (Ta2O5) and polyimide (PI) [1]. These materials were selected since the ALD process for depositing Ta2O5/PI nanolaminate films is already available [2] and both materials exhibit high third order nonlinearities [3-4]. In our nanolaminate ring resonators, the optical power is not only confined in the narrow central air slot but also in several parallel sub-10 nm wide vertical polyimide slots. This indicates that the mode profiles in the silicon slot waveguide can be accurately tuned by the atomic layer deposition (ALD) method. Our results show that ALD of organic and inorganic materials can be combined with conventional silicon waveguide fabrication techniques to create slot waveguide ring resonators with varying mode profiles. Secondly we demonstrate the integration of atomic layer deposited erbium-doped aluminum oxide (Al2O3) nanolaminates on silicon waveguides. This method provides an efficient way for controlling the concentration and distribution of erbium ions. We have applied this method on silicon strip and slot waveguides and show signal enhancement. Our results show that atomic layer deposited nanolaminates can potentially open new possibilities for various photonic applications, such as silicon photonic devices for light emission and amplification, optical sensing and all-optical signal processing. References [1] A. Autere, L. Karvonen, A. Säynätjoki, M. Roussey, E. Färm, M. Kemell, X. Tu, T.Y. Liow, G.Q. Lo, M. Ritala, M. Leskelä, S. Honkanen, H. Lipsanen, and Z. Sun, "Slot waveguide ring resonators coated by an atomic layer deposited organic/inorganic nanolaminate," Opt. Express 23, 26940-26951 (2015) [2] L. D. Salmi, E. Puukilainen, M. Vehkamäki, M. Heikkilä, and M. Ritala, “Atomic layer deposition of Ta2O5/polyimide nanolaminates,” Chem. Vap. Deposition 15, 221–226 (2009). [3] S. Morino, T. Yamashita, K. Horie, T. Wada, and H. Sasabe, “Third-order nonlinear optical properties of aromatic polyisoimides,” React. Funct. Polym. 44, 183–188 (2000). [4] C.-Y. Tai, J. Wilkinson, N. Perney, M. Netti, F. Cattaneo, C. Finlayson, and J. Baumberg, “Determination of nonlinear refractive index in a Ta2O5 rib waveguide using self-phase modulation,” Opt. Express 12, 5110–5116 (2004).

In a conventional slot waveguide structure, light is strongly confined in the slot region only either for the Quasi-TE (in a vertically oriented slot) or for the Quasi-TM (in a horizontally oriented slot) fundamental eigenmode, which enhances the optical forces, thus decreasing the optical power necessary to control the displacement of the device, but only for one eigenmode polarization at a time. In this work, we analyzed the optical forces in a cross-slot waveguide, which is formed by four suspended silicon waveguides separated by two orthogonal air slots. Cross-slot waveguides can strongly confine light in both Quasi-TE and Quasi-TM polarizations, thus enhancing the optical force and reducing the optical power for both. Our simulation results show that, by adjusting the optical power and the light polarization, it is possible to control the displacement of the waveguides in the vertical or in the horizontal direction almost individually, or in both directions simultaneously. The proposed NOMS device offers potential applications in active photonic devices, novel nanosensing and nanoactuators mechanisms, such as optical nanotweezers.

We propose germanium-rich silicon-germanium (SiGe) as a new platform for optical interconnects. The platform viability is experimentally and theoretically investigated through the realization of main building blocks of passive circuitry. Germanium-rich Si_1-xGe_x guiding layer on a graded SiGe layer is used to experimentally show 12μm radius bends by light confinement tuning at a wavelength of 1550nm. As a next step, Mach Zehnder interferometer with 10 dB extinction ratio is demonstrated. High Ge content of the proposed platform allows the coupling with Ge-based active devices, relying on a high quality epitaxial growth. Hence, the integration on Silicon of high speed and low power consumption Ge-rich active components is possible, despite the high lattice mismatch between silicon and germanium.

The development of a semiconductor laser compatible with silicon substrates and high-volume silicon integrated circuit manufacturing is a key requirement for monolithic silicon photonic transceivers. Tensile strained germanium is a promising material system which meets these criteria, and both optically pumped and electrically injected lasing have been reported[1,2]. It is well established that growth of thick (~1 micron) layers of germanium on silicon substrates by two-stage chemical vapour deposition followed by thermal annealing results in nearly-relaxed germanium with a residual biaxial tensile strain of typically 0.15-0.25% [3]. Several researchers have investigated methods of amplifying this built-in strain in order to increase the attainable optical gain. Increased uniaxial strain levels have been demonstrated in suspended linear bridge structures created by wet chemical underetching. However, uniaxial strain is less effective than biaxial strain in converting germanium from an indirect to a direct gap semiconductor and hence generating substantial optical gain. In this work, we have computationally investigated and optimised two-dimensional patterning and under-etching of germanium membranes in order to achieve biaxial strain amplification. Strain simulations were carried out using finite element methods and the shape of the suspended germanium structures was optimised to achieve the highest tensile strain whilst remaining below the empirically determined yield strength of the thin membranes. The net optical gain distribution across the membrane was calculated using 8 band k.p bandstructure to determine the full interband gain, the inter-valence-band absorption and the intervalley and intravalley phonon- and impurity-assisted free carrier absorption. Band-gap narrowing effects were included using empirical data. Biaxial strain values of ~1% can be achieved in the lasing region of the structure, which, although below the level required to convert germanium to a direct band-gap material, nonetheless result in net optical gain with practical (mid 1019cm-3) n-type doping densities, for a range of electrical injection conditions. References 1.J Liu et al, Opt. Lett 35 679 (2010) 2. R Camacho-Aguilera et al, Opt. Express 20 11316 (2012) 3. R Camacho-Aguilera et al, Appl. Phys. Lett 102 152106 (2013)

Hybrid structures are needed to fully exploit the great advantages of Si photonics and several approaches have been addressed where Si devices are bonded to different materials and nanostructures. Here we study the use of semiconductor carbon nanotubes for emission in the 1300 nm wavelength range to functionalize Si photonic structures in view of optoelectronic applications. The Si micro-rings are fully characterized by near field forward resonant scattering with 100 nm resolution. We show that both TE and TM modes can be addressed on the top of the micro-rings in a vectorial imaging of the in-plane polarization components. We coupled the Si micro-resonators with selected carbon nanotubes for high photoluminescence emission. Coupling nanotubes with the evanescent tails in air of the electric field localized in the photonic modes of the micro-resonators is demonstrated by sharp resonances over imposed to the nanotube emission bands. By mapping the Si and the nanotube emission we demonstrate that strong enhancement of the nanotube photoluminescence can be achieved both in the photonic modes of micro-disks and slot micro-rings, whenever the spatial overlap between nano-emitters and photonic modes is fulfilled.

Demanding applications such as video streaming, social networking, or web search relay on a large network of data centres, interconnected through optical links. The ever-growing data rates and power consumption inside these data centres are pushing copper links close to their fundamental limits. Optical interconnects are being extensively studied with the purpose of solving these limitations. Among the different possible technology platforms, silicon photonics, due to its compatibility with the CMOS platform, has become one of the preferred solutions for the development of the future generation photonic interconnects. However, the on-chip integration of all photonic and optoelectronic building blocks (sources, modulators and detectors…) is very complex and is not cost-effective due to the various materials involved (Ge for detection, doped Si for modulators and III-V for lasing). Carbon nanotubes (CNTs) are nanomaterials of great interest in photonics thanks to their fundamental optical properties, including near-IR room-temperature foto- and electro- luminescence, Stark effect, Kerr effect and absorption. In consequence, CNTs have the ability to emit, modulate and detect light in the telecommunications wavelength range. Furthermore, they are being extensively developed for new nano-electronics applications. In this work, we propose to use CNTs as active material integrated into silicon photonics for the development of all optoelectronic devices. Here, we report on the development of new integration schemes to couple the light emission from CNTs into optical resonators implemented on the silicon-on-insulator and silicon-nitride-on-insulator platforms. A theoretical and experimental analysis of the light interaction of CNTs with micro-ring resonators based on strip and slot waveguides and slot photonic crystal heterostructure cavities were carried out.

In this work, we present for the first time a partially slotted silicon ring resonator (PSRR) covered with an electro-optical polymer (Poly[(methyl methacrylate)-co-(Disperse Red 1 acrylate)]). The PSRR takes advantage of both a highly efficient vertical slot waveguide based phase shifter and a low loss strip waveguide in a single ring. The device is realized on 200 mm silicon-on-insulator wafers using 248 nm DUV lithography and covered with the electro-optic polymer in a post process. This silicon-organic hybrid ring resonator has a small footprint, high optical quality factor, and high DC device tunability. A quality factor of up to 10⁵ and a DC device tunability of about 700 pm/V is experimentally demonstrated in the wavelength range of 1540 nm to 1590 nm. Further, we compare our results with state-of-the-art silicon-organic hybrid devices by determining the poling efficiency. It is demonstrated that the active PSRR is a promising candidate for efficient optical switches and tunable filters.

Nonlinear periodically poled ridge LiNbO₃ waveguides have been fabricated on silicon substrates. Components are micromachined with a precision dicing machine and/or by grinding or polishing steps. They show efficient second harmonic generation at telecommunication wavelengths with normalized conversion reaching 600%/W in a 20mm long device. Influence of geometrical non uniformities of waveguides due to fabrication process is asserted. Components characteristics are studied notably their robustness and tunability versus temperature.

We present the use of femtosecond laser ablation for the removal of monolayer graphene from silicon-on-insulator (SOI) waveguides, and the use of oxygen plasma etching through a metal mask to peel off graphene from the grating couplers attached to the waveguides. Through Raman spectroscopy and atomic force microscopy, we show that the removal of graphene is successful with minimal damage to the underlying SOI waveguides. Finally, we employ both removal techniques to measure the contribution of graphene to the loss of grating-coupled graphene-covered SOI waveguides using the cut-back method. This loss contribution is measured to be 0.132 dB/μm.

Micro-disk resonators with high Q-factor have been experimentally demonstrated on germanium-on-insulator (GOI). GOI substrates fabricated by direct wafer bonding show better crystal quality that germanium films directly grown on Si. Sharp resonant peaks with Q-factor around 1000–4000 have been observed from micro-disks fabricated on GOI substrate by low-temperature photoluminescence measurements. The light emission properties against pump laser power and device temperature are also investigated. Our results indicating that GOI micro-disks are promising resonators for low threshold, ultra-compact Ge lasers on Si.

Energy-efficient integrated circuits for on-chip or chip-to-chip data transfer via photons could be tackled by monolithically grown group IV photonic devices. The major goal here is the realization of fully integrated group IV room temperature electrically driven lasers. An approach beyond the already demonstrated optically-pumped lasers would be the introduction of GeSn/(Si)Ge(Sn) heterostructures and exploitation of quantum mechanical effects by reducing the dimensionality, which affects the density of states. In this contribution we present epitaxial growth, processing and characterization of GeSn/(Si)Ge(Sn) heterostructures, ranging from GeSn/Ge multi quantum wells (MQWs) to GeSn quantum dots (QDs) embedded in a Ge matrix. Light emitting diodes (LEDs) were fabricated based on the MQW structure and structurally analyzed via TEM, XRD and RBS. Moreover, EL measurements were performed to investigate quantum confinement effects in the wells. The GeSn QDs were formed via Sn diffusion /segregation upon thermal annealing of GeSn single quantum wells (SQW) embedded in Ge layers. The evaluation of the experimental results is supported by band structure calculations of GeSn/(Si)Ge(Sn) heterostructures to investigate their applicability for photonic devices.

Germanium is a strong candidate as a laser source for silicon photonics. Despite the indirect nature of its bandgap, the application of several percent of tensile strain reduces the energy difference between its direct and indirect bandgaps [1]. It has been predicted that above a certain strain threshold, germanium transforms into an actual direct bandgap material [2]. However, the properties of this material at unprecedented levels of strain still raise issues. A recently introduced strain technology based on prestressed germanium layers [3] enables to fabricate micro-membranes at such high strain. We present here both a theoretical and experimental study of the band edge and Raman shift at such high strain level. For above reasons, we start from slightly tensile-strained germanium-on-insulator (GeOI) substrates obtained by the Smart CutTM technology [4]. By etching adequate pattern in the germanium layer, both uniaxial and biaxial stress conditions were obtained after etching away the sacrificial buried oxide underneath the germanium layer. We performed x-ray diffraction measurements at the ESRF synchrotron using Laue in combinations with rainbow filtering techniques on the micro-membranes revealing strain values of 4.9 % for the uniaxial strain and 1.9 % for the biaxial strain [5-6]. We then studied the relationship between strain and Raman shift. While the relationship remained linear for biaxial stress condition, a significant deviation from the linear behavior behavior was observed above 2.5 % uniaxial strain. Such nonlinearity becomes dominant at very high strain levels; indeed a 9.9 cm-1 Raman shift corresponds to 4.9 % strain instead of the 6.5 % predicted by the linear extrapolation [5]. We performed simulations of the band structure of germanium under various stress conditions using a tight-binding model. For uniaxial stress, the relation between the energy positions of the band edges differed significantly from the deformation potential models in [1]. Finally, we performed electro-absorption measurements on micro-membranes to determine the energy of the direct transitions (conduction band to light and heavy holes) in uniaxially stressed germanium. The relationship between strain and direct bandgap became nonlinear above 2.5 %, in agreement with our theoretical models. In conclusion, we show that under uniaxial strain level above 2 %, germanium exhibits significant nonlinear behaviors which have to be taken into consideration for the design and fabrication of future on-chip germanium laser sources compatible with CMOS technologies. [1] Van de Walle, Phys. Rev. B 39, 1871 (1989) [2] Lim et al., Opt. Express 17, 16358 (2009) [3] Süess et al., Nature Phot. 7, 466 (2013) [4] Widiez et al., ECS Transaction 64, 35-48 (2014) [5] Gassenq et al., submitted [6] Gassenq et al., submitted

Recently, rare-earth doped waveguide amplifiers (REDWAs) have drawn significant attention as a promising solution to on-chip amplification of light in silicon photonics and integrated optics by virtue of their high excited state lifetime (up to 10 ms) and broad emission spectrum (up to 200 nm) at infrared wavelengths. In the family of rare-earths, at least erbium, holmium, thulium, neodymium and ytterbium have been demonstrated to be good candidates for amplifier operation at moderate concentrations (< 0.1 %). However, efficient amplifier operation in REDWAs is a very challenging task because high concentration of ions (<0.1%) is required in order to produce reasonable amplification over short device length. Inevitably, high concentration of ions leads to energy-transfer between neighboring ions, which results as decreased gain and increased noise in the amplifier system. It has been shown that these energy-transfer mechanisms in highly-doped gain media are inversely proportional to the sixth power of the distance between the ions. Therefore, novel fabrication techniques with the ability to control the distribution of the rare-earth ions within the gain medium are urgently needed in order to fabricate REDWAs with high efficiency and low noise. Here, we show that atomic layer deposition (ALD) is an excellent technique to fabricate highly-doped (<1%) RE:Al2O3 gain materials by using its nanoscale engineering ability to delicately control the incorporation of RE ions during the deposition. In our experiment, we fabricated Er:Al2O3 and Tm:Al2O3 thin films with ALD by varying the concentration of RE ions from 1% to 7%. By measuring the photoluminescence response of the fabricated samples, we demonstrate that it is possible to incorporate up to 5% of either Er- or Tm-ions in Al2O3 host before severe quenching occurs. We believe that this technique can be extended to other RE ions as well. Therefore, our results show the exceptionality of ALD as a deposition technique for REDWA technology.

Silicon photonics technology is an enabler for the integration of complex circuits on a single chip, for various optical link applications such as routing, optical networks on chip, short range links and long haul transmitters. Quadrature Phase Shift Keying (QPSK) transmitters is one of the typical circuits that can be achieved using silicon photonics integrated circuits. The achievement of 25GBd QPSK transmitter modules requires several building blocks to be optimized: the pn junction used to build a BPSK (Binary Shift Phase Keying) modulator, the RF access and the optical interconnect at the package level. In this paper, we describe the various design steps of a BPSK module and the related tests that are needed at every stage of the fabrication process.

We propose a silicon dual-ring modulator consisting of two serially cascaded rings embedded with p-i-n-i-p junctions driven by one signal, in contrast to a differential signal pair. The simulations for optimizing the design of the phase shifter and the optical response of the both rings are performed. We show that the device performs better than the single ring. The modulator has a higher optical bandwidth, 3-dB modulation bandwidth and bit rate as compared to a single ring. In contrast to a cascaded ring modulator driven by a differential signal pair, it can be driven by a single p-i-n-i-p junction and RF signal. A serially coupled ring resonator has three times the resonance linewidth as compared to a single ring. Furthermore, it can support significantly higher data rates of up to 13 GHz.

We present two novel capacitive modulator phase shifter architectures using a vertical oxide. The first structure consists of a vertical oxide slot embedded in a silicon waveguide along the propagation direction. The second structure is based on a sub-wavelength approach, with several periodic vertical oxide layers along the propagation direction. This paper focuses on the design of the modulator devices and the simulation of their performances.

Modulators based on free-carrier plasma dispersion effect has attracted a lot of attention in the recent years. Particularly, a lot of effort is dedicated nowadays into optimizing carrier accumulation silicon modulators. In this work a novel structure of carrier accumulation modulator is proposed and analyzed. The structure is based on interleaved capacitors in a periodic segmented waveguide such as SubWavelength Gratings (SWG) waveguides. This new structure overcomes the major limitation of carrier accumulation structures, which is the limited overlap of the optical mode with the region where the carrier accumulation takes place. The analysis of this novel structure is presented in detail from the optical and the electrical point of view. Furthermore, the procedure to analyze and extract the modulator performance is also presented. Values of modulation efficiency and loss below 0.5Vcm and 5 dB respectively were obtained for the proposed modulator.

Electro-absorption and electro-refraction in Ge/SiGe coupled quantum wells (CQW) grown on Si have been investigated by means of optical transmission measurements. The separate confinement of electrons and holes in the heterostructure gives rise to an anomalous Quantum Confined Stark Effect (QCSE) that can be exploited to strongly enhance the electro refractive effect with respect to uncoupled quantum wells. A refractive index variation up to 2.3 x 10^-3 has been measured at 1.5 V, with an VπLπ of 0.046 V cm. This result is very promising for the realization of an efficient and compact phase modulator based on the Ge/SiGe material system.

Electronic circuit designers commonly start their design process with a schematic, namely an abstract representation of the physical circuit. In integrated photonics on the other hand, it is common for the design to begin at the physical component level, and create a layout by connecting components with interconnects. In this paper, we discuss how to create a schematic from the physical layout via netlist extraction, which enables circuit simulations. Post-layout extraction can also be used to predict how fabrication variability and non-uniformity will impact circuit performance. This is based on the component position information, compact models that are parameterized for dimensional variations, and manufacturing variability models such as a simulated wafer thickness map. This final step is critical in understanding how real-world silicon photonic circuits will behave. We present an example based on treating the ring resonator as a circuit. A silicon photonics design kit, as described here, is available for download at http://github.com/lukasc-ubc/SiEPIC_EBeam_PDK.

A simple and novel method to make a silicon ring resonator without the limit of free spectral range (FSR) is presented. For many applications, a ring resonator is desired with an ultra-wide FSR, a high quality factor, a large extinction ratio and low-fabrication complexity. In this paper we propose a novel method to obtain such a ring resonator, which is a single all-silicon micro-ring with a FSR as large as 150 nm around 1550 nm. It's based on the well-known phenomenon of resonance splitting.

We present an effective approach to evaluate the performance of multi-channel silicon (Si) photonic systems. The system is composed of strip Si photonic waveguides (Si-PhWs) with uniform cross-section or photonic-crystal (PhC) Si waveguides (Si-PhCWs), combined with a set of direct-detection receivers. Moreover, the optical field in each channel is the superposition of a continuous-wave nonreturn-to-zero ON-OFF keying modulated signal and a white Gaussian noise. In order to characterize the optical signal propagation in the waveguides, an accurate mathematical model describing all relevant linear and nonlinear optical effects and its linearized version is employed. In addition, two semi-analytical methods, time- and frequency-domain Karhunen-Loève series expansion, are used to assess the system bit-error-rate (BER). Our analysis reveals that Si-PhCWs provide similar performance as Si-PhWs, but for 100× shorter length. Importantly, much worse BER is achieved in Si-PhCWs when one operates in slow-light regime, due to the enhanced linear and nonlinear effects.

Abstract Integrated photonics, and especially silicon photonics, has been rapidly expanded its catalog of building blocks and functionalities. Now, it is maturing fast towards circuit-level integration to serve more complex applications in industry. However, performance variability due to the fabrication process and operational conditions can limit the yield of large-scale circuits. It is essential to assess this impact at the design level with an efficient variability analysis: how variations in geometrical, electrical and optical parameters propagate into components performance. In particular when implementing wavelength-selective filters, many primary functional parameters are affected by fabrication-induced variability. The key functional parameters that we assess in this paper are the waveguide propagation constant (the effective index, essential to define the exact length of a delay line) and the coupling coefficients in coupling structure (necessary to set the power distribution over different delay lines). The Monte Carlo (MC) method is the standard method for variability analysis, thanks to its accuracy and easy implementation. However, due to its slow convergence, it requires a large set of samples (simulations or measurements), making it computationally or experimentally expensive. More efficient methods to assess such variability can be used, such as generalized polynomial chaos (gPC) expansion or stochastic collocation. In this paper, we demonstrate stochastic collocation (SC) as an efficient alternative to MC or gPC to characterize photonic devices under the effect of uncertainty. The idea of SC is to interpolate stochastic solutions in the random space by interpolation polynomials. After sampling the deterministic problem at a pre-defined set of nodes in random space, the interpolation is constructed. SC drastically reduces computation and measurement cost. Also, like MC method, sampling-based SC is easy to implement. Its computation cost can be further reduced by choosing proper sparse grid. Meanwhile, it offers similar high accuracy and efficiency as gPC method. We applied SC method to analyze the variability of waveguides and directional couplers. They are both building blocks in the construction of larger photonics components such as higher-order ring resonator or Mach-Zehnder filters. These wavelength filters cannot be constructed without a proper control of waveguides and directional couplers properties. However, in silicon photonics, waveguides and directional couplers are susceptible to fabrication variations: a small change in linewidth or thickness of the core can dramatically vary effective indices and coupling coefficients. We applied SC technique to map the variability of waveguides and directional couplers. We verified its accuracy and efficiency by comparing the results with a classic MC analysis. In the simulation, this led to a speedup of 146×. We also applied the technique to analyze effective index of waveguides and directional couplers measured on wafer-scale. Predicted variability by SC shows good agreement with measured variability. The result of this research is not limited to coupling devices. Thanks to the flexibility in the choice of the interpolation schemes and the efficiency of sparse grid sampling to choose the collocation nodes for multiple-dimension, the proposed approach is flexible and can be applied to study a broad range of photonic devices.

A key enabler of the IT revolution of the late 20th century was the development of electronic design automation (EDA) tools allowing engineers to manage the complexity of electronic circuits with transistor counts now reaching into the billions. Recently, we have been developing large-scale nonlinear photonic integrated logic circuits for next generation all-optical information processing. At this time a sufficiently powerful EDA-style software tool chain to design this type of complex circuits does not yet exist. Here we describe a hierarchical approach to automating the design and validation of photonic integrated circuits, which can scale to several orders of magnitude higher complexity than the state of the art. Most photonic integrated circuits developed today consist of a small number of components, and only limited hierarchy. For example, a simple photonic transceiver may contain on the order of 10 building-block components, consisting of grating couplers for photonic I/O, modulators, and signal splitters/combiners. Because this is relatively easy to lay out by hand (or simple script) existing photonic design tools have relatively little automation in comparison to electronics tools. But demonstrating all-optical logic will require significantly more complex photonic circuits containing up to 1,000 components, hence becoming infeasible to design manually. Our design framework is based off Python-based software from Luceda Photonics which provides an environment to describe components, simulate their behavior, and export design files (GDS) to foundries for fabrication. At a fundamental level, a photonic component is described as a parametric cell (PCell) similarly to electronics design. PCells are described by geometric characteristics of their layout. A critical part of the design framework is the implementation of PCells as Python objects. PCell objects can then use inheritance to simplify design, and hierarchical designs can be made by creating composite PCells (modules) which consist of primitive building-block PCells (components). To automatically produce layouts, we built on a construct provided by Luceda called a PlaceAndAutoRoute cell: we create a module component by supplying a list of child cells, and a list of the desired connections between the cells (e.g. the out0 port of a microring is connected to a grating coupler). This functionality allowed us to write algorithms to automatically lay out the components: for instance, by laying out the first component and walking through the list of connections to check to see if the next component is already placed or not. The placement and orientation of the new component is determined by minimizing the length of a connecting waveguide. Our photonic circuits also utilize electrical signals to tune the photonic elements (setting propagation phases or microring resonant frequencies via thermo-optical tuning): the algorithm also routes the contacts for the metal heaters to contact pads at the edge of the circuit being designed where it can be contacted by electrical probes. We are currently validating a test run fabricated over the summer, and will use detailed characterization results to prepare our final design cycle in which we aim to demonstrate complex operational logic circuits containing ~50-100 nonlinear resonators.

The tremendous progress in the fabrication of highly confining silicon-on-insulator (SOI) waveguides has been very beneficial for four-wave-mixing (FWM)-based wavelength conversion applications. Nevertheless, to establish power-efficient and wideband FWM wavelength conversion, one typically requires long (cm-scale) SOI waveguides with dispersion-engineered cross-sections that do not comply with the fabrication constraints of multiproject- wafer-oriented silicon photonics foundries. In this paper, we numerically examine the opportunities for wideband wavelength conversion through FWM in a foundry-compatible SOI waveguide covered with the highly nonlinear two-dimensional material of graphene. When combining subwatt level pump powers with a short waveguide length of only a few hundreds of microns, perfectly phase-matched conversion with significant efficiencies close to 20 dB can be obtained over a more than 40 THz-wide signal band adjacent to the pump frequency. Because of the tunability of the graphene properties, it is also possible to obtain quasi-phase matched FWM conversion through a periodic sign reversal of the graphene third-order nonlinearity along the waveguide. Conversion efficiencies exceeding 30 dB can be achieved over a 3.4 THz-wide signal band that is situated as much as 58 THz away from the pump frequency. Finally, the graphene tunability also allows for switching between the perfectly phase-matched and quasi-phase-matched operation modes.

A silicon photonics platform that uses a CMOS foundry line is described. Fabrication process is following a modular integration scheme which leads to a flexible platform, allowing different device combinations. A complete device library is demonstrated for 1310 nm applications with state of the art performances. A PDK which includes specific photonic features and which is compatible with commercial EDA tools has been developed allowing an MPW shuttle service. Finally platform evolutions such as device offer extension to 1550 nm or new process modules introduction are presented.

A new technological platform aimed at making prototypes and feasibility studies has been setup at STMicroelectronics using 300mm wafer foundry facilities. The technology, called DAPHNE (Datacom Advanced PHotonic Nanoscale Environment), is devoted at developing and evaluating new devices and sub-systems in particular for wavelength division multiplexing (WDM) applications and ring resonator based applications. Developed in the course of PLAT4MFP7 European project, DAPHNE is a flexible platform that fits perfectly R&D needs. The fabrication flow enables the processing of photonic integrated circuits using a silicon-on-insulator (SOI) of 300nm, partial etches of 150nm and 50nm and a total silicon etching. Consequently, two varieties of rib waveguides and one strip waveguide can be fabricated simultaneously with auto-alignment properties. The process variability on the 150nm partially etched silicon and the thin 50nm slab region are both less than 6 nm. Using a variety of different implantation configurations and a back-end of line of 5 metal layers, active devices are fabricated both in germanium and silicon. An available far back-end of line process consists of making 20 μm diameter copper posts on top of the electrical pads so that an electronic integrated circuit can be bonded on top the photonic die by 3D integration. Besides having those fabrication process options, DAPHNE is equipped with a library of standard cells for optical routing and multiplexing. Moreover, typical Mach-Zehnder modulators based on silicon pn junctions are also available for optical signal modulation. To achieve signal detection, germanium photodetectors also exist as standard cells. The measured single-mode propagation losses are 3.5 dB/cm for strip, 3.7 dB/cm for deep-rib (50nm slab) and 1.4 dB/cm for standard rib (150nm slab) waveguides. Transition tapers between different waveguide structures are as low as 0.006 dB.

Silicon rib waveguides enable single-mode (SM) operation even with the combination of multi-micron core dimensions and high refractive index contrast. In such large waveguides the optical mode field is almost completely confined inside the Si core, which leads to small propagation losses and small polarization dependency. The unique SM condition of the rib waveguide also enables the use of an ultra-wide wavelength range, for example from 1.2 to <1.7 μm, without sacrificing either SM operation or low propagation loss. This makes micron-scale Si waveguides particularly well-suited for spectroscopy and extensive wavelength division multiplexing. However, rib waveguides require large bending radii, which lead to large circuit sizes. There are two solutions for this. So-called Euler bends in Si strip waveguides enable low-loss bends down to 1 μm bending radius with less than 0.1 dB/90° loss for both polarizations. Another alternative is a total-internal reflection mirror that can have loss as low as 0.1 dB for both polarizations in either strip or rib waveguides. The excitation of higher order modes in large strip waveguides is avoided by using adiabatic rib-strip converters and low-loss components. With rib and strip waveguides it is possible to reach a unique combination of low loss, extremely small footprint, small polarization dependency, ultra-wide bandwidth and tolerance to high optical powers.

The heterogeneous integration of SiGe/Ge and III-V semiconductors gives us an opportunity to enhance functionalities of Si photonics platform through their superior material properties which lack in Si. In this paper we discuss what SiGe/Ge and III-V can bring to Si photonics. We have predicted that the light effective hole mass in strained SiGe results in the enhanced the free-carrier effects such as the plasma dispersion effect and free-carrier absorption. We observed significantly larger free-carrier absorption in the SiGe optical modulator than in the control Si device. By fabricating asymmetric Mach-Zehnder interferometer (MZI) SiGe optical modulators, the enhancement of the plasma dispersion effect in strained SiGe has been successfully demonstrated. Mid-infrared integrated photonics based on Ge waveguides on Si have also been investigated. Since Ge is transparent to the entire mid-infrared range, Ge photonic integrated circuits on the Ge-on-Insulator (GeOI) wafer are quite attractive. We have successfully fabricated the GeOI wafer with 2-μm-thick buried oxide (BOX) layer by wafer bonding. The passive waveguide components based on Ge strip waveguides have been demonstrated on the GeOI. We have also demonstrated carrier-injection Ge variable optical attenuators. We have proposed and investigate the III-V CMOS photonics platform by using the III-V on Insulator (IIIV- OI) on a Si wafer. The strong optical confinement in the III-V-OI enables us to achieve high-performance photonic devices. We have successfully demonstrated InGaAsP MZI optical switch with the low on-state crosstalk on the III-V-OI. Ultra-low dark current waveguide InGaAs PDs integrated with an InP grating coupler are also achieved.

The mid-Infrared wavelength range (2-20 µm), so-called fingerprint region, contains the very sharp vibrational and rotational resonances of many chemical and biological substances. Thereby, on-chip absorption-spectrometry-based sensors operating in the mid-Infrared (mid-IR) have the potential to perform high-precision, label-free, real-time detection of multiple target molecules within a single sensor, which makes them an ideal technology for the implementation of lab-on-a-chip devices. Benefiting from the great development realized in the telecom field, silicon photonics is poised to deliver ultra-compact efficient and cost-effective devices fabricated at mass scale. In addition, Si is transparent up to 8 µm wavelength, making it an ideal material for the implementation of high-performance mid-IR photonic circuits. The silicon-on-insulator (SOI) technology, typically used in telecom applications, relies on silicon dioxide as bottom insulator. Unfortunately, silicon dioxide absorbs light beyond 3.6 µm, limiting the usability range of the SOI platform for the mid-IR. Silicon-on-sapphire (SOS) has been proposed as an alternative solution that extends the operability region up to 6 µm (sapphire absorption), while providing a high-index contrast. In this context, surface grating couplers have been proved as an efficient means of injecting and extracting light from mid-IR SOS circuits that obviate the need of cleaving sapphire. However, grating couplers typically have a reduced bandwidth, compared with facet coupling solutions such as inverse or sub-wavelength tapers. This feature limits their feasibility for absorption spectroscopy applications that may require monitoring wide wavelength ranges. Interestingly, sub-wavelength engineering can be used to substantially improve grating coupler bandwidth, as demonstrated in devices operating at telecom wavelengths. Here, we report on the development of fiber-to-chip interconnects to ZrF4 optical fibers and integrated SOS circuits with 500 nm thick Si, operating around 3.8 µm wavelength. Results on facet coupling and sub-wavelength engineered grating coupler solutions in the mid-IR regime will be compared.

We report on miniaturized optical spectrometers integrated on a photonic integrated circuit (PIC) platform based on silicon nitride waveguides and fabricated in a CMOS-compatible approach. As compared to a silicon- on-insulator PIC-platform, the usage of silicon nitride allows for operation in the visible and near infrared. Furthermore, the moderately high refractive index contrast in silicon-nitride photonic wire waveguides provides a valuable compromise between compactness, optical loss and sensitivity to phase error. Three generic types of on-chip spectrometers are discussed: the arrayed waveguide grating (AWG) spectrometer, the echelle grating or planar concave grating (PCG) spectrometer and the stationary Fourier transform spectrometer (FTS) spectrometer. Both the design as well as experimental results are presented and discussed. For the FTS spectrometer a specific design is described in detail leading to an ultra-small (0.1 mm²) footprint device with a resolution of 1 nm and a spectral range of 100nm. Examples are given of the usage of these spectrometers in refractive index biosensing, absorption spectroscopy and Raman spectroscopy.

Fluorescence detection is a commonly used technique to detect particles. Microscopes are used for the fluorescence detection of the micro-particles. However, the conventional microscopes are bulky. It is cumbersome to integrate all the equipment used for detection in one setup. They can be replaced by photonic chips for the detection of micro-particles such as cells. Most of the biological detection techniques require the utilization of the visible range of the spectrum. SiN as a waveguide material stands out for biological applications due to its transparency in the visible spectrum. Specifically designed grating couplers can be exploited to focus from inside SiN waveguides at a designated location above the chip. Those SiN focusing grating couplers can mimic microscope objectives for on-chip biological detection applications such as fluorescence and Raman spectroscopy. In this report, we present a 2D SiN focusing grating coupler. We study the effect of the grating design on the focus properties of visible light using finite-difference time-domain simulations.

The integration of the Whispering Gallery Modes (WGMs) resonators in a microfluidics platform represents an important feature towards the realization of a compact high performance label-free biosensor. These hollow resonant microstructures present the advantage to combine the WGM resonator properties with the intrinsic capability of integrated microfluidics. In this sense, optical microbubble resonators (OMBRs), intended as a hollow core spherical bulge realized in a glass microcapillary by a suitable fabrication process, with their high Q factors (< 107 in air) well satisfy this requirement. Their operation is based on the fact that, given a small enough wall thickness of the bubble, the WGM optical field extends on both sides of the wall, so that it is possible to couple light into the resonator from an outer waveguide, and at the same time to have interaction of the WGM field with the inner fluid and analyte. The biosensing mechanism of these devices is based on the WGMs morphological dependence: any change on the OMBR inner surface, due to some chemical and/or biochemical binding, causes a shift of the resonance position and reduces the Q factor of the OMBR. By measuring these changes, important information about the sensing capability of the device can be obtained. In order to develop an OMBR based biosensor and optimize its performance, a crucial step is represented by its chemical/biochemical functionalization. Here we present a novel technique able to guarantee that the chemical interaction occurs in the OMBR inner wall, leaving the other microfluidic parts completely inert from a biochemical point of view. The method is based on UV photoactivation, which allows to localize the biolayers only in correspondence of the OMBR inner wall. As a proof of concept, an immunoassay based on rabbit IgG/anti rabbit-IgG interaction was performed and. The anti rabbit-IgG antibody was labelled with Alexa Fluor 488 to verify, by a fluorescence characterization, the goodness of this procedure. Moreover, an anti mouse-IgG, labelled with the same fluorophore (Alexa Fluor 488) was used for specificity-tests of the IgG/anti-IgG interaction. The immunoassay based on fluorescence was characterized using an optical microscope (Zeiss AXIO inverted fluorescence microscope) working at the wavelengths of 470 nm for excitation of Alexa Fluor 488. The real time measurement of the resonance broadening after each functionalization step together with the high Q factor (< 105) measured after the IgG/anti-IgG interaction in water, gives a further proof for the method validity.

Bio-detection based on CMOS technology boosts the miniaturization of detection systems and the success on highly efficient, robust, accurate, and low coast Lab-on-Chip detection schemes. Such on chip detection technologies have covered healthy related harmful gases, bio-chemical analytes, genetic micro RNA, etc. Their monitoring accuracy is mainly qualified in terms of sensitivity and limit of the detection (LOD) of the detection system. In this context, recently developed silicon on insulator (SOI) optical devices have displayed highly performant detection abilities that LOD could go beyond 10-8RIU and sensitivity could exceeds 103nm/RIU. The SOI integrated optical sensing devices include strip/slotted waveguide consisting in structures like Mach-Zehnder interferometers (MZI), ring resonators (RR), nano cavities, etc. Typically, hollow core RR and nano-cavities could exhibit higher sensitivity due to their optical mode confinement properties with a partial localization of the electric field in low index sensing regions than devices based on evanescent field tails outside of the optical cores. Furthermore, they also provide larger sensing areas for surface functionalization to reach higher sensitivities and lower LODs. The state of art of hollow core devices, either based on Bragg gratings formed from a slot waveguide cavity or photonic crystal slot cavities, show sensitivities (S) up to 400nm/RIU and Figure of Merit (FOM) around 3,000 in water environment, FOM being defined as the inverse of LOD and precisely as FOM=SQ/λ, with λ the resonance wavelength and Q the quality factor of the considered resonator. Such high achieved FOMs in nano cavities are mainly due to their large Q factors around 15,000. While for mostly used RR, which do not require particular design strategies, relatively low Q factors around 1800 in water are met and moderate sensitivities about 300nm/RIU are found. In this work, we present here a novel slot ring resonator design to make breakthrough of the performance of slot ring resonator sensing ability. Different from the normal sensing regime by monitoring one specific resonance (λres) peak shift, the proposed approach stems from the sensitivity of the RR critical coupling. The critical coupling peak is auto-selected out by matching the following condition: the ring resonator’s round trip attenuation coefficient a(λ) being equal to the coupler self-coupling coefficient k(λ), thus resulting in the deepest extinction ratio (ER) among the spectrum RR comb. The obtained sensing comb, based on a V-shape spectrum envelop, is engineered by controlling a(λ) and k(λ) with opposite monotonicities. Both a(λ)and k(λ) are tuned to have a large dispersion along the wavelength, which means that |a(λ)-k(λ)| keeps rapidly increasing as λres is far away from λc, eliminating the resonance ER quickly down to 0. Experimentally, slot waveguide ring resonators with a radius of 50µm have been fabricated on a standard silicon platform with a Si thickness of 220nm, loaded by racetrack couplers with a straight coupling length of 20µm. Sensing experiments have been carried out by changing the top cladding material from a series of Cargille optical liquids with refraction index values ranging from 1.3 to 1.5. The Q factors of critical coupling resonances was monitored from 2,000 to 6,000, and measured wavelength shifts of this peak are from 1.41µm to 1.56µm. The maximum sensitivity of 1300nm/RIU is observed in the cladding index range 1.30-1.35. To conclude, a new sensing regime by tracking the critical coupling resonance λc of slot waveguide ring resonators is demonstrated. The reported sensitivity is up 1300nm/RIU around the water RI of 1.33, and the monitored sensing FOM is about 2300, which is very close to the FOM values achieved from nanobeam cavities. This work can thus contribute to future integrated optical sensing schemes based on slot RRs.

Fully automated, high precision, cost-effective assembly technology for photonic packages remains one of the main challenges in photonic component manufacturing. Next to the cost aspect the most demanding assembly task for multiport photonic integrated circuits (PICs) is the high-precision (±0.1 μm) alignment and fixing required for optical I/O in InP PICs, even with waveguide spot size conversion. In a European research initiative – PHASTFlex - we develop and investigate an innovative, novel assembly concept, in which the waveguides in a matching TriPleX interposer PIC are released during fabrication to make them movable. After assembly of both chips by flip-chip bonding on a common carrier, TriPleX based actuators and clamping functions position and fix the flexible waveguides with the required accuracy.

Designing photonic integrated circuits (PICs) with packaging in mind is important since this impacts the performance of the final product. In coherent optical communication applications there are a large number of DC and RF lines that need routed to connect the PIC to the outer packaging. These RF lines should be impedance matched to the devices, isolated from each other, low loss and protected against electromagnetic interference (EMI) over the frequency range of interest to achieve the performance required for the application. Multilevel low temperature co-fired ceramic (LTCC) boards can be used as a carrier board connecting the PIC to the packaging due to its good RF performance, machinability, compatibility with hermetic sealing, and ability to integrate drivers into the board. Flexibility with layer numbers enables additional layers for shielding against electromagnetic interference or increased space for routing electrical connections. In this paper the design, simulations, and measured results for a set of 4 phase matched transmission lines in LTCC that would be used with an IQ MZM are presented. The measured 3dB bandwidth for a set of four phase matched transmission lines for an IQ MZM was measured to be 19.8 GHz.

The packaging of high speed Photonic Integrated Circuits (PICs) should maintain the electrical signal integrity. The standard packaging of high speed PICs relies on wire bonds. This is not desirable because wire bonds degrade the quality of the electrical signal. The research presented in this paper proposes to replace wire bonds with an interposer with multilevel transmission lines. By attaching the PIC by flip chip onto the interposer, the use of wire bonds is avoided. The main concern for designing an interposer with multilevel transmission lines is the vertical transition, which must be designed to avoid return and radiation losses. In this paper, a novel design of a high speed vertical transition for Low Temperature Co-fired Ceramic (LTCC) is presented. The proposed vertical transition is simpler than others recently published in the literature, due to eliminating the need for additional ceramic layers or air cavities. A LTCC board was fabricated with several variations of the presented transition to find the optimal dimensions of the structure. The structures were fabricated then characterized and have a 3 dB bandwidth of 37 GHz and an open eye diagram at 44 Gbps. A full wave electromagnetic simulation is described and compared with good agreement to the measurements. The results suggest that an LTCC board with this design can be used for 40 Gbps per channel applications. Keywords: Photonics packaging, Low Temperature Co-Fired Ceramics.

This paper reports on the progress related to a multichannel photonic alignment concept, aiming for sub-micrometer precision in the alignment of the waveguides of two photonic integrated circuits (PICs). The concept consists of two steps: chip-to-chip positioning and chip bonding provide a coarse alignment after which waveguide-to-waveguide positioning and fixing result in a fine alignment. For the waveguide-to-waveguide alignment, an alignment functionality is developed and integrated in one of the PICs, consisting of mechanically flexible waveguides and MEMS actuators. This paper reports on the fabrication and characterization of a mechanically flexible waveguide array that can be positioned by two out-of-plane actuators.

Thermal actuators are integrated with mechanically flexible waveguide beams to enable positioning them with high precision. By adding a poly-Si pattern on top of SiO₂ beams, an out-of-plane bimorph actuator can be realized. An analytical model enables estimating the curvature and the deflection of a single bimorph beam. Acquiring a small initial deflection while having a large motion range of the actuator proves to have conflicting demands on the poly-Si/SiO₂ thickness ratio.

In this paper, we show that suspended waveguide arrays with integrated alignment functionality have an initial deflection- they curl up- due to residual stress in the materials. The actuators can be operated using a driving voltage between 0V to 45V, corresponding to ~50mW. Using higher voltages brings the risk of permanently changing the material properties of the heaters. The actuators can accomplish an out-of-plane crossbar translation up to 6.5 μm at ~50mW as well as a rotation around the propagation direction of the light ranging from -0:1° to 0.1°. At a constant actuation power of ~50mW, the crossbar shows a drift in vertical deflection of 0.16 μm over a time of 30 min.

Several 3D multilayer silicon photonics platforms have been proposed to provide densely integrated structures for complex integrated circuits. Amongst these platforms, great interest has been given to the inclusion of silicon nitride layers to achieve low propagation losses due to their capacity of providing tight optical confinement with low scattering losses in a wide spectral range. However, none of the proposed platforms have demonstrated the integration of active devices. The problem is that typically low loss silicon nitride layers have been fabricated with LPCVD which involves high processing temperatures (<1000 ºC) that affect metallisation and doping processes that are sensitive to temperatures above 400ºC. As a result, we have investigated ammonia-free PECVD and HWCVD processes to obtain high quality silicon nitride films with reduced hydrogen content at low temperatures. Several deposition recipes were defined through a design of experiments methodology in which different combinations of deposition parameters were tested to optimise the quality and the losses of the deposited layers. The physical, chemical and optical properties of the deposited materials were characterised using different techniques including ellipsometry, SEM, FTIR, AFM and the waveguide loss cut-back method. Silicon nitride layers with hydrogen content between 10-20%, losses below 10dB/cm and high material quality were obtained with the ammonia-free recipe. Similarly, it was demonstrated that HWCVD has the potential to fabricate waveguides with low losses due to its capacity of yielding hydrogen contents <10% and roughness <1.5nm.

Realization of a monolithically integrated on-chip laser source remains the holy-grail of Silicon Photonics. Germanium (Ge) is a promising semiconductor for lasing applications when highly doped with Phosphorous (P) and or alloyed with Sn [1, 2]. P doping makes Ge a pseudo-direct band gap material and the emitted wavelengths are compatible with fiber-optic communication applications. However, in-situ P doping with Ge2H6 precursor allows a maximum active P concentration of 6×1019 cm-3 [3]. Even with such active P levels, n++ Ge is still an indirect band gap material and could result in very high threshold current densities. In this work, we demonstrate P-doped Ge layers with active n-type doping beyond 1020 cm-3, grown using Ge2H6 and PH3 and subsequently laser annealed, targeting power-efficient on-chip laser sources. The use of Ge2H6 precursors during the growth of P-doped Ge increases the active P concentration level to a record fully activated concentration of 1.3×1020 cm-3 when laser annealed with a fluence of 1.2 J/cm2. The material stack consisted of 200 nm thick P-doped Ge grown on an annealed 1 µm Ge buffer on Si. Ge:P epitaxy was performed with PH3 and Ge2H6 at 320oC. Low temperature growth enable Ge:P epitaxy far from thermodynamic equilibrium, resulting in an enhanced incorporation of P atoms [3]. At such high active P concentration, the n++ Ge layer is expected to be a pseudo-direct band gap material. The photoluminescence (PL) intensities for layers with highest active P concentration show an enhancement of 18× when compared to undoped Ge grown on Si as shown in Fig. 1 and Fig. 2. The layers were optically pumped with a 640 nm laser and an incident intensity of 410 mW/cm2. The PL was measured with a NIR spectrometer with a Hamamatsu R5509-72 NIR photomultiplier tube detector whose detectivity drops at 1620 nm. Due to high active P concentration, we expect band gap narrowing phenomena to push the PL peak to wavelengths beyond the detection limit (1620nm) of the setup. Therefore, the 18× enhancement is a lower limit estimation. In this contribution, an extensive study of laser annealing conditions and their impact on material properties will be discussed. A major concern in using highly doped Ge as an active medium is the increase in free-carrier absorption (FCA). However, results reported in [4] suggest that FCA is significantly dominated by holes due to larger absorption cross-section of holes compared to electrons. The FCA results in [4] and JDOS modeling were used to calculate the gain spectrum for the highest doped Ge samples, including the typical 0.25% biaxial tensile strain of epitaxial Ge on Si. A carrier lifetime of 3 ns is required as shown in Fig. 3 for a target threshold current density of sub-20 kA/cm2 which represents at least tenfold reduction when compared to active P-doping level of 6×1019 cm-3. As a result, laser annealed highly doped Ge layers grown with Ge2H6 precursors are a promising approach for realizing a power efficient on-chip Ge laser source.

We present the technology steps to integrate an Echelle grating in the process flow of silicon-organic hybrid (SOH) modulators or related active devices. The CMOS-compatible process flow on SOI substrates uses a mix of optical i-line lithography and electron beam lithography (EBL). High speed optical data communication depends on wavelength divisions multiplexing and de-multiplexing devices like Echelle gratings. The minimum feature sizes vary from device to device and reach down to 60 nm inside a modulator, while the total area of a single Echelle grating is up to several mm² of unprocessed silicon. Resist patterning using a variable shape beam electron beam pattern generator allows high resolution. An oxide hard mask is deposited, patterns are structured threefold by EBL and are later transferred to the silicon. We demonstrate a 9-channel multiplexer featuring a 2 dB on-chip loss and an adjacent channel crosstalk better than -22 dB. Additionally a 45-channel Echelle multiplexer is presented with 5 dB on chip loss and a channel crosstalk better than -12 dB. The devices cover an on-chip area of only 0.08 mm² and 0.5 mm² with a wavelength spacing of 10.5 nm and 2.0 nm, respectively.

It is well known that the performance of semiconductor lasers is very sensitive to external optical feedback. This feedback can lead to changes in lasing characteristics and a variety of dynamical effects including chaos and coherence collapse. One way to avoid this external feedback is by using optical isolation, but these isolators and their packaging will increase the cost of the total system.

Semiconductor ring lasers nowadays are promising sources in photonic integrated circuits because they do not require cleaved facets or mirrors to form a laser cavity. Recently, some of us proposed to combine semiconductor ring lasers with on chip filtered optical feedback to achieve tunable lasers. The feedback is realized by employing two arrayed waveguide gratings to split/recombine light into different wavelength channels. Semiconductor optical amplifier gates are used to control the feedback strength. In this work, we investigate how such lasers with filtered feedback are influenced by an external conventional optical feedback.

The experimental results show intensity fluctuations in the time traces in both the clockwise and counterclockwise directions due to the conventional feedback. We quantify the strength of the conventional feedback induced dynamics be extracting the standard deviation of the intensity fluctuations in the time traces. By using filtered feedback, we can shift the onset of the conventional feedback induced dynamics to larger values of the feedback rate [ Khoder et al, IEEE Photon. Technol. Lett. DOI: 10.1109/LPT.2016.2522184]. The on-chip filtered optical feedback thus makes the semiconductor ring laser less senstive to the effect of (long) conventional optical feedback. We think these conclusions can be extended to other types of lasers.

Thermo-optical switches are widely used in integrated optics and various types of integrated optical structures have been reported in literature. These structures include, but are not limited to Mach-Zehnder-Interferometer (MZI) switches and digital optical switches. The thermo-optical effect depends on the refractive index, the polarizability and the density of a material. The polarizability effect can often be neglected and the change of refractive index is dominated by a density change due to the thermal expansion of the material.

We report herein a new method to measure the thermo-optical effect of waveguides directly, using integrated MZIs fabricated in polymer waveguide technology. Common methods rely on macroscopic samples, but the properties can differ significantly for micro-structured waveguides. Using a floodlight halogen rod lamp and metal-shields, we realized a radiation heater with a trapezoidal-shaped heating pattern. While the heating occurred from the bottom side, a thermocouple was placed on top of the sample. By dynamically measuring the temperature and the corresponding output-power of the MZI, the temperature difference between constructive and destructive interference can be determined. Multiple measurements of different sample MZIs exhibit an average thermo-optical coefficient (TOC) of 1.6 ∗ 10^-4 1/K .

In this paper, we use the nonlinear property of SiC multilayer devices under UV irradiation to design an optical processor for indoor positioning. The transducer combines the simultaneous demultiplexing operation with the photodetection and self-amplification. The proposed coding is based on SiC technology. Based on that, we present a way to achieve indoor localization using the parity bits and a navigation syndrome. A representation with a 4 bit original string colour message and the transmitted 7 bit string, the encoding and decoding accurate positional information processes and the design of SiC navigation syndrome generators are discussed and tested. A visible multilateration method estimates the position of the device by using the decoded information received from several, non-collinear transmitters. The location and motion information is found by mapping position and estimates the location areas. Since the indoor position of the LED light source is known from building floor plans and lighting plans, the corresponding indoor position and travel direction of a mobile device can be determined.

The device under study is an a-SiC:H/a-Si:H pinpin photodiodes produced by PECVD (Plasma Enhanced Chemical Vapour Deposition) and has a structure that consists of a p-i'(a-SiC:H)-n/p-i(a-Si:H)-n heterostructure with low conductivity doped layers. This device structure has been demonstrated useful in optical communications that use the WDM technique to encode multiple signals in the visible light range. We present in this work experimental results about C-V measurements of the device under complex conditions of illumination. Also it is presented an analysis based on the transient response of the device when illuminated by a pulsed light, with and without optical bias superposition. Rising and decaying times of the collected photocurrent will be outlined under the different conditions. A simulation study outlines the role played by each pin substructure on the response speed and gives some hint on the possible optimization of this device.

Time delay integration sensors are of increasing interest in CMOS processes owing to their low cost, power and ability to integrate with other circuit readout blocks. This paper presents an analysis of the noise contributors in current day CMOS Time-Delay-Integration image sensors with various readout architectures. An analysis of charge versus voltage domain readout modes is presented, followed by a noise classification of the existing Analog Accumulator Readout (AAR) and Digital Accumulator Readout (DAR) schemes for TDI imaging. The analysis and classification of existing readout schemes include, pipelined charge transfer, buffered direct injection, voltage as well as current-mode analog accumulators and all-digital accumulator techniques. Time-Delay-Integration imaging modes in CMOS processes typically use an N-number of readout steps, equivalent to the number of TDI pixel stages. In CMOS TDI sensors, where voltage domain readout is used, the requirements over speed and noise of the ADC readout chain are increased due to accumulation of the dominant voltage readout and ADC noise with every stage N. Until this day, the latter is the primary reason for a leap-back of CMOS TDI sensors as compared to their CCD counterparts. Moreover, most commercial CMOS TDI implementations are still based on a charge-domain readout, mimicking a CCD-like operation mode. Thus, having a good understanding of each noise contributor in the signal chain, as well as its magnitude in different readout architectures, is vital for the design of future generation low-noise CMOS TDI image sensors based on a voltage domain readout. This paper gives a quantitative classification of all major noise sources for all popular implementations in the literature.

A nanoscale layer of amorphized silicon is obtained by implantations with silicon ions through a P-doped FZ-silicon wafer material few nanometers below the wafer surface. After a controlled annealing, the amorphized silicon material is sandwiched between two layers of recrystallized silicon. Defects remain at the interface c-Si/a-Si/c-Si. Photoluminescence at very low temperature is experimented to determine the energy levels generated by this design. TEM pictures show that some nanocrystalline elements are located close to the interface surrounded by a-Si. However, the photoluminescence spectra do not present any signal of luminescence from them. This could be due to random sizes of nanocrystals. Then, a scan from energies below the silicon bandgap has been realized at 8 K. The spectrum is composed of multiple narrow peaks close to the conduction band and a broadband from 0.78 eV to 1.05 eV. In order to determine the origin of these signals, spectra of three distinct peaks were collected at different temperatures from 8 K to 120 K. The broadband collapses more rapidly by increasing the temperature than the narrow lines and theirs maxima of intensity differ.

In this paper, a novel design of hybrid silicon plasmonic transverse electric (TE) pass polarizer based on silicon-oninsulator (SOI) platform is reported and analyzed. The numerical results are obtained by using full vectorial finite element method. The suggested design depends on gold rods that are injected into the substrate in order to tolerate the function of the device and hence the required polarizing state can be obtained. The proposed SOI TE polarizer can achieve -0.19 dB insertion losses with compact device length of 18 μm. Further, the introduced device is easy for fabrication and is compatible with the standard CMOS fabrication process.

We present an algorithm for synthesizing generalized cascaded Mach-Zehnder interferometer (GCMZI) lattice filters. The algorithm is then used to design a tenth-order filter using a GCMZI filter with eight stages.