This PDF file contains the front matter associated with SPIE Proceedings Volume 10686, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and Conference Committee listing.

In this paper we describe our work on silicon photonics based advanced transceivers. This includes the realization of 100 GBaud OOK and duobinary transceivers and 56 GBaud PAM-4 transmitters for datacenter applications, 40 GBaud 16-QAM coherent receivers for metro and 20GBaud PAM-4 transceivers for access networks. Analog radio-over-fiber transceivers for next-generation 5G wireless access will be described as well, covering carrier frequencies ranging from 3.5 GHz to 28 GHz. These transceivers rely on both advanced photonic and electronic integrated circuits. In order to realize a fully integrated transceiver, the III-V laser source needs to be integrated as well. For this purpose we are developing a novel integration strategy: micro-transfer printing. In this talk I will elaborate on the underlying concepts as well as discuss first realizations of transfer printed III-V-on-silicon opto-electronic components.

Scaling in the electronics industry according to Moor's law has recently been slowing down. Further size reductions will ultimately lead to an atomic scale electronics. However, how viable is such an atomic scale technology? Photonics too has undergone quite some size reduction. The ultimate limit is set by the wavelength and the refractive index. The question then is if there is room for scaling beyond the Abbe diffraction limit? In this talk we will review recent advances in the field of an atomic scale electronics and photonics.

The dramatic optical property change of optical phase change materials (O-PCMs) between their amorphous and crystalline states potentially allows the realization of reconfigurable photonics devices with low power consumption, such as optical switches and routers, reconfigurable meta-optics, displays, and photonic memories. However, conventional O-PCMs, such as VO2 and Ge2Sb2Te5, are inherently plagued by their excessive optical losses even in dielectric states, limiting their optical performance and hence application space. In this talk, we present the development of a new group of O-PCMs and their implementations in novel photonic devices. Ge-Sb-Se-Te (GSST), obtained by partially substituting Te with Se in traditional GST alloys, feature unprecedented broadband optical transparency covering the telecommunication bands to LWIR. Capitalizing on the dramatically-enhanced optical performance, novel non-volatile, reconfigurable on-chip photonics devices and architectures are demonstrated. GSST-integrated Si photonics based on the material innovation and novel “non-perturbative” designs exhibit significantly improved switching performance over state-of-the-art GST-based approaches. The technology is further scalable to realize non-blocking matrix switches with arbitrary network complexity, paving the path towards high performance reconfigurable photonics chips.

With the fast growing demand of data, current chip-scale communication systems based on electrical links suffer rate limitations and high power consumptions to address these new requirements. In this context, Silicon Photonics has proven to be a viable alternative by replacing electronic links with optical ones while taking advantage of the well-established CMOS foundries techniques to reduce fabrication costs. However, silicon, in spite of being an excellent material to guide light, its centrosymmetry prevents second order nonlinear effects to exist, such as Pockels effect an electro-optic effect extensively used in high speed and low power consumption data transmission. Nevertheless, straining silicon by means of stressed thin films allows breaking the crystal symmetry and eventually enhancing Pockels effect. However the semiconductor nature of silicon makes the analysis of Pockels effect a challenging task because free carriers have a direct impact, through plasma dispersion effect, on its efficiency, which in turn complicates the estimation of the second order susceptibility necessary for further optimizations. However, this analysis is more relaxed working in high-speed regime because of the frequency limitation of free carriers-based modulation. In this work, we report experimental results on the modulation characteristics based on Mach-Zehnder interferometers strained by silicon nitride. We demonstrated high speed Pockels-based optical modulation up to 25 GHz in the C-band.

Digital-to-analog converters (DAC) are indispensable functional units in optical signal transmission and processing. The photonic DAC that converts electrical digital signals to an optical analog one will offer advantages in lowering system complexity, power, and cost. Especially with the required bandwidth increasing, it could mitigate the problems faced by its electrical counterparts in dealing with higher sampling rate. Achieving such a photonic DAC in silicon photonics is promising due to the integration capability of both electronics and photonics and large scale DAC-based photonic circuits can be further realized for on-chip optical signal processing. In this work, we demonstrate 2-bit D/A conversion for the simple proof of concept utilizing only one single silicon Mach-Zehnder modulator (MZM), which is much simpler than previously reported segmented MZM and microring resonator based DACs. One-single MZM capable of 2-bit DAC merits future higher bit resolution design and meanwhile guarantees wide spectral bandwidth. One arm of MZM is used for the MSB bit input, while the other for the LSB, both of them being accomplished by only one phase shifter. For each bit input, we utilize amplitude modulation, instead of phase modulation, by applying the carrier injection induced absorption in the phase shifters. For principle, by setting different bias points for two phase shifters, we can produce the condition at which the amplitude weighting ratio of LSB to MSB is 1/2 in order to obtain the linear amplitude DAC output. In other words, the output optical field has the analog linear amplitude levels (0,1,2,3) which corresponds to the power levels of (0,1,4,9) at the full extinction condition. For fabrication, this device was fabricated on a 220-nm SOI wafer with a 3-m buried-oxide layer at the AIST SCR 300-mm CMOS foundry. The 430-nm-wide fully etched channel waveguide was used for the components except for the pn phase shifter which adopted the shallow-etched rib waveguide structure with a slab thickness of about 110 nm and a width of about 600 nm. The doping density in the weak p/n regions was about 1.61018 cm-3. This MZM was arm-balanced with 2-mm-long phase shifters, adopting GSGSG configuration. Two 50- terminators were also integrated on-chip at the ends of two signal electrodes. For measurement, a two-channel pulse pattern generator produced bit sequences at various frequencies for both MSB and LSB which was applied to the signal electrodes through bias-tees and high-speed probes. The 1.55-m cw light at TE polarization was coupled into the chip via a tapered fiber and the optical output passed an EDFA and a bandpass filter and then was sent to a high-speed oscilloscope for examining DAC analog output. Using this device, we successfully achieved correct D/A conversions with the sampling rates up to 3 GS/s with <1 V peak-to-peak voltages. Note that this speed can be further enhanced to <10 GS/s by constructing the pn phase shifter into a MZM structure or replacing it with a SiGe electro-absorption modulator. In summary, this work verified the feasibility to realize high-sampling-rate 2-bit D/A conversion utilizing a single silicon MZM modulator.

In this paper, an ultra-thin buffer technology for the epitaxial growth of Si_xGe_1-x-ySn_y structures on Si or Si-on-Insulator substrates by using molecular beam epitaxy is presented. This technology builds the basis for integrated photonic devices as detectors, modulators and light sources. The paper discusses different device families with different material compositions, which all use a relaxed Ge virtual substrate with high quality. These are pseudomorphic Ge/Ge_1-ySn_y structures, Si_xGe_1-x-ySn_y structures lattice matched to Ge and (partially) relaxed Ge_1-ySn_y virtual substrates. The photonic devices consist of heterojunction diodes with vertical pin doping structures. As an example, Ge/Ge_1-ySn_y multi quantum well photodetectors which active regions made from Nx(Ge_0.93Sn_0.07/Ge) multi-quantum well structures are presented. Optical measurements at high frequencies are successfully performed on these photodetectors. A 3-dB bandwidth above 40 GHz is measured at the optical telecommunication wavelength of 1550 nm.

Mode-locked lasers provide extremely low jitter optical pulse trains for a number of applications ranging from sampling of RF-signals and optical frequency combs to microwave and optical signal synthesis. Integrated versions have the advantage of high reliability, low cost and compact. Here, we describe a fully integrated mode-locked laser architecture on a CMOS platform that utilizes rare-earth doped gain media, double-chirped waveguide gratings for dispersion compensation and nonlinear Michelson Interferometers for generating an artificial saturable absorber to implement additive pulse mode locking on chip. First results of devices at 1.9 μm using thulium doped aluminum-oxide glass and operating in the Q-switched mode locking regime are presented.

NRZ data with PRBS length 2³¹ − 1, was propagated over 25 km of standard single mode fibre at a rate of 10 Gbit/s and 12.5 Gbit/s in a repeater-less transmission system. Results show that, for a received optical power of −13.2 dBm, the BER were 2.7 × 10⁻⁶ and 3.3 × 10⁻⁴ respectively, with sufficient margin below the FEC limit of 1 × 10⁻³ . The packaged transmitter comprised an integrated DFB laser, electro-absorption modulator and semiconductor optical amplifier hybrid integrated on silicon wafer.

A high-efficiency and low-loss Mach-Zehnder modulator on a Si platform is a key component for meeting the demand for high-capacity, low-cost and low-power optical transceivers in future optical fiber links. We report a III-V/Si MOS capacitor Mach-Zehnder modulator with an ultrahigh-efficiency phase shifter, which consists of n-type InGaAsP and ptype Si. The main advantage of this structure is a large electron-induced refractive index change and low free-carrier absorption loss of the n-type InGaAsP. The heterogeneously integrated InGaAsP/Si MOS capacitor structure is fabricated by using the oxygen plasma assisted bonding method. The fabricated device shows V_πL of 0.09 Vcm, a value over three-times smaller than that of the conventional Si MOS capacitor Mach-Zehnder modulator, without an increase in the insertion loss. This clearly indicates that the proposed III-V/Si MOS capacitor Mach-Zehnder modulator overcomes the performance limit of the Si Mach-Zehnder modulator.

A room temperature four-channel hybrid laser array was realized by using selective area metal bonding method, which is able to evanescently couple the light from InGaAsP multi-quantum well laser diodes into the slotted silicon waveguide. The output wavelength of this hybrid laser can be accurately tuned to 1538.6 nm, 1540.5 nm 1544.9 nm and 1550 nm through controlling the waveguide width of each channel. A typical threshold current of one channel was of 20 mA and the corresponding side-mode suppression ratio was of 20 dB. Moreover, one advantage of such slotted structures is the ease in fabrication, thus making this type of devices as a promising candidate for wavelength division multiplex in future optical communication system.

With the recent progress in integrated silicon photonics technology and the recent development of efficient quantum cascade lasers (QCL), there is now a very good opportunity to investigate new gas sensors offering very high sensitivity, high selectivity (multi-gas sensing, atmosphere analysis) and low cost thanks to integration on planar Si substrates. Gas sensing generally requires a tunable source continuously covering the whole operational range of the QCL stack. This paper presents the design, fabrication and characterization of Array Waveguide Grating (AWG) devices aiming at the simultaneous detection of several gas using arrays of QCL sources. We have developed a new platform based on Ge cores surrounded by thick Si₈₀Ge₂₀ layers. The index difference between the core and the cladding is around 0.5 on the 3-13 μm spectral range. The core has a typical cross section of 2.5 × 2.5 μm and is surrounded by 6 μm thick SiGe cladding layers. As test vehicle device, we designed a 35 inputs multiplexer working in the 9.5 μm operation range (1050-1250 cm^-1 ). The design was carried out to match the inputs of a Distributed Feed-Back-QCL array. Preliminary measurements on the waveguide showed losses in the 3.5 dB/cm range. AWG devices were fabricated and tested. They showed results in good agreement with the modeling. An almost flat transmission over a full 200 cm^-1 operational range was obtained, with a peak-to-valley modulation of -5dB.

Since many important molecules have strong “fingerprints” in the mid-infrared (mid-IR, between 3μm and 15μm), this wavelength spectrum is currently gaining significant attention for applications ranging from pollution detection, quality control in the food industry, early cancer diagnosis, security and safety [1, 2]. Molecular sensing devices in the mid-IR are currently being developed and are in the process of commercialization. An appealing approach is to create molecular sensing devices in the mid-IR based on low cost integrated mid-IR chips. A key building block is a high brightness integrated broadband light source that would allow the detection of several molecules characterized by distinct absorption lines in parallel. Such an integrated broadband source, referred to as a supercontinuum source, has been already demonstrated in the mid-IR on a chalcogenide chip [3]. However, demonstrating mid-IR supercontinuum on group IV materials, in order to exploit the advantages of reliable CMOS fabrication technology, remains a challenge. So far, numerous CMOS-compatible supercontinuum sources have been demonstrated in silicon nitride-on-insulator [4, 5], silicon-on-insulator [6, 7], silicon germanium-on-insulator [8] and silicon-on-sapphire platforms [9]. However, these sources are limited up to 3.5µm and 6µm due to the absorption in the silica and sapphire substrate, respectively. More recently, the silicon germanium-on-silicon platform [10, 11], emerged as an attractive platform for mid-IR photonics, with transparency potentially extending up to 15μm depending on the Ge content [12]. Here we report experimentally the first octave spanning supercontinuum generation from a SiGe waveguide in the actual mid-IR with 5mW on-chip power exceeding that produced so far in any other Si-based platform (0.15mW in SiGe/SiO2 [8] and ~1mW in silicon-on-sapphire [9]). Our 4.25µm x 2.70µm cross-section air-clad SiGe-on-Si waveguide has been designed and manufactured to achieve single mode operation at 4µm, low anomalous dispersion and strong fundamental TE mode confinement in the core nonlinear material (~96% at 4μm). Losses as low as 0.4dB/cm were measured between 3.8 and 5µm. The achieved supercontinuum covered more than an octave between 2.95 and 6.0µm was generated by pumping a 7cm long waveguide with ~ 200fs pulses at 4.15µm and 63MHz repetition rate, as delivered by a Miropa-fs optical parameter amplifier. These results were supported by simulations and the related generation of a high brightness supercontinuum establishes silicon germanium-on-silicon as a promising platform for integrated nonlinear photonics in the mid-IR, with the potential to extend the operating range to beyond 8μm. References [1] R. Soref, Nature Photonics, 4, 495-497, (2010). [2] B. Mizaikoff, 42, 8683-99, (2013). [3] Y. Yu, et al., Opt Lett, 41, 958-61, (2016). [4] J. M. Chavez Boggio, et al., JOSA B, 31, 2846-2857, (2014). [5] A. R. Johnson, et al., Opt Lett, 40, 5117-20, (2015). [6] B. Kuyken, et al., Opt Express, 19, 20172-81, (2011). [7] R. K. W. Lau, et al., Optics letters, 39, 4518-4521, (2014). [8] M. A. Ettabib, et al., Optics letters, 40, 4118-4121, (2015). [9] N. Singh, et al., Optica, 2, 797-802, (2015). [10] L. Carletti, et al., Opt Express, 23, 8261-71, (2015). [11] L. Carletti, et al., Opt Express, 23, 32202-14, (2015). [12] J. M. Ramirez, et al., Opt Lett, 42, 105-108, (2017).

The mid infrared (MIR) region, which ranges from 2 μm to 20 μm, has attracted a lot of interest, particularly for novel applications in medical diagnosis, astronomy, chemical and biological sensing or security, to name a few. Most recently, Germanium-rich Silicon Germanium (Ge-rich SiGe) has emerged as a promising waveguide platform to realize complex mid-IR photonic integrated circuits. The Ge-rich SiGe graded buffer benefits from a wide transparency window, strong 3rd order nonlinearity, and the compatibility with mature large-scale fabrication processes, which in turn, paves the way for the development of mid-IR photonic devices that afford improved on-chip functionalities, altogether with compact footprints and cost-effective production. Albeit, low-loss waveguides and wideband Mach-Zehnder interferometers (MZIs) have been recently successfully demonstrated at mid-IR wavelengths, the coupling of light between external access ports, typically optical fibers, and integrated circuits remains challenging. Surface grating couplers provide technologically attractive scenario for light coupling, since they allow flexible placement on the chip, thereby enabling automatic testing of fabricated devices on a wafer-scale, preferred for large-volume developments. In this work, we report two designs for surface grating couplers implemented on the Ge-rich SiGe graded buffer. The grating couplers are designed for transverse electric (TE) and transverse magnetic (TM) polarizations, respectively, both operating at 7.5 μm wavelength. In particular, the TE-designed grating coupler with an inverse taper excitation arrangement yields a coupling efficiency of 6.3% (-12 dB), a 1-dB bandwidth of 300 nm, and reduced back-reflection less than 1%. Furthermore, the TM-designed grating coupler with a conventional taper injection stage predicts an improved coupling performance up to 11% (-9.6 dB), with a 1-dB bandwidth of 310 nm, and only 1% back-reflection. These results open up the way for the realization of complex and multifunctional photonics integrated circuits on Ge-rich SiGe platform with operation at midIR wavelengths.

The third-order nonlinear parameter of Ge-rich SiGe waveguides are experimentally retrieved using a bi-directional top hat D-scan at λ = 1.58 μm. The obtained values are then used to fit the theoretical equation, providing promising values in the mid-IR, where the nonlinear effects are no longer limited by two-photon absorption. New Ge-rich SiGe waveguide designs are provided to exploit the nonlinear properties in the mid-IR, showing a flat anomalous dispersion over one octave spanning from λ = 3 µm to λ = 8 μm and a γ parameter that decreases from γ = 10 W^-1m^-1 .

Bandwidth demands in optical communication systems are growing steadily and making Wavelength Division Multiplexing (WDM) reach its limit. New multiplexing techniques are required in order to fulfill future bandwidth demands in next generation optical communications. Mode Division Multiplexing (MDM) has been recently proposed as good solution to increase aggregate bandwidth by multiplexing on the spatial domain. In this work we discuss the propositions of ultra-compact mode converters based on periodically perturbed waveguides. A corrugation (perturbation) is periodically inserted on one side of the waveguide. Each time the fundamental mode propagates through a perturbation a part of the incident light is transferred to the second mode. Around 5 periods are only needed to achieve complete power transfer, enabling for ultra-compact devices. Insertion loss below 0.5 dB and extinction ratio higher than 13 dB in the C-Band have been evaluated in a device with a total length of only 12 μm.

In silicon photonics, optical coupling between optical fibers and silicon photonics chips still remains a big issue to be improved. Recently we have developed a new surface coupler consisting of a vertically-bent silicon waveguide using ion implantation as a bending method. This bending method enabled bending of silicon waveguide terminals to the vertical surface direction with curvature radii as small as several μm and bending of propagation direction of light to the surface direction with broadband wavelength property. In our initial studies, to couple the propagated light with optical fibers efficiently, we taper-narrowed the tips of the waveguide terminals from 440 nm to ~200 nm to expand the mode field and obtained a coupling loss of ~2 dB in TE-polarization mode for a lensed fiber with a beam diameter of 2 μm. We are now developing couplers with high coupling efficiencies for 5 μm mode field diameter fibers. Such mode field diameters are available by HNA fibers and position alignment accuracy can be mitigated to a micron level. In addition, HNA fibers can be fusion-spliced to standard single mode fibers. Using 3D-FDTD simulation we have demonstrated that <1dB coupling is possible if the tip of the bent waveguide is taper-narrowed to 50 nm and covered with a lens-shaped SiO₂ dome. In practice, such a structure could be fabricated successfully and 5 dB coupling with a 5 μm mode field diameter lensed fiber has been demonstrated in our initial experiment

Optical input/output interfaces between silicon-on-insulator (SOI) waveguides and optical fibers, allowing robust, costeffective and low-loss coupling of light, are fundamental functional elements in the library of silicon photonic devices. Surface grating couplers are particularly desirable as they allow wafer-scale device testing, yield improved alignment tolerances, and are compatible with state-of-the-art integration and packaging technologies. While several factors jointly contribute to the coupler performance, the grating directionality is a critical parameter for high-efficiency fiber-chip coupling. To address this issue, conventional coupler designs typically call upon comparatively complex architectures to improve light coupling efficiency. Increasing the intrinsic directionality of the grating by exploiting the blazing effects is another promising solution. In this paper, we report on our recent advances in development of low-loss grating couplers that afford excellent directionality, close to the theoretical limit of 100%. In particular, we demonstrate, by theory and experiments, several implementations of blazed grating couplers with layout features that are compatible with deepultraviolet (deep-UV) optical lithography. Devices can be advantageously implemented on various photonic platforms, including industry-specific and the offerings of publicly accessible foundries. The first experimental realizations of uniform deep-UV-compatible couplers yield losses of -2.7 dB at 1.55-µm and a 3-dB bandwidth of 62 nm. A subwavelength-index-engineered impedance matching transition is used to reduce back-reflections down to -20 dB.

In this work, we report on the modeling and the experimental characterization of a 6×400 GHz silicon Arrayed Waveguide Grating (AWG). The design of the device is based on the reduction of the background noise. The good characteristics of the AWG demonstrate that unwanted reflections have a detrimental role on its performance. We demonstrate a smoothing of the output channel shape of the AWG, as well as a reduction of the crosstalk level from −20.6(1) dB to −24.4(1) dB.

We report on the co-integration of an additional passive layer within a Silicon Photonic chip for advanced passive devices. Being a CMOS compatible material, Silicon Nitride (SiN) appears as an attractive candidate. With a moderate refractive index contrast compared to SOI, SiN based devices would be intrinsically much more tolerant to fabrication errors while keeping a reasonable footprint. In addition, it's seven times lower thermo-optical coefficient, relatively to Silicon, could lead to thermal-tuning free components. The co-integration of SiN on SOI has been explored in ST 300mm R and D photonic platform DAPHNE and is presented in this paper. Surface roughness of the SiN films have been characterized through Atomic Force Microscopy (AFM) showing an RMS roughness below 2nm. The film thickness uniformity have been evaluated by ellipsometry revealing a three-sigma of 21nm. Statistical measurements have been performed on basic key building blocks such as SiN strip waveguide showing propagation loss below 0.7dB/cm and 40μm radius bends with losses below 0.02dB/90°. A compact Si-SiN transition taper was developed and statistically measured showing insertion losses below 0.17dB/transition on the whole O-band wavelength range. Moreover, advanced WDM devices such as wavelength-stabilized directional couplers (WSDC) have been developed.

Silicon photonics is a very promising solution to achieve high-speed and energy-efficient optical interconnects at reduced cost of production. For several years now, Silicon photonic platforms have offered high performance passive and active compact components, with first 100 Gb/s silicon photonics commercial transceivers on the market. However, interferometric devices still remain sensitive to temperature changes. Silicon nitride appears as an appealing material for CMOS compatible, energy-efficient and cost-effective photonics. Its low optical index contrast with the SiO2 cladding provides low loss waveguide and a better tolerance to fabrication imperfections while its optical index is much less sensitive to temperature variations. In particular, the monolithic integration of multiple Si and SiN layers on the same platform provides a promising solution for Coarse Wavelength Division Multiplexing (CWDM) transceiver applications for which thermal stability is essential. We report here on a 200mm-CMOS-compatible platform where SiN is co-integrated with Si to benefit from the advantages of both materials. We present the fabrication flow of this enhanced platform and we show the wafer scale characterization of its passive Si, SiN and hybrid components, assessing their performances in term of 4-wavelength CWDM application in the O-band. Building on the CEA-LETI silicon photonics fabrication line, we present the monolithic integration of a low stress PECVD SiN deposited at 300°C, which make it compatible with the standard SOI platform with doped active devices. The success of the integration is confirmed by the preserved performances of Si components (waveguide, bend and fiber grating coupler losses) as well as propagation losses as low as 0.8dB/cm for a single mode SiN waveguide. Furthermore, characterizing micro-ring resonators, the resonance shift with temperature allows us to extract a 10-fold reduction of the SiN thermo-optic coefficient compared to Si. An essential building block of the Si-SiN platform is the Si-SiN interlayer transition. We present a complete study of such transitions with both simulation and experimental data of various geometries, and we demonstrate state-of-the-art insertion losses of 0.09 ± 0.01 dB over the O-band for the TE mode at the wafer scale. We show as well the realization of hybrid Si-SiN grating couplers, compatible with a standard packaging fiber angle of 8°. In such couplers, the main SiN grating is combined with a Si grating placed underneath and a longitudinal shift between gratings allows the tuning of the coupler’s directionality. Consequently, we achieve a dramatic improvement in terms of bandwidth with respect to the standard all Si fiber couplers, going from a 23nm -1dB bandwidth for Si couplers to more than 50nm for hybrid couplers. These wideband hybrid couplers are key for CWDM where a flat transmission spectrum is needed over the O-band. Finally, to complete the CWDM components review, we present SiN Echelles grating (de)multiplexer for 4-channels CWDM in the O-band showing quasi-absolute thermal insensitiveness and low insertion losses.

We present the design, fabrication and optical characterization of an ultra-long silicon ring resonator for opto-microwave applications. The investigated optical ring resonator has been fabricated under silicon-based processes using the silicon on insulator photonic platform. An add-drop ring resonator with integrated grating couplers was designed in the shape of a spiral in order to reduce the footprint of the device. After fabrication, the samples have been tested in the optical domain. At the wavelength of 1.55 μm, the optical transmission spectrum has shown a free spectral range (FSR) of 16 GHz which was in good agreement with the design target. The extinction ratio reached up to 19.5 dB for the drop port. Concerning the quality factor of the resonator, we got values up to 1.74 × 10⁵ . After these tests, a device has been selected to be introduced in the loop of an optoelectronic oscillator (OEO) system. Under these conditions it turned to be possible to generate oscillations with a peak at frequency of 16 GHz, showing a good agreement with the previously measured FSR.

The silicon-on-insulator (SOI) platform allows for a miniaturization of optical elements at the micron size. It is now a mature technology, with high quality material and well-known fabrication processes. Another advantage stems in its compatibility with the CMOS facilities. The SOI platform is used already for numerus applications such as datacom, sensing or manipulation of quantum objects. Bragg filters are often used for on-chip the rejection of pump lasers. They can also be used for sensing purposes. By periodically modulating a standard waveguide width, it is possible to realize a 1-D photonic crystal with a forbidden wavelength band. In principle, the bandwidth and central wavelength of this bandgap can be tailored just by a proper design of the introduced corrugation. However, the very large index contrast of Si-wires makes the realization of narrowband rejection filters a technological challenge, requiring multiple etching steps or corrugation widths of a few tens of nanometers. Sub-wavelength nanostructuration of Si waveguides has shown to allow narrowband operation with a single-etch process, but reported rejection levels remained limited. Here we present an innovative differential corrugation approach that allows the realization of narrowband rejection optical filters with relaxed fabrication constraints. By sub-wavelength engineering of the waveguide geometry we experimentally demonstrated simultaneous high rejection of 50dB and narrowband operations less than 3nm. We propose new subwavelength designs based on subwavelength structures having two subperiods in each Bragg period. We also investigate the equivalent asymmetric structure to reduce the index contrast in the periods and further reduce the bandwidth without adding any new fabrication constraints. We have fabricated the sub-wavelength engineered filters in standard SOI wafer with a 220 nm thick Si guiding layer and bottom oxide layer of 2 µm. We have used electron beam lithography with 5 nm step-size and have patterned the structurse by dry etching with an inductively coupled plasma etcher. Finally, we have covered the devices with PMMA to provide symmetric cladding. We report results showing that subwavelength Bragg filter geometries allow a drastic reduction of the operating optical bandwidth to the 0.6nm-2.5nm range if compared with regular Bragg filters (20 nm) while retaining still a strong rejection level of around 40dB. Similarly, each asymmetric version was observed to be bandwidth narrower. To sum up, this paper is an investigation of advanced SOI waveguide Bragg mirrors. We report that the use of subwavelength corrugations and a judicious of waveguide Bragg asymmetry allow to push the extinction ratio/operating bandwidth beyond its traditional limit.

Conference Presentation for "Harnessing photonic integrated circuits"

We demonstrate the automatic thermal alignment of photonic components within an integrated optical switch. The WDM optical switch involves switching elements, wavelength de-multiplexers, interleavers and monitors each one needing independent control. Our system manages rerouting of channels coming from four different directions, each carrying 12, 200GHz spaced, wavelengths into eight add/drop ports. The integrated device includes 12 interleavers, which can act either as optical de-interleavers to split the optical signal into odd and even channels or as optical interleavers that recombine the odd and even channels coming from the switching matrix. Integrated Ge photodiodes are placed in key positions within the photonic integrated circuit (PIC) are serve for monitoring. An electronic integrated circuit (EIC) drives the photonic elements by means of dedicated heating circuits (824 on-board heater control cells, 768 for the switching elements and 56 for the interleavers and the mux/de-mux) and reads out the Ge diodes photocurrent through TIAs. We applied a stochastic optimization algorithm to align the spectral response of the interleavers to the ITU grid. We exploit the thermo-optic effect to shift the interleavers pass-band in a desired spectral position. The interleavers are provided with dedicated metallic heaters that can be operated in order to tune the interleaver response, which is typically misaligned due to fabrication inaccuracies. The experimental setup is made of a tunable laser coupled with one input port of optical switch. The optimization algorithm is implemented via a software to drive the EIC till finding the best heating configuration (on the two branches of the interleaver) on the basis of the monitor diode-feedback. This way, the even and odd wavelengths input in the interleaver are directed toward the wanted lines within the switching matrix. Our method has been used for aligning the micro-ring based switching elements in the PIC as well. In that case, the integrated Ge photodiodes have been used to align the photonic components in the PIC in order to enable different pathways for the routing or the broadcasting operation of the optical switch. With no bias applied to the heaters of the switching elements, the optical signal is expected to be maximum at the through port. When the micro-ring heaters are biased, the feedback controller finds the best set of heating values that minimize the optical power at the through port of the switching node. This way, the optical signal is coupled in the drop port and the node is enabled for switching. The algorithm, implemented in LabVIEW, converges over multiple instances and it is robust against stagnation. This work aims at enabling the automatic reconfiguration/restoration of the whole WDW optical switch.

Photonic integrated circuits (PICs) have been demonstrated as a promising technology to implement flexible and hitless reconfigurable devices for telecom, datacom and optical interconnect applications. However, the complexity scaling of such devices is raising novel needs related to their control systems, and the automatic calibration, reconfiguration and operation of these complex architectures both during manufacturing and in service is still an open issue. In this contribution, we report our recent achievements on the automatic hitless tuning of a telecommunication-graded filter operating in the L band, fabricated on a commercial foundry Silicon Photonics (SiP) run. A novel channel labeling strategy is used to automatically identify the desired channel within a Dense Wavelength Division Multiplexing (DWDM) through a FPGA-embedded closed-loop control algorithm. The photonic architecture consists of a hitless third order Micro Ring Resonator (MRR) filter with 8 nm Free Spectral Range (FSR), integrating transparent detectors (ContactLess Integrated Photonic Probes - CLIPP) as power monitors and thermooptic actuators. Transparent detectors enable to control the input/output port of the filter without introducing any loss to the WDM channel comb. Hitless operation is achieved through a pair of switchable Mach-Zehnder interferometers used as input/output couplers of the MRR filter. The fabricated device has a 3 dB bandwidth of 40.7 GHz and provides a through-port in-band isolation of 23 dB and a drop port isolation of 25 dB at 50 GHz spacing from the dropped channel. Hitless reconfiguration is achieved with more than 30 dB isolation during channel selection. The automatic tuning and locking technique is based on the use of a pilot tone generated locally at the node site and applied as a low frequency (few kHz), small modulation index (< 8%), amplitude modulation on the channel to be added to the network. The effectiveness and robustness of the automatic controller for tuning and stabilization of the filter is demonstrated by showing that no significant bit-error rate (BER) degradation is observed in an adjacent channel while the filter is being reconfigured. In addition, the convergence of the algorithm is shown to require only few tens of iterations, each one requiring a few milliseconds. The FPGA-embedded control technique together with the compactness provided by SiP meets the integration requirements for high capacity networks and pluggable modules. In addition, the filter unit can be cascaded with other units to realize a multichannel reconfigurable add-drop architecture operating on several wavelengths at the same time with complete independency.

Dielectric, ohmic-loss-free, finite photonic crystals (PCs) may sustain the propagation of highly confined surface waves that propagate bound to the interface of the bulk structure and the free space. For many years the dielectric photonic crystals surface states have been treated as a subsidiary effect related to the inevitable finite size of the PCs in realistic implementations. However, in the recent years it has been realized that the features of the dielectric surface states and their impact to the wave exit from the photonic crystal structure render the relevant structures suitable components for a variety of applications, including optical spectroscopy, sensing, intercomponent coupling, etc. In this work we present the design of a silicon-based PC component that couples the modes that propagate through the bulk structure into outgoing, free space propagating beams with high directionality. In addition to previous works involving silicon-based PCs for beam collimation in the near infrared and optical regime, we demonstrate here, that the (frequency depended) emission angle of the generated beams can be controlled by properly engineering the PC termination. Thus the component may serve as a beam steering structure or a de-multiplexer in the optical telecommunications wavelength band (~1.5 μm). Our design takes into account the state-of-the-art nano-fabrication technology and all the constraints arising from the treatment of silicon-based periodic media in this frequency regime. It consists of air holes drilled through an infinite silicon slab, arranged in a standard hexagonal lattice with periodicity α = 320 nm. Within the bulk photonic crystal we assume a line-defect waveguide that leads to the PC-air interface, where a properly designed termination layer of air holes is imprinted. The termination is designed to induce surface states at the PC-air interface with desirable dispersion and spatial characteristics. The line-defect waveguide is an area of unperforated silicon slab and it is chosen since it is a widely used scheme for guiding energy through the reflective photonic crystals; therefore, our design is compatible with the majority of the silicon-based PCs circuit components. By properly designing the interfacial termination layer we can tailor the properties of the non-radiating, dark, surface states in order to adapt and match the waveguide propagating modes and the free-space modes. As a result, we demonstrate a silicon-based photonic crystal structure that provides (a) the generation of well-defined and highly directional beams at the exit of the photonic crystal structure; (b) efficient beam multiplexing, i.e., the formation of two well defined beams that exhibit sufficiently high spectral isolation, as well as sufficiently high spatial separation. In particular we present two design approaches; the first is able to generate two beams at the operation wavelengths λ1 = 1.37 μm and λ2 = 1.5 μm, with high spatial separation defined by the emission angles φ1 = +20 deg and φ2 = -23 deg and the second generates two beams at λa = 1.42 μm and λb = 1.52 μm with emission angle φa = +1 deg and φb = 23 deg. The wavelengths of operation, as well as the emission angle, can be engineered at will.

We report the design and analysis of a novel sub nanometer narrow bandwidth grating using sidewall corrugated SOI rib waveguide. Wavelength filters with narrow bandwidth, smaller device footprint, lesser complexity and higher fabrication tolerance are essential for on-chip Wavelength Division Multiplexing (WDM) applications. In this paper, we propose a simple and effective method to obtain ultra-narrow bandwidth. The design consists of a deeply etched rib waveguide with slab sidewall corrugation for index modulation. In order to understand the device characteristics, numerical investigations were employed. With a corrugation width of 60 nm, narrow 3-dB bandwidth of 0.5 nm at 1550 nm was achieved. Our results also serve as the precursor for a Dense WDM architecture with channel spacing of 0.8 nm, realizable by varying the grating pitch.

Nowadays, a huge data traffic requires a high-speed processing, so the use of optical memories is a logical solution for high speed data processing. In this paper, a large-scale parallel integration of wavelength addressable optical bit memories is presented based on three photonic crystal nanocavities (C₁, C₂, and C₃) filled with liquid crystal. Each cavity is storing two different wavelengths, where each wavelength is representing a single bit. We have calculated Q factors in basing and unbiasing states for C₁, C₂, and C₃. Also, the group velocities across the storage cell have been measured in the biased and unbiased cases for all cavities to confirm the storage and confinement. The maximum consumed power for six bits optical memory is only 13 nW.

In this paper, the results of the successful fabrication as well as the optical backscattering characterization of single plasmonic gold dipole nanoantennas on a SiO₂/Si layered substrate are shown. The nanoantennas were designed for a scattering resonance in the NIR range. In contrast to usually used glass substrates, a six inch Siwafer with a thermally oxidized SiO₂ layer in combination with an electron beam lithography lift-off fabrication process has been used for the sake of compatibility with microelectronics fabrication processes. In order to achieve high structural resolutions, a bilayer resist system with different exposure sensitivities was realized. In a second step, the entire resist thickness of 540 nm was reduced to 150 nm in a single layer. The SiO₂ thickness was chosen in a way that the optical near-field interactions of the nanoantennas with the silicon substrate are decoupled. The SEM characterization of the fabricated structures shows precise nanoantenna geometries with low edge roughness in the case of the bilayer resist system. The aspect ratio of the fabricated nanoantenna structures is slightly decreased compared to the desired value of five. Depending on the applied e-beam exposure dose, an increase of the structural cross-section, i.e. critical dimension of the dipole width, was observed. Furthermore, the single resist layer introduces some structuring issues. The spectral behavior of the nanoantenna structures was investigated with an optical confocal broadband backscattering measurement setup allowing the spectral characterization of single nanoantenna structures. The developed numerical models helped to understand the impact of the manufacturing imperfections providing improved designs.

In this paper, an input coupler to efficiently couple light from a silicon slab waveguide to a silicon slot waveguide is proposed. The slab waveguide is tapered to slot waveguide dimension for minimum loss. It also allows a miniaturized and compact design. Typical parameter such as taper length is numerically optimized to obtain minimum insertion and return losses as compared to other existing couplers.

Photonic lab-on-a-chip portable platforms have proved to be very sensitive, rapid in analysis and easy-to-use. However, they still rely on a bulk light source to operate, thus hindering the actual portability and potential for commercial realization. In the present paper we have proposed a design for a light emitting structure that could be easily implemented on chip. The design consists of a Si₃N₄ strip waveguide on SiO₂ substrate, with an active material that emits light as top and lateral cladding. The cross-section of the waveguide was optimised to support both excitation and emission as guided modes, with a high mutual overlap and high confinement to the cladding. This ensures an efficient light emission activation from the cladding and a stable propagation along the waveguide. The proposed structure shows to be operative along the visible range; demonstrated from 400nm to 633nm. The procedure we have followed along this report can be virtually used for designing the cross-section geometry of any strip waveguide system so that the performance is optimised for a given cladding refractive index and emission and excitation wavelengths. In addition we have proposed the use of polymeric quantum dots as the gain material to be used as active cladding. The ease of on-chip integration of this gain material via spin-coating, together with the simplicity of our light emitting waveguide, makes our light source design suitable for large-scale integration on Si chip. Specially, for lab-on-chip applications where multiplexed operation is essential.