This PDF file contains the front matter associated with SPIE Proceedings Volume 10923, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

Modern sub-28nm CMOS process nodes, namely FinFET and thin-body silicon-on-insulator have front-end layer thicknesses that are too thin to confine an optical mode. Integration of silicon photonics in these nodes necessitates the development of a deposition process that forms the waveguide structures with sufficient geometries after the CMOS front-end processing. As a step toward creating a photonics process module that can be added to these nodes, we demonstrate the integration of deposited polysilicon photonic platform in a low-power 65nm bulk CMOS process node in a 12” wafer foundry. This process module is designed with minimal number of additional masks to control the fabrication costs by optimizing the fabrication steps and reusing original process’s mask set (~5 additional masks among +40 masks required for the state-of-the-art CMOS nodes). The center of the platform is a polysilicon deposition step, which creates the waveguide layer, followed by a low-temperature crystallization process, which does not impact the electronics. All the passive and active photonic devices are fabricated by patterning and doping this layer. Transistor’s source/drain doping implantations are postponed after finishing and doping photonic polysilicon in order to avoid affecting transistors and reusing the implantation masks for doping active photonic devices as well. The waveguide loss ranges from 10-20dB/cm at 1310nm wavelength. To ease the loss optimization at wafer-scale, deep trench isolation has been added in photonic rows to optically isolate photonics from lossy silicon bulk. Grating couplers are used to couple in/out the light into the chip with 5dB loss. Micro-ring depletion-mode modulators achieved Q-factors of >5k and ~1.6THz free spectral range (FSR) enabling 10 channels in DWDM links. Resonant defect-based photodetectors are utilized on the receive side with 10% quantum efficiency at 5V reverse bias. Our first system demonstrations in this platform are O-band wavelength division multiplexed (WDM) optical transceivers using ring-resonators. Chips are designed in a modular fashion with 64 transceiver macros supporting 4 stand-alone transmit and receive WDM rows each with up to 16 individual channels. Each macro contains about 0.5 million transistors including transceiver’s analog custom front-ends, a digital backend, and microrings’ thermal tuners synthesized by original CMOS technology’s IP standard cells. We have used a variety of available transistor types with different oxide-thicknesses and threshold voltages to optimize energy-efficiency of the electronics. We have characterized the transistor performance across the die and wafer by measuring the frequency of the ring-oscillators embedded in each macro, and observed that the normal distribution is consistent with the foundry provided models for the native CMOS process. Electronics are operating using nominal supply voltage of 1.2V. We achieved 10Gb/s transmission with 4.7dB extinction ratio, and bit-error-rates of below 1e-10 at 7Gb/s with -3dBm sensitivity per channel. Total electrical energy-efficiency is about 600fJ/b (100fJ/b for Tx and 500fJ/b for Rx).

The field of silicon photonics has expanded rapidly over the past several decades. This has led to a degree of standardisation in the commercial device fabrication foundries that are available for universities and fabless companies alike. Whilst this is advantageous in terms of yield, repeatability etc., it is not conducive for researchers to develop new and novel devices for future systems. CORNERSTONE offers researchers a flexible device prototyping capability that can support photonics research around the world. The CORNERSTONE project (Capability for OptoelectRoNics, mEtamateRialS, nanoTechnOlogy, aNd sEnsing) is a UK Engineering and Physical Sciences Research Council (EPSRC) funded project between 3 UK universities: University of Southampton, University of Glasgow and University of Surrey. The project is based on deep-ultraviolet (DUV) photolithography equipment, installed at the University of Southampton, centred around a 248 nm Scanner, the first of its kind in a UK university. Utilising these facilities, CORNERSTONE will offer a multi-project wafer (MPW) service on several silicon-on-insulator (SOI) platforms (220 nm, 340 nm & 500 nm) for both passive and active silicon photonic devices. This talk will give an overview of the CORNERSTONE project, present some of its early data, and summarise future MPW offerings.

Directly interfacing a photonic integrated circuit allows at best an alignment tolerance of a few micrometer due to the small dimensions of optical (coupling) features on chip, but when using microlenses integrated on the substrate-side, alignment tolerances for interfacing the chips can greatly be relaxed. This is demonstrated on a 750 μm thick chip with standard grating couplers (operation wavelength around 1550 nm). Low roughness silicon microlenses were realized by transferring reflowed photoresist into the silicon substrate using reactive ion etching. The microlens allows interfacing the chip from the backside with an expanded beam, drastically increasing lateral alignment tolerances. A 1 dB alignment tolerance of ±8 μm and ±11 μm (along and perpendicular to the grating coupler direction, respectively) was experimentally found when a 40 μm mode field diameter beam was used at the input.

Transfer printing is an enabling technology for the efficient integration of III-V semiconductor devices on a silicon waveguide circuit. In this paper we discuss the transfer printing of substrate-illuminated III-V C-band photodetectors on a silicon photonic waveguide circuit. The devices were fabricated on an InP substrate, encapsulated and underetched in FeCl₃, held in place by photoresist tethers. Using a 2x2 arrayed PDMS stamp with a pitch of 500 μm in x-direction and 250 μm in y-direction the photodiodes were transfer printed onto DVS-BCB-coated SOI waveguide circuits interfaced with grating couplers. 83 out of 84 devices were successfully integrated

The compatibility of silicon photonics with existing CMOS fabrication processes enables the possibility of large scale manufacturing of integrated photonics for communication applications above 100Gbit/s. However, the design of silicon photonics devices that are able to fulfil the high-performance requirements of broad optical bandwidth ( 50nm), low loss (2dB), high-speed (25Gbit/s), low driving voltage (≥ –5V), large fabrication tolerance (±20nm) and high reproducibility across the wafer is a unique practical challenge. In this work, we present a library of passive and active integrated silicon photonic devices that meet the stringent design requirements for communication applications. The design, modelling and experimental results are presented for low loss edge couplers with insertion loss < 2dB, low-loss waveguides and bends, high-performance polarization beam splitter with extinction ratio above 25 dB, broadband directional couplers with optical bandwidth above 50nm, and arrayed waveguide gratings (AWGs) with low insertion loss < 1dB. Last but not least, high-speed silicon modulators (25Gbit/s) with phase efficiency VπL ≤ 2V.cm at DC reverse bias of ≥ –5V and low propagation loss of ≤ ∼1dB/mm are demonstrated.

Silicon photonics is one of the most promising candidates to achieve lab-on-a-chip systems. Making use of the evanescent-field sensing principle, it is possible to determine the presence and concentration of substances by simply measuring the variation produced by the light-matter interaction in the real part of the mode effective index (in the near-infrared band), or in its imaginary part in a specific range of wavelengths (in the mid-infrared band). Regardless of which is the operating wavelength range, it is essential to select the proper sensing waveguide in order to maximize the device sensitivity. In this work we will review the potential of diffractionless subwavelength grating waveguides (SWG) for sensing applications by demonstrating their powerful capability to engineer the spatial distribution of the mode profile, and thereby to maximize the light-matter interaction. Among other things, we will demonstrate that the SWG waveguide dimensions used until now in the near-infrared are not optimal for sensing applications. In the mid-infrared band, due to the unacceptable losses of silicon dioxide for wavelengths longer than 4 μm, an additional effort is required to provide a more convenient platform for the development of future applications. In this regard, we will also show our recent progress in the development of a new platform, the suspended silicon waveguide with subwavelength metamaterial cladding. A complete set of elemental building blocks capable of covering the full transparency window of silicon (λ < ∼8.5 μm) will be discussed.

Germanium has become a material of high interest for mid-infrared (MIR) integrated photonics due to its complementary metal-oxide-semiconductor (CMOS) compatibility and its wide transparency window covering the 2-15 μm spectral region exceeding the 4 μm and 8 μm limit of the Silicon-on-Insulator (SOI) platform and Si material respectively. Here, we present suspended germanium waveguides operating at wavelengths of 3.8 μm and 7.67 μm with propagation losses of 2.9 ± 0.2 dB/cm and 2.6 ± 0.3 dB/cm respectively.

Thanks to its high Kerr non-linearity and its low linear absorption, silicon is a material of choice for optical devices in the mid-infrared (from 3 to 5 microns) such as microresonators. In this wavelength range, the available optical sources such as quantum cascade lasers have a limited tunability. Tuning the refractive index of silicon can be achieved by a temperature change of the chip and has been previously demonstrated on ring resonators using integrated heaters or thermo-electric elements. We present a new method for thermo-optical tuning of silicon devices by directly using the light from a laser diode operating at 450 nm. The blue light focused on the silicon induces a local elevation of temperature and thus the refractive index locally increases. When applying this method on silicon ring resonator, the elevation of temperature leads to a decreasing free-spectral range and thus shift the resonances to lower frequencies. At 4.5 µm we measured a tuning efficiency of 200 MHz per mW of incident light. Numerical simulations of the thermo-optical effect show the locality of this tuning method, and confirm the experimental results. Finally a frequency study of the response of this method is performed and a time constant of the order of the micro-second is measured. In conclusion, we propose a fast, local, and non-invasive method for tuning silicon resonators operating in the mid-infrared that can be extended to any silicon-based device.

Silicon-on-Insulator devices are particularly sensitive to fabrication errors. As an example, a deviation in waveguide height or width of as little as 1nm translates directly to a 1nm offset in the transfer function of any interferometric devices (such as a ring resonator) constructed using the said waveguide. Therefore, even as fabrication tolerance continues to improve, post-fabrication treatment is often the only way of ensuring device uniformity for particularly demanding applications. This work proposes a novel approach for post fabrication trimming of SOI devices based on localised laser annealing of HSQ cladding layer. HSQ is a versatile material often used in fabrication of SOI devices as both the mask material for electron-beam lithography resist and as a cladding or planarization layer due to its similarity to conventional silica. However, unlike silica, the refractive index of HSQ can be changed significantly (up to ΔnHSQ = 3.26*10-2) by thermal processing. We utilise this property for trimming by cladding a conventional SOI waveguide optimised for TE propagation (height h=220 nm, width=500nm) with a layer of HSQ and then permanently changing the refractive index of the cladding via laser annealing. This approach allows us to select individual devices and only apply the change where necessary. As a demonstrator, we trim a resonance of a racetrack resonator by 1.3nm. The technique has proven to be robust with no parameter drift observed 7 days after trimming and no thermal cross-talk to neighbouring devices. Furthermore, unlike its predecessors, it is based on a standard fabrication process and does not require expensive specialised equipment.

Amorphous Al2O3 is an attractive material for integrated photonics, providing both active and passive functionalities. Al2O3 exhibits high solubility for rare-earth ions with moderate quenching of luminescence, a wide transparency window (150-7000 nm) and low propagation loss. It is therefore a very attractive material for visible, near- and mid-IR on-chip active devices. We have developed two different integration procedures to integrate Al2O3 onto passive photonic platforms. A double photonic layer integration scheme permits the low-loss integration of rare-earth ion doped Al2O3 onto the Si3N4 photonic platform. A single photonic layer integration scheme, based on the photonic damascene process, permits the creation of active and passive regions at the same level on a wafer, with the consequent reduction of the number of fabrication steps compared to the vertical integration of two materials. On-chip amplifiers on Si3N4 with more than 10 dB of net gain at 1550 nm as well as the realization of narrow linewidth lasers on active-passive Al2O3 for label-free sensing applications will be discussed.

Antimonide-based materials rely on the GaSb, InAs, AlSb, InSb binary compounds and their quaternary or pentanary alloys (AlGaAsSb, GaInAsSb, AlGaInAsSb,.. ). This technology exhibits several distinctive properties as compared to other semiconductors: type-I to type-III band alignments, giant band offsets, low effective masses of electrons and holes, direct bandgaps between 0.15 and 1.7 eV [1]. Conventional laser diodes (LDs) rely essentially on GaInAsSb type-I quantum wells (QWs) confined by AlGa(In)AsSb barrier layers. Low threshold currents and high T0 have been demonstrated between 1.5 and 3.4 µm [2]. The AlGaInAsSb pentanary barrier is needed to extend the wavelength beyond 3 µm while keeping a type-I band alignment [3] even though it makes the epitaxial growth complex. Single mode operation has been achieved with both DFB lasers [4-6] and VCSELs [7, 8] using the same active zone. At longer wavelength, interband cascade lasers (ICLs) based on GaInSb/InAs type-II p-n junctions stacked in series exhibit room temperature cw emission between 3.5 and 5 µm, including single mode operation of DFB lasers [9]. At still longer wavelength InAs/AlSb quantum cascade lasers (QCLs) benefit from the low InAs effective mass and giant conduction band offset. High performance have been demonstrated all the way from 2.6 µm up to 25 µm, particularly at long wavelength which is an asset of this technology [10]. Type-II InAs/GaSb superlattices (T2SLs), play an increasing role in the field of high performance IR photodetection [11]. The staggered type-II, i.e. type-III, alignment allows controlling the cut-off wavelength from the short- to the long-IR range simply by changing the individual layer thickness. This technology is now competing against established HgCdTe IR systems, particularly at long wavelength. Still, these GaSb-rich T2SLs suffer from GaSb native defects which limits their dark current above theoretical expectations [12]. This opened the way to the implementation of so-called "Ga-free" InAs/InAsSb T2SLs which exhibit improved carrier lifetimes [13]. The photodetector performance however still lag behind theory and work remains to be done [14]. It is noticeable that the Sb technology is very versatile. The whole NIR to LWIR wavelength range can be covered and multi-color photodetection systems can be achieved by engineering at will superlattices based on GaSb, InAs, AlSb and any combination of them [15]. Moreover, the growth of these structures in production systems has already been demonstrated [16, 17] opening the way to commercialization. On another ground, the evolution toward smart, integrated, sensors requires integrating III-V optoelectronic devices with Si-based platforms. The epitaxial growth of III-V compounds on Si has thus been the focus of renewed attention for about a decade now. We have shown that the Si substrate preparation and the III-Sb nucleation on Si are crucial steps [18, 19]. This allowed us demonstrating a variety of epitaxially integrated optoelectronic devices such as laser diodes [20, 21], photodetectors [22] and the first ever QCL grown on Si [23]. In this presentation we review the recent results obtained on the integration of antimonide-based optoelectronic devices epitaxially grown on Si substrates. We will show that this technology is very attractive for future III-V on Si integration, and we will discuss future integration schemes. Part of the work performed at Univ. Montpellier has been supported by the French program on “Investment for the Future” (EquipEx EXTRA, ANR-11-EQPX-0016), research agency (ANR) and defense agency (DGA) and by the European Union (FP6, FP7, FEDER, H2020). [1] I. Vurgaftman, J.R. Meyer, and L.R. Ram-Mohan, J. Appl. Phys. 89, 5815 (2001). [2] See, e.g., G. Belenky and L. Shterengas, M. V. Kisin, and T. Hosoda, in Semiconductor lasers: fundamental and applications, edited by A.N. Baranov and E. Tournié, pp. 441 – 486 (Woodhead Publishing, 2013). [3] M. Grau, C. Lin, O. Dier, C. Lauer, and M.-C. Amann, Appl. Phys. Lett. 87, 241104 (2005). [4] S. Forouhar, R. M. Briggs, C. Frez, K. J. Franz, and A. Ksendzov, Appl.Phys. Lett. 100, 031107 (2012). [5] P. Apiratikul, L. He, and C. J. K. Richardson, Appl. Phys. Lett. 102, 231101 (2013). [6] Q. Gaimard, M. Triki, T. Nguyen-Ba, L. Cerutti, G. Boissier, R. Teissier, A.N. Baranov, Y. Rouillard, and A. Vicet, Opt. Express 23, 19118 (2015). [7] A. Bachmann, K. Kashani-Shirazi, S. Arafin and M.-C. Amann, IEEE J. Sel. Top. Quantum Electron. 15, 933 (2009). [8] D. Sanchez, L. Cerutti, and E. Tournié, J. Phys D: Applied Physics 46, 495101 (2013). [9] See, e.g., I. Vurgaftman, R. Weih, M. Kamp, J. R. Meyer, C. L. Canedy, C. S. Kim, M. Kim, W. W. Bewley, C. D. Merritt, J. Abell, and S. Höfling, J. Phys. D: Applied Physics, 48, 123001 (2015). [10] See, e.g., A.N. Baranov and R. Teissier, IEEE J. of Select. Top. in Quant. Electron. 21, 1200612 (2015). [11] See, e.g., A. Rogalski, P. Martyniuk, and M. Kopytko, Appl. Phys. Rev. 4, 031304 (2017). [12] M. Delmas, J.-B. Rodriguez, R. Rossignol, A.S. Licht, E. Giard, I. Ribet-Mohamed, and P. Christol, J. Appl. Phys. 119, 174503 (2016). [13] E. H. Steenbergen, B. C. Connelly, G. D. Metcalfe, H. Shen, M. Wraback, D. Lubyshev, Y. Qiu, J. M. Fastenau, A. W. K. Liu, S. Elhamri, O. O. Cellek, and Y.-H. Zhang, Appl. Phys. Lett. 99, 251110 (2011). [14] E. H. Steenbergen, G. Ariyawansa, C.J. Reyner, G. D. Jenkins, C. P. Morath, J. M. Duran, J. E. Scheihing, V. M. Cowan, Proc. SPIE 10111, 1011104 (2017). [15] See, e.g., M. Razeghi et al. [16] D. Loubyshev, J.M. Fastenau, M. Kattner, P. Frey, A.W.K. Liu, M.J. Furlong, Proc. SPIE 10177, UNSP 1017718 (2017). [17] P.C. Klipstein et al., J. Electron. Mater. 46, 5386 (2017). [18] K. Madiomanana, M. Bahri, J.B. Rodriguez, L. Largeau, L. Cerutti, O. Mauguin, A. Castellano, G. Patriarche, and E. Tournié, J. Cryst. Growth 413, 17 (2015). [19] J.B. Rodriguez, K. Madiomanana, L. Cerutti, A. Castellano, and E. Tournié, J. Cryst. Growth 439, 33 (2016). [20] J.R. Reboul, L. Cerutti, J.B. Rodriguez, P. Grech, and E. Tournié Appl. Phys. Lett. 99, 121113 (2011). [21] A. Castellano, L. Cerutti, J.B. Rodriguez, G. Narcy, A. Garreau, F. Lelarge, and E. Tournié, APL Photonics 2, 061301 (2017). [22] Q. Durlin, J.P. Perez, L. Cerutti, J.B. Rodriguez, T. Cerba, T. Baron, E. Tournié, P. Christol, Infrared Phys. and Technol., to be published. [23] H. Nguyen-Van, A.N. Baranov, Z. Loghmari, L. Cerutti, J.B. Rodriguez, J. Tournet, G. Narcy, G. Boissier, G. Patriarche, M. Bahriz, E. Tournié, R. Teissier, Sci. Rep., 8, 7206 (2018).

Tellurite glasses have promising material properties in applications for linear and nonlinear integrated optical devices. Tellurite glasses have high rare earth solubilities for applications in rare earth doped lasers as well as high nonlinear refractive indices, Raman gain coefficients and acousto-optic figures of merit. However, it is difficult to take advantage of tellurite glass properties in silicon photonics, as the waveguiding materials available for use in silicon photonic devices are typically limited to silicon, silicon dioxide, silicon nitride, and germanium. Here, we report on a tellurium oxide whispering gallery resonator, integrated onto a silicon photonic chip and coupled to a silicon waveguide. The silicon waveguides are fabricated using a standard foundry process and the cladding oxide is etched in a ring shape with precise alignment to the bus waveguides at gaps from 0.2 to 1.0 μm to form the cavity. Post processing deposition of a tellurium oxide film coats the bottom of the etched oxide cavity, forming a tellurium oxide waveguiding layer, into which light can be coupled from the silicon waveguide. A resonator with a radius of 40 μm and a 1.1-μm-thick tellurium oxide coating is measured to have an internal Q-factor of greater than 1E5. These results illustrate the potential for integration of tellurite glass devices into silicon photonic microsystems. Applications of this cavity structure in optical sensing, design considerations and methods to improve performance will be discussed.

We present a U-bend design for traveling wave III-V gain devices, such as semiconductor optical amplifiers and laser diodes. The design greatly simplifies the butt-coupling between the III-V chip and silicon-on-insulator photonic circuit by bringing the I/O ports on one facet. This removes the need for precise dimension control otherwise required for 2-side coupling, therefore increasing the yield of mounted devices towards 100%. The design, fabrication and characterization of the U-bend device based on Euler bend geometry is presented. The losses for a bend with a minimum bending radius of 83 μm are 1.1 dB. In addition, we present an analysis comparing the yield and coupling losses of the traditionally cleaved devices with the results that the Euler bend approach enable, with the final conclusion that the yield is improved by several times while the losses are decreased by several dB.

Erbium (Er) has offered a means towards optical amplification around 1.5 µm due to the intra-4f transitions of Er3+ ions. Er silicates are of much interest due to a 3 order increase in the concentration of Er3+ ions in the film as opposed to different Er-doped materials. Unfortunately, the major hindrance toward optical gains in such erbium containing materials is the fast quenching of Er luminescence, mainly resulting from excitation energy dissipation at structural defects even with a small density, via resonant energy transfer processes among Er ions. In this work, we investigate effects of hydrogen passivation and micro/nano scale structures on the luminescence properties of Er silicates. Arrays of micron-sized erbium silicate structures are created via etching a silicon wafer followed by deposition of erbium metal onto the etched pits. After deposition, the photoresist is removed through lift off and the metal structures are subjected to high temperature oxygen annealing (1200˚C) for oxidation of the film. Hydrogen passivation is conducted in a H2 gas ambient between 500˚C and 900˚C. Rutherford backscattering spectroscopy (RBS) and x-ray diffraction (XRD) are used to determine the composition and crystal structure information of the resultant thin films and photoluminescence (PL) is measured for their luminescence properties. The results show a significant decrease of photoluminescence in the ultraviolet/visible (UV/Vis) range, accompanied by an increase in both the intensity and lifetime of the near-infrared (NIR) luminescence emission around 1.5 µm wavelength from Er oxide/silicate compound thin films, following passivation in a H2 gas. Furthermore, samples with arrays of micro-structured Er silicates exhibit stronger NIR luminescence than the thin film sample. Combining with computer simulations, we identify the possible mechanisms for the observed Er luminescence enhancement, and suggest promising routes toward optical amplification around 1.5 µm in Er compounds.

We present a chip-scale spectroscopic methane sensor, incorporating a tunable laser, sensor waveguides, and methane reference cell, assembled as a compact silicon photonic integrated circuit. The sensor incorporates an InP-based semiconductor optical amplifier/photodetector array, flip-chip soldered onto a silicon photonic substrate using highprecision waveguide-to-waveguide interfaces. The InP chip provides gain for a hybrid external cavity laser operating at 1650 nm. The sensor features a 20-cm-long TM-mode evanescent-field waveguide as the sensing element and is compatible with high-volume wafer-scale silicon photonics manufacturing and assembly processes. This sensor can be an enabling platform for economical methane and more general distributed environmental trace-gas monitoring.

The Mid-IR spectral range (2.5 μm up to 12 μm) has been considered as the paradigm for innovative silicon photonic devices. In less than a decade, chemical sensing has become a key application for Mid-IR silicon photonic devices because of the growing potential in spectroscopy, materials processing, chemical and biomolecular sensing, security and industry applications. Measuring in this spectral range, usually called molecule fingerprint region, allows to address a unique combination of fundamental absorption bands orders of magnitude stronger than overtone and combination bands in the near IR. This feature provides highly selective, sensitive and unequivocal identification of the chemicals.

Progress in Cascade Laser technology (QCL and ICL) allows to select emission wavelengths suitable to target the detection of specific chemicals. With these sources, novel spectroscopic tools allowing real-time in-situ detection of gasses down to traces are nowadays commercially available.

Mid-IR Si photonics has developed a novel class of integrated components leading to the integration at chip level of the main building blocks required for chemical sensing, i.e. the source, the PICs and the detector. Three main directions of improvement can be drawn: i) extend the range of wavelengths available from a single source, ii) move beam handling and routing from discrete optics to PICs and iii) investigate detection schemes for a fully integrated on-chip sensing.

This paper reviews recent key achievements in the miniaturization and the co-integration of photonics devices at chip and packaging level to address cost, size and power consumption. Perspectives on potential applications will also be presented.

This work describes the integration of mid-infrared (MIR) silicon photonics with PDMS microfluidics to perform absorption spectroscopy of IPA-water solutions. The MIR spectral region contains strong absorption bands for many molecules, and photonic devices operating in the MIR can be used in many sensing applications. In this work a preliminary demonstration of a silicon-on-insulator (SOI) device is carried out in which the transmission spectra of different concentrations of water-IPA solutions are measured at wavelengths between 3.725 μm and 3.888 μm. A PDMS microfluidic channel is integrated with the waveguides in order to improve the repeatability of sample handing, reduce reagent volumes and prevent evaporation of the analyte. A microfluidic channel with 3000 x 100 μm cross-section and 30 mm length is bonded to a SOI chip comprising 500 nm thick rib waveguides and a 2 μm thick SiO₂top cladding isolating the waveguide mode from the analyte. Trenches were patterned into the SiO₂ cladding to create sensing windows of varying lengths (10 μm to 3mm) along different waveguides. The devices were used to detect an expected IPA absorption peak at 3.77 μm, and concentration as low as 3% IPA in water (by volume) was detected. Further work will focus on increasing the sensitivity of the measurement by using increased interaction lengths, reduction of noise and instability, and on the detection of drugs using transmission measurements over a broader wavelength range.

The ever decreasing demand for bandwidth in optical communications has made silicon photonics one of the promising technologies as it can dramatically reduce energy consumption and footprint in photonic integrated circuits (PIC). Many research efforts have aimed to incorporate silicon into the PIC platform by using it as a resonant reflector in the form of a microdisk, racetrack resonator, ring resonator or photonic crystal (PhC) cavity. Tuning of these devices allow for modulation of the lasing frequency by means of the electro-optic or thermo-optic effect. Our solution utilises a III-V hybrid laser with a reflective semiconductor optical amplifier (RSOA) and a PhC cavity resonant reflector. Current research shows electro-optical modulation of a PN junction on the Si-reflector as a means of tuning the reflectance wavelength. This work focuses on the thermo-optical effect in silicon to achieve modulation of the lasing frequency. Modulation of the current to the PN junction on the Si-reflector of the external cavity laser will change the refractive index which will tune the reflectance wavelength and hence modulate the lasing frequency. PhC cavities are smaller in area than a typical ring resonator and have larger free spectral range that results in less severe mode competition effects. For trace gas detection a frequency modulated laser scanned across the absorption frequency of the target gas will result in change in the output power of the laser. The PhC laser we demonstrate shows to have a very small intensity modulation (IM) on the output offering it as an ideal candidate for this application. Experimental results show the laser to have a threshold current of 15 mA with output optical power of 300 µW. With an applied heating power of 25 mW, a frequency shift of 10 GHz was observed. At a modulation frequency of 10 kHz, a modulation depth of 2 GHz was observed.

For the past century, industrial temperature measurements have relied on resistance measurement of a thin metal wire or thin metal film whose resistance varies with temperature. Today’s resistance thermometers can routinely measure temperatures with uncertainties of 10 mK to 100 mK over a broad range of temperatures in varied settings ranging from a stove top to an industrial broiler to a nuclear power plant. However, for all their utility, resistance thermometers remain vulnerable to mechanical and thermal shock and attack from harsh chemicals. The resultant drift in sensor resistance necessitates frequent off-line, expensive, and time-consuming calibrations. These fundamental limitations of resistance thermometry, born of material properties, have generated considerable interest in developing photonic temperature sensors. Photonic approaches hold the promise of leveraging recent advances in frequency metrology and of achieving greater mechanical and environmental robustness. In recent years many groups including ours have demonstrated a suite of photonic devices including silicon photonic devices that can not only meet but exceed the state of art in temperature metrology

Astrophotonics is the application of photonics to astronomical instrumentation. This relatively young and rapidly developing Field can play a transformative role in astronomy, offering new functionality, increased efficiency and stability, and allowing traditional large bulk optics instruments to be replaced with miniature scalable modular instruments. Astrophotonics has reached a stage of maturity in which many prototypes and concepts are now being employed in facility instruments, and the field is transitioning from purely research and design to become part of mainstream astronomical instrumentation. As part of this process integrated optical components are playing an increasingly important role, often supplanting the fibre optics used in early prototypes. In this paper we provide a brief review of astrophotonics, paying particular attention to those areas in which silicon photonics can be beneficial, as well as noting the requirements and challenges for silicon astrophotonics.

Cavity optomechanics explores reciprocal interactions between light and mechanical vibrations, down to the quantum level. It constitutes a very promising research field on both fundamental and applied physics, in areas ranging from quantum coherent control over bulky mechanical objects to ultra-sensitive on-chip inertial sensors.

In most optomechanical systems, the resonance frequency of an optical cavity is shifted due to the oscillations of a mechanical resonator, which is in turn put into motion through optical forces such as radiation pressure or optical gradient forces. This interaction is quantified by the said optomechanical vacuum coupling rate, that is a measurement of the interaction between a single photon and phonon. Increasing this factor is mandatory for both quantum and classical applications. In order to do so, low volume and high quality factors cavities and resonators are required. The optomechanical strength is also proportional to the intra-cavity power, and it is then advantageous to avoid any non-linear or thermal effect to insure the stability of the system.

Here, we investigate the use of sub-wavelength grating (SWG) structures as a way to induce optomechanical coupling. This paper will present the design, realization, and characterization of various geometries fabricated on standard SOI wafers and working over the telecom C-band. We expect a strong optomechanical vacuum coupling rate, that we believe will open new perspectives towards highly sensitive and stable on-chip optomechanical systems, optical signal processing, or even quasi room temperature back-action cooling.

The temperature of earth depends upon the balance between the energy enterring and leaving the planet. The dynamic balance has been broken by the drastical increase of greenhouse gases generated by human activities during the past 150 years. Thus, monitoring of the global emission of greenhouse gases is urgent for human beings. Fourier transform spectroscopy (FTS) in infrared wavelength range is an effective measure for this purpose. An infrared spectrum represents a fingerprint of a material with absorption peaks corresponding to the vibration of the bonds of the atoms making up the material. Because each material is a unique combination of atoms, no two compounds produce the exact same infrared spectrum. Therefore, infrared spectroscopy can result in a positive identification (qualitative analysis) of every kind of materials. In addition, the size of the peaks in the spectrum is a direct indication of the amount of material present. Compared to dispersive optics or filter based spectroscopy approaches, FTS has a few significant advantages, such as high throughput, high signal-to-noise ratio, and high sensitivity. However, the size, weight and free space optics components make FTS a laboratory only instrument demanding extensive human involvement. In this paper, we report a demonstration of an on-chip Fourier transform spectrometer near 3.3 μm wavelength on silicon-on-sapphire. Propagation loss of 5.2 dB/cm has been experimentally demonstrated for strip waveguides. The on-chip FTS comprises an array of Mach–Zehnder interferometers (MZIs) with linearly increased optical path differences. The recovery of the spectrum of an inter-band cascaded laser has been demonstrated.

We propose, theoretically analyze and experimentally demonstrate a novel broadband operating Si-nanowire multistage delayed Mach-Zehnder interferometer (DMZI) based (De)MUXs utilizing wavelength insensitive couplers (WINCs) and the discrete phase shifters for the phase matching between cascade-connected DMZIs. Based on the coupled mode theory and transfer matrix method, the coupling characteristics of a Si-wire directional coupler (DC) and WINC are analytically discussed from the viewpoint of a wavelength sensitivity of coupling efficiency [κ_Coup(λ)]. We theoretically verify that the operating window of the proposed DeMUX can be as broad as >110 nm by introducing the WINC and the additional phase matching between multiple DMZIs. Based on the theoretical analysis, 300-mm waferscale ArF-immersion lithography process are used to fabricate the Si-wire-based 1×4Ch DeMUXs. It will be shown that >110-nm-wide operating window for the proposed DeMUXs with four kinds of channel spacings (Δν=400, 800, 1250, 1900GHz). The theoretically identified aspect of nearly constant κ_Coup(λ) by the WINC and the phase matching by the phase shifter is proven to be extremely effective way to make the production yield much better, because DeMUX spectral response keeps nearly constant even if insufficient fabrication accuracy makes spectral response be shifted toward longer or shorter wavelength side. In addition, it will be shown that operating spectral window could be made much wider by optimizing the WINC design parameters in the DeMUX configuration. The proposed scheme would be attractive for increasing available channel count without inducing any excessive losses, which makes the proposed scheme more practical in WDM optical transceivers.

Temperature sensitivity is an issue that severely affects many integrated silicon photonic devices. Proper circuit functionality is normally ensured by active thermal control at the expense of energy consumption. In some cases, athermal behavior can be achieved exploiting cladding materials with a negative thermo-optic coefficient to counterbalance the positive coefficients of silicon and silica. On the other hand, in echelle grating filters this method is not effective because in the slab free-propagation region the modal overlap with the cladding is small, especially for TEpolarized light. Moreover the need to add non-standard materials to the established silicon-on-insulator (SOI) fabrication process could make these solutions impractical. Here we present the design of a temperature-insensitive echelle grating demultiplexer with four channels operating in the TE polarization that does not use any materials with negative thermooptic coefficient and relies exclusively on standard processes for SOI photonics. The design exploits a temperaturesynchronized Mach-Zehnder interferometer as input to the echelle to compensate the shift of the imaged field with temperature. The device achieves a significant reduction in the temperature dependence of the overall transmission with a residual channel wavelength fluctuation smaller than 45 pm over a temperature range of 20 K, compared to a 1.6-nm shift for the same grating with a conventional waveguide input. The excess loss due to the use of the Mach-Zehnder input is no more than 0.7 dB for all four channels. Furthermore, the proposed design shows a very good tolerance to fabrication uncertainty, with minimum degradation of the performance for waveguide width variations of 10 nm.

In recent years, we have presented results on the development of a variety of silicon photonic devices such as erasable gratings and directional couplers, tunable resonators and Mach-Zehnder interferometers, and programmable photonic circuits using germanium ion implantation and localised laser annealing. In this paper we have carried out experiments to analyse a series of devices that can be fabricated using the same technology, particularly silicon-on-insulator racetrack resonators which are very sensitive to fabrication imperfections. Simulation and experimental results revealed the ability to permanently optimise the coupling efficiency of these structures by selective localised laser annealing.

Low loss Ge-on-Si waveguides are demonstrated in the 8 – 14 μm atmospheric transmission window, a technology that will enable detection and sensing of unique molecular vibrations. Such a low cost platform would have applications in key markets such as pollution monitoring, explosives detection and point of care diagnostics. Rib-waveguides are fabricated using electron beam lithography and dry etching. The waveguides propagation losses are characterized using the Fabry-Perot technique, and are found to be below 5 dB/cm across the measurement range of 7.5 to 11 μm wavelength, reaching as low as ~ 1 dB/cm. The contribution to the losses are analyzed using the experimentally measured Si substrate losses, and the calculated scattering losses from an analytical model. The results verify the feasibility of the Ge-on-Si platform for integrated mid-infrared photonics and sensing.

The broad functionality of the silicon-based photonic platform has led to a number of exciting demonstrations in both linear and nonlinear integrated photonics. Hydrogenated amorphous silicon (a-Si:H) films exhibiting nonlinear refractive indices an order of magnitude larger than c-Si can be deposited at a low temperature (typically 200 - 400 °C) and patterned by the same technology as c-Si, making them compatible with back-end-of-the-line (BEOL) CMOS technology. SiNx waveguides can be fabricated with extremely low losses providing long on-chip interconnects with high optical efficiency. Such specialty waveguiding layers can be combined into a multilayer silicon-based photonic platform for sophisticated, multi-material and multi-functional platforms. Here we will discuss our recent work in the a-Si:H waveguide platform demonstrating highly nonlinear interactions with mW-level peak pump powers, as well as the ability to integrate this waveguide platform in a multi-layer configuration with low-loss SiNx waveguides. We show four-wave mixing (FWM) frequency conversion in an a-Si:H waveguide addressed with a SiNx waveguide through interlayer coupling devices. Additionally, we will discuss a variety of demonstrations in silicon-based waveguides that exploit nonlinear optical interactions including frequency conversion and signal depletion. We will show the multilayer integration of these devices with their counterparts made from low-loss SiNx waveguides, etc. Furthermore, we will discuss these devices for a variety of applications including optical signal processing, logic, and security.

With silicon photonics going fabless, large-scale silicon photonic integrated circuits (PICs) have recently become a reality. Many of these PICs feature system reconfigurability to benefit from the cost-effective mass manufacture of a universal platform. However, reconfigurable silicon PICs relying on the weak, volatile thermo-optic or electro-optic effect of silicon usually suffer from a large footprint and energy consumption. Recently, phase-change materials have shown great promise for energy-efficient, ultra-compact and ultra-fast non-volatile integrated photonic applications. Here, by integrating phase-change materials, Ge2Sb2Te5 (GST) with silicon microring resonators, we demonstrate a non-volatile, programmable, energy-efficient, and compact platform over the telecommunication range. By measuring and fitting the output spectra of the microrings covered with various lengths of GST in the amorphous and crystalline states, we characterize the strong broadband attenuation (~7.3 dB/μm) and optical phase (~0.70 nm/μm) modulation effects of the platform. By adjusting the energy and number of free-space laser pulses applied to the GST, we perform reversible and quasi-continuous tuning of the GST state, and the subsequent tuning of the attenuation and resonance of the microring resonators enabled by the thermo-optically-induced phase changes. Designed to achieve near critical coupling of the microring resonators when the GST is in the amorphous state, a non-volatile 1×1 optical switch with high extinction ratio as large as 33 dB is demonstrated. Our research constitutes the first step towards future large-scale programmable silicon PICs. With appropriate design, a broadband low-loss 2×2 optical switch could be electrically controlled which would be the building block for a future non-volatile routing network and optical FPGA. Reference: J. J. Zheng, A. Khanolkar, P. P. Xu, S. Deshmukh, J. Myers, J. Frantz, E. Pop, J. Hendrickson, J. Doylend, N. Boechler, and A. Majumdar, "GST-on-silicon hybrid nanophotonic integrated circuits: a non-volatile quasi-continuously reprogrammable platform," Opt. Mater. Express 8(6), 1551-1561 (2018).

Mode splitting induced by coherent optical mode interference in coupled resonant cavities is a key phenomenon in photonic resonators that can lead to powerful and versatile filtering functions, in close analogy to electromagnetically-inducedtransparency, Autler-Townes splitting, Fano resonances, and dark states. It can not only break the dependence between quality factor, free spectral range, and physical cavity length, but can also lead to group delay response and mode interactions that are useful for enhancing light-material interaction and dispersion engineering in nonlinear optics. In this work, we investigate mode splitting in standing-wave (SW) resonators implemented by cascaded Sagnac loop reflectors (CSLRs) and demonstrate its use for engineering the spectral profile of integrated photonic filters. By changing the reflectivity of the Sagnac loop reflectors (SLRs) and the phase shifts along the connecting waveguides, we tailor mode splitting in the CSLR resonators to achieve a wide range of filter shapes for diverse applications including enhanced light trapping, flat-top filtering, Q factor enhancement, and signal reshaping. We present the theoretical designs and compare the performance of CSLR resonators with three, four, and eight SLRs fabricated in silicon-on-insulator nanowires. We achieve high performance and versatile filter shapes via diverse mode splitting that agree well with theory. The experimental results confirm the effectiveness of our approach towards realizing integrated multi-functional SW filters for flexible spectral engineering.

GeSn is discussed as solution to realize the dream of a group IV light source integrated on a Si chip. Sn added into a Ge lattice decreases the conduction band energies leading to a direct bandgap semiconductor band structure. However, the compressive strain increases the direct band energy imposing a large Sn content in the GeSn bulk. In spite of many difficulties regarding the growth of epitaxial GeSn alloys on Si, several hundred nm thick GeSn layers with various Sn concentrations up to 15% could be realized and used as gain material for lasers. Nowadays research concentrates on increasing the Sn content towards 20 at% as well as structural layout. The challenge here is the decreasing quality at high Sn contents and the isolation of the active layer from the mists formed at the interface with Ge/Si which increase the laser threshold. In this direction we discuss the influence on lasing and threshold of MQW SiGeSn/GeSn heterostructures with different quantum well thicknesses. Other solution proposed is the change of intrinsic strain type from compressive into tensile by introducing Si3N4 stressors and also GeSn on Insulator technology. These methods are well known in CMOS technology and can be applied to very low Sn content GeSn alloys. The discussion on the best way to reach room temperature laser is addressed both theoretical and experimental.

The recent advances in the development of quantum cascade laser with room temperature operation in the mid infrared paved the way for the realization of wideband communication systems. Particularly, two mid-infrared atmosphere transparency windows lying between 3-5 μm and between 8-14 μm exhibit great potential for further implementation of wideband free space communications. Additionally this wide unregulated spectral region shows reduced background noise and low Mie and Rayleigh scattering. Despite the development of a plethora of photonic components in mid infrared such as sources, detectors, passive structures, less efforts have been dedicated to investigate polarization management for information transport. In this work, the potential of Ge-rich SiGe waveguides is exploited to build a polarization insensitive platform in the mid-infrared. The gradual index evolution in SiGe alloys and geometric parameter optimization are used to obtain waveguides with birefringence below 2×10^-4 and an unprecedented bandwidth in both atmosphere transparency windows i.e. near 3.5 μm and 9 μm. Following waveguide birefringence optimization an ultra-wideband and polarization insensitive multimode interference coupler was designed. The optimized structure shows a 4.5 μm wide bandwidth in transverse electric and transverse magnetic polarization at 9 μm wavelength. The developed ultra-wideband polarization insensitive photonic building blocks presented in this work pave the way for further implementation of free space communication systems in the mid infrared spectral region.

Mid-infrared (MIR) resonators with high quality (Q) factors play crucial roles in a variety of applications in nonlinear optics, lasing, biochemical sensing, and spectroscopy by virtue of their features of long photon lifetime as well as strong field confinement and enhancement. Previously, such devices have been mainly studied on silicon integration platforms while the development of high-Q germanium resonators is still in its infancy due to quality limitations of current germanium integration platforms. Compared with silicon, germanium possesses a number of advantages for MIR applications, such as a wider transparency window (2 - 15 µm), a higher refractive index (~4), and a higher third-order nonlinear susceptibility. Here we present our experimental demonstration of two types of MIR high-Q germanium resonators, namely, a microring resonator and a photonic crystal nanobeam cavity. A maximum Q factor of ~57,000 is experimentally realized, which is the highest to date on germanium platforms. Moreover, we demonstrate a monolithic integration of the high-Q germanium resonators with suspended-membrane waveguides and focusing subwavelength grating couplers. Our resonators pave a new avenue for the study of on-chip light-germanium interactions and development of on-chip MIR applications in sensing and spectroscopy.

Germanium (Ge), thanks to its CMOS compatibility and near direct bandgap configuration -140 meV offset between the conduction band states at Gamma and L - has been for long in the race for an all-group-IV laser solution. In the GeSn alloy system, such demonstration has been achieved recently [1]. For Ge, the evidences were much less apparent, in spite of the fact that by applying strain [2], a true direct bandgap configuration is expected and thus the prospect for lasing operation is valid. Here, we explored for the first time the regime where (i) we excite the strained micro bridges at an energy much below the Ge bandgap to reduce the optical loss for modes propagating in the unstrained region of the cavity, (ii) the excitation pulse is 100 ps long, a time shorter than the carrier lifetime of > 5 ns and also shorter than the thermal constant of the suspended bridges but (iii) longer than any thermalization and carrier equilibration times. Under these conditions, using <100> uniaxial loading of strain in the range of 5 %, we obtain unambiguous lasing operation near 3.65 µm at low temperatures with linewidths down to 50 GHz with (a) thresholds at carrier concentration of typically 1E18 cm-3, (b) several orders of magnitude raise of the emission efficiency under lasing and (c) spectrally single mode operation, confirming the expected mode/gain competition behaviour. [1] S. Wirths, R. Geiger, et al. NP 2015;9(2):88-92. [2] M.J. Süess, R. Geiger, et al. NP 2013;7(6):466-472.

We have studied a Si photonics non-mechanical beam steering device for LiDARs. We exploit a doubly periodic Si photonic crystal waveguide (PCW) with a collimator lens, which emits a single-peaked optical beam. Thanks to the slow light effect in the PCW, wide range beam steering can be obtained in the longitudinal direction with maintaining a small beam divergence by a small change of the wavelength and/or index of the PCW. However, due to the symmetric crosssection of the PCW, the emission occurs in both upward and downward directions, which causes a 3-dB loss in the transmission of the optical beam. The downward beam is partly reflected by the substrate, and the reflected beam interferes with the upward beam and modifies the far field pattern, which further increases the loss at particular beam angles. In LiDARs, this loss is repeated at the reception of returned light, resulting in a severe loss penalty. In this study, we investigated the unidirectional upward emission in some PCW structures with vertical asymmetries. We found theoretically that a shallow etched grating on top of the Si layer, which overlaps with the PCW holes significantly increases the upward emission. We fabricated such a device using Si photonics CMOS process and observed 2-8 times stronger upward emission as compared with that of the symmetric PCW. Furthermore, we integrated 32 PCWs in parallel configuration and selected one working PCW so that its relative position against a collimator lens is switched and the beam is steered in the lateral direction. We observed over 400×32 resolution points.

Silicon photonics micro-ring resonator (MRR) and Mach-Zehnder waveguide based sensors have attracted much attention in recent years because of their capacity for high sensitivity, small footprint and mass-scalable (low cost) potential. This type of sensor is based on the detection of changes in optical amplitude/phase due to small changes in local, near-field refractive index (RI) in the environment surrounding the waveguide device. Sensitivity to ever smaller changes in RI are sought, e.g. for vapour/gas based sensing, which may be realised by designing devices based around the slot waveguide. Furthermore, tailoring resonant line-shapes to generate asymmetric (or Fano-like) modes through series, parallel or ‘nested’ arrangements of coupled MRRs also demonstrates the potential for such sensitivity enhancement. This type of device is likely to be of interest, for example where sensing of volatile organic compounds (VOCs) is important, e.g. in industrial process and environmental monitoring. We demonstrate a number of such photonic sensing platforms, combining both the slot waveguide and both established and novel ‘photonic molecule’ structures, fabricated on silicon-on-insulator using standard foundry fabrication processes. Integrated TiN heaters provide the capacity for thermal tuning in order to manipulate the spectral characteristics of our devices and the sensitivity of the devices to a range of VOCs; benzene, toluene and xylene, are investigated as exemplars using a custom-made vapour delivery system. Sensor performance is established with the assistance of device modelling and comparison made with conventional single MRR devices as a reference. The potential of adding functional layers to the devices as a method for achieving chemical selectivity will also be discussed.

Monolithic growth of III-V semiconductors on silicon is a promising path for the development of silicon-based lasers. The GaP binary has a lattice constant very close to that of silicon and can be grown defect free without anti-phase domains (APDs) or stacking faults on (001) exact orientated silicon substrates. These GaP on Si templates provide the base for growth and investigation of III-V lasers. The addition of boron can be used to partially replace Ga and further reduce the lattice constant. This can be balanced to match the lattice constant of silicon by adding As to partially replace P. The alloying also provides control of band gaps and band offsets as well as refractive index. The BxGa(1-x)P and BxGa(1-x)AsyP(1-y) alloys are being explored to provide lattice matching/ strain compensation, cladding and the Separate Confined Heterostructure (SCH). The effects of the inclusion of boron on device related alloy properties have not been studied extensively and are not well understood. We investigate the refractive index and extinction coefficient dispersion relation and the electronic band structure properties of these boron containing alloys using spectroscopic ellipsometry to provide inputs for device modelling and optimisation. Results from the spectroscopic ellipsometry are presented for a series of BGaP and BGaAsP alloy samples with boron fractions in the range 0-6.6% and arsenic fractions from 0-17% on GaP substrates and GaP/ Si templates. These results provide important information for the design of lasers with strong optical and electronic confinement, as shall be discussed.

Silicon photonics is a promising solution for next generation of short-range optical communication systems. Silicon modulators have driven an important research activity over the past years, and many transmission links using on-off keying modulation format (OOK) were successfully demonstrated with a large diversity of modulator structures. In order to keep up with the demand of increasing bitrates for limited bandwidths in Datacom applications, higher modulation formats are explored, such as quadrature phase shift keying (QPSK) or 4-level pulse amplitude modulation (PAM-4). However, driving the modulators to generate PAM-4 signals commonly require expensive and power-hungry electronic devices such as digital-to-analog converters (DACs) for pulse-shaping and digital signal processors (DSP) for nonlinearity compensation. Lastly, new solutions were studied to overcome this issue, including new driving methods based on the use of two different input binary sequences applied directly on the modulator. While most of the reported works are focused on the C-band of communication, the O-band can present a definitive advantage due to the low dispersion of standard single-mode (SSMF) fiber. For those reasons, we demonstrate the generation of a 10-Gbaud DAC-less PAM-4 signal in the O-band using a depletion-based silicon traveling wave Mach-Zehnder modulator (TWMZM). An open eye diagram was obtained, and a bit error rate (BER) of 3.8×10^-3 was measured for a received optical power of about -6 dBm.

Silicon photonic (SiPho) Mach-Zehnder modulator (MZM) working in carrier depletion mode has been demonstrated for its high speed linear response capabilities, promising high bitrate optical transceivers with PAM or QAM modulation formats in Data Center interconnect and communication applications. However, a number of key elements, such as PN junction, phase shifter (PS), traveling waveguide (TW), and termination, have impacts on performances including bandwidth (BW), 𝑉π, optical loss, extinction ratio, etc. Therefore accurate models are needed for design optimization including phase and impedance matches and various trade-offs, which are critical in different high bitrate applications. Modelling high speed SiPho TW modulators is challenging with traditional methods. The high speed response of TW needs electromagnetic (EM) model; but PN junction requires distributed circuit model to align with characterization test. In this paper, we developed a hybrid model with an innovated segmental method, which allows us to combine EM and circuit models to accurately represent TW and PN distributed characterizations for SiPho TW MZM modulator. By using this model, the impacts of critical design parameters are studied. We fabricated MZMs with design optimizations using the model with commercial processes in open foundries. The test results agreed well with the simulation data. 6dB EO BW of 42-56GHz and 43-61GHz without notches and roll-off are achieved at -3V PN bias with 3.5-2mm long PS, by adjusting PN doping levels to achieve 𝑉π * 𝐿 of 12V*mm and 16V*mm, respectively. These MZMs have great potentials for 50- 100Gbaud PAM and QAM optical transceiver applications at 400Gb/s and beyond.

We demonstrate 50 Gbps operation of a compact Si photonic crystal optical modulator, employing meanderline electrodes which compensate phase mismatch between slow light and RF signals. Although low dispersion slow light increases the modulation efficiency, the phase mismatch becomes a limiting factor of the operation speed. The meanderline electrodes and termination resistors broke this limit and enhanced the cut-off frequency up to 31 GHz. The quality of the modulation characteristics was improved at 25, 32 Gbps, and a clear eye opening even observed at 40–50 Gbps.

The depletion-type Si ring modulator (RM) is of great interest among many Si photonic devices for optical interconnect applications because it has a small size, low power consumption, and large modulation bandwidth. Although the major application of the Si RM are digital optical interconnect systems, there is another application of importance, namely microwave photonics in which the modulation linearity is a key performance parameter. We investigate the modulation linearity performance in terms of spurious-free dynamic range (SFDR) of a RM device fabricated by IHP Si PIC foundry. The device has 8-um radius, 290-nm coupling gap and the nominal peak doping concentration of 7×1017 cm−3 for p-region and 3×1018 cm−3 for n-region. The measured SFDR is 78.7 dB·Hz2/3. The major sources of non-linearity of this device are the nonlinear free-carrier plasma dispersion effect in PN junction as well as the nonlinear resonance characteristics. We also perform the numerical simulation of RM SFDR using key device parameters extracted from measurement. The simulation results match well with the measurement results. With this numerical model, we are able to identify the exact cause of RM nonlinearity and come up with suggestions for improving RM linearity.

In this paper a subwavelength waveguide-to-fiber coupler based on two dimensional periodic grating is proposed. In this approach Silicon-On-Insulator (SOI) based structure is employed to couple the radiated mode field inside the core of optical fiber. Rectangular photonic integrated circuit (PIC) waveguide having standard SOI height technology of 220 nm is considered for guiding optical field inside on chip waveguide.

Vernier effect, series-coupled microring resonators (MRRs) are used to extend the free-spectral-range (FSR) of MRRs. In this work we demonstrate integrating two MRRs in a compact Vernier configuration (compact as compared to previously demonstrated Vernier effect devices). our design was realized by using two waveguide crossings to form a major, outer ring that was coupled to a minor ring nested within the major ring. The ratio of the path length of the major ring to the path length of minor ring was 5:2. The spectral response of the device had an FSR of 27.94 nm, a drop port 3 dB bandwidth of 0.87 nm, a minimum extinction ratio of 16.1 dB, a minimum interstitial peak suppression of 10.6 dB, and a footprint of only 540 μm² .

Grating coupler is one of the most basic integrated photonic components. Due to the excellent performance of compact non-mechanical beam steering, it has attracted a lot of research interest. Here we propose a new compound period grating coupler, which can couple the waveguide mode into two radiation modes with different angles by combining two grating structures with different periodicities. Therefore, the extra beam doubles the beam steering range. We numerically demonstrate this idea, and a steering range of 26.20 degree is observed within the wavelength tuning range of 1500 nm to 1600 nm. The compound period grating structure with DBRs (distributed Bragg reflectors) as the substrate has also been demonstrated numerically, and its energy leakage to the substrate is highly suppressed. Furthermore, investigation of fabrication tolerance shows the new structure can be fabricated with the current CMOS technology.

In this paper a subwavelength ring resonator based on array of silicon nanowire is proposed. Silicon-on-Insulator (SOI) based structure is employed to couple resonance wavelength from one waveguide to another waveguide in opposite direction via ring waveguide structure. Silicon nanowire waveguide having standard SOI height technology of 220nm is considered for designing of proposed ring resonator structure. Replacing silicon nanowire technology with conventional integrated optics takes the photonics to a new level, where the cavity between nanowires serves as a nice optical confinement region. The advantage of using silicon nanowire over ring resonator can be seen in the narrowness of the filtered wavelength ranges and hence the sharpness of the filtering process. This structure finds its application as a notch filter to filter out the C-band wavelengths by adjusting waveguide geometrical parameters.

Recently, integrated silicon photonics have attracted much interest in mid-infrared applications for many reasons, such as gas sensing. Unfortunately, traditional silicon on silica waveguides suffer from huge silica absorption losses beyond the wavelength of 3.7 μm, where the absorption fingerprint of many gases, such as carbon dioxide, exist. Also, power leakage from the waveguide core to the substrate is significant for the standard 2 μm-thick buried silica underneath the 220 nm thick silicon layer at such long wavelengths. Therefore, efforts were exerted to find alternative materials, such as sapphire and silicon nitride, which offer lower absorption while keeping a strong refractive index contrast with silicon, and hence small fabrication footprints. In this work, though, we show that an optimized design of single-mode silicon on silica waveguides could push this technology limits deeper into the mid-infrared zone. A challenging wavelength of 4.28 μm, where CO₂ possesses a strong absorption peak, is chosen for this study. We show that the leakage loss can be eliminated using 5 μm and 4 μm thick buried silica layers for 220 nm and 300 nm thick silicon layers, respectively. The penalty of silica absorption is a propagation loss of approximately 6 dB/cm and 4 dB/cm for 220 nm and 300 nm thick silicon layers, respectively. The propagation loss can be further reduced using thicker silicon layers. Fortunately, the scattering loss decreases as the wavelength increases. Therefore, such a mature technology could still play a role in mid-infrared applications.

Integrating an optical receiver in CMOS optimized for near infrared light (NIR) remains appealing but at the same time challenging due to the deep photon penetration depth. A novel implementation of a light detector is demonstrated in a 350 nm CMOS technology, whereby, through adding a majority current with associated electric field distribution in the silicon detection volume, photo-generated minority electrons get quickly guided to the center of this volume. In the center, a tiny PN junction collects the photo-electrons. The detection speed subsequently increases, NIR light is received with improved responsivity and the detector capacitance gets drastically reduced to femtofarad level. The latter improvement also increases signal-to-noise performance and can be used to trade-off with other design parameters to improve global performance of the opto-electronic system. An optical datacom receiver at 1 Gbps is demonstrated at NIR-wavelength for proving useful Current-Assisted Photodiode detector operation in an actual CMOS system

Silicon photonics technology has emerged as a viable solution for the demonstration of highly functional Photonic Integrated Circuits (PICs) relying on the mixture of light sources with silicon based waveguides. However, the incorporation of the laser sources in all PICs has always been at the center of industrial and research attention. To date, the vast majority of such merging schemes focus on either flip chip bonding of external III-V dies or hybrid-integration techniques that feature very good optical performance at the expense of fabrication cost. The next evolution of PICs, however will rely on the monolithic integration of the III-V lasers on the silicon substrates for simultaneous optimization of cost and circuit performance. In this work two low-loss coupling interface schemes are presented for efficient light transition between monolithically integrated InP-based laser sources and a Si₃N₄ passive circuitry through an intermediate waveguiding layer. For both coupling interface schemes, the light is butt-coupled from the III-V source into an intermediate waveguide that in turn couples the light into the final Si₃N₄ waveguide platform utilizing an evanescent coupling scheme. Two approaches are investigated towards this direction: The first approach is based on a purely stoichiometric Si₃N₄ waveguide, while the second one is based on a Si-Rich Nitride (SRN) acting as the intermediate layer. In both cases 2D-FDTD simulations verified by 3D-FDTD simulation results reveal total transition losses of less than 1.7dB for the pure-Si₃N₄ and less than 1dB for the SRN approach.

Interferometers are one of the basic devices in many photonics applications. Interferometers can be used in the design of optical filters, wavelength de-multiplexing (WDM), electro-optical modulators and optical sensors. They can also form the building block of optical digital signal processor (DSP). In this work, we propose novel integrated Michelson interferometer based on the Silicon on Insulator (SOI) technology with 220nm silicon device layer and working in the near infrared region. The Interferometer consists of input splitter directional coupler, two waveguide arms and directional coupler combiner with loop reflector. The interferometer transfer function and its parameters including the free spectral range (FSR), the full width half maximum (FWHM) and sensitivity were derived analytically. Using our proposed interferometer instead of the conventional Mach Zehnder Interferometer (MZI) as optical filter, electro-optical modulator or sensor will reduce the size of the device needed by a factor of two while achieving the same performance. Here, we use our Michelson Interferometer with four different path length differences resulting in FSR from 0.8nm to 6.4nm. A strip waveguide with 500nm width platform is used. These devices are suitable for optical filtering as well as wavelength de-multiplexing WDM applications. The simulation results of the proposed designs are extracted using Lumerical MODE and INTERCONNECT software tools that use scattering matrices of optical components to determine the transfer function of photonic integrated circuits (PICs). The designs were verified with three-dimensional finite-difference-time-domain (3D-FDTD) solver and show good agreement. Finally, the designs were fabricated using Electron Beam Lithography (EBL) and characterized showing also good matching with the numerical simulations results.

Gas sensors have been widely used for different applications including chemical detection, quality assurance, environmental monitoring and medical diagnostics. Optical gas sensors exhibit higher sensitivity and wider dynamic range than their electrical counterparts. This work demonstrates a novel design for a gas sensor based on conventional Silicon-on-insulator (SOI) platform. The sensor design is based on interferometer working in the near-infrared (NIR) region where directional couplers were used in splitting and combining the input power to and from the two arms of the interferometer with 50/50 splitting ratio at 1550nm. Slot-waveguide is used in the sensing arm of the interferometer and strip-to-slot and slot-to-stip converters with high coupling efficiency were used for transforming the optical mode. Finite difference eigenmode (FDE) solver was used to calculate its mode field profiles, effective index, and loss to optimize the waveguide dimensions and to achieve a waveguide sensitivity of 0.7 at 1550nm for 220nm silicon thickness. Three-dimensional finite-dif-ference-time-domain (3D-FDTD) method was used in the analysis and optimization of the proposed gas sensor. Results show significant improvement in the figure-of-merit (FOM) and reduction of device area. The sensor also exhibits low insertion loss (IL) leading to a low detection limit. The proposed sensor is easily fabricated using CMOS technology which is essential for mass-scale fabrication, and thus a low-cost sensor can be integrated with optical fiber communication systems and optoelectronic systems. Therefore, the proposed sensor has the potential to be a key component in lab-on-a-chip (LOC) systems.

In this work, silicon-based plasmonic nanoantennas was realized. Using silicon instead of metals as the material of choice in building such nanoantennas is advantageous as it enables the integration of nanoantennas-based structures into integrated-optoelectronics circuits built using the standard fabrication techniques in the electronic industry. It also allows for low cost mass production of the proposed devices. Upon light incidence on an array of nanoantennas, Localized Surface Plasmon Resonance (LSPR) is generated which causes an enhancement in the localized field inside the structure and in the near field zone. The enhanced localized field is manifested as an enhancement in the absorbed as well as the scattered field. Varying the surrounding material causes variations in the wavelength of the enhancement peak as well as the enhancement level itself. Hence, sensors can be built to facilitate sensing molecules with its characteristic vibrational transitions. In this paper, dipole and bowtie silicon nanoantennas are investigated. It is found that when using silicon with high excess carrier concentrations as the material of choice, the enhancement occur in the mid-IR spectral range which is red shifted compared to the enhancement produced when using metal such as gold or silver. Working in mid-IR is advantageous for sensing applications as the characteristic vibrational transitions of the majority of bio-chemical molecules happens in the mid-IR.

Electro-optical modulator is a key component in data-communication, telecommunication and optical interconnects. In this paper we propose a novel electro-optical modulator design that utilizes Michelson Interferometer based on the widespread Silicon-on-insulator (SOI) technology with 220nm thickness of the silicon device layer. The proposed modulator is working at the telecommunication wavelength 1550nm. Due to its high Pockels coefficient and CMOS compatibility electro-optical polymer (EOP) is used as an active material where its refractive index changes with the applied electric field. The Michelson Interferometer consist of directional couplers which are used in splitting and combining the input power to and from the interferometer arms with 50/50 ratio at 1550nm. Slot waveguide with EOP clad is used in the interferometer arms to achieve high optical field confinement in the EOP which maximizes the mode effective index change of the interferometer arms when applying voltage. Finite Difference Eigen mode (FDE) solver was used to calculate the mode field profiles, effective index and loss of the slot waveguide. By optimizing the waveguide dimensions, we have achieved a waveguide sensitivity S_wg=dn_eff/dn_EOP of 0.9135 at 1550nm. Three-dimensional finite-difference-time-domain (3D-FDTD) method was used in the analysis and optimization of our Michelson Interferometer electro-optical modulator. Results show that our Michelson Interferometer modulator exhibit lower V_πL_π product than previously published SOI based modulators. Moreover, the modulator exhibit low insertion loss (IL) leading to high extinction ratio (ER) in addition to its CMOS compatibility. Thus, our proposed modulator allows for compact, high performance and low cost modulators.

In this work, we present the realization of a novel configuration of dual-coupler nested coupled ring resonator on silicon photonics technology. The waveguide height and width are 220 nm and 500 nm, respectively, surrounded by air from the top and by silicon-oxide on the three other sides. The design assumes TM mode with group refractive index of 3.17 at 1550 nm. The design consists of two mini-racetrack resonators of the smallest lengths of 150.8 and 182.13 μm corresponding to resonance, and bending radius of 25 μm, to minimize bending losses. The directional couplers are designed for a coupling ratio of 97/3. The proposed configuration and the single cavity ring resonator are fabricated using IMEC- ePIXfab passive technology. The measured response shows that the new configuration enhances the finesse by up to 10 times. The proposed high finesse resonator can boost the performance in many applications such as gas sensing, rotation sensing and optical filters.

Mid-infrared (MIR) region is an important region for sensing applications because it contains vibrational resonance for many gases such as methane, carbon monoxide, carbon dioxide, sulfuric acid, ammonia, and acetone. Doped silicon with negative permittivity in MIR region can be used in plasmonic technology to design gas sensors which combining both benefits of silicon and plasmonic technology in MIR region. Fabricating plasmonic integrated devices became easier with current progress in Nanotechnology. Small foot print could be achieved by using Plasmonics technology. Additionally, silicon is CMOS compatible, tunable, and it has high mobility. In this paper we proposed a Fabry-Perot resonator made of doped silicon. Moreover, we studied the response of the Fabry-Perot resonator as a gas sensor in the presence of air, methane and carbon dioxide gases. Consequently, the sensitivity, quality factor and the figure of merit are calculated.

A Mach-Zehnder Interferometer based optical modulator composed of photonic crystals as phase shifters is presented. The switching process relies on the phase modulation along the arms of interferometer resulted from the photonic band shift of the photonic crystal regions. A small bulk index change leads to a large effective index difference between two arms of the interferometer and a small foot-print device can operate as a modulator. Presented MZI-based optical modulator is shown to have tunable bandwidth up to 15.7nm. This interesting approach provides possibility of broadband, low-voltage, low-loss, high-speed and tunable optical integrated switches, modulators and multiplexers.

In optical integrated circuits (OICs), inverted tapered spot-size converters (SSC) provide a high coupling efficiency between a silicon nanowire and an optical fiber. However, to reduce the packaging costs of these OICs, it is beneficial to use a SSC with a mode field diameter that matches to that of a conventional single-mode fiber (SMF). Such a SSC offers high alignment tolerance for butt-coupling with a SMF, without the need for a specialized lensed or tapered fiber. In this work, we propose a novel SSC which is not only broadband but also polarization insensitive. The proposed SSC is composed of a stack of Si₃N₄/SiO₂ layers deposited on top of a silicon nanowire. The most optimal modal overlap of our SSC with a conventional SMF of radius 4.2 μm showed 94% and 99%, for TE and TM polarization, respectively. This multilayer stack is tapered along the propagation direction to transfer power to a tapered silicon nanowire. We have studied the adiabatic transfer of power by optimizing the taper lengths such that the minimum loss can be achieved for both polarizations. Our design demonstrates a record low overall coupling loss (including losses due to mode overlap, tapered design, and reflection) of 0.71 dB (for TE) and 0.48 dB (for TM) between a conventional SMF and the silicon nanowire at the wavelength of 1.55 μm. The variation in coupling loss due to the tapered design is less than 0.6 dB over the wavelength range from 1.5 μm to 1.6 μm.

In this letter, we present a novel optical power splitter having an arbitrary split-ratio that can be tuned over a wide range by employing relatively low voltage levels. It is based on a slotted ring resonator. A 120 nm electro-optic polymer-filled slot is created throughout the circumference of the ring. The hybrid ring resonator is made to work between the full and off resonance states, allowing it to work as a power splitter. This is done by changing the refractive index of the electrooptic polymer inside the slot by the application of an external electric field. The splitter combines the electro-optic functionality of the polymer with the high index contrast of the silicon, resulting in a low tuning voltage power splitter. Over a small voltage range of 0-1 V, it is possible to change the split-ratio of this splitter from 0.031-16.738, making it 10 times better than other competing designs. In addition, it takes less than 500 ps to reconfigure the splitter.

In this paper, we present a new design for an electro-optic modulator ⎯ operating at the telecomm wavelength of 1550 nm and having a very high extinction ratio ⎯ based on photonic crystal (PhC) slab waveguide and phase change material Germanium Selenide (GeSe) embedded in core silicon layer. The device is based on the shifting of the photonic bandgap of the PhC slab waveguide when the refractive index of the GeSe layer changes on application of electric field. Since GeSe changes from its phase crystalline to amorphous on application of an electric field, its refractive index also changes when this phase transition occurs. As a result of a large refractive index contrast between the two phases, the change in the effective refractive index in the PhC slab waveguide is also very high. With two self-sustainable states, the hybrid modulator shows broadband switching capability and an On/Off extinction ratio > 37 dB around a wavelength of 1550 nm.