Search

In this paper, we propose and validate a new approach for increasing the optical bandwidth of waveguide grating couplers by using a center-symmetric grating structure. Our results demonstrate state-of-the-art performance in increasing the 1 dB optical bandwidth to 108 nm and experimentally validated coupling efficiency of −3.0 dB. The fabrication uses standard MPW fabrication service offered by commercial foundry with a minimum feature size above 181 nm. Such a flat wideband and high-efficiency performance is suitable for large-volume manufacturing and will enable the use of grating couplers in wideband applications such as wavelength division multiplexed communications, broadband applications of optical frequency combs and optical sensors.

We describe a novel high coupling efficiency multimode waveguide grating coupler which uses a polysilicon overlay numerically optimized to improve both the directivity and mode-matching of the grating coupler with different modes. The device was realized experimentally and designed with a large minimum feature size of 181 nm for fabrication by deep ultraviolet photolithography in the commercial foundry. Simulations predicted the coupling efficiencies to be −1.1 dB for LP01 mode and −1.56 dB for the LP11 mode. The experimentally measured coupling efficiencies were −1.36 dB and −2.21 dB for LP01 mode and LP11 modes, respectively. These results demonstrate a significant advance over previous designs, and have the potential for use in silicon photonic transceivers for mode division multiplexed communications in multimode fibers.

Novel microring based optical phase-shifters on Si-photonic platforms for microwave photonic applications are demonstrated. By using an add-drop microring structure in combination with an additional multimode interferometer (MMI), pure phase modulation with an ultrahigh modulation efficiency can be realized. Comparison is made to the reference all-pass ring design with both the same radius (5 μm) and coupling coefficient (κ: ∼0.2). The demonstrated structure minimizes the residual amplitude modulation (0.9 vs. 1.8 dB) for nearly 2π phase shifting over a wide optical window (1520 to 1570 nm) at the expense of a larger insertion loss (0.06 dB) due to the additional MMI. The demonstrated phase shifter structure is implemented in a Mach-Zehnder interferometer (MZI) for further study of its dynamic and linearity performance. Under forward bias, it exhibits a low Vπ/Iπ(0.4 V/12.7 mA), fast rise-time (3 ns), low insertion loss (0.2 dB), low residue-amplitude-modulations (<0.9 dB), and can cover wide optical windows (1560–1568 nm) with a small footprint (37 $ \times $ 37 μm2). The measured spurious-free dynamic range (SFDR: 70–75 dB) of our novel phase shifter-based MZI is comparable to that of the traditional traveling-wave MZI (∼75 dB) under reverse bias and tested using the same setup. [ABSTRACT FROM AUTHOR]

We have been developing silicon-on-insulator (SOI) pixel detectors with a pinned depleted diode (PDD) structure, named "XRPIX", for X-ray astronomy. The PDD structure is formed in a thick p-type substrate, to which high negative voltage is applied to make it fully depleted. A pinned p-well is introduced at the backside of the insulator layer to reduce a dark current generation at the Si-SiO 2 interface and to fix the back-gate voltage of the SOI transistors. An n-well is further introduced between the p-well and the substrate to make a potential barrier between them and suppress a leakage current. An optimization study on the n-well dopant concentration is necessary because a higher dopant concentration could result in a higher potential barrier but also in a larger sense-node capacitance leading to a lower spectroscopic performance, and vice versa. Based on a device simulation, we fabricated five candidate chips having different n-well dopant concentrations. We successfully found out the best n-well design, which suppressed a large leakage current and showed satisfactory X-ray spectroscopic performance. Too low and too high n-well dopant concentration chips showed a large leakage current and degraded X-ray spectroscopic performance, respectively. We also found that the dependency of X-ray spectroscopic performance on the n-well dopant concentration can be largely explained by the difference in sense-node capacitance. [ABSTRACT FROM AUTHOR]

This paper addresses the low-frequency noise characterization of Z2-FET structures. These double-gated p-i-n diode devices have been fabricated at STMicroelectronics in an ultrathin body and box (UTBB) 28-nm FDSOI technology and designed to operate as 1T-DRAM memory cells, although other applications, as for example electro static discharge (ESD) protection, have been reported. The experimentally extracted power spectral density of current reveals that the high-diode series resistance, carrier number fluctuations due to oxide traps, and gate leakage current are the main noise contributors at high-current regimes. These mechanisms are expected to contribute to the degradation of cell variability and retention time. Higher flicker noise levels have been reported when increasing the vertical electric field. A simple model considering the contribution of the main noise sources is proposed.

We review high baud rate operation of silicon photonic modulators with emphasis on short-reach and metro-haul applications. The current status of both devices and digital signal processing are discussed. The basic design of a typical carrier-depletion-based dual-drive traveling-wave Mach-Zehnder modulator (TW-MZM) is studied by simulation. Based on an all-silicon TW-MZM, we report the transmission experiments of high baud rate single sideband 4-ary pulse amplitude modulation (PAM-4) signals with linear equalization and artificial neural network based equalization, respectively. Our work indicates all-silicon modulators can address the requirements of future optical communication systems operating at 100 Gbaud. [ABSTRACT FROM AUTHOR]

We present carrier-injection-based photonic switches, engineered for optical pulse distribution with maximum energy efficiency. We apply small-signal analysis and for the first time large-signal modelling to methodically optimize the switches for minimum energy consumption and to classify the electronic contributions from resistance, capacitance, and diode. We present optimized electronic switch activation, which yields a sixfold reduction in energy consumption and we show how static power consumption becomes a negligible factor for optical pulse switching. We demonstrate that with adjusted phase shifter dimensions, MZI-based switches can operate with additional 50% enhanced energy efficiency with down to 4 pJ per switching operation. We show even further efficiency improvement using ring-based designs, allowing an additional improvement of 50% in energy efficiency and we discuss the trade-off between efficiency and optical bandwidth associated to the Q-factor. We benchmark carrier-injection-based switches together with comparable technologies of the silicon photonics platform and identify carrier-injection to be the most suitable technology for pulse switching applications. [ABSTRACT FROM AUTHOR]

Compact and efficient optical antennas are fundamental components for many applications, including high-density fiber-chip coupling and optical phased arrays. Here we present the design of grating-based micro-antennas with perfectly vertical emission in the 300-nm silicon-on-insulator platform. We leverage a methodology combining adjoint optimization and machine learning dimensionality reduction to efficiently map the multiparameter design space of the antennas, analyse a large number of relevant performance metrics, carry out the required multi-objective optimization, and discover high performance designs. Using a one-step apodized grating we achieve a vertical upward diffraction efficiency of almost 92% with a 3.6 $\mu {}$ m-long antenna. When coupled with an ultra-high numerical aperture fiber, the antenna exhibits a coupling efficiency of more than 81% (−0.9 dB) and a 1-dB bandwidth of almost 158 nm. The reflection generated by the perfectly vertical antenna is smaller than −20 dB on a 200-nm bandwidth centered at $\lambda$ = 1550 nm. [ABSTRACT FROM AUTHOR]

We demonstrate a top-down fabricated reconfigurable field effect transistor (RFET) based on a silicon nanowire that can be electrostatically programmed to p- and n-configuration. The device unites a high symmetry of transfer characteristics, high ON/OFF current ratios in both configurations and superior current densities in comparison to other top-down fabricated RFETs. Two NiSi2/Si Schottky junctions are formed inside the wire and gated individually. The narrow omega-gated channel is fabricated by a repeated SiO2 etch and growth sequence and a conformal TiN deposition. The gate and Schottky contact metal work functions and the oxide-induced compressive stress to the Schottky junction are adjusted to result in only factor 1.6 higher p- than n-current for in absolute terms identical gate voltages and identical drain voltages. [ABSTRACT FROM AUTHOR]

We recently reported for the first time 200 Gbit/s/ $\lambda$ net rate data transmission with a single silicon photonic modulator in an intensity modulation direct detection link. Our demonstration relied on an in-house designed, O-band segmented-electrode MZM. In this article, we detail the optimization of the modulator driving scheme, bandwidth/efficiency trade-off and modulation format enabling this result. We use two single-ended RF signals, each having 2.3-Vpp amplitude only, to drive two 1.5-mm modulator segments and transmit data at 80 Gbaud PAM-8 (240 Gbit/s) across 2 km under the 20% soft-decision forward error correction BER threshold of 2E-02. We further decrease the BER using maximum likelihood sequence detection (MLSD), and discuss the related implications. Additionally, we present the system margins regarding modulator drive voltage, received optical power and equalizer taps. At 240 Gbit/s line rate, the modulator energy consumption is only 73 fJ/bit at the SD-FEC threshold. We finally discuss practical implementation considerations for our system. [ABSTRACT FROM AUTHOR]

We demonstrate up to 100 Gbaud quadrature-amplitude modulated (QAM) signal generation with monolithic silicon in-phase quadrature (I/Q)-modulator based on silicon-germanium (SiGe) electro-absorption modulators (EAM). The modulator is in a differentially driven Mach–Zehnder (MZ) interferometric configuration similar to the conventional MZ modulator (MZM) based (I/Q) modulators. In particular, single-polarization quadrature phase-shift keying (SP-QPSK) signaling at symbol rates of 100 Gbaud with BERs below the hard-decision forward error correction (FEC) limit is shown. SP-QPSK optical signal generation and transmission over 80 km single-mode fiber (SMF) are demonstrated at symbol rates of 50 Gbaud with BER of less than $\mathbf {10^{-5}}$. The effect of the frequency roll-off used for Nyquist pulse shaping with raised cosine frequency response is studied for SP-QPSK 50 Gbaud signals. Furthermore, 16QAM signals at symbol rates of 50 Gbaud are generated. [ABSTRACT FROM AUTHOR]

We demonstrate a novel optical phase shifter with a bipolar junction transistor (BJT) type of device structure based on the silicon photonics foundry platform. By operating such a device in saturation mode, we obtain measured output IEC-VEC characteristics very similar to those of an ideal diode, which has a nearly zero turn-on voltage and an extremely small differential resistance. The huge amount of injected current under a small voltage swing (0.1 V) in the operation window makes it possible to obtain significant plasma induced change of the refractive index in the optical waveguide with an extremely small driving-voltage. The device has a small foot-print (500 μm in length), but exhibits small static power consumption for the π phase-shift (Pπ: 4.1 mW), reasonable propagation loss (0.02 dB/μm), small driving-voltage (Vπ: 0.12 V), fast switching time (rise/fall <1.6/1 ns), a residue amplitude modulation (RAM) as small as 0.18 dB, and a very-low power consumption (0.45 mW) during dynamic operation. [ABSTRACT FROM AUTHOR]

The high efficiency on-chip optical phase shifters are required in almost every photo-electronic integration systems. In this work, we present a novel silicon optical circuit for the efficient phase shifiting. A polarization multiplexing scheme is proposed to improve the effective phase tuning length. The circuit is fabricated and in depth analyzed, exhibiting a 3.57 times effective group optical length enhancement compared with the traditional design. Moreover, A thermal-tuning phase shift based on such polarization multiplexing structure was experimentally demonstrated. Within the size of $75 \times 320~\mu \text{m}^{2}$ , a 63.8% power saving for thermal phase shift was measured in the proposed circuit. With considerable efficiency, such circuit is expected to serve as a general underlying structure for the silicon based integrated phase tuning. [ABSTRACT FROM AUTHOR]

Abstract We have been developing event driven X-ray Silicon-On-Insulator (SOI) pixel sensors, called "XRPIX", for the next generation of X-ray astronomy satellites. XRPIX is a monolithic active pixel sensor, fabricated using the SOI CMOS technology, and is equipped with the so-called "Event-Driven readout", which allows reading out only hit pixels by using the trigger circuit implemented in each pixel. The current version of XRPIX has lower spectral performance in the Event-Driven readout mode than in the Frame readout mode, which is due to the interference between the sensor layer and the circuit layer. The interference also lowers the gain. In order to suppress the interference, we developed a new device, "XRPIX6E" equipped with the Pinned Depleted Diode structure. A sufficiently highly-doped buried p-well is formed at the interface between the buried oxide layer and the sensor layer, and acts as a shield layer. XRPIX6E exhibits improved spectral performances both in the Event-Driven readout mode and in the Frame readout mode in comparison to previous devices. The energy resolutions in full width at half maximum at 6.4 keV are 236 ± 1 eV and 335 ± 4 eV in the Frame and Event-Driven readout modes, respectively. There are differences between the readout noise and the spectral performance in the two modes, which suggests that some mechanism still degrades the performance in the Event-Driven readout mode. [ABSTRACT FROM AUTHOR]

We demonstrate a four-channel de-multiplexer based on third-order cascaded elliptical micro-rings on silicon-on-insulator. By applying cascaded elliptical micro-rings structure, the fabrication tolerance was released. The spectral responses of four drop-ports over the S, C, and L bands (1450–1630 nm) were measured, which showed an enhanced performance. The four-channel de-multiplexer has box-like transmission spectra with an ultra-wide 3 dB bandwidth of about 2.6 nm, high out-of-band rejection ratio of more than 40 dB, low-channel cross talk of about −38 dB, small footprint of about 0.0012 mm2, and large operational spectral range of more than 180 nm. The de-multiplexer can support high-data-rate signal transmission capacity. [ABSTRACT FROM AUTHOR]

Photonic integrated circuits with compact size and low power consumption have opened the possibility of realization of ultrahigh-speed and energy-efficient optical computing in an integrated system that may be comparable to CMOS-based electrical integrated circuits in many aspects. Directed logic is an innovative paradigm that can make full use of the advantages of electronics and photonics for optical computing. In this paper, we propose various designs of directed-logic-based electro-optic ripple-carry adders in integrated silicon photonics, which replace the electrical components in the critical path using optical counterparts. All control signals are applied simultaneously through ultralow-power microdisk modulators so that the propagation delay could be reduced significantly. A two-bit thermal-optic full adder based on microdisk modulators is demonstrated as a proof of concept along with a projection of high-speed performance. The proposed electro-optic full adder paves the way to future low-power-consumption and large-bandwidth optical computing in integrated silicon photonics. [ABSTRACT FROM AUTHOR]

Carrier depletion ring modulators offer low power, low footprint, and high modulation speed in the perspective of future silicon photonic interconnects. However, the optical power penalty, which stands for the impact of this device on power budget, is sensitive to the ring power coupling coefficient. In this paper, we derive a relationship between the optical power and this device parameter using a simple analytical model. This model is first described and then compared to ring modulators fabricated in a 300-mm silicon photonic platform through systematic characterization of resonance depth, quality factor, and power penalty. A significant power penalty sensitivity to coupling coefficient variability as high as ∼0.6 dB/% is evaluated from this model and confirmed by measurements. The limitations of this model are also discussed. [ABSTRACT FROM AUTHOR]

Optically transparent polymer waveguides are employed for interfacing silicon photonics devices to fibers. The highly confined optical mode in the nanophotonic silicon waveguide is transferred to a fiber-matched polymer waveguide through adiabatic optical coupling by tapering the silicon waveguide. The polymer waveguides are either processed onto the silicon photonics wafer or bonded to individual chips. Fibers are interfaced to the polymer waveguides through butt-coupling. We show polarization and wavelength-tolerant fiber-to-chip coupling loss of less than 3.5 dB across the O-band. The polymer waveguide-to-silicon-chip alignment tolerance is 2 μm for a loss increase of only 1 dB. Reflection losses are well below −45 dB and the scalability to large numbers of channels is demonstrated. These results open a path to broadband and polarization-tolerant optical packaging of silicon photonics devices for ultrahigh bandwidth applications employing wavelength division multiplexing across multiple channels as envisioned for future data-center interconnects. [ABSTRACT FROM AUTHOR]

Reconfigurable photonic filters are promising candidates to meet the requirements of future microwave communication systems. However, sensitivity to process and temperature variations necessitates an automatic calibration solution to enable robust operation. This paper demonstrates fully automatic tuning of silicon photonic all-pass filter (APF)-based pole/zero filters using a monitor-based tuning method that calibrates the initial response by controlling each pole and zero individually via microheaters. The proposed tuning approach calibrates severely degraded initial responses to the designed elliptic filter shapes and allows us for automatic bandwidth and center-frequency reconfiguration of these filters.This algorithm is demonstrated on second- and fourth-order filters fabricated in a standard silicon photonics foundry process. After the initial calibration, only 300 ms is required to reconfigure a filter to a different center frequency. Thermal crosstalk between the microheaters is investigated, with substrate thinning demonstrated to suppress this effect and reduce filter calibration to less than half of the original thick substrate times. This fully automatic tuning approach opens the possibility of employing silicon photonic filters in real communication systems. [ABSTRACT FROM AUTHOR]

In this paper, a technology for top-down single-gated Schottky barrier transistor is presented exhibiting the highest symmetry of on-currents for n- and p-conductance of such silicon-on-insulator-based devices. The symmetry in the current-voltage-characteristics is a mandatory requirement to realize circuits with reconfigurable nanowire field effect transistors (RFETs) whose channel can be switched electrostatically between n- and p-conductance. It was achieved by an oxidation-induced stressor layer covering the nanowire. Together with the demand for only a single gate potential level, this opens the route to build top-down RFET circuits. Our device features an atomically sharp Schottky junction between intruded nickel silicide and the intrinsic nanowire channel. [ABSTRACT FROM PUBLISHER]

In this paper a silicon-on-insulator nanorib waveguide with reverse-biased p-i-n-junction is used to experimentally demonstrate phase-sensitive amplification based on four-wave mixing for the realization of an in-line four-level phase-regenerator. Phase regeneration of a 14-GBd quadrature phase-shift keying signal is demonstrated for the first time employing a silicon waveguide as phase sensitive amplifier. The regenerator is placed in a 1040-km dispersion-managed link and it is shown that phase noise accumulation in the transmission link can be reduced by the use of the in-line silicon-based phase-regeneration. [ABSTRACT FROM PUBLISHER]

In this paper, we review a variety of hybrid III/V-on-silicon integration platforms with focus on the optical coupling between III/V and silicon waveguides. Numerical simulations with regard to the coupling efficiency are conducted for various mode-overlapped and adiabatic-coupled structures for future design guideline. As a highlight, we show a novel wafer-scale dies-transfer bonding technology that features the merits of high bonding efficiency and process scalability for potential manufacturability of hybrid III/V-on-silicon photonic integrated circuit. Exemplary demonstration of up to 100 dies bonding to an 8-in processed silicon wafer is shown with ∼80% bonding yield, according to the C-mode scanning acoustic microscope characterization. As a proof-of-concept, hybrid silicon mode-locked lasers using the bonded wafer are demonstrated. [ABSTRACT FROM PUBLISHER]

A silicon microring resonator with an embedded Bragg grating for orbital angular momentum (OAM) generation is designed and experimentally demonstrated. The device consists of a 3.9 \mu \textm -radius ring resonator with cylindrical silica holes along the central path and exploits the radially polarized component of the propagating electric field in the resonator’s strip waveguide, enabling both the bandwidth broadening and the emission efficiency improvement, while reducing the footprint. The experimental results show that the device achieves a 3dB bandwidth of 134 GHz and an efficiency up to 6.5%. The interference patterns are measured for different OAM orders in which the number of spirals are clearly visible proving that the device can support different topological charges. Results are in good agreement with simulations. [ABSTRACT FROM PUBLISHER]

We propose and demonstrate a polarization beam splitter (PBS) based on a tapered directional coupler (DC). The tapered DC structure has the advantage of being insensitive to variations of the coupling length and the local coupling coefficient, and thus can significantly increase the bandwidth and polarization extinction ratio (ER). The fabricated PBS with one waveguide width tapered from 0.55 to 0.41 \mu \textm and another tapered reversely in a length of 29 \mu \textm shows a polarization ER above 16 dB and an insertion loss less than 0.4 dB over a wavelength range of 100 nm. The fabrication tolerance of the proposed PBS is also analyzed to be ±50 nm, ensuring the compatibility with the standard 180-nm complementary metal–oxide–semiconductor process. [ABSTRACT FROM PUBLISHER]

We have been developing Kyoto's X-ray astronomy SOI pixel sensors, called “XRPIX”, aiming to extend the frontiers of X-ray astronomy with the wide-band imaging spectroscopy in the 0.5–40 keV band. A dead layer on the X-ray incident surface should ideally be as thin as possible to achieve a high sensitivity below 1 keV, and the depletion layer is required to be thick enough to detect 40 keV X-rays. Thus, we have started developing fully-depleted back-side illuminated (BI) types of XRPIXs. This paper reports on our first two BI devices and their X-ray evaluation (2.6–12 keV). The device named “XRPIX2b-FZ-LA” successfully reaches a full depletion with a thickness of 500 μm. On the other hand, it has a dead layer with a thickness of 1.1–1.5 μm and struggles to achieve the requirement of 1.0 μm. The other device named “XRPIX2b-CZ-PZ”, which is applied with a thin Si sensor-layer and an improved back-side process, is found to satisfy the requirement with its thickness of 0.9–1.0 μm, including Al optical blocking filter of 0.2 μm, although the Si sensor-layer is rather thin with 62 μm. We also describe in this paper the X-ray calibration system that we have built for the X-ray evaluation of XRPIXs. [ABSTRACT FROM AUTHOR]

An array of 200 silicon-on-insulator slot waveguide devices of varying slot widths, ribs widths, taper lengths, and slot lengths were created in each cell of a wafer fabricated at a commercial foundry. The cells were cleaved into individual chips after fabrication. Some chips were coated with thin films of polymers that fully infiltrated the slots. Measurements of spectral loss were made on the grating coupler waveguide devices of both coated and uncoated chips. Individual devices exhibited insertion losses varying from several dB up to values so great that the response was below the noise floor of the optical spectrum analyzer (OSA) employed as a receiver. The chips that failed the transmission test were primarily uncoated ones. Nominally identical devices on different chips exhibited nominally identical behavior. A commercial software program was used to simulate each of the structures that were included in the 200 device test. The simulations were seen to agree quantitatively well with the experimental results and to show a degree of qualitative agreement. Comparison of the experiment and the simulations indicate that the loss inherent in a slot waveguide is quite low. Also near loss free couplers from ridges to slots are achievable. Use of a surface roughness model in comparison with analytical results for slow mode propagation indicates that the excess loss that slots exhibit with respect to a ridge mode counterpart arise almost solely from surface scattering off the surface roughness. The increased loss in the case of the slot guide arises from the higher electromagnetic energy density at the surface of the guide due to the electric field discontinuity that is employed as a guidance mechanism in slot modes in contradistinction to ridge modes that are index-guided. Conclusions include some speculation as to the limits on the loss that can be achieved by variation aware design of slot guides without any improvement in surface roughness over what is now available with fabrication in commercial foundries. [ABSTRACT FROM PUBLISHER]

We numerically examine the opportunities for wideband wavelength conversion through four-wave mixing (FWM) in a foundry-compatible 220\text-\textnm-thick silicon-on-insulator (SOI) waveguide covered with the highly nonlinear two-dimensional material of graphene. As a case study, we consider a foundry-compatible SOI waveguide shaped as a double spiral and covered with two separate graphene sheets, which are covered in turn by two solid polymer electrolyte gates. When combining subwatt-level pump powers with a short waveguide length of only a few hundreds of micrometers, “perfectly phase-matched” conversion with significant efficiencies close to -20\;\textdB can be obtained over a more than 40\text-\textTHz-wide signal band adjacent to the pump frequency. Because of the tunability of the graphene properties using the electrolyte top gates, it is also possible to obtain “quasi-phase matched” FWM conversion through a periodic sign reversal of the graphene third-order nonlinearity along the waveguide. Conversion efficiencies exceeding -30\;\textdB can be achieved over a 3.4\text-\textTHz-wide signal band that is situated as much as 58\;\textTHz away from the pump frequency. Finally, the tunability of the graphene also allows for switching the converter from the perfectly phase-matched to the quasi-phase-matched operation mode. [ABSTRACT FROM PUBLISHER]

The jitter performance of different photonic sampling techniques are investigated. Two basic sampling mechanisms are identified. Depending on analog-to-digital converter (ADC) architectures, the sampling instant is defined either by a time instant at the rising edge of the detected pulses or by their center of mass. A generalized noise model is introduced and used for performance analysis. It is shown that using the center of mass of optical pulses for the sampling process is about one order of magnitude more precise in time than using a time instant at the rising edge. But the increased timing precision of down to few fs is accompanied by a higher amplitude uncertainty decreasing the photonic ADC resolution to $\approx 9$ effective number of bits. [ABSTRACT FROM PUBLISHER]

Phase-sensitive photonic devices fabricated in silicon-on-insulator platform can have their response significantly modified by wafer height fluctuations. In this letter, we demonstrate a novel metrology technique for measuring on-chip thickness variations using long integrated chirped Bragg gratings. Thickness variations are recovered from the measurement of the complex transmission and reflection spectra using an inverse scattering algorithm. With 1-cm long sidewall gratings on 1.2- \mu \textm wide waveguides, we measure waveguide height variations with nanometer-scale precision and tens of micrometer of longitudinal resolution. The results are confirmed by atomic force microscopy measurements. [ABSTRACT FROM PUBLISHER]

This paper reports on fabrication and characterization of Schottky barrier diodes, which were obtained by Schottky contacts between aluminum electrodes and phosphorous-doped single crystalline silicon layer of silicon-on-insulator (SOI) structure. Our photodetectors (PDs) with this structure have advantages, which are simple fabrication process and compatible with complementary metal–oxide–semiconductor devices. The detector performed with the highest external quantum efficiency of 0.397 at 360 nm in ultraviolet spectral range and compared with other devices in previously reported papers. The light sensitivity of 116.2 mV/mW $\cdot \,$ cm2 and stable response time of $\sim 1.3$ ms were achieved at 3 V applied voltage. In addition, operation of an SoI-based $1\times 6$ PD array was demonstrated to verify the availability of ultraviolet line-scanner. [ABSTRACT FROM PUBLISHER]

Recent progresses on hybrid integration of III–V on silicon widely tunable laser technologies are presented for achieving high-performance colorless optical network unit. We demonstrate a symmetric 60-Gb/s aggregate rate wavelength division multiplexing (WDM) passive optical network. The device is tuned on six wavelengths over the $C$ -band. The use of the high output power hybrid silicon transmitter enables up to 30 dB (27 dB) optical budget at BER = 1.10^{-3} (in-band forward error correction assumed) under 2.5 Gb/s (10 Gb/s) direct modulation over 25 and 50 km. Long-reach transmission over 100-km single-mode fiber is also demonstrated. We believe that integrated hybrid III–V/Si devices will have important applications for WDM transmission for both short range and long haul optical transmission. [ABSTRACT FROM PUBLISHER]