Rozprawy doktorskie na temat „Silicon photodetector”

In this paper, the performance simulations of a novel ultra-compact balanced coherent photodetector for operation at a wavelength of 1.5 mu m are presented and design proposals for future fabrication processes are provided. It consists of a compact 2x2 MMI that is evanescently coupled into a germanium MSM photodetection layer. The simulations demonstrate dark current less than 10 nA, capacitance less than 20 fF, and optical bandwidth in the 10-30 GHz range. We propose utilizing the simplicity of direct wafer bonding to bond the detection layer to the output waveguides to avoid complicated epitaxial growth issues. This ultra-compact device shows promise as a high-speed, low-cost integrated silicon photonics solution for the telecommunications infrastructure.

The design, simulation, and initial fabrication of a novel ultra-compact 2x2 silicon multimode-interference device evanescently coupled to a dual germanium metal-semiconductor-metal (MSM) photodetector is presented. For operation at the standard telecom wavelength of 1.5 µm, the simulations demonstrate high-speed operation at 30 GHz, low dark current in the nanoamp range, and external quantum efficiency of 80%. Error analysis was performed for possible tilt error introduced by hybrid integration of the MSM layer on top of the MMI waveguides by use of surface mount technology (SMT) and direct wafer bonding.

Thesis: S.B., Massachusetts Institute of Technology, Department of Physics, 2019
Cataloged from PDF of thesis. "The Table of Contents does not accurately represent the page numbering"--Disclaimer page.
Includes bibliographical references (pages 45-47).
The growing demand for efficient infrared sensors for light ranging, thermal-cameras, and soon, free-space optical communications has yet to be answered. In this study, we use polycrystalline silicon in conjunction with a photonic crystal cavity (PhCC) to enhance light absorption for efficient sensing. We present a cost-effective alternative to the current III-V detectors. By adding a 2D-PhC resonator layer, surface-illuminated light can be confined within a 10 micron region with great intensity, leading to a higher effective path-length and improved detector responsivity. More than 1000 variants of this detector are designed and implemented in a 65nm CMOS process. Using a nearest neighbor method, we find the optimized designs. We validate experimental findings by simulating mode behavior of the PhCC structures using FDTD models. In addition, a numerical study on cavity parameter optimization for achieving high Q-factors and extinction ratios specifically for surface-illumination is presented. We report polysilicon PhCC-enhanced sensors with Q-factors of 6500 resulting in responsivities at 1300nm up to 0.13mA/W -a 25x improvement over non-resonant surface-illuminated Silicon detectors.
by Ebrahim Dakhil Al Johani.
S.B.
S.B. Massachusetts Institute of Technology, Department of Physics

Conventional image sensors achieve color imaging using absorptive organic dye filters. These face considerable challenges however in the trend toward ever higher pixel densities and advanced imaging methods such as multispectral imaging and polarization-resolved imaging. In this dissertation, we investigate the optical properties of vertical silicon nanowires with the goal of image sensor applications. First, we demonstrate a multispectral imaging system that uses a novel filter that consists of vertical silicon nanowires embedded in a transparent medium. Second, we demonstrate pixels consisting of vertical silicon nanowires with integrated photodetectors. We show that their spectral sensitivities are governed by nanowire radius, and perform color imaging. In addition, we demonstrate polarization-resolving photodetectors consisting of silicon nanowires with elliptical cross sections. Finally, we discuss a dual detector device. Each pixel consists of vertical silicon nanowires (incorporating photodetectors) formed above a silicon substrate (that also incorporates a photodetector). Our method is very practical from a manufacturing standpoint because all filter functions are defined at the same time through a single lithography step. In addition, our approach is conceptually different from current filter-based methods, as absorbed light in our device is converted to photocurrent, rather than discarded. This ultimately presents the opportunity for very high photon efficiency.
Engineering and Applied Sciences

During the past twenty years, the photoelectronic industry has grown to incredible proportions, producing devices that have become staples of our every-day life and have completely transformed the way we live. The core technology at the heart of the photoelectronic industry is the “photodetector”. Traditional photodetectors’ performance has dramatically improved after the introduction of nanostructured materials. Among these enhancement materials, plasmonic metal structures are of particular interest due to their peculiar and useful properties, such as optical resonances, high bulk, and surface sensitivities and sub-wavelength light confinement. Plasmonic materials can lead to the development of detectors for novel applications in nano-photonics, biosensing, integrated optics, and lasers. Here, we want to propose an evolution of the common “waveguide detector” design that enables the fabrication of a plasmonic structure, capable of enhancing the absorption efficiency at any desired wavelength, on top of the active area of any kind of traditional semiconductor-based photodetector, including single-photon silicon photomultipliers (SiPMs). Our structure works in a simple tri-layer configuration, metal-dielectric-semiconductor (MIS), by coupling an optical resonance phenomenon with a traditional plasmonic resonance excited on the metal/dielectric interface by means of hole-coupling transmission through an open-slits type of grating. The resulting hybrid opto-plasmonic resonance can drastically enhance the responsivity of a photodetector.

Avec la croissance forte de ces dernières années des objets connectés les technologies de communication optique et radio voient davantage d’opportunités de s’associer et se combiner dans des technologies bas-couts Photoniques-Microondes (MWP). Les réseaux domestiques en sont un exemple. La bande millimétrique notamment, de 57GHz à 67GHz, est utilisé pour contenir les exigences des communications sans fils très haut-débit, néanmoins, la couverture de ces systèmes wireless est limitée en intérieur (indoor) essentiellement à une seule pièce, à la fois du fait de l’atténuation forte de l’atmosphère dans cette bande de fréquence, mais aussi de fait de l’absorption et de la réflexion des murs. Ainsi il nécessaire de déployer une infrastructure pour diffuser l’information au travers d’un système d’antennes distribuées. Les technologiques optiques et photoniques-microondes sont une des solutions envisagées. Les technologies MWP se sont également étendues et couvrent une gamme très large d’applications incluant les communications mobiles 5G, les analyses biomédicales, les communications courtes-distances (datacom), le traitement de signal par voie optique et les interconnexions dans les véhicules et aéronefs. Beaucoup de ces applications requièrent de la rapidité, de la bande-passante et une grande dynamique à la fois, en même temps de demander des dispositifs compacts, légers et à faible consommation. Le cout d’implémentation est de plus un critère essentiel à leur déploiement, en particulier dans l’environnement domestique ainsi que dans d’autres applications variées des technologies MWP.Ce travail de thèse vise ainsi le développement de composants photonique-microondes (MPW) intégrés en technologie BiCMOS ou Bipolaire SiGe/Si, à très bas coût, incluant les phototransistors bipolaires à hétérojonctions (HPT) SiGe/Si, les Diodes Electro-Luminescentes (LED) Si et SiGe, ainsi que l’intégration combinées des composants optoélectroniques et microondes, pour l’ensemble des applications impliquant des courtes longueurs d’ondes (de 750nm à 950nm typiquement).Ces travaux se concentrent ainsi sur les points suivants :La meilleure compréhension de phototransistors SiGe/Si latéraux et verticaux conçus dans une technologie HBT SiGe 80GHz de Telefunken GmbH. Nous traçons des conclusions sur les performances optimales du phototransistor. Les effets de photodétection du substrat et de la dispersion spatiale des flux de porteurs sont analysés expérimentalement. Cette étude aide à développer des règles de dessin pour améliorer les performances fréquentielles du phototransistor HPT pour les applications visées.Dans l’objectif de développer de futures interconnexions intra- et inter- puces, nous concevons des lignes de transmissions faibles-pertes et des guides d’ondes optiques polymères sur Silicium faible résistivité. Il s’agit d’une étape afin d’envisager des plateformes Silicium dans lesquelles les HPT SiGe pourront potentiellement être intégrés de manière performante à très bas coût avec d’autres structures telles que des lasers à émission par la surface (VCSEL), afin de construire un transpondeur optique complet sur une interface Silicium. Le polymère est utilisé comme une interface diélectrique entre les lignes de transmission et le substrat, pour les interconnexions électriques, et pour définir le gain du guide d’onde optique dans les interconnexions optiques.La conception, la fabrication et la caractérisation du premier lien photonique-microonde sur puce Silicium sont menées en se basant sur la même technologie HBT SiGe 80GHz de Telefunken dans la gamme de longueur d’onde 0,65µm-0,85µm. Ce lien optique complétement intégré combine des LEDS Silicium en régime d’avalanche (Si Av LED), des guides d’ondes optiques Nitrure et Silice ainsi qu’un phototransistor SiGe. Un tel dispositif pourrait permettre d’accueillir à l’avenir des communications sur-puce, de systèmes micro-fluidiques et des applications d’analyse biochimiques
With the recent explosive growth of connected objects, for example in Home Area Networks, the wireless and optical communication technologies see more opportunity to merge with low cost MicroWave Photonic (MWP) technologies. Millimeter frequency band from 57GHz to 67GHz is used to accommodate the very high speed wireless data communication requirements. However, the coverage distance of these wireless systems is limited to few meters (10m). The propagation is then limiting to a single room mostly, due to both the high propagation attenuation of signals in this frequency range and to the wall absorption and reflections. Therefore, an infrastructure is needed to lead the signal to the distributed antennas configuration through MWP technology. Moreover, MWP technology has recently extended to address a considerable number of novel applications including 5G mobile communication, biomedical analysis, Datacom, optical signal processing and for interconnection in vehicles and airplanes. Many of these application areas also demand high speed, bandwidth and dynamic range at the same time they require devices that are small, light and low power consuming. Furthermore, implementation cost is a key consideration for the deployment of such MWP systems in home environment and various integrated MWP application.This PhD deals with very cheap, Bipolar or BiCMOS integrated SiGe/Si MWP devices such as SiGe HPTs, Si LEDs and SiGe LEDs, and focused on the combined integration of mm wave and optoelectronic devices for various applications involving short wavelength links (750nm to 950nm).This research focused on the study of the following points:The better understanding of vertical and lateral illuminated SiGe phototransistors designed in a 80 GHz Telefunken GmbH SiGe HBT technology. We draw conclusions on the optimal performances of the phototransistor. The light sensitive Si substrate and two-dimensional carrier flow effects on SiGe phototransistor performance are investigated. This study helps to derive design rules to improve frequency behavior of the HPT for the targeted applications.For future intra /inter chip hybrid interconnections, we design polymer based low loss microwave transmission lines and optical waveguides on low resistive silicon substrate. It is a step to envisage further Silicon based platforms where SiGe HPT could be integrated at ultra-low cost and high performances with other structures such high-speed VCSEL to build up a complete optical transceiver on a Silicon optical interposer. The polymer is used as dielectric interface between the line and the substrate for electrical interconnections and to design the core and cladding of the optical waveguide.The design, fabrication and characterization of the first on-chip microwave photonic links at mid infrared wavelength (0.65-0.85μm) based on 80 GHz Telefunken GmbH SiGe HBT technological processes. The full optical link combines Silicon Avalanche based Light Emitting Devices (Si Av LEDs), silicon nitride based waveguides and SiGe HPT. Such device could permit hosting microfluidic systems, on chip data communication and bio-chemical analysis applications

The increasing demand for efficient energy systems in the last decade has brought about the development of advanced sensor systems that utilize advance detection methods to help in preventive maintenance of these essential systems. These usually are needed in hard to access environments where conditions are extreme and unfit for human interaction. Thin film based sensors deposited directly on the surfaces exposed to harsh environments can serve as ideal means of measuring the temperature of the component during operation. They provide the basic advantage of proximity to the surface and hence accurate measurement of the surface temperature. The low mass size ratio provides the additional advantage of least interference to system operation. The four elements consisting of Si, C, B, and N can be used to form binary, ternary and quaternary compounds like carbides, nitrides, which are chemically and thermally stable with extreme hardness, thermal conductivity and can be doped n- and p-type. Hence these compounds can be potential candidates for high temperature applications. This research is focused on studying sputtering as a candidate to obtain thin SiCBN films. The deposition and characterization of amorphous thin films of silicon boron carbon nitride (SiCBN) is reported. The SiCBN thin films were deposited in a radio frequency (rf) magnetron sputtering system using reactive co-sputtering of silicon carbide (SiC) and boron nitride (BN) targets. Films of different compositions were deposited by varying the ratios of argon and nitrogen gas in the sputtering ambient. Investigation of the oxidation kinetics of these materials was performed to study high temperature compatibility of the material. Surface characterization of the deposited films was performed using X-ray photoelectron spectroscopy and optical profilometry. Studies reveal that the chemical state of the films is highly sensitive to nitrogen flow ratios during sputtering. Surface analysis shows that smooth and uniform SiCBN films can be produced using this technique. Carbon and nitrogen content in the films seem to be sensitive to annealing temperatures. However depth profile studies reveal certain stoichiometric compositions to be stable after high temperature anneal up to 900ºC. Electrical and optical characteristics are also investigated with interesting results. Finally a metal semiconductor metal structure based optoelectronic device is demonstrated with excellent performance improvement over standard silicon based devices under higher temperature conditions.
Ph.D.
School of Electrical Engineering and Computer Science
Engineering and Computer Science
Electrical Engineering PhD

The Silicon Photomultiplier (SiPM) is a novel solid state photon counting detector consisting of a parallel array of avalanche photodiodes biased beyond their breakdown voltage. It has known a fast development in the last few years as a possible alternative to vacuum photomultiplier tubes (PMTs) and conventional avalanche photodiodes (APDs). Indeed, current research in photodetectors is directed toward an increasing miniaturization of the pixel size, thus both improving the spatial resolution and reducing the device dimensions. SiPMs show high photon detection efficiency in the visible and near infrared range, low power consumption, high gain, ruggedness, compact size, excellent single-photon response, fast rise time and reduced sensitivity with temperature, voltage fluctuations, and magnetic fields. Furthermore, solid-state technology owns the typical advantages of the planar integration process, therefore, they can be manufactured at low costs and with high reproducibility. SiPMs performances in photon counting regime have been deeply investigated in literature, using picosecond pulsed lasers. In this regime, they can be used in applications like positron emission tomography, magnetic resonance imaging, nuclear physics instrumentation, high energy physics. An optical characterization performed via continuous wave (CW) sources has seldom been reported even though this kind of excitation seems to be very useful in several fields such as low power measurements, near-infrared spectroscopy and immunoassay tests. In this Thesis, I perform an electrical and optical analysis of two novel classes of SiPMs in the CW regime. After a brief introduction about the SiPM operating principle, parameters and properties (Chapter 1), I describe my responsivity measurements made with an incident optical power down to tenths of picowatts, monitoring the temperature of the device packages, and on a spectrum ranging from ultraviolet to near infrared (Chapter 2). These measurements allowed to define an innovative criterion to establish the conditions necessary for the device to be usable in CW regime. Chapter 3 continues with an investigation of the SiPM signal-to-noise ratio. Measurements employed a 10 Hz equivalent noise bandwidth, around a tunable reference frequency in the range 1 - 100 kHz, and were performed varying the applied bias and the temperature of the SiPM package. These results were compared with similar measurements performed on a PMT. Once the SiPM is characterized, Chapter 4 reports an innovative application: an optical characterization of a class of photonic crystals infiltrated with a new ethanol responsive hydrogel employing the SiPM as a reference photodetector. This activity shows innovative developments for the ethanol sensing to be applied into inexpensive and minimally invasive breathalyzers. Finally, Appendix A shows an electro-optical characterization of a novel class of Silicon Carbide (SiC) vertical Schottky UV detectors. I performed responsivity measurements as a function of the wavelength and the applied bias, varying the temperature of the SiC package, in the 200 - 400 nm range. The results of this work show a new approach to investigate the SiPM capabilities, the CW regime, demonstrating its outstanding performances and innovative applications. This Thesis was made in collaboration with the "Advanced Sensors Development Group" of STMicroelectronics and partially supported by the Project HIGH PROFILE (HIGH-throughput PROduction of FunctIonaL 3D imagEs of the brain), which is funded by the European Community under the ARTEMIS Joint Undertaking scheme.
The Silicon Photomultiplier (SiPM) is a novel solid state photon counting detector consisting of a parallel array of avalanche photodiodes biased beyond their breakdown voltage. It has known a fast development in the last few years as a possible alternative to vacuum photomultiplier tubes (PMTs) and conventional avalanche photodiodes (APDs). Indeed, current research in photodetectors is directed toward an increasing miniaturization of the pixel size, thus both improving the spatial resolution and reducing the device dimensions. SiPMs show high photon detection efficiency in the visible and near infrared range, low power consumption, high gain, ruggedness, compact size, excellent single-photon response, fast rise time and reduced sensitivity with temperature, voltage fluctuations, and magnetic fields. Furthermore, solid-state technology owns the typical advantages of the planar integration process, therefore, they can be manufactured at low costs and with high reproducibility. SiPMs performances in photon counting regime have been deeply investigated in literature, using picosecond pulsed lasers. In this regime, they can be used in applications like positron emission tomography, magnetic resonance imaging, nuclear physics instrumentation, high energy physics. An optical characterization performed via continuous wave (CW) sources has seldom been reported even though this kind of excitation seems to be very useful in several fields such as low power measurements, near-infrared spectroscopy and immunoassay tests. In this Thesis, I perform an electrical and optical analysis of two novel classes of SiPMs in the CW regime. After a brief introduction about the SiPM operating principle, parameters and properties (Chapter 1), I describe my responsivity measurements made with an incident optical power down to tenths of picowatts, monitoring the temperature of the device packages, and on a spectrum ranging from ultraviolet to near infrared (Chapter 2). These measurements allowed to define an innovative criterion to establish the conditions necessary for the device to be usable in CW regime. Chapter 3 continues with an investigation of the SiPM signal-to-noise ratio. Measurements employed a 10 Hz equivalent noise bandwidth, around a tunable reference frequency in the range 1 - 100 kHz, and were performed varying the applied bias and the temperature of the SiPM package. These results were compared with similar measurements performed on a PMT. Once the SiPM is characterized, Chapter 4 reports an innovative application: an optical characterization of a class of photonic crystals infiltrated with a new ethanol responsive hydrogel employing the SiPM as a reference photodetector. This activity shows innovative developments for the ethanol sensing to be applied into inexpensive and minimally invasive breathalyzers. Finally, Appendix A shows an electro-optical characterization of a novel class of Silicon Carbide (SiC) vertical Schottky UV detectors. I performed responsivity measurements as a function of the wavelength and the applied bias, varying the temperature of the SiC package, in the 200 - 400 nm range. The results of this work show a new approach to investigate the SiPM capabilities, the CW regime, demonstrating its outstanding performances and innovative applications. This Thesis was made in collaboration with the "Advanced Sensors Development Group" of STMicroelectronics and partially supported by the Project HIGH PROFILE (HIGH-throughput PROduction of FunctIonaL 3D imagEs of the brain), which is funded by the European Community under the ARTEMIS Joint Undertaking scheme.

La diffusion électromagnétique par des surfaces rugueuses concerne un ensemble très vaste de problèmes actuels en optronique (maîtrise des signatures/cloaking, analyse des signatures laser et infrarouge, imagerie active, localisation de la lumière, imagerie optique haute résolution, modélisation des interactions lumière matière et des signatures optiques, applications photovoltaïques et détecteurs infrarouges, biotechnologie). Les travaux réalisés dans le cadre de la thèse "Contrôle de la lumière par des éléments de surface désordonnés'' consistent à réaliser expérimentalement des surfaces rugueuses dont les propriétés statistiques sont contrôlées. Pour cela, un banc expérimental de photofabrication a été développé, utilisant sles propriétés statistiques des figures de speckle issues d'un faisceau laser mis en forme spatialement.Les surfaces réalisées présentent ainsi des propriétés statistiques qui n'existent pas à l'état naturel (fonction d'autocorrélation non gaussienne). Au delà de ces surfaces photofabriquées, le travaux de cette thèse s'intéressent également aux propriétés de surfaces de silicium fortement rugueuses caractérisées de Black Silicon
Scattering of electromagnetic waves from rough surfaces is involved in a wide area of research in optronics (cloaking, laser and infrared signature analysis, active imaging, light localisation, high resolution optical imaging, laser-matter interaction and optical signature modelling, photovoltaics, infrared sensors, biotechnologies). Studies performed during the thesis ``Ligth control by random surface elements'' consist in the experimental fabrication of rough surfaces which statistical properties are controled. An experimental setup has been implemented, using the properties of a spatially shaped laser speckle pattern. The photofabricated surfaces show statistical properties that do not exist in nature such as non Gaussian autocorrelation function. Beyond these photofabricated surfaces, we also studied very rough surfaces of semi-conductor known as Black Silicon

In this study, detailed characteristics and performance assessment of 3&
#8722
5 µ
m p-i-n InSb photodetectors on Si substrates are reported. The detector epilayers were grown on GaAs coated Si substrates by molecular beam apitaxy (MBE). Both homojunction and single heterojunction (AlInSb/InSb) detector structures were investigated. Arrays of 33x33 µ
m2 detectors were fabricated and flip-chip bonded to a test substrate for detailed electrical and optical characterization. A peak detectivity as high as 1x1010 cmHz1/2/W was achieved with InSb homojunction detectors on Si substrate in spite of the large lattice mismatch between InSb and Si (%19). In both homojunction and single heterojunction structures the differential resistance is significantly degraded by trap assisted tunneling (TAT) under moderately large reverse bias and by ohmic leakage near zero-bias. While the heterojunction structures provide a higher 80 K zero bias differential resistance, the responsivity of this structure is significantly lower than that of homojunction InSb photodiodes. In both homojunction and heterojunction photodetectors, 80K 1/f noise is dominated by TAT processes, and the noise current at 1 Hz follows the empirical relation in= &
#945
TAT(ITAT) &
#946
with &
#945
TAT&
#8764
1.1x10&
#8211
6 and &
#946
&
#8764
0.53.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2012.
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (p. [107]-115).
The challenge of overcoming energy efficiency and bandwidth limitations in interconnects between components in computer systems (e.g. between memory and processors) has motivated the development of short-range optical interconnects, which in many approaches require optical devices and waveguides fabricated within the same CMOS environments as the electronics. This thesis centers on developing photodetectors for infrared light within the silicon of commercial CMOS processes; silicon's lack of strong absorption at the wavelengths of interest makes this challenging. The approach uses defect-state mediated linear absorption and two-photon absorption (TPA) in small mode-volume resonators to generate photocarriers. Such resonators allow efficient linear absorption in short devices despite low absorption coefficients, and a greater TPA rate than in bulk material due to the large energy densities achievable. The devices here are made in the polysilicon layer of a commercial DRAM process, and characterization of this material, different from crystalline Si in both its linear and nonlinear absorption, forms a starting point. The design, fabrication, and testing of electrically addressable photonic crystal resonators subject to the constraints associated with working in a CMOS process are then presented. The best resonators made were able to reach Qs of 70,000, limited by linear loss in the polysilicon. Linear absorption is dominant in the devices made to date, and allowed quantum efficiencies of a few tens of percent on resonance. However, high biases of around -20 V were required to achieve these QEs, and the bandwidth of the devices was limited to only approximately 500 MHz. Improvements to the electrical structure of the devices are likely to improve these characteristics. The ability to fabricate high-Q photonic crystal resonators within full CMOS flows, and the QEs allowed by defect-assisted absorption in the devices measured, indicate promise for this approach to photodetection in integrated CMOS photonic systems.
by Karan K. Mehta.
S.M.

This thesis presents the development of integrated silicon photonic devices. These devices are compatible with the present and near future CMOS technology. High-khorizontal grating couplers and waveguides are proposed. This work consists of simulations and device design, as well as the layout for the fabrication process, device fabrication, process development, characterization instrument development and electro-optical characterizations. The work demonstrates an alternative solution to costly silicon-on-insulator photonics. The proposed solution uses bulk silicon wafers and thin film deposited waveguides. Back-end deposited horizontal slot grating couplers and waveguides are realized by multi-layers of amorphous silicon and high-k materials. The achievements of this work include: A theoretical study of fully etched slot grating couplers with Al2O3, HfO2 and AIN, an optical study of the high-k films with spectroscopic ellipsometry, an experimental demonstration of fully etched SiO2 single slot grating couplers and double slot Al2O3 grating couplers, a practical demonstration of horizontal double slot high-k waveguides, partially etched Al2O3 single slot grating couplers, a study of a scheme for integration of the double slot Al2O3  waveguides with selectively grown germanium PIN photodetectors, realization of test chips for the integrated germanium photodetectors, and study of integration with graphene photodetectors through embedding the graphene into a high-k slot layer. From an application point of view, these high-k slot waveguides add more functionality to the current silicon photonics. The presented devices can be used for low cost photonics applications. Also alternative optical materials can be used in the context of this photonics platform. With the robust design, the grating couplers result in improved yield and a more cost effective solution is realized for integration of the waveguides with the germanium and graphene photodetectors.

QC 20141114

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 1998.
Includes bibliographical references (leaves 129-132).
Strain-balanced silicon-germanium superlattices grown on high quality compositionally graded buffers, or virtual substrates. make a complete range of alloy composition and biaxial strain combinations accessible. This structure is a unique way to achieve high quantum efficiency near IR photodetection for silicon-based optical interconnects. The growth of the strain-balanced superlattices by molecular beam epitaxy (MBE) and ultra high vacuum chemical vapor deposition (UHV-CVD) is presented and the role of adsorbed hydrogen during UHV-CVD growth is addressed. Hydrogen adsorption 0,1 the growth surface proved a useful technique to minimize coherent strain relaxation at the higher growth temperatures required for UHV-CVD silicon-germanium growth. The near IR absorption spectrum of the strained silicon-germanium materials possible using strain-balanced superlattices is critically required in the design of a photodetector. A model based on deformation potential theory and semiconductor absorption physics is used to predict the absorption coefficient as a function of strain and alloy composition. Photocurrent junction spectroscopy of strain-balanced silicon­germanium materials is used to confirm the results of the model. The effects of threading dislocations associated with the compositionally graded buffers on the bulk leakage current of photodiodes is determined using electron-beam induced current imaging techniques to measure dislocation density. The correlation between dislocation density and leakage current yielded a current per dislocation line length of 200 pA [mu]m·1. Coupling strategies for the integration of high dielectric contrast polycrystalline silicon/ Si02 strip waveguides and silicon-germanium photodetectors are presented. The high optical power densities possible with the polycrystalline silicon waveguides permits the miniaturization of photodetectors. The effects of integration and miniaturization on photodetector performance are discussed.
by Laura Marie Giovane.
Ph.D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 2010.
Cataloged from PDF version of thesis.
Includes bibliographical references (p. 162-172).
Multispectral infrared (IR) detection has been widely employed for numerous applications including hyperspectral imaging, IR spectroscopy, and target identification. Traditional multispectral detection technology is based on the combination of broadband focal plane arrays (FPA) and spectral filters, grating spectrometers, or Fourier transform spectrometers, which requires bulky, high-cost mechanical scanning instruments and have a slow response. Hybrid structures integrating FPA and silicon readout integrated circuits (Si ROIC) greatly limit the yield and result in extremely expensive devices. Single photodetectors capable of detecting multiple wavebands simultaneously and monolithic integration with Si ROIC, however, enable dramatically simplified system design with superior mechanical robustness, and thus attract a lot of interest around the world today. In this thesis, we focus on the development of novel IR sensitive material and resonant-cavity- enhanced (RCE) photodetector devices that address the emerging need in the field of IR radiation detection. Polycrystalline PbTe films have been identified as the IR absorbing layers due to their high photosensitivity and fabrication flexibility; on the device side, we have established a universal design theory for multispectral detection and demonstrated fully functional mid-IR RCE photodetectors capable of monolithic integration with Si ROIC. We have developed room-temperature-sensitized, polycrystalline PbTe films using single source thermal evaporation for detecting IR light up to 5 pm in wavelength. Thinner PbTe layers yields enhanced performance than thicker layers due to strong thickness dependence of both photo-responsivity and detectivity. Structural, electrical, and optical property studies reveal photoconductivity mechanism in the films and point out directions of further material optimization. We have established a versatile and scalable design theory for cavity-enhanced multispectral photodetectors using phase-tuned propagation. Critical coupling condition is identified as the prerequisite to achieve near unity quantum efficiency in RCE photodetectors. Coupling-matching layers are positioned between cascaded planar resonant cavities for controlling optical phase and coupling strength between incident light and resonant modes to obtain critical coupling condition. After developing another two IR transparent layers as low and high index materials, evaporated As2S3 and sputtered Ge, we design and fabricate distributed Bragg reflectors (DBR) for mid-IR resonant cavities. In our design example of dual waveband RCE photodetectors, peak quantum efficiencies over 80% have been realized in both wavebands (1.55 pm and 3.6 pm) with only 50 nm and 100 nm thick PbTe IR absorbers, and spectral cross talk as low as 0.1% is obtained. Preliminary results on our first attempt of fabricated dual waveband RCE photodetectors demonstrate the two resonant cavity modes at 1.61 pm and 3.70 pm. And quantum efficiencies as high as 92% and 68% have been achieved in two wavebands respectively. We have developed and optimized a multi-step lift-off patterning technique to fabricate RCE photodetectors on a Si platform. Single waveband RCE photodetectors for mid-IR (3.5 pm) have been designed according to critical coupling condition to achieve near unity quantum efficiency. The fabricated devices show high quantum efficiency (90%) and peak responsivity at the resonant wavelength of 3.5 pm, which is 13.4 times higher than blanket PbTe film of the same thickness. We demonstrate detectivity as high as 0.72x0 cmHzmW~l, comparable with commercial polycrystalline mid-IR photodetectors. As low temperature processing (150 'C) is accomplished in the entire fabrication process, this demonstration paves the way for monolithic integration of RCE photodetectors with Si ROIC. Lastly, for the first time, we fabricate and test integrated devices of single waveband (3.6 pm) RCE photodetectors and Si ROIC. Both hybrid and monolithic integration structures are investigated. We have developed the fabrication process to accommodate Si ROIC chips of only 3 mm x 5 mm in area, and successfully integrated RCE photodetectors on Si ROIC directly. Our preliminary results show high promise for monolithic integration of RCE photodetectors and Si ROIC in the future.
by Jianfei Wang.
Ph.D.

The predictable quantum efficient detector (PQED) is a cheaper, more practical alternative to the current radiometric primary standard, the cryogenic radiometer. The PQED is made of an induced diode, a rectifying junction based on a positively charged dielectric film which induces an n-type inversion layer on a p-type silicon substrate. Increasing the fixed charge Qf in the dielectric has been theoretically predicted to improve the quantum efficiency of the diode by decreasing the surface recombination velocity (SRV) at the dielectric-silicon interface, as well as improving the performance of the diode at high intensities. In this work, we purpose the replacement of silicon oxide (SiOx) with silicon nitride (SiNx) as the inversion-inducing dielectric layer as a means of increasing Qf.Amorphous PECVD SiNx has been shown to have a significantly higher Qf than thermally grown SiOx. A sixfold increase of Qf in SiNx by means of charge injection via bias soaking has also been demonstrated. The dielectric ideality of the nitride increases with the atomic nitrogen concentration x, and a high-x film has shown superior charge retention properties. The absorption in the high-x film is far less than 1% in the applicable wavelength ranges, but is larger in the low-x film. The impact of Qf on the SRV has been experimentally demonstrated, and may be fit to the extended SRH theory with the inclusion of a sub-surface damage term. Finally, it has been demonstrated that the aluminium contacts deposited for the purpose of charging the SiNx may be chemically removed while the injected charge remains. SiNx has thus been shown to be an interesting candidate for the PQED application.

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Materials Science and Engineering, 1998.
Includes bibliographical references (leaves 137-147).
Silicon germanium layers on silicon substrates (SiGe/Si) are useful for a variety of microelectronics applications. The most successful technique for growing low defect density relaxed SiGe/Si layers is relaxed compositional grading. However, with increasing Ge content in the graded SiGe layers, the following materials concerns need to be addressed-{i) a high surface roughness, (ii) the formation of dislocation pile-ups, (iii) an increase in the threading dislocation density, (iv) tensile strains and micro-cracking due to thermal mismatch, and (v) particulate contamination from germane gas phase nucleation. We have grown relaxed graded SiGe/Si layers using ultra-high vacuum chemical vapor deposition (UHVCVD) at growth temperatures ranging between 500°-900° C and pressures between 30-500 millitorr. The SiGe growth rates at different temperature and Ge content regimes agree with previously proposed theories. By applying both a dislocation blocking criterion and surface roughness effects to graded SiGe/Si structures, we have proposed a model to explain and predict the formation of dislocation pile-ups in graded structures. We have discovered that there is a substantial improvement in the surface roughness and dislocation pile-up density in graded SiGe layers by growing on miscut Si(00l) substrates. It was found that the array of 60° dislocations that usually forms to relieve the misfit stress could transform into a novel lower energy hexagonal dislocation network consisting of all edge dislocations. High resolution X-ray diffraction measurements indicate that there is a decrease in the rate of epilayer tilting in the Ge-rich layers of the graded buffer in agreement with the observed dislocation structure. We have designed an optimized relaxed buffer (ORB) process that allows us to grow high quality Ge layers on Si substrates. By employing a chemical mechanical polishing (CMP) and regrowth step within the epitaxial structure, we have minimized the formation of dislocation pile-ups. Compressive strain has been incorporated into the graded layers to overcome the thermal mismatch problem. The ORB process eliminates dislocation pile-ups, decreases gas-phase nucleation of particles, and eliminates the increase in threading dislocation density. Germanium p-n photodiodes were fabricated to assess the diectronic quality and prove the feasibility of a high quality infrared detector on a Si substrate. The dark current in the diodes was at least two orders of magnitude lower than any previously reported value for Ge photodiodes on Si substrates. Capacitance measurements indicate that the devices are capable of high speed operation. Dislocation filtering experiments were conducted to reduce threading dislocation densities on pattered mesas. It was found that at the strain levels from small compressive mismatch, dislocation nucleation from the mesa edges dominates over dislocation filtering.
by Srikanth B. Samavedam.
Ph.D.

Thesis (M.S. in Combat Systems Technology)--Naval Postgraduate School, Dec. 2004.
Thesis Advisor(s): Gamani Karunasiri, Richard C. Olsen. Includes bibliographical references (p. 57-60). Also available online.

[EN] The processing of radiofrequency signals using photonics means is a discipline that appeared almost at the same time as the laser and the optical fibre. Photonics offers the capability of managing broadband radiofrequency (RF) signals thanks to its low transmission attenuation, a variety of linear and non-linear phenomena and, recently, the potential to implement integrated photonic subsystems. These features open the door for the implementation of multiple functionalities including optical transportation, up and down frequency conversion, optical RF filtering, signal multiplexing, de-multiplexing, routing and switching, optical sampling, tone generation, delay control, beamforming and photonic generation of digital modulations, and even a combination of several of these functionalities. This thesis is focused on the application of vector processing in the optical domain to radiofrequency signals in two fields of application: optical beamforming, and photonic vector modulation and demodulation of digital quadrature amplitude modulations. The photonic vector control enables to adjust the amplitude and phase of the radiofrequency signals in the optical domain, which is the fundamental processing that is required in different applications such as beamforming networks for direct radiating array (DRA) antennas and multilevel quadrature modulation. The work described in this thesis include different techniques for implementing a photonic version of beamforming networks for direct radiating arrays (DRA) known as optical beamforming networks (OBFN), with the objectives of providing a precise control in terrestrial applications of broadband signals at very high frequencies above 40 GHz in communication antennas, optimizing the size and mass when compared with the electrical counterparts in space application, and presenting new photonic-based OBFN functionalities. Thus, two families of OBFNs are studied: fibre-based true time delay architectures and integrated networks. The first allow the control of broadband signals using dispersive optical fibres with wavelength division multiplexing techniques and advanced functionalities such as direction of arrival estimation in receiving architectures. In the second, passive OBFNs based on monolithically-integrated Optical Butler Matrices are studied, including an ultra-compact solution using optical heterodyne techniques in silicon-on-insulator (SOI) material, and an alternative implementing a homodyne counterpart in germanium doped silica material. In this thesis, the application of photonic vector processing to the generation of quadrature digital modulations has also been investigated. Multilevel modulations are based on encoding digital information in discrete states of phase and amplitude of an electrical signal to enhance spectral efficiency, as for instance, in quadrature modulation. The signal process required for generating and demodulating this kind of signals involves vector processing (phase and amplitude control) and frequency conversion. Unlike the common electronic or digital implementation, in this thesis, different photonic based signal processing techniques are studied to produce digital modulation (photonic vector modulation, PVM) and demodulation (PVdM). These techniques are of particular interest in the case of broadband signals where the data rate required to be managed is in the order of gigabit per second, for applications like wireless backhauling of metro optical networks (known as fibre-to-the-air). The techniques described use optical dispersion in optical fibres, wavelength division multiplexing and photonic up/down conversion. Additionally, an optical heterodyne solution implemented monolithically in a photonic integrated circuit (PIC) is also described.
[ES] El procesamiento de señales de radiofrecuencia (RF) utilizando medios fotónicos es una disciplina que apareció casi al mismo tiempo que el láser y la fibra óptica. La fotónica ofrece la capacidad de manipular señales de radiofrecuencia de banda ancha, una baja atenuación, procesados basados en una amplia variedad de fenómenos lineales y no lineales y, recientemente, el potencial para implementar subsistemas fotónicos integrados. Estas características ofrecen un gran potencial para la implementación de múltiples funcionalidades incluyendo transporte óptico, conversión de frecuencia, filtrado óptico de RF, multiplexación y demultiplexación de señales, encaminamiento y conmutación, muestreo óptico, generación de tonos, líneas de retardo, conformación de haz en agrupaciones de antenas o generación fotónica de modulaciones digitales, e incluso una combinación de varias de estas funcionalidades. Esta tesis se centra en la aplicación del procesamiento vectorial en el dominio óptico de señales de radiofrecuencia en dos campos de aplicación: la conformación óptica de haces y la modulación y demodulación vectorial fotónica de señales digitales en cuadratura. El control fotónico vectorial permite manipular la amplitud y fase de las señales de radiofrecuencia en el dominio óptico, que es el procesamiento fundamental que se requiere en diferentes aplicaciones tales como las redes de conformación de haces para agrupaciones de antenas y en la modulación en cuadratura. El trabajo descrito en esta tesis incluye diferentes técnicas para implementar una versión fotónica de las redes de conformación de haces de en agrupaciones de antenas, conocidas como redes ópticas de conformación de haces (OBFN). Se estudian dos familias de redes: arquitecturas de retardo en fibra óptica y arquitecturas integradas. Las primeras permiten el control de señales de banda ancha utilizando fibras ópticas dispersivas con técnicas de multiplexado por división de longitud de onda y funcionalidades avanzadas tales como la estimación del ángulo de llegada de la señal en la antena receptora. En la segunda, se estudian redes de conformación pasivas basadas en Matrices de Butler ópticas integradas, incluyendo una solución ultra-compacta utilizando técnicas ópticas heterodinas en silicio sobre aislante (SOI), y una alternativa homodina en sílice dopado con germanio. En esta tesis, también se han investigado técnicas de procesado vectorial fotónico para la generación de modulaciones digitales en cuadratura. Las modulaciones multinivel codifican la información digital en estados discretos de fase y amplitud de una señal eléctrica para aumentar su eficiencia espectral, como por ejemplo la modulación en cuadratura. El procesado necesario para generar y demodular este tipo de señales implica el procesamiento vectorial (control de amplitud y fase) y la conversión de frecuencia. A diferencia de la implementación electrónica o digital convencional, en esta tesis se estudian diferentes técnicas de procesado fotónico tanto para la generación de modulaciones digitales (modulación vectorial fotónica, PVM) como para su demodulación (PVdM). Esto es de particular interés en el caso de señales de banda ancha, donde la velocidad de datos requerida es del orden de gigabits por segundo, para aplicaciones como backhaul inalámbrico de redes ópticas metropolitanas (conocida como fibra hasta el aire). Las técnicas descritas se basan en explotar la dispersión cromática de la fibra óptica, la multiplexación por división de longitud de onda y la conversión en frecuencia. Además, se presenta una solución heterodina implementada monolíticamente en un circuito integrado fotónico (PIC).
[CAT] El processament de senyals de radiofreqüència (RF) utilitzant mitjans fotònics és una disciplina que va aparèixer gairebé al mateix temps que el làser i la fibra òptica. La fotònica ofereix la capacitat de manipular senyals de radiofreqüència de banda ampla, una baixa atenuació, processats basats en una àmplia varietat de fenòmens lineals i no lineals i, recentment, el potencial per implementar subsistemes fotònics integrats. Aquestes característiques ofereixen un gran potencial per a la implementació de múltiples funcionalitats incloent transport òptic, conversió de freqüència, filtrat òptic de RF, multiplexació i demultiplexació de senyals, encaminament i commutació, mostreig òptic, generació de tons, línies de retard, conformació de feix en agrupacions d'antenes i la generació fotònica de modulacions digitals, i fins i tot una combinació de diverses d'aquestes funcionalitats. Aquesta tesi es centra en l'aplicació del processament vectorial en el domini òptic de senyals de radiofreqüència en dos camps d'aplicació: la conformació òptica de feixos i la modulació i demodulació vectorial fotònica de senyals digitals en quadratura. El control fotònic vectorial permet manipular l'amplitud i la fase dels senyals de radiofreqüència en el domini òptic, que és el processament fonamental que es requereix en diferents aplicacions com ara les xarxes de conformació de feixos per agrupacions d'antenes i en modulació multinivell. El treball descrit en aquesta tesi inclou diferents tècniques per implementar una versió fotònica de les xarxes de conformació de feixos en agrupacions d'antenes, conegudes com a xarxes òptiques de conformació de feixos (OBFN), amb els objectius de proporcionar un control precís en aplicacions terrestres de senyals de banda ampla a freqüències molt altes per sobre de 40 GHz en antenes de comunicacions, optimitzant la mida i el pes quan es compara amb els homòlegs elèctrics en aplicacions espacials, i la presentació de noves funcionalitats fotòniques per agrupacions d'antenes. Per tant, s'estudien dues famílies de OBFNs: arquitectures de retard en fibra òptica i arquitectures integrades. Les primeres permeten el control de senyals de banda ampla utilitzant fibres òptiques dispersives amb tècniques de multiplexació per divisió en longitud d'ona i funcionalitats avançades com ara l'estimació de l'angle d'arribada del senyal a l'antena receptora. A la segona, s'estudien xarxes de conformació passives basades en Matrius de Butler òptiques en fotònica integrada, incloent una solució ultra-compacta utilitzant tècniques òptiques heterodinas en silici sobre aïllant (SOI), i una alternativa homodina en sílice dopat amb germani. D'altra banda, també s'ha investigat en aquesta tesi tècniques de processament vectorial fotònic per a la generació de modulacions digitals en quadratura. Les modulacions multinivell codifiquen la informació digital en estats discrets de fase i amplitud d'un senyal elèctric per augmentar la seva eficiència espectral, com ara la modulació en quadratura. El processat necessari per generar i desmodular aquest tipus de senyals implica el processament vectorial (control d'amplitud i fase) i la conversió de freqüència. A diferència de la implementació electrònica o digital convencional, en aquesta tesi s'estudien diferents tècniques de processament fotònic tant per a la generació de modulacions digitals (modulació vectorial fotònica, PVM) com per la seva demodulació (PVdM). Això és de particular interès en el cas de senyals de banda ampla, on la velocitat de dades requerida és de l'ordre de gigabits per segon, per a aplicacions com backhaul sense fils de xarxes òptiques metropolitanes (coneguda com fibra fins l'aire). Les tècniques descrites es basen en explotar la dispersió cromàtica de la fibra òptica, la multiplexació per divisió en longitud d'ona i la conversió en freqüència. A més, es prese
Piqueras Ruipérez, MÁ. (2016). Photonic Vector Processing Techniques for Radiofrequency Signals [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/63264
TESIS

碩士
國立中山大學
電機工程研究所
84
A high sensitivity porous silicon (PS) photodetector is fabricated through rapid thermal oxidation (RTO) and ra- pid thermal annealing (RTA) processes. A RTO process is used to passivate the surface of PS and therefore to en- hance the photoresponsivity of the PS layer. A RTA proc- ess is employed to improve the quality of the oxide layer on the surface of PS layer and hence to reduce the dark current of the PS photodetector. Under our optimum prep- aration conditions, photocurrent can reach 21 mA(under 22.92mW tungsten lamp illumination) and dark current is about 5.4 uA(at the reverse bias of 10V). The quantum e- fficiencies of above 90% at the wavelength shorter than 750nm and of 80% to 70% in the range of wavelength from 750nm to 1050nm are obtained. It is demonstrated that a RTA-RTO-PS photodetector behaves as a blue- enhanced metal -i-p photodetector. Of utmost importance is the incidence of photoresponsivity at the wavelength longer than 1600nm. We attribute it to the absorption of photons with the help of band offset at the natural heterointerface between PS and Si.

Driven by the exponential growth of data tra c, current infrastructure and standards are evolved to meeting the requirements. In long haul communication, 1550/1310 nm based singlemode ber technology is a commercially viable platform. For short-reach optical interconnects for rack-to-rack communication and within buildings, the matured 850 nm VCSEL based multimode ber (MMF) technology is an industry-standard. IEEE has recently proposed a 400 Gbps roadmap for data centers to scale up short-reach infrastructure. However, the current shortreach datacom infrastructure is not scalable to support a 400 Gbps data rate. Integration of all the functional components of an optical interconnect on a single platform can meet the requirement of a scalable, energy-e cient, and a ordable system. Additionally, CMOS compatibility can leverage electronic and photonic circuits' co-existence on a single platform and low-cost mass manufacturing. Integrating optical functionalities on a single-chip also o ers application in sensing as well. Due to the low absorption in water, the 850 nm wavelength window is also attractive for realizing Lab-on-a-chip biosensors. Integrated photonic circuits at 850 nm band can therefore be useful for a lab-on-a-chip biosensor platform as well. This thesis presents an integrated photonic platform comprising silicon nitride (SiN) waveguide, SiN surface grating coupler, silicon photodetector, and wavelength lters integrated monolithically on the SiN-on-SOI platform at 850 nm wavelength. Our primary focus is to overcome the limitation of lower responsivity and bandwidth of silicon photodetector. We extensively study various techniques to integrate silicon photodetector with passive SiN waveguides e ciently suitable for future short-reach datacom and lab-on-a-chip biosensors. In the rst part, we realize a single-mode SiN waveguide along with high-e ciency surface grating couplers. We have demonstrated a uniform and apodized grating coupler with a bottom Bragg re ector. Apodized gratings provide higher coupling e ciency than uniform gratings due to better mode pro le matching between Gaussian-shaped ber mode and the apodized grating eld pro le. Distributed Bragg re ector (DBR) reduces the optical loss due to high order di racted light directed towards the bottom substrate. SIN apodized grating coupler with DBR as the bottom re ector achieves the highest ever coupling e ciency of 2.19 dB/coupler and 3dB bandwidth of 40 nm at 876nm wavelength. In the second part, we demonstrate various architectures to integrate high-speed silicon photodetector with SiN waveguide. First, we demonstrate the integration of SiN waveguide with high-speed, lateral silicon pin photodetector. Compared to the silicon photodetector realized on bulk silicon, photodetector on an SOI has higher bandwidth due to the lower cross-section. We use silicon inverse taper to improve the coupling from SiN to silicon, which results in better responsivity of silicon photodetector. We have achieved the highest responsivity of 0.44 A/W and bandwidth of 15 GHz for the integrated silicon pin. Bandwidth improvement without degradation of responsivity is attributed to the lateral collection of photocarriers transverse to the propagation direction, and low RC time-limited bandwidth due to the thin silicon. To enhance the photodetector responsivity further, we propose a SiN ring resonator enhanced silicon metal-semiconductor-metal (MSM) photodetector. Compact, cavity-enhanced silicon- MSM photodetector responsivity is estimated to be 0.81 A/W at 5 V, which is 100 times higher than the conventional waveguide photodetector. Moreover, the photodetector's compact size (6X6 m2) can o er high bandwidth due to reduced RC time-limited bandwidth. In this section, we also discuss the integration of SiN waveguide with a thin silicon-MSM photodetector (70 nm thick). In this con guration, the SiN waveguide is placed on top of the silicon-MSM. Since the silicon's thickness is low SiN, the waveguide does not su er from mode mismatch losses between silicon and SiN. Such con guration is attractive due to its high responsivity and bandwidth, along with ease of fabrication. We have shown the DC measurements with a maximum responsivity of 0.56 A/W at 10 V bias. Finally, we have demonstrated the integration of wavelength division multiplexer with silicon photodetector since the shortwave wavelength division multiplexing (SWDM) at 850 nm wavelength band is considered one of the viable solutions to attain 400 Gbps roadmap. We have realized the WDM using SiN Echelle gratings and integrated the output channel waveguides with a silicon-MSM photodetector. Experimentally, we have shown that the Echelle grating has the insertion loss of 4.3 dB and adjacent channel cross talk of 22 dB for the channels having wavelength separation of 10 nm. Future exploration of the demonstrated device can lead to precise wavelength ltering with on-chip detection useful for both high-speed short-reach datacom and lab-on-a-chip biosensors. In summary, we have demonstrated the capability of realizing a scalable, energy-e cient, and cost-e ective silicon nitride based integrated photonic receiver in the 850 nm wavelength band.

碩士
國立中山大學
電機工程研究所
82
We applied porous silicon to the fabrication of photodetect- or. Normally, the cut-in voltage of PS photodetector is about 0.8 V and ideality factor is 0.7 - 1.7. The short wavelength (ultraviolet) spectrum response of PS photodetector centered at 300 nm is dominated by the optical absorption of Al-PS Schottky junction. And the quantum efficiency at -2 V is usually higher than 6000% at 300 nm. The highest one of our works is 66100%. Porous silicon is a direct energy bandgap material, and its PL spectrum is ranging from 580-950 nm. The spectrum response with wavelength ranging 500-1100 nm (band- edge of Si) is due to the photon absorption of the PS layer. The quantum efficiency in this wavelength region is about 90%-100%. The quantum efficiency centered at 1700 nm is resulted by ΔEc, the conduction-band offset. And the optical response from about 1300-1400 nm is due to Er, the interface residual gap. Because of the quantum effic- iency effect, different Si wire size leads to different energy bandgap. ΔEc and Er are of a distribution, too. So, all of the spectrum response are broad-band, and the PS photodetector would detect a broad wavelength band at the same time. The PS photode- tector therefore is a full color photodetector. The textured su- rface of PS layer would reduce the reflection of incident light and it will increase the quantum efficiency. In addition, the PS layer is an intrinsic layer and it acts as the "i" layer in p-i-n photodetector. The voltage drop in the Si wire of PS layer would result in a locally strong field. The gain of the PS phot- odetector would be greatly enhanced by avalanche in PS layer. The PS photodetector hence exhibits very high sensitivity to full color, and it is a high sensitivity and wide band photodet- ector.

碩士
國立清華大學
電機工程學系
85
The purpose of this thesis is to fabricate a-SiGe:H IR detectors. Two mainefforts of this thesis are (1)fabrication of a-SiGe:H films. (2)NIPIN IR photodetectors. In order to reduce the influence of the particulate formation in thepreparation of a-SiGe:H films, high hydrogen dilution ratio which is about93 % and pulse RF power is used for a-SiGe:H films. The particulate formationis decreased by using pulse RF power. The device structure of a-SiGe:H photo detector is NIPIN barrier type phototransistor. The I layer which light is first incident is the visible lightabsorption region and the other is the IR light absorption region.The photo to dark current ratios of this device under 780 nm LED illumination is over 100 and the photocurrent is over 1 uA which is enough for the requirement for the circuit design.

碩士
國立暨南國際大學
光電科技碩士學位學程在職專班
100
In recent decades, variety of the porous silicon material components was highly discussed, investigated, & intentional developed, such as photo detectors, light emitting diodes, and solar cells.As for this case study focus the surface of the silicon substrate on its electrochemical etching experiments, together utilizing the material properties of porous silicon to further improved the rate of light absorption, thereby increasing degrees of its responsively. However, the use of the current density adjustment, by etching of the different holes in its correspondences, further through the optical conductivity measurement methodology of its ongoing light reflection & measurement. In contrast, the strength characteristics of the porous silicon surface area on its volume ratio, achieved better performance in light absorption. Simultaneously, carrying out nanoparticle quantum dots, appropriately implementing correspondent holes, moreover, further highlighted the light response degrees in the optical absorption spectrums via the phenomenon of the Band-gap Alignment, as in further illustrated the porous silicon of hybrid quantum dot will significantly improve its responsiveness. In concluded the experiment & findings of this case study, I strongly believed the porous silicon of its future application by ways of porous silicon optical sensors developments will play a great assist enhancement in a new coming era.

碩士
國立中央大學
電機工程學系
103
Silicon/Germanium integration receives a lot of attention for both electronic and photonic devices; because pure germanium has a narrow band gap with electronic and hole mobility four times greater than that of silicon. These out-performed characteristics make germanium an ideal candidate for high speed electronics. However pure germanium grown on silicon is critical due to the lattice mismatch (4.2%) between these two semiconductors. Generally germanium epitaxy technique was high cost which use high vacuum machine. However it growth slow and take a lot of time, then can get high quality crystal germanium. It will enhance silicon/germanium integration difficulty. In this article, we use liquid-phase-epitaxy and rapid-melting-growth germanium technique, then can get high quality germanium. It was take low heat cost. Then we use TEM, SEM and Raman spectrum to observe the growth germanium quality and discuss silicon/germanium hetrojuction dislocation. Then produce single crystal germanium PIN photodector devices and discuses responsibility.

碩士
國立中央大學
電機工程學系
104
Recently, silicon(Si)-based electronics integrated with germanium(Ge) are widely adopted due to the fact that Ge has a narrow band gap as absorption material for infrared wavelength and CMOS-compatible process, while Si has a favorable ionization coefficient ratio as avalanche multiplication material. However, it’s difficult to grow Ge on Si because of the large lattice mismatch (4.2%) between Si and Ge.Generally, epitaxial growth of Ge on Si relies on ultra-high vacuum chemical vapor deposition, which was expensive and time-consuming. In this report, we used rapid-melting-growth method to grown high-quality Ge on SOI (silicon-on-insulator), and then we can obtain a high-quality Ge absorption layer. Defect analysis was conducted by TEM, SEM and Raman Spectrometer.Furthermore, the SACM Ge/Si avalanche photodetector by rapid-melting-growth was fabricated and measured I-V characteristics of the fabricated device were studied.

碩士
國立中央大學
光電科學與工程學系
103
In this research, we propose the three different structures of Si-based photodetector for the light source of 850 nm. We modulated thickness of film to improve efficiency of the photodetector at low temperature with brief fabrication steps and successfully integrated it to the Si-based optical waveguide bench. There is great potentiality to apply this technique to the MEMS system in the future. Nowadays, MEMS is getting popular in the industry but Ⅲ-Ⅴcompound is high cost and difficult to fabricate to the active components which are used in MEMS system. Therefore, the Si-based photodetector is a good choice to promote the development of MEMS system because of low cost, mature fabrication technology, convenient to reduce scale and integration. In this investigation, we focused on the structure of Metal-Semiconductor-Metal Si-based Photodetector. Annealing which can improve film quality or good ohmic contact is the common way to optimize the components but we used DC sputter system to sputter athe layer of transparent conducting oxide, In2O3: Sn (ITO) which provides good anti-reflection and efficient electrons or holes collection. The layer is used to fabricate photodetector components and the best responsivity is 0.79 A/W with anti-reflection layer for 850nm. In order to improve MSM structure by inducing the Iphoto/ Idark ratio、decreasing dark current without impacting on responsivity, we inserted the layer about 3.5nm SiO2 between ITO and N-type Silicon to form Metal-Insulator-Semiconductor structure. Because of the insulator, the Schottky barrier height was induced from 0.53eV (SiO2=0nm) to 0.64eV (SiO2=3.5nm) and the dark current was reduced by two orders of magnitude and the Iphoto/ Idark ratio was enhanced from 13.7 to 309, the responsivity of MIS photodetector was 0.51 A/W. Then, we optimized MIS photodetector at 120 oC process for 2 minutes. The result was that the dark current density was reduced from 1.89×〖10〗^(-3) A/cm2 to 4.04×〖10〗^(-5) A/cm2 and the photo current was increasing from 1.42×〖10〗^(-3) A to 1.84×〖10〗^(-3) A. In addition to MSM and MIS structures, we choosed PN photodetector that formed diode junction by traditional ion implantation as a reference. The responsivity of PN photodetector was 1.5 times larger than the one without anti-reflection layer. The responsivity of PN photodetector with anti-reflection was 0.32 A/W. We have compared three different structure of photodetectors; in addition it was successful to integrate the PN photodetector to the Si-based optical waveguide bench. Then, the PN photodetector component was compared with it that integrating in system. The responsivity of PN photodetector component was 0.33 A/W and the dark current was 2.74×〖10〗^(-7)A. Otherwise, the responsivity of photodetector in system was 0.28 A/W and the dark current was 6.49×〖10〗^(-6) A.

碩士
國立中山大學
光電工程學系研究所
103
In this paper, we design and study 1550 nm band Germanium photodetector based on silicon photonics platform. In optical fiber communication systems, Most of the silicon photonic components are using SOI wafers as the basic architecture. It has a high refractive index difference between the core and cladding. The refractive index of Silicon and Oxide are about 3.47 and 1.44 respectively. It has the advantage of reducing the loss and the size of component. In recent years, the development of siliconphotonics is quickly. It can fabricate optical switch, optical modulator, optical laser and photodetector individually. In the future, it will combine all the component into a chip. In this condition, the integrated of photodetector play a important role. The combination of germanium and silicon substrate as germanium photodetector, is one of the main issue of the future integrated. There are various structures of photodetector. Among them, we set the PIN photodetector for the research project. In PIN photodetector, evanescent coupling need longer absorption length. We design the butt coupling mode photodetector Type2 and Type3 to reduce the absorption length. In simulation, the optical signal can directly inject into germanium and be absorbed.Reducing the length from 25μm to 14 μm. The fabrication of Germanium photodetector in NDL can get the I-V curve.There is characteristic of diode in three types of photodetector. In XRD analysis, the test wafer FWHM is about 250 arcsec. Moreover, we will fabricated the real Germanium photodetector and measure it. This research and development will still ongoing.

碩士
國立中央大學
電機工程研究所
100
This work demonstrates a lateral avalanche photodetectors (APDs) by standard 0.18?m CMOS technology. In the proposed APD, we used n-implant/p-well to compose the avalanche region, and used metal layer to distinguish the absorption region and avalanche regions. At reverse bias of 11.27 V, the 3-dB bandwidth of APD is 3.0 GHz with the responsivity of 3.75 A/W. As in the proposed APD structure with Deep-N-Well layer, the 3-dB bandwidth is decreased to 1.0 GHz due to the extra depletion region at the p-well/DNW and p-substrate/DNW. The extra depletion regions could collect more photo-generated carriers but including diffusion carriers. The measured 3-dB bandwidth of proposed APDs will be improved when the reverse bias is within the avalanche region. At reverse bias of 11.3 V, the 3-dB bandwidth can be increased to 4.8 GHz. The possible reason may be attributed to the peaking effect in the frequency response. From the small signal analysis, we establish an equivalent circuit model including a parallel connected capacitance and an inductance due to the avalanche process. When the APD operates at high reverse bias voltages, the included LC equivalent circuit provides a resonance from the avalanche effect, which introduces a peaking phenomenon to further improve the frequency response. Finally, we discuss the differences between APDs with different absorption region widths. The highest M-Factor is about 148.84, and the best 3-dB bandwidth is 5.4 GHz at reverse bias of 11.73 V for the APD with 0.43 ?m absorption region width. As the avalanche region size decreases, the M-Factor decreases to 65.25 for the APD with 3.44 ?m absorption region width. And the 3-dB bandwidth is decreased to 0.7 GHz due to the diffusive photogenerated carriers as the absorption region width increases.

碩士
國立中央大學
電機工程學系
105
With the rapid development of optical fiber communication systems, the integration of germanium with silicon has attracted much attention for both electronic and photonic devices, taking the advantage of narrow band gap as well as high electron and hole mobility. On the other hand, germanium tin(GeSn) alloys was expected to improve the photo-response of a near-infrared photodetector. However the growth of single-crystalline GeSn alloys on silicon is very challenging because of the low equilibrium solubility (<1%) of Sn in Ge and the large lattice mismatch（greater than 4.2％）between silicon and GeSn alloys. Therefore, traditional GeSn alloys grown on silicon rely on ultra-high-vacuum chambers, which are expensive and time-consuming. In this article, we use rapid-melting-growth technique to grow GeSn alloys on silicon substrate, expect to get high-quality GeSn alloys. The as-grown GeSn alloys were employed for PIN photodetector devices. At the last, physical characteristics, including the I-V behavior, photoresponsibility, and materials quality were investigated by electrical analyzer, SEM, TEM and Raman spectroscopy.

碩士
國立臺灣海洋大學
光電科學研究所
102
The thesis studies the fabrication of efficient broadband(380 nm ~ 940 nm) photodetector(PD) base on graphene/CdSe QDs/Si multiple junctions. At 2 V bias for the graphene side, the external quantum efficiency(EQE) up to 218 % at 510 nm and the responsivity as high as 0.9 A/W for the range from 520 nm to 660 nm were achieved. The response time and recovery time are 0.24 ms and 0.28 ms, respectively. Under bias of -2 V for the graphene side, the dark current, photocurrent and on/off ratio was fou- nd to be 1.77#westeur024#10-6 A, 2.06#westeur024#10-4 A and ~10+2, respectively. It was found that when CdSe QDs were added, the EQE and responsivity of graphene/CdSe QDs/Si PD with respect to that of graphene/Si PD were enhanced together with the increase in the absorption band of the detector ranging from ultraviolet to near-infrared. Thus the graphene/CdSe QDs/Si multiple junctions can form the efficient broadband phtodetector.

博士
國立臺北科技大學
機電學院機電科技博士班
104
ZnO is a promising semiconductor material for many kinds of functional optoelectronics applications due to its wide direct band gap of 3.37 eV and high exciton binding energy of 60 meV at room temperature. ZnO is a hexagonal wurtzite structure and exhibits a non-centrosymmetric structure, which causes that the ZnO possesses polar and nonpolar plane. The polar and nonpolar planes show the different surface atomic configurations and physic and chemistry properties anisotropic, making that the ZnO attracts extensive research interests and diversity applications. The polar plane for ZnO is (0002) plane, which is a lowest surface energy plane. The most of studies are focus on growth of polar c-orientation ZnO nanostructures and explore its material properties and relative applications. Compared with polar ZnO, the nonpolar planes of (101 ̅0) and (112 ̅0) ZnO are seldom reported due to the difficulty of preparation. However, nonpolar ZnO is considered as a candidate material for next generation high efficiency optoelectronic device due to absence of spontaneous polarization effect in the crystal. Therefore, this present study used plasma enhanced chemical vapor deposition (PECVD) system to synthesize polar and nonpolar ZnO thin films. By adjusting the synthesis temperature, synthesis pressure, and precursor gas flow rate ratio to obtain the high quality ZnO thin films are the first step in this study. The experimental results indicate that the synthesis temperature is a dominated process parameter for controlling the crystallographic orientation of ZnO, and the crystal quality can be improved by altering the synthesis pressure and precursor gas flow rate ratio. Moreover, the possible growth mechanism of ZnO with different crystallographic orientations have been proposed based on the OES analyses and SEM observations. Both the polar and nonpolar ZnO thin films are used as a sensing layer for UV photodetectors applications. The interdigitated Pt thin films were deposited onto the polar and nonpolar ZnO thin films as a contact electrodes via conventional lithography process and RF sputtering. The performance including responsivity, reliability, and sensitivity of both detector were determined by typical current-voltage (I-V) characterization under dark and UV light illumination and time-dependent photoresponse measurement. The photoresponse results indicated that the nonpolar detector possesses better responsivity (4708.88 μA/W) and faster response (0.141 s) and recovery times (0.125 s) than the polar one (3367.73 μA/W, the response time cannot be determined). The performance of the polar ZnO-based UV photodetector can be improved by using RTA system annealed in different ambients. The photodetector annealed in air revealed the largest responsivity at operating temperature of 25 oC while the detector annealed in nitrogen showed a stable responsivity. The nonpolar ZnO-based UV photodetector with Pt as a contact electrode natively possesses good sensitivity and acceptable responsivity and reliability, but the polar ZnO-based UV photodetector have to be improved through a post-annealing assistance to exhibit a good performance.

碩士
國立成功大學
電機工程研究所
84
In this thesis , the amorphous silicon-germanium / crystal silicon heterojunction high-speed IR photodetector was studied in detail. In preparing the samples , the amorphous silicon- germanium alloys were grown on the crystal silicon subtrate by plasma enhanced chemical vapor deposition (PECVD). Both advantages of the low resistivity and high mobility characteristics of crystal silicon and the low temperature preparation processing,high optical absorption ,large area device feasibility and low cost of amorphous silicon are employed to prepare the heterojunction structure photodetector for faster response speed and lower cost. Compared with the traditoinal amorphous silicon germanium structure,the device with the structure of Al/n-a-Si:H/i-a- Si0.6 Ge0.4:H/p-c-Si has the following advantages: 1.The absorption wavelength peak moves to a higher value(865 nm ) than that 710 nm of the traditional amorphous silicon germanium structure. 2.The device has a faster response speed ( with a rise time of 195μs ) than that ( with a rise time of 465μs ) of the traditional amorphous silicon germanium structure . 3.The dark current has been decreased to a lower value (3.3μA under a reverse bias of 5V ) than that ( 50μA under the same bias)of the amorphous silicon germanium structure.

碩士
國立成功大學
電機工程研究所
83
In this thesis, an high optical gain amorphous silicon germanium alloy IR photodetector with Bragg reflectors has been developed by plasma enhanced chemical vaporphase deposition( PECVD). By using Bragg reflectors ,we can reabsorb the unabsorbed light to increase absorption efficiency , thus we can increase the optical gain. In addition, the Bragg reflector structure will serve as a barrier to block the outdiffusion of defects from the substrate into the active region of the detector which contributes to lower dark current. Therefore, under the same dark current condition, the PIN photodetector with Bragg reflectors can be operated with large voltage bias to increase the avalanche multiplication which will contribute higher optical gain. Based on experiment results, we find that the photodetector with Bragg reflectors has better characteristis than the conventional photodetector The full width of half magnitude(FWHM) can be reduced from 250nm to 150nm .The optical gain under 25uA can be attained as high as 328 with an incident light power of 1uW . Additionally,the rise time of this device is increased from 750us to 942.5us .

碩士
國立新竹教育大學
應用科學系碩士班
104
In this work, the strategy of Al-doped ZnO thin film (~100 nm) deposited on n-type silicon in atomic layer deposition (ALD) for high efficiency photodetector is provided and demonstrated successfully. Investigating the correlations between different amount of Al-doped within ZnO and material properties like the work function, resistivity, and crystallinity benefits to create the lower barrier height of Schottky hetero-junction among the interface of Al-doped ZnO thin film and n-type silicon. This Schottky hetro-junction plays an critial role for light detection. Lower barrier height of Schottky hetero-junction introduces the higher both dark current and photocurrent impacting the ratio of photocurrent to dark current typically. According to the verification results from XRD or UV-Visible spectrum of different amount of Al-doped within ZnO, the highly prefer C-axis and insignificant diversity in optical reflection behavior, respectively, provide the evidences for high quality Al-doped ZnO thin film with similar optical reflection. The Al-doped ZnO thin film possesses the extra capabilities of optical antireflection itself and lower resistivity compared with previously studies in this field. Besides, by the electrical or optical verification, the desired low barrier height (~0.35 eV) Schottky hetero-junction among the interface of Al-doped ZnO thin film and n-type silicon or the variant film thickness for modurated absorption spectrum can also be confirmed in this work. In our sample, as the film thickness of 100 nm, the carrier concentration and resistivity is about 1020/cm2 and 10-4 Ω/cm, respectively. Under the halogen illumination, our photodetector without external bias reveals the ratio of photo short current to dark current about 10000 implying feasibility and invention of the device design concept in photodetection application further.

碩士
國立中央大學
電機工程學系
104
High-speed communication devices have been developed for many years for the synergy of photonic and electronic signal transmission. Because of the narrow bandgap for infrared light absorption of germanium and its strain-engineered band gap modulation, the incorporation of germanium into silicon-based optoelectronics has attracted more attention. However, germanium has lattice mismatch of 4.2% with silicon. Generally, germanium epitaxy technique is high-cost by MBE and will compound the difficulty of silicon/germanium integration. In this thesis, we use rapid-melting-growth to grow germanium thin film on Si, and then we can obtain high-quality germanium mesa. The film was investigated by TEM and SEM microscopy and Raman spectrum. Moreover, the high-quality silicon/germanium/silicon PIN-photodetector was fabricated and measured to study its I-V characteristics and study the photocurrent and responsivity at near-infrared frequency.