Atomic layer epitaxy continues to emerge as a promising epitaxial technique when atomic layer thickness control is required in semiconductors device structures. However, ALE still faces several problems such as low growth rate, relatively high carbon background and difficulties in growing high bandgap ternary alloys. This paper will outline recent progress in these areas illustrated by the state-of-art devices recently grown by the ALE technique.
Due to the material property restrictions of LiNb03 and III-V compound material systems a sizable portion of the research work on guided wave devices has been shifted to polymer-based materials. Low material dispersion, flexible material preparation process, unlimited device size and cost effectiveness are the major factors that can not be provided using conventional inorganic materials. By definition, polymer matrix is formed by linking an array of monomers. Therefore, there are infinite number of polymeric materials can be generated. The Polymeric materials suitable for guided wave device research are the ones with desired optical and electrooptic properties.
In this paper, we report the research status of the photolime gel superpolymer. In contrast to any artificial polymer that are synthesized according to a predesigned formula, the polymer we employed is a class of biopolymer which consists of thousands of 1 to 2 nmlong amino acids. A myriad of passive and active guided wave devices has been successfully fabricated using the photolime gel polymer. These include high density linear and curved channel waveguide arrays, electrooptic modulator and modulator array, highly multiplexed waveguide holograms for wavelength division demultiplexing and optical interconnects, waveguide lens, and rare earth ion doped polymer waveguide amplifier. A single-mode linear channel waveguide array with device packaging density of 1250 channels/cm is achieved. The first 12-channel wavelength division demultiplexer working at 830, 840, 850, 860, 870, 880, 890, 900, 910, 920, 930, and 940nm on a GaAs substrate is also described in this paper. A polymer-based electrooptic traveling wave modulator with 40 GHz electrical bandwidth is further delineated. A rare earth ion doped polymer waveguide amplifier working at 1.06pm with 8.5dB optical gain is also achieved using this polymer matrix.
The tunability of the waveguide refractive index allows the formation of a graded index (GRIN) layer. As a result, these active and passive guided wave devices can be realized on any substrate of interest. High quality waveguide (loss<0.1dB/cm) has been made on Glass, LiNb03, Fused Silica, Quartz, PC board, GaAs, Si, Al, Cu, Cr, Au, Kovar, BeO, AI2O3 and AIN.
X-ray diffraction provides a means of measuring the residual stress tensor in films, their bonding to the substrates, yield stress, stress relaxation, and stress gradients, all without removal from the substrate. Examples of such studies are presented for interconnect material, sputtered film and epoxied layers.
These stresses often far exceed those of the same material in the bulk, even when they are microns thick. Grain boundary diffusion controls stress relaxation and, therefore, also film failure by void formation. There are no gradients through the thickness of interconnect material.
LiNb03 integrated optics is a sufficiently mature technology that several devices are close to commercial realization. These include high speed modulators, switch arrays, and optical circuitry for the fiber gyroscope and for optical time domain reflectometry. Although the physical principles which underlie the operation of these devices are well understood, predictable and reproducible performance is often difficult to achieve and maintain. This is due to an absence of a detailed understanding of the materials science of LiNb03 and of the processes associated with integrated optics device fabrication. In this presentation I focus on the materials issues which affect the performance, yield, and the design predictability of these devices.
Atomic structure of defects and interfaces in semiconductor heterostructures determine properties and operation of device structures of these materials. In order to minimize their adverse effects, we need to understand the atomic structure of these defects and correlations with physical properties and device operation. A three-step iterative procedure has been developed to determine atomic structure of dislocations, twins, stacking faults and grain boundaries. These steps consist of : (a) calculation of atomic structure of defects using appropriate interatomic potentials, (b) simulation of high-resolution TEM images, and (c) comparison with experimental images. By manipulating atomic structure of defects, electrically active defects can be converted into inactive defects, thus reducing their effectiveness as trap or recombination centers. We also discuss various methods which can be employed to reduce the number density or preferably eliminate the process-induced defects.
Proton exchange process is an attractive technique for the fabrication of low-loss optical waveguides and devices in ferroelectric materials. It has some distinctive advantages, such as simplicity, large index change and relatively low exchange temperature compared to titanium indiffusion. Due to their unique properties, protonexchanged waveguides have numerous applications in integrated optical devices. This paper will review and discuss the proton-exchange technology for integrated optics application.
Integrated optical components based on ion-exchanged glass waveguides are attractive for several reasons. Significant progress has occurred in the past decade in the understanding of the ion-exchange process and its relationship to the dopant profile. It is now possible to model the thermal as well as electric field enhanced ion exchange with reasonable accuracy and to design ion-exchanged waveguides and devices with prespecified characteristics.
This paper first describes the theory of the ionexchange process and its application to planar and 2-D (channel) waveguides. Following this, we review the latest results for several passive devices based on single-mode channel waveguides. These include couplers, power dividers, MUX/DEMUX, and polarization splitters. Finally, we point out the latest trends and novel emerging applications of ionexchanged waveguides, such as waveguide lasers and amplifiers, all optical switching, and integration of passive waveguides on glass with optoelectronic devices. We conclude by presenting some of the issues to be addressed for the technology to make an impact on the market.
With the rapid development and application of fiber and integrated optic guided wave technologies, concerns for the operation of these technologies in adverse environments have arisen. Recent radiation effects data pertaining to guided wave components are of great interest to planners of space and military systems using these technologies. Key research results and brief descriptions of some ongoing studies are highlighted within this review in order to familiarize the reader with the state-of-the-art, and to facilitate access to authoritative reference materials.
Between bandgap energies of 1.77 and 3.35 electron volts (eV) lies the visible spectrum while the ultraviolet (UV) spectrum lies between 3.35 and 12.39 eV. Over 99% of all semiconductor research and development expenditures during the past 50 years has been expended on semiconductor materials exhibiting bandgaps below these ranges. Many good reasons for the relative lack of attention to these higher bandgap semiconductors exist and are addressed herein, but various new growth techniques and incentives have resulted in research funding support in these materials recently experiencing greater increases than any other category of semiconductors.
In addition to the obvious applications of visible and UV light emission and detection, these materials offer other significant advantages including negative electron affinity, extremely high dielectric strength, low dielectric constants, very high charge carrier velocities (at high electric field strengths) and extremely low intrinsic leakage. The latter attribute enables such new concepts as non-volatile memories and extremely low loss charge coupled devices.
In order to put research and development of optoelectronic integrated circuits (OEICs) for fiber-optic telecommunications in perspective, the determinant issues and market forces driving the evolution of telecom networks are first briefly analyzed. Advanced transfer links require high-performance OEICs, whereas switching nodes call for compact solutions, and optical distribution to the subscriber will demand low cost components. A review of state-of-the-art InP OEIC technologies shows that OEIC receivers are now able to compete with hybrids in terms of performance, with both the HFET and the HBT being candidates for a generic OEIC technology, since high bit rate transmitter OEICs can also be made using these electronic devices, albeit with a more complex technology. Cost requirements will drive the realization of subscriber modules, which need more development of photonic integration technology before adjunction of electronic devices, especially to lasers. For the nearer term, hybrid flip-chip mounting of optoelectronic devices onto electronic ICs could prove to be an economically viable alternative, with challenging performances as well. Epitaxial lift-off has more remote and uncertain possibilities. For low cost OEIC components to ever reach the market, mass manufacturable packaging technologies need urgently to be developed.
As we enter the Information Age, it is imperative to accelerate the evolution of many relevant supporting technologies. Virtually all of these essential technologies are hindered by inherent, unresolved, critical materials related impediments, if not limitations. Challenging materials problems await resolution in the areas of information transmission (photonics, electronics and optoelectronic integration) and information storage devices. Selected recent research activities, aimed at addressing these technological bottlenecks are discussed.
III-V semiconductor devices, most notably light-emitting diodes, lasers and photodetectors lie at the heart of modem optoelectronics. The efficiency of the conversion process between electrons and photons, in either direction, can be remarkably high in the best laser and photodiode devices. For communications purposes, the high modulation rate capability of semiconductor lasers and the high bandwidth detection capability of photodiodes, both traceable in part to the intrinsically small size of typical devices, are important aspects. Because of the importance of the communications (in particular fibre-optical communications) applications of optoelectronic integrated circuits (OEICs), this review will concentrate on that area, but other potential areas of application will also receive some attention.
Technological compatibility between photodetectors and field effect transistors or optical waveguides is one of the key factors for the development of integrated circuits associating such components. In this paper, we present recent advances in Europe such as the use of lattice mismatch epitaxy, selective and shadow masked growth, planar structure and dielectric optical waveguide to reduce the technological difficulties inherent to the fabrication of these integrated circuits.
Low temperature epitaxy (LTE) of Si and SiGecanbe performed at a temperature of 550 C or lower. Very promising applications can be opened. Such as high speed/high frequency operations at 90GHZ by constructing heterojunction bipolar transistors. High performance FET'slikepseudomorphic p-channel orn-channel high mobility field effect transistors are presented which canbe composed to perform CMOS operations.
Optoelectronic devices such as IRdetectors (1-12um), mutiple quantum well (MOW), disordered superlattice (d-SL) which are the potential candidatesof IR detector and optical sources (e.q. LED, LD etc.)
Various physical insights regarding to SiGe heterostructures are presented which includeswave function filter, mass filter as well as band mixing are introduced.
Researchesat National Nano Device Laboratory (NDL) which processes the capability of 0.3um Si ULSI technologies and SiGe works as well as lll-V, a-Si/SiGe lines are also presented.
As device dimensions in nanoscale structures and mesoscopic devices are reduced, the characteristics and interactions of dimensionally-confined longitudinal-optical (LO) phonons deviate substantially from those of bulk polar semiconductors. This account emphasizes the properties of LO-phonon modes arising in polar-semiconductor quantum wells and quantum wires. In particular, this review highlights recent results of both microscopic and macroscopic models of LO phonons in polar-semiconductor quantum wells and quantum wires with a variety of cross sectional geometries. Emphasis is placed on the dielectric continuum model of confined and interface phonons. In addition, this review provides brief discussions of how carrier-LO-phonon interactions change in the presence of dynamical screening. Finally, the use of metal-semiconductor heterointerfaces to reduce unwanted inelastic scattering in nanoscale electronic and optoelectronic structures is discussed.
The objective of this review is to present the reader with a roadmap to the development of guided wave phase and amplitude modulators with emphasis on the economic aspects of the applications, the device requirements, and some of the technical issues that will impact the demand for modulators in the near future.
Attachment of optical fiber to planar waveguides is an important issue in production of integrated optical circuits for practical applications. To produce low-cost components, it is essential to use an effective process for fiber-waveguide connection. In this paper, we review some techniques proposed for this purpose and discuss their suitability.
High speed, high frequency optical interconnections are needed for computers and for controlling microwave and millimeter wave phased arrays in radar and communications systems. Direct modulation of semiconductor lasers to 3db frequency limits of 28-30 GHz(1,2) has been achieved, and this limit may be doubled in the next five years. External modulation of continuous optical signals has been used to achieve, for example, a 3db bandwidth of -40 GHz with ~2V-mm modulation sensitivity using multiquantum- well, waveguide modulators. This frequency response may also be doubled in the next five years. In addition, narrow band operation can be achieved, by mode locking, at center frequencies up to 350 GHz(3), and this also can be doubled in the next five years.
In the twenty first century, optoelectronics will join with electronics and establish new photon electron collaborating systems, Photo-Electronic Integrated systems, which will become the major type system for information processings. The optical interconnection in computers which has started recently will develop deeper into the computer system and will reach optical interconnections inside of ULSIs, resulting the first step of Photo-Electronic Integrated System. Many micro optoelectronic devices will be integrated into ultra-LSIs to form ultra-large scale opto-electronic integrated circuits, U-OEICs.
Close collaborations of photons with electrons are thus achieved and a wide variety of new systems will be developed based on U-OEIC technologies. By using U-OEIC scheme electronic systems will become faster, lower power, lower noise and also expand into three dimensional systems. Three dimensional super-parallel systems will be developed, which will be used for ultra high speed computers, neural systems and many "optical systems". Two dimensional images can be analyzed by the Photo-Electronic Integrated Systems, in which high density micro-electronics will be utilized for logical operations in the three dimensional optically interconnected circuits.
In the last decade, semiconductor technology has been advanced to a great extent in terms of electronic and photonic discrete devices. One of the main reasons for such a progress, is the result of advancement in the epitaxial growth techniques such as molecular beam epitaxy (MBE) and metalorganic chemical vapor deposition (MOCVD), where device quality films can be grown with great control over composition, uniformity and thickness. MOCVD has proven to be one of the best growth methods for many IH-V semiconductor thin films 1. Its flexibility and potential to yield a broad range of growth rates resulted in the layers featuring the thicknesses from tens of microns down to several nanometers. Planar structures containing quantum wells with atomically flat interfaces, superlattices, strained or graded-index layers were successfully grown by MOCVD. Furthermore, MOCVD proved its efficiency in producing a laser devices by overgrowth and epitaxy on patterned substrates. The importance of MOCVD is strongly enhanced by the possibility of large-scale production by simultaneous growth on several substrates in one process. Several III-V semiconductor films with bandgaps ranging from infrared to ultraviolet (15 to 0.2 μm) have been successfully grown by MOCVD.