Computer Science Workbench is a monograph series which will provide you with an in-depth working knowledge of current developments in computer technology. Every volume in this series will deal with a topic of importance in computer science and elaborate on how you yourself can build systems related to the main theme. You will be able to develop a variety of systems, including computer software tools, computer graphics, computer animation, database management systems, and computer-aided design and manufacturing systems. Computer Science Workbench represents an important new contribution in the field of practical computer technology. TOSIYASU L. KUNII Preface The evolution of database systems research is itself a story. Long after the emergence of systems derived from practical applications, Codd's relational data model has gradually occupied the theoretical domain of database systems and is moving into the realms of practical use. Certainly, the theoretical foundation makes database design, validation, and testing easier. However, Cod d's model allows only fiat tables to be handled, while most business and engineering data in practice are in nested table forms. Thus, a recent major obstacle in database systems development is the large gap between the theory and the practice.

The purpose of this book is to present analysis and design principles, procedures and techniques of analog integrated circuits which are to be implemented in MOS (metal oxide semiconductor) technology. MOS technology is becoming dominant in the realization of digital systems, and its use for analog circuits opens new pos­ sibilities for the design of complex mixed analog/digital VLSI (very large scale in­ tegration) chips. Although we are focusing attention in this book principally on circuits and systems which can be implemented in CMOS technology, many con­ siderations and structures are of a general nature and can be adapted to other promising and emerging technologies, namely GaAs (Gallium Arsenide) and BI­ MOS (bipolar MOS, i. e. circuits which combine both bipolar and CMOS devices) technology. Moreover, some of the structures and circuits described in this book can also be useful without integration. In this book we describe two large classes of analog integrated circuits: • switched capacitor (SC) networks, • continuous-time CMOS (unswitched) circuits. SC networks are sampled-data systems in which electric charges are transferred from one point to another at regular discrete intervals of time and thus the signal samples are stored and processed. Other circuits belonging to this class of sampled-data systems are charge transfer devices (CTD) and charge coupled dev­ ices (CCD). In contrast to SC circuits, continuous-time CMOS circuits operate continuously in time. They can be considered as subcircuits or building blocks (e. g.

Semiconductor device modelling has developed in recent years from being solely the domain of device physicists to span broader technological disciplines involved in device and electronic circuit design and develop­ ment. The rapid emergence of very high speed, high density integrated circuit technology and the drive towards high speed communications has meant that extremely small-scale device structures are used in contempor­ ary designs. The characterisation and analysis of these devices can no longer be satisfied by electrical measurements alone. Traditional equivalent circuit models and closed-form analytical models cannot always provide consis­ tently accurate results for all modes of operation of these very small devices. Furthermore, the highly competitive nature of the semiconductor industry has led to the need to minimise development costs and lead-time associated with introducing new designs. This has meant that there has been a greater demand for models capable of increasing our understanding of how these devices operate and capable of predicting accurate quantitative results. The desire to move towards computer aided design and expert systems has reinforced the need for models capable of representing device operation under DC, small-signal, large-signal and high frequency operation. It is also desirable to relate the physical structure of the device to the electrical performance. This demand for better models has led to the introduction of improved equivalent circuit models and a upsurge in interest in using physical models.

The era of the personal computer has, without doubt, permanently altered our life style in a myriad of ways. The "brain" of the personal computer is the microprocessor (together with RAM and ROM) which makes the decisions needed for the computer to perform in the desired manner. The microprocessor continues to evolve as increasingly complex tasks are required. While not sharing the limelight of the microprocessor, the "heart" of the personal computer, namely the power supply, is equally important since without the necessary source of power the microprocessor would be a useless piece of silicon. The power supply of twenty years ago was much different than its modem day equivalent. At the dawn of the personal computer era in the late 1970s, de power was obtained from a simple diode bridge. However, the need for smooth, regulated DC at low voltage required at the same time both a bulky input transformer and a large dc side ftlter. Those computer fans present at the birth of this industry can remember the large boxes housing our Altair, Cromemco and Northstar computers which was made necessary largely because of the huge power supply. It is not well appreciated but certainly true that the huge sucess of the Apple II computer in those days was due, at least in part, to the relatively slim proftle of the machine. This sleek appearance was largely due to the adoption of the then new and unproven switched mode power supply.

The aim of this book is twofold: it is intended for use as a textbook for a ~ourse on electronic materials (indeed, it stems from a series of lectures on this topic delivered at Milan Polytechnic and at the universities of Modena and Parma), and as an up-to-date review for scientists working in the field ::>f silicon processing. Although a number of works on silicon are already available, the vast amount of existing and new data on silicon properties are nowhere adequately summarized in a single comprehensive report. The present volume is intended to fill this gap. Most of the examples dealt with are taken from the authors' every­ day experience, this choice being dictated merely by their greater knowl­ edge of these areas. Certain aspects of the physics of silicon have not been included; this is either because they have been treated in standard textbooks (e.g. the inhomogeneously doped semiconductor and the chem­ istry of isotropic or preferential aqueous etching of silicon), or because they are still in a rapidly evolving phase (e.g. silicon band-gap engineering, generation-recombination phenomena, cryogenic properties and the chem­ istry of plasma etching). In line with the standard practice in microelectronics, CGS units will be used for mechanical and thermal quantities, and SI units for electrical quan­ tities. All atomic energies will be given in electronvolts and the angstrom will be the unit of length used for atomic phenomena.

Solid-state physics has for many years been one of the largest and most active areas of research in physics, and the physics of metals and semiconductors has in turn been one of the largest and most active areas in solid-state physics. Despite this, it is an area in which new and quite unexpected phenomena - such as the quantum Hall effect - are still being discovered, and in which many things are not yet fully understood. It forms an essential part of any undergraduate physics course. A number of textbooks on solid-state physics have appeared over the years and, because the subject has now grown so large, the books too have usually been large. By aiming at a more limited range of topics, I have tried in this book to cover them within a reasonably small compass. But I have also tried to avoid the phrase 'It can be shown that. . . ', as far as possible, and instead to explain to the reader just why things are the way they are; and sometimes this takes a little longer. I hope that some readers at least will find this approach helpful. 1 The free-electron model 1. 1 THE CLASSICAL DRUDE THEORY The characteristic properties of metals and semiconductors are due to their conduction electrons: the electrons in the outermost atomic shells, which in the solid state are no longer bound to individual atoms, but are free to wander through the solid.

Modern electronics is the most visible result of research in solid state physics. Transistors and integrated circuits are used everywhere in ever increasing numbers. The microprocessor controlled coffee-pot exists. Most experimental physicists, and, indeed, experimental scientists in most disciplines, study their subject with the aid of apparatus containing significant amounts of electronics and much of that electronics is digital. In order to design experiments and apparatus or simply to understand how a piece of equipment works, an under­ standing of electronics has become increasingly important. In recognition that electronics has pervaded so many areas, courses in digital electronics are now a recommended part of physics and many other science degree courses. At the introductory level, digital electronics is, primarily, a practical subject with relatively few basic concepts and any complex­ ity arises from the coupling together of many simple circuits and the extensive use of feedback. Designing an electronic circuit and then getting it to work correctly provides an experience, and a sense of achievement, which is significantly different from most undergradu­ ate work as it more closely resembles project work than standard laboratory practicals.

Large computational resources are of ever increasing importance for the simulation of semiconductor processes, devices and integrated circuits. The Workshop on Computational Electronics was intended to be a forum for the dis­ cussion of the state-of-the-art of device simulation. Three major research areas were covered: conventional simulations, based on the drift-diffusion and the hydrodynamic models; Monte Carlo methods and other techniques for the solution of the Boltzmann transport equation; and computational approaches to quantum transport which are relevant to novel devices based on quantum interference and resonant tunneling phenomena. Our goal was to bring together researchers from various disciplines that contribute to the advancement of device simulation. These include Computer Sci­ ence, Electrical Engineering, Applied Physics and Applied Mathematics. The suc­ cess of this multidisciplinary formula was proven by numerous interactions which took place at the Workshop and during the following three-day Short Course on Computational Electronics. The format of the course, including a number of tutorial lectures, and the large attendance of graduate students, stimulated many discussions and has proven to us once more the importance of cross-fertilization between the different disciplines.

Control systems particularly designed to manage uncertainties are called robust control system. Choosing appropriate design methods, the influence of uncertainties on the closed-loop behaviour can be reduced to a large extent. Most of the important areas of robust control are covered. The aim of the book is to provide an introduction to the theory and methods of robust control system design, to present a coherent body of knowledge, to clarify and unify presentation of significant derivations and proofs. The book contains a thorough treatment of important material of uncertainties and robust control which is scattered throughout the literature.

Electronic materials are a dominant factor in many areas of modern technology. The need to understand'them is paramount; this book addresses that need. The main aim of this volume is to provide a broad unified view of electronic materials, including key aspects of their science and technology and also, in many cases, their commercial implications. It was considered important that much of the contents of such an overview should be intelligible by a broad audience of graduates and industrial scientists, and relevant to advanced undergraduate studies. It should also be up to date and even looking forward to the future. Although more extensive, and written specifically as a text, the resulting book has much in common with a short course of the same name given at Coventry Polytechnic. The interpretation of the term "electronic materials" used in this volume is a very broad one, in line with the initial aim. The principal restriction is that, with one or two minor exceptions relating to aspects of device processing, for example, the materials dealt with are all active materials. Materials such as simple insulators or simple conductors, playing only a passive role, are not singled out for consider­ ation. Active materials might be defined as those involved in the processing of signals in a way that depends crucially on some specific property of those materials, and the immediate question then concerns the types of signals that might be considered.

To be perfect does not mean that there is nothing to add, but rather there is nothing to take away Antoine de Saint-Exupery The drift-diffusion approximation has served for more than two decades as the cornerstone for the numerical simulation of semiconductor devices. However, the tremendous speed in the development of the semiconductor industry demands numerical simulation tools that are efficient and provide reliable results. This makes the development of a simulation tool an interdisciplinary task in which physics, numerical algorithms, and device technology merge. For the sake of an efficient code there are trade-offs between the different influencing factors. The numerical performance of a program that is highly flexible in device types and the geometries it covers certainly cannot compare with a program that is optimized for one type of device only. Very often the device is sufficiently described by a two­ dimensional geometry. This is the case in a MOSFET, for example, if the gate length is small compared with the gate width. In these cases the geometry reduces to the specification of a two-dimensional device. Here again the simplest geometries, which are planar or at least rectangular surfaces, will give the most efficient numerical codes. The device engineer has to decide whether this reduced description of the real device is still suitable for his purposes.