IEE Books

Properties of crystalline silicon

Edited by Robert Hull, University of Virginia, USA

FOREWORD

Information technology is one of the key drivers of progress into the 21st century, transforming the way in which we communicate, learn, work, live. It will be essential for the growth of our economy as well as for the solutions of some of the most critical problems facing society. The electronic systems industry needed to support this vision has now overtaken the automotive industry in terms of contribution to the global economy. Micro- and optoelectronic sensors and transducers are the basis of a rapidly growing business segment. But to a very large extent electronic systems are based upon the technology for producing integrated circuits. And this technology - that just celebrated its 40th anniversary - is synonymous with silicon technology: its material properties combined with its abundance in nature have given silicon a unique position in the fabrication of miniaturized devices that is not likely to be challenged soon.

The business based upon this silicon integrated circuit itself is unique and particularly challenging. It is unique because over more than 3 decades it has sustained a market growth of 15% per year. It is at the same time extremely challenging, because this growth is based on an exponential technology development, resulting in electronic functions with continuously improved performance and reliability, reduced power consumption and a 25-30% per-year cost reduction.

Supporting this productivity increase is the growth of circuit complexity by a factor of approximately 2 every year as predicted by Gordon Moore in 1975 and resulting from feature scaling, die size increase and circuit cleverness. Some of the historical contributions to increased productivity – such as manufacturing yield improvement – are no longer available, and device scaling still remains the largest contributor to productivity growth. The Semiconductor Industry Association has coordinated the consensus building on the future technology requirements for maintaining the historical rate of advancement. The resulting vision is documented in the "National Technology Roadmap for Semiconductors" – the next update of which is likely to become known as the "International Technology Roadmap for Semiconductors" – predicting 50-nm features by 2012.

The challenges faced by the semiconductor community as it moves into production of metal-oxide-semiconductor (MOS) transistor devices with dimensions below 100 nm are enormous. As technology scales and materials are improved, they are stressed to their intrinsic limits and a thorough understanding of the fundamental properties is needed. At the same time new materials, new patterning techniques and new device structures will be required: new gate dielectrics and gate electrode materials and processes, as well as new interconnect materials and processes to replace the familiar metals, dielectrics and contact materials used in device interconnect structures. Also when removing 'bottlenecks' process technologists will require insight into the fundamentals.

In early stages of technology development modelling and simulation are extensively used in order to obtain insight into the directions of technology. It is clear that when devices are further scaled, tools will be needed to model physical and chemical processes at an atomic level and to simulate carrier transport under non-equilibrium conditions. The development of such tools will require advanced understanding of the fundamental processes.

Competitiveness of the IC industry has always been linked to advances in material preparation and process technology. But the IC industry has grown up in recent years. Among others things, this means that the costs together with the risks have risen. This has resulted in much more fragmentation and specialization as illustrated by the current successful business model based on the collaboration of "fab-less" design companies with "foundries" specializing in cost-effective and state-of-the-art manufacturing. The trend towards the "system-on-a-chip" has led to a growing need for truly multidisciplinary teams of experts with a deep technical insight in specific areas such as process modules or integration combined with a good understanding of system and electronic circuit design, or vice versa. Communication among the members of such teams has created an increased need for up-to-date and in-depth reviews of materials properties and technology.

The 18 chapters of this new book from the EMIS Datareviews series are the result of an ambitious project bringing together contributions from a worldwide group of experts covering a broad spectrum from silicon physics and material properties to device technology. The balance of coverage and presentation should make this volume a valuable working information resource for academics, semiconductor process developers and electronic device simulation engineers needing meaningful, usable data as they seek to advance their understanding and exploitation of silicon.

Inspired by the increased use of the Internet the editors will hopefully revive their plan to launch an online system. We believe such a service allowing regular additions and revisions would be of value to the thousands of materials and technology specialists who need instant access to a growing amount of data and knowledge. But for now and in the years to come, the present book will serve the needs of the "silicon community" more than adequately.

Roger de Keersmaecker
IMEC, Leuven, Belgium
May 1999

[Top of file]