Lithography The ability to control interactions between features, assess image quality and compensate for lateral diffusion spreading during lithography is essential to enhancing the fundamental limits on SFR. In this project, we focus on creative, high-risk, high-payoff techniques involving metrology, control, masks, optics and resists for enhancing lithography resolution and capability.
Transient Enhanced Diffusion Our studies focus on understanding the changes of Silicon self- and dopant diffusion when the material turns from intrinsic to extrinsic at the high diffusion temperatures. This effect has the potential of limiting the vertical miniaturization of next generation transistors, and it is therefore an important SFR subject matter.
Sensor Integration The effects of process variability can be greatly reduced by the intelligent use of all available process state information. Sensor data in conjunction with models enable optimal operating point selection, intelligent process diagnosis, and run-to-run or real time process control. In this research thrust, we will invent and integrate metrology, data mining and modeling to realize these options.
Plasma Etching Plasma instabilities and high electron temperatures in inductively driven plasma reactors are important factors limiting process reproducibility, uniformity, and selectivity during etch/deposition for larger size wafers. This project focuses on novel methods to sense, measure, model and control plasmas used in the manufacture of deep-submicron features over large substrates.
Education The educational component of our program is addressing both the UC student body, as well as the personnel of participating industries. The traditional UC model is “education through research.” Our program has followed this tradition by addressing “output-end” educational upgrades at the junior, senior and graduate years. Two major thrusts are (1) an upgraded semiconductor processing course at Berkeley, increasing capacity and making the course accessible to all science and engineering majors, and (2) the creation of new Berkeley laboratory course on semiconductor processing equipment, interdisciplinary across four engineering disciplines.
Chemical Mechanical Planarization A comprehensive model of CMP will allow the prediction of feature pattern evolution, the design of optimal process recipes and the understanding of the CMP-imposed limits on SFR. Such a model is the focus of our comprehensive, multi-disciplinary approach to this problem.
Objective
Our objective is to introduce innovative process steps, new sensors, and effective control strategies that will make IC manufacturing factories more efficient. During the first year (2002-03) of this study we introduced several specific innovations. During the second year (03-04) we worked to integrate these innovations into a broad control system, and we are presently focused on technology transfer and field trials.
Abstract
The dramatic progress in producing amazingly capable, yet inexpensive Integrated Circuits (ICs) hinges on our ability to craft minute electronic devices into silicon. Today, these devices are so small that about 5,000 of them would fit side by side across one human hair. IC production has experienced a dramatic improvement in productivity in the last 30 years, and the results have transformed society. This trend will end unless a systematic approach is used to enable the reliable reproducibility of the very small devices used in future IC products. In the past we have assembled a comprehensive research program that brings fundamental understanding to this issue. Our present focus is to distill this understanding into a practical foundation that will enable the industry to produce even more capable, even more cost effective ICs.
(You need a login and password to access this site.
Please, email the
for login/password information.)
