Dr. Gary W. Rubloff
Home
Research Topics
Process diagnostics, sensing, metrology, and control
Semiconductor manufacturing processes and equipment
Systems modeling, simulation, and optimization
Engineered learning systems

Recent Results

Publications
Presentations

Teaching

Current class
Experience

Other

Facilities
Research group
CV
Connections

Contacts & maps

Collaborations

Collaborators' corner
Internal-group
External-restricted
 

Other:
Consumer reports

Counter by digits.com

All materials at this site are copyrighted 2003 by the University of Maryland
1996-2003, all rights reserved

Fall 1999: Programmable Reactor Design

 

Instructions

Overview

Instructions

Organization

Teams

Results

 

Return to Teaching

Problem statement
The instructor identified a specific problem area for the course design project.

In the highly competitive semiconductor manufacturing industry, the productivity of capital equipment is a key factor in maintaining the industry's historical growth curve (25% improvement per year in function per unit cost). This challenge is pervasive in industry thinking, highlighted in the trade press, emphasized in The National Technology Roadmap for Semiconductors, and serves as a major driver for the entire supply chain of the industry.

With this backdrop, the instructor focused the technology domain of the course on the nature of manufacturing process equipment design, in which physics and chemistry conspire to determine the properties of submicron and atomic-scale materials and structures in these processes, while manufacturing cost considerations dictate that atomic-level control be achieved uniformly over increasingly larger wafer sizes (now moving from 8" to 12"). This uniformity requirement imposes constraints on acceptable process parameters, leading to tradeoffs between process performance at the material and device scale and manufacturing cost. Despite major efforts to optimize equipment design, the final design of the equipment hardware imposes fixed relationships between process performance and manufacturing cost. Furthermore, process adjustment is required in manufacturing and development to achieve effective integration of sequences of process steps for high yield (and rapid yield learning), so the performance-cost tradeoff can become even more problematic.

The challenge of the project was to assess a very different strategy for equipment design, namely one in which the design of the equipment might be adjustable and software programmable. Specifically, are there approaches by which one might achieve software programmability of across-wafer uniformity for at least some of the key processes? If equipment designs, in-situ sensors, and control strategies could be developed which allowed proper alteration of key process parameters as a function of position across the wafer, then uniformity and its cost implications might be decoupled from process performance. This might minimize constraints on usable process parameter regimes and improve material and device performance, since uniformity could be guaranteed at any desired process design point. In addition, it might be possible to exploit intentional nonuniformity to more rapidly explore process optimization (i.e., multiple experiments from a single wafer run), both in development and during process adjustment in manufacturing.

It was intended that the potential impact of this problem on semiconductor technology and its relevance to research interests at the University of Maryland would enhance the realism and meaning of the course experience for both systems and materials students.

Operations
Project teams were expected to meet regularly, outside of class as well as in class, to pursue their responsibilities. Team update presentations were given to the entire class several times through the course. Part of the class time, particularly earlier in the course, was devoted to lectures aimed at conveying key technology concepts and the relevance of systems engineering, particularly to that technology. The rest of class time was devoted to team meetings, with the instructor circulating between them, and to full-class discussions on the entire project, moderated by the instructor.

Expectations
Primary course deliverables were:
Final presentation, structured as an executive overview
Final report, prepared as a Word document

Both are available at this site.

A final exam concentrating on the key issues of the project was given as a take-home individual exercise. Grading was 40% team project, 40% individual contributions to project and class participation, and 20% final exam.