INDUSTRY NEWS

California R&D program wants state's IC makers at head of class

Participants in a new semiconductor research partnership between industry and the University of California system hope to forge advancements in CMP, plasma, materials science, and lithographic applications through the program's first major project. Defect reduction and yield enhancement concerns are integral parts of this effort, as well as other process-related research projects begun under the cooperative plan, they say.

The major project titled "Small-Feature Reproducibility" is one of 15 awards issued this year under an inaugural program called UC Smart. Launched in February 1998, UC Smart is the first program in the state-run university system to specifically target semiconductor manufacturing. Recognizing the importance of the semiconductor industry to the California economy, program organizers have two goals: to improve chipmaking processes through basic research and to educate future process engineers in order to keep the state at the forefront of the worldwide industry.

The small-feature reproducibility project involves research conducted at three University of California campuses. Sponsored by 15 companies, the two-year award brings together faculty experts in lithography, plasma processes, chemical-mechanical polishing, and sensors and metrology, says David Dornfeld, a professor of mechanical engineering at UC Berkeley and former chairman of the UC Smart Task Force. He says the range of disciplines and the cross-campus cooperation indicate the strength of the UC Smart program. It recognizes not only the multifaceted nature of the chipmaking process but also the benefits of combining the expertise of professors and students who may have been unknown to each other prior to the program's inception. In addition to UC Berkeley, UC Davis and UCLA are the other campuses involved in the major project.

Participating researcher Olav Solgaard, a professor of electrical and computer engineering at UC Davis, says "extended collaboration between universities" is essential to develop the synergy necessary to successfully complete the project. For one thing, "it's too expensive to have a fully functional IC lab at every campus." The project makes use of labs at Berkeley, a main contributor to the project, and at Davis, which is near Sacramento. The Davis site has "a large cleanroom facility where we hope to do. . . a significant part of the educational aspect of this program." The Berkeley campus uses a microfabrication laboratory, where research is conducted under the directorship of Costas Spanos, a professor of electrical engineering and computer science.

Solgaard says the "key issues" in the major project award are photolithography modeling, plasma processing and diagnostic capabilities, and CMP. "In one sense I'm bridging all three," he adds, because his university work involves developing an optical interconnect communications system for sensors "that are going to be used for diagnosing plasma processes."

The idea is to place "a number of sensors on some kind of carrier that looks like a regular silicon wafer and. . . to communicate with those sensors. We want to get information from them, not to them," using simple modulators. The concept involves a laser and is called grading light valve technology, or GLV, says Solgaard. Although he has conducted some of the work with optical modulators before becoming involved in this project, this particular application of "very precise measurement inside a plasma environment" is new. "There are a number of basic issues that are being investigated for the first time."

Dornfeld, who has expertise in CMP, says he will focus on two areas. The first is the "invention of a material-removal mechanism which ultimately will lead to the development of simulation tools...that would allow a process engineer to run these 'what-if' scenarios." The second "has to do with the development of sensor technology in process monitoring [related to] yield management issues or just uniformity across the wafer or from wafer to wafer." A third area "that's a little sexy these days is the whole issue of environmentally conscious manufacturing and CMP."

Before becoming involved in the major project, Dornfeld had already noticed the intense interest the industry showed in his students at Berkeley. In particular, he was struck by the fact that his area of expertise, mechanical engineering, was in demand. "The more we talked with industry folks like Applied Materials and others, especially for mechanical engineers who have typically stood outside with their noses pressed against the window, we'd say, 'Isn't this great? They hire mechanical engineers to design process equipment.'

"It's really amazing," Dornfeld continues, "because there's this tremendous industry out there with an insatiable appetite for mechanical engineers... I started to see the companies trying to steal people out of my lab." Already involved in a National Science Foundation—funded CMP research project, Dornfeld plans to hold a "benchmarking" meeting on December 4 at the campus to discuss strategic issues in CMP processes "and especially to see how we can more closely integrate the research interest and thrust with the educational mission."

In its first year, UC Smart contributed $3.5 million from the state and the university system. Program administrators said they will match joint research funds on a dollar-for-dollar basis with California businesses. Equipment donations can be applied toward the matching funds at a value of as much as 20%. Among the industry participants are Applied Materials, Intel, Lam Research, and Novellus Systems.

Downturn or no downturn, layoffs or no layoffs, the semiconductor industry will still need qualified engineers when the bleeding stops. That's one reason that Applied Materials contributed to the UC Smart program, says Sam Broydo, managing director responsible for university relations at the world's leading semiconductor equipment supplier. Applied recently announced plans to lay off approximately 2000 workers and trim executive salaries by approximately 10%.

"We donated $100,000 to UC Berkeley and $100,000 to UCLA this year with the obligation to give the same amount of money next year," Broydo points out. "From our point of view this is basically to help them better educate themselves in areas relevant to our industry." Although some of the funds are aimed at joint R&D projects, the money was stipulated "for the general cause of making sure that we have more people coming out of the university into our pool from which we can hire engineers for our industry."

The overall goal of Applied's involvement is to help both undergraduate and postgraduate students become more familiar with integrated processing technology and the equipment involved, says Broydo. This knowledge would cover areas such as device physics, process chemistry, and process equipment. "Within those categories I think defect reduction is such a central or dominant theme of semiconductor technology that whatever you mention automatically involves being able to understand and implement all the requirements to minimize defects."

Julie Stein, UC Smart program director, says she has seen no drop in interest in the program because of the semiconductor industry's recent misfortunes, especially for the first round of proposals for awards that began in June. If a decrease in private sponsor participation does occur, she expects it to happen when the second set of proposals are due this month. "We're not sure that this is going to be a heavy round of proposals," says Stein.

UC Smart grew out of the success of another UC-related program. It is part of President Richard Atkinson's Industry-University Cooperative Research Program, which was established in 1996 to further joint research in areas critical to the California economy. The success of a biotechnology matching-grants program led Atkinson and colleagues "to set about targeting other areas within the state of California with a high rate of innovation and an impact on the California economy," Stein recalls. They spent close to a year looking for their new target, soliciting proposals from UC campuses. They settled on chip manufacturing and digital media.

"The faculty members with the strongest relations with the industry were the ones who were able to pull proposals together quickly and on such short notice," says Stein of the compressed schedule. For the 1999—2000 fiscal year, award proposals are due on February 5, 1999. Award projects will begin on June 1.

Six campuses are participating in the 15 project proposals awarded in the first round of funding, Stein says, adding that the goal is to include all of the UC system. In addition to the schools already named, the other UC campuses with inaugural projects are San Diego, Santa Barbara, and Irvine. "The Smart program is undertaking outreach in order to spread the word and plan activities that will create networking environments for faculty members at all the campuses."

Faculty and students in the physics departments and the electrical, chemical, and mechanical engineering departments "have participated so far" in the award projects, Stein notes. She adds that program administrators strongly believe that industrial engineering, chemistry, and computer science also can be tapped. "We hope to move into some of the broader aspects [of semiconductor manufacturing] like environmental and economic issues [as well]."

"I'm ecstatic about these programs," says UC Davis's Solgaard, who believes they're underpublicized. For Broydo at Applied, closer cooperation between the world of the mortarboard and the world of the circuit board is vital. "We consider it as a very crucial thing to the future of our company, regardless of today's business climate."

(For more information on the UC Smart program, call Julie Stein at 510/642-1896. The program's Web site is http://uc-smart.berkeley.edu. David Dornfeld may be reached at 510/642-0906 or by e-mail at dornfeld@me.berkeley.edu.)

Illustration by James Schlesinger

MicroHome | Search | Current Issue | MicroArchives
Buyers Guide | Media Kit