INDUSTRY NEWS

European chip equipment initiative succeeds in putting region's toolmakers on world map

A four-year-old initiative has succeeded in reversing the perception that semiconductor manufacturing tools made by European companies, while innovative, lack the productivity capabilities needed to compete with equipment made by their counterparts in the United States and Asia.

The Semiconductor Equipment Assessment (SEA) initiative accomplished this goal in two ways, asserts Jeff Bruchez, head of project interface and dissemination in the SEA dissemination office. First, the program has included extended equipment testing, primarily at IC manufacturers and research institutes. The tests involved the efforts of global partners from the United States, Japan, and South Korea. Second, it has ensured that realistic targets were set and sought "via independent assessment."

"A high level of success has been achieved within SEA projects to date, with significant improvements in performance, equipment reliability, productivity, and user satisfaction," Bruchez says. Because the results are produced by users--primarily chipmakers--instead of suppliers, then independently evaluated by the SEA dissemination organization, the results arguably have more credibility than if a supplier alone provided its own assessment.

Initially sponsored by Esprit at its inception in 1996, SEA is working within the Information Society Technologies (IST) program of the European Community (EC) for the 1999­2002 period. SEA seeks "to provide a platform for European equipment and materials companies to demonstrate their emerging equipment to global users, not with quickie trials, but in a prolonged, phased manner," Bruchez points out. The assessment plan first sets up the "relevant state-of-the-art requirements" in the technological application, then establishes "productivity metrics during a 12- to 18-month project." The organization disseminates the results in the form of technical papers, project articles, and wrap-up workshops for interested parties.

According to an internal poll, the approach is working to participants' satisfaction, Bruchez notes. Ninety-five percent of IC manufacturers involved in SEA viewed favorably the long evaluations with industrial involvement, and 86% actually saw performance gains. Eighty-one percent of responding manufacturers say the new technology or equipment offers more benefits than the tools they used previously.

Bruchez points out that a recent survey by VLSI Research, the San Jose­based market research firm, shows "nine citations for European Community­based equipment companies that are now rated within the top 10 in their relevant categories." Seven of the companies "have had their equipment assessments completed in a total of 13 separate SEA projects," he notes.

Among the participants in the "user-driven" organization are tool users from Europe, the United States, South Korea, Taiwan, and Japan. The list includes AMD, Analog Devices, Applied Materials, MEMC, Samsung, TSMC, TI, and Xerox PARC. Suppliers are based in Germany, the Netherlands, United States, Belgium, France, Israel, United Kingdom, and Austria. Eleven industrial research organizations based in Europe are also taking part.

More non-European companies take part in the program now than European participants, a development that Bruchez calls "very significant" for three reasons. One, there are more equipment users and a larger market outside of Europe than within Europe. Two, non-indigenous IC manufacturers with European plants emulate the decisions of their fabs back home in making purchasing decisions. Three, European tool users will not purchase European equipment unless it has a large installed base and worldwide recognition. Before SEA was established, Bruchez notes that the involvement of non-European tool users in EC projects was scarce.

"The list of participants shows that non-European users are interested in European equipment and are actively participating in what they see as worthwhile projects which they can both contribute to and gain useful information from," Bruchez maintains, "not only on the equipment but also related materials, applications, and solutions."

One major success that SEA points to is a centrifugal cleaner for 200- and 300-mm wafer containers. Called the Milestone III, the system was developed by Dynamic Micro Systems (DMS) to clean FOUPs, 200-mm open cassettes, and wafer transport boxes. Evaluated at STMicroelectronics fab in Crolles, France, between August 1998 and July 1999, the Milestone removes hardened photoresist, CMP by-products, and metallic residues. According to an SEA results bulletin, the system is designed as a bridging tool that will make the transition from 200- to 300-mm processes cost-effectively.

The results from tests at STMicroelectronics and I300I in Austin, TX, found the system to have a very low level of contamination with multiple cleaning usage over the nine-month testing period "even better than at the outset of usage of the equipment." Complete cleaning cycles totaled 3177, or the equivalent of 1332 hours of operation with no breakdowns and no observable electrostatic discharge effects. The bulletin concludes that the Milestone III offers combined benefits of centrifugal force and optimized quartz halogen drying to provide high throughput and effective cleaning of particulate and ionic contamination without distortion of the containers.

Of the 43 projects undertaken under Esprit's aegis, 15 focused on 300-mm technologies. Eight of these are in beta tests at user sites, and seven are in alpha tests at equipment company facilities, Bruchez says. "We decided in 1996 to be cautious but flexible when defining a 300-mm SEA policy. At the time the market was buoyant, and 300-mm was being strongly urged by the users. We didn't want to push the equipment companies too far too soon by insisting on 300-mm equipment only. We also noted that double productivity gains could be gotten from improvements to 200-mm equipment...at lower financial risk, so we devised a phased approach."

A three-way approach enabled SEA to keep its hand in 300-mm development. The approach involved SEA300 projects that evaluated alpha tools at equipment company sites at lower budgets, while conducting higher budgeted assessments at a few user sites such as I300I and SC300. SEA also set up a silicon services center at Gressi CNET to provide structured wafers for process and measurement trials.

SEA's annual budget is not fixed, but depends instead on the number of projects presented and accepted each year. In the first four years of its existence SEA had a $100-million budget, half of which was provided by the European Community. Partners in the projects pitch in 50% of the cost for each project. Proposals for 2001 will be evaluated in November, and the selected projects will begin early next year.

Three of four projects recently begun in the new IST program have a defect analysis or yield improvement focus. Project APC 300 will examine incorporating the latest in situ diagnostic OEM equipment from ASI into an advanced subsystem for total monitoring of plasma etch parameters. The goal is to introduce real-time process control using the Hercules/APC sensor diagnostics at a modular level. Project Oxepicle will assess the Steag MicroTech 300-mm wafer cleaner for surface preparation before key process stages such as gate oxide low-temperature silicon epitaxy. Finally, Project Summit will look at the requirements for next-generation 300-mm in-line wafer surface analysis with integrated TXRF. The equipment supplier is GeMeTec.

Only a few disappointments have marred SEA's record of success, Bruchez says. One project "did not meet the users' aspirations. The other was a small company that went out of business for financial reasons during its equipment evaluation." He did not wish to identify the companies involved, adding that one of the SEA300 alpha projects did not start because the company deferred its 300-mm activities during the industry recession. It now says it will reapply to the program. Another firm had good technology but its tool's throughput was unacceptable and could not be improved for the targeted application. The company eventually decided to find a different market for the equipment.

"Overall," Bruchez maintains, "SEA's success rate has been very high."

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