ISMI 450 Program Moves to Albany, Expands Goals
By David Lammers
The 450-mm wafer program at the International Sematech Manufacturing Initiative (ISMI) is moving from Austin to Albany, armed with $20 million in New York state funding to advance the program’s goals, said program manager Tom Jefferson.
After several years of working out wafer handling and other standards at the Interoperability Test Bed (ITB) in Austin, the 450-mm program is moving – along with the rest of the ISMI consortium – to the CNSE Albany Nanotech Complex. The 450 program will occupy a portion of the 14,000 sq. ft. of space being renovated to accommodate ISMI’s needs.
The $20 million subsidy is being disbursed now, specifically in support the 450-mm program. The grant is a one-time affair, though future funding is a possibility, which Jefferson declined to discuss.
A large portion of the $20 million will be spent to expand the 450-mm wafer bank. Last year, the program largely used sintered polysilicon wafers to test out the mechanical handling systems developed within the ITB. This year, the program plans to expand the bank of single-crystal silicon to thousands of wafers, sending most of them out to equipment vendors. The program has single-crystal wafers on order from suppliers, and expects to start receiving them in the second quarter, he said.
“They are loans, free of charge” to equipment companies developing 450-mm tools. “The wafer bank is an example of how ISMI is making this a cost-effective transition,” Jefferson said. A single-crystal test wafer could be sent out to a company that might deposit a blanket film, for example, which would then be returned for measurement, and perhaps sent out again to a different link in the supply chain.
The program is working with wafer suppliers to gradually improve the quantity and quality of the wafers. “The more wafers we make, the more we learn. With the increased quantity we see an increase in quality, both for bare and processed silicon,” he said.
At the Semicon Japan show in December, the program discussed its early findings on localized defects, wafer flatness, roughness, edge profile, wafer sag, stress in the wafer, and other metrics. The program has metrology tools in Albany and elsewhere, and will step up the number of wafer tests as the quantity of wafers in the wafer bank increases.
The major operational change this year will be the shift to wafer processing. Jefferson said the number of companies involved in the program has tripled over the last 16 months, and as an example he pointed to Oxford Instruments (Oxfordshire, U.K.), which has developed a PECVD oxide tool adapted to 450-mm wafers. “Oxford is not the only company, certainly. Some choose to be public; others don’t,” he said.
An ongoing activity is getting the infrastructure in Albany ready for 450-mm capable processing equipment. A FOUP (front-end opening universal pod) washer has been purchased, and a stocker for the larger wafers is on order. Improved person-guided vehicles (PGVs) are being purchased. “We are busy putting together some of the factory infrastructure pieces to go along with increased operational skills,” he said.
Supplier companies are playing a role. After wafers were broken during shipment last year, Entegris Inc. developed a box-within-a-box shipping container – a multi application carrier, or MAC — that better fits the needs of the transport firms.
Jefferson said the work on wafer handling and other issues at the ITB resulted in a speedup in the SEMI standards-setting process. All of the major standards needed for a 450-mm wafer transition are done, with the exception of the prime or production-grade wafer standard – the SEMI M1 standard — needed for actual chip production.
Over the past few years the consortium has been working with the supplier community to develop the equipment productivity metrics (EPMs) for some 60 tool types, aimed at 32nm and 22nm design rules. That process continues, with the always-evolving EPMs regularly updated on the ISMI 450 program site.
“Our goal is to validate unit processing capability of the equipment, so we can put down a pattern, etch and so on, and can it do it reliably, as determined through marathon testing. We want to get to the point of complete equipment demonstrations, so ultimately device makers can make equipment selections for their pilot lines,” Jefferson said.
