Ready or Not, When Will 157 Really Be Needed?
Aaron Hand, Managing Editor -- Semiconductor International, 5/1/2001
 |
The Photolithography Technical Group Meeting at the recent SPIE Microlithography conference brought together 10 industry players to address the issue of what is needed to get 157 nm lithography ready on time. Although panelists discussed concerns involving resists, inspection capabilities, process optimization, pellicles and optical materials, the outlook was relatively optimistic. At least some of that optimism comes from the assumption that 193 nm (ArF) lithography will be extended in much the same way that 248 nm (KrF) lithography has been (see "KrF Lithography: Not Finished Yet," Semiconductor International, February 2001).
The panelist predictions ran the gamut from "definitely ready" to "definitely not." But the general feeling seemed to be that — although they were certainly trying to get 157 ready, and they were making good progress — the chances are that they will be relying on ArF lithography as a backup. By 2005, the industry will probably still be using ArF lithography, but with extremely low k values, predicted Jim Wiley, technical director of KLA-Tencor's Reticle and Photomask Inspection Division (RAPID, San Jose). Although both Kurt Ronse, director of lithography for IMEC (Leuven, Belgium), and Christian Wagner, who heads the system engineering division at Carl Zeiss (Oberkochen, Germany), expressed hope for 157's readiness, they expected the industry to have to rely on 193 for a little longer.
Burn Lin of Taiwan Semiconductor Manufacturing Co. (TSMC, Hsinchu, Taiwan) went so far as to say that it might not be worth using 157 nm technology at all. It might be better, he said, to extend 193 nm lithography and concentrate on developing extreme ultraviolet (EUV) techniques to follow it up with. Yan Borodovsky, director of advanced lithography for Intel Corp. (Santa Clara, Calif.), noted that the timing for EUV lithography also continues to be pushed further out, leaving chipmakers no choice but to rely on 157.
Nikon (Tokyo), for one, is keeping its options open. Masaomi Kameyama discussed extending ArF with 0.8-NA lenses and alternating phase-shift masks (PSMs), as well as the company's involvement with e-beam projection lithography (EPL). For 157, he said, problems remain with resist outgassing, environmental control, CaF2 (or BaF2) procurement, and pellicles, to name a few.
Among other problems with pellicles, the protective covers make inspection difficult, noted KLA's Wiley. Current techniques cannot effectively inspect through the thick quartz pellicle, he said, without corrections. He recommends developing a solution that uses either no pellicle or a removable pellicle. Asahi Glass has been working on a pellicle made of modified fused silica, according to Shinya Kikugawa. Gravitational bending is the most critical issue, he said, but a thicker pellicle (800 µm) reduces the bend to <10 µm.
Zeiss, working closely with ASM Lithography (Veldhoven, Netherlands), is studying the lens options for the 70 nm technology node. Looking at how CaF2 affects imaging performance, Wagner said that they need to achieve a 50% improvement over a two-year period. "It's possible, but we'll have to work hard," he said. In response to a question, Wagner said he thought it was reasonable to say that CaF2 would be available in the quantity and quality necessary by 2004.
The big question, according to Will Conley of Motorola and International SEMATECH (Austin, Texas), is whether the industry can really afford to implement 157 nm lithography. One economic consideration he pointed out is the fact that different resist processes have to be designed for specific applications. To make 157 nm technology work, he said, there needs to be extensive cooperation among industry players. Joseph Gordon of DuPont Photomasks Inc. (DPI, Round Rock, Texas) reiterated the cost-of-ownership issue, also calling for increased partnering throughout the supply chain.
From the chipmaker's perspective, Intel's Borodovsky said, "If 157 nm technology will be available in time (for the 100 nm node), we will insert it. I'm not sure it will be."
Also participating in the discussion were Akiyoshi Suzuki of Canon Inc. (Tokyo), with Mircea Dusa of ASML's U.S. Technology Development Center (San Jose) as chair, and Nigel Farrar of Cymer Inc. (San Diego) as co-chair and moderator.
For additional information on lithography, go to www.semiconductor.net/lithography
