What's Beyond 193 nm?

Ruth DeJule, Associate Editor -- Semiconductor International, 10/1/1999

Over the past year, 157 nm lithography has attracted considerable interest as a potential successor to 193 nm technology. In a recent interview Will Conley, a member of Motorola's 157 nm program in Austin, Texas, discussed the viability of 157 nm technology.

SI: You said the easiest post-193 nm technology is 157 nm. What needs to be done?

Conley: There's a lot that needs to be done. Starting with resist technology, there are some ideas in the works, but there is no working single-layer 157 resist. So we need solutions for single-layer resist technologies, as well as bi-layer resists. And there are programs being started at various resist companies, and SEMATECH has sponsored a program in conjunction with the University of Texas.

SI: What about materials for mask and lens technologies?

Conley: In terms of mask technology, it looks like fluorinated fused silica. For the lens, most of the work is focusing around CaF₂. But things are so fluid. Right now, there are more questions than answers.

SI: Can this be done with the industry supporting several other post-193 nm approaches?

Conley: Virtually all the suppliers are getting on the 157 bandwagon. Most have stated publicly that they'll support other NGL technologies in addition to 157. SVGL has both 157 and EUV programs; ASML has 157 and SCALPEL. There are similar scenarios at Canon and Nikon. Support is coming from the large fabs that are starting to put money in preferred suppliers.

SI: What kind of timeline are you looking at?

Conley: It all depends on the end-user. Memory folks probably will introduce it at a larger geometry than the logic folks. I think logic will be looking at the 70, 60, 50 nm node.

SI: 157 is the newest of post-193 technologies. Do you think a manufacturable 157 process will be developed before EUV or SCAPEL?

Conley: I think it will be available before SCAPEL, but it's up in the air whether or not it will be available before EUV. It all depends on what happens over the next four to six months, on where companies decide to put their money. If there's more momentum behind 157, companies will pour more money into 157, and then it becomes a self-fulfilling prophecy that 157 will be ready before the other NGL techniques. We'll have to wait and see.

SI: How much impact is there in being first?

Conley: It has a huge impact. Whatever emerges first, of course it will be embraced sooner, giving it a head start.

SI: The industry is being stretched financially. Are there any solutions?

Conley: I think the solutions lie in extending 248 nm and 193 nm technologies. Look where we are right now. There are just a few 193 tools available. So fabs are continuing to push 248 with optical extensions like hard phase-shift masks, and the same thing's going to happen at 193. I think the push will continue. Some people are thinking if they can push 193 far enough, EUV will be ready, and they won't have to do 157. One of the things we're noticing is that a number of companies are reverting back to i-line lithography; they're pushing it further than ever for some back-end levels. Many of the fabs, especially MPUs and ASICs, are pushing KrF down to 130 nm design rules. If they can do this, then 193 probably will be introduced around 100 nm and pushed down to maybe 80 or 70 nm. By then 157 lithography and/or EUV will be ready.

SI: How much of 193 nm technology development can be carried over into 157 nm?

Conley: Some of the lens and mask technologies will carry over, and a lot of the resist technology. It won't be a 1:1 carryover, but some of the learning will be transferrable.

SI: Where does the industry need to be a year from now?

Conley: In a year from now, we need to have some benchtop 157 systems ready to expose new photoresists. We'll need to have some idea of the reticle technology, including a pellicle. We have no pellicle solutions at all for 157. Nobody knows what that material should be. Is it going to be pellicle-less? In a year from now, we need to have some idea.

SI: What does the industry need to do to get there?

Conley: The programs that are in progress, like those at SELETE and SEMATECH, will require more support from member companies. Also, many of the stepper companies are talking to their customers about supporting development programs. Cost sharing happened with both 248 and 193 technologies; this has become a necessary trend.