X-ray Lithography: Back for Round 2

In a recent Editorial space, I said that X-ray lithography as a mainstream next-generation lithography (NGL) technique was dead (see X-ray Lithography Is Just an NGL Dream, Semiconductor International , June 2002). A technique based in traditions of huge money- and space-eating synchrotrons, after all, really had no place in the future of commercial silicon processes, and anybody still dreaming that dream needed to, well, wake up.

I'm still not sure that any X-ray lithography technique will ultimately be able to overcome the stigma that has come to dog that end of the spectrum, and win the hearts necessary to make it a marketable NGL success. But at least one company seems to have a shot at it. JMAR Technologies Inc. (San Diego) has already made great strides in the practical development of a unique and viable NGL technique that overcomes the problems associated with traditional XRL.

It does away with the large, expensive synchrotron source in favor of a point-source system that is actually smaller than the DUV steppers in use today. JMAR's X-ray system overcomes significant cost concerns as well, with an expected price tag of $8M-$10M, well below those for EUV and even 157 nm lithography systems, according to Daniel J. Fleming, president of JMAR/SAL NanoLithography Inc. (South Burlington, Vt.), JMAR's lithography systems division (JMAR acquired X-ray system supplier SAL Inc. in August of last year).

	The CPL source field illumination and the demonstration of sub-100 nm lithography. (Source: JMAR Technologies)

They have also been making progress on the mask front, which has been an area of concern for X-ray lithography. "I won't say masks are a slam dunk," Fleming said, but he is confident that the company's roadmap will get them where they need to be, and he is convinced that X-ray masks will be the least expensive NGL masks available.

JMAR is currently assembling a beta version of its system. The company's roadmap calls for the first system, printing critical linewidths of 100-130 nm, to be available by next year. This system would have a single laser module, handling 150 mm wafers at 5 wph. By the third generation, JMAR expects to be able to print 300 mm wafers at 50 wph. Scheduled for 2006, this system would stack three laser modules, interlacing the pulses at the X-ray generator chamber for a higher average intensity.

Although the company has its work cut out for it to gain industry acceptance, it's taken the first step by shedding the X-ray moniker, following in the footsteps of current NGL darling soft X-ray — or rather, extreme ultraviolet (EUV) — lithography.

"Obviously, one of our concerns is the largely negative opinion the industry has for the synchrotron-based technology pioneered at IBM and continuing in Japan," Fleming said. JMAR recently changed the name of its particular technology from proximity X-ray lithography (PXL) to collimated plasma lithography (CPL) — its collimated laser plasma generation of 1 nm light being what differentiates it from other techniques in the X-ray regime. "While in the X-ray region, this light is highly absorbed in air and all other materials, making for easy shielding of the emissions that leave the wavelength conversion chamber," Fleming said.

While the name change is little more than a marketing technique, the strength behind JMAR's chances for the big time lie behind its technological developments. The company's short-term goal for its CPL technology is the patterning of high-performance GaAs devices produced on 150 mm wafers, with the vision extending to 50 nm features printed on 300 mm silicon wafers.

For additional information on lithography, go to www.semiconductor.net/lithography.