Maskmakers and lithographers need to
08/01/1998
Maskmakers and lithographers need to "Just Do It"
Picking up a slogan dropped by sportswear manufacturer Nike, Kathy Cleveland of Lucent Technologies, says, "Our strategy has been to put together a `Just Do It` attitude. We have a mask utilization group that coordinates all mask purchases for all Lucent facilities, and also all of our foundries and partnerships."
A mask-technologist veteran, Cleveland participated in the 4th Annual Advanced Reticle Symposium. Her presentation of the near future of maskmaking and lithography captured the opinions expressed by many of the symposium`s attendees.
"At Lucent, we are introducing new technology faster. The mask has become a crucial part of our business. Mask manufacturability is a crucial part of total manufacturability," said Cleveland. Acknowledging the often joked about "maskmaker`s vacation" (i.e., the ease in technology when the industry went from 1? to 5? lithography), she said, "Now the push is really on to improve the mask. We can have mask problems that make products unmanufacturable."
Jim Nulty of Cypress Semiconductor added a degree of economy and expediency to the future role for maskmaking. "We see masks really helping us to extend our lithography tools; at $10 million a cell, anything we can do to keep it another generation, we are going to do it. But at the same time, we think we are going to need 130-nm [lithography] a year sooner than the roadmap target of 2003." Nulty expressed the immediate need for manufacturable and aggressive (i.e., fast cycle time) optical proximity correction (OPC). "Basically, drop-in OPC is what we need," he said.
 |
Nulty also expressed concern about the application of dry etch in maskmaking. "The industry still doesn`t know how to do it manufacturably. I have some concern about how the mask industry is going to start up so quickly, and start cranking out mask with dry-etch technology."
Much of the symposium`s content focused on application of OPC and phase shift masks (PSMs) to extend current lithography equipment-set lifecycles. Lars Liebmann of IBM Microelectronics stressed, "PSM manufacturing insertion is very much need-driven, not capability-driven. Manufacturing success of alternating PSMs may depend on the failure of 193-nm lithography. We have to start communicating."
Cleveland could not have agreed more on communicating. Interspersed with her check list (see table) for "just doing it," Cleveland said, "We have developed cross company and cross functions teams. For example, a very active critical dimension (CD) team that involves the strategy of seven different groups - technology developers, design library developers, designers, maskmakers, wafer technologists, wafer lithographers, and metrologists. We are doing real data analysis of the current capabilities of the mask vendors. And we believe that we need a very strong leadership empowered to make cross organizational decisions. That means that we have to educate our leadership."
Jim Northup of Photronics emphasized, "It is pretty easy to educate the purchasing guy and the lithography guy, but when you get back into the design group, that is where the real education is needed. It`s partly an inertia problem: Most design rules involve years of evolutions, a handshake, and a partnership that is somewhat different with every customer of every mask shop."
The concept of industry cooperation was also expressed in an answer posed by M. David Levenson, one of the originators of PSM technology and chief editor of Microlithography World magazine. He asked, "Where is the maskmaking industry going to get the funds necessary for the needed R&D?" Dave Markle of Ultratech Stepper, echoed other opinions, "No single mask shop will be able to come up with the funds. The answer is in partnership with tools suppliers, semiconductor manufacturers, SEMATECH, and the like. The [funding] question becomes extreme when you start to look at post optical [lithography] technologies." A true visionary in lithography, Markle also shared his views on the potential and advantages for maskless lithography technologies in the future, such as the emerging "step and squish" technique being worked on at several universities. "Step and squish" is the common term given to contact lithography with injected photoresists.
Cleveland added, "We do feel strongly that mask manufacturers `own` the capability and improvement issues. And it is an effort that has to concentrate on R&D and certainly a lot of practical investment."
One estimate generally agreed to at the symposium is that the cost of a mask could easily get to be 5% of the cost of a chip.
Cleveland`s summary? "We need to operate at the elbow: We want the most we can get, but the mask has to be able to be made and it has to be cost competitive and timely."- P.B.