INDUSTRY NEWS

University-backed project shoots for "green" cleaning breakthrough

Backed by a University of California program, a team of researchers is inching closer to the commercial introduction of an environmentally friendly tool for removing both organic contaminants and surface particles from wafers and advanced photomasks.

Researchers from UCLA's chemical engineering department are working with Beta Squared, a Texas-based vendor of plasma etch and wafer cleaning tools, to develop the system. It combines nonthermal atmospheric-pressure plasma jet (APPJ) technology discovered at Los Alamos National Laboratories and an in situ dry plasma cleaning technology for removing submicron-sized particles using inert gases, mechanical vibration, and plasma forces.

Trailblazing: This nitrogen plasma-source prototype at UCLA lab holds promise as economical, "green" method for cleaning substrates. Photo courtesy of UCLA Plasma Process Engineering Laboratory.

Called Plasmax, the latter technology is being developed by Beta Squared and Los Alamos. The technology was invented and demonstrated by UCLA and the DOE laboratories. The university is working with Beta Squared to develop plasma process applications that operate at atmospheric pressure. This design would eliminate the need for the vacuum components and assemblies used in traditional plasma-based systems.

Under the terms of a cooperative R&D agreement signed early this year, Beta Squared, UCLA, and Los Alamos will combine the APPJ technology and Plasmax to create a single tool capable of cleaning all kinds of contaminants. The goal is to have a prototype available for beta-site tests by the end of 1999, says John Festa, vice president of Beta Squared's plasma division.

The project is one of the first to receive funds from the UC Smart program. Established in 1998, the program matches semiconductor manufacturers and tool suppliers through the University of California system to support R&D partnerships. The program is designed to foster basic research through multidisciplinary approaches. UCLA received an award for the project after submitting a proposal.

Beta Squared requested one year of funding for the cleaning project, says Robert Hicks, a professor of chemical engineering at the university. Hicks, who oversees one postdoctoral scholar and one graduate student working on the research, says UCLA has received more than $1 million in Smart project funding for several projects. None of the participants would disclose the project's funding level. However, Hicks notes UCLA "is chipping in almost as much as Beta Squared. It's nearly a 50-50 match."

This way: Schematic shows design of atmospheric pressure plasma jet cleaning tool.

"The project end date is a little fluid," says Hicks. "The funding period is until August 15, but it could be extended to December 31. We always have the option to renew the agreements in order to do more work."

UC Smart is Hicks's brainchild, although he no longer is involved in administering it. The program's primary benefit is that it "has both an educational and a research component." Chipmakers "view training of students as important, and the industry likes that part a lot." One goal is to help train graduates "in all levels of education in the needs of the equipment and materials industry." One of the program's supporters is Applied Materials. A postdoctoral candidate working on the Beta Squared project, for example, "has since gone back and taken a job at Balzers," Hicks points out.

UCLA began working with Los Alamos in 1995 on the cleaning technology, he says. The DOE laboratories in Albuquerque noted that the technology efficiently removed surface radioactive contaminants at DOE sites. "The government was very excited about our technology," relates Hicks, adding, "It became clear that this technology could be used for semiconductor processing." UCLA is coordinating the joint development program.

"A long time ago at UCLA we were thinking about the idea of developing a dry process for cleaning," Hicks notes. "Our department has long been associated with environmental research and environmentally benign technologies, and we knew back then that the DOE and industry were very interested in technologies that could eliminate solvent-based cleaning operations using water and organics." That goal of pursuing "green" R&D is written into UC Smart's charter, he says.

As for the process itself, "there is one breakthrough aspect to it," Hicks maintains. "This plasma process can operate at atmospheric pressure and at a low temperature at the same time. You can treat material at a low temperature but with a very reactive gas species, allowing you to do cleaning applications that you couldn't do otherwise."

Lucent Technologies is developing the technology for use with Scalpel, its advanced electron-beam lithography program. Beta Squared claims the system can clean both wafers and masks in 25 seconds or less. Festa says the 25-second claim is based on a comparison "with what he knows" of other green dry cleaning technologies, such as the Radiance Process, which uses either a deep-UV excimer, an Nd:YAG, or a CO₂ laser and an inert gas flow to mechanically remove the contaminants dislodged by the laser from the substrate surface. The time element is also based on typical wet cleans, entailing a series of baths involving acids and DI water, Festa explains.

The Beta Squared technology "is a true in situ process," the company executive asserts. Tacked onto an etch or deposition process, it could drastically reduce a chipmaker's overall manufacturing time, he adds.

"Our goal at Beta Squared is to merge Plasmax with the APPJ 'jet'. . . to have an all-encompassing cleaning solution. . . that removes particles, organics, and photoresists with one system," says Festa. The main benefit of such a system is that "it's totally environmentally benign, as opposed to the incumbent technology or to aqueous cleaning technologies, which literally use millions of gallons of water and acids per fab annually. The state of California for one is making a large effort to reduce the pollutants generated by the semiconductor industry."

Festa estimates the market for substrate cleaning at "somewhere around $2 billion a year," depending on how one defines it. "It's becoming a larger market as the industry shrinks die geometry sizes and increases the substrate size. The amount of cleans per device is increasing. Sematech estimated that between 30 and 35% of manufacturing steps will be, or are, clean steps. And if you look at high-end devices with up to 300 manufacturing steps, 80 or 90 of those are cleaning steps." Festa notes that a typical killer particle "is half the size of a geometry, so that at 0.18-µm linewidth, a killer particle is 0.09 µm."

After examining the existing cleaning technologies, Lucent chose to work with Beta Squared for its next-generation Scalpel project, and Bell Laboratories is in the late stages of its own mask-cleaning project. Says Festa: "We are looking at possibly integrating Plasmax and our cleaning technologies into lithography steppers and possibly metrology tools. The more systems you can integrate without exposing the mask or the wafer to the cleanroom atmosphere, the better."

Ignition: Like the nitrogen prototype, this oxygen plasma source operates at 100°C. Photo courtesy of UCLA Plasma Process Engineering Laboratory.

The company, which is based in Allen, TX, is making "a lot of efforts to merge multiple applications at present," says the executive. "I think if you look at what Applied Materials is doing, you can see the direction the industry is headed in."

Festa sees the Radiance Process as a possible competitor with Beta Squared's process. "In terms of the technologies being developed I would consider that a competitor, although we feel we have an advantage." The edge is cost of ownership, according to Festa, who argues that the Radiance system "is a standalone tool with an excimer laser and a large footprint." He reasons that chipmakers are trying to reduce the amount of space taken up by equipment in cleanrooms, because, at a cost of $3000 per square foot, manufacturers need to reduce fab size and lower their overhead. Integrated into process tools, the Beta Squared technology would not take up any more room than the tool already requires, Festa argues.

That claim, however, misses the mark, replies Donna Bethell, president of Radiance Services. "It's not true we're limited to a stand-alone system." She points out that the Radiance concept can be integrated into a cluster tool and therefore "not increase footprint." Certainly, she adds, the Radiance Process requires much less space than a typical wet-process station. As for the laser, it can be placed anywhere, even outside the cleanroom area, provided that the light reaches the substrate.

"The idea would be to integrate us into the tool," Bethell continues, "to have two cleaning methods in one cleaning system. We're quite likely complementary. I don't think there's anyone who has come up with a method to clean everything. They may be a competitor," although, she notes, "we're not making equipment, so we're not competing with them on that score."

In fact, she concludes, the Maryland-based firm would "be happy to complement Beta Squared. Put us together, and we can clean everything."

While Beta Squared and Radiance sort out their differences, if any, chipmakers probably are pleased to have two environmentally friendly and economical cleaning technologies to choose from. Hicks says that early discussions with Gary Selwyn, a principal investigator at Los Alamos, focused on making the technology a low-cost alternative to traditional wafer cleans. "We said it would be a great thing, would be a low-cost [system], and to do low cost you had to get rid of vacuum chambers." Selwyn "had an idea of how to do that," recalls Hicks. Working as a team, "we actually came up with devices that can do the job, and we're really excited we were able to do it." It was Selwyn's idea to work in atmospheric pressure, says the professor. "There were many years of engineering involved. . . in how to get this thing to work on large-scale materials such as the 8-in. wafers relevant to the semiconductor industry. It wasn't clear then that that could be done."

MicroHome | Search | Current Issue | MicroArchives
Buyers Guide | Media Kit