NEW YORK--Chip makers may have made a firm commitment to reduce their perfluorinated (PFC) emissions, but the industry and its suppliers face major technology challenges on several fronts before they will be able to meet the environment goals.
"The WTO (World Trade Organization) pact represents an aggressive goal," concedes Chuck Fraust, San Jose-based director of environment health and safety for the Semiconductor Industry Association. "The suppliers must provide large-scale and basic equipment and new gas chemistries," he points out, "and this will not happen overnight."
The World Semiconductor Council (WSC) agreed recently to reduce PFC emissions by at least 10% below 1995 levels by 2010. This pact is in line with increasingly stringent EPA regulations and a growing public concern over greenhouse gas emissions in the U.S.
The EPA maintains that PFC gases emitted by the chip industry contribute to global warming with atmospheric lifetimes ranging from 2,000 to 50,000 years. Chip makers use the gases for plasma etching thin films and cleaning chemical vapor deposition (CVD) tool chambers.
The government agency has been after the semiconductor industry for some time. "In 1994 the EPA wanted us to reduce PFCs," acknowledges Walter Worth, an environmental safety and health program manager for Sematech in Austin. "While as an industry we represent a small amount of the emissions, the EPA mandates that all players do their share."
The semiconductor industry has "set a stake in the ground," maintains Jim Snow, a research center manager of the Microelectronics Gas Process Division of Millipore Corp., in Allen, Tex. "As an equipment supplier," he says, "we have to work enough ahead to be in touch so when chip makers bring in new gases for manufacturing, we have to have new mass flow controllers (MFCs), filters, and purifiers ready."
To reduce PFC emissions, companies are working on a wide variety of tools. This includes new abatement processes, recycling, new chemistries, and boosting the efficiencies of new processes. "Conventional etch chemistries combined with optimized abatement technologies will coexist with advanced chemistries for some time," predicts Christophe Fontaine, vice president of strategic development for Air Liquide Electronics in Paris.
"Remote NF3 (nitrogen triflouride) plasma cleaning appears to be the optimal cleaning solution today despite its high cost and the difficulties in properly handling its exhaust byproducts," Fontaine says. "ClFx or BrFx (chlorine or bromine fluorides) chemistries remain strong potential solutions in terms of process effectiveness, cost, throughput, and environmental aftereffects," he adds, "provided that certain safety concerns in the fab operation are first resolved."
For the silicon oxynitride etch chemistries, "it seems indispensable to maintain a carbon-based, fluorine-oxygen-hydrogen chemistry in order to reach the process required etch rates, selectivities, and aspect ratios," Fontaine says. This is still true, he says, "despite a huge amount of recent development work on alternative etch chemistries."
Some recently developed etch chemistries, such as C4F6 (hexafluorobutadience) or chemically similar compounds, appear to satisfy both the process criteria and the global warming gas reduction requirement, points out Fontaine. "However, the extra costs involved by such alternative chemistries in comparison to conventional techniques seem unacceptable to satisfy the productivity roadmap," he says.
At the same time this development work is going on, equipment suppliers also are developing abatement solutions and processes to handle the new gas mixes. For abatement, suppliers are developing new purification schemes. They will eliminate PFC gases from gas streams, redesign filtration systems, and simplify processes by going to more calibrated mass flow controllers (MFC) to provide accurate delivery of the new gases.
It won't be easy to develop these new systems, notes Millipore's Snow. They pose development challenges on several fronts, he says. "Since the new gases are inherently low pressure gases, there is an issue of condensation at minimal pressures," he notes. "In purifiers, sometimes you have problems with gases becoming liquefied which can mix or react with process gases."
But commercialization of new PFC reducing processes and equipment should be coming in the near future, Snow predicts. "C4F8 (octofluorocyclobutane) and other gases and equipment should be commercialized and in operation before the end of this year," he says.
The phase out of PFCs will correspond largely to the startup of 300-mm wafer fabs, says Shaunna Sowell, vice president/manager of worldwide facilities for Texas Instruments Inc. in Dallas. "By 2010," she says, "the majority of chip production will be 300-mm units which will more than just meet the emission standards' goals." Already, she says, current equipment for cleaning chemical vapor deposition gear has reduced PFC gases by 90% for 300-mm processes. But for 200-mm processes, "TI has decided to install point-of-use abatement for PFC emissions with all 200-mm tools we install."
Air Liquide is developing PFC-reduction processes of its own. "Our patented technology is based on the same principle as air separation membrane enabling N2 (nitrogen) or 02 (oxygen) to be extracted from the air," Fontaine points out. While the R&D investment is high for the beta testing and providing these gases and equipment to reduce PFC emissions, the costs will be collectively shared by the industry internationally, says the SIA's Fraust. "If everybody does what they have agreed to do, then the extra costs will be equally shared. There is a level playing field," he adds.
It will take a close working relationship between equipment and gas suppliers and chip makers, Fraust points out. "There is a shared responsibility. This involves chip makers meeting with their key suppliers to communicate their shared needs."
This development work is now being done on-site by chip producers, gas suppliers, and equipment makers. "The technical performances of our PFC reclaim technology have been beta tested at several sites worldwide to test capture and separation efficiencies as well as uptime and reliability," Air Liquide's Fontaine says. "Our solutions are developed and beta tested jointly with semiconductor manufacturers and OEMs."
At the same time, Millipore is collaborating with the research teams of the two
largest gas suppliers worldwide--Air Liquide and Air Products--as well as the major chip makers and OEMs, Snow says. "Everybody is doing what it takes to get these new gases and processes validated."