New CO2 process for higher-density microchip fabrication
|
LOS ALAMOS, N.M., July 2002 -- A process designed around the use of supercritical carbon dioxide (CO2) promises to enable the production of higher-density microchips than can be made today, allowing the global integrated circuit industry to achieve its goal of producing increasingly higher-density microchips. Developed at the Department of Energy's Los Alamos National Laboratory, it would also eliminate the need for vast amounts of water and many of the hazardous chemicals used in current fabrication processes.
Patented by the Laboratory, the technology now available to integrated circuit makers is licensed to SC Fluids, Inc. of Nashua, N.H., which is beginning to market it commercially in partnership with IBM.
The work won the 2002 Presidential Green Chemistry Challenge, an award that recognizes the development of new and cost effective, environmentally friendly technologies. Winners of the annual award, sponsored by the Environmental Protection Agency, are selected by an independent panel of technical experts from the American Chemical Society and represent innovative ways to significantly reduce pollution at its sources.
"While the integrated circuit industry is one of the most successful growth stories in history, currently available fabrication technology has built-in limits to growth that threaten the industry's growth objectives," said Craig Taylor, of the Laboratory's Applied Chemical Technology (C-ACT). "A principal limit to growth is existing technology that inhibits the volume production of microchip "feature" sizes smaller than 0.10 microns even though the market for chips is expected to demand a six-fold or greater reduction in size."
The existing process is based on use of large amounts of corrosive chemicals and huge quantities of distilled, de-ionized water – as much as six million gallons of water a day in some fabrication facilities – to bring the microchip from beginning to end of production. Taylor says the water/chemicals methodology loses its effectiveness as microchip features become smaller, because water cannot penetrate them to remove fabrication residue, particles and chemicals.
The researchers named the new technology SCORR –Supercritical CO2 Resist Removal.
Microchip manufacturing relies on photolithography, a process similar to photography, to define the shape and pattern of individual surface features. A film of photo-reactive polymer, known as a photoresist, is applied to the surface of a silicon wafer where a latent image is generated in the photoresist by exposure to light. The latent image is developed using hazardous chemicals leaving patterned areas of photoresist, which must then be rinsed to remove undesirable residues. Kirk Hollis, of C-ACT and a team leader on the project, noted that some of the features created by photoresist are 1,000 times smaller than a human hair and not penetrable by water because of water's classic propensity for beading on and not always penetrating small surface features – a physical property known as "surface tension."
SCORR eliminates the surface tension problem because it avoids use of water. Instead, it employs supercritical CO2, or carbon dioxide under pressure and at elevated temperatures, which has no surface tension. When mixed with a small amount of an environmentally friendly co-solvent, it forms a compound capable of deep penetration. Deep within the photoresist features, it causes the undesirable residue or particulate matter to swell, crack and detach, facilitating capture and disposal. The new technology also replaces end-of-cycle rinsing and drying procedures.
Environmental and worker safety advocates have been critics of the use of dangerous solvents and volumes of water used in the current microchip fabrication process. Together with being a cost-effective fabrication technology, the SCORR process uses a closed-loop system that recycles all of its components. As a result, no CO2 or co-solvents are released to the atmosphere. "We believe the application of this technology provides a unique and creative solution for the integrated circuit industry, a solution that addresses technical as well as environmental issues," Taylor concluded.
The research leading to SCORR was launched in 1998 with research development support from Hewlett Packard, now Agilent Technologies, Inc. In addition to Taylor, members of the Los Alamos research group are Kirk Hollis, Jerry Barton, Leisa Davenhall, Gunilla Jacobson, Laurie Williams and Jim Rubin).