DCU has just received equipment worth $1 million to help to find better ways to manufacture computer chips. Dick Ahlstrom reports
Modern microchips are getting so small that the changes happening on their surfaces can be a mystery. A research team at Dublin City University is trying to overcome this problem and to help manufacturers to build a better microchip.
DCU's Plasma Research Laboratory (PRL) heads the research effort and recently received a donation of equipment worth $1 million from the California-based Lam Research Corporation. The firm is a key supplier of wafer fabrication equipment to the semiconductor industry and hopes the PRL will improve its production systems.
DCU has developed particular expertise in the creation and use of plasmas through the PRL, which is part of the campus-based National Centre for Plasma Science and Technology.
A plasma in this context is the term used to describe a gas that has been stripped of its electrons to become ionised. The colourful lighting displays seen on the fronts of many shops are based on ionising argon or neon gas trapped in a thin glass tube using electricity.
Plasmas of bromine, chlorine and fluorine are now widely used to etch the surfaces of microchips, says Dr Bert Ellingboe, director of the PRL. "We do basic plasma physics but we are particularly interested in the interface between a plasma and a surface," he explains.
Plasmas are used by a wide range of manufacturers to change the surfaces of materials in a process described as "plasma surface modification technology". Biomedical implants such as artificial joints can be treated to toughen them but also to control how they react with the body. One treatment might encourage bone to knit into the metal while another might help prevent bone from building up on a surface.
Teflon can be made to coat plastic parts and plasma based manufacturing produces the reflective metal-coated windows used on modern buildings, says Dr Ellingboe. Finding new uses isn't the challenge however, it is understanding what happens between the plasma and the surface. "You actually need to know what you have done," he says. The PRL uses radio waves to produce plasmas that can then be contained in a purpose-built chamber. The radio wave knocks off electrons which then go on to knock off other electrons, producing a plasma in a chain reaction that persists for as long as the radio waves are applied.
Plasma technology is ideal for silicon wafer production because it is more precise but it also works better, he says. "We are able, by a large factor, to increase the reactivity of the gas which increases the efficiency of exposing it to materials." The three gases used in the process are highly reactive but become even more so when ionised.
"By producing a plasma, not only does the gas hit the surface randomly, but it also injects the wafer with high energy." The ions strike the surface with an extra kick of 100 to 500 electron volts, a measure of the strength of the impact. "That is sufficient to break chemical bonds on the surface," explains Dr Ellingboe.
EARLIER etching techniques relied on liquid chemical processes but these are nowhere near as accurate as plasma manufacture. Liquid etching produced bulb-shaped channels but plasmas are cut clean as though by a drill, he says. Liquid etching produced electronic junctions about 5,000 billionths of a metre across, but with plasmas this can be reduced to 70 to 90 billionths of a metre across.
Dr Ellingboe and his team are developing automated systems to help identify and correct any flaws that may arise during wafer manufacture. "We will use the equipment donated by Lam to better understand the plasma surface interaction and consequently to develop suites of diagnostics for the purpose of real-time plasma control," he says.
Dr Ellingboe used to be Lam's chief technologist for this equipment and in a sense the donation followed him from the US. He wanted to return to research in an academic setting and knew of the advanced plasma research being done at DCU. He agreed to come here to set up the PRL. He wanted a challenge and knew the Dublin team would give it to him: "I wanted people I could argue with".