Environmentally Safe Semi-Conductor Production

February 12, 1993

By Lynn Yarris, [email protected]

The word "fat" may be a source of vexation for millions of Americans, but for two researchers in LBL's Materials Science Division (MSD), it could prove to be the key to solving a serious problem for the semiconductor industry.

Employing what they call a "big ball of fat," materials scientist Edith Bourret-Courchesne and chemist John Arnold have developed a cost- efficient and environmentally benign technique for making semiconducting thin films out of notoriously toxic and volatile chemical compounds.

For more than a decade now, the semiconductor industry has been searching for a low-cost means of making thin films that have selected optical as well as electronic properties. Metal-chalcogen compounds (elements from the II and VIa columns of the periodic chart) possess the desired properties, but their production--through a technique called organometallic vapor phase epitaxy or OMVPE--has always meant working with metal alkyls and hydride gases. Because of their toxicity and volatility, these compounds are heavily regulated, hence very costly.

Bourret-Courchesne and Arnold overcame conventional production drawbacks by synthesizing their materials from single molecule precursors. These precursors are made by surrounding metal and chalchogen compounds with groups of hydrocarbon molecules--the ball of fat.

Says Arnold, "The hydrocarbon groups help to make the compounds volatile and, at the same time, provide the molecules with sufficient solubility for their use in growing thin films. Because we work with a single solid precursor instead of two gases, handling is easier and the toxicity is much lower."

The by-product that Bourret-Courchesne and Arnold get after pyrolysis of their precursor is a liquid organic compound that is easily trapped and disposed of with no hazard to the environment.

In addition to eliminating the environmental problems and reducing the high costs of cleanup associated with conventional production methods, the new technique also promises to yield higher quality thin films.

Says Bourret-Courchesne, "The technical difficulties are much greater when you are trying to decompose two sources simultaneously. In a single molecule precursor, in which the stoichiometry is defined at the molecular level, we have a better opportunity for controlling the composition and thickness of the films we grow."

To date, Bourret-Courchesne and Arnold have successfully made thin films of zinc telluride, zinc selenide, and cadmium telluride using organometallic chemical vapor deposition. They also synthesized a mercury telluride compound but it was so stable it proved difficult to pyrolize. The films produced were crude, the scientists say, but prove the single molecule precursor approach is "perfectly suited" for OMVPE.

"Our results strongly indicate that single molecule precursors can be designed for the whole family of metal-chalcogen compounds as well as other major classes of materials," says Bourret-Courchesne.

She and Arnold were especially pleased that their films were deposited at relatively low temperatures--ranging from 250 to 350 degrees Celsius. At these temperatures, the electronic and optical properties of the films should be optimal.

Bourret-Courchesne and Arnold have been awarded Laboratory Directed Research and Development programs for FY93 to continue their research. One of their next steps will be to build a new OMVPE reactor that will allow them to make thin films whose electronic and optical properties can be measured. They will also look into extending their technique to ternary materials, such as copper indium selenide, which could be used in solar cells.

Says Arnold, "No one has reported single-molecule precursors to mixed metal compounds that are this complex. Making a ternary single molecule source is a major challenge, although another approach might be to combine two single-source precursors, one or both of which would be a binary system."

The MSD researchers say their working together is almost as unique as the project they are working on.

"We were lucky to find one another," says Arnold. "The two fields have not worked together until relatively recently."

Says Bourret-Courchesne, "Usually chemists are busy making compounds and materials scientists are busy testing them to see what they will do with little or no interaction between the two."