Now, an electrical engineer, a biochemist and a geneticist are collaborating at the company they founded--NimbleGen Systems Inc. (Madison, Wis.)--to harness the diatom, as an alternative to lithography, to produce the intricate patterns on future semiconductors. By identifying the 75 genes, out of 13,000, that control silicon dioxide pattern formation in diatoms, the researchers hope to precisely control diatoms to pattern chips. Their results will be published in an upcoming issue of the prestigious "Proceedings of the National Academy of Sciences."

"By genetically controlling [diatom patterning], we potentially have a whole new way of performing the nanofabrication used to make computer chips," said Michael Sussman, a University of Wisconsin-Madison biochemistry professor and director of UW-Madison's Biotechnology Center.

Because the tiny animals are microscopically small, they can potentially assemble submicron patterns on semiconductors that exceed the limits of even the most optimistic predictions for future photolithographic techniques.

Sussman founded NimbleGen with electrical engineer Franco Cerrina and geneticist Fred Blattner, both at the UW-Madison, and successfully sequenced the genome of the diatom Thalassiosira pseudonana with oceanography professor Virginia Armbrust, of the University of Washington.

DNA sequencing chip

Using living organisms to pattern semiconductor chips is not as far fetched as it sounds, since the intricate patterns they produce are unique to each species of diatom, of which there are over 100,000 varieties. So far, NimbleGen has created a DNA sequencing chip to identify all the genes in this particular diatom species, but the same technique should be able to similarly sequence other species, as well. Since each diatom species has unique patterning capabilities, sequencing them will create a vast catalog of patterns that can be mixed and matched.

Diatoms build their shells by successively depositing submicron-sized lines of silicon dioxide, the most common insulator used in the production of semiconductor chips. By identifying the 75 genes responsible for silicon dioxide line production in diatoms, the researchers plan to turn on and off specific genes to precisely control the lines as they form atop future chips. Of those 75 genes, the researchers have already identified 30 (25 of which are completely new, with no similarities to previously known genes) that can be controlled by growing the diatoms in low levels of silicic acid, the raw material they use to make silicon dioxide.

Now the researchers are using the tedious process of trial-and-error to catalog the effects on pattern production as they turn on and off each gene. As the catalog grows, NimbleGen will also be able to improve the DNA chip they invented to sequence the first diatom species, hopefully allowing other species of diatoms to be sequenced more and more quickly, which in turn should grow their catalog of patterns more and more quickly. Nevertheless, it will be several years before EEs will have enough control over diatoms to attempt patterning real chips with the living organisms.