Portland, Ore. — Using chemical deposition and other techniques, Georgia Institute of Technology researchers have demonstrated that microscopic algae — silicon-dioxide diatoms — can be converted into semiconductors without changing the intricate structure of the organism's shell. The ability to mass-produce the shellfish easily and form thin films of them in uniform arrays prompted the researchers to convert them into semiconductors.
"It's taken a long time to get the infrastructure together to grow our own diatoms and now to use gases and other means to change their electrical characteristics," said professor Kenneth Sandhage. "But now we are ready to test out any number of ideas as to how to make useful devices out of them." Sandhage is the director of the Biologically Enabled Advanced Manufacturing Center at the School of Materials and Science Engineering, and the Institute for Bioengineering and Biosciences at the Georgia Institute of Technology.
Nearly 20 percent of living things are 10-micron-diameter unicellular algae called diatoms — microscopic silicon-dioxide shellfish that float in every sea, lake, river and stream. Diatoms, which eat carbon dioxide and give off oxygen, generate 40 percent of the 50 billion tons of organic carbon in the sea.
They multiply by doubling, enabling enormous populations to grow quickly. By doubling their number every generation, diatoms can be grown in almost any amount necessary — 40 reproduction cycles yield 1 trillion replicas.
Sandhage's method first evicts the tenants, leaving only the intricately patterned silicon-dioxide shell with thousands of compartments inside, measured in nanometers. Ideally, the internal structure would be programmed by genetic engineering so that each 10-micron diatom shell would be about as smart as a microprocessor.
"Our next step will be to try to pattern arrays with them, as well as to solicit ideas on how to turn diatoms into useful devices," said Sandhage.
Today, the researchers are content to have proven that chemical vapor deposition could alter the silicon dioxide's chemistry enough to transform it into a semiconductor.
Diatoms naturally grow a silicon-dioxide shell that is very similar to the silicon dioxide used as an insulator for semiconductors, but in a detailed pattern that's more intricate and precise. And instead of the layer upon layer of silicon that characterizes current chips, the features of the diatom's shell are grown in three dimensions.
In addition, the diatom replicates quickly and with unparalleled nanoscale precision, so that the regularity of the grown arrays approaches the angstrom level.
"We believe the intricate 3-D internal structures make them vastly superior to trying to make 3-D structures by building up layer after layer, like semiconductor makers do today," said Sandhage. "There are thousands of different types of diatoms and many different ways readily available to make films from them, plus the entire genome for the diatom has been mapped."
The genome for diatoms was completely mapped only last year. In contrast to the 30,000 genes that humans have, diatoms have 11,500, which is 15 percent more than plants. Mapping the genome is the holy grail of genetic engineering, since theoretically all possible functions can now be turned on or off, enabling researches to try out a vast variety of different, repeatable patterns.
Sandhage thinks that early on, the diatoms may be used to pattern a silicon wafer with a nanoscale array. The diatoms would then be covered with silicon and later sacrificed. Now he is building a library of new functional chemistries that will preserve the 3-D structure while converting the diatom shell's function.
The ultimate goal of Sandhage's research group is to develop genetically engineered microdevices for applications in biomedical, computing, environmental cleanup and defense.