David Lammers

Three Innovations to Watch
October 31, 2008

Innovation is the fourth great economic input, along with labor, capital, and machines, and the hardest to predict. Three innovations which may pay off involve the semiconductor manufacturing and equipment industries: new transistors, printed electronics, and single electron trapping (SET) medical devices.

Yale Professor T.P. Ma has suggested Unipolar CMOS, which could be a game changer. The idea rests on a design which uses electrons for both channels, rather than the faster electrons for the n-channel and the slower holes for the p-channel. If it works, Ma's idea would extend silicon’s performance roadmap, provide for higher densities due to shared contacts, and simplify manufacturing by not requiring additional steps and materials for the strained PFETs.

Also, Ma has a roadmap, going from silicon Unipolar CMOS to III-V based NFETs which would take advantage of the much-higher mobilities in InGaAs, perhaps InSb which Intel is investigating, and other compound devices. Will Ma find a backer, i.e. funding at Yale itself to take his idea to the prototype stage, or a chip company willing to see if his simulations prove out in reality? The potential benefits are too great to allow this idea to wither on the vine.

Printed electronics already are at the demonstration phase. Kovio Inc. (Sunnyvale, Calif.) has developed a series of inks, including a critical silicon ink, which can be applied with inkjet and other printing techniques. While the linewidths are an order of magnitude larger than leading-edge CMOS, the costs are much less. If RF and other circuits can be printed on cheap metal and other flexible substrates, it opens up a whole range of possibilities for embedded electronics. Kovio has demonstrated RFID chips and expects to take this innovation to mass production starting next year. Other markets that depend on reducing costs to pennies per chip, such as sensors, are out there.

At a recent seminar organized by SEMI Austin, another mind-boggling innovation was presented. Again, this is an idea that requires the semiconductor manufacturing infrastructure to support a technology that could expand to a multi-billion-dollar industry.

Louis Brousseau III, the president of Quantum Logic Devices (Austin, Texas) described his startup’s work to develop devices that seem almost Jules Verne-ish. Brousseau and his colleagues have developed prototype devices which can trap single electrons on a gate, thus modulating the flow of current. The devices are designed to detect specific biological molecules, such as tainted blood, DNA, or the protein markers which indicate the presence of cancerous tumors.

These quantum dot (QD) devices have dimensions <10 nm, and employ gold connections to the source and drain, making them difficult to fabricate in a conventional fab, Brousseau said. While he has been able to create prototypes at a small fab at the University of Texas at Austin, Brousseau is still searching for a commercial fab that will be capable of commercial production. Also, the small company perseveres despite the early death, by cancer last year, of its principal engineer David Metzer, who Brousseau said was blessed with a genius for “figuring out ways to get it done.”

Quantum Logic Devices has a contract with NASA’s Johnson Space Center to develop a PDA-sized detector that would be used to monitor metabolic changes in astronauts in space, where the g-forces and optical detectors don’t mix well. Again, the potential implications are huge. Brousseau said biologists now spend in the range of $500-$1000 per experiment on the diagnostic tools needed to detect important molecules. “We can create black box diagnostics that reduce the cost per experiment to cents, and eventually microcents,” he told the SEMI audience. 

Posted by David Lammers on October 31, 2008 | Comments (1)