A start-up called Alphabet Energy in San Francisco says it has developed a chip able to serve as a more efficient thermoelectric sensor. The device is based on theoretical work done at the Lawrence Berkeley National Laboratory to lower the thermal conductivity of the sensor material and increase the amount of electricity it puts out using a smaller amount of heat.
The thermoelectric chip is fabricated using methods resembling those for producing integrated circuits. All in all, Alphabet Energy figures it can lower the cost of installing such systems to under $1 per watt. The company plans on a pilot test at an industrial facility next year where the thermoelectric devices will try to harvest heat from flu gases. Commercialization may start in 2012.
The thermoelectric chips use a technique wherein oxygen impurities get introduced into a special class of semiconductors known as highly mismatched alloys (HMAs). The result is high electric conductivity and low thermal conductivity An increase in thermopower has typically come at the cost of a less electric conductivity. HMAs get around this conundrum and are formed from alloys that are highly mismatched in electronegativity, a measurement of their ability to attract electrons. A small amount of doping in HMA material causes the partial replacement of anions with highly electronegative isoelectronic ions. Anions are negatively charged atoms and isoelectronic ions are different elements that have identical electronic configurations. Though the explanation in complicated, the end result is high electric conductivity in an electronic structure that greatly benefits thermoelectricity.
The exact details of the Alphabet Energy devices are sketchy. However, one might surmise they would greatly resemble those studied at the Lawrence Berkeley Lab. There, researchers started with the semiconductor zinc selenide and simulated the introduction of two dilute concentrations of oxygen atoms to create model HMAs. The results suggested, say Berkeley researchers, that highly electronegativity-mismatched alloys can be designed for high performance thermoelectric applications.
Berkeley researchers wrote about their HMA work here: http://prl.aps.org/abstract/PRL/v104/i1/e016602
Science Daily also wrote about the Berkeley work: http://www.sciencedaily.com/releases/2010/01/100127113755.htm
The Alphabet Energy site itself is pretty bare bones: http://www.alphabetenergy.com/