200-year-old discovery offers breakthrough in thermoelectric sustainability.

Researchers at the National Institute of Standards and Technology, and collaboration with the University of Colorado, have created a novel method using gallium nitride that could drastically increase the conversion of heat into electricity. If perfected, this technology could help recoup some of the roughly $100 billion of heat energy lost in the U.S. each year, according to NIST.

The new heat-to-electricity findings build upon an intriguing phenomenon first observed 200 years ago by German physicist Thomas Seebeck.

While studying magnetic fields observed by earlier famed physicist Hans Ørsted, Seebeck found that a junction of dissimilar metals would produce a deflection on the needle of a compass when exposed to a temperature gradient.

Upon witnessing this phenomenon, he recognized that the reaction of the compass occurred due to the temperature difference, which induced a voltage between the two regions, causing a current to flow from the hotter region to the colder one.

The current created a magnetic field that caused a reaction of the compass needle.

Using what is now called the Seebeck effect, scientists have long predicted it could be an ideal way to recycle heat energy that would otherwise be lost.

Despite this possibility, there has always been a major obstacle – a material must conduct heat poorly to maintain a temperature difference between the two regions yet conduct electricity extremely well to convert the heat to a significant amount of electrical energy to make it worthwhile.

For most materials, heat and electrical conductivity go hand in hand; a poor heat conductor makes for a poor electrical conductor, and vice versa.

Approaching the experimentation with the foreknowledge established by Seebeck, University of Colorado researcher Mahmoud Hussein discovered that the property of thermoelectric conversion could be decoupled in a thin membrane covered with nanopillars – standing columns of material no more than a few millionths of a meter in length, or roughly one-tenth the thickness of a human hair.

It was this finding that led to the collaboration between the University of Colorado and NIST researcher Kris Bertness.

This discovery ultimately led to the fabrication of a technique developed by the collaboration, which comprises hundreds of thousands of microscopic columns of gallium nitride atop a silicon wafer.

Microscopic astroturf

Publishing their findings in "Advanced Materials" on March 23, the paper outlines that the interaction between the pillars and the silicon sheet actually slows the transport of heat in the silicon, essentially allowing more heat to convert to electric current.

Using the nanopillars, Hussein, Bertness and colleagues were successful in uncoupling the heat conductivity from electrical conductivity by using a silicon sheet – a first for any material and a milestone in itself for enabling efficient conversion of heat to electrical energy.

Continuous adjustment allowed the researchers to reduce the heat conductivity of a silicon sheet by roughly 21% without lowering its electrical conductivity or changing its Seebeck effect.

In silicon and other solids, atoms are constrained by bonds and thus cannot move freely to transmit heat. This is practically a law in the physical universe. As a result, the transport of heat energy takes the form of phonons instead – the collective vibrational movement of the atoms.

Both the gallium nitride nanopillars and the silicon sheet carry phonons; however, those within the nanopillars are standing waves, pinned down by the walls of the tiny columns, much like how a guitar string, once plucked, will vibrate but hold fast end-to-end.

This interaction between the phonons traveling in the silicon sheet and the vibrations in the nanopillars slows the traveling phonons, making it harder for heat to pass through the material, thus reducing its thermal conductivity.

Just as importantly, the phonon interaction accomplishes this feat while leaving the electrical conductivity of the silicon sheet unchanged.

With the fundamental technique established, the team is now working on structures fabricated entirely of silicon and with a better geometry for thermoelectric heat recovery. Once perfected, the scientists envision these sheets could be wrapped around steam or exhaust pipes to convert heat emissions to electricity that could power nearby devices or be delivered to a power grid.

Another potential application would be the cooling of computer chips.

Finally, the researchers expect to soon demonstrate a heat-to-electricity conversion rate high enough to make their technique economically viable for industry.