TravelCenters of America LLC planning to install 50 on-island DEF dispensers in 2011
DOE issues RFI on algae biomass supplies for advanced biofuels RD&D efforts

U Michigan team boosting efficiency of thermoelectric materials; potential of reaching 15-20% efficiency could result in practical applications

Example of the structure of a partially filled CoSb skutterudite. Source: C. Uher. Click to enlarge.

Researchers at the University of Michigan and Hong Kong University of Science and Technology are working to improve the thermoelectric (TE) capabilities of filled skutterudites to the point that they could improve the performance of thermoelectric materials from about 5% efficiency to about 15 or 20%, according to Massoud Kaviany, professor in the Department of Mechanical Engineering, and co-author of a paper on the work published in the journal Physical Review Letters.

At that level of efficiency, TE materials become useful in practical applications such as waste heat recovery in vehicles, Kaviany noted. Earlier this year, the Directorate for Engineering at the National Science Foundation in partnership with the US Department of Energy Vehicle Technologies Program issued a solicitation for proposals for thermoelectric waste heat recovery devices for automotive applications. (Earlier post.)

Skutterudites are one of several promising novel thermoelectrics—materials that convert a heat differential to electricity—that have been developed and pursued for more than a decade, notes Ctirad Uher, co-author of the PRL paper and author of a chapter on Skutterudite-based thermoelectrics in the 2006 Thermoelectrics Handbook. The basic skutterudite is a cubic structure of the formula MX3, where M=CO, Rh, Ir and X=P, As, Sb. Skutterudites are some of the highest mobility semiconductors, and have outstanding electronic properties. However, for thermoelectrics, their thermal conductivity is too high, Uher notes.

Efficient thermoelectric materials want to conduct electricity well, and conduct heat poorly. Filling the skutterudite voids with the right material can dramatically reduce the thermal conductivity, and optimize the thermoelectric properties.

In their work, the team showed that the phonon conductivity of filled skutterudites is greatly influenced by the topology of the filler species. (Basically, scattering phonons lowers thermal conductivity, earlier post.) The team identified certain configurations of barium that drastically increase the skutterudites’ thermoelectric efficiency.

We predict (ab initio) the phase diagram of BaxCo4Sb12 and find several stable configurations of Ba ordering over the intrinsic voids. The phonon conductivity predicted using molecular dynamics shows a minimum in the two-phase mixture regime, dominated by significantly reduced long-range acoustic phonon transport.

—Kim et al.

We knew that skutterudites are promising materials. But we did not know what features we could manipulate to maximize the conversion of heat into electricity. In this paper, we propose that certain configurations of the filler element barium will be very effective in lowering the materials’ thermal conductivity and thus increasing their conversion efficiency.

This is an important advancement in the sense that it provides guidance for the experimentalists to focus as they try to synthesize highly efficient thermoelectric materials.

—Ctirad Uher

Car companies are extremely interested in this technology, Uher said. The ideal environments for these materials are spots where large differences in temperatures exist. One such place is the pipe system of a car between the motor and the catalytic converter.

The research is funded by the University of Michigan’s Center for Solar and Thermal Energy Conversion, which is funded by the US Department of Energy's Office of Basic Energy Sciences.

We explained the physics of these materials for the first time. This will help to advance the development of these materials. If you are designing them based on fundamental physics and materials and not just by trial and error, then you need to know how they work so you can avoid haphazard experimentation.


  • Hyoungchul Kim, Massoud Kaviany, John C. Thomas, Anton Van der Ven, Ctirad Uher, and Baoling Huang (2010) Structural Order-Disorder Transitions and Phonon Conductivity of Partially Filled Skutterudites. Phys. Rev. Lett. 105, 265901 doi: 10.1103/PhysRevLett.105.265901

  • Ctirad Uher, Recent Progress in the Development of N-type Skutterudites, 2009 DoE Thermoelectrics Applications Workshop

  • Chapter 34. Skutterudite-Based Thermoelectrics Ctirad Uher, Thermoelectrics Handbook Macro to Nano; Edited by D. M. Rowe CRC Press 2006. Pages 34-1–34-17 doi: 10.1201/9781420038903.ch34



Might be a breakthrough with huge potential applications since low quality heat wastes are plentiful. So far I was not very positive on thermoelectric material because they are too inefficient, too fragile and too expensive for practical energy recovery. But 15% efficiency would definitively change that. Let's see if it is more than a lab prowess


Yup. Nothing like the lure of free energy.

But of course they are just;
" .. working to improve the thermoelectric (TE) capabilities . . to the point that they could [further?] improve the performance . . to about 15 or 20% . .".

That's at least a double wishy-washy.


From a mere 5% to 20% efficiency would represent a major jump. If it can be done cheaply, recovering some of the 60+% of the heat (energy) wasted in ICE could be very interesting. It could also make PHEVs more efficient. It could also be useful in Domestic, Industrial, Vehecular HVAC system.

Secondly, if it can be improved from 5% to 20% it could probably be improved again.


If this can be done cheapo, it could even give solar panels, Stirling and other CSP a run for their money.

Huge amount of heat can be easily produced even without concentrating the sun (simple black metal absorbers).


Heat sinks on electrical devices,any chimneys, Kitchen sink drains.
"If we look after the pennies, the pounds will look after themselves"

It certainly would be good to see 15 or 20% for starters.

Still cant see why a Crower cycle (5 stroke) steam application could be so difficult the ice efficiency should be able to gain as much as 30% improvement.


We may see heat recovery in buses and trucks sooner or later. You can only make an internal combustion engine just so efficient, you can help it with hybrid motors, but using the waste heat is a good idea.

The comments to this entry are closed.