Nanostructured Thermoelectric Materials



Thermoelectric energy conversion devices offer cooling and electricity generation capabilities in compact, solid-state devices. The efficiency of these devices, however, has not been sufficiently high to propel this technology from niche applications to more ubiquitous refrigeration and waste heat conversion applications. The problem is fundamentally a materials limitation that is apparent in the basic materials figure-of-merit, ZT. The dimensionless quantity ZT is the product of the electrical conductivity, the temperature, and the square of the Seebeck coefficient, divided by the thermal conductivity. To increase ZT substantially, one must introduce "defects" (e.g., interfaces, barrier layers or scattering centers) in the material that enhance the power factor (the numerator of ZT), while suppressing the thermal conductivity (denominator). Without engineering these defects carefully, the effect will likely be to reduce both the electrical and thermal conductivity proportionately - a result that will not improve ZT.

Ideally, we would like to use quantum confinement to increase the density of states at the Fermi level, using these same interfaces to selectively scatter or confine phonons (lattice vibrations) to reduce the thermal conductivity.

The HEMI group, along with the Gronsky group (Berkeley MSE) and the Stacy group (Berkeley Chemistry) have joined efforts to form the Berkeley Thermoelectrics Research Group. We are part of a DoD MURI on Quantum Structures for Thermoelectric Applications that is led by Professor Gang Chen at UCLA. Also participating in the MURI are groups from MIT, UCLA, JPL, Hi-Z Technologies, Inc., and Marlow, Inc. The Berkeley Thermoelectrics Research Group focuses is engaged in the following research efforts: Recent accomplishments include the first epitaxial growth of skutterudite (e.g., CoSb3) thin films, and the development of an electrochemical method for the synthesis of dense thermoelectric nanowire arrays with wire diameters down to 20 nm. This latter project is based on the well-known process for producing self-organized pore arrays in alumina by anodization of aluminum. The schematic diagram at left below shows the resulting double-layer oxide. The figure at right is an SEM image of porous alumina produced by this process at Berkeley. Pore aspect ratios (length divided by diameter) can be as high as 1000.



These alumina membranes are metallized and the barrier oxide is removed to create templates for the electrodeposition of Bi, Bi-Sb, CoSb3 and Bi2Te3 into the pores. Since the electrodeposition only proceeds on conductive surfaces, the nanowires nucleate at the pore bottom. All nanowires are thus electrically continuous as deposited. The SEM figures below (courtesy M. Sander) show low and high magnification images of the filled templates - crystalline Bi2Te3 in this example. Note that approximately 90% of the pores are filled (the alumina template is medium grey, the Bi2Te3 wires are light grey, and the black regions are pores that were not filled).





Publications





Last updated 6/19/02 - TDS

back to HEMI front page | back to HEMI Projects page | back to top of this page