ICT2014 Abstracts

The vast majority of thermoelectric materials research focuses on achieving high zT values. However, the zT of a homogenous material peaks in a narrow temperature range, which reduces the efficiency when the material is operated in a module under a large temperature gradient. Grading the sample properties along the heat gradient direction can shift the peak in zT in accordance with the temperature profile, and represent a promising complimentary technique for improving the efficiency of thermoelectric generators.

Material gradients inherent to directional crystal growth techniques provide a one-step route to functionally grade several state of the art thermoelectric materials. It has previously been shown that gradients in either band gap¹, or doping² can be achieved in this way. However, while a combination of the two effects is suggested to significantly improve the performance,² it has not yet been demonstrated.

We present a proof-of-principle study aimed at identifying and testing material systems that can be graded in both doping and composition, in such a way that these effects enhance each other, through either Czochralski or Bridgman/Stockbarger crystal growth.

Boron doped Ge_1-xSi_x (x= 0 to ~0.25) samples graded in both band gap and carrier concentration have been prepared by the Czochralski method. Along the length of the Ge_1-xSi_x samples x changes continuously giving rise to changes in the band gap from 0.87 eV to 0.65 eV. Similarly, gradients in the boron content results in continuous carrier density changes along the sample. This simultaneous grading of the band gap and the carrier concentration results in gradings in all material parameters relevant for thermoelectric performance.

¹M. Christensen, S. Johnsen, M. Søndergaard, J. Overgaard, H. Birkedal, and B. B. Iversen, Chem. Mater 21, 122 (2009).

²V. Kuznetsov, in Thermoelectrics handbook : macro to nano, edited by D. M. Rowe (CRC/Taylor & Francis, Boca Raton, 2006).