Modeling study of thermoelectric SiGe nanocomposites

Abstract

Nanocomposite thermoelectric materials have attracted much attention recently due to experimental demonstrations of improved thermoelectric properties over those of the corresponding bulk material. In order to better understand the reported data and to gain insight into transport in nanocomposites, we use the Boltzmann transport equation under the relaxation-time approximation to calculate the thermoelectric properties of n-type and p-type SiGe nanocomposites. We account for the strong grain-boundary scattering mechanism in nanocomposites using phonon and electron grain-boundary scattering models. The results from this analysis are in excellent agreement with recently reported measurements for the n-type nanocomposite but the experimental Seebeck coefficient for the p-type nanocomposite is approximately 25% higher than the model's prediction. The reason for this discrepancy is not clear at the present time and warrants further investigation. Using new mobility measurements and the model, we find that dopant precipitation is an important process in both n-type and p-type nanocomposites, in contrast to bulk SiGe, where dopant precipitation is most significant only in n-type materials. The model also shows that the potential barrier at the grain boundary required to explain the data is several times larger than the value estimated using the Poisson equation, indicating the presence of crystal defects in the material. This suggests that an improvement in mobility is possible by reducing the number of defects or reducing the number of trapping states at the grain boundaries.

Comparison of thermoelectric properties of nanostructured Mg2Si, FeSi2, SiGe, and nanocomposites of SiGe–Mg2Si, SiGe–FeSi2

“…3,4 The following experiments demonstrated the role of nanostructuring on increasing the figure of merit both in thin films 5 and nanostructured bulk materials. 6 Although nanostructuring of some bulk materials such as BiSbTe, 7 Si, 8 SiGe, 9 and PbSrTe 10 showed significant improvement in ZT, similar trend was not observed in some other materials like Mg 2 Si, 11,12 FeSi 2 13 and MnSi 1.7 . 14,15 Multiphase composite materials have been also considered for thermoelectric applications.…”

On the tuning of electrical and thermal transport in thermoelectrics: an integrated theory–experiment perspective

“…117 Besides the focus on κ L , calculations were also made to estimate ZT values for some composite systems such as SiGe composites, nanotubes and quantum dot superlattices. [120][121][122] A great progress has been achieved in reducing the κ L by scattering phonons over a wide range of frequencies. Advances in computational methodologies have been instrumental in promoting the understanding of the lattice dynamics in TE materials as exemplified by prolific studies of PbTe.…”

Frequency-Dependent Monte Carlo Simulations of Phonon Transport in Nanostructures

“…Because Henry and Chen only provided phonon-phonon scattering rates from 300 to 1000 K for our simulations, the lattice thermal conductivity at 1173 K is linearly extrapolated from simulation results for 973 and 773 K. Slight inaccuracy is expected in this treatment due to the weak temperature dependence of lattice thermal conductivities in 10-nm-grain-size samples. For charge carrier transport inside nanocomposites, we have developed a model based on the BTE under the relaxation-time approximation (Minnich et al, 2009b). The total relaxation time for all the scattering mechanisms is obtained by adding up the scattering rates i using Matthiessen's rule 1 − = 1 i i − ∑ (Lundstrom, 2000;Chen, 2005).…”

“…On grain interfaces, charge carriers are scattered by a potential barrier created by charges trapped on the interfaces. The grain interface scattering of charge carriers is included into our model by identifying a scattering potential and calculating the corresponding scattering rate, which is then added to the scattering rates of other two mechanisms (Minnich et al, 2009b). Assuming the parameters in Table 1 are unchanged for 10 nm grain sizes, we use the same model to predict e k and power factors 2 S (Fig.…”