| How can we boost thermoelectric properties of oxides by nano? |
| Yoshiaki KINEMUCHI |
| National Institute of Advanced Industrial Science and Technology (AIST) |
|
Harvesting energy with low exergy, such as wasted heat with low temperature, is fundamentally challenging. As with thermoelectric (TE) energy conversion, ¡°nano¡± is regarded as one of the key solutions for the challenge. Among the TE materials, oxides provide versatile platform for nanostructuring, which may contributes to the efficiency of energy conversion of TE.
In TE, expected outcomes of nanostructuring are low phonon thermal conductivity, high thermopower and low electric conductivity. The former two effects raise the efficiency, while the last one deteriorates the performance. Thus the tuning or decoupling of the effects is crucial, yet it is not established so far. Here, we will revisit the fundamental idea of nanostructuring and will evaluate the effect independently with examples in oxides.
1. THERMAL CONDUCTIVITY ()
Most of oxides possess rigid framework in their crystal structure, thus their tend to be high due to the phonon contribution. For the practical application, less than 1 W/(mK), including electron thermal conductivity, is the common limit. Nanostructuring provides the method to control phonon scattering rather than the modification of bonding nature: reducing phonon mean free path (MFP) while maintaining both sound velocity and heat capacity. Practically, this is achieved by rattling or boundary scattering. Nowadays, the boundary scattering is quantitatively evaluated based on Debye-Callaway model and is understood that the strategy can apply any kinds of material. Recent progress in TE efficiency exceeding ZT of 1.5 is mostly owing to this effect. For instance, ZnO of single crystal shows of about 100 W/(mK), while the value largely reduces to 5 W/mK for the crystal size of 30 nm [1].
2. ELECTRICAL CONDUCTIVITY ()
Although reduction in phonon thermal conductivity via boundary scattering is rational approach, the drawback of reduction in must be taken in consideration. As a simple picture, it is said that MFP of charged carrier is shorter than that of phonon, and thus the scattering event induced by nanostructuring may be negligible or not seriously influence to . Quantitatively, the influence can be incorporated through Mattisen¡¯s rule, and we can understand that a certain size range exists to enhance ZT. Practically, band bending near the boundary, which is caused by band offset or impurity trap, must be solved for a gain in ZT, as seen in In2O3 [2].
3. THERMOPOWER (S)
Nanostructuring is able to tune the density of state (DOS), which dramatically modify S. So far, experimental proof of this idea has been demonstrated mostly in two-dimensional system, namely super lattice approach. On the other hand, practical application of TE power generation requires bulk form; thus realization of the DOS tuning in bulk material is strongly demanded. Recently, we found that surface structure of nano-particles (NPs) of SrTiO3 can be a playground to achieve the DOS tuning of bulk materials. SrTiO3 is non-polar because of its cubic symmetry, while the surface of NPs shows polar caused by the surface relaxation, which leads to obvious first order Raman scattering. By means of chemical pressure, we can modulate the surface relaxation rationally, and thus the potential barrier near the surface vary accordingly. The barrier has the effect to cut the low energy carrier, resulting in apparent DOS tuning, hence we can observe large S enhancement [3].
References:
[1] Y. Kinemuchi et al., J. Appl. Phys., 108, 053721 (2010).
[2] Y. Kinemuchi et al., J. Appl. Phys., 110, 12304 (2011).
[3] Y. Kinemuchi et al., J. Electr. Mater., 43, 2011(2014). |
| Acknowledgements : |
|
