Understanding size and interface dependent anisotropic thermal conduction in correlated multilayer structures

最近已经确定，一种非常有效的提高热电效率的方法是通过在纳米尺度上结构来抑制热导率，例如在多层体系中。然而，这种效应的机制，即界面对不同波长声子的散射和阻挡还不是很清楚。这一联合项目的目标是进一步加深对氧化物多层膜系统中各向异性和尺寸相关的热传导的理解。我们优化了顶部和底部电极几何形状的3-omega方法，以便可靠地测量导热系数κ。将神经网络与热电耦合条件的详细有限元建模相结合，将能够高精度地确定平面内和平面间的贡献、κ┴和κ||。我们对声子色散的连续模型为我们的进一步研究选择了有前景的多层热电氧化物系统，即掺杂的SrTiO_3和Pr1-xCaxMnO_3，以及失配的层状化合物Ca3Co4O9。从界面、无序和掺杂的不同散射贡献中分离出层的声阻抗失配的影响，将通过仔细的实验研究，结合使用非平衡分子动力学和声子Boltzmann方程的热导率的原子水平模拟来实现。

It has recently been established that a very effective method for enhancement of thermoelectric efficiency is through suppression of thermal conductivity by structuring at the nanoscale, such as in multilayer systems. However, the mechanisms of this effect, i.e. the scattering and blocking of phonons at various wavelengths by interfaces are not well understood. The objective of this proposed joint project is to further develop the understanding of anisotropic and size dependent thermal conduction in oxide multilayer systems. We have optimized the 3-omega method in top and bottom electrode geometries for reliable measurements of thermal conductivity κ. The use of neural networks in combination with detailed finite element modelling of the coupled thermal and electrical conditions will allow for high-precision determination of in-plane and cross-plane contributions, κ┴ and κ||. Our continuum modelling of phonon dispersion has led to the selection of promising thermoelectric oxide systems for multilayers, namely doped SrTiO3 and Pr1-xCaxMnO3 and the misfit layered compound Ca3Co4O9, for our further studies. Separating out the influence of acoustic impedance mismatch of layers from various scattering contributions of interfaces, disorder, and dopants will be performed by careful experimental studies combined with atomic level modelling of thermal conductivity using non-equilibrium molecular dynamics and the phonon Boltzmann equation.