Quenching Thermal Transport in Aperiodic Superlattices: A Molecular Dynamics and Machine Learning Study

Abstract

Random multilayer (RML) structures, or aperiodic superlattices, can localize coherent phonons and therefore exhibit drastically reduced lattice thermal conductivity compared to their superlattice counterparts. The optimization of RML structures is essential for obtaining ultralow thermal conductivity, which is critical for various applications such as thermoelectrics and thermal barrier coatings. A higher degree of disorder in RMLs will lead to stronger phonon localization and, correspondingly, a lower lattice thermal conductivity. In this work, we identified several essential parameters for quantifying the disorder in layer thicknesses of RMLs. We were able to correlate these disorder parameters with thermal conductivity, as confirmed by classical molecular dynamics simulations of conceptual Lennard-Jones RMLs. Moreover, we have shown that these parameters are effective as features for physics-based machine learning models to predict the lattice thermal conductivity of RMLs with improved accuracy and efficiency.

Improving thermal conductivities of textile materials by nanohybrid approaches

“…For instance, a textile composed of an outer layer with low thermal conductivity and an inner layer with high thermal conductivity may be promising for cold weather conditions. Moreover, thermal transport can be quenched in textile materials by using aperiodic superlattices in the case of nano-enhanced textile fibers ( Chakraborty et al., 2020 ; Hu et al., 2020b ). Fabrication methods such as vacuum-assisted filtration and spraying can be preferred for the production of the high thermally conductive layers, whereas methods such as coating, wet spinning, and chemical coupling may be more suitable for the fabrication of low thermally conductive layers.…”

Machine Learning for Harnessing Thermal Energy: From Materials Discovery to System Optimization

“…In addition to homogeneous materials, the effects of compositional and structural factors on thermal conductivity can also be efficiently predicted with ML algorithms, mainly for nanostructures, composites, and porous materials. By using period and layer thickness as material descriptors and MD simulation results as training data for ANN, the thermal conductivity of superlattice can be minimized. , The thermal transport in a porous medium can be modeled with the finite element method and heat diffusion equation once the porous structure size is much larger than the heat carriers’ mean free path. By training a limited data set of finite element method simulation results with appropriate structural features like shape and bottleneck thickness, the structure–thermal conductivity relationship can be found .…”

Thermal Metamaterial: Fundamental, Application, and Outlook

“…Some recent works about using these two kinds of phononic crystals and their comparison were reported ( Sledzinska et al., 2020 ; Hussein et al., 2020 ). Besides the Bragg-type crystal and locally resonant crystal, recently, aperiodic superlattices or random multilayer structures were also found effective systems for tuning the wave nature of phonons ( Chakraborty et al, 2020 ; Hu et al, 2020 ). For example, thermal conductivity can be significantly reduced in GaAs/AlAs superlattices with clean interfaces ( Hu et al, 2020 ).…”