2018/11/16 by Keishu Utimula, Tom Ichibha, Ryo Maezono +1 · 21 citations
Chemistry · Engineering · Materials Science · Physics and Astronomy · #Ab initio #Ab initio quantum chemistry methods #Advanced Memory and Neural Computing #Chemical physics #Chemistry #Composite material #Conductivity #Machine Learning in Materials Science #Materials science #Molecule #Organic chemistry #Physical chemistry #Polymer #Thermal conductivity #Thermal properties of materials #cond-mat.mtrl-sci
paper · pdf · open access · doi:10.1021/acs.chemmater.9b00020
published in Chemistry of Materials 31(13), 4649-4656 (American Chemical Society)
arxiv created 2018/11/16 · openalex created_date 2018/11/29 · openalex publication_date 2019/06/05 · arxiv updated 2020/08/21 · openalex updated_date 2026/08/06
High Resolution Image Download MS PowerPoint Slide We investigated the lattice thermal conductivity (LTC) of a subset of polymer crystals from the polymer genome library to explore high LTC polymer systems. We employed a first-principles approach to evaluate the phonon lifetime within the third-order perturbation theory combined with density functional theory and then solved the linearized Boltzmann transport equation with a single-mode relaxation time approximated by the computed lifetime. Typical high LTC polymer systems, namely, polyethylene (PE) crystal and fiber, were benchmarked to validate our approach. In addition to PE, we evaluated the LTC of polyphenylene sulfide (PPS) and poly(ethylene terephthalate) (PET) because, although their experimental LTC values are not obtained from their “perfect” crystals, they are mostly available. Our simulations reproduced an experimentally observed LTC ordering (PE ≫ PPS > PET). Applying our scheme to a number of polymer crystals for the first time, we discovered that the β-phase of a poly(vinylidenesurely fluoride) (PVDF-β) crystal at low temperatures has the highest LTC among all of the cases considered. We also found the LTC correlating to the curvature of an energy–volume plot. This curvature can be used as one of the descriptors of constructing modern machine learning models to further explore high LTC polymer crystals by means of a data-driven approach beyond a human-based one.