Design of Polymer Blend Electrolytes through a Machine Learning Approach

Bill K. Wheatle,Erick F. Fuentes,Nathaniel A. Lynd,Venkat Ganesan
DOI: https://doi.org/10.1021/acs.macromol.0c01547
IF: 5.5
2020-10-21
Macromolecules
Abstract:We apply a machine learning (ML) technique to the multiobjective design of polymer blend electrolytes. In particular, we are interested in maximizing electrolyte performance measured by a combination of ionic transport (measured by ionic conductivity) and electrolyte mechanical properties (measured by viscosity) in a coarse-grained molecular dynamics framework. Recognizing the expense of evaluating each of these properties, we identify that the anionic mean-squared displacement and polymer relaxation time can serve as their proxies. By employing the ML approach known as Bayesian optimization, we identify a trade-off between ion transport and electrolyte mechanical properties as a function of varied design parameters, which include host molecular weight and polarity. Our results suggest that blend electrolytes whose hosts have unequal molecular weights, such as gel polymer electrolytes, rarely maximize electrolyte performance. Overall, our results suggest the potential of a framework to design high-performance electrolytes using a combination of molecular simulation and ML.The Supporting Information is available free of charge at <a class="ext-link" href="/doi/10.1021/acs.macromol.0c01547?goto=supporting-info">https://pubs.acs.org/doi/10.1021/acs.macromol.0c01547</a>.Core simulation parameters; sample core simulation MSDs and <i>S</i>(<i>q</i>,<i>t</i>); Bayesian optimization theory; optimization run radius of gyrations and simulation energies for <i>w</i><sub>D</sub> = 0.5; <i>f</i><sub>p</sub>(<b>x</b>) evaluations; long-run simulation results of <i>G</i>(<i>t</i>) for optimal design parameters; and perturbation MSD complement (<a class="ext-link" href="/doi/suppl/10.1021/acs.macromol.0c01547/suppl_file/ma0c01547_si_001.pdf">PDF</a>)This article has not yet been cited by other publications.
polymer science
What problem does this paper attempt to address?