Multiobjective Solid Electrolyte Design of Tetragonal and Cubic Inverse-Perovskites for All-Solid-State Lithium-Ion Batteries by High-Throughput Density Functional Theory Calculations and AI-Driven Methods
Randy Jalem,Yoshitaka Tateyama,Kazunori Takada,Seong-Hoon Jang
DOI: https://doi.org/10.1021/acs.jpcc.3c02801
2023-08-28
The Journal of Physical Chemistry C
Abstract:Solid electrolytes (SEs) are crucial materials to realize highly safe and practical all-solid-state Li+-ion batteries. Here, we performed a large-scale computational SE screening on a chemical space of >10 000 Li-rich inverse-perovskite (ip) compounds with tetragonal and cubic structures by high-throughput density functional theory (DFT) and AI-driven methods. A total of 1413 novel candidate compounds were predicted to be synthesizable based on thermodynamic decomposition energy (E d) and machine-learned experimental synthesis likelihood (L s). These compounds were further screened using a Pareto-front approximation set of a multiobjective Bayesian optimization tasks for k = 3 DFT-calculated SE properties (f k , with k = 1, 2, and 3): (i) electrochemical window from electronic band gap energy (f 1: E g), (ii) chemical stability by reaction with moisture (f 2: E h), and (iii) 400 K bulk Li+-ion conductivity (f 3: Λ). As a result, the compound list was reduced down to 24 candidate ip SEs, and examples include Cm Li8O2Cl3Br (E d = 0, L s > 0.5, E g = 4.74 eV, E h = −33.22 kJ/mol, and Λ = 9.0 × 10–4 S/cm), Amm2 Li8OSCl4 (E d = 0.070 eV/atom, L s > 0.5, E g = 4.14 eV, E h = −40.70 kJ/mol, and Λ = 9.2 × 10–2 S/cm), and Cmcm Li12O3SeClBr3 (E d = 0.097 eV/atom, L s > 0.5, E g = 3.36 eV, E h = −86.88 kJ/mol, and Λ = 7.8 × 10–1 S/cm). Possible solid-state synthesis routes for the screened SE candidates were also explored using thermodynamic phase competition analysis and classical nucleation theory reaction barrier. Aside from providing a well-informed list of potentially novel ip-type SEs, our work also reports on an effective calculation methodology for tiered large-scale material screening which, at the same time, incorporates “small data” learning on target property datasets that are computationally expensive to obtain. The generated datasets are expected as well to be of great utility for future data-driven material design efforts.
chemistry, physical,nanoscience & nanotechnology,materials science, multidisciplinary