Theoretical Analysis of Materials, used in Energy Storage Applications: the Quest for Robust and Accurate Computational Methodologies

Abstract

Study of materials, used as constituents for energy storage devices (e. g. Li‐ or Na‐ion batteries) has been greatly enhanced by the first principles theoretical modeling. As a prerequisite, accurate computational framework should be employed in order to provide adequate description of studied materials in close agreement with available experimental data. In this account we present our approach to formulation of a suite of computational techniques, that can be used for accurate evaluation of key materials properties with a moderate computational cost. Using the results of our calculations, we show that Hubbard corrected DFT+U method with self‐consistent evaluation of U parameters employing linear response approach can be used for accurate evaluation of redox potentials of cathode materials, although in some cases (e. g. Fe sulfates), addition of other corrections may be required. Moreover, we also provide examples of applications of developed techniques for construction of phase diagrams of cathode materials and analysis of the effects of doping of NaMnO2 cathodes. Finally, we discuss possible directions for future development of computational methodologies and propose new avenues for their applications in battery research.

Redox, Magnetic, and Structural Properties of α-NaMnO2 Cathode Material Analyzed by Fitting-Free DFT+U Calculations, Parameterized by the Linear Response Approach

“…Evaluation of the U parameter computationally, particularly using the linear response method, is an alternative approach, − which permits determination of U , fully avoiding fitting to experimental data. Previous works, including several performed by us, , have shown that average redox voltages can be calculated with a reasonable accuracy for various cathode materials, adopting the values of U parameters determined using linear response parametrization.…”