First Principle Calculations for the Study of Solid-state Electrolytes

Abstract

The first principle calculations based on the density functional theory is a very useful technique for understanding and designing battery materials including all-solid-state batteries. It is widely used to study of solid electrolytes, which are core materials of all-solid-state batteries, such as, the chemical and electrochemical stability chemical and electrochemical reactions at interfaces, and diffusion mechanisms of ions. In particular, the recently developed ab initio molecular dynamics simulation successfully clarified the diffusion mechanisms of various ion conductors by analyzing the behavior of ions in a solid electrolyte with the accuracy of quantum mechanics. Although pioneers are developing excellent computational methodologies, it is still in the early stages of research, so it is necessary to raise and solve various problems and sometimes develop new methods. Chapter 2 of this thesis describes in detail the computational methods used in the research on solid electrolytes developed so far and the newly developed computational methods. Specifically, a method of constructing a phase diagram and a grand potential phase diagram based on DFT was presented. In addition, a method for evaluating the chemical and electrochemical stability of solid materials and chemical and electrochemical reactions at the interface using the phase diagrams is introduced. Based on the grand potential phase diagram, we developed a method to evaluate the degree of emission of toxic H2S gas due to hydrolysis reaction, which is the main problem of sulfide-based solid electrolytes. Using this, the reaction energy of H2S gas formation was calculated for various compositions of argyrodite sulfide-based solid electrolyte and the results were consistent with those measured in the experiment. In addition, the ionic conductivity calculation method, which must be the most basic property in the development of solid electrolytes, was introduced. The ionic conductivity calculation method of a solid electrolyte using AIMD simulation has been developed in 2012, and various guidelines have been proposed, but it is still being improved and developed. Here, a standard method with following several guidelines to obtain ionic conductivity at room temperature is introduced. We presented a method to effectively determine the total simulation time, which is one of the problems encountered in this process. Moreover, we developed correction methods to reduce the gap in ionic conductivity obtained from experiments and calculations which caused from site disorder and crystallinity of experimentally synthesized crystal structures. And a method for calculating the melting point of a solid was developed by quantitatively analyzing the observation of severe structural displacement while performing AIMD simulations at high temperature. The calculated melting point was in good agreement with that observed in the experiment, and is expected to provide useful information for synthesizing new solid materials. In Chapter 3, Li-argyrodite, a representative sulfide-based solid electrolyte, was analyzed at the atomic level and presented with a new perspective of structural properties that had not been studied before. It was suggested that altering the halogen distribution in Li argyrodites during synthesis could increase the Li-ion conductivity of these materials due to site disorder of S2-/X- single anions. Inspired by this work, we systematically investigated the “composition-structure-property” relationship in Li6-xPS5-xX1+x (0 ≤ x ≤ 1 and X=Cl, Br or I) model structures. Our results show a close correlation between the Li-ion conductivity and the cage-like Li sublattice structure around the S2-/X- single anions. We particularly found that the size of the Li-ion cage becomes uniform with increasing the halogen doping level, and the inter-cage diffusion of Li ions is accelerated to increase Li-ion conductivity. Therefore, we propose a standard deviation (STD) of Li-cage size around S2-/X- single anions as a descriptor for the screening of argyrodite-based superionic conductors. Our results will provide a novel approach for tuning the compositional change of Li argyrodites based on “composition-structure-property” relationships that accelerate inter-cage diffusion to increase Li-ion conductivity. In Chapter 4, the ionic diffusion characteristics of 600 argyrodite compositions, Li8−x−y([A]1−y[B]y[C]4)[C]2−x[X]x (A=Group 4, B=Group 5, X=Group 7, x=0 or 1 , y=0 or 1), were evaluated using a newly developed descriptor instead of long AIMD simulations. As expected, when the size difference of anions constituting argyrodite was small, the Li-cage size STD was calculated to be significantly lower, which caused fast Li-ion diffusion. The top-ranking compositions evaluated based on the Li-cage size STD were attempted to be synthesized through experiments. Among them, Li6PSe5Br was successfully synthesized, and the ionic conductivity of 1.57 mS/cm was measured. It was similar to calculated Li-ion conductivity (1.36 mS/cm) using AIMD simulation. In halogen-free argyrodites, the Li-cage size STD shows a very low value because the same type of element is located at both anion Wyckoff positions, which leads to high ionic conductivity. However, these compositions failed to be synthesized through experiments. If a new synthesis method is developed and able to synthesize halogen-free argyrodite at room temperature, it is expected to show very high Li-ion conductivity. These results demonstrate that the new descriptor called Li-cage size STD developed in Chapter 3 was successfully applied. Since this method can predict the ionic diffusion performance of various argyrodites with a short calculation, it can be very practically used for argyrodite research using computational science.