(Invited) Theoretical Analysis of Microscopic Interface Ionics at Heterogeneous Solid-Solid Interfaces in Batteries

Abstract

All-solid-state battery (ASSB) has attracted considerable attention for decades, and the developments of many new materials appropriate for ASSB have been made. In contrast, less is known about the interfacial processes, especially at the solid electrolyte (SE)-electrode interfaces. The microscopic origins of interfacial resistance and degradation are still crucial issues for the efficient ASSBs, which are closely related to the interface chemistry and electrochemistry on the ionic and electronic scales. Therefore, understanding of the microscopic interface ionics and electronics is indispensable for the cutting-edge ASSB development.We have addressed establishing a universal calculation framework based on density functional theory (DFT) for heterogeneous solid-solid interfaces. Typical DFT geometry optimization or molecular dynamics does not work well for the atomistic structure search due to the coexistence of collective and local geometry relaxations inherent to the solid-solid interfaces. The methods with empirical force field are also insufficient because their parameters are usually fitted to bulk properties, not interfacial ones.To overcome these issues, we have developed a novel computational technique, “Heterogeneous Interface (HI)-CALYPSO method” [1], for systematic sampling of heterogeneous solid-solid interface structures. The breakthrough was achieved by applying CALYPSO structure prediction method for all types of interfacial geometry relaxations, allowing to explore a variety of interfacial disordered structures [1]. This approach is compatible to massive parallel calculations on the supercomputers like Fugaku.Then, the method was applied to a variety of solid-solid interfaces in ASSBs. As a representative model, we investigated the system with LiCoO2 (LCO), β-Li3PS4 (LPS), and LiNbO3 (LNO) acting as a cathode, a sulfide electrolyte, and a buffer layer, respectively [1-3]. For the LCO/LPS interfaces, we sampled over 20000 configurations and found several stable disordered structures involving mixing cations and anions, leading to the formation of a reaction layer. On the other hand, Li-ion sites that can be preferentially depleted upon charging always exist around the cathode-SE interfaces irrespective of the interfacial order/disorder. Therefore, we conclude that the dynamic Li-ion depletion is related to suppression of successive Li-ion transport, leading to the interfacial resistance. Through the electronic states analysis, we also deduced a probable origin for the interfacial Li-ion depletion and a mean to suppress this problematic behavior. Finally, we provided ionic potential surface across the interfaces, governing ion dynamics across the interfaces. In consequences, sampling probable interface structures enabled to analyze statistically probable characteristics of the ionics and electronics at the solid-solid interfaces and provide a possibility of space-charge layer as well as electric double layer.We have also applied the method to the interface between Li metal anode and Li7La3Zr2O12 electrolyte [4]. Again, we succeeded in sampling of the probable interface structures, which have many interfacial Li-O bonds and unsaturated (i.e. coordination number < 6) Zr sites. These undercoordinated Zr sites are reduced once the LLZO surface is in contact with the Li metal, leading to chemical instability of the LLZO/Li interface.Recently, the interfaces between the Li1.17Al0.17Ti1.83(PO4)3 SE and LCO cathode as well as Li metal anode have been investigated as well. Though it’s not related to ASSB, we examined the heterogeneous interfaces between the graphite anode and Li2CO3 model SEI layer [5]. All of these studies suggest a unified insight into the interfacial ionic transport in ASSBs. In this talk, I’ll introduce our recent methodological and application studies on the ASSB interfaces and discuss the unified insight obtained.Acknowledgements: These works were done in collaboration with Dr. Bo Gao, Dr. Hong-Kang Tian, and Dr. Randy Jalem in NIMS. The works were partly supported by MEXT as “Program for Promoting Researches on the Supercomputer Fugaku” (Fugaku Battery & Fuel Cell Project) and JSPS KAKENHI “Interface IONICS” Grant Number JP19H05815.