Investigating the Role of Transition Metal Deposition on Solid Electrolyte Interphase Formation and Growth Using Redox Shuttles

Abstract

Listen

Translate article

Transition metal based electrodes are attractive positive electrode materials for lithium ion-batteries due to their high energy densities and large capacities. However, these positive electrode materials are prone to dissolution of the transition metal structure into the electrolyte. The subsequent result of the dissolution is accelerated capacity and power fade. Investigation into the dissolution phenomena shows that after dissolution, the transition metals transport to the negative electrode. It is there where the metals ions deposit on the negative electrode surface, catalyzing solid electrolyte interphase (SEI) growth and blocking active sites for intercalation (1-3). In spite of these studies, much remains unclear about variations in the morphology and electronic passivity of the SEI due to transition metal deposition on the negative electrode surface. In this work, we investigate changes in the morphology and electronic passivity of the SEI using redox shuttles. We introduce manganese, nickel, and cobalt-based transition metal salts over a range of concentrations to the primary electrolyte. SEIs are formed on glassy-carbon electrodes by applying potentiostatic holds and cyclic voltammetry (CV). After formation, the SEI is probed using the ferrocenium-ferrocene redox shuttle. A mathematical model of the electrode-electrolyte interface is used to quantify the thickness and porosities of formed SEIs. Figure 1(a) shows the effectiveness of the redox shuttle in characterizing the magnitude of passivation on the working electrode surface. The working electrode is glassy carbon and primary electrolyte is (1:1 wt%) ethylene carbonate : diethyl carbonate with 1 M LiPF6 salt. CVs were measured at ν = 10 mV s-1 around the equilibrium redox potential, ~ 3.24 VLi/Li +. Increasing passivation time reduces redox peak currents, suggesting that surface passivation limits redox kinetics. Figure 1(b) shows experimentally obtained and simulated CVs, demonstrating that the system can be effectively modeled with an SEI.