High Throughput Approach to Accelerate Electrolyte Discovery for Post Li-Ion Batteries

Abstract

Transforming transportation and electricity grid with high performance, low cost energy storage require development of post Li-ion technology and innovations in electrodes and electrolytes, alike. A non-aqueous multivalent (e.g., Mg2+, Ca2+ and Zn2+) metal cell is one of the potential candidates for a post-lithium-ion battery. The theoretical volumetric capacity of a metal anode coupled with the lack of dendrite formation at a multivalent metal anode provide an attractive opportunity in energy storage.1 The ability of Mg batteries to provide much higher volumetric capacity, particularly on the anode side where the Mg metal can theoretically provide 3833 mA h/cc as compared to the Li counterpart graphite (∼ 800 mA h/cc) at a lower cost makes the technology an attractive candidate for future batteries. On the other hand Zn metal anodes coupled with a reversible intercalation cathode chemistry have a number of promising features: (1) highly efficient (≥ 99%) reversible Zn deposition on Zn metal anode in high performance nonaqueous Zn electrolytes (e.g., high anodic stability (maximum ∼ 3.8 V) and ionic conductivity) (2) relatively lower activation barrier energy for migration of Zn2+ ions in a variety of cathode materials (e.g., FePO4, NiO2, and V2O5) (3) similar ionic radius compared with Li+ and Mg2+ ions (4) much higher volumetric capacity compared to other multivalent metals, such as Mg and Ca. However, development and commercialization of multivalent batteries require not only improved electrode discovery and development but also novel electrolytes, which are compatible with both the multivalent metal as well as the multivalent cathode. Thus a fundamental understanding of molecular level properties of these electrolytes is required to improve the electrochemical stability and the charge transfer properties. An automatic High-throughput computational infrastructure has been constructed for the electrolyte genome project supported by the US Joint Center for Energy Storage Research (JCESR) coupled with experimental analysis2. In this work, we present a multi-scale modelling approach for multivalent salts in various solvents. We uncover a novel effect between concentration dependent ion pair formation and anion stability at reducing potential, e.g., at the metal anode1. We elucidate systematic correlations between molecular level interactions and composite electrolyte properties, such as electrochemical stability, solvation structure, and dynamics. We find that Mg electrolytes are highly prone to ion pair formation, even at modest concentrations, for a wide range of solvents with different dielectric constants, which have implications for dynamics as well as charge transfer. Specifically, we observe that, at Mg metal potentials, the ion pair undergoes partial reduction at the Mg cation center (Mg2+→Mg+), which competes with the charge transfer mechanism and can activate the anion to render it susceptible to decomposition. Specifically, TFSI− exhibits a significant bond weakening while paired with the transient, partially reduced Mg+. We uncover the origin of anodic stability for a range of nonaqueous zinc electrolytes. By examination of electrochemical, structural, and transport properties of nonaqueous zinc electrolytes with varying concentrations, it is demonstrated that the acetonitrile−Zn(TFSI)2 , acetonitrile− Zn(CF3SO3)2 , and propylene carbonate− Zn(TFSI)2 electrolytes can not only support highly reversible Zn deposition behavior on a Zn metal anode (≥ 99% of Coulombic efficiency) but also provide high anodic stability (up to ∼ 3.8 V vs Zn/Zn2+ ).3 The combination of multi-scale modeling with experimental techniques provides unprecedented insight into the origin of the electrochemical, structural, and transport properties in multivalent electrolytes.