Data-Driven Prediction of Formation Mechanisms of Lithium Ethylene Monocarbonate with an Automated Reaction Network

Interfacial reactions are notoriously difficult to characterize and robust prediction of the chemical evolution and associated functionality of the resulting surface film is one of the grand challenges of materials chemistry. The solid–electrolyte interphase (SEI), critical to the operation of Li-ion batteries (LIB), exemplifies such a surface film and despite decades of work, considerable controversy remains regarding the major components of the SEI as well as their formation mechanisms. In this work, we present pioneering results of a newly developed data-driven reaction network addressing the recent question whether lithium ethylene monocarbonate (LEMC) or lithium ethylene dicarbonate (LEDC) is the major organic component of the LIB SEI. Our data-driven, automated methodology is based on a systematic generation of relevant species using a general fragmentation/recombination procedure which provides the basis for our vast thermodynamic reaction landscape, calculated with density functional theory (DFT). The graph implementation of the reaction landscape is subsequently explored using shortest pathfinding algorithms, identifying reactions to LEMC from EC, Li+ and H2O under the electron chemical potential of Li metal. Confirming the viability of our approach, the reaction network automatically recovers previously-proposed formation mechanisms of LEMC from EC and LEDC through hydrolysis, among which the direct hydrolysis of EC under basic conditions is found to be the most kinetically favorable. We also identify several other new reaction pathways to LEMC, illustrating the complex and competitive landscape of possible electrochemical electrolyte decomposition reactions. For example, we recover a LEMC formation mechanism that generates lithium hydride as a by-product and a radical mechanism through breaking the (CH2)O–C(–O)OLi bond in LEDC, neither of which has been proposed previously. Most importantly, we find that all identified paths, which are also kinetically favorable under the explored conditions, require water as a reactant. This condition severely limits the amount of LEMC that can form, as compared to LEDC, a conclusion that has direct impact on our understanding of SEI formation in Li-ion energy storage systems. Finally, we emphasize that our framework demonstrates robust, automated, data-driven predictions of novel interfacial reaction mechanisms and this framework is generally applicable to other reactive systems.