No abstract

Metallic lithium deposition on graphite anodes is a critical degradation mode in lithium-ion batteries, which limits safety and fast charge capability. A conclusive strategy to mitigate lithium deposition under fast charging yet remains elusive. In this work, we examine the role of electrode microstructure in mitigating lithium plating behavior under various operating conditions, including fast charging. The multilength scale characteristics of the electrode microstructure lead to a complex interaction of transport and kinetic limitations that significantly governs the cell performance and the occurrence of Li plating. We demonstrate, based on a comprehensive mesoscale analysis, that the performance and degradation can be significantly modulated via systematic design improvements at the hierarchy of length scales. It is found that the improvement in kinetic and transport characteristics achievable at disparate scales can dramatically affect Li plating propensity.

The development of next-generation batteries with high areal and volumetric energy density requires the use of high active material mass loading electrodes. This typically reduces the power density, but the push for rapid charging has propelled innovation in microstructure design for improved transport and electrochemical conversion efficiency. This requires accurate effective electrode property estimation, such as tortuosity, electronic conductivity, and interfacial area. Obtaining this information solely from experiments and 3D mesoscale simulations is time-consuming while empirical relations are limited to simplified microstructure geometry. In this work, we propose an alternate route for rapid characterization of electrode microstructural effective properties using machine learning (ML). Using the Li-ion battery graphite anode electrode as an exemplar system, we generate a comprehensive data set of ∼17 000 electrode microstructures. These consist of various shapes, sizes, orientations, and chemical compositions, and characterize their effective properties using 3D mesoscale simulations. A low dimensional representation of each microstructure is achieved by calculating a set of comprehensive physical descriptors and eliminating redundant features. The mesoscale ML analytics based on porous electrode microstructural characteristics achieves prediction accuracy of more than 90% for effective property estimation.

Li-ion batteries (LIB) are ubiquitous in today’s world with applications ranging from portable electronic devices to electric vehicles. These set of applications require the batteries to have diverse energy and power dense configurations. For these reasons the LIB exists in multiple form factors such as cylindrical, pouch cell configurations and different chemistries. The most commercialized and successful LIB are based on intercalation mechanism, which requires shuttling of Li-ion between anode and cathode with redox reaction occurring on the surface of the electrodes and ultimately leading to the Li-ions to diffuse into the host structure of the electrode. Such electrodes are naturally made to be porous, to maximize the surface area for redox and electrochemical reaction, also giving rise to multiple pathways for Li-ion diffusion and migration. These electrode structures are also integrated with additives to increase the electronic conductivity and mechanical strength. This gives rise to the microstructure as a matrix of different phases with multi-length scale structure and characteristics at different scales. The performance and operation of battery is closely intertwined with these electrode microstructure, and thus microstructural characteristics significantly influence the transport process or kinetics of the reaction. This relation between performance and electrode microstructure is not limited to the composition of electrode rather even with fixed chemical composition the widely different structural arrangement of the phases alters the short- and long-range interactions within the electrode. Specifically based on electrochemical performance model, the electrochemical-thermal interactions can be captured with few important effective electrode properties such as interfacial area, tortuosity and conductivity. Here we develop a framework for the accurate prediction of these effective electrode properties as a function of the microstructural properties. For this purpose of effective electrode property prediction, we develop an integrated framework, including information from the physics informed mesoscale model along with the data driven models. The accurate effective electrode properties are evaluated from pore-scale characterization of microstructure, to serve as output for data set. Our main objective with developing such a framework is to describe the effective electrode properties with simple yet meaningful correlations, which can accurately capture physics while maintaining high predictive accuracy, thus also providing directions for better electrode manufacturing.

No abstract