UC Berkeley

Fast charging of lithium-ion batteries (LIBs) is affected by the electrolyte in many ways. The rate at which Li+ can shuttle between the electrodes is determined by bulk electrolyte transport properties, such as the ionic conductivity and cation transference number (t+), which influence the Li+ concentration gradients that form in the cell and determine the attainable energy density at a given charge rate. The electrolyte also must form a stable electron-insulating, Li+-conducting interface on both the anode and cathode in order to prevent continual degradation of the salt and solvent. The interfacial layer on the anode, the so-called 'SEI,' also fundamentally affects fast charging, as Li+ must transport through the SEI layer, strip from its solvation shell, and ultimately insert into graphite. The influenceof both of these aspects - bulk electrolyte transport properties and SEI formation - on fast charge capability are the focus of my dissertation.

The standard 'Gen 2' electrolyte employed in conventional batteries (1.2 M LiPF6 in 3:7w ethylene carbonate:ethyl methyl carbonate (EC:EMC)) has an ionic conductivity of 10 mS/cm and a cation transference number (t+, defined as the cation diffusivity divided by the sum of the cation and anion diffusivity) of ~0.4, meaning the strongly solvated Li+ actually diffuses slower in solution than the bulky PF6- anion. This relatively low t+ results in large concentration gradients during fast charge, which can lead to high required overpotentials and can result in Li plating on the graphite. As will be discussed in Chapter 2, if t+ of the electrolyte can be engineered to be modestly higher, even with a substantial reduction in conductivity, charge performance can be improved. Chapter 2 also presents useful targets for electrolyte transport properties and outlines the benefit of high t+ electrolytes in preventing Li plating, a hazardous side reaction that is exacerbated by fast charging.

The reason Gen 2 electrolyte is so commonly used, though, is not necessarily due to its transport properties, but rather due to the EC solvent's ability to form a stable SEI on graphite and allow Li+ to strip its solvation shell before inserting into graphite. This dissertation outlines a titration procedure (called mass spectrometry titration, or MST) with which we quantify the amount of the solid carbonates that deposit on graphite as a result of EC reduction during SEI formation. The titration is also extended to quantify plated Li that deposits on the graphite as well as other SEI components, such as Li2C2, and the contributions of the various irreversibly formed species to the observed capacity fade during fast charge are determined.

Finally, differential electrochemical mass spectrometry (DEMS) is used to measure gases that are evolved during battery cycling, and these gases provide quantitative insights into the amount of SEI species deposited on graphite. The DEMS technique is especially powerful when combined with MST, which provides complementary ex situ information about the extent of conversion of initially deposited SEI to other species. The holistic picture of the SEI provided by DEMS and MST is then used to determine the influence of the SEI on fast charge performance for various electrolyte compositions. In all, this dissertation quantitatively probes submicron-scale phenomena, including Li plating and SEI formation, to better understand the challenge of fast charging.