In-Situ Characterization of the Li-Al System at Elevated Temperatures

Abstract

We will describe our platform for ex- and in-situ characterization of lithium ion electrode and electrolyte materials over a wide range of temperatures, and our use of this platform to characterize phase formation and dissociation in the Li-Al system. Aluminum electrodes have been studied for use in lithium electrochemical cells for more than 40 years. Electrochemical insertion of lithium in to aluminum at very high temperatures (>400 °C) follows the equilibrium phase diagram; studies at e.g. room temperature only demonstrate the formation of LiAl. The formation of the other expected Li-Al phases, Li3Al2 and Li9Al4, is thought to be inhibited by slow diffusion of Li in LiAl. Aluminum is available in a wide variety of morphologies and manufacturing-induced microstructures and as such is a model test platform for deconvoluting the effects of microstructure, diffusion, and phase formation in metal alloy electrodes. Our core enabling technology is an adaptable high-temperature compatible cell design inspired by high-temperature compatible primary cell hardware and adapted from earlier work. Our double-ended glass-to-metal Conflat cells are compatible with low and high voltage electrode materials, adjustable (and commercially-relevant) electrode thicknesses and stack pressures, and can maintain ultra-high vacuum (<10-7 Pa) tight seals at temperatures up to 415 °C. Results from in-situ cycling of a Conflat cell with lithium and aluminum foil electrodes at 110 °C are presented in Fig. 1. Three voltage plateaus are evident, and the step changes between plateaus roughly correspond with the expected phase transition compositions (i.e. LiAl for Li/Al = 0 <–> 1, Li3Al2 for Li/Al = 1 <–> 1.5, Li9Al4 for Li/Al = 1.5 <-> 2.25). XRD patterns collected at room temperature (ex-situ) at three indicated compositions (x1, x2, x3) are also provided in Fig. 1. Patterns collected from samples cycled to 200 mV vs. Li metal on insertion (x1) and 150 mV vs. Li metal on removal (x3) indicate LiAl and Li3Al2 are present, as expected. However, the XRD pattern collected from the sample cycled to 40 mV vs. Li metal on removal (x2) indicates Li1.77Al, rather than Li9Al4, is present at a composition of Li/Al = 2.2. Li9Al4 has been detected in samples cycled in different conditions (temperature, current density, voltage limits, sample morphology etc.) The discrepancy between measured and expected phase composition at x2 (Li/Al = ~2.2 vs. ~1.8, respectively) is attributed to consumption of lithium in solid-electrolyte interphases (SEI). Li3Al2 (at composition x3, Li/Al = 1.8) has a similar discrepancy (expected composition of Li/Al = 1.5). Known phase transitions can therefore be used as guideposts to quantify SEI formation. XRD results shown in Fig. 1 and described above were collected by cooling and then disassembling Conflat cells and incorporating the lithiated electrode in a polypropylene encapsulated sample holder. A major advantage of Conflat cell technology is the ability to bolt-on new functionality without compromising cell performance. Conflat flanges incorporating x-ray transparent windows are now being incorporated in to our cells. We will present ex- and in-situ (variable temperature cycling, room temperature XRD, with and without cell disassembly) and in-operando (integrated variable temperature cycling and XRD, without cooling) results from aluminum thin films and (much thicker, but still thin) aluminum foils to map out phase formation as a function of temperature, sample morphology, and static and dynamic stress. This data, when coupled with insight in to Li diffusion rates and nucleation barriers for the various Li-Al phases, will enable a more complete understanding of ion mobility and phase formation and stability in the Li-Al and other metal alloy systems. Figure 1