Dendrite Growth in Lithium/Polymer Systems

Abstract

Dendrite growth in a parallel-electrode lithium/polymer cell during galvanostatic charging has been modeled. The growth model is surface-energy controlled, incorporating the effect of dendrite tip curvature into its dendrite growth kinetics. Using data representative of the oxymethylene-linked poly͑ethylene oxide͒/LiTFSI system, it is shown that dendrites accelerate across cells under all conditions, and that growth is always slowed by lowering the current density. Cell shorting occurs during typical charges at current densities above 75% of the limiting current. Increased interelectrode distance slows failure, but the advantages decrease as distance lengthens. A factor of 1000 increase in surface forces delays cell failure by only 6%. While larger diffusion coefficients usually extend the time to cell failure, this trend is not consistent at high transference numbers.Lithium batteries were proposed more than 30 years ago as an ideal system for high energy-density applications. In the early 1970s, organic electrolytes were discovered to form passivating films that chemically protect pure lithium electrodes over time. 1 Although this solid-electrolyte interphase ͑SEI͒ prevents chemical reaction at a state of no current, a number of failure mechanisms have been discovered that occur during cycling of lithium rechargeable batteries. One of these mechanisms is dendritic growth, first hypothesized to occur in the lithium/polymer system in 1974. 2 Dendritic lithium deposits, sometimes called mossy lithium after being subjected to multiple charges and discharges, were first directly observed in 1980. 3 Since then, scanning electron microscopy has been used to characterize more thoroughly the appearance of lithium dendrites after cell failure. 4 While work on the lithium electrode is recent, empirical observation and theoretical analysis of dendrite growth in other electrochemical systems has been pursued for years. The most significant of these results are presented here and used to prepare a model that simulates lithium dendrite growth under galvanostatic conditions. The name dendrite derives from the treelike shape of the deposit, which can range from nearly linear and pointed ͑needle-like͒ to highly branched ͑bush-like͒. Past work has primarily considered the propagation rate and morphology of dendrites. 5 Most early experiments rely on microscopy, but qualitative pictures resulting from computer simulation have been presented in the last twenty years. Aogaki 6 simulated the electrode appearance during dendrite formation using a three-dimensional linear stability analysis. Chen and Jorné 7 used a fractal growth model to show that macroscopic shapes and branching can be mimicked by changing the dendrite resistance.Many references in the current literature show the effect of additives on dendritic deposits. Most of the work on additives tends to be simple characterization, and is relatively system-specific. The goals of these experiments include developing more inert solvent/ solute mixtures, using novel anions, and in...

Rate-Dependent Morphology of Li2O2 Growth in Li–O2 Batteries

“…1). This has since been confirmed on different carbon substrates at low surface specific rates [7,9,11,22,23]. Although the disc-like particles reach 100 nm sizes, toroid-like particles can grow much larger, and the electron transport path and growth mechanisms are just beginning to be understood [10].…”

A dendrite-suppressing composite ion conductor from aramid nanofibres

“…The results obtained for copper are relevant for lithium because the Young's modulus of copper is E ¼ 129.8 GPa, while that for lithium is E ¼ 4.91 GPa. The theory of electrochemical dendrite growth 34 indicates that if local mechanical properties of ICM are sufficient to prevent mechanical stress from dendrites from copper, it will also suppress dendrites from lithium, which are much softer. Notably, we do not claim that copper and lithium are electrochemically similar.…”

Quantifying the dependence of dead lithium losses on the cycling period in lithium metal batteries

“…4,5 This drawback not only compromises the reliability but ultimately decreases the capacity of Li 0 batteries. 2,[6][7][8][9][10] Work on dendrite growth has mainly focused on the effects of charging protocol, 11,12 current density, 13,14 electrode surface morphology, 15,16 temperature, 17,18 solvent and electrolyte chemical composition, [19][20][21] electrolyte concentration 22,23 and evolution time 24,25 on dendrite growth. Some strategies included the use of powder electrodes 26 and adhesive polymers.…”

“…27 The empirical nature of these approaches, however, reflects the fact that current models of dendrite growth exist based on various simplifying assumptions that have fallen short of capturing the essentials of this phenomenon. 14,[28][29][30][31][32] We view DLC formation as a manifestation of the intrinsic sponginess of Li 0 electrodeposits, 33 and the non-uniform dissolution rates of such deposits upon discharge. 5,34,35 In this paper, we report experimental results on the effect of the duration of the charging period t on the amount of DLCs produced at constant charge.…”