Lithium metal is theoretically the best anode material for Li-based batteries. However, it is notorious for irregular growth upon repeated battery operation, thus limiting its practical relevance. Accordingly, graphite anodes that exhibit smaller but sustained performance have been used in traditional Li-ion batteries for the past four decades. The pursuit of more energy-dense batteries for the next-generation electrification challenges has redirected researchers’ attention to lithium anodes. The latest advances in experimental imaging and computational capabilities have started to provide new insights into the underlying fundamental mechanisms accounting for the idiosyncrasies of electrochemically grown lithium.
Recent research reported in Nature Materials (doi:10.1038/s41563-020-0729-1) explores the initial stages of lithium electrodeposition in liquid electrolytes – mimicking the behavior of Li-metal anodes in batteries. The study finds that the crystallinity of the freshly grown lithium strongly affects its operation. The work was carried out by Boryann Liaw, Y. Shirley Meng, and colleagues from Idaho National Laboratory, University of California, San Diego, and Pacific Northwest National Laboratory. To track the initial stages of nucleation and growth, at every current density (i.e., growth rate), electrodeposition was carried out for progressively longer times and the resultant growth structures were analyzed using cryogenic transmission electron microscopy to assess crystallinity of the new phase and scanning electron microscopy (SEM) to visualize the growth morphology.
The nucleated phase (very early times) is found to be amorphous. As the deposited clusters grow, they gradually transition to an ordered (crystalline) phase. Such a disorder–order phase transition is found to be strongly sensitive to the growth rate where faster deposition (i.e., higher current) leads to more crystalline deposits. Reactive molecular dynamics (MD) simulations were performed to probe the origins of such a phase transition, revealing that large initial nuclei were crystalline while smaller ones were amorphous. Thus, higher currents lead to larger initial nuclei that grow as crystalline deposits, while smaller ones at low currents grow as a disordered amorphous phase.
Since a crystalline phase grows to conform to the crystal shape (thermodynamic Wulff shape), sharp growth features are observed. On the other hand, amorphous nuclei grow in a relatively uniform fashion. Such differences are apparent in SEM images of growth morphology. Classical nucleation theory does not account for the disorder–order phase transition that is important to describe the differences in the growth morphologies. Similar current-dependent disorder-to-order behavior was observed when the researchers electrodeposited sodium, potassium, magnesium, and zinc—other metal anodes of interest.
Visualizing the interface structure is a challenging aspect, in part because the lithium surface is tricky to preserve going from battery to the imaging station. The high reactivity of lithium forms new phases at the interface of interest as soon as the electrodes are extracted from cells and come in contact with the ambient. The cryogenic technique developed by Meng's group cleverly circumvents this and lets one observe relevant growth features. Liaw points out that the researchers were initially surprised by the glassy deposits because one would expect the post-nucleation growth to transition to the crystalline state. MD calculations helped identify the underlying “long incubation times” that meta-stabilized the glassy phase.
These scanning electron microscope images show that when lithium is electrodeposited at different rates, resultant growth morphology varies and is intrinsically related to the crystallographic order in the electrodeposited phase. The identical amount of lithium is deposited in these three tests. Credit: Xuefeng Wang, Y. Shirley Meng, and Boryann Liaw.
This study reveals that the amorphous lithium leads to desired growth morphology. To further test this hypothesis, the researchers examined electrodeposition at identical currents but using two different electrolytes. It is observed that the electrolyte that gives more amorphous deposits also exhibits superior performance and reversibility, further bolstering the importance of amorphous growth.
While discussing this study, Jürgen Janek and Thorben Krauskopf from Justus-Liebig-Universität Giessen, Germany (not related to the work), mention that controlling the evolution of nanostructure and microstructure of lithium is the key to reversible metal anodes. They are optimistic about the new opportunities this work promises via tailoring morphology of the electrochemical growth. They suggest that this work demonstrates a poorly understood coupling between reaction kinetics and growth that unlike classical nucleation theory predictions, exhibits larger cluster sizes for faster growths. A possible explanation is the role of surface diffusion (Energy Environ. Sci., doi:10.1039/C8EE01448E) that occurs concurrently with reaction and growth. Other exploratory opportunities are to scale up this mechanism to any current of interest and explore whether the amorphous growth is a strict criterion for homogeneous deposition.
The research team is excited about examining the regimes of glassy electrochemical growth to enable anodes made of other metals. Given its correlation to small current densities, (i.e., growth rate per unit area), one interesting possibility is to examine three-dimensional structures whose higher area-to-volume ratio can facilitate desired glassy growth for the same total current.
