Non-sticky Li-alloy leaves for long-lasting secondary batteries

Can Zhang,Huimin Fan,Xinlong Chen,Hui Xu,Jiatao Lou,Yao Li,Yunhui Huang,Sa Li
DOI: https://doi.org/10.1039/d2ee02135h
2022-01-01
Abstract:Metal foils are used for packaging and decoration, and gold leaf down to 0.1 mu m thickness has been made since ancient Egyptian times. The chemical inertness of gold, causing non-stickiness, and its malleability, underlie the success of "Goldbeating". In contrast, lithium metal is extremely reactive and sticks to rollers as well as degrades in air. How to make one-to-few mu m thick, large-area lithium foil is a key challenge for the battery industry, as such foils can provide extraordinarily longer cycle life to rechargeable batteries when standard anodes such as graphite or silicon are prelithiated right before cell assembly, by mechanically pressing the thin foil to the anode. Here, we show that the addition of a small alloy-element on the order of 1 at% to the body-centered cubic (BCC) phase drastically changes the oxidized surface layer composition, making the Li alloy much less sticky towards the rolling machine. By forming a protective passivation layer, the air stability is dramatically enhanced, so it can be stored for up to 12 hours without significant performance degradation. Also, there is a 10x strengthening effect, making the sheet much less prone to ductile rupture damage during rolling, while maintaining malleability. As a result, we are able to fabricate large-area lithium foil in a roll-to-roll manner down to 1 mu m thickness. Such an ultra-thin Li foil anode demonstrates superior electrochemical performance directly in pouch cells (>320 W h kg(-1)), in which the 0.8 A h pouch-cell achieved a 100 cycle life under lean electrolyte conditions of 2.67 g A h(-1). Furthermore, we applied 5 mu m Li to graphite prelithiation, having an extra 1 mA h cm(-2) lithium inventory (the amount of cyclable lithium in a full cell), the capacity of the LFP//graphite full cell increases by 18%, with an average coulombic efficiency of >99.99% over more than 300 cycles.
What problem does this paper attempt to address?