The Impact of Silver Content on Lithium-Silver Alloy Electrode Performance with Garnet Solid Electrolyte in All Solid State Batteries

Lithium (Li) metal is one of the most promising anodes to enable the development of high energy density all solid-state batteries (ASSBs). To fully exploit its role as a practical anode the problem of Li dendrites (filaments) formation at the interface between the Li and solid electrolyte (SE) and subsequent propagation within the SE leading to short circuits must be solved. Unlike other anode materials such as graphite, Si, or Li₄Ti₅O₁₂, Li is an alkali metal that can penetrate the SE if i) there are surface defects at the interface or ii) at high charge/discharge rates. It is thus important to improve the performance of the negative electrode without the use of high operational or stack pressures. Since Li metal anode has a solid-to-solid contact with SE, it is a challenge for Li to remain in full contact with the SE during cycling especially at high current densities (rates) creating electrochemically inactive spaces (cavities/gaps/pores/voids). Our investigations show that pore formation in a Li|L_6.6La₃Zr_1.6Ta_0.4O₁₂|Li symmetric cell happens at a very low current density of 0.3 mA/cm². In literature, this is attributed to the low speed at which Li atoms move to fill these pores. A materials level approach is taken where a combination of solid-solution LiAg binary alloys and single directional stripping and plating experiments are investigated for solving this problem. We find that Li₂₀Ag alloy increase the diffusion kinetics of Li atoms. Our analysis shows that pores in ASSBs can be eliminated up to 0.8 mA/cm² in a single cell for a total capacity of 1 mAh/cm² during multi-cell single directional stripping test. Similarly, dendrites formation is suppressed till 0.8 mA/cm² as demonstrated by both single-directional and bidirectional critical current density (CCD) tests. Pore suppression is supported by Galvanostatic Intermittent Titration Technique (GITT) measurements which indicate a high Li diffusion coefficient with 2-3 orders of magnitude increase compared to reported values for pure Li consistent with previously published experimental and modeling findings. An increase in diffusion is noted by Nuclear magnetic resonance (NMR) spectroscopy with Li₁₀Ag having a Li bulk self-diffusion coefficient of 1.43 x 10^-8 cm²/s at 373 K compared to that for pure Li. Electrochemical impedance spectroscopy (EIS) shows that the despite the bulk impedance being the same, the interfacial impedance increases dramatically after stripping and this discrepancy increases with a decrease in Ag content i.e. Li₁₀Ag. The results demonstrate that a controlled amount of Ag in Li can effectively mitigate pore formation during stripping. The method and design principles discussed in our model can help guide the development of dense, pore-free anodes for suppressing dendrites in ASSBs.