Evaluating Li Anode Performance in Coin Cells

CR2032-type coin cells were assembled to deposit Li on the current collectors to evaluate the Coulombic efficiency, electrochemical impedance spectra and other properties. The cells were assembled in an argon-filled glove box. The coin cell was composed of a Li foil as the counter/reference electrode, a Celgard separator, and a current collector as the working electrode. The electrolyte was 1 M LiTFSI in DOL/DME (1:1 by volume, 30 μl, BASF) without any additives unless noted otherwise. The Coulombic efficiency was tested at 0.5 mA cm⁻² on a LAND electrochemical testing system at room temperature. The batteries were first cycled at 0–1 V (versus Li⁺/Li) at 50 μA for five cycles to stabilize the SEI and remove surface contaminations. After that, 1 mA h cm⁻² of Li was deposited onto the current collector and then charged to 0.5 V (versus Li⁺/Li) to strip the Li at 0.5 mA cm⁻² for each cycle. The Coulombic efficiency was calculated based on the ratio of Li stripping and plating. Electrochemical impedance spectra measurement was performed using an Autolab workstation (Metrohm) in the frequency range of 100 kHz to 100 mHz after specific cycles.
Symmetric cells were employed to evaluate the cycling stability and cycle life (short-circuit time T_sc) of the Li anodes on different current collectors. The symmetric cell was assembled using a hollow spacer substituting for the Celgard separator in a CR2032-type coin cell, as illustrated in Supplementary Fig. 11. The electrolyte (1 M LiTFSI in DOL/DME, 200 μl) was carefully charged into the spacer without entrainment of bubbles. For the long-term galvanostatic discharge/charge test, 2 mA h cm⁻² of Li was first deposited on the current collectors at 0.5 mA cm⁻² and the cells were then charged and discharged at 0.2 mA cm⁻² for 2.5 h in each half cycle. For the unidirectional galvanostatic polarization (accelerated test for T_sc), Li was continuously plated onto the current collectors from the counter electrode at 0.5 mA cm⁻² until short circuit. The average T_sc was obtained from at least three cells for each current collector.
For full cells with 3D Cu-based Li-metal anodes, LiFePO₄ (Sanxin Industrial) was employed as cathode material. LiFePO₄ was casted on an Al foil with an areal capacity density of ∼0.5 mA h cm⁻². The 3D Cu was first assembled into a half cell using a Li foil as counter electrode. After plating 1 mA h cm⁻² of Li metal into the 3D current collector, Li anode was extracted from the half cell and reassembled into a full cell against LiFePO₄ cathode. The electrolyte was the same as that in the half cells (1 M LiTFSI in DOL/DME, 30 μl). Assembly of pouch cells was similar to that of the coin cells. The electrodes (∼42 cm² in area) were stacked and assembled in a pouch cell. Li anodes plated in the 3D current collector were assembled into a pouch cell against LiFePO₄ cathodes with 4 ml of the electrolyte to gain the pouch full cell with a capacity of ∼40 mA h.