Reversible Li plating regulation on graphite anode through a barium sulfate nanofibers-based dielectric separator for fast charging and high-safety lithium-ion battery

Poor Li plating reversibility and high thermal runaway risks are key challenges for fast charging lithiumion batteries with graphite anodes. Herein, a dielectric and fire-resistant separator based on hybrid nanofibers of barium sulfate (BS) and bacterial cellulose (BC) is developed to synchronously enhance the battery's fast charging and thermal-safety performances. The regulation mechanism of the dielectric BS/BC separator in enhancing the Li* ion transport and Li plating reversibility is revealed. (1) The Max-Wagner polarization electric field of the dielectric BS/BC separator can accelerate the desolvation of solvated Li* ions, enhancing their transport kinetics. (2) Moreover, due to the charge balancing effect, the dielectric BS/BC separator homogenizes the electric field/Li* ion flux at the graphite anode-separator interface, facilitating uniform Li plating and suppressing Li dendrite growth. Consequently, the fast-charge graphite anode with the BS/BC separator shows higher Coulombic efficiency (99.0% vs. 96.9%) and longer cycling lifespan (100 cycles vs. 59 cycles) than that with the polypropylene (PP) separator in the constantlithiation cycling test at 2 mA cm-2. The high-loading LiFePO4 (15.5 mg cm-2)//graphite (7.5 mg cm-2) full cell with the BS/BC separator exhibits excellent fast charging performance, retaining 70% of its capacity after 500 cycles at a high rate of 2C, which is significantly better than that of the cell with the PP separator (retaining only 27% of its capacity after 500 cycles). More importantly, the thermally stable BS/BC separator effectively elevates the critical temperature and reduces the heat release rate during thermal runaway, thereby significantly enhancing the battery's safety. (c) 2024 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier B.V. and Science Press. All rights are reserved, including those for text and data mining, AI training, and similar technologies.