Are Solid-State Batteries Safe? A Thermodynamic Analysis

A.M. Bates, Y. Preger, L. Torres-Castro, K.L. Harrison, S.J. Harris, J. Hewson
Sandia National Laboratories,
United States

Keywords: solid-state battery, Li-ion battery, safety, modeling


Solid-state battery technology has garnered substantial attention in recent years due to the promise of enhanced safety, energy density, and charge rate. Research has shown that small format, single layer solid-state batteries can withstand such abuse as nail penetration, cutting, and bending without going into thermal runaway. These demonstrations, along with the understanding that the flammable liquid electrolyte (LE) is replaced by a more stable solid electrolyte (SE), have contributed to the assumption of inherent safety. However, to date, the safety of large format (>1 Ah) solid-state batteries has not been probed significantly. In the battery safety community, it is well recognized that cell size directly correlates to abuse response. As cell size is increased, a point may be reached at which the abuse response is significantly different. In this work, we probe the question of safety in solid-state batteries by bounding the upper limit of heat release through thermodynamic modeling. These heat release values, in comparison with a conventional Li-ion battery (LIB), provide an indication of the potential abuse response we may expect. We also consider the impact of including a small volume of LE, as high interfacial resistance between electrode and SE is one of the major challenges in SSB technology, preventing widespread commercialization. In this work, we calculated the thermodynamics of failure for three battery configurations: all-solid-state battery (ASSB), solid-state battery with some LE (SSB), and a conventional LIB. These battery configurations were studied in different failure scenarios including thermal runaway from external heating, internal short circuit, and catastrophic mechanical failure of the SE. Thermal runaway from external heating included the following well-documented reactions: cathode decomposition, LE reaction with O2, and LE reaction with the lithiated anode. Internal short circuit considered the complete conversion of stored electrochemical energy to heat. Catastrophic mechanical failure allowed for O2 generated at the cathode to move through a broken SE to react with a Li-metal anode. The volume of LE in the SSB and LIB was varied, and the resulting heat release and potential temperature rise was compared for different ASSB, SSB, and LIB formats of increasing energy density. For the external heating failure scenario, the potential temperature rise of the SSB with LE was lower than the threshold for cascading propagation. We suggest that the volume of LE will be a question of safety risk versus performance and large-scale manufacturability. As energy densities improve, it is clear from the high potential temperatures calculated for short circuit failure, that safety in SSBs of all types must be evaluated. In addition, we found that significant heat release pathways do exist for even the ASSB.