(Invited) Thermal Runaway Propagation Resistance in Lithium-Ion Batteries

Thermal runaway (TR) is a commonly-known mode of failure in lithium-ion (Li-ion) cells. TR generates large amounts of heat and fire as well as toxic gases that are potentially destructive. A Li-ion battery consists of multiple cells in which TR from a single cell could propagate creating cascading failures across multiple cells, generating high-energy shrapnel and a large quantity of toxic and corrosive materials. TR propagating between cells in multi-cell batteries causes more destruction than TR in a single-cell battery, the most striking examples being electric vehicle batteries in which fires are typically much larger than fires in single-cell mobile phones. Conventional understanding of TR propagation has been limited to the perception that heat generated from a cell under TR is carried to the adjacent cells: a.) predominantly by ejected high-temperature gases and solids; and b.) to a lesser degree by conduction through metallic parts within the battery such as cell interconnects. If the conventional understanding is correct, then removing/venting the high-temperature gases and solids ejected at the time of TR from the battery compartment will primarily prevent cell-to-cell TR propagation. In fact, although batteries have been designed with vents that provide opportunity for hot matter to exhaust away from the rest of the cells, they have not successfully prevented TR propagation. In other words, the conventional model used in describing TR propagation needs to be revised. We recently observed by infrared spectroscopy that when a cell temperature rises, but before the start of TR, flammable organic compounds such as carbonate esters are ejected from the cell. This observation sheds new light on the thermal propagation processes in multi-cell Li-ion batteries. The ester is ejected as a gas through vent valves in the cell, and it condenses outside the cell which is evident in its absorption spectra. Although the ester can be highly flammable (similar to gasoline), it requires a spark to ignite and an oxidizer to burn. If either is missing, the ester is deposited on adjacent cells, either as a liquid or a mixture of liquid and trapped gas, and is not vented out of the battery housing. The condensation of esters on adjacent cells is supported both by experimental data and computational fluid dynamics models. When the cell goes into TR, both ignition spark and oxidizers are present, which ignites the ester. The ester starts burning on the surface of the adjacent cells, sending them also into TR. This finding ultimately has implications for innovative technical solutions aimed at preventing large multi-cell battery fires.