Quantification and Simulation of Thermal Decomposition Reactions of Li-Ion Battery Materials By Simultaneous Thermal Analysis Coupled with Gas Analysis

Abstract

Internal short-circuit is one of the major safety hazards in lithium-ion battery systems, possibly leading to single cell thermal runaway (TR) and propagation to neighbouring cells. Therefore, developing tests to assess safety properties of batteries is essential to prevent potential safety hazards to the users. The main goal of this work is to understand the thermal effects of the most important parameters for triggering thermal runaway, induced by internal short-circuit, in a Graphite-Lithium Nickel-Manganese-Cobalt-Oxide (NMC 111) battery cells. This investigation will allow the reduction of parameters to the most significant ones on the tests outcome, reducing the number of test needs. The first part of this work, presented here, aims at the identification of the main decomposition processes, using Differential Scanning Calorimetry (DSC) and Thermal Gravimetry (TGA) combined with gas analysis (FTIR and GC-MS) and the subsequent development of reaction kinetics model. The activation energy, frequency factor and heat of reaction of the different sub-processes taking place in a thermal event are calculated using 3 different heating rates: 5, 10 and 15ºKmin-1. It is found that both the anode and cathode thermally decompose in multiple parallel and consecutive reactions between 100-600 ºC. A double breakdown mechanism of the SEI layer is suggested to describe the anode thermal decomposition profile. The experimental data show a partial breakdown of the primary SEI layer along with secondary SEI formation. With increasing temperature, a second breakdown occurs with decomposition of the organic/polymeric part from the secondary SEI layer, leading to the full consumption of intercalated Li. Processes like EC (ethylene carbonate) evaporation with simultaneous EC decomposition and Li-electrolyte reactions have been identified. In a third phase, the binder and the more stable degradation products decompose. A combination of diffusion-type kinetic and Arrhenius-type kinetic with Kissinger analysis has been suggested to model the anode decomposition profile. With respect to the cathode decomposition, the following processes have been identified: EC evaporation, NMC(111) decomposition with release of oxygen and combustion of EC. At a later stage, the decomposition of binder and combustion of conductive carbon additive with the evolved oxygen from NMC decomposition have been observed. The thermal decomposition profile of cathode has been described using Arrhenius-type kinetic and Kissinger analysis. The simulated heat flow curves for both electrodes show good agreement with experimental ones. In particular for the anode, the simulated double breakdown mechanism of the SEI layer fits better the experimental curve than the single breakdown mechanism, in agreement with the literature. The simulated cathode curve shows a good fit with the experimental one when taking into account the electrolyte. From the thermal data and measurement uncertainties, an approximate kinetic model is developed to simulate the dynamic of a thermal runaway in a realistic way. The second part of this work (not shown here) is to simulate TR and to calculate the probability of thermal runaway as a function of the parameters of a triggering method.