Degradation of lithium-ion batteries under automotive-like conditions: P2D model-based understanding and ex-situ validation

Despite its worldwide commercialisation, the degradation of lithium-ion battery technology is still a hot research topic. Batteries are known to decrease in capacity and increase in internal resistance, but it is quite uncommon to further investigate the performance decay, distinguishing classes of ageing mechanisms (resistive, kinetic and mass-transport) and relating them with the operation. This work exploits the P2D model to understand the performance decay of ageing cells from a physical perspective. A complex experimental campaign combining 13 different automotive-like cycles, applied to commercial battery samples to recreate the degradation of batteries under realistic conditions, is analysed with such methodology. Along the ageing tests, physical models' parameters are periodically identified by means of particle swarm optimisation applied over characterisation tests. Parameter evolution is then correlated with specific degradation mechanisms, related operating conditions and performance decay. A significant decrease in electrolyte conductivity and lithium solid-state diffusivity within the positive electrode are detected, progressively inducing heterogeneous operation and worsening of both efficiency and capacity retention. Particularly, cycle depth appears to promote particle cracking and loss of positive electrode material. Post-mortem analyses are then performed to support the interpretations on degradation mechanisms, confirming the degradation of electrolyte and positive electrode. • Exploiting EIS, P2D model interprets physical aging of 13 automotive-cycling tests. • Cycles depth promotes particle cracking and loss of positive electrode material. • Electrolyte conductivity and Li inventory linearly decrease with number of cycles. • Low ionic and bulk diffusivity progressively drive heterogeneous cell operation. • Post-mortem analyses confirm both the interpretations and methodology reliability.