Ex-situ Li plating detection on graphite anodes in extremely fast-charged lithium-ion batteries using simultaneous neutron and X-ray tomography

Abstract

Problem: The long charging times required for lithium-ion batteries (LIBs) constitute a major bottleneck in the widespread deployment of battery electric vehicles (BEVs). Currently available BEVs cannot charge at rates that offer a similar experience to that of refueling a gasoline car at a gas station. Among the state-of-the-art BEVs, Tesla vehicles take ~ 1-12 hours to charge at the fastest recharge rates of 120 kW through Supercharger stations. 1 Therefore, there is a global push to enable extreme fast charging (XFC) that would reduce LIB charging times to 10-15 minutes. 2 However, existing LIBs cannot achieve this XFC goal without significantly reducing battery performance. One of the identified XFC failure pathways is a phenomenon known as "lithium (Li) plating," which severely limits battery capacity over lifetime and eventually leads to battery failure. 3 Thus, understanding the origin and characteristics of Li plating on graphite anodes is crucial to developing XFC batteries. Approach: In this work, we used simultaneous neutrons and X-ray-based tomography (NeXT) 4 as a non-destructive imaging modality to visualize Li plating across three dimensions on graphite anode ex-situ after XFC in LIBs. Since X-rays are sensitive to the electron density and neutrons to the nuclear density of the material, NeXT readily separates battery anode components such as Li, graphite, and copper, due to the complementary interaction of the two imaging probes with matter. Higher-energy X-rays are needed to penetrate the metallic components in a battery such as the copper current collector. However, X-rays lack the sufficient imaging contrast to differentiate low-Z elements, especially at high energies. Here, neutrons provide the sensitivity to differentiate graphite from Li due to the larger difference in their relative neutron cross-sections. Methodology: We performed proof-of-concept multi-modal imaging experiments at the NeXT system located on the BT-2 imaging beamline at the National Institute of Standards and Technology Center for Neutron Research (NCNR). 4 We characterized pristine and cycled graphite anode strips containing plated Li. For cycled anode strips, we disassembled the battery pouch cells and harvested graphite anodes at fully discharged condition after these were cycled under XFC conditions, specifically 9 Crate for 450 cycles. 5 Our spatial resolution was ~15-20 μm, which was sufficient to pinpoint the location of Li plating within the thickness (80-100 μm) of the graphite anode. Data analysis/Discussion: For data analysis, we first denoised the neutron and X-ray images. Then, we used Livermore Tomography Toolbox, 6 a fast and user-friendly tomographic reconstruction package developed in LabView, for 3D iterative cone beam reconstruction. Next, we used bivariate histogram phase segmentation 7 on the reconstructed neutron and X-ray