Application of pulse EPR to investigate degradation processes in Li-ion batteries

Li-ion batteries suffer from degradation processes arising particularly during long-term usage or operation outside of recommended cycling conditions. Similarly, degradation hinders the safe implementation of 'beyond lithium' batteries, which promise significant performance enhancements. To mechanistically understand degradation and test counter-measures, a variety of analytical techniques have been applied. However, certain degradation phenomena such as the morphology evolution of metallic lithium deposits and the behavior of dissolved transition metal ions in solution are hard to address by established analytics, and are therefore still largely unexplored. In this thesis, pulse Electron Paramagnetic Resonance (pEPR) is demonstrated to contribute a complementary perspective on battery degradation, addressing metallic lithium formation and evolution as well as transition metal complexation and precipitation. In the first part, pEPR of conduction electrons in metallic lithium that has formed at the battery anode is discussed, where the skin effect and electron mobility are key characteristics. The measurement of relaxation times and electron spin nutations enables semi-quantitative approximation of the lithium morphology, which is irregular and microstructured when deposited electrochemically. The dynamics of lithium deposits are studied during operation with a maximum time resolution of 100 ms. At lithium metal anodes, pEPR reveals the continuation of morphology changes for several seconds, even after the current flow terminated, which is attributed to surface diffusion leading to a smoothened morphology. At graphite anodes operated at C-rates of up to 18C, detrimental lithium plating is monitored and pEPR provides details on the plating onset, the time-dependent plating rate, the partial intercalation following the charging process, and the relative amounts of dead lithium. In the second part, electron-nuclear spin interactions in transition metal complexes are exploited to derive solvation preferences, using pEPR and Density Functional Theory (DFT). Experiments were performed at cryogenic temperatures to enable spin manipulation and detection. First, dissolved Mn${}^{2+}$ ions from cathode dissolution and Cu${}^{2+}$ ions from current collector dissolution are modelled by dissolving salts in premixed electrolyte solutions. Mn${}^{2+}$ and Cu${}^{2+}$ ions are both found to be mainly coordinated by cyclic carbonates. However, if stored at 35 °C for 24 h, their behavior diverges. Mn${}^{2+}$ is selectively coordinated by fluorophosphate ligands that formed via heat-induced decomposition of the electrolyte salt LiPF${}_6$ with trace H${}_2$O. In contrast, relaxation and hyperfine data of Cu${}^{2+}$ species indicate partial precipitation and unaltered complexation of Cu${}^{2+}$ in solution. To investigate dissolved transition metal ions in operating batteries, V${}_2$O${}_5$ cathodes are used, which exhibit severe dissolution already during the first discharge, liberating vanadyl ions (VO${}^{2+}$). Among others, glycol dianion ligands formed through cyclic carbonate decomposition are found to selectively coordinate to VO${}^{2+}$. Additionally, chemisorption at conductive carbon surfaces can be postulated based on measurements with deuterated electrolyte solvents. If stored at 45 °C for a week, fluorophosphate and fluorophosphite ligands with phosphorus oxidation states +V and +III are determined. The obtained complexes exhibit remarkably large distributions of hyperfine coupling constants over around 40 MHz that allow the analysis of spin delocalization pathways.