Open Access Version

Abstract:
Silicon offers a nine times higher theoretical storage capacity than graphite anodes, which dominate the current commercial lithium-ion battery (LIB) landscape. Being comparably cheap and earth-abundant makes silicon a promising candidate to leverage LIB technology and meet ever increasing demands. To compete with graphite, silicon anodes need to be operated in a potential range outside the thermodynamic stability window of currently available organic electrolytes. Thus, cyclability requires the electrolyte to be kinetically stabilized by the so-called Solid Electrolyte Interphase (SEI), a layer that ideally forms once from in-/organic decomposition products of the electrolyte during the first charging cycles. While it works for graphite, the SEI on silicon fails to stabilize the electrolyte sufficiently, mainly due to silicon’s large volume swings upon de-/lithiation. Due to its physical and chemical volatile nature as well as its inaccessibility being burrowed in the cell assembly, the picture of the SEI on silicon remains elusive. To gain a deeper understanding of the SEI and ultimately improve its efficiency, in-situ and operando approaches are necessary. X-ray Absorption Spectroscopy (XAS) is an element-specific technique, which allows to probe the electronic and chemical structure of the SEI. In this work, I introduce a novel approach for transmission XAS on liquids and thin-film battery electrode materials under in-situ conditions in the soft X-ray regime. Thematically, this work is divided into two parts: 1) the introduction of this novel method and 2) its application to investigate the SEI on silicon thin film anodes. The presented technique is based on an electrochemical half-cell equipped with a sandwich of two soft X-ray transparent silicon nitride membrane windows to encapsulate the electrolyte. One of the membranes acts as substrate for the silicon thin-film anode, which is cycled with a metallic lithium counter-electrode. After the SEI has formed, a gas bubble is intentionally introduced through radiolysis by a high intensity X-ray to push out excessive electrolyte and stabilize a thin electrolyte layer on top of the SEI, keeping it in-situ. The obtained stack comprised of bubble, electrolyte thin-layer, SEI and anode, is sufficiently thin to be probed with transmission XAS in the soft X-ray regime. To illuminate the process of bubble formation, I present a Finite Element Method simulation of the temperature and concentration fields which evolve within the electrolyte under the high intensity X-ray beam. Additionally, I simulate the bubble growth through Computational Fluid Dynamics and demonstrate that the bubble ends up sitting steadily between the membranes. The second part of this work utilizes the presented method to investigate the SEI on 50 nm thick amorphous silicon thin-film anodes in 1M LiPF6 ethylene carbonate:dimethyl carbonate (50:50 v/v) electrolyte at the BESSY II synchrotron facility in Berlin. The anodes’ electrochemical characterization shows all significant features of silicon’s de-/lithiation. Complementarily, the cell is extended to allow operando Optical Microscopy (OM) and operando UV/Vis Spectroscopy (UV/Vis), where the degree of lithiation can be monitored through visible absorption. The OM reveals a reversible wrinkling of the membrane/silicon bilayer upon de-/lithiation, indicating a swing between tensile and compressive stress within each cycle. Both UV/Vis and OM verify that the anodes remain crack-free for the cycle count of the in-situ XAS investigation. The transmission in-situ XAS investigation at the silicon L-edge as well as the oxygen and fluorine K-edge unveils every component in the beam’s path. Carbon monoxide and carbon dioxide are identified as the main radiolysis products and bubble constituents. Additionally, it is shown that the degree of lithiation can be monitored at the silicon L-edge.