Development of Monolithic Silicon Based Anodes for Lithium Ion Batteries

Abstract

Traditionally lithium-ion batteries (LIBs) use graphite as the anode material because it is very stable and has a very well understood insertion/extraction lithiation mechanism. These batteries are used in many applications around the world due to the high power-energy ratio of these batteries when compared to most other modern chemical batteries. The maximum theoretical capacity of graphite is 372 mAh/g, which works decently for most applications, and it is stable enough to last well over 1000 charge-discharge cycles before significant capacity loss. This has been a standard for many years, but with increasing demand for higher energy
density battery applications that can last longer between charging cycles and some concern in the mining and refining of graphite, new materials are being investigated and silicon is a promising contender. Silicon has a theoretical capacity of 4200 mAh/g as an anode in lithium ion batteries, but is much less stable and often has significant capacity loss after 100 or fewer cycles. This is due to the material swelling between 200-300% of its initial volume when lithiated, causing several forms of degradation to occur much faster than in graphite anodes. A potential solution to this issue is using a silicon sub-nitride (SiNx) anode material, which
has been shown experimentally to have capacities on the order of 1500+ mAh/g and higher stabilities of 200- 300 cycles or more. There is room for improvement before these batteries can rival those with graphite anodes, but this thesis aims to move one step closer to bridging this gap and making silicon-nitride anodes commonplace in lithium ion batteries. Compared to graphite, silicon is a very abundant element with a massive amount of research and industry already in place that could potentially aid in making these anodes more readily available on a large scale. In this work, plasma enhanced chemical vapor deposition (PECVD) is used to deposit layers of silicon and silicon nitride with varying compositions and mass loadings onto textured copper foil current collectors that are then made into anodes in LIBs. PECVD is an effective and scalable technology, allowing for high precision control over deposition conditions of thin films. The composition of the film is determined by the flow rate ratio of the precursor gases silane and ammonia that are then ionized together to deposit SiNx onto the foil. Physical
and electrochemical analyses are then performed to determine the specific compositions of these materials and address how these two parameters affect their electrochemical performance as anodes. After thorough testing is done, a final objective is explored by looking into how using a thin layer of the most stable SiNx on top of a layer of pure silicon might improve battery performance as an artificial solid electrolyte interface (SEI). Three different deposition times of 30 minutes, 1 hour and 1.5 hours along with five flow rate ratios (roughly)
corresponding to pure Si, SiN0.2, SiN0.4, SiN0.6 and SiN1.1 were used to create fifteen different sets of anode material. These materials were tested physically to determine the chemical composition and mass loading of the material, then tested as anodes to determine their specific capacities and stability in batteries. It was found that the lower mass loadings corresponded to higher specific capacities and more stable anodic performance, and the most stable materials for the lowest mass loadings were SiN0.4 and SiN0.6 with average specific capacities of 2204.7 mAh/g and 1135.6 mAh/g and capacity retention of 105% and 117% over 100 cycles, respectively.
The bi-layer depositions made with the SiN0.6 as a thin top layer did not perform as well as the pure SiN0.6 anodes, leading the author to recommend research into other methods of enhancing battery performance.