Synthesis and application of semiconducting nanomaterials as high-performance lithium-ion battery anodes

This thesis involves the synthesis and application of high-performance lithium-ion battery (LIB) anode materials, based on nanostructured electrode designs; specifically, Li-alloying silicon and germanium nanowires (NWs) and Li-intercalation/conversion tungsten diselenide nanocrystals (NCs). Carbon-based Si nanocomposite materials are similarly investigated. The core chapters are formatted as research articles with introductory summaries at the beginning of each chapter.

The modest gravimetric capacity of commercial graphite (372 mAh g-1) is becoming increasingly incompatible with the next generation of high-energy, fast-charging LIB formulations. Poor conductivity of graphite (Gt) limits fast-charging capabilities, unsuitable for high-power applications. Additionally, the low operating voltage of Gt (0.05 V vs Li/Li+) represents an ever-present risk of Li plating under stringent operating conditions (i.e., high current, low temperature). High-capacity alternatives like Si (3579 mAh g-1) and Ge (1384 mAh g-1) have gained major interest over recent years, boasting capacities several multiples that of Gt. Also, low operating voltages (~0.4 V vs Li/Li+)  benefit both cell output and stability, elevating full cell (FC) voltage while limiting Li plating issues.

The electrochemical instability of bulk Si and Ge has greatly hindered their utility as practical LIB anodes. High capacity comes at the cost of severe volume fluctuations during charge/discharge, expanding by ~300 % when fully lithiated. Nanostructuring of bulk material has been shown to significantly enhance cell stability, providing suitable porosity and void spaces to accommodate volume expansion and ease mechanical stress build-up. Additionally, larger surface area to volume ratios shorten the Li+diffusion length, enhancing both rate performance and coulombic efficiency (CE). To date, the incorporation of such nanomaterials into commercial-grade cells has been hindered by several critical roadblocks, including electrochemical instability, loading limitations of bottom-up growth techniques, and inefficient scalability of synthetic protocols. Overcoming such issues is the driving force behind advancements in research into Li-alloying anode materials and as such, this work focuses on further developing these critical components.

Chapter 3 describes the application of binder-free Ge NWs as compatible wide?temperature anodes in LIB FCs vs LiCoO2 (labelled LiCoO2||Ge), with their performance contrasted to LiCoO2||Graphite (labelled LiCoO2||C) cells between the temperature limits of -40 oC to 40 oC. LiCoO2||Ge reached capacities 30× and 50× that of LiCoO2||C cells at -20 oC, and -40 oC, respectively. At low temperatures (LTs), propylene carbonate (PC)  enhanced performance by suppressing electrolyte freezing and Li carbide formation, as well as increasing the LiF, Li ether and poly(VC) contents of the SEI. At high temperatures (HTs), a dual addition of ethyl methyl carbonate (EMC) and lithium bis(oxalato)borate (LiBOB) improved capacity retention threefold, slowing the fade rate from 0.22 % cycle-1 to 0.07 % cycle-1 at 40 oC. The incorporation of both additives increased the LiF content while concurrently lowering the overall fluorine content through boron substitution. LT cycling was found to positively impact cell stability with capacity retention improving as cell temperature was lowered. Improved capacity retention at LT was accredited to a  combination of delayed NW amorphization, suppressed high-order Li15Ge4 formation and  incomplete lithiation, limiting complete expansion of the active material.

Chapter 4 tackles loading saturation issues of binder-free Si NWs by substituting planar stainless steel (SS) with a highly textured copper silicide (Cu15Si4) NW substrate. Cu is largely regarded as the optimal anode current collector for LIBs due to its superior conductivity over SS and other carbon-based substrates. However, direct growth of Si on Cu leads to the formation of electrochemically inactive Cu-Si compounds, like CuSi NWs. A one-pot solvent vapour growth (SVG) process was developed to grow CuSi NWs from Cu foil for use as high surface areas substrates for Si NW growth (> 1.6 mg cm-2). In this one-pot approach, CuSi NWs were synthesised via a vapour-solid-solid (VSS) process, behaving as textured surfaces for vapour-liquid-solid (VLS) growth of Si NWs (denoted Si/CuSi). Si/CuSi demonstrated excellent long-term stability, delivering a stable areal capacity of 2.2 mAh cm-2 after 300 cycles. This stable capacity coincided with the formation of a highly interwoven Si mesh, showing excellent adhesion to the underlying CuSi layer. Overall, complete anode fabrication is achieved within a single reaction, eliminating the requirements for preliminary catalyst deposition techniques and post-synthetic plasma-enhanced chemical vapour deposition (PECVD) a-Si coating techniques.

In chapter 5, directly-grown silicon nanowire-on-graphite composite anodes (denoted SiNW/Gt) were formulated to synergistically combine the high capacity of Si with the stability of Gt. Five different compositions (12-45 wt% Si) were tested as suitable battery anodes in both half-cell (HC) and FC configurations and their electrochemical performances were compared to pure Gt and Gt-free SiNW. Low-level Si additions (12-18 wt%) best accessed the high capacity of Si while minimising inactive mass gains from supplemental binder and additive additions. A high Si content was encumbered by weakened a capacity retention and excess dead weight, negatively impacting specific energy of FC analogues. Notably, the specific energy of pure SiNW fell below pure Gt after 200 cycles, highlighting the inadequacies of Si-rich slurries for stable cell performance. Direct growth of Si NWs on Gt allowed for a robust contact at the Si-Gt interface, with improved stability over a comparable composite mixture of Si NWs and Gt.

Chapter 6 investigates the effects of both crystal phase and morphology on the electrochemical behaviour of WSe2 NCs. Firstly, to analyse the effect of crystal phase on the electrochemical properties of WSe2, polymorphs of 2 different crystal phases (namely metallic 1T’ and semiconducting 2H) were analysed. The cycling behaviour of 2H nanosheets (denoted NS-2H) was characterised by a single step intercalation/deintercalation process, delivering a specific capacity of 498 mAh g-1. Metallic IT’ nanosheets (denoted NS-1T’) showed no distinct redox peaks with the low capacity of 156 mAh g-1 over the wide cycling potential attributed predominantly to non?faradaic pseudocapacitance of the W redox couple. Secondly, to analyse the effect of morphology on the electrochemical behaviour of WSe2, 2H nanoflowers (denoted NF-2H) were compared to NS-2H. NF-2H exhibited a distinct two-step charge/discharge process, involving intercalation into WSe2 followed by conversion of the intermediate LixWSe2 structure to metallic W and Li2Se, delivering 982 mAh g-1 after 100 cycles. A continual capacity increase was noted over 100 cycles, due to the continued precipitation of metallic Se from the charged Li2Se structure for further alloying with Li.