Enhancement of the Thermal Stability and Electrochemical Performance of Silicon-based Anode Materials in Lithium-ion Batteries

Abstract

Si-based anode materials, specifically silicon (Si) and silicon monoxide (SiO) have attracted increasing attention in achieving high energy density in lithium-ion batteries (LIBs) owing to their high specific capacity, which is more than six times higher than the conventional graphite anode. A Si anode delivers a specific capacity of 3579 mAh g^-1, but it experiences significant volume expansion (>300%) when alloying with Li, which leads to problematic issues such as particle pulverization, continuous solid-electrolyte interphase (SEI) formation, electrode delamination, etc., resulting in rapid capacity fading and poor cyclability. SiO was later utilized as the alternative anode material due to its better cycle performance and smaller volume change (<200%) than Si. However, it suffers from a low first Coulombic efficiency (FCE) due to the reaction between the oxygen matrix in SiO and Li, which forms irreversible lithium oxide and lithium silicates during the initial lithiation process. The goal of the thesis is to understand how battery performance with Si-based anodes, including capacity, cyclability, and safety, is affected by the active materials, electrolyte design, and interphase between electrolyte and electrode, and by doing so, develop strategies to improve the stability of the batteries so that Si-based materials can be used in practical applications in the future.

Chapter 1 and 2 summarize the primary background works on anode materials by various researchers and most of the standard experimental methods in this study, respectively.

Our earlier work demonstrates two contributions to the irreversible capacity of SiO, depending on the reaction products during lithiation. Disproportionation reduces the irreversibility of SiO by forming some electrochemically inactive SiO₂ in the material, though the overall capacity is diminished. With the understanding of the contributions to the irreversibility of SiO electrodes, Chapter 3 presents the method to overcome the low FCE of SiO. We incorporate sodium carbonate (Na₂CO₃) through thermal annealing to deactivate the O matrix in SiO, accommodating the formation of irreversible lithiation products during the 1^st discharge without supplying additional Li sources. Moreover, this approach is straightforward and commercially ready, which needs no inert atmosphere, dry room, chemical solvents, or disassembly/reassembly of the cells. By adding only 5wt% Na₂CO₃, the FCE of SiO is increased from 61% to 86%. This is attributed to the incorporation of Na in the O matrix, which facilitates the SiO disproportionation into crystalline Si and SiO₂, where the SiO₂ is thought to be inactive towards Li, thus inhibiting its reaction with Li and the formation of irreversible capacity. From the cycle performance of pre-sodiated SiO half cell and full cell, this pre-treatment does not affect the original electrochemical behavior but makes a high energy density of 900 Wh L^-1 available.

When achieving such high energy density, the thermal stability of a battery cell becomes another prominent aspect because it tightly relates to LIBs commercialization and human safety. So, in Chapters 4-6, we further study the thermal stability of Si-based materials with DSC. Thermal stability is expected to be affected by the type of active materials, the type of SEI on the surface of the material, and also the electrolyte. We therefore investigate the effect of material, states of charge (SOC), electrolyte additives, and type of electrolyte solvents.

Preliminarily, thermal behavior is different among materials due to their complex structure or composition. In Chapter 4, to pinpoint the differences originating from material structure, the thermal stability of Si and SiO is compared by their exothermic reaction temperature and heat generation using differential scanning calorimetry (DSC) under the same conditions, including SOC and electrolyte composition. The DSC results confirm that SiO is safer than Si under the same conditions, owing to the less reactive lithiation products of lithium oxide and lithium silicates than lithium silicide (Li_xSi). Furthermore, the amount of fluoroethylene carbonate (FEC), a widely used electrolyte additive, is studied in relation to thermal stability. It is shown that 12.5vol% FEC gives the best thermal stability of both Si and SiO because of the formation of a LiF-rich SEI, which prevents the direct contact of lithiated electrode and electrolyte. At the same time, a higher amount of FEC will reversely worsen the thermal reaction due to the higher reactivity of FEC than ethylene carbonate (EC).

Exploring the effect on thermal performance from material difference and electrolyte co-solvent speculates that the thermal reactivity of bare electrolyte and the stability of SEI derived from electrochemical and thermal decomposition of electrolyte are of great interest. In Chapter 5, to understand the effect of electrolyte and generated SEI on thermal behavior, 1wt%, 3wt%, and 5wt% of (3-aminopropyl) triethoxysilane (APTES) are then used as the electrolyte additive in the Si anode, which does not affect much of the original electrolyte system and cyclability but impacts thermal stability with a traceable trend. NMR and XPS characterizations suggest three possible reasons. First, APTES serves as a PF₅/HF scavenger, stabilizing the electrolyte and suppressing its decomposition at a high temperature. Second, APTES may polymerize with moisture, contributing to a network covering Si particles which prevents direct contact between electrode and electrolyte. Third, APTES enables the formation of a SiO₂-rich solid-electrolyte interphase on the surface of Si particles, leading to better thermal stability.

With the information provided by the APTES additive, the essential role of electrolytes in thermal stability is further verified. The existing carbonate-based commercial electrolyte is always the major problematic component of the poor thermal stability in LIBs. To study and alleviate this issue, sulfolane (SL) is used to substitute the original electrolyte solvents in a Si anode as presented in Chapter 6. Specifically, SL-based electrolyte is prepared with different LiPF₆ salt concentrations of 1M, 1.5M, and 2M. The DSC profiles elucidate an improving trend in thermal stability of both bare electrolytes and lithiated electrodes in terms of less heat generation with increasing salt concentration. In contrast, the cycle performance is similar for 1M and 1.5M electrolytes. When the salt concentration is 2M, the capacity will decay after 60 cycles. Raman spectroscopy shows that different salt concentrations will form different solvation structures with the SL solvent. The higher salt concentration is usually thought to affect electrolyte decomposition and form salt-derived SEI, which is dense and thermally stable. XPS verified the formation of sulfurous compounds originating from SL decomposition, which might also improve thermal behavior. As the salt concentration increases from 1M to 2M, the electrolyte viscosity also rises. The conductometer speculates a poor ionic conductivity of 2M electrolyte as 2.6 mS cm^-1, which might be the reason for high polarization and fast capacity decay.

Chapter 7 presents the conclusions and suggestions for future works. Overall, this thesis provides a better understanding of Si-based anode materials and possible strategies to enhance their FCE and thermal behaviors. Material structure, SEI conformation, electrolyte composition, the type and amounts of additives all influence the thermal stability based on our study. Specifically, the electrolyte is thought to have the most significant effect. By feasible adjustment of materials together with proper electrolytes, a cell with good cycle performance and thermal stability will become possible.

Date of Award	13 Jun 2022
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Yau Wai Denis YU (Supervisor)