A Method for Controlled Oxide and Carbon Coating of Silicon Nanoparticles as Anode in Lithium-Ion Batteries

Abstract

Silicon has emerged as a promising alternative to graphite in lithium-ion batteries. Silicon can, theoretically, provide up to ten times higher capacity due to its high alloying capabilities with lithium. This improvement in initial capacity is however followed by capacity fade during the early cycles, which is undesirable for commercial applications. The low cycle life is in turn due to the high alloying with lithium, as this leads to a large volume expansion and subsequent contraction of the material.

In this work, a method for coating silicon has been investigated to eventually be able to buffer the volume change associated with cycling. Double core-shell particles were fabricated by wet chemical methods, and can serve as a base for creating yolk-shell particles in the future. Silicon nanoparticles from a Free Space Reactor with diameters between 50 nm and 1 ¹m have successfully been coated with a 50-80nm thick oxide layer by the use of tetraethyl orthosilicate, designed to be a sacrificial layer to be etched away at a later state. Following the oxide layer, a thin carbon-coating has been applied by a resorcinol-formaldehyde resin and pyrolysis, which interconnects the oxide coated particles in a web-like morphology. The coating steps have been thoroughly characterized by electron microscopy, energy dispersive spectroscopy, x-ray diffraction, thermogravimetric analysis, dynamic light scattering and nitrogen adsorption. Problems related to the upscaling of this process were revealed during carbon-coating, as it did not give sufficient control and reproducibility with coating thickness and morphology for the chosen starting material.

Li-ion battery half-cells were prepared with the reference silicon and electrochemically cycled, and one half-cell exhibited a superior lifetime (up to 1,000 cycles) when cycled at a reversible capacity of 700 mAh/g, while others could retain 1,200 mAh/g for only 150 cycles. The cycling of both amorphous and crystalline silicon was done, but no lifetime-dependence on crystallinity was found. Crystalline silicon anodes with a loading of 0.4 mg/cm2 displayed an irreversible capacity loss (ICL) in the first cycle of 8-9%, and a reversible capacity of 1,750 mAh/g for 120 cycles, while a higher loading of 0.5 mg/cm2 displayed the same ICL but a capacity of 2,500 mAh/g for only 50 cycles. Amorphous silicon anodes with a loading of 0.5 mg/cm2 displayed a low ICL of 1%, and a reversible capacity of 2,000 mAh/g for 70 cycles. The lifetimes are clearly more dependent upon loading and cell fabrication, than of initial crystallinity of the material.

In sum, this work has provided a starting point for the fabrication of yolk-shell particles by using Si nanoparticles from an FSR. A homogeneous coating of oxide has been applied, and a carbon coating was observed for a few samples. Electrochemical results from the cycling of pure silicon showed a high dependence on loading and cell fabrication, and it is therefore imperative to keep the electrode fabrication and cell assembly as similar as possible for comparison purposes.