(603d) First Principles Study of Aluminum Doped Polycrystalline Silicon As a Potential Anode Candidate in Li-Ion Batteries

Addressing sustainable energy storage remains crucial for transitioning to renewable sources. While Li-ion batteries have made significant contributions, enhancing their capacity through alternative materials remains a key challenge. Polycrystalline silicon is a promising anode material due to its tenfold higher theoretical capacity compared to conventional graphite. However, its practical implementation is limited by a main bottleneck, the huge volumetric expansion after lithiation (310% for Li22Si5) which causes a large build-up of stress, resulting in the pulverization of the material and immediate capacity loss during cycling. We propose a novel approach to mitigate this issue by doping trace amounts of aluminum into the polycrystalline silicon electrode using ball milling. We employ density functional theory to establish a theoretical framework elucidating how grain boundary (GB) sliding, a key mechanism involved in the volumetric expansion, is effected by the presence of aluminum. The prevalence of Σ3 {111} GB in polycrystalline silicon is revealed in our GB characteristics quantification by electron backscattered diffraction. Hence, we consistently employed the Σ3 {111} GB (see Figure 1 and 2) in all our simulations. Basin hopping, a global optimization technique, is used to identify the most favorable aluminum segregation configuration. Subsequent sliding simulations (see Figure 3 for the results of undoped silicon) revealed facilitation of GB sliding and prevention of stress build up in doped compared to undoped silicon, reducing the likelihood of failure. Notably, a strong dependence on aluminum concentration is observed at lower concentrations and diminishing dependence at higher concentrations, confirming the effectiveness of trace doping. To validate our theoretical predictions, we conducted capacity retention experiments on undoped and Al-doped polycrystalline silicon samples. The results demonstrate significantly reduced capacity fading in the doped sample, corroborating the theoretical framework and showcasing the potential of aluminum doping for improved Li-ion battery performance.