Publikationsrepositorium - Helmholtz-Zentrum Dresden-Rossendorf

Si as anode material for lithium-ion batteries promises 10x the capacity of state-of-the-art graphite. However, Si anodes suffer from pulverization and electrode collapse due to large volume increase during lithiation. It has been shown that Si structures with sizes below about 200 nm remain stable [1]. Therefore, we try to understand the formation of Si nanosponge in µ-sized particles during quenching of AlSi droplets. Subsequently Al is removed from the as-produced particles by etching. Phase separation of Si and Al upon solidification of the molten AlSi alloy occurs in two stages: First nucleation and growth of primary Si grains and second formation of eutectic sponge in the Si depleted melt, with faster cooling leading to finer structures. Through modelling and simulation, the reaction pathway can be understood, allowing to optimize process parameters. For this, a model was developed, which has as initial state a fully liquid, spherical droplet with random distribution of atom species Al, Si, vacancies and oxygen impurities. A many body angular-dependent potential (ADP) has been employed which reproduces the Al-Si phase diagram quite reasonable. As the melt cools below the liquidus temperature, precipitation of primary Si takes place, followed by spinodal demixing of the melt upon reaching the eutectic. Nucleation is influenced by trace oxygen which modifies surface energies and leads to formation of sites for heteronucleation. The diffusion-reaction behavior of the species, including nucleation and/or spinodal decomposition are simulated with a 3D kinetic lattice Monte Carlo program [2] using the ADP-potential for the Al-Si system [3] with modifications added to model surface oxidation. This program enables large scale calculations by a bit-encoded lattice and lattice jumps via bit-manipulation. Our simulations qualitatively reproduce the Al-Si phase diagram, as well as composition dependent interface energies of solid Si to Al-Si melt and the nucleation behavior. The simulation results agree with the experimentally found Si nanostructures and highlight the relevance of oxygen impurities for their formation.
This work is supported by the German federal ministry for economic affairs and climate protection under grant number 01221755/1.
[1] Su et al., Adv. En. Mat. 4 (2014) 1300882
[2] Strobel et al., Phys. Rev. B 64 (2001) 245422
[3] Starikov et al., Comp. Mat. Sc. 184 (2020) 109891