Synopsis: Predicting crystal structures
Replacing tedious hours of trial and error in the lab, computer simulations can, in principle, predict the myriad structures that atoms, molecules, and nanoparticles can form. The most powerful algorithms for these simulations mimic a realistic growth environment, including the effects of temperature, the range of the particle interactions, and the existence of multiple states that lie close in energy, without costing inordinate amounts of computation time.
For a suspension of nanoparticles, which almost only interact when they come in direct contact, entropy can be the dominant factor that determines how the nanoparticles will solidify. With this type of system in mind, Laura Filion and colleagues at Utrecht University in The Netherlands present a method that can predict crystal structures at a finite temperature. Their Monte Carlo simulations, which are published in Physical Review Letters, extend earlier molecular dynamics calculations so that the crystal structure can emerge out of the fluid phase.
Filion et al. compare their predictions against those from established methods for a wide range of systems, including mixtures of hard spheres and star shaped polymers, which interact over a longer range. And, to encourage the exploration of something new, they predict candidate crystal structures for nanoparticles with asymmetric dumbbell, bowl-like, and platelet shapes. – Jessica Thomas