(231a) Disorder and Entropy in Molecular Crystals

Many computational methods for predicting polymorphic behavior generate many possible crystal structures and rapidly evaluate the lattice energies of these candidate structures to determine the most likely to be observed experimentally. However, lattice energies alone do not fully represent the stability of systems that are at room temperature and sample a range of configurations due to the thermally-induced motions at non-zero temperatures, which have both harmonic and anharmonic character.

In this talk, I survey our work using molecular dynamics to examine the free energy differences between enantiotropic crystalline systems, where the approximation that the lowest lattice energy crystal is the most stable must necessarily break down. We have found that although small rigid crystals are well-described by quasiharmonic approximations, even moderately conformationally flexible molecules can have extremely complex potential energy surfaces, making it hard to find appropriate structures from which to compute quasiharmonic approximations. We also find that many of these minima are accessible at room temperature, and proper annealing can help identify appropriate ensembles of structures to use for quasiharmonic analysis, and that bulk crystals differ thermodynamically in important ways from single unit cells. We also describe the extent to which there is moderate entropy-enthalpy compensation between different models and computational methods, potentially opening the door for using lower levels of computational detail to determine stability at room temperature even if classical potentials do not provide sufficient accuracy to determine lattice energies.