Validating empirical force fields for molecular-level simulation of cellulose dissolution

Abstract

The calculations presented here, which include dynamics simulations using molecular mechanics force fields and first principles studies, indicate that the COMPASS force field is preferred over the Dreiding and Universal force fields for studying dissolution of large cellulose structures. The validity of these force fields was assessed by comparing structures and energies of cellobiose, which is the shortest cellulose chain, obtained from the force fields with those obtained from MP2 and DFT methods. In agreement with the first principles methods, COMPASS is the only force field of the three studied here that favors the anti form of cellobiose in the vacuum. This force field was also used to compare changes in energies when hydrating cellobiose with 1–4 water molecules. Although the COMPASS force field does not yield the change from anti to syn minimum energy structure when hydrating with more than two water molecules – as predicted by DFT – it does predict that the syn conformer is preferred when simulating cellobiose in bulk liquid water and at temperatures relevant to cellulose dissolution. This indicates that the COMPASS force field yields valid structures of cellulose under these conditions. Simulations based on the COMPASS force field show that, due to entropic effects, the syn form of cellobiose is energetically preferred at elevated temperature, both in vacuum and in bulk water. This is also in agreement with DFT calculations.