Cellulose dissolution in mixtures of ionic liquids and molecular solvents: The fruitful synergism of experiment and theory

•Cellulose dissolves in 1-butyl-3-methylimidazolium ionic-liquids (ILs)/molecular solvents.•Biopolymer dissolution depends on IL anion (acetate and chloride), binary solvent composition and temperature.•Molecular dynamics simulations show preferential solvation of cellulose by IL anion.•Conclusions of molecular dynamic simulations corroborated with conductivity and turbidity of cellulose solutions.•Synergism between experiment and theory can be used to screen cellulose solvents, saving labor and material. We investigated the dependence of dissolution of microcrystalline cellulose (MCC) in binary solvent mixtures of ionic liquids (ILs) and the molecular solvents (MSs) N,N-dimethylacetamide (DMAc) and dimethyl sulfoxide (DMSO). The ILs we used are based on 1-(n-butyl)-3-methylimidazolium cation, BuMeIm+, with acetate (AcO−) and chloride (Cl−) anions. As experimental variables, we changed the mole fraction of the MS (χDMAc or χDMSO) and the temperature (T). MCC dissolution was determined using an optical microscope and expressed on the mass fraction scale (m%). The order of maximum m% was BuMeImAcO-DMSO > BuMeImAcO-DMAc ≫ BuMeImCl-DMSO. To rationalize this dependence, we used molecular dynamics (MD) simulations employing, as a model for cellulose, a crystallite of eight chains, each containing ten anhydroglucose units (AGUs). We calculated the root-mean-square-deviation (RMSD) and the radial distribution function, g(r). Values of RMSD showed that the separation of decamer-chains follows the same order as the maximum m%; more chain separation was taken to indicate greater crystallite dissolution. Additionally, the maximum values and the sharpness of the g(r) peaks followed the same order of dissolution, due to a combination of hydrogen bonding between the hydroxyl groups of the AGU and ions of the IL. We also recorded the turbidity of MCC-IL-DMSO solutions and the conductance of cellobiose-IL-MS solutions; the results of both experiments corroborated our interpretations of the MD simulation results. Our study suggests that MD simulations can be fruitfully employed to predict the best candidates among a series of solvents, e.g., IL-MS. This screening saves labor and material in the search for optimal solvents for cellulose dissolution.