Ionic liquid assisted nanocellulose production from microcrystalline cellulose: Correlation between cellulose solubility and nanocellulose yield via COSMO-RS prediction

[Display omitted] •Cellulose solubility in ionic liquids predicted via COSMO-RS.•The predicted results revealed a negative correlation between cellulose solubility and NC yield.•Increasing the alkyl chain side group of the cation increased the yield of NC.•The hydrogen bond acidity (α) of the ILs played a major role in nanocellulose yield. The use of ionic liquids (ILs) in nanocellulose (NC) production from cellulosic materials has gained significant interest due to the intrinsic physical solubility of cellulosic materials in many ILs. However, given the numerous possible cation–anion combinations, selecting potential candidates from thousands of ILs for NC production is rather difficult. In addition, NC yield depends on the types of ILs and their capability in dissolution of cellulosic materials. In this study, we investigate NC production from microcrystalline cellulose (MCC) using various ILs as a reaction medium. Firstly, the prediction of MCC solubility in 300 ILs (15 cations and 20 anions) was obtained by deploying the conductor-like screening model for real solvents (COSMO-RS) which has emerged as a reliable and promising tool for solvents screening based on the dominant interactions of H-bonds, misfits, and Van der Waals energies of the fluid mixture. Experimental measurement of NC production from MCC in six ILs, namely [Emim][OAc], [Bmim][OAc], [Bmim][Cl], [Emim][HSO4], [Bmim][HSO4], and [Hmim][HSO4], was conducted. Based on the logarithmic activity coefficient prediction at infinite dilution, there was a negative correlation between cellulose solubility and experimental NC yield obtained. [Hmim][HSO4] IL showed the highest nanocellulose yield (65%) due to a relatively weaker H-bonding ability with cellulose. Moreover, the multiparameter scale of the Kamlet-Taft of six ILs revealed that the increase in alkyl side chain length of the IL cation resulted in a decrease of hydrogen bonding acidity, which contributed to higher NC yield. The quantitative prediction of COSMO-RS, coupled with the experimental study, can therefore be utilized to screen, and classify the potential ILs to prepare nanocellulose.