Using molecular structure of predict purification of hemicellulose polymers

Solvent selection is the primary step in the recovery of hemicellulose using centrifugal partition chromatography (CPC). Hemicellulose is a long chain polymer consisting of both five and six carbon sugar units and ubiquitous in biomass. The hemicellulose polymer chain can be depolymerized via hydrolysis into shorter chain polymers and monomers using either hot water or dilute sulfuric acid and the short chain polymers can be separated for conversion into variety bio-renewable products such as xylitol, furfural.  In this research, the aim is to purify the hemicellulose-based compounds are from the hydrolysis mixture using CPC. The first step in hemicellulose recovery is to select a solvent system based on how the compounds partition between the two liquid phases in the CPC.  Currently, solvent systems are determined by a series of trial and error experiments, which requires a significant amount of time and resources.  This research uses a computer simulation to calculate the partition coefficient of a compound based on its structure and the phase equilibria in the solvent system. This theoretical calculation will promote the development of a more efficient, economically feasible method to determine the best solvent system for CPC separation.  Experimental methods to determine the partition coefficient of five birchwood xylan hemicellulose polymers were compared to the theoretical results from the simulation and the best solvent was selected.  This solvent system was tested in the CPC and the results indicated that the simulation was a reliable means to determine the partition coefficient for CPC separations.