Solvent system selection for xylooligosaccharides separation by centrifugal partition chromatography using conductor-like screening model for real solvents

He Zhang, Purdue University

Abstract

The production of value-added, bio-based industrial commodity chemicals is an important area in science right now and this study provides an initial step in the recovery of bio-based chemicals from hemicellulose. Hemicellulose is a long chain polymer mostly consisting of xylose, which is a five-carbon sugar, and a variety of other compounds that are ubiquitous in plant life. The hemicellulose polymer chain can be depolymerized into smaller components, called xylooligosaccharides (XOS) with different chain lengths of xylose linked by beta-1-4 glycosidic bonds, using either hot water, or dilute sulfuric acid. The xylooligosaccharides can then be purified from one another from one another using a variety of separation techniques, one such being centrifugal partition chromatography (CPC). The individual XOS components can be accumulated from successive separations for analysis to determine their best use as a value-added product. The initial step in isolating xylooligosaccharides is to determine a method to theoretically calculate the partition coefficient, KD, of a CPC solvent system. Currently, solvent systems are determined by a series of trial and error experiments. This research builds on a theoretical method to determine the activity coefficient of a solute in a solvent system. The method used is called the Conductor-like Screening Model for Real Solvents (COSMO-RS) and was used to calculate the partition coefficients of xylose and two major functional xylooligosaccharides – xylobiose and xylotriose, based on the structure of the compounds and the composition of solvents in two phases of the solvent system. The COSMO-RS calculation allows for a faster and more economical approach to hone in on the best solvent system for a CPC separation. The experimental partition coefficient determination requires high purity standards which are not always readily available or affordable. For this work, the two XOS compounds studied, xylobiose and xylotriose, were relatively costly but available for purchase. The COSMO-RS method developed in this work can alleviate the requirement for such pure standards and allows a user to determine the partition coefficient based solely on the structure of the compound. The results of experimental solubility of xylose were in relative agreement with COSMO-RS predicted capacity in different chemical groups except for water and ketones groups; in this case when choosing potential solvents for making ternary biphasic solvent systems, water and ketones will not be considered, which means only non-aqueous solvent systems were tested. The comparison of the experimental results and predicted data for xylose in 12 non-aqueous solvent systems demonstrated the potential of using the COSMO-RS method as a tool for solvent system selection. Furthermore, for xylose and xylobiose, a close match between experimental results and the COSMO-RS estimated data in three heptane/n-butanol/acetonitrile systems were obtained. The results for xylotriose predicted a lower partition coefficient than the experimental findings. Considering the partition coefficients of the three target compounds – xylose, xylobiose, and xylotriose – the upper/lower phase ratio and also the separation factors between three compounds, the solvent system containing heptane/n-butanol/acetonitrile with volume ratio of 9:4:5 was deemed to be the most appropriate solvent system for purifying the xylooligosaccharides in a CPC separation.

Degree

M.S.E.

Advisors

Engelberth, Purdue University.

Subject Area

Agricultural engineering|Chemical engineering|Industrial engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS