Recovery of monosaccharides from lignocellulosic hydrolysates by ion exclusion chromatography.

The production of sugars from lignocellulosic biomass is the key to a sustainable, renewable chemical industry. Glucose, xylose and other monosaccharides can be easily produced by hydrolyzing cellulose and hemicellulose, the primary polysaccharides in biomass. However, the hydrolysis of biomass generates byproducts that, together with the mineral acid normally added in the hydrolysis step, have to be removed before the downstream conversion processes. In this work, the recovery of monosaccharides from lignocellulosic hydrolysates by means of Ion Exclusion Chromatography (IEC) has been studied. The analyzed process relies on new pretreatment and hydrolysis steps, involving the neutralization of the hydrolysate with sodium hydroxide. The adsorption behavior of the main components involved in the separation has been experimentally investigated. Pulse tests at the high loading encountered in preparative conditions have been performed for a selected group of model components found in the hydrolysates. For all the electrolytes, the retention volume fraction was always between the interparticle porosity and the total column porosity, confirming that ion exclusion was the dominant retention mechanism. On the other hand, sugars eluted before the total column porosity, indicating partial steric exclusion from the resin pores. This observation was then confirmed by size-exclusion experiments with polyethylene glycol standards, from which the distribution coefficient of the studied sugars has been determined. The comparison between the elution profiles of the same sugars in pure form and as a mixture present in the hydrolysate showed differences in both peak shape and retention times. Therefore, an investigation of the influence of the main electrolytes contained in the hydrolysates on sugars adsorption has been performed through the pulse on a plateau method. The electrolytes were found to enhance the sugars retention by promoting their adsorption onto the resin. However, this effect was not sufficient to explain the observed differences, which were effectively explained in terms of viscous fingering, due to the high viscosity differences between the eluent and the sample. A previously developed model for IEC has been updated to take into account all the observed phenomena and applied to simulate the experimental results. The proposed model was in good agreement with the batch-column elution profiles both for the pure components and for the actual hydrolysate, allowing a quantitative description of the separation.