Understanding the role of the structure and chemical composition on the pyrolysis of xylan-based hemicelluloses

Introduction: A number of experimental and modeling works are continuously ongoing on the pyrolysis of biomass as well as of their single organic components (cellulose, hemicellulose and lignin) for assessing at which extent the variability of the biomass composition, in terms of both organic and inorganic matrices, affects the pyrolysis characteristic temperatures and products yields. Despite its abundance in plants (about 20-30 wt.%), hemicellulose pyrolysis is scarcely studied because of its chemical heterogeneity, complexity and less defined structure. Moreover, the hemicellulose isolation techniques can induce polymer modifications and cause salts inclusions, thus making even harder the understanding of the pyrolytic behaviour of this complex polymer. Aim: A research activity devoted to the characterization of hemicellulose pyrolytic behaviour under slow pyrolysis conditions has been carried out since a few years, taking into account also the effect of AAEMs presence. This work aims at elucidating the mechanisms involved in the pyrolysis of xylan-based hemicellulose, emphasizing the role of the inorganics and of the composition and the structure of the polymeric chain. Methods: Two commercially available xylans isolated from beechwood (BW X) and corncob (CC X) and one hemicellulose extracted from grape pruning residues (GP X) have been selected. The hemicellulose was extracted through alkaline treatment after a proper purification and its structure was fully characterized. The thermal behaviour of the BW X, CC X and GP X samples has been evaluated by thermogravimetric (TG) analysis and slow pyrolysis tests have been performed up to 700 °C. Weight loss profiles and gas release rates as a function of the temperature have been compared and the differences between products yields, gas composition and the yields of the main liquid species have been discussed. Results: The results of TG analyses indicate that the GP X follows a different reaction pathway characterized by a lower weight loss rate with respect to BW X and CC X during the main devolatilization stage. Consistently, higher yields of char have been obtained from the GP X pyrolysis. At the same time, high amounts of gaseous species (CO, CO2 and H2) have been produced and a lower yield of the liquid product has been obtained. Conclusion: Since the GP X sample has the highest ash content it is likely that the observed differences were due to the catalytic effect of alkali metals, well known in cellulose pyrolysis, on the ring scission reactions with the consequent formation of light gases. However, the composition and the structure of the starting material may contribute to the differences in char yields. The comparison between the raw and demineralized samples is now ongoing to highlight the relative role of ashes and of the polymer composition and structure.