On the primary devolatilization of hemicellulose

Abstract

This paper presents validation cases on a reaction mechanism for the primary devolatilization of hemicellulose called “hemi-FC.” The analysis incorporates depolymerization via random scission; charring via monomer decomposition, semichar decomposition, and thermal annealing; fragmentation statistics; and the flash distillation analogy. It predicts the yields and elemental compositions of bio-oil and char, oil molecular weight distributions, and the yields of CO, CO2, H2O, H2, and light oxygenates. A single set of model parameters accurately interprets the partitioning of hemicellulose with cation levels from 5000 to 30,000 ppmw into oil, noncondensables, and solid residue for temperatures to 900 °C, heating rates from 10 °C/min to 6000 °C/s, contact times after heatup to 1800 s, and pressures of vacuum and 0.1 MPa. The model results are generally within useful quantitative tolerances of the measured values across the domain of test conditions. Nonetheless, the commercial xylans that have been tested devolatilize at much cooler temperatures than cellulose and produce more gas and char and less bio-oil primarily because their decomposition is strongly catalyzed by excessive residual amounts of Na and Ca from the hemicellulose extraction process. These cation loadings are definitely heavy enough to grossly distort the dynamics of the spontaneous reaction channels and ultimate devolatilization behavior. Consequently, the spontaneous, uncatalyzed reaction channels for hemicellulose devolatilization remain uncharacterized even while they are bound to be prominent in the devolatilization of many natural hemicellulose forms.