Kinetic Model of Xylose Dehydration for a Wide Range of Sulfuric Acid Concentrations

Abstract

This paper presents a robust kinetic model for the dehydration of xylose in concentrated sulfuric acid (i.e., 0.1– 2 M) at 120–160 °C, i.e., conditions that were not yet explored in the literature and hold promise in terms of process intensification. The model is built on an extensive set of batch experiments and an integral analysis method of the kinetic data. Direct condensation of furfural and xylose is not a major degradation route, but the former reacts with other dehydrated intermediates. The kinetic constants of xylose dehydration and furfural degradation show a non-linear dependency with respect to the hydronium concentration at higher acid concentrations (>1 M). This trend could be attributed to a simultaneous attack of two hydronium ions on the formyl group of the C1 atom and on the hydroxyl group of the C3 atom occurring under high acid concentrations (>1 M). Unlike previously reported models, the developed kinetic model is able to predict the experimental results (xylose conversion and furfural yield) within a 95% confidence interval under a wide range of temperatures and sulfuric acid concentrations. Even more, this model is also able to predict the experimental results reported in the literature obtained with sulfuric acid with high accuracy.

Efficient synthesis of furfural from xylose over HCl catalyst in slug flow microreactors promoted by NaCl addition

“…Table S3), of which among others the three most or O/A ratio in biphasic systems, would lead to different overall xylose conversion rates. 31,32 This, however, was not observed under the prevailing conditions in this work (Figures S8 and S9), and as such this mechanism can be ruled out. In addition, the reaction between dehydration intermediates and furfural (Scheme 1C) was reported to play only a minor role compared with the direct xylose condensation (Scheme 1A), 33 and since the intermediate is not quantifiable, the relevant kinetic parameters were usually estimated based on an overall fitting.…”

“…In addition, the reaction between dehydration intermediates and furfural (Scheme 1C) was reported to play only a minor role compared with the direct xylose condensation (Scheme 1A), 33 and since the intermediate is not quantifiable, the relevant kinetic parameters were usually estimated based on an overall fitting. [32][33][34] Generally speaking, all these three reaction networks or Moreover, the fluid is under laminar flow with a parabolic velocity profile, which however can be compensated by the fast radial diffusion (characterized by the Fourier numbers of xylose and furfural being [much] larger than 1) to afford an approximate plug flow behavior. 35 Thus, here a simplified plug flow model was adopted for kinetic modeling.…”

“…where α mono,aq (ranging from 1.06 to 1.14 within 120-180 C; cf. Figure S2a) is the ratio of the densities of the aqueous phase at 24,32 is adopted in the current work…”

“…Given the scarcity of the existing models incorporating the effect of NaCl on the xylose conversion, we compared only the representative experimental results without NaCl addition in this work to the predictions from some typical literature models (using HCl, H 2 SO 4 , or formic acid as the catalyst). 24,[31][32][33] A detailed discussion is provided in Section S10 of the Supporting Information. Generally, for reactions in the monophasic water system, models of Marcotullio et al, 24 Weingarten et al, 31 and Krzelj et al 32 with mineral acid catalysts (HCl or H 2 SO 4 ) largely underestimate the reaction rate (i.e., both the xylose consumption and furfural formation rates) in microreactors, though they can capture more or less satisfactorily the measured furfural yield/selectivity as a function of the xylose conversion (Figure S19a-c).…”

“…24,[31][32][33] A detailed discussion is provided in Section S10 of the Supporting Information. Generally, for reactions in the monophasic water system, models of Marcotullio et al, 24 Weingarten et al, 31 and Krzelj et al 32 with mineral acid catalysts (HCl or H 2 SO 4 ) largely underestimate the reaction rate (i.e., both the xylose consumption and furfural formation rates) in microreactors, though they can capture more or less satisfactorily the measured furfural yield/selectivity as a function of the xylose conversion (Figure S19a-c). This implies that additional effects and unpractical furfural selectivity close to 100% was predicted by models of Weingarten et al 31 and Krzelj et al 29 that do not include the direct xylose condensation in the reaction network (Figure S20c).…”