We perform the first extensive experimental kinetic studies of fructose dehydration and 5-hydroxymethyl furfural (HMF) rehydration at low temperatures over a wide range of conditions (T ∼ 70–150 °C; pH values 0.7–1.6 and initial concentrations of fructose (5–20%w/v) and HMF (2.5–10%w/v)). Guided from insights from our first-principles calculations, we perform kinetic isotope effect (KIE) experiments of labeled fructose to validate the rate-limiting step. Subsequently, we develop the first skeleton model for fructose dehydration and HMF rehydration that integrates the fundamental kinetic experiments and accounts for the KIE, as well as the distribution of fructose tautomers, which changes significantly with temperature, and a direct path of fructose conversion to formic acid. It is shown that the skeleton mechanism of two steps consisting of fast protonation and dehydration followed by intramolecular hydride transfer as the rate-limiting step can capture the experimental kinetics and KIE experiments well. Fructose dehydration is found to result in stoichiometric excess of formic acid relative to levulinic acid, produced directly from fructose. All reactions are shown to be pseudo-first order in both catalyst and substrate. These insights are incorporated in a continuous flow reactor model; higher temperatures improve the optimum yield of HMF, while HMF selectivity at low conversions is less sensitive to temperature.

No abstract

5-hydroxymethylfurfural (HMF) can be produced from the acid-catalyzed dehydration of fructose but its yield is limited due to subsequent HMF degradation to side products. A reactive adsorption process is proposed to improve the yield to HMF. Separate experimental singlecomponent isotherms of fructose, HMF, formic acid, and levulinic acid on carbon BP2000 and reaction kinetics of the fructose dehydration to HMF in aqueous solution of HCl are presented to develop empirical isotherms and kinetic rate constants, respectively. These sub models are subsequently integrated in an adsorptive reactor at a range of temperatures (100-180°C) with different loadings of adsorbent. It is shown that the adsorbent improves HMF yield compared to the single solution phase (adsorbent-free case). Low temperatures and high adsorbent loadings improve HMF yield. Under certain conditions both reactive adsorption and the commonly used reactive extraction can result in a similar improvement in HMF yield. HMF recovery from the solid adsorbent has been identified as a major challenge that can be ameliorated through