Catalysis of Organic..

198 Fructose HydrogenationHOHCCH 2OHHCHCCHOHOHOHCH 2OHSorbitolH 2catalystH 2catalystOHOOHOHHOHOβ-pyranose53%HOOHOβ-furanose32%OHOHOHHOCH 2OHCHCHCCHOOHOHCH 2OHD-Fructose3%HOOHO OHOHHOHOα-pyranose2%OHOOHα-furanose10%OHOHH 2catalystHOH 2catalystHOCH 2OHHCHCCHCHOHOHCH 2OHMannitolFigure 1. Fructose mutarotation (at 80°C) and hydrogenation (assuming retention).fructose (4) meaning that a less abundant but more active species can influence thereaction’s selectivity more than its equilibrium concentration would predict. Theabove mentioned 1:1 mannitol:sorbitol ratio found with Ni confirms this finding andsuggests that not all of the tautomeric forms are equally active. This and otherfindings have led to the proposal of various fructose hydrogenation mechanisms.Ruddlesden et al. (7) have shown that fructose hydrogenation on Ni and Cu does notoccur by either the open ketose or the 1,2-enediol forms. They proposed that itproceeds via the hydrogenolysis of the bond between the ring oxygen and theanomeric carbon. They also showed that a hydroxyl group must be on the anomericcarbon for this reaction to take place indicating that this is not a typicalhydrogenolysis reaction. Although they admittedly could not identify the adsorbedactive species, they did propose that ketal hydrogenolysis occurs with retention at theanomeric carbon. Makkee (6) and colleagues (8) studied the hydrogenation offructose over Cu, Ni and other metals. They found that the reaction is first order tohydrogen and initially zero order to fructose with a shift to first order atconcentrations lower than 0.3 M implying that the rate determining step at higherconcentrations is the attack of hydrogen on the adsorbed fructose. They also statedthat the pyranose rings are preferentially adsorbed over the furanose rings and thatthe furanose species were far more active, meaning that most of the adsorbedfructose reacts slowly. A similar trend was found for Ru/C by Heinen et al. (4),where they claimed that pyranose and furanose rings adsorb with equal strengths andthat only furanose is active. While the products did not inhibit fructose adsorptionon Ni or Cu (6), they were found to do so on Ru (4). Makkee et al. (6) went on tosay that adsorbed fructose forms an ionized ketal species where the anomeric carbonto ring oxygen bond becomes less than single and the hydroxyl oxygen to anomericcarbon bond becomes more than single. This ionized adsorbed species is thenthought to be attacked by a hydride-like species in a S N 2 fashion to give the polyolwith inversion at the anomeric carbon. The term hydride-like species is not clear,because hydrogen dissociates on these metals to form chemisorbed hydrogen atomsthat are active for hydrogenation (9). While these metals may form metallichydrides (10), they will not form active surface hydrides (11) under the conditionsused here (10,12). Moreover, these metals form the metallic type of hydrides and