Sustainable Chemistry for Biomass Utilization

Vanadium catalyzed deoxydehydration (DODH) of vicinal diols to alkenes have been investigated. The aim of the project was to examine the possibility of using an alcohol as both the solvent and the reductant. Development of a process that functions in a polar solvent was thought to be especially important, as the final goal was to use the method to convert oxygen-rich molecules from biomass into valuable compounds.

Initial experiments found the DODH reaction to proceed in an autoclave at 230 °C, when 1,2-decanediol was allowed to react with 5 mol% ammonium metavanadate (NH4VO3) in isopropanol. After 17 hours, a yield of 1-decene that corresponded to 50 mol% of the starting substrate, could be detected using gas chromatography.

Different alcohols were tested as solvents along with their ability to reduce the vanadium center back to the oxidation state in which it is active. Secondary alcohols gave much better yields than their primary analogues and isopropanol was thus selected for further research. Catalyst optimization was carried out by comparison of the products and yields obtained when a range of different vanadium compounds were used.

Most of the screened catalysts gave the same yields and product distributions, and NH4VO3 was judged to be a good choice based on its availability and ease of handling. Even though low turn-over numbers were found in batch experiments, the catalyst proved to be much more sustainable if reused before it gets deactivated.

Vanadium catalyzed DODH was found to have unique selectivity towards substrates containing exactly one primary and one secondary hydroxyl group. In contrast, internal diols such as 3,4-hexanediol gave no yield of alkene and both stereoisomers of 1,2-cyclohexandiol were almost completely unreactive. However, substrates that are stabilized by conjugation, such as hydrobenzoin, were found to undergo oxidative cleavage to form two aldehydes.

The reactivity of the diols also depends strongly on the orientation of the hydroxyl groups relative to one another. A sharp decline in product yield was thus observed when cis-1,2-cyclohexanediol was added to a reaction of 1,2-hexanediol, which was in contrast to addition of the trans stereoisomer.

A possible reason is that the trans isomer cannot simultaneously coordinate to the metal with both hydroxyl groups and thereby inhibit the reaction. Glycerol proved to be unreactive as well, whereas 1,2-propanediol and 3-isopropoxy-1,2-propanediol did undergo DODH to yield propene and 3-isopropoxy-1-propene, respectively.

The water that is formed as the reaction proceeds was found to inhibit the DODH so strongly that  removing it might be the key to achieve alkene yields above 50 mol%. Molecular sieves, 2,2-dimethoxypropane, triethyl orthoformate and various other methods to remove water were tested and analyzed. The vanadium catalyzed reactive distillation of glycerol has also been developed.

The reaction setup and conditions were optimized to ensure a quick separation of the products from the mixture. Only 1 mol% NH4VO3 was enough to achieve total conversion of 23 g of glycerol, with direct collection of a mixture containing allyl alcohol, acrolein and water. The best results were obtained at 275 °C, with yields of 22 mol% allyl alcohol and roughly 4 mol% acrolein obtained after 5 hours.

A black material remained in the reaction flask and its composition was determined by elemental analysis. This material was later proposed to be a polymer mainly built from acrolein and possibly also allyl alcohol and glycerol monomers. The catalytic performance of MeReO3, (NH4)6Mo7O24•4H2O and NH4VO3 was compared and vanadium found to give more allyl alcohol than rhenium, which is surprising due to the increased DODH reactivity of the latter.