(426a) Leaching of Rhenium Species from Oxide-Supported Deoxydehydration Catalysts for Biomass Upgrading: Support, Solvent and Reactant Effects

Leaching of active species from solid catalysts is a common phenomenon in liquid phase processing and can be a key challenge to developing stable solid catalysts for conversion of non-volatile reactants. There are multiple examples of dissolution of supported noble metal species [1,2]. The concerns for catalyst activity and stability in solution phase processing are of particular importance in relation to upgrading reactions for biomass-derived feedstock [3], as these molecules often require a solvent because their size and polarity makes solid-gas phase reactions less feasible.

One such biomass upgrading reaction is deoxydehydration (DODH), named such because it combines a dehydration with a deoxygenation utilizing a sacrificial reductant to convert a vicinal diol into an olefin in a single-step catalytic reaction. The transformation was discovered using methyltrioxorhenium as a homogeneous catalyst [4]. This reaction is of great interest in biomass upgrading because it can highly selectively convert polyols, such as sugar-derived alcohols, into olefins, which could be of use as a renewable source of fuels, or of polymer and rubber feedstocks, such as butadiene.

A key challenge in making this reaction industrially feasible is the cost and rarity of the rhenium catalyst. Recently, efforts have been made to increase the industrial viability of this reaction by developing solid catalysts [5-7] and exploring the use of alternative metals to rhenium as the active species. While leaching of active rhenium species from solid catalysts is a persistent issue, reports in the literature digress on the extent of leaching and on the best supports to heterogenize oxo-rhenium species. In this work, we present an investigation into the primary causes of leaching and use this insight to devise strategies to mitigate leaching.

A series of oxide-supported oxo-rhenium DODH catalysts was prepared via the incipient wetness method uding ammonium perrhenate as the precursor, with subsequent calcination at mild temperatures of less than 420 °C. Reactions were conducted in batch, with a typical diol-to-rhenium ratio of 20:1, and a slight excess of the reductant, triphenylphosphine. Solvents included toluene, dioxane, isopropanol and 3-octanol, where the latter two can also serve as reductants. The standard reaction temperature was 150 °C. Rhenium leaching was monitored by hot and cold filtration tests, followed by ICP-MS and reactivity analysis of the supernatants.

All catalysts were highly active and selective for conversion of 1,2-diols such as decane diol or hexane diol to the corresponding alpha-olefins. Leaching was observed for all catalysts and was found to be the main cause for catalyst deactivation. Leaching was much more pronounced for SiO₂ and TiO₂ supports than for ZrO₂, Al₂O₃ and Fe₂O₃ and, over several uses of the catalyst, led to complete loss of all rhenium from the SiO₂ support. Comparison of solid activity with supernatant activity after separation demonstrated that catalysis is predominantly, albeit not entirely, homogeneous. A leaching hypothesis was formulated on the basis of the published mechanism [8]. The reaction has three fundamental steps, a reduction step where the catalyst is reduced, a condensation step in which the diol coordinates with the oxo-metal site to form a glycolate, and an elimination step in which the catalytic site is re-oxidized and the olefin is eliminated. We hypothesized that the chelate complex with the diol is highly soluble in organic media because of the hydrophobic tail of the diol. Indeed, it was found that a longer hydrocarbon tail promoted leaching in nonpolar solvents. However, more polar solvents were better able to solubilize the perrhenate anion, suggesting that for each reactant, a solvent can be found that implies minimal leaching. Moreover, it could be shown that after complete consumption of the diol, some rhenium redeposits, opening avenues for a ‘release and catch’ strategy.

[1] S. Hübner, J. G. de Vries, V. Farina, Adv. Synth. Catal. 358 (2016) 3– 25.

[2] D. B. Eremin, V. P. Ananikov, Coord. Chem. Rev. 346 (2017) 2–19.

[3] I. Sádaba, M. López Granados, A. Riisager, E. Taarning, Green Chem. 17 (2015) 413.

[4] G. K. Cook, M. A. Andrews, J. Am. Chem. Soc. 118 (1996) 9448.

[5] A. L. Denning, H. Dang, Z. Liu, K. M. Nicholas, F. C. Jentoft, ChemCatChem 5 (2013) 3567.

[6] L. Sandbrink, E. Klindtworth, H. -U. Islam, A. M. Beale, R. Palkovits, ACS Catal. 6 (2016) 677.

[7] S. Tazawa, N. Ota, M. Tamura, Y. Nakagawa, K. Okumura, K. Tomishige, ACS Catal. 6 (2016) 6393.

[8] P. Liu, K. M. Nicholas, Organometallics 32 (2013) 1821.