(461c) Protic Solvents Catalyzed Synergistic H-Adatom and Proton Transfer during Methoxyphenol Hydrodeoxygenation on Ru Clusters

Authors: 
Shangguan, J., University of Toronto
Chin, C., University of Toronto
Hensley, A., Washington State University
Gradiski, M., University of Toronto
Bray, J., Washington State University
Morris, R., University of Toronto
McEwen, J. S., Washington State University
Protic Solvents Catalyzed Synergistic H-adatom and Proton Transfer during Methoxyphenol Hydrodeoxygenation on Ru Clusters

Junnan Shangguan,1 Alyssa J. R. Hensley,2 Matthew V. Gradiski,3 Jacob E. Bray,2 Robert H. Morris,3 Jean-Sabin McEwen,2 and Ya-Huei (Cathy) Chin1*

1 Department of Chemical Engineering and Applied Chemistry, University of Toronto, Toronto, Ontario M5S 3E5, Canada

2 The Gene & Linda Voiland School of Chemical Engineering and Bioengineering, Washington State University, Pullman, Washington 99164, United States

3 Department of Chemistry, University of Toronto, Toronto, Ontario M5S 3H6, Canada

*cathy.chin@utoronto.ca

Supported transition metal clusters catalyze hydrodeoxygenation of phenolic compounds, converting them to value-added alkanes and alkanols. Hydrogenation and hydrodeoxygenation rates vary sensitively with the solvent identity [1]; for example, hydrodeoxygenation rate of guaiacol in water, a protic solvent, is more 3 orders of magnitude higher than that in ethyl acetate, a non-protic solvent. The molecular fate of phenolic compounds on transition metal surfaces and the different extents of solvation towards different reactive precursors and transition states, resulting the apparent rate promotions, have not been unambiguously established. Through kinetic measurements, H-D isotopic exchange, nuclear magnetic resonance (NMR) spectroscopy studies, and density functional theory (DFT) calculations, we propose an elementary reaction sequence for the hydrodeoxygenation of 2-methoxyphenol, 3-methoxyphenol, and 4-methoxyphenol at water-Ru (14 nm Ru cluster size) interfaces, identify the direct participation of solvent molecules in the formation of the respective, kinetically relevant transition states, and from which we rationalize the apparent solvent effects on the reactivity. We establish the substitution effects by using methoxyphenols and identify solvent dielectric permittivity as a kinetic descriptor of hydrogenation and hydrodeoxygenation reactions.

At 423 K and 10-55 bar of H2 pressure, reactions of methoxyphenol [C6H4(OCH3)(OH)] and H2 on Ru clusters occur via the desired C-OCH3 cleavage route (rC-O), forming products predominantly as cyclohexanol. A portion of the methoxyphenol undergoes the undesired H-addition (rH-A), producing methoxycyclohexanol. Irrespective the substituent (i.e., hydroxyl and methoxy) positions on the benzylic ring, rC-O and rH-A exhibit identical, near zero-order dependences on the phenolic reactant (e.g., ~0.22 for 2-methoxyphenol [2]), indicating that phenolic derived species saturates Ru cluster surfaces. Rates rC-O and rH-A both exhibit ~1 and ~1.5 orders with respect to H2 pressure, respectively, for all methoxyphenol (e.g., 2-methoxyphenol [2]). Since H2 dissociation is quasi-equilibrated, as confirmed from rapid H2-D2 scrabmling, these apparent H2 reaction orders indicate that two and three H-addition steps are required to transform the phenolic derived most abundant surface intermediates (MASI) into the respective transition states for C-OCH3 and H-addition routes. The identical phenolic and H2 dependences among all methoxyphenols suggest similar number of reactive hydrogens required to evolve the transition states along the respective reaction coordinates for C-OCH3 cleavage and H-addition routes.

DFT calculations on a model 2-methoxyphenol on Ru(0001)-H2O interface reveal the mechanistic detail of these competitive routes. On Ru surfaces, 2-methoxyphenol preferentially dehydrogenates to form surface guaioxy species [C6H4(OCH3)(O)*] as the MASI, consistent with near zero order dependence on the reactant substrate. For C-OCH3 cleavage route, C6H4(OCH3)(O)* undergoes two quasi-equilibrated H* additions on its oxygen atom and then on its meta carbon atoms (relevant to the hydroxyl group) to form reactive, partially saturated surface enol [i.e., C6H5(OCH3)(OH)*]. Mass spectrometry and 1H-NMR analyses have shown that 2-methoxyphenol-D2 (1 bar)-D2O reaction on Ru/C at 423 K leads to selective H/D exchange on the meta carbon, thereby confirming the quasi-equilibrated H* addition on these specific carbon atoms. The formed C6H5(OCH3)(OH)* undergoes kinetically relevant C-OCH3 bond rupture via an intramolecular proton transfer, during which the proton from its hydroxyl group shuttles through the adjacent water molecule to the methoxy group, forming a surface keto [i.e., C6H5(OH)*] and a methanol molecule. The elementary activation free energy of this intramolecular proton transfer is less than 65 kJ mol-1 at 423 K, >40 kJ mol-1 smaller than the direct oxidative metal insertion to cleave the C-OCH3 bond. Bader charge analyses reveal that significant charge separation occurs at the transition state complex compared to the initial state during the proton-transfer induced C-O cleavage, which is significantly stabilized in the polar protic environment. As a result, more polar solvents with higher dielectric permittivities (e.g. water) promote the hydrodeoxygenation reaction as compared to solvents with lower dielectric permittivities (e.g., ethyl acetate); the C-OCH3 bond strength dictates the rC-O among different substrates. For the competitive H-addition route, C6H4(OCH3)(O)* (MASI) undergoes two consecutive quasi-equilibrated H-additions on one of the meta carbon and then on the para carbon, leading to the formation of partially saturated surface keto [i.e., C6H6(OCH3)(O)*], prior to the kinetically relevant H* addition on the other meta carbon to form C6H7(OCH3)(O)* (activation free energy of ~90 kJ mol-1 at 423 K). DFT calculations reveal that the subsequent H* additions onto the benzylic ring of C6H7(OCH3)(O)* occurs rapidly, forming methoxycyclohexanol. The progressive H* addition, however, renders insignificant charge transfer (<0.1 Bader charge) due to the homolytic nature of C-H bond formations. As a result, solvent does not significantly affect the rate (and equilibrium) constant of H* addition, in sharp contrast to the rate constant of kinetically relevant C-O bond cleavage route.

In summary, polar protic solvent shuttles the proton, preferentially stabilizes the transition state over the reactant state during the kinetically relevant C-OCH3 bond cleavage step, and thus lead to significant hydrodeoxygenation rate enhancements. In contrast, the homolytic nature of H* additions on the carbon ring leads to the insignificant solvent effect on ring saturation reaction route. Such mechanistic knowledge is generalizable towards all methoxyphenols and potentially towards other model lignin structures with both phenolic hydroxyl group(s) and strong aryl C-O bond(s).

References:

  • Shangguan, Y-H. Chin, “Proton-Electron Transfer during Condensed Phase Reduction of Carbonyls on Transition Metal Clusters,” ACS Catal., 9 (2019) 1763-1788.
  • Shangguan, N. Pfriem, Y-H. Chin, “Mechanistic Details of C-O Bond Activation in and H-addition to Guaiacol at Water-Ru Cluster Interfaces,” J. Catal., 370 (2019) 186-199.