Abstract
Nickel phosphide catalysts (Ni12P5 and Ni2P) preferentially cleave sterically hindered 3C–O bonds over unhindered 2C–O bonds, and Ni2P is up to 50 times more selective toward 3C–O bond cleavage than Ni. Here, we combine kinetic measurements, in situ infrared spectroscopy, and density functional theory (DFT) calculations to describe the mechanism for C–O bond rupture over Ni, Ni12P5, and Ni2P catalysts. Steady-state rate measurements and DFT calculations of C–O bond rupture within 2-methyltetrahydrofuran (MTHF) show that quasi-equilibrated MTHF adsorption and dehydrogenation steps precede kinetically relevant C–O bond rupture at these conditions (1–50 kPa MTHF; 0.1–6 MPa H2; 543 K). Rates for 3C–O and 2C–O bond rupture are inhibited by H2, and the ratio of these rates increases with [H2]1/2, suggesting that the composition of the reactive intermediates for 3C–O and 2C–O rupture differs by one H atom. Site-blocking CO*, NH3*, and H* inhibit rates without altering the ratio of 3C–O to 2C–O bond rupture, indicating that these C–O bond rupture precursors and transition states bind to identical active sites. DFT-based predictions suggest that these sites are exposed ensembles of 3 Ni atoms on Ni(111) and Ni2P(001) and 4 Ni atoms on Ni12P5(001) and that the incorporation of P disrupts extended Ni ensembles and alters the reactivity of the Ni. Increasing the phosphorus to nickel ratio (P:Ni) decreases measured and DFT-predicted activation enthalpies (ΔH ‡, 473–583 K) for 3C–O bond rupture relative to that of 2C–O bond rupture. Selectivity differences between specific C–O bonds within MTHF reflect differences in the H content of reactive intermediates, activation enthalpy barriers, and P:Ni of Ni, Ni12P5, and Ni2P nanoparticles.