Solvent effects in liquid-phase reactions: I. Activity and selectivity during citral hydrogenation on Pt/SiO2 and evaluation of mass transfer effects

The effect of the solvent on the liquid-phase hydrogenation of citral on a Pt/SiO2 catalyst was examined by comparing the specific activity and the product selectivity in eight nonreactive solvents—n-amyl acetate, ethyl acetate, n-hexane, cyclohexane, tetrahydrofuran, p-dioxane, ethanol, and cyclohexanol—that have significantly different physical and electronic properties. Appendix A describes a detailed approach to calculating the properties of complex molecules that are required to determine an accurate Weisz–Prater criterion (or a Thiele modulus) to evaluate the presence or absence of pore diffusion limitations. These properties include viscosity, heat of vaporization, specific volume, and gas solubility in pure liquids or mixtures. Their utilization to calculate accurate effective diffusivities in porous catalysts is presented. All rate data utilized here were obtained in the kinetic regime. The absence of Cl in the catalyst prevented side reactions producing acetals. In a regime of kinetic control between 298 and 423 K, the turnover frequency (TOF) varied by a factor of approximately 3 at any temperature, with the highest value always obtained in p-dioxane. The variation in TOF did not correlate with either the solvent dielectric constant or its dipole moment. These catalysts deactivated by 1–2 orders of magnitude over a 24-h period at 298 or 373 K, presumably due to CO adsorption caused by a decarbonylation side reaction, and the total number of turnovers at 373 K was highest in ethanol and p-dioxane (ca. 2000), which gave conversions of 50–60%. At 423 K, the highest conversions after 24 h increased to about 95% (in ethanol or cyclohexanol). When compared at conversions near 30%, the solvent did not markedly influence the product distribution, although p-dioxane gave the lowest selectivity to geraniol and nerol. Lower citral concentrations lowered selectivity to these unsaturated alcohols, whereas H2 pressure did not impart any significant trends. In these eight solvents, a one-half to first-order dependence on citral occurred, whereas the exponential dependence on H2 pressure was relatively invariant around 0.3.