Although hydrogen spillover is often invoked to explain anomalies in catalysis, spillover remains a poorly understood phenomenon. Hydrogen spillover (H*) is best described as highly mobile H atom equivalents that arise when H₂ migrates from a metal nanoparticle to an oxide or carbon support. In the 60 years since its discovery, few methods have become available to quantify or characterize H*-support interactions. We recently showed in situ infrared spectroscopy and volumetric chemisorption can quantify reversible H₂ adsorption on Au/TiO₂ catalysts, where adsorbed hydrogen exists as H* and interacts with titania surface hydroxyl (TiOH) groups. Here, we report parallel thermogravimetric analysis and Fourier transform infrared spectroscopy methods for systematically manipulating the surface TiOH density. We examine the role of surface hydroxylation on spillover thermodynamics using van't Hoff studies to determine apparent adsorption enthalpies and entropies at constant H* coverage, which is necessary to maintain constant H* translational entropy. Although surface TiOH groups are the likely adsorption sites, the data show removing hydroxyl groups increases spillover. This surprising finding, that adsorption increases as the adsorption site density decreases, is associated with improved thermodynamics on dehydroxylated surfaces. A strong adsorption enthalpy–entropy correlation implicates the changing surface entropy of the titania support itself (i.e., an initial state effect) is deeply intertwined with the H* configurational entropy. These effects are surprising and should apply to all low-coverage adsorbates where entropy terms dominate more traditional enthalpic considerations. Moreover, this study points toward a kinetic test for invoking spillover in a reaction mechanism: namely, in situ dehydroxylation should enhance spillover processes.

Ian Loscher

Appearances