Molecular bonding-based descriptors for surface adsorption and reactivity

• Bonding analysis highlights the surface–adsorbate interactions mediating catalysis.
• Natural Bond Orbital analysis enables a novel class of bonding-based descriptors.
• Molecular bonding-based descriptors quantify adsorbate “pre-activation.”
• “Pre-activation” correlates strongly with adsorption strength and reactivity.
• Bonding-based descriptors provide a “chemical” perspective and mechanistic insight.

Descriptors, quantities that correlate with catalytically interesting quantities, are widely used in computational catalyst screenings. For example, adsorption energies of atomic species (O, C, etc.) are often employed as descriptors of catalytic activity, in conjunction with appropriate scaling and Brønsted–Evans–Polanyi relations. While very successful in terms of their predictive value, the reaction-specific mechanistic insight gained from such relations can be limited since adsorption energy descriptors often correlate indirectly with catalytic activity. To this end, we have developed a novel class of descriptors based on molecular bonding, characterizing the intramolecular bonding of adsorbed reactants in terms of two-center bonding derived on the basis of Natural Bond Orbital analysis. We demonstrate this approach by examining a diverse set of reactions across a variety of transition metal surfaces. We find that these bonding-based descriptors have a direct connection with the surface–adsorbate interactions that mediate catalysis, explicitly treating the bond breakage/formation that accompanies the rearrangement of atomic contacts. In many cases, we can interpret these interactions in terms of adsorbate “pre-activation”, reflecting the intrinsic locality of surface reactions. We anticipate that this molecular bonding-based approach can help elucidate more comprehensive mechanistic insight into surface-catalyzed reactions and facilitate the unification of concepts from homogeneous and heterogeneous catalysis.

Figure optionsDownload high-quality image (80 K)Download as PowerPoint slide