(217d) Sequential Quadratic Programming Framework for Ab Initio Structure Property Relationships in Amorphous Catalysts and Supports

Methods for modeling isolated sites on amorphous catalyst supports lag far behind methods for modeling catalytic sites on metal surfaces, zeolites, and other crystalline materials.  Typical strategies for modeling isolated sites on amorphous supports use cluster models with arbitrarily fixed atoms to model the rigid amorphous support.  The arbitrarily fixed atoms impart arbitrary properties to the site.  Alternatively no constraints are used, which results in sites with unrealistic flexibility.  To avoid arbitrary constraints while still retaining influences from the extended solid structure, we introduce a quenched disorder embedding scheme based on the reversible work theorem.  A sequential quadratic programming framework then helps us test mechanistic hypotheses and relate chemical properties, such as the activation energy, to active site structure. The algorithm is illustrated on an Empirical Valence Bond model energy landscape and its limitations are discussed. We then use the algorithm to model an off-pathway kinetic trap in olefin metathesis by isolated Mo sites on amorphous SiO₂.