The interaction between catalytic nanoparticles (NPs) and their supports, which are often amorphous oxides, has not been well characterized at the atomic level, although it is known that, in some cases, NP−support interactions dominate the catalytic activity of the system. Furthermore, there is a lack of understanding of how support preparation affects both the stability of the NP (resistance to sintering) and the catalytic activity. We present first-principles density functional theory (DFT) calculations on amorphous silica supported Pt NPs of various sizes. Our calculations predict that support preparation methods that lead to higher hydroxyl density when NPs are deposited on the support will lead to higher resistance to sintering. We find that the total charge on supported NPs, which can affect catalyst activity, depends linearly on the number of Pt−silica bonds formed during NP deposition. The number of bonds between an NP of a known geometry and the silica support with a known hydroxyl density can be estimated from very fast discrete element method simulations, enabling the prediction of both the net charge and the adhesion energy of the particle from a linear fit correlation derived from DFT calculations of a series of differently sized Pt clusters. This work quantifies interactions between Pt NPs and amorphous silica supports and demonstrates a new method for rapid estimation of NP−support interactions on amorphous supports.
■ INTRODUCTIONRecent developments in both computational methods 1−11 and experimental characterization techniques 12−14 have facilitated detailed characterization of catalytic materials at the atomic level. Insights relating structural and electronic properties to catalytic reactivity, selectivity, and stability have provided some general rules for tailoring catalyst properties via what is now ubiquitously termed "rational design." 7−11,14,15 Many computational results have been shown to agree quite well with experimental observations, in particular, correlations involving simple properties (e.g., coordination number and adsorbate binding energies) and simple systems (e.g., extended metal surfaces and diatomic reactants). Catalytic materials research, however, has recently shifted toward ever-smaller nanoparticles (NPs) and increasingly complex support materials. Whereas approaches to understanding simple catalytic materials have been well-established and researched in recent years, 7−11,14−16 atomistic studies of supported NP systems have been limited to idealized, perfectly crystalline supports. 17−21 Numerous studies have exploited geometry-constrained calculations to isolate the effects of coordination number and NP size. 7,8,11 This approach often provides clear trends and powerful insight regarding catalyst design. However, distortion of the support structure has been shown to significantly alter the structure and electronic properties of supported NPs. 17,22 Amorphous supports exhibit a diverse range of local surface structures, which result in a distribution of catalyst−support...