Francisco Zaera

Conspectus: Much industrial catalysis employs noble metals such as Pt, Pd, and Rh to facilitate the activation of key diatomic reactants such as H2 and O2, a requisite for the promotion of hydrogenation and oxidation reactions, respectively. However, these metals are often unselective for the conversion of organic feedstocks, so the addition of another, moderating metal is usually needed. The resulting bimetallics have gained prominence recently with an emphasis on so-called single-atom alloys (SAAs), in which the active component (the guest; Pt, Pd) is highly diluted within a matrix of a milder second metal (the host; Cu, Au, Ag). Empirically, SAAs have been shown to improve the selectivity of many catalytic reactions, but a full explanation of how this occurs is still lacking. A simple model has been advanced, based on “modern surface-science” studies using model surfaces under ultrahigh vacuum (UHV), where the guest atoms are isolated on the surface of the host and are presumed to not be significantly affected by it. They are proposed to promote the activation of the diatomic molecule (H2, O2), after which the resulting atoms (H, O) spill over onto the host surface where the rest of the catalytic conversion takes place. Unfortunately, this mechanism fails to explain certain observations from studies of SAAs under catalytic conditions, where the metals exist as nanoparticles (NPs) dispersed on high-surface-area supports, typically porous oxides, and where the surfaces are exposed to atmospheric pressures of the reactants. Our studies have focused on bridging these materials and pressure gaps, using the selective hydrogenation of unsaturated aldehydes with Pt–Cu alloys as a prototypical system. We have determined that, in the very dilute limit, the Pt atoms are located not on the exposed NP surface but rather at the interface between the metal NP and the oxide support, where they are inaccessible to the reactants. Furthermore, their effect on catalytic performance was found to be indirect, via the remote modification of the electronic properties of the surface Cu atoms where the reactions take place. Specifically, the activation barrier for molecular hydrogen dissociation at the Cu sites is decreased, thereby facilitating the production of the atomic hydrogen required for the hydrogenation of the organic adsorbates. Moreover, the added Pt modifies both the adsorption mode and the energetics of the reactants and products on Cu sites and the rate constants of the hydrogenation steps; like pure Cu once it is provided with atomic hydrogen, the Cu–Pt SAAs enhance the production of the desired products (unsaturated alcohols from unsaturated aldehydes), likely by favoring a unique reactant adsorption geometry through the oxygen atom. Our findings contradict the proposed independence of the two metals that constitutes the essence of the current working model for SAA and provide a new explanation of how SAAs function that may lead to better formulations of new bimetallic catalysts.

This is the final report for this project. The long-term objective of our project is to develop a general molecular-level picture of the surface chemistry associated with ALD processes. Our central hypothesis is that the chemistry of ALD precursors can differ significantly from that seen in solution. Awareness of such differences should provide general guidelines on what to watch for when synthesizing new ALD precursors and designing and optimizing new ALD processes, especially in terms of minimizing the incorporation of impurities in the growing films and of carrying out depositions under mild pressure and temperature conditions. The main objective of our research project has been to advance the fundamental knowledge of the surface chemistry of ALD precursors needed for the design and optimization of film deposition processes. A modern surface-science approach has been implemented to both elucidate the mechanism of the reactions of the precursors on the surface and characterize the composition and morphology of the growing films. In general, emphasis is being placed on: • Identifying the primary reactions that may lead to the deposition of clean films; •Identifying the secondary reactions that may help the ALD process, by, for instance, helping with the reduction (or oxidation) of the metal atom; • Identifying the secondary reactions that may lead to the deposition of undesirable impurities in the growing films; • Determining the kinetic parameters of the relevant surface reactions in order to define the optimum conditions for film deposition and to minimize impurity deposition. • Characterizing the nature of the resulting films, with focus on their stoichiometry and on the final oxidation states of the constituent elements; and • Using the information obtained to propose better precursors for given ALD processes.

The strong cross-linked network structure of fiber-reinforced epoxy composites gives them excellent mechanical properties but also makes them difficult to recycle after disposal. Expensive catalysts or undesirable oxidants are often required to degrade such stable material. In this work, we report a DMSO-dissolution method that can recycle amine-cured epoxy resins as well as commercial composites using only dimethyl sulfoxide in air. By selectively breaking C–N bonds via an oxidation-Cope elimination mechanism, fiber recovery and resin matrix upcycling can be achieved. A high products mass recovery (> 98%) can be attained from the degradation of resin matrices, which can be used as wood adhesives with adhesive strength comparable to commercial glues. Our results can be easily extended to degradation of other polymers with tertiary amine structures.

The performance of catalysts made out of Pt supported on TiO₂ thin films grown on SBA-15 (a silica mesoporous material) by atomic layer deposition (ALD) was characterized systematically by combining in situ infrared absorption spectroscopy (IR) with other techniques including electron microscopy and adsorption–desorption isothermal measurements. The titania films in the resulting high-surface-area catalysts were evenly distributed throughout the inner surface of the SBA-15 mesopores, and their thickness could be controlled at a submonolayer level, with 3 to 4 TiO₂ ALD cycles needed for the complete coverage of the silica sites. The titania films could be deposited either before or after adding the metal (Pt), which was dispersed in the form of small nanoparticles (NPs) approximately 4–6 nm in diameter, in order to exert some control on the density and nature of the Pt/TiO₂ interface sites. One important lesson deriving from this work is that such an order of deposition leads to significantly different catalysts in spite of the fact that most of their structural properties are similar. If the Pt is deposited on the titania films, the resulting metal NPs are slightly smaller than those grown on silica and display CO adsorption sites with lower surface Pt coordination numbers. On the other hand, when TiO₂ is deposited on the Pt/SBA-15 starting material, some titania grows on the metal and partially blocks its surface while also creating new interface sites where CO binds more weakly and displays lower C–O stretching frequencies. In terms of catalytic performance, the results from in situ IR CO site titration and kinetic measurements combined suggest a mechanism where CO first adsorbs on Pt atop sites and then migrates to Pt/TiO₂ interface sites, where oxidation takes place. Both types of sites appear to be similar in all the catalysts tested, but catalytic performance could be optimized by tuning their surface densities. Maximum catalytic activity was obtained when the TiO₂ films were deposited first and with TiO₂ coverages of at least half a monolayer, that is, after at least 2 ALD cycles.

The effectiveness of silylation treatments of silicon oxide surfaces using N-(trimethylsilyl)­dimethylamine (TMSDMA) for the passivation of atomic layer depositions (ALDs) was evaluated by using X-ray photoelectron spectroscopy (XPS). It was determined that such silylation does indeed block the silanol (Si–OH) nucleation centers where the ALD precursors react and therefore inhibits film growth, but only temporarily; after a few ALD cycles, deposition becomes evident. By testing this chemistry on two types of SiO₂ surfaces, prepared by plasma-enhanced chemical vapor deposition (PE-CVD) and by chemical (RCA) treatment of Si(100) wafers, it was concluded that the nature of the initial substrate does not play a crucial role in the silylation or ALD blocking processes. The material being deposited, on the other hand, does make a difference: TiO₂ film growth can be blocked for almost 10 ALD cycles, whereas HfO₂ starts building up on the surface after less than 5 ALD cycles. Moreover, the steady-state deposition rate reached for TiO₂ on the silylated surfaces is lower than that seen for the untreated substrate, whereas with HfO₂ not only is that not the case, but the new films may in fact grow as 3D nanostructures. One of the key findings of this work is that the silylation can be carried out using either gas- or liquid-phase treatments. It was found that the extent of silylation and the inhibition of the subsequent ALD were comparable in both cases, but the gas-phase method was determined to be cleaner and to deposit less carbon contaminants. This silylation-based ALD inhibition is expected to be selective with respect to the nature of the substrate and therefore useful for the design of area-selective ALD (AS-ALD) processes.

The use of photocatalysts to promote the production of molecular hydrogen from water, following the so-called water splitting reaction, continues to be a promising route for the green production of fuels. The molecular basis of this photocatalysis is the photoexcitation of electrons from the valence band of semiconductors to their conduction band, from which they can be transferred to chemical reactants, protons in the case of water, to promote a reduction reaction. The mechanism by which such a process takes place has been studied extensively using titanium oxide, a simple material that fulfills most requirements for water splitting. However, photocatalysis with TiO₂ tends to be highly inefficient; a cocatalyst, commonly a late transition metal (Au, Pt) in nanoparticle form, needs to be added to facilitate the production of H₂. The metal is widely believed to help with the scavenging of the excited electrons from the conduction band of the semiconductor in order to prevent their recombination with the accompanying hole formed in the valence band, a step that cancels the initial photon absorption and competes with the photolytic chemical reduction. Here we review and analyze the molecular basis for that mechanism and argue for an alternative explanation, that the role of the metal is to help with the recombination of the atomic hydrogen atoms produced by proton reduction on the semiconductor surface instead. First, we summarize what is known about the electronic structure of these photocatalysts and how the electronic levels need to line up for the reduction of protons in water to be feasible. Next, we review the current understanding of the dynamics of the steps associated with the absorption of photons, the de-excitation via electron-hole pair recombination and fluorescence decay, and the electronic transitions that lead to proton reduction, and contrast those with the rates of the chemical steps required to produce molecular hydrogen. The following section addresses the changes introduced by the addition of the metal cocatalyst, comparatively evaluating its role as either an electron scavenger or a promoter of the recombination of hydrogen atoms. A discussion of the viable chemical mechanisms for the latter pathway is included. Finally, we briefly mention other associated aspects of this photocatalysis, including the possible promotion of H₂ production with visible light via resonant excitation of the surface plasmon of Au nanoparticles, the use of single-metal (Au, Pt) atom catalysts and of yolk-shell nanostructures, and the reduction of organic molecules. We end with a brief personal perspective on the possible generality of the concepts introduced in this Critical Review.