Hybrid Nanomaterials for Environmental and Chemical Catalysis

Hybrid Nanomaterials for Environmental and Chemical Catalysis

The CO₂ Challenge

Like other small molecule feedstocks (e.g. N₂), CO₂ is thermodynamically quite stable (DH^o_f = -394 kJ/mol), however it is susceptible to acid-base thermal (heterogeneous and homogeneous), electrochemical, and photoredox activation schemes in the presence of functional catalysts. Within these schemes, the general properties associated with CO₂ activation are distortion of the linear structure, redistribution of charge to create relative acidic (carbon) and basic (oxygen) sub-sites, reduction of the molecule by electron transfer into the lower-lying LUMO of the distorted structure, and proton transfer to the compromised molecule. On surfaces, the adjacent atoms of a close-packed structure can serve to dissociate CO₂ to generate reduced CH₄, or partially reduced products such as CO and formate, as well as surface carbonate byproducts. Advanced catalysts will need to use these and other building blocks of this type in atom economical ways to assemble higher-order, value-added chemicals.

In stark contrast to kinetically challenging, 8-electron reduction to methane strategies for CO₂ repurposing, nature has taken a unique approach to managing the critical greenhouse gas CO₂. The Lewis acid (Mg²⁺) enzyme Rubisco catalyzes the carboxylation of ribulose-1,5-bisphosphonate as the primary approach to recycling CO₂. Rather than a catalytic cycle based on metal-based reduction of the substrate at Mg²⁺, this transformation is carried by acid-base incorporation of CO₂ at the non-redox active metal and internal electron push to CO₂ without the need to manage multiple redox equivalents externally and high energy, one-electron steps. Thus, our vision is to use insertion (Lewis acid-base) strategies at heterogeneous catalytic active sites guided by computational AI to repurpose CO₂ into value-added chemicals that when uncatalyzed, require high energetic costs to construct. Moreover, we will develop photothermal activation methods of non-equilibrium catalyst heating to learn to drive transformations at reduced external temperatures.

We have also adhered gold nanorods to glass coverslips through electrostatic interactions by treating the glass coverslips with (3-aminopropyl)trimethoxysilane and polystyrene sulfonate. By exploring a condition matrix approach, we were able to select specific nanorod (NR) catalyst loadings that give optimal single monolayer coverage. These Au NR-functionalized coverslips exhibit sustainable, catalytic reduction of 4-nitroaniline by sodium borohydride. Moreover, Au NR coverslips were also found to have pseudo-first-order photochemical rate constants two orders of magnitude faster than the only other affixed catalyst, Au NP in polymer. Au NR coverslips are also recyclable at ~65% of the original rate after 5 cycles and can be carried from solution to solution with no catalyst mobility or loss of function.

Figure 1. TEM images of AgPd nanodendrite-
modified Au nanoprisms with different amounts
of Ag coating (A,F,K) and H₂PdCl₄ (A-E, F-J, K-O).

Of higher profile than the other two reactions, we have also demonstrated that our synthetic routes to multifunctional, hybrid nanoparticles in the form of Au/Ag/Pd prisms, octahedra, and hexagonal plates produce excellent catalysts for electocatalytic CO₂ reduction. By controlling the under-potential deposition of Ag/Pd, novel nanostructures with photoactive Au cores and reactive dendritic tips can be controllably produced. The prisms exhibit very low potential for CO₂ reduction at the nanoscale and form high yields of CO (~87%), HCOOH (~50% in aq. soln.), as well as C-C coupling products (Figure 1). These hybrid materials combine multiple functionalities into single particle catalysts and are clearly the direction of the future, especially for tandem reactions and in-line processes.

Multi-Metallic Satellite Nanocatalysts for C(sp)-CO₂ Transformations

Our experience in generating satellite-based, multi-metallic nanoarchitectures such as Au@AgPd and Fe₃O₄-M (M = Au, Pd, Ag, and PtAg) for electro and photo-chemical reduction of CO₂ and nitroaromatics has taught us that these catalytically active sites are potent for small molecule reduction/insertion and alkyne transformations. Literature precedence also demonstrates the adsorption properties of small molecules (e.g. CO₂, NH₃) and activation of alkynes on Pd, Cu, Ag (111) and Ni(110) metal facets, and Fe³⁺/Fe²⁺ active sites. This allows for design of nanocatalysts for small molecule alkyne insertions to generate diverse products despite the paucity of published examples.

Figure 2. Methodology for hybrid
nanocatalyst development.

We are currently refining our surfactant free, hydrothermal method for 3D growth of metallic silver with Fe₃O₄ octahedra (Fe₃O₄@Ag) in bulk scale (0.5-0.8 g). Microscopy (SEM/EDX) analysis and elemental mapping demonstrate formation of 0.5-1 mm Fe₃O₄@Ag octahedra upon reaction of FeCl₃ and AgNO₃ (2-40 wt%) in the presence of NaOH in aqueous ethylene glycol at 200°C for 12 h (Figure 2). Detailed SEM/EDX characterization of bulk particles as well as microtome and FIB slices illustrate that Ag(0) is present both inside the particle (low [Ag]) and on the octahedral surface facets, corners, and edges (high [Ag]).

Figure 3. Fe₃O₄@Ag catalyzed CO ₂ insertion in
alkynes to afford lactone isomer derivatives (top).
Control table for substrate R = H; Ar = Ph (bottom).

With a sound structural characterization of catalyst architecture in hand, our efforts focused on demonstrating the composition/function relationship for the catalyst toward CO₂ feedstock insertion into alkynes with consequential cyclization to yield value-added, higher-order lactone products. Reaction of functionalized phenylacetylenes (Hammett substitutions) with modestly reactive 3-phenylpropargyl chloride in the presence of Fe₃O₄@Ag with varying Ag (2-40 wt%) amounts at 90°C for 8 h yields major (43-55%) and minor (15-30%) lactone isomers in up to 85% (Figure 3). In the presence of just the Fe₃O₄ Lewis acid component of the catalyst, the reaction generates the diyne coupled product as the major species with a low amounts of the CO₂ insertion isomers. Alternatively, if the reaction is performed using just Ag NPs as the active catalyst, the only products detected are the CO₂ insertion isomers, but with markedly lower yields under the same reaction conditions. These critical control experiments suggest that there is a synergistic or at least cooperative effect between Fe₃O₄ and Ag that contributes to a catalyst system that affords significant reaction yields with good selectivity for the CO₂ inserted products under modest conditions.

Figure 4. Proposed mechanism for lactone
formation at Fe₃O₄@Ag.

These key observations have allowed us to generate a unifying catalytic cycle that accounts for the reactivity observed using the different catalyst components. Since diyne formation is the dominant, while not exclusive product in the Fe₃O₄-only condition, this suggests that Fe-acetylide formation with oxidative addition to form the diyne (Figure 4, Path I) is a more facile pathway than CO₂ insertion (Figure 4, Path II). Additionally, the observation that only lactone isomers are detected in Ag NP-catalyzed trials, albeit in low yields, also illuminates the CO₂ activation capability of Ag as a key mechanistic determinant. Thus, Fe₃O₄ and Ag clearly complement each other as catalyst components for lactone isomer formation via CO₂ insertion. We propose that Fe₃O₄ serves as the Lewis acid to form the key Fe-acetylide bond, which may or may not be buffered by electron density from neighboring metallic Ag to enhance the subsequent CO₂ insertion step from activated CO₂ on Ag. While the presence of metallic Ag alone can carry out both Ag-acetylide and CO₂ insertion step, the overall reaction is clearly sluggish but yet exhibits no diyne formation. Thus, the CO₂ insertion step must be facile with the slow step most likely involving binding of the anionic phenylacetylide to metallic Ag. In further support of the role of each catalyst component, subsequent S_N2 reaction of the catalyst-bound carboxylate with the 3-phenylpropargyl chloride gives the ester, which subsequently cyclizes to the lactone. While catalyst may assist in these last steps, control experiments have shown that reactions proceed efficiently in the absence of catalyst, indicating that the role of the catalyst is paramount only for the initial steps in the transformation. Therefore, this hybrid catalyst offers complementary Lewis acidic and CO₂ activating sites that enhance the rate of the initial steps in the formation of tandem reaction products under moderate conditions.