Scott L. Anderson

Sub-nanometer supported clusters are of interest in catalysis due to nearly complete precious metal atom utilization, generally high catalytic activity, and attractive tunability via size and cluster-support interactions. The stability of such clusters remains a challenge, however. We demonstrate anchoring and stabilization of size-selected Ptn, n = 2–10 clusters soft-landed on highly oriented pyrolytic graphite (HOPG), achieved by modifying the HOPG with 100 eV kinetic energy Nn+ (n = 1, 2) and Ar+ ions, with and without O2 oxidation of the ion-modified surfaces, prior to Ptn soft-landing. Ion modification is found to significantly improve the sticking probability for soft-landed Ptn, implying increased adsorption energies, and O2 exposure prior to cluster deposition increases the sticking probability to ∼unity. Density functional theory (DFT) finds that Ptn bind preferentially to under-coordinated carbon sites created by the ion impacts, rather than to implanted N atoms, and X-ray photoelectron spectroscopy and scanning tunneling microscopy both show that the defect-anchored clusters are stable against sintering. There is net electron transfer from the defect-bound clusters to the support, and the local defect structure may undergo significant restructuring due to Pt–C bond formation. To probe the effects of ion-modified HOPG on the chemical properties of the adsorbed Ptn clusters, we tested electrodes prepared by Ptn deposition on ion-modified HOPG for the oxygen reduction reaction (ORR). The ion-modified electrodes showed ORR activity consistent with cluster size being maintained, and O2 modification enhanced both cluster anchoring and ORR activity. Thus, ion modification provides means to stabilize subnano cluster electrodes, while oxygen functionalization of the anchoring sites tunes the electrocatalytic activity.

Sub-nanometer supported clusters are of interest in catalysis due to nearly complete precious metal atom utilization, generally high catalytic activity, and attractive tunability via size and clustersupport interactions. The stability of such clusters remains a challenge, however. We demonstrate anchoring and stabilization of size-selected Ptn, n = 2 - 10 clusters soft-landed on HOPG, achieved 2 by modifying the HOPG with 100 eV kinetic energy N(𝑛 1, 2) and Ar+ ions, with and without O2 oxidation of the ion-modified surfaces, prior to Ptn soft-landing. Ion modification is found to significantly improve the sticking probability for soft-landed Ptn, implying increased adsorption energies, and O2 exposure prior to cluster deposition increases the sticking probability to ~unity. DFT finds that Ptn bind preferentially to under-coordinated carbon sites created by the ion impacts, rather than to implanted N atoms, and X-ray photoelectron spectroscopy and scanning tunneling microscopy both show that the defect-anchored clusters are stable against sintering. There is net electron transfer from the defect-bound clusters to the support, and the local defect structure may undergo significant restructuring due to Pt-C bond formation. To probe the effects of ion-modified HOPG on the chemical properties of the adsorbed Ptn clusters, we tested electrodes prepared by Ptn deposition on ion-modified HOPG for the oxygen reduction reaction (ORR). The ion-modified electrodes showed ORR activity consistent with cluster size being maintained, and O2 modification enhanced both cluster anchoring and ORR activity. Thus, ion modification provides means to stabilize sub-nano cluster electrodes, while oxygen functionalization of the anchoring sites tunes the electrocatalytic activity.

In this work, 3-atom clusters, Ru ₃ and Pt ₃ , were deposited onto radio frequency RF-sputter deposited TiO ₂ , treated with Ar ⁺ ion sputtering. Pt ₃ shows discrete energy levels and are most likley not agglomerated. The Ru clusters have agglomerated.

Electrodes are exposed to air prior to use in most electrochemistry experiments, but the resulting adventitious contaminants are easily removed by cycling the electrode potential to drive reactions or desorption. In most cases, the resulting changes to the electrode surface are probably inconsequential, however, for electrodes with catalytic centers consisting of sub-nanometer clusters or single atoms, potential cycling may drive dissolution, migration, or other processes that significantly alter the catalyst properties. We have developed the capability to prepare clean electrodes in ultrahigh vacuum, decorated with precisely controlled clusters or atoms, and to study their electrochemistry within the vacuum system after controlled exposures to pure gases or to air. Here, we examine the effects air exposure on the activity for HER, ORR, and oxidation of ethanol and propanol, catalyzed by Pt4 clusters deposited on both ITO and nitrogen-implanted HOPG electrodes. Air exposure was found to have no significant effects on HER but did have substantial effects on the other reactions. Some effects were transient, i.e., they decreased over the course of a few dozen CV cycles, but there were also effects such as increased overpotentials and changes in peak currents or potentials, that persisted after extensive potential cycling.

Mass-selected Pt_n⁺ (n ≤ 7) were deposited at variable energies on highly oriented pyrolytic graphite (HOPG), creating highly active hydrogen evolution reaction (HER) electrocatalysts. HER mass activities were ∼2 to >10 times higher than those for the surface atoms in bulk Pt and for Pt_n deposited on several other supports. Thus, high activity reflects the Pt-C structures formed by energetic Pt_n─HOPG impacts, in addition to high Pt surface availability. The Pt_n/HOPG electrodes were probed by X-ray photoelectron spectroscopy, low energy ion scattering, and electron microscopy. Born-Oppenheimer molecular dynamics (BOMD) was used to simulate Pt_n─HOPG impacts, revealing the types of structures formed at different energies, then DFT was used to probe their most important HER pathways. For low deposition energies, the Pt_n deposit onto the HOPG surface with subunit sticking probability, aggregating at defects. With increasing deposition energy, the sticking probability initially decreases, then rises to unity as subplantation and defect creation allow formation of strongly bonded platinum-carbon structures. Barriers for HER on these structures were found to be low and weakly dependent on Pt_n size, consistent with experiment. The activities were highest for small covalently bonded Pt-C structures created at high deposition energies. The larger aggregated structures formed at low energies were less active, but still substantially better than the bulk Pt surface monolayer. The catalysts were stable in repeated potential cycling at reducing potentials, but electrodes containing subplanted Pt became more active when scanned to oxidizing potentials, due to emergence of subplanted Pt onto the surface.

Mass- and energy-selected Ptn+ (n ≤ 7) clusters were deposited at variable energies on highly oriented pyrolytic graphite (HOPG) to create highly active hydrogen evolution reaction (HER) electrocatalysts with clusters stabilized by interaction with carbon defects and subplanted Pt atoms. HER mass activities for the Ptn/HOPG electrodes were between 2 and >10 times higher than that for the surface monolayer of bulk Pt, and also substantially higher than those for Ptn deposited on several other supports. Thus, high activity is a property of the Pt-C structures formed by Ptn – HOPG impacts, rather than simply being due to high Pt surface availability. The physical properties of the Ptn/HOPG electrodes were probed by X-ray photoelectron spectroscopy, low energy He+ ion scattering, and a variety of electron microscopy methods. Born-Oppenheimer molecular dynamic (BOMD) trajectories were used to simulate the Ptn – HOPG impacts, revealing the types of structures formed at different energies, then DFT was used to simulate different HER pathways for the most important Ptn-carbon structures. Experiment and theory lead to the following mechanism. For low deposition energies, the Ptn deposit onto the HOPG surface with sticking probability below unity, and aggregate at defects and steps. With increasing deposition energy, the sticking probability initially decreases, then approaches unity as subplantation and carbon defect creation increase the Pt-surface binding. The rate-limiting barriers for HER at the structures formed in the trajectories were found to be low and weakly dependent on the Ptn size, consistent with the observation of high HER activities that were weakly size dependent. The activity of the surface atoms in Ptn/HOPG were highest for small covalently-bonded Pt-C structures created at high deposition energies. The activity of the larger aggregated structures form at low energies was lower, but still substantially higher than the surface monolayer of bulk Pt. The catalysts were stable in repeated potential cycling, unless the upper potential ranged into oxidizing conditions. In that case, electrodes with Pt all on the surface were still stable, but electrodes containing subplanted Pt became more active as Pt emerged into electrolyte-accessible surface sites.

Methodology is presented for non-destructive, optically-detected single nanoparticle (NP) mass spectrometry, with the goal of extracting surface reaction kinetics for single NPs at high temperatures. Methods for determining the NP charge, mass, and temperature as a function of time are discussed, and the data are used to extract both the absolute kinetics for mass change, as well as the efficiencies of the surface processes that cause them. Factors that contribute to the uncertainties in absolute and relative mass determination, and in the resulting kinetic parameters, are discussed. The method allows the NP-to- NP variations in initial reactivity to be measured directly, along with the time evolution of reactivity resulting from NP structural/compositional changes that occur under reaction conditions. The strengths and limitations of single nanoparticle mass spectrometry as a high temperature surface kinetics tool are discussed in the context of sublimation and O2 oxidation kinetics experiments for single hafnium (Hf) NPs at temperatures ranging above 2400 K. The Hf oxidation kinetics are compared to analogous oxidation experiments for silicon, graphite, and carbon black NPs. In all four cases, the oxidation chemistry was dominated by processes that result in net mass loss, and the distinct mechanisms responsible are discussed. All four NPs also eventually passivated, i.e., the efficiencies for oxidative etching decreased by at least two orders of magnitude, relative to the initial efficiencies. The passivation mechanisms, which are quite different for carbon, compared to silicon or hafnium, are discussed. Carbon NP passivation is attributed to structural isomerization leading to fully coordinated, fullerene-like NP surfaces, while for silicon and hafnium, passivation results from delayed formation of an oxide layer, triggered by accumulation of oxygen in the NP sub-surface region.

The reactivity and adsorbate binding properties of alumina-supported Ptn and PtnGem (n/m = 4/1, 7/2, 11/4) model catalysts were studied using a combination of ethylene and CO temperatureprogrammed desorption (TPD), He+ ion scattering (ISS), and detailed density functional theory (DFT). The Ptn/alumina catalysts gradually deactivated in repeated CO TPD runs and deactivated more rapidly in repeated ethylene TPD runs due to the effects of sintering and carbon deposition (coking). PtnGem/alumina catalysts were dramatically more stable against deactivation by both sintering and coking. Carbon addition to Ptn/alumina also increased stability against sintering, but not coking. Ion scattering was used to probe the nature of adsorbate binding, including the effects of both carbon and Ge addition on the accessibility of strong Pt-associated binding sites for ethylene and CO. DFT was used to examine the electronic and geometric structures and ethylene and CO binding properties of Ptn (n = 4, 7) with added Ge, carbon, or both. Consistent with the ISS results, DFT shows that loss of strong adsorbate binding sites mostly resulted from changes to the electronic properties of the clusters, rather than simple blocking of binding sites by Ge or carbon

Ru metal clusters can get encapsulated into TiO₂ upon heat treatment or cluster source deposition.