AVS2015 Session SS+AS+EN-WeA: Metals, Alloys & Oxides: Reactivity and Catalysis

The nucleation, growth and chemical activity of bimetallic Pt-Re clusters on titania have been investigated as model systems for understanding Pt-Re catalysts for oxidation reactions. Scanning tunneling microscopy studies demonstrate that exclusively bimetallic clusters can be grown from the sequential deposition of Pt on Re or Re on Pt, provided that the deposition of the first metal creates a high enough cluster density for the nucleation of the second metal. Low energy ion scattering experiments indicate that the bimetallic clusters are Pt-rich regardless of the order of deposition. However, X-ray photoelectron spectroscopy (XPS) suggest that a Pt-Re alloy is formed from deposition of Re on Pt but not from Pt on Re. Furthermore, Re interacts more strongly with the titania support than Pt, resulting in reduction of titania. Temperature programmed desorption studies for CO desorption and methanol reaction confirm that the Re clusters have lower activity than Pt despite their higher surface area, and this behavior is attributed to oxidation of Re by the titania support. The alloyed clusters exhibit new activity for CO and H₂ evolution that is not observed for the pure or unalloyed clusters. Methanol oxidation activity of these model surfaces are studied in a microreactor attached to an ultrahigh vacuum chamber so that the surfaces can be characterized by XPS before and after reaction. Specifically, changes in the oxidation states of Re in the bimetallic and pure Re clusters are investigated.

Transition metals that are used to catalyze reactions of hydrocarbons are often deactivated by the deposition of unreactive carbon on the catalyst surface. The structure and properties of the deposited carbon are often poorly defined. We have investigated the reactivity of carbon deposited onto a Pt(111) surface through exposure to acetylene at 750 K. At this temperature the acetylene is completely dehydrogenated leaving only carbon the surface. Earlier work had shown that the carbon deposited in this way largely consists of C₂ molecules. We have used reflection absorption infrared spectroscopy (RAIRS) to characterize the reactivity of the deposited carbon under ambient pressures of H₂(g) up to 10 torr. The results show that C₂ can be hydrogenated to ethylidyne (CCH₃) and that the ethylidyne is slowly hydrogenated to ethane, which desorbs thus removing carbon from the surface. The maximum coverage of the C₂ molecules can be deduced from comparison with the peak areas measured with RAIRS for ethylidyne formed from ethylene exposure, which is known to give an ethylidyne coverage of 0.25 monolayer. Auger electron spectroscopy confirms that surface carbon is removed by hydrogenation under these conditions. In separate experiments based on comparisons of s- and p-polarized RAIR spectra in which both surface and gas phase species can be simultaneously monitored, we have shown that surface ethylidyne is a spectator species as gas phase acetylene is converted first to gas phase ethylene and then to gas phase ethane. Although ethylidyne is a spectator species in the hydrogenation of ethylene and acetylene to ethane over Pt(111), in the case of C₂ hydrogenation, ethylidyne plays the role of a reaction intermediate.

Experimental and theoretical results concerning the thermodynamics and kinetics of surface chemical reactions of importance in late transition metal catalysis will be reviewed. Topics include: (1) calorimetric measurements of the adsorption energies of small molecules and molecular fragments on single crystal surfaces, and their comparison to different DFT methods; (2) measurements of the entropies of adsorbates and their trends, (3) using these together with elementary-step rate measurements to build microkinetic models for multi-step catalytic reactions, and a method for analyzing these that quantifies the extent to which each elementary step and intermediate controls the net rate; and, (4) measurements of the energies of transition metal atoms in nanoparticle catalysts as a function of particle size and support, which correlate with catalytic activity and sintering rates.

· Work supported by NSF and DOE-OBES Chemical Sciences Division.

TiO₂ is an important photocatalyst with many practical applications. It is also a good model system for fundamental studies of thermal and non-thermal reactions, including photocatalytic water splitting. Our understanding of water’s reactions on TiO₂ surface is limited. In this paper, we have investigated temperature-dependent reaction of water molecules with bridging oxygen vacancy (V_O) on rutile TiO₂(110) surface using three independent methods: i) infrared reflection absorption spectroscopy (IRAS) to monitor the bridging hydroxyl (OH_B or OD_B) formation, ii) electron-stimulated desorption (ESD) of molecular water to monitor the water coverage,¹and iii) photon-stimulated desorption (PSD) of CO₂ – which is a product of CO photooxidation – to monitor the unoccupied V_O coverage. ² Narrow, distinct peaks for isolated OD_B and OH_B at ~2736.5 cm^-1 and ~3711.5 cm^-1 are detected in P-polarized mode for the samples exposed to D₂O and H₂O, respectively. If water is dosed at low temperature and annealed, bridging hydroxyl peaks appear above 150 K, growing with temperature until ~ 250 K, then saturate. In the same temperature range, molecular water and V_O coverages from the ESD and PSD data decrease in correlated fashion according to the reaction H₂O_Ti + V_O -> 2OH_B. The temperature range for this conversion appears to be too broad to be fitted with a single Arrhenius term and a reasonable pre-factor. On the other hand, the data can be fitted well using a “normal” prefactor (ν = 10¹² s^-1 ) and a distribution of activation energy (E_a) centered at 0.545 eV with ΔE_a(FWHM) = 0.125 eV. These parameters are close to those obtained from STM data³ and theory⁴ for the water monomer diffusivity on Ti sites, which most likely controls the water – vacancy reaction. This work was supported by the US Department of Energy, Office of Science, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences & Biosciences.

(1) Zhang, Z.; Du, Y.; Petrik, N. G.; Kimmel, G. A.; Lyubinetsky, I.; Dohnalek, Z. Water as a Catalyst: Imaging Reactions of O₂ with Partially and Fully Hydroxylated TiO₂(110) Surfaces. J. Phys. Chem. C 2009,113, 1908-1916.

(2) Petrik, N. G.; Kimmel, G. A. Off-Normal CO₂ Desorption from the Photooxidation of CO on Reduced TiO₂(110). J. Phys. Chem. Lett. 2010,1, 2508-2513.

(3) Matthiesen, J.; Hansen, J. O.; Wendt, S.; Lira, E.; Schaub, R.; Laegsgaard, E.; Besenbacher, F.; Hammer, B. Formation and Diffusion of Water Dimers on Rutile TiO₂(110). Phys. Rev. Lett. 2009, 102, 226101.

(4) Hammer, B.; Wendt, S.; Besenbacher, F. Water Adsorption on TiO₂. Top. Catal. 2010, 53, 423–430.

Pd-Au alloys are known to catalyze a wide range of hydrogenation and oxidation reactions; however, the size of Pd ensembles in Au required for small molecule activation is not well understood. On the atomic scale, we investigate size effects of Pd atoms in Au for the catalytic activation of H₂ and O₂. We show that isolated Pd atoms are capable of catalyzing the dissociative adsorption of H₂, a process which was previously thought to require contiguous Pd sites. H spillover from active Pd sites to the Au surface can be induced the adsorption of CO. Conversely, single Pd atoms are not capable of O₂ dissociation. Small Pd clusters on Au enable O₂ activation and adsorption at Pd-Au interface sites. Since weakly bound H and O atoms are capable of enhancing reaction selectivity on Au substrates, this Pd-Au system serves as an ideal model system with which to probe selective hydrogenation and oxidation reactions at both single and ensemble active sites.

Selective decomposition on metal catalysts is a critical step in formic acid’s application as a hydrogen storage molecule and for its use in direct formic acid fuel cells. Depending on the metal, formic acid can decompose via a dehydrogenation pathway to produce CO₂ and H₂, or a dehydration pathway to produce CO and H₂O. For most applications, very high selectivity to dehydrogenation is preferred as reactively formed CO from dehydration can poison the catalyst. The Cu(110) surface is known to selectively decompose formic acid via dehydrogenation, however, despite being the most dominate facet of nanoparticles, Cu(111) has received little study. Pt surfaces exhibit greater reactivity to decomposition, but are not as selective resulting in increased catalyst poisoning. We report that formic acid on Cu(111) and Pt/Cu(111) selectively decomposes via dehydrogenation. We find the bare Cu(111) surface to be 100% selective towards dehydrogenation, but not very active. Substitution of 1% of a monolayer of Pt into the Cu(111) surface results in a single atom alloy (SAA) that maintains this high selectivity and is ~six times more reactive than Cu(111). Higher coverages of Pt improve reactivity further, but beyond the single atom regime the selectivity towards dehydrogenation decreases and dehydration is observed. Our results show that Pt/Cu SAAs significantly improve the reactivity of Cu, while also maintaining high selectivity towards dehydrogenation, therefore avoiding catalyst poisoning by CO. Based on our results, real nanoparticle catalysts designed on the SAA principle are expected to be promising candidates for formic acid dehydrogenation.