Papers

AVS1997 Session SS-WeM: Activation and Dynamics of Reactions

Wednesday, October 22, 1997 8:20 AM in Room A3/4

Wednesday Morning

Start	Invited?	Item
8:20 AM		SS-WeM-1 Interaction of Gas-Phase D Atom with O₂ Chemisorbed on Pt(111): O₂ Desorption and D₂O Formation at 85 K J.M. Lee, J.-Y. Kim (Seoul National University, Korea) The surface processes induced by gas-phase D atoms incident on O₂-adsorbed Pt(111) at 85 K have been studied by real-time monitoring of the desorbing species with a mass spectrometer. Upon subjecting a Pt(111) surface covered with chemisorbed O₂ at 85 K to a flux of D atoms, desorption of O₂ and D₂O was observed. While the desorption signal of O₂ shows a step-like jump at time zero followed by an almost exponential decay, that of D₂O slowly increases to reach a maximum at finite time. A series of post-TPD spectra taken as a function of D atom exposure showed three D₂O desorption peaks at 205, 180, and 167 K, which is due to OD recombination, recombination between adsorbed OD and D, and desorption of adsorbed D₂O, respectively. Furthermore, the 205 K peak appears first and then is successively replaced by a lower temperature peak. Based on these observations and the kinetics of D₂O evolution, we conclude that (1) D₂O is formed by a series reaction via an adsorbed OD intermediate and (2) most of D₂O remains on the surface and only a small fraction desorbs directly. The angular distribution of desorbing O₂ is peaked in the surface normal and can be best represented by cos(θ) squared. A possible mechanism of O₂ desorption involving a hot reaction intermediate will be discussed in conjunction with low temperature CO oxidation triggered by D atom.
8:40 AM		SS-WeM-2 Reaction of Gas-Phase Atomic Oxygen with Preadsorbed Deuterium on Ru(001) M.J. Weiss, C.J. Hagedorn, W.H. Weinberg (University of California, Santa Barbara) Ultrahigh vacuum (UHV) studies on well-defined surfaces have yielded considerable insight into catalysis at surfaces. The existence of the well known "pressure gap" has, however, often prevented the study of chemical reactions under UHV conditions. One example of a reaction inhibited by the pressure gap is the hydrogen oxidation reaction on Ru(001). Under UHV conditions, oxygen and hydrogen can be coadsorbed on Ru(001) at temperatures below 300 K. Heating of a coadsorption layer does not lead to the formation of water, but rather results in the desorption of hydrogen. One method of bridging the pressure gap that has increasingly been utilized is the use of reactive gas-phase atomic species to effect reactions that previously have been inaccessible to UHV investigators. Most investigations of this type have utilized atomic hydrogen, which can be easily produced using a hot tungsten filament or microwave cavity. Previous work in our group utilized gas-phase atomic hydrogen to produce water and hydroxyl species from oxygen overlayers on Ru(001). This work takes a different approach to the hydrogen oxidation reaction that should be applicable to a wide range of catalytic oxidation reactions. Here, we react gas-phase atomic oxygen (which is produced in a microwave cavity) with deuterium overlayers on Ru(001) to produce water. (Deuterium is used to eliminate any contribution from background water adsorption.) This reaction produces fairly sharp D₂O desorption features between 150 and 225 K in temperature programmed desorption studies. The detailed kinetics of this reaction provide additional insight into the hydrogen oxidation reaction which is an important model reaction for the understanding of catalytic oxidation reactions. Supported by the Department of Energy (Grant No. DE-FG03- 89ER14048); MJW and CJH acknowledge support by NSF predoctoral fellowships.
9:00 AM		SS-WeM-3 Dynamical Simulations of the Molecular Adsorption of Alkanes on Platinum Single Crystal Surfaces R.J. Madix, J.F. Weaver, J.A. Stinnett (Stanford University) Many surface reaction processes occur via molecular precursor states. In such processes the molecule is first adsorbed; subsequent reaction then competes with desorption. The overall reaction rate is given by the adsorption probability times the branching ratio - the ratio of the reaction rate constant to the sum of the rate constants for desorption and reaction. With accurate potential functions it should be possible to predict adsorption probabilities from dynamical simulations for a wide range of conditions. A significant experimental effort has been expended recently to understand the energetics of adsorption processes using molecular beams. In such experiments the translational and rotational energies of the molecules are controlled, and the adsorption probabilities measured as a function of the incident energies and angles. A methyl- platinum potential has been developed that accurately accounts for the dynamics of adsorption of ethane on Pt(111) and successfully predicts the adsorption probabilities of ethane on Pt(110) (including the azimuthal angle dependence) and the adsorption probability of propane on Pt(111) and Pt(110). In addition this potential accurately predicts the adsorption probabilities of n-butane, isobutane and neopentane on Pt(111) over a wide range of incident angles and energies. The success of this potential affords some confidence in the dynamical features of the adsorption process that are revealed in the trajectory simulations. These features will be discussed.
9:40 AM		SS-WeM-5 Kinetics and Dynamics of O₂ Displacement from Pt(111) by Incident CO Molecules C. Åkerlund, I. Zori@aa c@, B. Kasemo (Chalmers University of Technology, Sweden) We have investigated the O₂ displacement kinetics and dynamics, from a Pt(111) surface at 100K, induced by incident CO molecules. The central issue we explore is how the efficiency of this process depends on E_CO and on O₂ coverage. Simultaneous measurements of the O₂ displacement rate and CO adsorption rate are performed for a range of E_CO between 0.06 eV and 1.91 eV, and for surface temperatures between 95 K and 110 K. The dynamics and the kinetics of the process are found to behave distinctly different above and below E_CO=0.7eV. In the low energy regime a simultaneous decrease of the O₂ displacement rate and CO adsorption rate with increasing E_CO is observed. In contrast, when E_CO>0.7 eV the displacement rate increases while at the same time the CO adsorption rate remains constant, independent of E_CO. This behaviour suggests an opening of a new displacement channel at higher energies. The initial removal efficiency is constant at low energies at about 0.1 O₂ per adsorbed CO. For E_CO>0.7 eV that number increases with increasing E_CO, and at 1.91 eV it is almost 0.2. The accumulating CO coverage on the surface enhances the efficiency of the O₂ removal process. We suggest that the process, both for low and high E_CO, is primarily driven by the energy released during the CO chemisorption, and that the differences between the two energy regimes are related to the differences in the CO adsorption dynamics, rather than to different mechanisms for O₂ desorption. A kinetic model in support of this is presented.
10:00 AM		SS-WeM-6 CO Oxidation on Pt(111) at 85 K Triggered by Gas-Phase D Atoms J.-Y. Kim, J.M. Lee (Seoul National University, Korea) Low temperature CO oxidation on a CO+O₂/Pt(111) surface triggered by deuterium atoms incident from the gas-phase has been studied with a pyrolytic hydrogen atom source and a differentially pumped mass spectrometer. When a Pt(111) surface saturated first with O₂ and then with CO at 85 K was suddenly subjected to a beam of D atoms, the desorption signals of both CO₂ and O₂ showed an instantaneous jump to a maximum followed by a monotonic decay. Direct desorption of D₂O with relatively a smaller intensity was also observed, but it slowly increased from zero to a maximum and then decayed. A series of post exposure TPD spectra taken for increasing D atom exposures revealed that D₂O molecules were produced by a series reaction via an adsorbed OD intermediate, and most of D₂O thus produced are left on the surface. The angular distributions of CO₂ and O₂ are sharp and peaked in the surface normal. They could be best represented by a cos(θ) to the power n with n = 9 and n = 3 for CO₂ and O₂, respectively. CO oxidation observed here does not occur by the Langmuir-Hinshelwood mechanism, but is initiated by the gas-phase D atom. We rule out O₂ as the oxidizing species based on the faster signal decay of O₂ than that of CO₂. We propose a mechanism in which D atom first reacts with chemisorbed O₂ to form hot OD and O, whereby the hot O atom in turn reacts with a neighboring CO admolecule before it becomes accommodated.
10:20 AM		SS-WeM-7 Ammonia Decomposition Induced by Gas-Phase Atomic Hydrogen on Ru(001) C.J. Hagedorn, M.J. Weiss, W.H. Weinberg (University of California, Santa Barbara) Historically, as new techniques and concepts in the field of catalysis have been developed, they have been applied to the study of the ammonia synthesis reaction. One relatively new technique in ultrahigh vacuum (UHV) studies of catalysis is the use of energetic reagents such as gas-phase atomic hydrogen to bridge the "pressure gap" which has inhibited the study of industrially important reactions under UHV conditions. Here, we present what is, to the best of our knowledge, the first use of gas-phase atomic hydrogen in the study of the ammonia synthesis reaction. Although this work deals largely with the ammonia decomposition reaction, considerable insight into the ammonia synthesis reaction can be gained through consideration of the microscopic reversibility of the individual reaction steps as well as through the study of the overlayers formed following ammonia decomposition. In this work, we have effected the (partial) decomposition of adsorbed ammonia at 100 K using gas phase atomic hydrogen. By exposing ammonia layers to atomic hydrogen, we are able to create high surface coverages of both nitrogen and hydrogen. The decomposition of the ammonia presumably occurs by abstraction of one or more hydrogen atoms from the molecule by the impinging atomic hydrogen. Temperature programmed desorption studies show the desorption of large amounts of nitrogen containing species (which appear to be predominantly hydrazine and molecular nitrogen) in the temperature range of 500-1100 K. This is novel chemistry since in previous studies of the ammonia synthesis reaction, dinitrogen is the only desorption product above 500 K. Further characterization of the overlayers formed in these studies will enable us to investigate nitrogen and hydrogen coverages which have previously been inaccessible under UHV conditions. Supported by the National Science Foundation (Grant number CHE- 9626338). CJH and MJW were supported by NSF predoctoral fellowships.
10:40 AM		SS-WeM-8 Investigation of the Effects of Low-Energy Electron Bombardment on Coadsorbed Layers of Hydrogen and Ammonia on Pt(111) C. Bater, J.H. Campbell, J.H. Craig (University of Texas, El Paso) The effects of electron-impact on ammonia covered Pt(111) and coadsorbed ammonia and deuterium on Pt(111) have been studied using temperature programmed desorption (TPD) and electron-stimulated desorption (ESD). For coverages at or below one monolayer, ammonia adsorbs on the surface in two distinct TPD states: the α-state is broad and desorbs over the temperature range 160-400K, and the ß-state appears as a sharper peak at 150 K. The ß-state was found to be more susceptible to electron-beam damage than the α-state, resulting in formation of atomic nitrogen adatoms as indicated by the mass 28 recombinative nitrogen TPD peak appearing at 550 K. ESD kinetic energy distributions (KEDs) were obtained for m/e = 1 and 14 amu, which exhibited broad peaks generally. The H⁺ KEDs were analysed using empirical curve fits, with the resulting conclusion that the H⁺ KEDs contain contributions from at least three different hydrogen-containing surface species. We believe that these three H⁺ KED peaks are due to ESD from adsorbed NH₃(a), NH₂(a), and H(a). The ESD cross-section for NH₃ removal from clean Pt(111) was measured in three different ways, all of which were found to be in general agreement, and which gave an averaged cross-section value of Q = 4±1 x 10^-17 cm². Recombinative deuterium desorption from clean Pt(111) is compared to D₂, HD, and H₂ TPD spectra for ammonia adsorbed on a Pt(111) surface with varying exposures of preadsorbed deuterium. Deuterium-containing ammonia fragments are also sought. Additionally, the effects of electron-irradiation on the coadsorbed layers is investigated using both TPD and ESD.
11:00 AM		SS-WeM-9 Theory of Current-induced Bond Breaking by the STM: O₂/Pt(111) M. Persson, S. Gao, B.I. Lundqvist (Chalmers Univ. of Techn. and Göteborg Univ., Sweden); B.C. Stipe, M.A. Rezaei, W. Ho (Cornell University) We have developed a simple theory for bond breaking by tunneling electrons from a tip of an STM. The bond is activated by vibrational excitations induced by resonant inelastic electron tunneling. The theory explains the measured power-law dependence of the dissociation rate, R_d, of single O₂ molecules chemisorbed on Pt(111) on the tunneling current, I, for various applied voltages V¹. In contrast to Eigler’s atomic switch, the vibrational relaxation rate is much larger than the tunneling rate so that the dissociation is dominated by the processes that involves a minimum number of transitions among the bound state levels. Thus at voltages for which there are tunneling electrons with energies larger than the dissociation barrier, R_d is dominated by single transitions and R_d is linear in I whereas at lower voltages the number of transitions and the accompanied power-law dependence of R_d on I depends on the bound state level structure. These studies demonstrate that bond breaking by inelastic electron tunneling enables us to manipulate matter on the atomic scale and with single atom precision. ¹Stipe et al., Phys. Rev. Lett. June 9, 1997
11:20 AM		SS-WeM-10 Low Energy Dynamics for Mixtures of S and CO on Cu(111) C.J. Hirschmugl (Lawrence Berkeley National Laboratory); G.P. Williams (Brookhaven National Laboratory) Low-energy interactions between adsorbates on surfaces can now be probed with high resolution using Far-IRAS. In the present study, CO on Cu(100) surfaces predosed with sulfur has been investigated with TPD, Auger and Far-IRAS. The latter technique, which employs synchrotron radiation, can extend traditional IRAS measurements to below 400 cm ^-1 with noise levels of approximately .01% attainable in 100 seconds measuring time. As previously reported ¹, Far-IRAS measurements for CO/Cu(100) have revealed for the first time the C-metal stretch (344 cm ^-1) and the hindered rotation (285 cm ^-1), with a concomitant broadband absorption. In the present studies, we monitor the changes induced to these features (and to the C-O stretch) due to a controlled coverage of S as a sensitive indicator of the surface dipole fields in the presence of S. The intensities of the C-O stretch and hindered rotation decrease linearly with S coverage, corresponding to the decreasing amount of CO adsorbed as a function of adsorbed S; however, the dynamic- dipole strength of the C-metal stretch exhibits a more profound change upon S adsorption. Along with these changes in intensity, the widths of the C-O stretch and the C-metal stretch both show monotonic broadening trends as a function of increasing S. The contributions of inhomogeneous broadening and lifetime effects to these trends will be examined. ¹C.J. Hirschmugl, G.P. Williams, B.N.J. Persson, and A.I. Volokitin, Surface Science 317, L1141 (1994).