Laurens K. Verheij

Temperature dependency of the initial sticking probability of H2 and CO on Pt(111)

The initial sticking probability of the reactive gases H2 and CO on a nearly defect free Pt(111) surface is studied in the temperature range 90–300 K by means of Thermal Energy Atom Scattering (TEAS). By means of TEAS relative initial sticking probabilities can he measured with great accuracy. H2/Pt(111): The initial sticking probability is found to increase with increasing surface temperature. The important role in the chemisorption process played by defects, even at concentrations < 10−3 is emphasized. A two-stage model is proposed to explain these results. CO/Pt(111): The initial sticking probability is found to decrease with increasing surface temperature. This observation is explained with a precursor model.

New phenomena in homoepitaxial growth of metals

The growth of Pt(111) by Pt vapour deposition is studied by He diffraction as a function of substrate temperature and deposition rate. At a deposition rate of about 2.5×10−2 monolayers/second several growth modes are observed: layer-by-layer (2D-) growth at 450 K≲Ts≲800 K, multilayer (3D-) growth at 340 K≲Ts≲450 K and reentrant layer-by-layer (2D-) growth at Ts≲340 K. The observed growth modes and in particular the reentrant 2D-growth are shown to be characteristic of growing Pt(111) under clean conditions, i.e. not influenced by contaminants. The influence of the intra- and interlayer mass transport on the growth mode is discussed in the light of experimental and simulation results. The 3D-growth mode is attributed to the existence of an activation barrier which suppresses the descent of adatoms from the top of the growing adatom islands onto the lower terraces. The barrier can be overcome by thermal adatoms at Ts≳450 K enabling interlayer mass transport which leads to 2D-growth. The reentrant 2D-growth occurs due to a break down of this barrier for small, irregularly shaped islands.

Autocatalytic behavior and role of oxygen diffusion in the hydrogen-oxygen reaction on Pt(111)

Abstract A short review of the hydrogen-oxygen reaction on Pt(111) is given in which the emphasis is put on the apparent contradictions between experiments reported in the literature. We consider these results in the light of our recent observation that certain defects appear to be very active sites for this reaction. Making the basic assumption that the reactive sites play the central role in the hydrogen-oxygen reaction, it is concluded that two fundamentally different mechanisms for the overall reaction are possible: (1) A process in which diffusion of oxygen to the reactive sites is essential for the reaction, and (2) an autocatalytic process in which hydrogen reacts at the reactive sites and is transferred to oxygen on the terrace via molecular reactions between adsorbed H2O, the reaction intermediates and unreacted oxygen. Based on these ideas, a model for the hydrogen-oxygen reaction on Pt(111) is proposed. With the model a consistent explanation of the reported measurements can be given, i.e., it can resolve the apparent contradictions between the different experiments. A transition between these mechanisms is expected in the temperature range where H2O starts to desorb and adsorbed oxygen atoms become mobile, i.e., around 250 K. We present experiments obtained in this temperature range using molecular beam titration and helium diffraction. The measurements clearly show the transition between the autocatalytic and the diffusion controlled mechanism and they are considered, therefore, as a strong support for the model.

A molecular beam study of the interaction between hydrogen and the Pt(111) surface

Hydrogen adsorption on and desorption from a Pt(111) surface is investigated in the temperature range 570 45°). The results can be described by a desorption function which is proportional to the square of the perpendicular energy: Sd = S1E⊥2. According to the principle of detailed balance the sticking probability Sa should show a cos4ϑi dependence in that case. However a marked deviation from this behaviour is observed which shows that a second interaction mechanism is involved which is not (or hardly) dependent on E⊥: Sa = S1E⊥2 + S2. It is found that S2 increases strongly with temperature whereas S1 decreases by about 35% when heating the surface from 670 to 1070 K. The present results are in good agreement with adsorption experiments performed at low temperature (160 K) which will be presented elsewhere. Both interaction mechanisms are discussed in terms of atomistic models. The temperature dependence of S1 seems to be in conflict with the activation barrier model, which has been proposed previously, suggesting that another process is responsible for the observed behaviour. The second mechanism (S2) cannot be explained by surface defects. We attribute this mechanism to adsorption in a molecular precursor state which becomes more efficient with increasing surface temperature.

A molecular beam study of the interaction of CO molecules with a Pt(111) surface using pulse shape analysis

Abstract CO adsorption/desorption on a clean Pt(111) surface has been studied using molecular beam relaxation spectroscopy (MBRS). In contrast to conventional MBRS experiments, lock-in tecniques (or Fourier analysis) have been used here only for a qualitative survey. Pulse shape analysis, which allows the deduction of more detailed information from the experimental data, is discussed in detail, compared to conventional Fourier analysis and used for the results presented here. Detailed analysis of the shape of the MBRS pulse waveform has been used to determine the rate constant for CO desorption, the fraction of scattered signal attributable to chemisorption and the time-of-flight distribution of the non-chemisorbed fraction. The rate constant for CO desorption was measured with MBRS in the temperature range 530–650 K. Complementary measurements under quasi-equilibrium conditions using thermal energy atom (helium) scattering (TEAS) were also performed to extend the desorption rate constant determination down to 430 K, allowing accurate determinations of k over six orders of magnitude. On the clean surface (coverage k = 1.5 × 10 5 T 3 s exp (−28.8/ RT s ) s −1 (or in Arrhenius form, k = 4.3 × 10 14 exp (−32.0/ RT s ) s −1 ), with R in kcal mol . The MBRS measurements have also afforded identification of three distinct interactions of CO molecules with the clean Pt(111) surface: chemisorption, direct scattering and a third interaction showing all characteristics which are expected for physisorption. Approximately 3% of the molecules incident at zero coverage desorb from this state independently of temperature in the range 530–650 K. The angular distribution of the directly scattered molecules shows a peak with its maximum shifted away from the specular direction toward the surface normal. The time-of-flight distribution of the directly scattered molecules is similar to that of the incident beam, though somewhat broadened. Both the shift away from the specular direction and the broadening are ascribed to inelastic effects. Measurements of the relative intensities of the chemisorbed and physisorbed signals as a function of surface temperature provide no support to the assignment of the clean surface physisorption state as a precursor to chemisorption.

Adsorption, 2D phase transition and commensurate 2D phase of Xe on Pt(111)

Abstract Xe on a (nearly) defect free Pt(111) surface has been studied using Thermal Energy Atom Scattering (TEAS) and Low Energy Electron Diffraction (LEED) between 80 and 130 K. Xe adsorbs with a sticking probability close to unity. The isosteric heat of adsorption, about 6.4 kcal mol at zero coverage, increases with coverage indicating attractive mutual interactions. A dilute-condensed phase transition is observed at low coverages ( θ Xe ≌ 0.01 ). Upon θ Xe ≈ 0.27 the condensed Xe 2D overlayer is commensurate with the Pt(111) substrate exhibiting a (√3 × √3) R 30° structure. At monolayer completion (θ Xe ≈ 0.38) an incommensurate hexagonal compression structure is formed. The triple point temperature is 98 ± 2 K and the critical temperature is approximately 120 K.

Hydrogen adsorption on oxygen covered Pt(111)

Abstract Adsorption of hydrogen on oxygen covered Pt(111) is investigated in the temperature range 300–600 K by titration of adsorbed atomic oxygen with hydrogen from a supersonic beam. In most experiments the conditions were such that hydrogen adsorption was rate limiting for oxygen coverages larger than 10% of the saturation coverage. In that case, the hydrogen sticking probability is equal to the water formation rate per incident molecule. Activated and non-activated adsorption are observed. The two processes show qualitatively different dependences on oxygen coverage. The probability for non-activated adsorption does not depend on the coverage of disordered oxygen, but it increases with increasing order (on a scale of 2–4 atoms) of the adsorbed oxygen layer. The probability for activated adsorption decreases with oxygen coverage and is not sensitive to the order of the oxygen layer. Atomic steps change, above all, the adsorption characteristics at low coverages. We can exclude that steps are involved in the non-activated process. Under the experimental conditions in which hydrogen accumulates on the surface to a non-negligible coverage, we observe a phase separation between the hydrogen and oxygen, indicating a decreased binding energy of H atoms inside O islands on the surface.

Investigation of a randomly stepped Pt(111) surface using thermal energy atom scattering (TEAS)

Abstract The angular distribution of He atoms scattered from a randomly stepped surface differs in two aspects from a distribution obtained for a flat surface. The total yield in the specular beam is decreased due to diffuse scattering from the areas around the steps (1) and the angular distribution of the scattered atoms is broadened under anti-phase scattering conditions due to interference between scattering from terraces at different levels (2). Both these effects were used to investigate a Pt(111) surface on which steps were produced by high temperature sputtering. The angular distribution of the He atoms scattered in the specular beam is analysed with the facet model. A justification for applying this model is given. It is shown that any surface model with which the experimental angular distribution can be reproduced, may be applied to analyse the peak profile in terms of a step density. Determination of the step density demands correcting for a background (diffuse scattering) which introduces an uncertainty in the result of 5–10%. By comparing the step density obtained from this analysis with a total yield measurement, a cross section per unit length for diffuse scattering from a step edge D = 12 A is found for 16 meV He atoms.

Kinetic modelling of the hydrogen-oxygen reaction on Pt(111) at low temperature (< 170 K)

Abstract Recently a kinetic model was proposed for describing the hydrogen-oxygen reaction on Pt(111). This model is based on the reactive-site mechanism, i.e. only a very limited number of Pt sites are considered to be catalytically active for the actual water formation reaction. Here we consider the implications of the model for the H 2 O 2 reaction at low temperatures ( 2 O and reaction intermediates are taken into account which enable the complete conversion of oxygen to H 2 O. The model is consistent with the observed transition from a reaction mechanism through which H 2 O can be formed at temperatures as low as 130 K to the much slower diffusion-controlled mechanism which was found to dominate at low oxygen coverages at higher temperatures (> 250–300 K). Using the same reaction parameters, data sets of Ogle et al. and Germer et al. can be simulated quite well with the model. The actual water-formation reaction (at the reactive site), the reaction between H 2 O and O atoms (at the “reaction front”) and the adsorption of hydrogen appear to be the rate-limiting steps. The simulations indicate that the adsorption of hydrogen proceeds via a rather complex process, which is difficult to incorporate correctly in the model.

The hydrogen-oxygen reaction on the Pt(111) surface; influence of adsorbed oxygen on the sticking of hydrogen

The hydrogen-oxygen reaction on the Pt(111) surface was investigated with a slowly modulated (v ⩽ 0.5 Hz) molecular beam: oxygen atoms, adsorbed on the surface from the ambient gas, are periodically titrated by a pure D2 or a mixed H2D2 beam. The reaction is characterized by monitoring the desorption of D2O and HD. The adsorbate coverage is also continuously measured by monitoring the specular H2 intensity. LEED and AES were used to characterize the clean and oxygen covered Pt(111) surface. LEED measurements show that the (2 × 2)O structure observed at 400 K becomes disordered above 500 K. In the same temperature range a change in the dependence of the hy/fdrogen sticking probability SH2 on the oxygen coverage θO is found: at 400 K SH2(θO) is independent of oxygen coverage, whereas at 550 K SH2(θO) decreases with θO. We propose that the decrease of SH2 is due in the oxygen adlayer. With this model, one can also explain the extreme sensitivity of the reaction to contamination, which we observe, and the discrepancies with respect to previously reported results.