Low Oxygen Coverage Research Articles

A quantitative LEED analysis of the oxygen-induced p(3×1) reconstruction of Pt 25Rh 75(100)

Pt 25Rh 75(100) forms a p(3×1) reconstruction at saturation coverage of oxygen (23 L O 2, 600°C). A previous STM study on O/Pt 50Rh 50(100) suggests that every third row of the first substrate layer is shifted by half a lattice constant (“shifted rows”). We present a LEED I( E) analysis of Pt 25Rh 75(100) confirming the shifted-row model and find that oxygen resides in threefold-coordinated sites on both sides of the shifted rows. The adsorbate occupies those of the threefold-coordinated sites that are directly separated by the metal atom in the shifted row. Further I( E) calculations exclude the alternative threefold-coordinated adsorption site beside the shifted row as well as the fourfold-coordinated site symmetrically in between the shifted rows. We achieve a Pendry R-factor of 0.14 for the best-fit structure. Oxygen has equal bond lengths to its three metal neighbours, amounting to 1.95 Å. The first substrate layer relaxes outward by 8.8% of the bulk value to 2.08 Å, but we do not observe any significant relaxations of deeper layer spacings. The shifted rows pop out of the surface by 0.38 Å. After determination of the oxygen adsorption site with LEED, we examine local adsorption structures on Pt 25Rh 75(100) at low oxygen coverage with STM. We resolve the shifted rows in real-space, and for special tip conditions, we find maxima of apparent height at in-plane positions that coincide with the oxygen position as established by quantitative LEED. We determine chemical-composition depth-profiles by quantitative LEED for three surface preparations occurring during sample preparation. While the first substrate layer of clean and annealed Pt 25Rh 75(100) is enriched in Pt (76%) as compared to the bulk value (25%), that in p(3×1)-O/Pt 25Rh 75(100) is enriched in Rh (90%). Oxygen adsorption at a moderate temperature (600°C) and formation of the p(3×1) structure reverse segregation on Pt 25Rh 75(100). Finally, oxygen can be removed at room temperature by exposure of the surface to hydrogen. This lifts the reconstruction but keeps the Rh enrichment of the first substrate layer.

On the mechanism of the hydrogen–oxygen reaction on Pt(111)

The reaction between hydrogen and oxygen on Pt(111) is investigated with molecular beam relaxation spectroscopy in the temperature range between 300 and 1200 K. Three procedures were used for preparing the surface, resulting in different reactivities for the hydrogen–oxygen reaction. The observations are not compatible with a reaction mechanism which occurs homogeneously on the surface. Therefore we interpret them in terms of a non-homogeneous surface reaction model in which water formation is catalyzed by special reactive sites (reactive site model). Oxygen diffusion is found to be the rate-limiting process by which water is formed on Pt(111) at low oxygen coverages (<0.03 ML) and temperatures below 800 K. The activation energy for diffusion E a is 11.3±0.5 kcal mol −1. Between 800 and 1200 K we find an effective activation energy for the reaction which is negative (−14.8 kcal mol −1). A consistent explanation of all our results can be given with the reactive site model when it is assumed that two O atoms take part in the formation of one water molecule. We estimate that the density of reactive sites is about 10 −5 on a well-prepared surface. Experimental results that have been reported previously are discussed. It is shown that the reactive site model can explain also those results quite well.

Insights into surface reactivity: formic acid oxidation on Cu(110) studied using STM and a molecular beam reactor

Using a combination of STM and molecular beam reactor data we summarise some important features of a model reaction (formic acid oxidation on Cu(110)) which is of general significance to surface reactivity and to catalysis. Three such features are highlighted here. The first concerns the role of weakly held species (possibly physisorbed) in surface reactions. These species, although of very short lifetime on the surface, can, nevertheless, diffuse over long distances to “find” a sparse distribution of active sites. Thus a very low coverage of oxygen on the surface of Cu(110) increases the sticking probability of all the formic acid molecules which strike the surface to high value (0.82), even though the clean surface is relatively unreactive. The important concept here is the “diffusion circle” or “collection zone” which represents the area of surface visited by the molecule in its short sojourn in the weakly held state. The second theme concerns the concept of the “flexible surface”. We show that the involvement of surface atoms in reactions directs the structure and reactivity for a particular reaction. For formic acid oxidation the liberation of Cu atoms during the removal of oxygen as water leads to gross restructuring of the surface and can lead to “compression” of one reactant (the oxygen in this case) into a lower area, higher local coverage, unreactive state (the c(6×2) oxygen structure). Thirdly, and finally, it is proposed that, for many surface reactions, the surface acts in an analogous way to a solvent, supporting a “dissolved” (highly mobile and fluxional) phase of intermediates at low coverage, which crystallise out above a critical coverage (the 2D “solubility limit”).

The use of low-energy ion scattering spectroscopy for the quantitative determination of adsorption sites in surface chemistry studies

Ion scattering spectroscopy (ISS) was used in conjunction with temperature-programmed desorption (TPD) and X-ray photoelectron spectroscopy (XPS) for the study of the adsorption of hydrogen, oxygen, ethyl iodide, and 2-propyl iodide on Ni(100). In all those systems the decrease in Ni ISS signal intensity seen during the uptake could be explained by shielding of the Ni surface atoms by the adsorbate. In addition, since the ISS data could be correlated nicely with results from other techniques, the conclusion was reached that matrix effects and changes in neutralization probabilities (due to variations in work function) were negligible for the systems studied here. Additional ISS studies for the adsorption of 2-propyl iodide on oxygen-precovered Ni(100) indicated that those molecules bond preferentially to Ni (not O) sites. This confirms previous results which showed that the products of that system associated with chemistry on the Ni sites — propane, propene, and hydrogen — dominated the TPD spectra for low coverages of either 2-propyl iodide or oxygen. In a broader sense, this work also shows that ISS can be used to determine adsorption sites in complex systems where other techniques are not applicable.

Thermal Conversion of 2-Propyl Iodide on O/Ni(100): Changes in Product Distribution with Varying Oxygen Coverages

The oxidation of 2-propyl iodide on oxygen-covered Ni(100) surfaces has been studied with temperature-programmed desorption (TPD), X-ray photoelectron spectroscopy (XPS), and ion-scattering spectroscopy (ISS). It was found that the product distribution is strongly dependent on oxygen preexposure, with partial oxidation being favored at low oxygen coverages and total oxidation dominating on thin oxide films. XP I 3dcore-level spectra indicate that the adsorption of 2-propyl iodide below 100 K is molecular, and ISS data strongly suggest preferential bonding to Ni sites, Annealing the 2-propyl iodide adsorbed on O/Ni(100) surfaces between 120 to 200 K generates 2-propyl groups on the nickel sites via dissociation of the C–I bond, the same as on the clean surface. For submonolayer oxygen coverages the 2-propyl groups then follow one of two reaction pathways: they either undergo hydrogenation–dehydrogenation on the nickel sites to form propane, propene, and hydrogen, or, in the case of those formed near the oxygen sites, migrate and react to form 2-propoxide groups. The 2-propoxide moieties are stable on the surface up to ∼325 K, at which point some undergo β-hydride elimination to yield acetone. The fact that the rate of desorption of acetone from the reaction of 2-propyl iodide with oxygen is reaction limited is supported by the observation that the desorption of molecular acetone from Ni(100) occurs below 300 K. Also, other TPD experiments indicate that propene does not react to yield acetone on O/Ni(100) surfaces, and that the peak shapes and temperatures for acetone desorption from the reaction of 2-propyl iodide and 2-propanol on O/Ni(100) are nearly identical, suggesting that they form from a common intermediate, 2-propoxide.

A photoemission and XAS study of oxygen coadsorbed with a (2 × 2) layer of K on graphite

We have studied the coadsorption of oxygen with a (2 × 2) monolayer of K on graphite. At least three different adsorption phases for oxygen have been found. Different spectroscopic techniques have been used in order to identify the different oxygen species adsorbed on the surface. For low oxygen coverage, the molecules dissociate and form a K 2O species. At higher oxygen doses, K 2O 2 and KO 2 are present on the surface up to saturation coverage. XAS yields an O 2 bond length of about 1.39 Å for the KO 2 in this phase. This species is found to form an ionic complex which is weakly bound to the graphite.

STM studies of oxygen-induced reconstruction on a PtRh(100) alloy surface

Oxygen-induced reconstruction of a PtRh(100) alloy surface was studied by scanning tunneling microscopy (STM). When a clean Pt-enriched PtRh(100) surface was exposed to O 2 at more than 600 K, oxygen-induced reconstruction took place with segregation of Rh atoms to the surface. Depending on the coverage of oxygen, two types of reconstruction on the PtRh(100) alloy surface are induced. A p(3 × 1)-O structure appeared at relatively low oxygen coverage, and the TPD spectrum of the p(3 × 1)-O surface gave a desorption peak of O 2 at ca. 930 K. The p(3 × 1)-O structure showed several different STM images depending on the tip condition, from which we could distinguish the two component metals, RhO rows and Pt rows. Based on this, a model structure was deduced. At a high oxygen coverage, more Rh atoms were segregated to the surface and the STM image suggests a c(2 × 20)-O structure. The c(2 × 20)-O surface changed into the p(3 × 1)-O by heating with the concomitant desorption of O 2 at ca. 830 K. The STM images for this c(2 × 20)-O surface is analogous to the arrangement of Rh atoms on the top layer of a Rh(111) hexagonal lattice. Due to the lattice mismatch between the quasi-Rh(111) overlayer and PtRh(100) substrate, a moire pattern with 20 times periodicity along the 〈111〉 directions was observed. Both the p(3 × 1)-O and c(2 × 20)-O surfaces are easily reduced by H 2 even at room temperature.

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

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 (< 300 K) in more detail. Additional autocatalytic reaction steps between oxygen, H 2O and reaction intermediates are taken into account which enable the complete conversion of oxygen to H 2O. The model is consistent with the observed transition from a reaction mechanism through which H 2O 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 2O 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.

Reconstruction kinetics of a stepped metallic surface: step doubling and singling of Ni(977) induced by low oxygen coverages

Helium atom diffraction has been used to study the reconstruction kinetics of a stepped metallic surface, Ni(977), which sequentially undergoes step-doubling and -singling upon dosing with low coverages of oxygen. Over the temperature range 390–470 K it was found that less than 2% of a monolayer of oxygen was sufficient to transform the initially prepared single-stepped surface to a new steady state having a double-stepped structure. The thermal range over which the doubled phase exists extends to higher temperatures when more oxygen is present. At low oxygen exposures this doubled interface reverts to the single-stepped surface above 470 K. Singling can also be driven by more extensive levels of oxygen adsorption. The kinetics of the step-doubling transformation which occurs below 470 K was determined to be second order with respect to single-step density while, for temperatures above 470 K, step-singling followed first-order kinetics with respect to the double-step density. Oxygen atoms adsorbed at step edges play a crucial role in these transformations. Arrhenius analysis is used to extract energetics for the step-doubling and -singling reconstructions. These results delineate the sequence of mechanistic stages which occur during the initial stages of oxidation of a stepped metallic interface which precede the onset of bulk oxidation, findings which are important for developing an improved understanding of metallic oxidation and corrosion.

Polysilicon gate etching in high density plasmas. II. X-ray photoelectron spectroscopy investigation of silicon trenches etched using a chlorine-based chemistry

We have used x-ray photoelectron spectroscopy (XPS) to study the chemical constituents present on the surfaces after etching of poly-Si features masked with photoresist. Poly-Si etch rates were determined using in situ HeNe laser ellipsometry. The wafers were etched in a low pressure, high density plasma, helicon source using a chlorine/helium gas mixture. The XPS analysis chamber is connected to the reactor chamber via an ultrahigh vacuum chamber allowing quasi in situ analysis of the etched samples. Electrostatic charging of insulating surfaces and geometric shadowing of photoelectrons by adjacent photoresist lines have been used to differentiate photoemission signals from the tops, sidewalls, and bottoms of the features. Blanket and photoresist masked poly-Si samples were etched without biasing the substrate. The poly-Si etch rate decreased as a function of etching time and finally stopped due to the deposition of an oxide film. When biasing the sample, etching occurred and anisotropic etching profiles could be obtained. A thin oxide film was found on the sidewalls of the photoresist and poly-Si after etching. The thickness of the oxide film increased with etching time. No silicon was found on top of the photoresist. The silicon oxide film deposited on the surfaces resulted from the sputtering of the quartz tube located in the plasma generation region. Substantial amounts of chlorine were found on the tops, sidewalls and bottoms of the features. The chlorine concentrations on the sidewalls of the features etched at high bias power were comparable to those found on the horizontal surfaces when a low bias power was applied. The chlorine concentrations on the tops and bottoms of the features increased with increasing bias power. A low coverage of oxygen was found on all surfaces. The oxygen concentration on the sidewalls of the features also increased with etching time. Less oxygen was found on the bottom of the poly-Si trenches. Small amounts of carbon were detected on the bottoms of the features. A substantial coverage of carbon was present on the poly-Si sidewalls and was about half that found on the tops and sidewalls of the photoresist. These results indicate that the etching of the photoresist and the erosion of the quartz tube may enhance anisotropic etching.

Einfluß der Übertragungseigenschaften von Aufkohlungsatmosphären auf die Aufkohlungsschicht

Carburizing in gas atmospheres generated in situ by introducing a fuel-air mixture into the furnace, results in uniform and reproducible carburized layers with reduced internal oxidation. The reasons for this gas properties are investigated. The dominating influencing factor is the adsorped oxygen layer on the surface of the workpieces. Its effect is explained. Moreover, steps are demonstrated, which result in a low oxygen coverage. Also a method is proposed for the calculation of the carbon transfer rate for different degrees of oxygen coverage. Further, the mechanism of reaction is investigated, which makes it possible to get equal carburizing results with loads of different sizes without any external interference.

Einfluß der Übertragungseigenschaften von Aufkohlungsatmosphären auf die Aufkohlungsschicht

Carburizing in gas atmospheres generated in situ by introducing a fuel-air mixture into the furnace, results in uniform and reproducible carburized layers with reduced internal oxidation. The reasons for this gas properties are investigated. The dominating influencing factor is the adsorped oxygen layer on the surface of the workpieces. Its effect is explained. Moreover, steps are demonstrated, which result in a low oxygen coverage. Also a method is proposed for the calculation of the carbon transfer rate for different degrees of oxygen coverage. Further, the mechanism of reaction is investigated, which makes it possible to get equal carburizing results with loads of different sizes without any external interference.

Oxygen enhanced secondary ion emission of Fe and Co by TOF-SIMS and ISS/DR

Using secondary ion mass spectrometry (SIMS), ion scattering spectroscopy (ISS), direct recoil (DR) and X-ray photoelectron spectroscopy (XPS), we have studied the mechanism of oxygen enhancement of secondary ion emission. The secondary ion intensities were measured as a function of oxygen exposure for Fe and Co samples under static- and dynamic-SIMS conditions. In the static-SIMS, the intensity of Fe+ increased up to 60 L oxygen exposure and then slowly saturated. However, the intensity of Co+ had an additional peak at low oxygen coverage when compared with the result of Fe+. It sharply increased up to 5 L, dropping off around 20 L and increased again to saturation. Also the intensities of Fe+ and Co+ under dynamic-SIMS conditions showed similar changes to those under static-SIMS conditions. According to XPS spectra, the polycrystalline Fe surface was oxidized immediately on oxygen exposure. However the oxygen on the Co surface was changed from the chemisorbed state at low oxygen doses (0–5 L) to the oxide state at oxygen doses above 5 L. The relative oxygen coverages on Fe and Co samples estimated by DR and ISS intensities showed trends similar to the secondary ion intensities of Fe+ and Co+ as a function of oxygen dose. We applied the bond breaking model and the electron tunneling model for interpretation of the oxygen enhancement effect in the positive secondary ion emission for Fe and Co.

Kinetics of oxygen dissociation on Si(111)7 x 7 investigated with optical second-harmonic generation.

Optical second-harmonic generation is used as a sensitive, nondestructive probe for the thermally activated dissociation of molecular oxygen on Si(111)7\ifmmode\times\else\texttimes\fi{}7. Isothermal measurements show that the dissociation kinetics strongly depend on the total oxygen coverage. At room temperature the 1/e lifetime of chemisorbed ${\mathrm{O}}_{2}$ is \ensuremath{\sim}15 min for saturation coverage but \ensuremath{\sim}50 min for low oxygen coverage. The temperature dependence of the dissociation rates between 80 and 500 K indicates that more than one dissociation channel and/or molecular oxygen species exist on the surface.

The nature of chemisorbed hydroxyl radicals

The energetics of hydroxyl adsorption on Pt, Pd, Rh and Ni have been investigated using the bond order conservation-Morse potential approach (BOC-MP). The heat of adsorption of hydroxyl (QOH) was calculated both under high oxygen coverage and oxygen depleted conditions. Adsorbate-adsorbate interactions must be taken into account in order to explain the observed variation of activation energies for OH desorption reported in the literature for each metal and under different experimental conditions. High oxygen coverages lower the heat of hydroxyl adsorption, while at low oxygen coverages island formation between hydroxyls and hydrogen bonding explain the increase in the heat of adsorption of hydroxyl radicals. The calculated heats of adsorption for the different conditions are in quantitative agreement with experimental results.

Oxidation of the GaAs(110) surface

The principal absorption site of oxygen on the GaAs(110) surface has been deduced from scanning tunneling microscopy (STM) measurements (for low oxygen coverages of the surface) to be in the interchain trough and above the surface. In contrast, photoemission measurements of Ga and As core-level shifts and of oxygen-derived energy levels led to the conclusion that absorption occurs on or below the surface (for high oxygen coverages), with Ga and As atoms about equally oxidised. To understand these observations, we have performed total-energy calculations using the first-principles pseudofunction method for a model with a periodic half-monolayer of oxygen, adsorbed to a few selected sites in the surface unit cell. Our results, although computed for models with half-monolayer coverages of oxygen, strongly suggest that the oxygen image observed in STM measurements indeed is due to an oxygen atom in the interchain trough, but with the oxygen atom located slightly below the surface and bonded to a Ga atom. For high coverages, our results imply in-surface-plane multi-coordinated adsorption of oxygen, which accounts for the observed oxygen-derived spectral features about 5 eV below the valence band maximum.

HREELS study of subsurface oxygen on the Ni(100) surface

New high frequency vibrational loss features have been measured on O/Ni(100) for low oxygen coverages, using high-resolution electron-energy-loss spectroscopy (HREELS). These high frequency losses only appear after thermal annealing. Oxygen atoms are migrating underneath the surface with increasing temperature, giving rise to the high frequency modes. The modes are shown to result from subsurface oxygen atoms, squeezed in interstitial sites. We have detected five different high frequency loss signals, while five different interstitial sites are precisely available below the Ni(100) surface. Model calculations have been performed in order to relate frequencies to interstitial sites. The matching technique approach has been applied to analyze the case of periodic structures, whereas a cluster approach has been used to obtain the localized vibrational states associated with interstitial oxygen atoms.

Reaction pathways in the oxydehydrogenation of ammonia at Cu(110) surfaces

The activation of ammonia by oxygen at Cu(110) has been investigated by X-ray photoelectron and electron energy loss spectroscopies. The chemistry observed is dependent on the temperature, whether oxygen is preadsorbed and its surface coverage, or whether the oxygen is coadsorbed with ammonia. Amide species NH 2(a) are formed only when adsorbed ammonia is exposed to dioxygen at low temperatures. With increasing temperature further step-wise dehydrogenation occurs to give imide NH(a) and nitrogen adatoms N(a). For an ammonia-rich dioxygen-ammonia mixture a facile reaction to form exclusively bent imide species occurs at 295 K with no evidence for chemisorbed oxygen being present until θ NH approaches unity. A hot transient O −(s) species is implicated in the reaction mechanism. On the other hand for θ oxygen → 1.0 the oxygen overlayer is relatively unreactive, imide formation being kinetically slow and limited in extent. Furthermore there is no evidence in the HREEL spectra for a loss peak characteristic of δ nh although a ν NH loss peak is present. This suggests a linear form of NH(a) in contrast to the bent form generated by coadsorption of ammonia and dioxygen. Two different oxygen species can exist at the copper surface: one that is highly reactive to ammonia and undergoes chemisorptive replacement, the other inactive. We suggest that the former is O −-like and associated with isolated oxygen atoms and the latter O 2−-like and associated with multi-oxygen atom copper nuclei. High catalytic oxydehydrogenation activity can be maintained during the coadsorption of dioxygen and ammonia, provided the development of O 2− species (oxide growth associated with surface reconstruction) is suppressed. The latter has been shown to occur even at low oxygen coverages ( θ ≈ 0.1) the ammonia molecule acting as a sensitive and specific probe for the isolated O −-like species. The O −(s) species are therefore transients in the development of the chemisorbed oxygen overlayer and characterised by high chemical reactivity. Support for this model comes from recent scanning tunnelling microscope studies of the Al(111)-oxygen system of Ertl and coworkers [Phys. Rev. Lett. 68 (1992) 624] (ref. [1]).

Energy partitioning in the reaction 2H2+ O2? 2H2O on Pd(111)

(2 + 1)-Photon resonance-enhanced multiphoton ionization (REMPI) has been used to detect desorbed H2O molecules in a state-specific manner. We have studied the reaction on Pd(111) at temperatures between 500 and 700 K in the regime of low oxygen coverages ( <0.05 ML) resulting from adsorption of ambient O2. D2 is dosed using a supersonic molecular beam with pulse fluxes resulting in stoichiometric coverages of D and O atoms. We find that D2O molecules are desorbed with a translational energy of ca. 300 K with weak, if any, dependence on the surface temperature. The angular distribution is nearly cosine in form. The spectroscopic data indicate that the D2O is desorbed with an average energy in the rotational degree of freedom close to the surface temperature. Experiments probing vibrationally excited states yielded negative results setting an upper limit for their population. These data show that most of the excess energy of the final OD + D reaction step of ca. 100 kJ mol–1 is not channelled into any of the degrees of freedom of the desorbed molecules. The data are interpreted such that the product water molecules reside on the surface for a time sufficient to dissipate most of the energy.

Oxygen on Mo(110): low-temperature adsorption and high-temperature oxidation

High-resolution electron energy loss spectroscopy (HREELS), Auger electron spectroscopy (AES) and thermal desorption mass spectrometry (TDS) have been used to study the low-temperature and high-temperature oxidation of a Mo(110) single crystal. Oxidized overlayers prepared at 90 and 1000 K were studied at temperatures up to 1400 K. A low-oxygen coverage on Mo(110) at 90 K produces only one vibrational loss at 530 cm −1, indicating dissociative oxygen adsorption into the long-bridge site on this surface. Upon increased exposure, oxygen undergoes conversion to a quasi-threefold adsorption site, causing two characteristic losses at 395 and 605 cm −1 to become spectroscopically visible. An additional loss at 990 cm −1 appears at saturation oxygen exposure. An overlayer prepared by exposure of Mo(110) at 90 K to a low coverage of oxygen followed by heating to 1200 K regenerates a “clean” surface, as observed in HREELS; this is caused by oxygen diffusion into the subsurface region. Heating an oxygen-saturated overlayer to 1200 K regenerates a vibrational spectrum that is similar to a low coverage of oxygen on Mo(110). Heavy oxidation of Mo(110) at 1000 K produces an oxide layer characterized by the appearance of a fourth vibrational mode at 740 cm −1 and a twofold increase in the O(KLL)/Mo(LMM) Auger peak-to-peak ratio. Heating this O/Mo(110) layer to 1400 K regenerates a surface that is similar to the oxidized molybdenum overlayer produced by a saturation-oxygen exposure at 90 K.