XPS Intensity Research Articles

XPS Studies on Ce/Al2O3 and on CoRh/Nb2O5 Catalysts

AbstractThe surface properties of supported catalysts are studied. XPS data are used to identify the different surface species but, alone, it is not sufficient to provide enough information to describe the correct form and distribution of these species and phases on the surface. However, XPS supported by an accurate mathematical formalism to model catalysts surface structure, may become a powerful tool to help understanding surface properties. First, a new mathematical model for supported bimetallic Co‐Rh/Nb2O5 catalysts is proposed based on quantitative analysis of XPS intensities assisted only by TPR data. XPS peak intensities are compared to the values predicted from different structural models of the oxide particles. Secondly, the Ce/Al2O3 system generated from different precursors of cerium salts and contents is examined using different techniques. XPS results show the effect of the precursor on the Ce/Al ratio while TEM the morphology. In addition, the EPR data show the Ce3+ species distributed in regions underneath the topmost surface layer. Then, an improved model is proposed based on quantitative analysis of XPS intensities, XRD, and TEM results.

XPS and static SIMS studies of copoly(ether–esters) containing mixed polyether soft blocks

AbstractThe surface compositions for a series of copoly(ether–esters) are determined from XPS and static SIMS measurements. The copolymers consist of a polybutylene terephthalate (PBT) hard block and a mixed soft block of two of the following polyethers—poly(propylene glycol) (PPG), poly(tetramethylene) oxide (PTMO) and poly(ethylene oxide) (PEO). The surfaces of all the copolymers are enriched in the total polyether soft blocks. Quantification of XPS intensities shows the presence of both polyethers at the surface of these mixed soft‐block copolymers, in relative amounts close to the bulk formulations. In the static SIMS measurements, direct observation of characteristic fragment ions related to the specific polyethers confirms their presence and relative abundance in the surface.

New aspects in Volmer-Weber 3D growth: an XPS intensity study applied to thin films of Au and Ce on polypropylene

A new method of analyzing X-ray photoelectron spectroscopy or Auger electron spectroscopy intensity data is proposed, which allows an easy distinction between the established growth modes (Frank-van der Merwe, Stranski-Krastanov, Volmer-Weber) of thin films. The method, applied to thin films of Au or Ce on polypropylene (PP), reveals detailed information about the growth mode. For Ce one observes Volmer-Weber growth at fixed substrate coverage, while for the less reactive gold, the film growth is island-like. The different chemical reactivity of the two metals towards PP is the probable cause for the film growth mode.

Features of Li-Mn-MgO Catalysts and Their Relevance in the Oxidative Coupling of Methane

This work describes relevant structural features of the ternary Li-Mn-MgO system useful for the reaction of methane oxidative coupling. For a Mn-MgO catalyst, XRD, XPS, and ESR techniques revealed that manganese is incorporated into a Mg 6MnO 8 phase in which the formal oxidation state of manganese is Mn 4+. Addition of Li 2CO 3 in small amounts (0.48-1.60 wt.% Li) to the base Mn-MgO catalyst results in partial destruction of the Mg 6MnO 8 phase. TPR and XPS data showed that the reduction of manganese is easier in the bulk Mg 6MnO 8 compound than in the Mn-MgO catalyst and that it is inhibited by lithium. Moreover, the increase in Mn/Mg XPS intensity ratios in the lithium-doped counterparts indicates that lithium incorporation into the base Mn-MgO catalyst enhances manganese dispersion, in agreement with the increase of Mn 4+-Mn 4+ distance derived from ESR spectra.

XPS quantitative evaluation of the overlayer/support intensity ratio in particulate systems

The XPS intensity ratios M M/ I Zr were measured for supported MO x /ZrO 2 systems (where M is Cr, Mo, Na, K, Re, designated as the “promoter”) and compared with the intensity values calculated for different models. When the support is the high surface area (about 300 m 2 g −1) hydrous zirconia (HZ) treated at 383 K, the Kerkhof and Moulijn model fits the data very well. If the zirconia support is heated at T between 773–923 K before promoter adsorption, the observed intensity ratios are roughly twice as high as those predicted by a non-attenuating overlayer on a semi-infinite support (referred to as “planar”) model. To achieve a correct fit for the experimental data, the contribution of ejected electrons from low angles must be taken into account. A model has accordingly been tested, which can fit the observed intensity ratios independently of the nature of the promoter.

Effects of elastic and inelastic electron scattering on quantitative surface analyses by AES and XPS

A review is given that describes the complications due to elastic and inelastic electron scattering in quantitative surface analyses by Auger-electron spectroscopy and x-ray photoelectron spectroscopy. Four principal topics are addressed. First, the simple formulae for surface analyses are based on a model that ignores elastic scattering. Recent work assessing the effects of elastic scattering is summarized which shows that the simple formulae are valid in certain analytical situations but with an appropriate choice of the parameter describing inelastic scattering. Second, we review measurements of effective attenuation lengths and point out many sources of significant systematic error in these measurements. Third, we describe recent calculations of inelastic mean free paths (IMFPs) in over fifty materials that have been utilized to develop a predictive IMFP formula. Finally, we discuss the complicating effects of inelastic scattering on reliable measurements of AES and XPS intensities.

Adsorption of ethylene oxide on Ni(110): structure, thermal and radiation-induced decomposition

The adsorption and decomposition of ethylene oxide on Ni(110) at 95–350 K was investigated by LEED and mainly X-ray and UV photoelectron spectroscopy. Thermal as well as photon-induced decomposition were discovered, with X-rays being about 10 times more efficient than UV photons under the chosen experimental conditions. Photon-induced desorption of ethylene oxide or of the decomposition products was negligible. The decomposition itself was characterized by changes in valence band as well as O Is and C Is core level spectra. A possible reaction pathway involves intermediates, such as methoxy, CH x and the final (at > 300 K) species CO and carbon. A further possible intermediate, adsorbed acetaldehyde, appears to be short-lived on Ni(110), as indicated by auxiliary measurements with that species adsorbed at 95–150 K. The orientation of ethylene oxide at 95–120 K is determined by angle-resolved core level XPS intensity patterns, via the detection of O Is or C Is intensity enhancements. A tilted adsorption geometry is found, with the symmetry axis of the molecule tilted towards the [001] or [001̄] direction by 48°. The orientation is symmetric with respect to the [001] azimuth.

Structural and catalytic properties of several potassium-doped nickel/α-alumina solids

Studies of the chemical preparation, activation energies of reduction, temperature-programmed reduction (TPR), X-ray diffraction (XRD), X-ray photoelectron (XP) spectra and catalytic activities of several potassium-doped nickel/α-alumina catalysts have been carried out for the catalytic hydrogenation of hexanedinitrile, in a continuous process at 1 atm pressure, 443 K, and in the absence of ammonia. Activation energies of reduction for the potassium-doped non-stoichiometric NiO/α-alumina precursors were higher than those obtained for the potassium-free NiO/α-alumina and for the unsupported non-stoichiometric NiO. TPR results show that there is an inhibiting effect of α-alumina and potassium on NiO reduction. XRD measurements reveal the presence of α-alumina, NiO and Ni phases, the latter being present in larger quantities at higher reduction temperatures and lower potassium contents. XRD spectra also show fairly constant nickel particle sizes for all catalysts tested, XPS results confirm the inhibiting effect of potassium upon the reducibility of NiO. The XPS intensity ratios of surface Ni : NiO match the BET area ratios Ni : NiO of the samples. Conversions of the catalysed reactions decrease with increasing potassium content of the catalysts, probably because of poisoning of the latter with a Thorpe cycling formation of an enamine at high basic K2O contents. Selectivities towards 6-aminohexanenitrile (which may reach 100% at 85% conversion, thereby of potential interest for the manufacture of nylon-6,6) increase with potassium content of the catalysts. The less selective catalysts also yield 1,6-diaminohexane and azacycloheptane. A lack of oligomerization reactions is observed, and a mechanism is proposed.

Quantitative XPS of NiO, CoO and MnO. The effects of elastic and inelastic electron scattering

In order to check whether the method for background correction developed by Tougaard can be applied to oxides we measured XPS spectra of single crystalline NiO, MnO and CoO samples. The materials for which background treatment was developed and tested originally were metals and alloys. In contrast to that, the oxides we study in the present work are insulators. We have chosen these compounds because their structure is simple and by cleaving them in vacuo it is possible to obtain surfaces with well-defined stoichiometry. The inelastic background for these oxides is likewise described well by the above mentioned model. The ratio of metal to oxygen concentration calculated from the XPS intensities so obtained oscillates around 1.0 with a maximum deviation of about ±20% in the worst case. These deviations originate from elastic electron scattering, namely forward focusing (X-ray photoelectron diffraction). Taking this into account, the perfect stoichiometry is obtained within a few percent.

Fine structure of supported and promoted catalysts as deduced from XPS intensity data

AbstractFor model calculations it is assumed that metal spheroids are randomly distributed on spherical support particles of nearly closest‐packed arrangement. Both the metal and the support can be covered with layers or submonolayers of promoter(s) and carbon. The average diameters as well as n, the average number of metal particles on one support sphere, are estimated from total surface area and selective chemisorption measurements and preferably by electron microscopy too. In the computer programme two cases are distinguished depending on whether n < 4 or n > 4, since in the latter case the partial shadowing of coated metal particles by others should also be considered. Electron beams emitted by the metal, the support and their layers are calculated as a function of the ‘angle of latitude’ of the support spheres and the ‘angle of latitude and longitude’ of the metal spheroids. The computation variables are the thickness and sequence of different layers above the metal and the support, the extent of ‘immersion’ of metal particles into the support and/or its layers and the extent of diffusion of promotor cations into the support matrix. The calculated intensities are compared with the measured ones. By this method the probable fine structure of an alkali‐promoted Pd/SiO2 (n < 4) and a sodium‐promoted Ni/(Al2O3 + SiO2) catalyst (n > 4) has been determined.

Evaluation of take‐off‐angle‐dependent XPS for determining the thickness of passivation layers on aluminium and silicon

AbstractInvestigating the effect of a cleaner (used to remove lubrication oil residues) on the surface composition of rolled aluminium foils, we applied angle‐dependent XPS to determine the thickness of the passivation layer on these foils. We found the simple uniform overlayer model, which has frequently been applied to translate XPS intensity ratios into a value for the overlayer thickness, to be inapplicable on these aluminium foils. However, data obtained from an Si(100) single‐crystal surface were in rather good agreement with the model. SEM images from the aluminium foils show that their surface is rough on the (sub)micron scale. We believe this roughness to be the reason for the inapplicability of the uniform overlayer model. Simulations of XPS intensity ratios for model rough surfaces support this point of view. Both experiment and simulations make it clear that one should not base XPS thickness determinations on one measurement only, because in that case the applicability of the uniform overlayer model cannot be checked.

Quantitative x-ray photoelectron spectroscopy analysis of alkali ions / zirconia systems

Zirconia-supported alkali ions, containing Na, from 0.74 to 3.61 wt.%, and K, from 0.91 to 4.76 wt.%, have been studied by X-Ray Photoelectron Spectroscopy, X-ray diffraction and chemical analyses. The results show that the alkali ions remain dispersed on the zirconia surface. After treatment at 773 K in air, 5h, the Na containing samples maintain a good dispersion at low Na content, whereas for the more concentrated samples, agglomeration and incorporation in solid solution have been detected. The K based system shows neither agglomeration nor incorporation in solid solution of potassium ions. The XPS intensity ratios are compared with calculated values, according to different models, to account for the changes in surface morphology.

Determination of substrate and overlayer intensity ratios for metal-metal systems in AES and XPS by a crystallographic electron attenuation model

A model has been presented which gives AES and XPS intensity ratios for metal/metal systems. Three interface constants, a substrate escape function, a substrate maximum-electronic-escape-depth and an overlayer maximum-electronic-escape-depth are introduced to characterize the interface. The normalized substrate and overlayer intensity ratios initially depend simple-exponentially on coverage, but from a certain coverage the dependence is not simple exponential. It is suggested, that this breaking point is not necessarily associated with the point of monolayer coverage, and a method to determine the monolayer position is given. No further breaking points are expected. The exponential decay constant is to good accuracy equal to the inverse overlayer attenuation length for coverages below one monolayer. It is shown, that the sum of normalized substrate and overlayer intensity ratios may be constant over a wide range of coverages.

Quantitative analysis of discontinuous silver films with XPS and TEM

Mathematical formulae of angle-dependent integral XPS intensities based on simple geometrical models of metal particles grown upon a suitable substrate were derived. Measured diameters of silver particles deposited on a-C substrate were substituted into these formulae and XPS intensities calculated. The morphological data were obtained from TEM micrographs and measured by image analyzer. Calculated angle-dependent intensities were then compared with the measured ones. The differences between calculated and measured data are explained by a combined influence of several factors: invisibility of a certain amount of the deposited material in TEM, deviations from the hemispherical particle shape, and surface roughness of the substrate.

Chemical element of the outermost layer of AgBr single crystal

Surface elements of octahedral microcrystals of AgBr grown in an emulsion were investigated by comparing signal yields of Auger and X-ray photoelectron spectroscopy with those of cubic microcrystals. Signal yields of surface-sensitive MNN Auger electrons of bromide ions were weak, and it was not possible to identify the surface elements from either the Auger or XPS intensity ratios of bromide to silver ions. However, evidence that bromide ions occupy the surface layer of the octahedral crystals was obtained by comparing he Auger signal intensities of oxygen from adsorbed CO or H 2O molecules on the two types of crystals.

Depth profiling of catalyst samples: An XPS-based model for the sputtering behavior of powder materials

A model is presented to describe the depth profiles obtained by ion-bombarding polycrystalline multicomponent samples. Different calculations for idealized catalyst structures illustrate the possibilities of using the ion-sputtering technique for the characterization of catalysts. These calculations show that the depth profiles obtained using dispersed phase XPS intensities may provide semi-quantitative information on particle size, existence of solid solution, or formation of mono- or bimetallic particles in bimetallic catalysts. Examples with the dispersed phase in metallic or oxide forms that illustrate these applications are presented.

Structure and catalytic activity of alumina-supported platinum-cobalt bimetallic catalysts. 1. Characterization by x-ray photoelectron spectroscopy

A series of Pt{sub 1{minus}x}Co{sub x}/Al{sub 2}O{sub 3} bimetallic catalysts has been studied by x-ray photoelectron spectroscopy following in situ hydrogen reduction of the samples previously calcined in oxygen at 570 K. Platinum was found to be in zerovalent state for the entire composition range and its dispersion increased with decreasing platinum loading. When combined with platinum, cobalt stayed partly in a hardly reducible cobalt oxide surface phase (denoted by CSP) and partly in the metallic state, whereas without platinum CoO was the predominant form. At low cobalt concentration the variation of the XPS intensity changes of the cobalt 2p{sub 3/2} line in the different states was interpreted by assuming that the cobalt surface phase was being covered by platinum. On the other hand, at high cobalt contents saturation of the alumina surface by the cobalt surface phase was concluded from the occurrence of a constant, high amount of cobalt in the surface phase. On the basis of the comparison of surface platinum and cobalt contents first platinum-like property and then cobalt-like is predicted for these catalysts with increasing cobalt content.

DV-Xα calculation on XPS for SF 6 valence levels

We have performed molecular orbital calculations for SF 6, using the discrete-variational Xα method. The obtained theoretical XPS spectra for valence levels agree well in intensity with the experimental one except for two levels which have strong covalent interactions. Two types of conventional approximations for the XPS intensity are compared and give similar spectra. The present results confirm the assignment of all the valence levels proposed by Dehmer et al. and Gelius.

Restoration of the depth–concentration profile from the angle‐resolved relative intensities of x‐ray photoelectron spectra

AbstractA numerical method for the restoration of the depth–concentration profile from the angular distribution of XPS intensities is described. The proposed method is based on the regularization theory. A simpler formulation of the basic relationships than has been given in previous works2–8 enables the inclusion of more general a priori information. A new approach is also given for the choice of the most suitable profile for the experimental data, which enables restoration of profiles with up to two extremes. Calculations performed for all typical profiles show good agreement with the simulated data. The applicability of the method is demonstrated on the depth–concentration profile restoration of the SiO2/Si interface in the silicon surface barrier detector.

Depth profiling of elements in surface layers of solids based on angular resolved X-ray photoelectron spectroscopy

A numerical method is proposed for the non-destructive restoration of the depth—concentration profiles for four components, based on the angular distribution of XPS intensities. Calculations performed for a number of model systems demonstrated that ratios of concentrations obtained by using simplified procedures, i.e. from relative XPS intensities for a single escaping angle, led to large errors (up to 200%) for the upper surface layer. A number of restored profiles are presented using experimental angle-resolved XPS data for samples containing 3–4 components.