Step Edge Sites Research Articles

Adsorbate induced kink formation in straight step edges on Cu(533) and Cu(221)

Scanning tunneling microscopy (STM) and temperature programmed desorption (TPD) measurements have revealed that the straight step edges of Cu(5 3 3) surfaces are reconstructed by the room temperature adsorption of R-3methylcyclohexanone (R-3-MCHO). STM and low energy electron diffraction (LEED) investigations have shown that the clean Cu(5 3 3) surface is composed of arrays of narrow (1 1 1) terraces separated by straight (1 0 0) steps. During room temperature adsorption of R-3-MCHO, the (1 0 0) step edges on the Cu(5 3 3) surface reconstruct, adopting a zigzag structure that exposes kinked, high Miller index step edges. The density of step edge and kink adsorption sites on Cu surfaces can be titrated by TPD of R-3-MCHO. TPD experiments also reveal the formation of kinks on the Cu(5 3 3) surface during room temperature adsorption of R-3-MCHO. Furthermore, TPD experiments on the Cu(2 2 1) surface having straight (1 1 0) step edges indicate that these also reconstruct to expose kinked step edges during R-3MCHO adsorption at room temperature. Although the kinks on such surfaces are chiral, the R- and S-forms appear to be produced in a roughly racemic (equimolar) ratio. Thus, although the reconstruction of the initially achiral Cu surfaces in the presence of a chiral adsorbate could, in principle, lead to a surface with a net chiral structure, R-3MCHO does not seem to be an effective chiral imprinting agent. � 2004 Elsevier B.V. All rights reserved.

Energetics of surface atomic processes and effects on growth of ultrathin films and nanoislands

Real surfaces of a crystal, especially those of nanometer in size, have many steps and all types of lattice imperfections. Moreover, there are atoms of different atomic environments: adatoms sit at a flat terrace, kink-site atoms sit in kink sites, and step-edge atoms of various coordination numbers sit in different step-edge sites. Using field-ion and atom-probe field-ion microscopy, it is possible to measure site specific binding energies of these surface atoms, and determine the mechanisms and activation energies of the various elementary surface atomic processes, i.e. surface diffusion. For kink-site atoms, our measurements with several metals find the binding energy to be equal to the cohesive energy of these metals. Atoms at other sites have slightly smaller binding energies. However, they are not proportional to the coordination numbers of the atoms, as assumed in many calculations. Such information will not only shed light on the binding mechanisms of solids, but is also needed for understanding the mechanisms of growth phenomena, such as crystal growth and epitaxial growth of thin films and nanoislands.

Energetics of surface atomic processes and effects on growth of ultrathin films and nanoislands

Real surfaces of a crystal, especially those of nanometer in size, have many steps and all types of lattice imperfections. Moreover, there are atoms of different atomic environments: adatoms sit at a flat terrace, kink-site atoms sit in kink sites, and step-edge atoms of various coordination numbers sit in different step-edge sites. Using field-ion and atom-probe field-ion microscopy, it is possible to measure site specific binding energies of these surface atoms, and determine the mechanisms and activation energies of the various elementary surface atomic processes, i.e. surface diffusion. For kink-site atoms, our measurements with several metals find the binding energy to be equal to the cohesive energy of these metals. Atoms at other sites have slightly smaller binding energies. However, they are not proportional to the coordination numbers of the atoms, as assumed in many calculations. Such information will not only shed light on the binding mechanisms of solids, but is also needed for understanding the mechanisms of growth phenomena, such as crystal growth and epitaxial growth of thin films and nanoislands.

Prefactors for interlayer diffusion on Ag/Ag(111)

Interlayer diffusion controls the transfer of atoms between layers in a film and isthe key factor determining whether growth is layer-by-layer or two-dimensional.Analysis of recent experimental data on Ag/Ag(111) taken at two sets oftemperatures provides similar values for the step edge barrier ΔEsbut not for the ratio of the prefactors νs/νt where νtis the prefactor for diffusion on a terrace and νs the prefactorat a step edge site. A prefactor ratio larger than 1, νs/νt > 1,is extracted from the measurements at low temperature (T < 150 K),while νs/νt ∼ 1 isfound in the experiments done at a higher temperature (T ∼ 300 K).Since the conclusion νs/νt ∼ 1 isbased on a steady state analysis of island decay, we examine whether thiscondition is fulfilled for Ag/Ag(111). We show that the low value of the terracediffusion barrier on Ag(111) (Et ≈ 0.1 eV)offers an alternative interpretation to the steady state analysis, i.e.an independent detachment model. Alternatively, we propose anindependent detachment model. Re-analysis of the data in terms ofthis model results also in νs/νt > 1, ingood agreement with the low temperature experiments.

Structure determination of Cu( [formula omitted])–O using X-ray diffraction and DFT calculations

The Cu(4 1 0)–O surface, involving a 0.5 monolayer (ML) coverage of oxygen, is known to be extremely stable and a range of Cu(1 0 0) vicinal surfaces facet to (4 1 0) in the presence of adsorbed oxygen. A new surface X-ray diffraction investigation of this surface has been conducted to determine its structure, and the detailed structural parameter values obtained are compared with the results of a density functional theory (DFT) calculation. The results show that the metal structure is unreconstructed, with the oxygen forming an overlayer with 0.25 ML O atoms at near-colinear step-edge sites and 0.25 ML O atoms at mid-terrace hollow sites, approximately 0.6 Å above the terraces. The large number of independent structural parameters potentially relevant to this vicinal surface presents a significant challenge for unique structural optimisation, but various missing row reconstruction models can be clearly excluded. Two detailed structural solutions are identified which give equally acceptable fits to the X-ray diffraction data after imposition of a Lennard-Jones penalty factor. These models differ especially in the O positions, but one is found to be more favoured by comparison with the results of the DFT calculations, and by considerations based on bond lengths and valence. Substantial relaxations from the bulk metal positions occur for the outermost Cu atoms; the ability of the vicinal surface to relax in this way may help to account for its stability compared with the missing row reconstruction induced by oxygen chemisorption on the Cu(1 0 0) surface.

Decay of pyramidal nano-island formed on Si( [formula omitted]) studied by kinetic Monte Carlo simulation

By using kinetic Monte Carlo simulation based on a modified solid-on-solid model, the decay process of pyramidal nanostructure island formed on Si(1 0 0) surface is studied. The decay proceeds primarily due to mass transport of atoms on the surface from a kink or a step edge site of the island to another step edge in the lower layer. The basic process is the same for both two- and three-dimensional (3D) islands. The decay rate of the 3D island is, however, larger than that of the two-dimensional (2D) island by a factor of 2–4. The difference is attributed to different distance from the starting step edge to the terminating one. The diffusion length in the case of 2D island decay is much smaller than that estimated by the terrace length of the underlying layer. The diffusion occurs through atom vacancies formed on the terrace in the underlying layer. The vacancy-mediated diffusion is an important process for the decay of 2D island.

Computed energetics for etching of the Si( [formula omitted]) surface by F and Cl atoms

Energetics and limited reaction pathways have been computed using density functional theory with the B3LYP functional for the reaction of F and Cl atoms with the Si(1 0 0) surface. The energetics are in good agreement with the earlier ab initio calculations of Wu and Carter for the bonding of 1–3 F atoms to a surface dimer. The present energetics are also in good agreement with the results of Srivastava et al. using the Stillinger–Weber potential. We find no barrier for adding a third F atom to a monofluorinated dimer. A significant new result is a low energy pathway to SiF 4, which has not been previously characterized. We find that Cl differs from F in having smaller binding energies and much larger non-bonded repulsions. Adding a third Cl to a monochlorinated dimer leads to a barrier of about 0.6 eV (for a symmetry constrained search) and step edge sites are strongly favored in the Cl case, but not in the F case. The energetics obtained for Cl on Si(1 0 0) are in good agreement with results obtained by Halicioglu and coworkers using the Feil potential.

Oxygen driven reconstruction dynamics of Ni(977) measured by time-lapse scanning tunneling microscopy

Time-lapse scanning tunneling microscopy (STM) has been used to observe the oxygen induced reconstruction behavior of Ni(977), a stepped metallic surface. Previous studies using helium atom diffraction resolved the macroscopic kinetics for the reversible step-doubling and -singling of this vicinal surface. Sequential STM imaging recorded at elevated temperature has now elucidated atomic-level mechanistic details for the merging of steps in the presence of small amounts of adsorbed oxygen, less than 2% of a monolayer. Point contact between neighboring steps decorated with chemisorbed oxygen facilitates rapid step coalescence by means of zippering. An optimal oxygen concentration of step edge saturation was found to enable the step merging to proceed most rapidly. Excess oxygen was found to hinder the coalescence of neighboring steps through the possible growth of overlayer structures on the terraces. At sufficiently high temperatures, the surface is driven back to single steps due to oxygen dissolution. The departure of oxygen from the surface through dissolution, as well as the associated presence of oxygen in the selvedge region, may both play a role in destabilizing the double steps. Local step density influences the coalescence behavior by defining the number of available step edge sites. The microscopic details made available by time-resolved STM imaging illuminate some of the mechanistic steps related to the initial stages of metallic oxidation, and the sensitivity of surface morphological transformations to local surface structure and adsorbate coverage.

Step edges on galena (100): Probing the basis for defect driven surface reactivity at the atomic scale

Scanning tunneling microscopy (STM) images of the PbS (100) surface with a step and several kinks were obtained with atomic resolution. These images show an increased tunneling current at step edge sites and an apparent deformation of the lattice near the step. The experimental images are compared with theoretical ab initio calculations for which we developed a hybrid method of constant current and constant height mode STM image simulation. With these calculations, we find that the apparent deformation is mainly an electronic effect rather than relaxation of atoms. In addition, with the help of these calculations, we can identify the changes of individual terrace-like and step-like orbitals that are observed using the STM in terms of the energy, density and shape of these states. This detailed knowledge of the electronic behavior of the PbS surface near a step can be used as a basis for explaining adsorption, acid/base, and redox behavior on PbS terraces and at steps, and the differences between the two.

Combined Computational and Experimental Investigation of SOx Adsorption on MgO

One of the main obstacles to the use of alkaline-earth-oxide-based NOx adsorbents for emissions control is the ready poisoning of these materials by SOx. To shed light on the mechanisms of poisoning, density functional theory (DFT) calculations and infrared spectroscopy experiments are used to study SOx adsorption on MgO as a model alkaline earth oxide. DFT plane-wave, pseudopotential results are presented for SOx adsorption on MgO(001) terraces and steps. Both SO2 and SO3 are found to exhibit both weakly bound physisorbed and strongly bound chemisorbed forms, with minimal activation barriers separating the two. Chemisorption is dominated by interactions between Lewis acidic sulfur and Lewis basic oxide anion sites to form surface sulfites and sulfates from SO2 and SO3, respectively. Within the generalized gradient approximation to DFT, the SO2 adsorption energy ranges from 25 kcal mol-1 on terrace sites to 46 kcal mol-1 on lower-coordinated step edge sites; analogous SO3 adsorption energies range from 49 to 76 kcal mol-1. Surface sulfite is readily observed upon exposure of calcined MgO powder to SO2 and produces vibrational signatures consistent with that calculated for chemisorbed SO2. Sulfation is accomplished by exposure of the same powders to SO2 and O2 at elevated temperatures. Comparisons to calculated vibrational spectra indicate that both chemisorbed SO2 and SO3 are present under these conditions; further heating induces formation of a bulklike sulfate that is difficult to remove. The results demonstrate the high reactivity of SOx toward metal oxides and illustrate the challenges in developing sulfur-resistant oxide-based NOx traps.

Surface defects and self-diffusion on pyrite {100}: An ultra-high vacuum scanning tunneling microscopy and theoretical modeling study

Various defects on {100} cleavage surfaces of pyrite (FeS 2 ) are observed directly using ultra high vacuum scanning tunneling microscopy. Step edges are aligned along 〈10〉 and 〈11〉 surface directions. Atomic scale images indicate that the atomic structure, with respect to the Fe lattice, and local density of occupied states is unchanged at a step edge, including kink and corner sites. The inferred presence of monosulfides at step edges, based on X-ray photoelectron spectra on similar surfaces elsewhere, does not lead to occupied states higher in energy than d z2 dangling bond states at Fe sites. A sequence of consecutive images at the atomic scale captured evidence of dynamic structural changes at defects on this surface at room temperature. Step edges appear generally stable over the course of the STM observations, whereas vacancies, their surrounding sites, and corner step edge sites are not. Theoretical maps of the attachment energy for an Fe adatom over a {100} surface cell indicate the presence of low energy diffusion channels along the topology of the closest S atoms in the uppermost atomic S monolayer. Calculation of the activation energy barriers for the self-diffusion of an Fe adatom over a {100} terrace predict low 0.10 eV diffusion barriers along channels and 0.24 eV across channels. Subsequently, calculated Fe adatom mobilities over the time scale of the STM observations are very high, ranging from 10 5 -10 6 Å over the course of one minute, calculated for room temperature and depending on the diffusion direction. The structural changes documented in the STM images are attributed to the natural process of surface self-diffusion.

Chlorination of Si surfaces under strain conditions

Energetics for the chlorination process of Si(100)-(2×1) surfaces with an A-type single step edge were investigated under strain conditions. Strains applied uniaxially in the direction parallel to the exposed surfaces produced little or no effect on the energetics of the first chlorination step leading to fully Cl-covered terraces (one Cl attached to every exposed Si atom). The energy of the second chlorination step (leading to the formation of attached –SiCl2 units) taking place near a step-edge site, however, is affected considerably by the applied strain. The effect was found to be largest for strains applied in the direction perpendicular to the step edge.

Atomic structure of theSi(112)7×1−Insurface

The atomic structure of the $\mathrm{Si}(112)7\ifmmode\times\else\texttimes\fi{}1\ensuremath{-}\mathrm{In}$ surface has been studied with scanning tunneling microscopy (STM) and low-energy electron diffraction (LEED), and a model is proposed on the basis of the high resolution dual bias STM images and the systematic weakening of the LEED fractional-order beams, for further investigation. In sharp contrast to what has been reported for the $\mathrm{Si}(112)6\ifmmode\times\else\texttimes\fi{}1\ensuremath{-}\mathrm{Ga}$ surface, indium atoms in the $\mathrm{Si}(112)7\ifmmode\times\else\texttimes\fi{}1\ensuremath{-}\mathrm{In}$ surface do not occupy the step-edge site.

Theoretical study of chemical reactions on CVD diamond surfaces

The β-scission growth mechanism at the diamond (100)(2×1) surface is studied by a combination of nanoscale ab-initio LDA/GGA and semiempirical tight-binding techniques to provide the necessary input into the mesoscale variable time step Kinetic Monte Carlo (KMC) simulations of CVD diamond growth. The reaction path of the β-scission reaction is critically examined, and the activation barrier of the reverse etching of the methylene adsorbate is deduced. We also considered the energetics of possible adsorbate configurations and showed that the removal of methylene (CH2) moieties from terrace and step edge sites is not appreciable. This ‘preferential etching’ of isolated CH2 from the (100) surface by atomic hydrogen effectively creates a barrier to (100) layer nucleation.

Surface speciation of calcite observed in situ by high-resolution X-ray reflectivity

High-resolution, in situ X-ray reflectivity measurements were made of the calcite (104)–water interface in calcite-saturated aqueous solutions at pH values ranging from 6.8 to 12.1 and low PCO2. The X-ray reflectivity data, taken over a momentum transfer range of 6 Å−1, indicate that the calcite surface does not vary significantly over this range of experimental conditions. From an analysis of the data at pH 8.3, the best-fit reflectivity model requires the presence of 1.0 ± 0.4 monolayer of a hydroxyl species (OH or OH2) at 2.50 ± 0.12 Å above the surface Ca ions and involves rotations of the surface carbonate groups toward the (104) plane. The X-ray reflectivity data for pH 6.8 and 12.1 can be explained without invoking changes other than protonation reactions in the surface speciation of terrace areas. This is consistent with scanning force microscopy studies of calcite growth and dissolution near equilibrium, which show that attachment and detachment of Ca and CO3 ions occurs primarily at step-edge kink sites. These results demonstrate how high-resolution X-ray reflectivity can be used for direct, in situ measurement of mineral surface structure, to provide strong constraints on chemical speciation and reactivity at the mineral-fluid interface.

Calculation of energy barriers for the chlorine etching of Si(100) surface with step edges

Calculations were carried out to determine energies and barriers involved in the chlorination process of Si(100)-(2×1) surfaces and in the subsequent desorption of Si-containing species. First and second chlorination steps for Si atoms located at step edge sites were investigated. Energy calculations for exposed terraces include extremely low and high surface Cl concentrations. No energy barriers were found for the first chlorination step, which also displayed little or no dependence on the surface chlorine concentration. However, the second chlorination step (leading to the formation of an attached SiCl 2 unit) involves a barrier, and calculated energetics depend on the Cl concentration of the terraces. One of the important outcomes of this study is related to the composition and the structure of the exposed surfaces. Results strongly indicate that the role played by surface defects and by the attached Cl concentration (in the vicinity of the adsorption sites) on reaction energetics is significant. They influence local reactions and, in turn, affect the overall etching process.

Energetics for bonding and detachment steps in etching of Si by Cl

Simulation calculations were carried out to investigate the energetics of adsorption and desorption processes taking place during etching of the Si(100) surface by Cl atoms. Clean and fully chlorinated surfaces with step edges were taken into consideration. Attachment energies of Cl atoms at step edge sites and subsequent detachment (etching) energies of SiClx species were calculated. Configurational characteristics of Cl bonded regions were also investigated. Results provide a systematic description of Cl adsorption and etching processes.

The prewave in CO oxidation over roughened and Sn alloyed Pt surfaces: possible structure and electronic causes

Atom superposition and electron delocalization molecular orbital (ASED-MO) calculations with cluster models of stepped and rough (111) Pt and Pt–Sn alloy surfaces support a picture wherein the easily-oxidized CO(ads), that is present at high coverges and has an oxidation onset potential 0.4–0.5 V (H 2) lower than the main, hard-to-oxidize CO(ads) peak, goes by a mechanism wherein a H 2O molecule attacks the CO. This is followed by oxidation and deprotonation. Results indicate that this reaction is possible at step edge sites of rough surfaces and at edges of substitutional surface islands of two or more Sn atoms in alloy surfaces. The potential dependencies for these reactions as calculated using the ASED-MO band-shift include decreasing activation energies with increasing potential and the prediction of a lower oxidation onset voltage for the alloy compared to the rough surface, in agreement with experimental measurements from the literature. A subsurface Sn layer is calculated to greatly weaken the H 2O but not the CO adsorption bond strengths to surface Pt and therefore such sites would not be expected to contribute to the prewave.

Molecular-dynamics study of the structural rearrangements of Cu and Au clusters softly deposited on a Cu(001) surface

The soft deposition of copper and gold clusters on the Cu(001) surface is studied by constant energy molecular-dynamics simulations. The atomic interactions are mimicked by a many-body potential based on the tight-binding model. The influence of cluster size, substrate temperature, and incident kinetic energy (in the low-energy limit) is analyzed. Some of the simulations are extended up to a few hundred picoseconds and within this simulation time the cluster flattens and partially spreads over the surface, and the degree of spreading depends on the above variables. The influence of surface steps has also been studied and the deposited atoms seem to have a strong affinity for step-edge sites.

Nanomorphology of Kaolinites: Comparative SEM and AFM Studies

AbstractNanomorphological structure of well-crystallized Georgia and poorly crystallized North Queensland kaolinite particles have been compared using field emission scanning electron microscopy (SEM) and atomic force microscopy (AFM). In general, there is good agreement in information from the 2 very different imaging techniques. AFM gives more detailed information on step and ledge dimensions, microvalleys and crystallographic orientation of irregularities on basal planes and edges of the crystallites. There are major differences in nanomorphology and surface structure between the 2 kaolin samples with the Georgia kaolin showing 200–500-nm, relatively flat basal planes with some cascade-like step growth 50–100 nm wide. The edges, apparently flat and right-angled in SEM images, appear beveled in AFM images due to artifacts from the aspect ratio of the AFM tip. The North Queensland kaolinite has much more complex surface structure with anhedral crystallites attached to larger particles, high density of steps and nm-scale irregularities (often crystallographically directed). The additional step edge site contribution from the attached crystallites is estimated as a minimum of 6%, giving a total edge contribution above 30% of the kaolinite total surface area. These structures will generate a substantial pH-dependent charge across the surfaces of the North Queensland kaolinite platelets. An idealized, uniform, pH-independent, negatively charged basal plane cannot be assumed from these structures. There is also some evidence, from both SEM and AFM images, of curvature in the thinner, poorly ordered structures of the North Queensland kaolinite particles.