Adatom-substrate Interaction Research Articles

The initial stages of copper deposition on Ag(111): an STM study

The initial stages of Cu deposition on Ag(111) from a sulfate-containing solution has been studied by in-situ scanning tunneling microscopy (STM) for low and high overpotentials. Very surprisingly, at low overpotentials Cu is preferentially deposited at the terraces of the substrate rather than at steps and it grows two-dimensional. At high overpotentials, Cu nucleates at surface defects such as monoatomic high steps and forms three-dimensional clusters, very much like in the case of Cu on Au(111). At low overpotentials the first layer of Cu grows pseudomorphic on Ag(111) and it may cover the whole surface before the second Cu layer starts to grow on top. Yet, no underpotential deposition (UPD) is observed for Cu on Ag(111). This very unusual nucleation-and-growth behaviour is explained by the specific adsorption of sulfate at the steps of Ag(111), which blocks these sites for Cu deposition, and by a delicate balance of adatom-adatom and adatom-substrate interactions, which causes the UPD shift to disappear.

Vibrational properties of the Si(111)Ga( sqrt 3 x sqrt 3 ) surface.

The adatom-substrate interaction on the Si(111) Ga(\ensuremath{\surd}3 \ifmmode\times\else\texttimes\fi{} \ensuremath{\surd}3 )R30\ifmmode^\circ\else\textdegree\fi{} structure was studied with high-resolution electron-energy-loss spectroscopy and ab initio total-energy cluster calculations. Energy losses at 100, 200, 340, 480, and 550 ${\mathrm{cm}}^{\mathrm{\ensuremath{-}}1}$ were identified as the folded-back part of the Rayleigh wave, the Ga adatom vibration, a breathing mode of a ${\mathrm{GaSi}}_{4}$ cluster, and two different Si-Si stretching vibrations, respectively. The dispersion of these modes, measured along the \ensuremath{\Gamma}\ifmmode\bar\else\textasciimacron\fi{}K\ifmmode\bar\else\textasciimacron\fi{} direction of the (1\ifmmode\times\else\texttimes\fi{}1) surface Brillouin zone, was found to be weak.

Binding energy per adatom in metal clusters adsorbed on metal substrates

In this paper the dependence of the binding energy per atom on the cluster size for metal clusters supported on metal substrates is studied. A semi-empirical Anderson-Newns-Grimley model is applied to calculate qualitatively the binding energy for neutral adatoms. The influence of the support on the “even-odd” oscillations is revealed and the competition between the adatom-adatom and the adatom-substrate interaction is discussed. The connection of the obtained results with some experimental measurements is shown.

Surface phases of reconstructed tungsten (100) and molybdenum (100)

After several decades of experimental and theoretical work it has become clear that on most solids, including metals, the surface is subject to reconstruction, i.e. the outermost substrate atoms rearrange in response to changes in temperature, adsorbate coverage, or other variables. Studies of W(100) and Mo(100) have been very useful in this development. Early LEED studies of adsorption on these surfaces revealed a large number of distinct phases and it proved impossible to explain their formation in terms of conventional models which assumed the substrate to be “passive”. The subsequent discovery that the surface is reconstructed in these phases necessitated a reinterpretation of the results and showed that the effective adatom-substrate and adatom-adatom interactions depend on metal atom displacements. It has been found that, in general, substrate structural changes have a strong — sometimes dominant — influence on the kinetic and thermodynamic surface properties.

Adsorption parameters of Cu and Ni on the {110} plane of W and Mo

This paper reports on the embedded-atom method (EAM, developed by Johnson) calculations of adsorption parameters for Cu and Ni on {110} unrelaxed substrates of Mo and W. The following are calculated: (i) the equilibrium height hcof an adatom; (ii) the optimum coefficients W0Ahkof a truncated Fourier representation of the adatom-substrate interaction potential; (iii) the desorption energies Edes; and (iv) the activation energies Q of surface migration. It is shown that (a) the calculated values of Edesagree satisfactorily with available empirical data; (b) the scale factor W0 is approximately proportional to the bonding as measured by Edes, whereas the normalized coefficients Ahkare determined by the {110} substrate symmetry; (c) Edesand Q are respectively dominated by the embedding energy and pair potential of the EAM; and (d) W0 and Edesfor an atom in a monolayer are only about half as much as the corresponding quantities for an isolated atom.

Electronic structure of [formula omitted]R30°-Na and -K on Al(111): Comparison of “normal” and substitutional adsorption sites

Recently, it was shown that Na and K on Al(111) will kick out a surface Al atom and occupy the substitutional site. For K also the on-top position is possible. In this contribution, we give an analysis of the adsorbate wave functions, surface density of states, and surface band structures. Different adsorption sites are considered, and it is shown that the mechanism of bonding in the substitutional geometry on Al differs from that in the “normal” on-surface geometry. In the first case the adatom-substrate interaction dominates, and the surface electronic structure is largely determined by the character of the Al surface vacancy, whereas in the second case the adatom-adatom interaction dominates. The theoretical results are compared with experimental studies of direct and inverse photoemission, suggesting the need of a re-analysis of the data.

A molecular dynamics study of a nearly incommensurate overlayer on a metal substrate

We have examined the dynamics of a nearly incommensurate metal overlayer on a metal substrate using the technique of molecular dynamics. We have chosen the parameters in our model of the interatomic forces between atoms to fit the experimentally observed perpendicular vibration frequency of a silver overlayer on Ni(100). The corrugation of the silver-nickel interaction is not known. With this model it is possible to vary the corrugation of the adatom-substrate interaction potential. Thus we can study, in a general way, the effects of corrugation on the phonon spectrum and the dynamics of adatoms. At high temperatures (room temperature and above) we find the metal atoms undergo large displacements with respect to the substrate. The effects of these displacements on the phonon spectra are examined as a function of temperature and the corrugation of the interaction potential. The results will be of use to experimentalists studying the vibrations of nearly incommensurate systems.

Low-temperature wetting transitions in rare-gas monolayers adsorbed on graphite.

Interfacial wetting transitions are located in a two-dimensional model system describing rare-gas monolayers adsorbed on a graphite substrate. The adatom-adatom interactions are of the Lennard-Jones type with parameters \ensuremath{\varepsilon} and \ensuremath{\sigma}, while the corrugation amplitude of the adatom-substrate interaction is given by ${\mathit{V}}_{1}$. We calculate numerically the domain-wall energies of the four types of domain walls occurring in the registered (\ensuremath{\surd}3 \ifmmode\times\else\texttimes\fi{} \ensuremath{\surd}3 )R30\ifmmode^\circ\else\textdegree\fi{} structure. This structure is stable in a region of the (\ensuremath{\sigma},${\mathit{V}}_{1}$/\ensuremath{\varepsilon}) plane where all wall energies are positive and is further subdivided by a number of wetting lines marking the stability of a given interface against interfacial adsorption.

Chemisorbed halogen adlayers on Fe(110): electronic band structure and surface geometry

Abstract A series of overlayer structures, starting with a c(3 × 1) geometry are formed during the dissociative chemisorption of chlorine, bromine, and iodine on Fe(110). These adlayers provide an opportunity to systematically examine the corresponding changes in electronic band structure as a function of coverage and adsorbate. At coverage levels greater than θ = 1 3 low-energy electron diffraction patterns obtained from the halogen-covered Fe(110) surface evolve continuously with coverage. The electronic band structure was examined as a function of coverage and adsorbate species via angle-resolved ultraviolet photoelectron spectroscopy. It is concluded that the electronic band structure measurements support an incommensurate structural model for halogen overlayer geometry on Fe(110) in which an overlayer lattice, that is in general incommensurate with the substrate, rotates and compresses with increases in coverage. An alternative interpretation of the diffraction results based on antiphase domain boundaries within the overlayer is not consistent with the photoemission data. The incommensurate model emphasizes adatom-adatom interactions over adatom-substrate interactions while the reverse is true for the antiphase boundary model.

Chemisorbed halogen adlayers on Fe(110): Electronic band structure and surface geometry

A series of overlayer structures, starting with a c(3 × 1) geometry are formed during the dissociative chemisorption of chlorine, bromine, and iodine on Fe(110). These adlayers provide an opportunity to systematically examine the corresponding changes in electronic band structure as a function of coverage and adsorbate. At coverage levels greater than θ = 1 3 low-energy electron diffraction patterns obtained from the halogen-covered Fe(110) surface evolve continuously with coverage. The electronic band structure was examined as a function of coverage and adsorbate species via angle-resolved ultraviolet photoelectron spectroscopy. It is concluded that the electronic band structure measurements support an incommensurate structural model for halogen overlayer geometry on Fe(110) in which an overlayer lattice, that is in general incommensurate with the substrate, rotates and compresses with increases in coverage. An alternative interpretation of the diffraction results based on antiphase domain boundaries within the overlayer is not consistent with the photoemission data. The incommensurate model emphasizes adatom-adatom interactions over adatom-substrate interactions while the reverse is true for the antiphase boundary model.

Incommensurate epitaxial growth of Cu on Pd{111}

A low-energy electron diffraction (LEED) study of the growth of Cu on Pd{111} shows that the Cu films are epitaxial but are incommensurate with the substrate and have the equilibrium fcc Cu structure, in contrast to the growth on Pd{001}, where the Cu films are both epitaxial and pseudomorphic. The different behavior of the growth of the same material on two different surfaces of the same substrate is probably due to different magnitudes of the ratio of adatom-adatom interactions to the adatom-substrate interactions on the two surfaces.

Description of adsorption phenomena within the framework of the embedded atom method

We have calculated and compared the self-adsorption parameters (desorption energy, activation energy of adatom surface migration, vibration frequencies of adsorbed atoms, and truncated Fourier representation of the adatom-substrate interaction) for a W adatom on a W {110} surface using embedded-atom methods (EAM) introduced by Gollisch (G), by Finnis and Sinclair (FS) and by Johnson and Oh (JO), using the fitting parameters of the authors for bulk crystals. A small modification noticeably improved the results from the FS model. The equilibrium adsorbate-substrate binding energy, and height were mapped within a unit cell to generate the necessary data. The G and FS models both display double minima.

Energies and symmetries in interface formation: In/GaP(110) and Ga/InP(110)

We have used high-resolution synchrotron radiation photoemission to study adatom–substrate interactions and growth morphologies for In/GaP(110) and Ga/InP(110). Room-temperature experiments reveal extensive adatom clustering, but also substrate disruption and cation segregation for both interfaces, with greater amounts for Ga/InP(110). Ga deposition in InP(110) at 60 K also results in substrate disruption, but with kinetic trapping of the released In atoms close to the interface and a greater tendency toward layer-by-layer growth. In constrast, the deposition of preformed metallic Ga clusters shows no evidence for substrate disruption. We conclude that atom condensation and coalescence are responsible for disruption, with different activation barriers being present for cluster deposition and atom deposition because of the details of surface and interface bonding. For Ga deposition on n-type InP(110) at 60 K, the appearance of states at the Fermi level is correlated to changes in band bending. Metal cluster deposition leads to pinning positions that can be related to surface unrelaxation around the perimeter of the clusters; we find no evidence for metal induced gap states or defect levels.

An alternative potential model for one-dimensional commensurate-incommensurate transitions

A revised cosine potential is introduced to describe the adatom-substrate interaction for an array of interacting gas atoms adsorbed on a crystalline substrate. Starting from this potential, a nonlinear equation which the ground state satisfies and its solutions are obtained and discussed. The results obtained are then used to explain the observed structural phase transitions in TTF-TCNQ and found to be in good agreement with the experiment.

Structural phase transitions in krypton monolayers adsorbed on graphite

An alternative substrate potential to describe the adatom-substrate interaction is applied to study the two-dimensional commensurate-incommensurate (CI) transitions in krypton monolayers adsorbed on graphite. It is shown that in this overlayer system the CI transition is continuous and the hexagonal symmetry is maintained in the incommensurate phase, the critical exponent β for this transition is equal to 1 3 . These results explain the recent experiments on krypton on graphite.

Equilibrium interaction potential and height of an argon adatom on an Ar(001) crystal surface

A computational procedure to obtain optimum truncated two-dimensional Fourier representations of both the equilibrium adatom-substrate interaction potential and the equilibrium height of an adatom on a crystal surface, is presented. The procedure is applied to an adsorbed argon atom on a (001) argon crystal surface. A suitable Lennard-Jones potential is used to model the atomic interaction. The calculations show the following: (i) the Fourier coefficients in both representations tend rapidly to zero with increasing harmonic order; second-order terms are very small and third-order terms are negligible; (ii) the derivation of reliable force constants from the series requires higher-order truncations; (iii) the upper atomic plane contributes more than 80% not only towards the total adatom-substrate interaction potential energy U, but also towards the constant and first-order coefficients; (iv) the contributions of consecutive atomic planes to U and to the coefficients decay exponentially. The considerations therefore justify the implementation of truncated Fourier series including only low-order harmonic terms. This is of great importance, e.g. in epitaxy, because both the tendency to epitaxy and the realization of a specific orientation can be correlated directly with the Fourier coefficients and their values. The modulation in equilibrium height of about 30% is also of great significance, e.g. in molecular dynamic studies of adsorbed phases.

Molecular dynamics simulation of atomic O on Pt(111)

A molecular dynamics computation was used to study the dynamic behavior of atomic oxygen on a crystalline platinum FCC(111) surface. A Lennard-Jones (12-6) potential function was used to describe pairwise interactions of both substrate (Pt) and adatom (O) interactions. Using published values for the platinum potential, the adatom-substrate interactions were determined by an iterative process. Calculations of resultant frequency spectra were compared to experimental EELS data found in the literature. The simulation based upon the Lennard-Jones potential was able to reproduce the experimental frequency spectrum with good agreement. The model was fit to three different published PtO bond lengths (2.1, 1.98, and 1.91 Å), without altering the frequency spectra; also adsorption energies for the adatom were in accordance with published values. The simulation showed versatility to fit several sets of results suggesting applicability to several experimental environments. Observations were made as to the motion of the adatom on the Pt(111) surface. Diffusion measurements were made as a function of temperature and the results compared to similar work which utilized a different method.

Epitaxy at {111}fcc/{110}bcc metal interfaces

This paper reviews a theory that justifies “lattice match” assumptions in epitaxy and the application of the theory to epitaxy at {111}fcc/{110}bcc metal interfaces. The basic assumptions of the model are (i) that the adatom-substrate interaction has the periodicity and symmetry of the substrate surface, (ii) that the overlayer is crystalline, (iii) that the adatom-adatom interaction is harmonic and (iv) that the ideal epitaxial configurations are the same as for rigid half crystals. The theory correlates epitaxy with lattice match and predicts the occurrence of the Nishiyama-Wassermann and Kurdjumov-Sachs orientations in thick overlayers with astonishing accuracy. In thin overlayers there is good qualitative agreement with empirical data. The theory stresses the need for quantifying the Fourier coefficients.

Epitaxy at $lcurlb;111$rcurlb;fcc/$lcurlb;110$rcurlb;bcc metal interfaces

This paper reviews a theory that justifies “lattice match” assumptions in epitaxy and the application of the theory to epitaxy at {111}fcc/{110}bcc metal interfaces. The basic assumptions of the model are (i) that the adatom-substrate interaction has the periodicity and symmetry of the substrate surface, (ii) that the overlayer is crystalline, (iii) that the adatom-adatom interaction is harmonic and (iv) that the ideal epitaxial configurations are the same as for rigid half crystals. The theory correlates epitaxy with lattice match and predicts the occurrence of the Nishiyama-Wassermann and Kurdjumov-Sachs orientations in thick overlayers with astonishing accuracy. In thin overlayers there is good qualitative agreement with empirical data. The theory stresses the need for quantifying the Fourier coefficients.

Helium scattering studies of the surface structure and dynamics of rare-gas crystals

We review our experiments involving He scattering from the surfaces of ordered rage-gas overlayers physisorbed on Ag(111). The rare gases studied were Ar, Kr and Xe. The measurements were performed on thin-film structures which were grown on a layer-by-layer basis, from monolayer to thick film, in order to study how the surface properties changed with increasing distance from the substrate. Three types of experiments are discussed; elastic diffraction, selective adsorption and inelastic single-phonon scattering by angle-resolved time-of-flight. These experiments, taken together, provide a broad overview of the static and dynamic surface properties of rare-gas overlayers. They also provide information about the forces which govern the adatom–adatom and adatom–substrate interactions.