Low Coverage Limit Research Articles

Chemisorption of In and Al on GaAs(110)

A quantum molecular dynamics study of the interface formation of the Al/GaAs(110) and In/GaAs(110) interfaces has been performed. It is found that, at very low coverages, both Al and In atoms chemisorb on threefold coordinated sites, bonded to two As atoms and one Ga atom. The metal atoms attract each other on the GaAs surface and tend to form islands. The electronic structure for one monolayer is found to be semiconducting for both In and Al. The results presented here are consistent with the temperature-dependent evolution of Fermi level from the very low metal coverage limit to the overlayer limit for the deposition of In and Al on GaAs(110).

Coadsorption of hydrogen and CO on Pt(335): Structure and vibrational Stark effect

We have studied CO and H coadsorbed on the stepped Pt(335) surface in vacuum to learn about their interaction at steps and to compare with previous electrochemical studies. Both electroreflectance and conventional reflectance absorption vibrational spectra were obtained of atop bonded CO. Its stretch vibrational frequency ν, Stark tuning rate (dν/dE), and integrated absorbance S were all studied as functions of CO and H coverage. With CO only on step edges, S for atop CO decreases to zero with increasing H coverage. The CO affected by H is in a mixed phase and is bridge bonded. Atop CO’s (dν/dE) decreases with increasing CO coverage but is unaffected by coadsorbed H. In the low CO coverage limit (dν/dE)=88±9 cm−1/(V/Å), in agreement with theory and with previous measurement. The standard model of dipole–dipole screening is unable to explain the CO coverage dependence of both (dν/dE) and S. The vibrational polarizability of CO increases with coverage. We also compare our results with spectroelectrochemical data from CO on Pt(335) in aqueous electrolyte. Our measured (dν/dE) is too small to explain the variation of CO’s ν with electrode potential φ in the electrochemical experiments if (dE/dφ) in the double-layer is taken from conventional models. The effect of coadsorbed H is also different; in vacuum H has no effect on CO’s (dν/dE) but in electrolyte, at low CO coverage, (dν/dφ) drops to zero at the potential where H adsorption begins.

Binding energies for carbon atoms and clusters deposited on the Si(100) surface

Binding energies of carbon atoms adsorbed on a (2×1) reconstructed Si(100) surface were calculated as a function of sub-monolayer coverages. Also calculated were energies for C atoms deposited on the surface as small clusters, C n (with n = 3 and 4). All calculations were conducted considering a model potential function developed recently for systems containing C and Si atoms. For the low coverage limit (representing coverages up to one monolayer), carbon adatoms were considered as occupying only low energy surface sites. Owing to relatively large separations among these sites, C-C interactions are negligible and only Si-C interactions are found to be contributing to binding energies. The lowest binding energy in this case corresponds to a coverage of θ = 0.25. In the case of small C clusters deposited on the surface, however, binding energies were found to be stronger because of the increasing importance of C-C interactions. From an energetic viewpoint, results obtained in this study indicate that adsorbing C atoms on a Si(100) surface are more likely to form clusters than a layer-by-layer growth leading to a smooth and uniform surface coverage. This outcome was found to be consistent with various experimental results.

Long- and short-range stabilization of water by fluorine on Ag(110)

Water and fluorine were co-adsorbed on Ag(110) in an effort to model the metal/electrolyte interface and to compare the hydration of adsorbed fluorine with previous studies of hydroxyl and chlorine hydration. Thermal desorption of coadsorbed water reveals a series of fluorine-stabilized states with peak temperatures above that for the clean surface. At low water coverage water desorbs in a broad peak (B) at 260 K, representing strong hydration of fluorine, greater than that of Cl or OH. ESDIAD (electron stimulated desorption, ion angular distribution) measurements showed emission from this adsorption state along the [11̄0] direction. With increasing water coverage, a (1 × 2) low-energy electron diffraction pattern and ESDIAD emission along [001] coincide with a new water desorption peak (A 1) at 216–230 K. Two more thermal desorption peaks, A 2 at 190–200 K and A 3 at 165–175 K, develop with additional water coverage. The solvation number N s, representing the average number of stabilized water molecules per fluorine atom, ranges from 13 in the limit of low fluorine coverage to 2 at high fluorine coverage. For solvation numbers greater than 6, the fluorine/water interactions must extend beyond nearest neighbors forming either a second hydration shell or some form of ordered water layer. As fluorine coverage increases, the form of stabilization shifts from a long-range to a more direct, short-range interaction. The surface solvation of fluorine is compared with that of OH and Cl on Ag(110), as well as with the gas phase, single-ion hydration of F −, OH − and Cl −.

Simulations of femtosecond laser-induced desorption of CO from Cu(100)

Trajectory calculations of laser-induced desorption of CO from Cu(100) were performed employing molecular dynamics with electronic frictions and fluctuating forces to account for nonadiabatic energy exchange between nuclear and electronic degrees of freedom. The simulations were performed with an initial surface temperature of 95 K in the low coverage limit, with rapid electronic heating and cooling modeled via the electronic fluctuating forces. The calculated desorption yields are in rough agreement with available experiments, supporting the underlying mechanism of direct nonadiabatic energy transfer from hot electrons to adsorbate motion. Preferential nonadiabatic energy flow into frustrated rotational modes of CO on Cu is identified as a significant contributor to desorption.

An ex situ study of electrodeposited lead on platinum (111): II. vacuum characterization and thermal desorption of the emersed adlayer

The vacuum and thermal properties of Pt(111) electrodes emersed from Pb-containing HClO 4 solutions were studied with Auger electron spectroscopy, thermal desorption spectroscopy, Ar ion sputter depth profiling, and low energy electron diffraction. Following emersion of the electrode and transfer to the vacuum chamber the adsorbate consists of Pb, ClO 4, O, OH, and Cl, which form a number of (√3 × √3)R30° diffraction patterns. Lead is the primary thermal desorption product of submonolayer deposits and evolves in a single peak that shifts from 1300 K in the limit of low coverage to 1150 K at saturation of the first layer. The activation energy for Pb desorption varies linearly from 324 kJ mol in the limit of zero coverage to 273 kJ mol at monolayer saturation. Heating bulk Pb deposits produces, in addition to Pb desorption, a number of Cl- and O-containing Pb compounds, the most prominent being PbO and PbO 2, which desorb at 975 and about 800 K, respectively. PbO appears to result from oxidation during the emersion step, whereas PbO 2 forms during the temperature ramp by a thermally driven reaction with coadsorbed oxidants (e.g., ClO 4). Multilayer Pb desorbs in a peak at 960 K with an activation energy of 180 kJ mol . Auger and depth profile measurements gave reasonably consistent estimates of emersed adlayer thickness ranging from 2.5 to 25 Å depending on emersion potential.

Desorption kinetics of CO from epitaxial Pd films on W(110) in the low coverage limit

Desorption kinetics of CO from epitaxial Pd films on W(110) in the low coverage limit

Desorption kinetics of CO from epitaxial Pd films on W(110) in the low coverage limit

The adsorption-desorption kinetics of CO from ultrathin epitaxial Pd films has been studied using modulated molecular beam techniques. We find that the rate constant for desorption, k d, is nearly four orders of magnitude higher for the monolayer Pd film than that for the bilayer film. The rate decreases further by a factor of 7 between 2 and 6 monolayers at which k d attains the value seen on Pd single crystals. We conclude that although a bilayer film is similar to a bulk Pd(111) surface in binding CO, true bulk-like behavior is only reached above six atomic layers of Pd. Temperature programmed desorption at low coverage can be fit using the beam determined kinetic parameters, showing that defects play no major role in the beam experiments. Significant departures in the values of the activation energy for desorption and the preexponential factor for the monolayer film from expected values are attributed to a possible adsorbate induced surface reconstruction.

Alkali adsorbates on metal surfaces: observable consequences of the ionic K/Cu(100) interaction

We have examined the consequences for several observable properties of the adsorption and chemical bonding of alkali atoms on metal surfaces. The properties considered are vibrational frequencies, core level binding energies, BE's, and work functions. Our results are based on the use of molecular orbital, electronic wavefunctions for cluster models of K/Cu(100). Our principle concern is for the low coverage limit where our cluster model wavefunctions show that the Cu-K bond is dominantly ionic; however, we also consider models for covalently bonded K to represent high coverage adsorption of the alkali. All the properties that we have studied are consistent with a fully ionic alkali adsorbate at low coverage. In particular, we show: (1) the large dynamic dipole moment of ionic K on a metal surface leads to a large observed intensity of the K vibrations for the frustrated translational mode normal to the surface; (2) a small BE shift for the core levels of the surface atoms of the substrate is obtained even for an ideal ionic point charge model to represent the K + adsorbate; and, (3) the adsorbate BE's are smaller by about 1.5 eV when the Cu-K bond is covalent than when it is ionic. This is shown to be consistent with the observed decrease in the core level BE's of alkali adsorbates as their coverage on the substrate increases. Recent proposals that the interaction between alkali adsorbates and metal surfaces are covalent at all coverages, including the limit of zero coverage, are not supported by the present study.

Schottky barrier formation for In deposited on GaAs(110): the low coverage limit

An ab-initio LCAO method is used to calculate the electronic properties of different low coverages of indium deposited on GaAs(110). Results for a monolayer, half a monolayer, a quarter of a monolayer and the isolated indium atom are presented. Chemisorption energies and the most favourable adsorption sites are calculated for each case. The interface Fermi energy and the Schottky barrier formation are discussed as a function of the metal coverage.

Si and Ge Nanocrystallites Embedded in CaF2 by Molecular Beam Epitaxy (MBE)

AbstractThin films of CaF2 containing layers of Si and Ge nanocrystallites were grown epitaxially on Si( 111) substrates by MBE while varying the substrate temperature. The sticking coefficient of both Si and Ge to CaF2 are less than unity at the temperatures studied. Evidence that chemical reactions are minimal between Si and CaF2 leading to the formation of volatile species such as SiFx is presented and it is surmised that the dominant mechanism for re-evaporation is thermal desorption. Both Si and Ge sticking coefficients vary exponentially with 1/T and activation energies are determined. A cluster growth model describing the evolution of the amount of Si in clusters over time is given. Solving in the limit of low cluster coverage, a solution that varies linearly with time and exponentially with 1/T is obtained. Weak room temperature photoluminescence from Si nanocrystallites in CaF2 is seen, however, it is unclear whether the luminescence is coming from the Si nanocrystallites or the CaF2. Second harmonic generation is observed from samples containing single layers of Ge in CaF2.

Schottky barrier formation in the low metal coverage limit

A theoretical discussion of the state of the art in the field of Schottky barrier formation is presented. The thin metal layer limit is analyzed for Al and In on GaAs(110), and the adsorption sites are determined as a function of the metal coverage. Intrinsic and extrinsic charge neutrality levels are also discussed for the In-case.

Lateral Interaction between Adsorbates at Ionic Surfaces: Theoretical Investigation of the CO Vibrational Frequency Shift at MgO(001)

AbstractThe adsorption of CO molecules at the MgO(001) surface is investigated by the LCGTO‐LDF cluster method for the low‐coverage limit as well as for a monolayer coverage. A static adsorbate‐adsorbate contribution to the partial red shift of the C–O vibrational frequency with increasing surface coverage is calculated in agreement with experiment. This red shift is related to a small adsorbate‐to‐substrate σ charge transfer which modifies the electrostatic field on the surface and thus is a “solvent” effect. No transmission of the donated charge through the substrate is found. Both the low‐coverage and the monolayer models predict an approximate doubling of the C–O vibrational intensity as compared to free CO. This intensity enhancement derives to a large extent from a change of the π component of the CO dynamical dipole moment due to a Pauli repulsion between the adsorbate and the nearest‐neighbour surface anions.

The reaction of atomic oxygen with Si(100) and Si(111): I. Oxide decomposition, active oxidation and the transition to passive oxidation

The reactions of atomic oxygen with the (100) and (111) surfaces of silicon have been investigated by employing supersonic molecular beam techniques and X-ray photoelectron spectroscopy. The kinetics and mechanism of the active oxidation reaction, i.e., O x(g) + Si(s) → SiO(g) where x = 1 or 2, has been evaluated by employing modulated molecular beam reactive scattering (MMBRS). On both surfaces, the reaction of atomic oxygen involves the formation of a single stable surface intermediate, which reacts via first-order kinetics to produce SiO(g). The reaction of molecular oxygen, however, involves two stable surface intermediates that are formed sequentially, the second of which is identical to that formed by the reaction with atomic oxygen. We propose that the first intermediate formed in the molecular oxygen reaction is chemisorbed O 2(a), e.g., a peroxy radical or a peroxide bridge. The intermediate formed in the atomic oxygen reaction is assigned to either an isolated oxygen adatom or adsorbed SiO (a surface silanone complex). Oxide decomposition in the mono- and multi-layer regime has been examined with temperature-programmed desorption (TPD). Both increasing oxygen coverages and higher adsorption temperatures lead to higher decomposition temperatures for the oxygen adlayers formed. Above the monolayer regime, adlayers formed on Si(100) decompose at higher temperatures than those on Si(111). On Si(100), simple first-order decomposition kinetics is only observed in the monolayer regime, and in the limit of low coverage (θ < 0.3 ML). The implied rate coefficient for desorption in this regime is at least 2 orders of magnitude smaller than that measured by MMBRS at the same temperature. At higher coverages (θ ⩾ 4 ML) and on both surfaces, the decomposition reaction appears to be heterogeneous, involving void nucleation and growth. The transition between active and passive oxidation has been examined employing in situ, real-time mass spectrometric and XPS measurements. The transition involves a competition between highly activated ( E d ∼ 80 kcal mol −1) desorption of SiO, and nearly unactivated nucleation of surface oxide.

Kinetics of water formation on Pt(111)

Measurements using modulated molecular beam relaxation spectroscopy identify ODa consumption as the rate‐limiting step for D2O formation on Pt(111). In the limit of low surface coverage, D2O is produced via two parallel pathways, Da+ODa→D2Oa with activation energy Ea≊16 kcal/mol, and 2ODa→D2Oa+Oa with Ea≊18 kcal/mol. From these results the reactions Da+Oa →ODa, Da+ODa→D2Oa, and D2Oa+Oa→2ODa are found to be exothermic with enthalpy changes ΔH≊−16, −9, and −8 kcal/mol, respectively, and ΔHOHa≊−61 kcal/mol is obtained for the heat of adsorption of hydroxyl on platinum. The general features of the reaction mechanism cause complex kinetic behavior, manifest as changes in the apparent order of the rate function and the effective activation energy as surface temperature and reactant pressures are varied. The implications for transient and steady‐state rate measurements are discussed.

Collision-induced desorption of physisorbed CH4 from Ni(111): Experiments and simulations

The desorption of CH4 physisorbed on Ni(111) is observed to be induced by collision with Ar atoms incident with energies less than 2 eV. The absolute cross section for collision-induced desorption of CH4 in the low coverage limit of an isolated CH4 molecule and from a saturated CH4 monolayer is measured as a function of the kinetic energy and incident angle of the Ar beam. The dominant mechanism for collision-induced desorption is determined to involve the direct collision of the incident Ar with the physisorbed CH4. Indirect, surface mediated desorption processes and multiple desorptions are found to be unimportant. Three-dimensional, classical molecular dynamics simulations based upon a hard sphere/hard cube model of the direct collision mechanism show that the complicated dependence of the desorption cross section at low CH4 coverage on the Ar energy and incident angle is the result of two competing dynamical effects: the increase in the geometrical collision cross section and the decrease in the Ar kinetic energy that can be transferred to CH4 motion normal to the surface as the Ar incident angle increases. Multiple Ar–CH4 collisions and mirror collisions are found to make relatively minor contributions to the cross section for collision induced desorption. Normal energy accommodation during the CH4-surface collision plays a significant role in determining the threshold energy for desorption. At high CH4 coverage, the obstruction of small impact parameter, head-on Ar–CH4 collisions by neighboring CH4 molecules at large angles of incidence is the origin of the difference in the cross section observed for low and high CH4 coverage.

Investigation of the dynamics of the O2/Si(001) adsorption system by small-angle ion-surface scattering

The small non-adiabaticity associated with the slow collision of a 450 eV to 3 keV O2+ beam incident at 5° to a carefully flattened Si(001) surface was used to probe the adiabatic potential energy surface for the O2/Si(001) adsorption/reaction system in the limit of low coverage. The scattering products include O2− and O−. From the observation of efficient O2− formation and th dependence of the negative-ion yields on beam energy, we infer that O23− is an intermediate or precursor in the dissociative chemisorption of O2 on Si(001). The total yield of negative ions (O2− and O− ) is high in comparison with similar experiments on th O2/Ag(111) system. This may be explained by a reduced reneutralisation probability on the exit trajectory due to the band gap of Si.

O 2− formation at Si(001)

A 450 eV−3 keV O 2 + beam incident at 5° to a carefully flattened Si(001) surface was used to probe the adiabatic potential energy surface for the O 2 Si(001) adsorption/reaction system in the limit of low coverage. Over the full incident energy range, O 2 - an are formed as scattering products. We argue that the O 2 − product correlates with an O 2 −-like region on the potential energy surface and therefore propose that an O 2 −-like species is an intermediate or precursor in the dissociative adsorption of O 2 on Si(001). Classical trajectory calculations indicate that a large amount of collision induced dissociation occurs at all energies.

Potential Energy for the CO2‐NaCl(100) and the Lateral Adsorbate Interaction

AbstractAn empirical adsorption potential VAD of a single CO2 molecule on the (100) face of NaCl is reported. The molecule‐surface interaction is calculated for different orientations of the adsorbed molecule. The lowest energy minimum is found for a molecule parallel to the surface above the cell center, having an energy of adsorption in good agreement with the measured value in the low coverage limit. Using the potential the vibrational v2‐splitting is calculated for the different orientations of the molecule. — To determine the structure of the CO2‐monolayer on NaCl(100), intermolecular Coulomb and van der Waals terms are added to the molecule‐surface potential. Two basic overlayer structures, namely (2 × 2) and (2 x 1), are investigated and the potential minima determined under variation of the adsorbate configuration. In both cases we find an angle ϕ 90° between neighbouring molecules and a tilt angle ϑ of 30–40 between the CO2 molecular axis and the NaCl(100) plane for adsorption sites above Na+. The deepest potential energy minimum for a (2 x 1) structure with 2D space group pg is found for CO2 shifted by 0.07 Å from the center of the line between adjacent Na+ and Cl− towards Cl− and ϕ = 85.6°,ϑ= 26.2°. According to our spectral line shape analysis and LEED measurements [12] the adsorbate has the (2 x 1) structure.

Alkali metal adsorbates on W(110): Ionic, covalent, or metallic?

The photoemission signal from the first atomic layer of W(110) is used to assess the nature of the interaction between the surface atoms of the metal substrate and the adsorbates Na, K, and Cs for coverages up to 1 atomic layer. Our results indicate that there is little or no charge transfer from the alkali metal to the W surface, even in the limit of low coverage. The satellite structure of the photoemission lines of the outermost p shell of the alkali metals confirms this conclusion. While contrary to the conventional picture of alkali-metal-charge donation, these findings fully support recent theoretical calculations.