Island formation and condensation of a chemisorbed overlayer

The condensation of a chemisorbed overlayer from a lattice gas into a particular ordered structure in discussed in terms of the lattice-gas theory of Lee and Yang. The formation of islands of ordered structure is identified with the condensation phenomenon predicted by the theory. The phase diagram (transition temperature versus coverage) based on the theory of a two-dimensional Ising ferromagnet in zero magnetic field is constructed for the condensation of a lattice gas system with net attractive interactions between the particles. It is demonstrated that critical points at coverages other than θ = 0.5 are achieved for overlayer systems with unit meshes larger than (1 × 1). Low-energy electron diffraction results of the thermal disordering (island dissolution) for oxygen chemisorbed on W(110) are compared with the theory, and the effect of substrate surface heterogeneity on the phase diagram is discussed.     

The bonding of oxygen to Al(111)

Oxidation of the Al(111) surface is a two-stage process in which the formation of an ordered oxygen overlayer precedes the creation of a bulk-like amorphous oxide. An electronic structure calculation is reported here for the clean and oxygen-covered Al(111) surface and for bulk A1₂O₃. The calculation uses an atomic-orbital basis and the metal surface is modelled by an infinite two-dimensional crystal, containing seven layers of aluminium atoms. Oxygen atoms occupy three-fold sites, with an Al-O separation of 1.9 Å. The oxygen 2p resonance in the (1 × 1) chemisorbed overlayer is about 3 eV wide, compared to 1.9 eV for an equivalent isolated layer of oxygen atoms unhybridized with the metal. The valence band of A1₂O₃ is about 1.5 eV wider than the chemisorbed oxygen resonance, but in both cases most of the states are concentrated in the top 1.5 eV of the band. The results are related to available ultraviolet photoemission spectra, including the recent angular-resolved spectra of Martinson and Flodström.     

A LEED/UPS study on the interaction of oxygen with a Ni(111) surface

Exposure of a Ni(111) surface to oxygen leads at first to the formation of a chemisorbed overlayer which is characterized by a 2 × 2-superstructure and a maximum in the photoemission spectrum (hv = 40.8 eV) centered at 5.6 eV below the Fermi level E_F. The emission from the Ni d-states is nearly unaffected at this stage of interaction. After high oxygen exposures the epitaxial growth of NiO can be identified from the LEED pattern. The corresponding photoelectron spectrum is strongly altered and exhibits close agreement with the transition energies as calculated by Messmer et al. for a NiO₆^10- -cluster.     

Characterization of oxygen,carbon, and sulfur adlayers on W(211)

Adlayers of oxygen, carbon, and sulfur on W(211) have been characterized by LEED, AES, TPD, and CO adsorption. Oxygen initially adsorbs on the W(211) surface forming p(2 × 1)O and p(1 × 1)O structures. Atomic oxygen is the only desorption product from these surfaces. This initial adsorption selectively inhibits CO dissociation in the CO(β₁) state. Increased oxidation leads to a p(1 × 1)O structure which totally inhibits CO dissociation. Volatile metal oxides desorb from the p(1 × 1)O surface at 1850 K. Oxidation of W(211) at 1200 K leads to reconstruction of the surface and formation of p(1 × n)O LEED patterns, 3 ? n ? 7. The reconstructed surface also inhibits CO dissociation and volatile metal oxides are observed to desorb at 1700 K, as well as at 1850 K. Carburization of the W(211) surface below 1000 K produced no ordered structures. Above 1000 K carburization produces a c(6 × 4)C which is suggested to result from a hexagonal tungsten carbide overlayer. CO dissociation is inhibited on the W(211)?c(6×4)C surface. Sulfur initially orders into a c(2 × 2)S structure on W(211). Increased coverage leads to a c(2×6)S structure and then a complex structure. Adsorbed sulfur reduces CO dissociation on W(211), but even at the highest sulfur coverages CO dissociation was observed. Sulfur was found to desorb as atomic S at 1850 K for sulfur coverages less than $\text{7}\text{6}$ monolayers. At higher sulfur coverages the dimer, S₂, was observed to desorb at 1700 K in addition to atomic sulfur desorption.     

Structural and reactive properties at the metal-halogen interface: The interaction of bromine with chromium (110)

The chemisorption and subsequent reaction of bromine on Cr(110 has been studied by Auger spectroscopy, LEED, Δφ, and thermal desorption measurements. For gas doses of < 7.5 × 10¹⁸ molecules m^?2, very efficient dissociative chemisorption leads to a series of well-ordered, out-of-registry compression structures. Uniquely, however, the overlayer falls back into registry at saturation coverage; at this point the appearance of glide symmetry indicates that the three-fold coordinated adsorption sites are occupied exclusively. Bromine → metal charge transfer occurs during adsorption (in contrast to Cr(100)). On raising the temperature at low coverages, the surface phase decomposes by evaporation as CrBr molecules; at higher coverages the desorption product switches to CrBr₂. Continuous growth of bulk CrBr₂ sets in at high gas exposures, this corrosion reaction proceeding at a rate which is ten times slower than the rate of overlayer formation. The chromium dibromide layer also evaporates as CrBr₂(g). Structural relationships with related metal-halogen systems are discussed.     

Ordered oxygen overlayers on Cr(100) observed by leed and photoemission

Chemisorbed oxygen atoms on Cr(100) induce strong O(2p) derived surface resonances which are studied by angle resolved photoemission. Well ordered structures are observed after annealing (300°C). In the submonolayer range (θ_O < 1) a study of the symmetry and dispersion of the O(2p) derived features shows the two-dimensional Bloch character associated with either a c(2 × 2)-O surface at low coverages (θ_O?0.25) or a (1 × 1)-O structure at high coverages (θ_O?0.9). When combined with LEED observations and work function data this study indicates that both structures coexist around θ_O = 0.5 and chemisorbed oxygen is probably incorporated into the fourfold hollow sites. At θ_O > 1, the onset of oxidation is clearly shown in the valence band and core level spectra and the data support the existence of a thin spinel-like oxide layer.     

Low energy electron diffraction observations on Au-enriched Ni-0.8%Au alloy(110) surfaces prepared by surface segregation

Au-enriched (110) surfaces with Au coverages θ up to 1 monolayer (ML) were prepared by heating a Ni-0.8%Au crystal at temperatures T > 400°C. Low energy electron diffraction (LEED) indicated the following domains at T ≈ 20°C: 5× 1 (prominent for θ ≈ 0.7 ML), c(2 × 2) (0.7−0.9 ML). 7 × 7 (0.8−1.0 ML). The 7 × 7 domain was shown to contain c(2 × 4) subunits. The observations are shown to be consistent with a nearly close-packed 7 × 7 overlayer structure in which each c(2 × 4) subunit is associated with a hexagonal formation of six Au atoms, with a seventh Au atom at the center. Ni(210)-1 × 1 facets and Au(321)-3 × 1 facets due to Au crystallites were present on θ ≈ 1 ML surfaces. The effect of increasing T on θ and on LEED intensities was observed for θ ≈ 1 ML surfaces. θ was found to decrease by 15% on increasing T from 300 to 800°C. A first-order 7 × 7 → 2 × 7 phase transition was observed at T = 340 ± 10°C. Specific nearly close-packed overlayer models with AuAu equilibrium interatomic distance 2.7 Å were proposed for both the 7 × 7 and 2 × 7 surfaces. The models were shown to explain the observed LEED patterns qualitatively. Other LEED observations were interpreted tentatively to indicate a transition to a disordered overlayer structure for T > 550°C, disappearance of facets for T > 600°C, and one-dimensional melting of Au overlayers for T ≈ 800°C.     

Hydrogen-induced reconstruction of the W(100) surface,studied by surface core level spectroscopy

We report W(4?) surface core level shifts which yield new information on the energetics of the W(100) (1 × 1) → C(2 × 2)H phase transition. At small hydrogen coverages we find two co-existing surface core levels from atoms on normal lattice sites and from atoms in reconstructed domains. These surface levels are shifted to smaller binding energy (toward E_F) by 0.35 eV and 0.13 eV relative to the bulk level, respectively. The most stable configuration is obtained at a fractional coverage θ_H ? 0.2, at which all surface atoms are shown to be paired with neighboring atoms in the surface plane.     

The Pt(110) phase transitions: A study by rutherford backscattering,nuclear microanalysis,LEED and thermal desorption spectroscopy

The adsorbate induced (1×2) (1×1) (2×1)p1g1 phase transitions on Pt(110) have been studied by Rutherford backscattering (RBS), nuclear microanalysis (NMA), LEED and thermal desorption spectroscopy. RBS data indicate that any displacement of the surface atoms from their expected bulk-like lattice sites in the (1×2) phase is ? 0.002 nm laterally and ? 0.007 nm vertically. This contraint eliminates models for the reconstruction which involve significant lateral displacements (e.g., the paired-atom or hexagonal overlayer models). The RBS data are consistent with both the rumpled model with up/down displacements not exceeding ~0.007 nm and the missing row model with an unrelaxed surface in which the out-of-plane vibrational amplitude is slightly enhanced. A c(8×4) phase, produced by CO (or NO) exposure at T?250 K, has also been characterized by RBS which demonstrated that 0.92×10¹⁵ Pt cm^?2 move on average by ~0.017 nm laterally out-of-registry with the bulk upon formation of this phase. The values of the saturation adsorbate coverages at T?200K were determined by NMA to be 0.92 ± 0.05×10¹⁵, 1.0 ± 0.06×10¹⁵ and 1.07 ± 0.10×10¹⁵ CO molecules, NO molecules and D atoms, respectively, per cm². The value of the saturation coverage by CO (θ = 1.0) supports recent models of the (2×1)p1g1 overlayer. The isosteric heat of adsorption of CO is 160 ± 15 kJ mol^?1 in the range 0.2?θ?0.5.     

Unoccupied electronic states on Ni(100) induced by A c(2 × 2) oxygen overlayer

Bremsstrahlung isochromat spectra at hω₀ = 9.7 eV for electrons normally incident on a clean Ni(100) surface are compared to emission from Ni(100) with an ordered c(2 × 2) oxygen overlayer. We observe strong emission due to adsorbate induced antibonding electronic states near E_F and a simultaneous attenuation of previously identified direct bulk interband transitions. The results are in accord with theoretical predictions.     

Adsorption-desorption properties,coadsorption, and surface structural chemistry of chlorine and oxygen on Ag(331)

At 300 K oxygen chemisorbs on Ag(331) with a low sticking probability, and the surface eventually facets to form a (110)?(2 × 1) O structure with ΔΦ = +0.7 eV. This facetting is completely reversible upon O₂ desorption at ~570 K. The electron impact properties of the adlayer, together with the LEED and desorption data, suggest that the transition from the (110) facetted surface to the (331) surface occurs at an oxygen coverage of about two-thirds the saturation value. Chemisorbed oxygen reacts rapidly with gaseous CO at 300 K, the reaction probability per impinging CO molecule being ~0.1. At 300 K chlorine adsorbs via a mobile precursor state and with a sticking probability of unity. The surface saturates to form a (6 × 1) structure with ΔΦ = +1.6 eV. This is interpreted in terms of a buckled close-packed layer of Cl atoms whose interatomic spacing is similar to those for Cl overlayers on Ag(111) and Ag(100). Desorption occurs exclusively as Cl atoms with E_d ~ 213 kJ mol^?1; a comparison of the Auger, ΔΦ, and desorption data suggests that the Cl adlayer undergoes significant depolarisation at high coverages. The interaction of chlorine with the oxygen predosed surface, and the converse oxygen-chlorine reaction are examined.     

Synchrotron X-ray study of sub-monolayers of CF4 on graphite

The quasi-two-dimensional solid and fluid phases of CF₄ physisorbed on the (001) surface of graphite were studied by means of X-ray diffraction at the electron storage ring DORIS, Hamburg. Four crystalline phases were observed, one of which is a registered 2 × 2 structure. On variation of temperature or coverage, the system experiences a phase transition from the 2 × 2 structure to a uniaxially compressed or “stripe” phase. This is unusual among gas-monolayer systems. Only at higher coverages does an isotropically compressed phase appear. At low temperatures we see a complicated structure giving three closely spaced diffraction peaks. We present diffraction data which characterize the melting of the stripe and the hexagonal incommensurate phases.     

The GaAs (001)−c(4×4) and (2×4) reconstructions: A comparative photoemission study

Surface sensitive photoemission measurements of As(3d) core levels give new experimental evidence that the MBE-grown GaAs (0 0 1)?c(4×4) surface is richer in As than the 2×4 surface and that it is produced by an As overlayer chemisorbed onto the As atoms terminating the polar GaAs(0 0 1) surface. A study of As thermal desorption shows that the c(4×4) structure persists over a broad range of surface stoichiometry.     

A LEED and AES study of the adsorption of bromine on W(100) at room temperature

Bromine gas adsorbs atomically on W(100) at room temperature to a saturation concentration of θ = 0.88 relative to the surface tungsten atom density (10¹⁹ m^?2). Below θ ～ 0.4, a c(2 × 2) overlayer is formed. Beyond this a ($\text{3}\text{4}$√2 × √2)R45° structure is preferred and this saturates at θ = 0.67. Higher surface bromine concentrations result in hexagonal variable compression structures on W(100). The sequence begins w structures on W(100). The sequence begins with a c(4 × 2) coincidence mesh which at higher coverages is compressed in one 〈0,1〉 substrate direction. At certain compressions the overlayer achieves p(5 × 2), c(6 × 2), p(7 × 2) coincident configurations and perhaps c(8 × 2) at saturation. This would correspond to θ = 0.875 and is the closest coincidence structure to a perfect hcp overlayer. Bromine prefers a rectangular overlayer geometry on W(100) and compression into an hexagonal array greatly reduces the overlayer stability. The nn repulsions incurred limit room temperature adsorption as the overlayer compresses to perfect hep. Halogen behaviour on W(100) is compared with that on Fe(100). Most differences can be explained in terms of geometrical and bond strength differences but chlorine on W(100) appears to be an exception to this rule.     

The adsorption of PH3 on the Si(111)-(7 × 7) surface: an example of autocatalytic dissociative chemisorption

The dissociative chemisorption of phosphine, PH₃, on the Si(111)-(7 × 7) surface has been examined employing supersonic molecular beam techniques. The initial probability of reaction, S_R,0, has been found to be sensitive to substrate temperature, T_s, where S_R,0 increases sharply by approximately a factor of 4–5 as T_s is increased above 800°C, which corresponds well with the (7 × 7) ↔ “(1 × 1)” phase transition. The reaction probability, S_R, measured as a function of dose for PH₃ reacting on Si(111)-(7 × 7) at T_s < 800°C, exhibits a dramatic increase as the surface is exposed to the PH₃ molecular beam. This unique autocatalytic behavior is consistent with a mechanism in which submonolayer coverages of P(a) are capable of lifting the (7 × 7) reconstruction thus giving rise to a more reactive “(1 × 1)-like” phase. The reaction probability of Si₂H₆ on Si(111)-(7 × 7) is also observed to pass through a maximum with increasing P(a) coverages, and can be explained by considering similar changes in surface structure and reactivity.     

Phase transitions in adsorbed films of barium on W(011)

Phase transitions in barium submonolayers adsorbed on W(011) are studied in a wide range of temperatures and coverages by the LEED technique, including the temperature measurement of the diffraction intensity. The regions of ordered and disordered structures are determined, the result is presented in the form of phase diagram. The temperature dependence of the adfilm Bragg intensity in the low temperature limit (the lowest temperature is 5 K) shows an appreciable slope for all incoherent and almost all coherent structures, except for (3×2). The fact is discussed in terms of the adfilm long- and quasi-long-range order. The disordering of the (3 × 2) lattice near T_c=130 K is the second-order phase transition with the order parameter critical exponent β=0.16. the adfilm is two-phase in the range n=(3.2?3.8)×10¹⁴cm^?2 and singlehase for the rest of the coverages. The effect of the first-order phase transition on the character of the work function change in the two-phase region is discussed.     

A study of the adsorption of oxygen on Ni(111) using auger electron spectroscopy: Chemical shifts and valence spectra

Auger electron spectra have been recorded when oxygen is adsorbed on a Ni(111) single crystal surface. For the coverage range θ < 1, an analysis of the plot of the peak to peak height (H) of the oxygen KVV (516 eV) transition versus the total number of molecules cm^2? impinging on the surface (molecular beam dosing) shows agreement with the kinetic mechanism proposed by Morgan and King [Surface Sci. 23 (1970) 259] for the adsorption of oxygen on polycrystalline nickel films. In this coverage range, no energy shifts of the nickel or oxygen Auger peaks were recorded.At coverages θ > 1 (standard dosing procedure) shifts in the valence spectra M_{2, 3}VV (61 eV) and L₃M_{2, 3}V (782 eV) of ?2.3 eV and ?1.8eV respectively are recorded at 1.4 × 10^?2 torr-sec. Up to these coverages no shift of the L₃VV transition (849 eV) is observed. A chemical shift of ?2.1 eV is recorded in the L₃M_{2, 3}M_{2, 3} Auger transition (716 eV) at 1.4 × 10^?2 torr-sec.In the coverage range θ > 1, shifts in the energy of the oxygen Auger peaks are observed. At 5.8 × 10^?3 torr-sec. the KVV (516 eV) and KL₁V (495.2 ± 0.3 eV) transitions show shifts of ?1.5 eV and ?(1.0 ±0.3) eV respectively. No shift up to this coverage is recorded in the KL₁L₁ (480.6 ± 0.3 eV) transition.     

Sulfur dioxide adsorption and reaction with atomic oxygen on the Ag(110) surface

The adsorption of SO₂ on Ag(110) and the reaction of SO₂ with oxygen adatoms have been studied under ultrahigh vacuum conditions using low energy electron diffraction, temperature programmed reaction spectroscopy and photoelectron spectroscopy. Below 300 K, SO₂ adsorbs molecularly giving p(1×2) and c(4×2) LEED patterns at coverages of one half and three quarter monolayers. respectively. At intermediate coverages, streaked diffraction patterns, similar to those reported for noble gas and alkali metal adsorption on the (110) face of face-centered cubic metals were observed, indicating adsorption out of registry with the surface. A feature at low binding energy in the ultraviolet photoemission spectrum appeared which was assigned to a weak chemisorption bond to the surface via the sulfur, analogous to bonding observed in SO₂-amine charge transfer complexes and in transition metal complexes. SO₂ exhibited three binding states on Ag(110) with binding energies of 41, 53 and 64 kJ mol^?1; no decomposition on clean Ag(110) was observed. On oxygen pretreated Ag(110), SO₂ reacted with oxygen adatoms to form SO_3(a), as determined by X-ray photoelectron spectroscopy. Reacting preadsorbed atomic oxygen in a p(2 × 1) structure with SO₂ resulted in a c(6 × 2) pattern for SO_3(a). The adsorbed SO_3(a) decomposed and disproportionated upon heating to 500 K to yield SO_2(g), SO_4(a) and subsurface oxygen.     

X-ray study of molecular oxygen adsorbed on graphite

We have performed a detailed X-ray diffraction study of O₂ adsorbed on UCAR-ZYX and Le Carbon Lorraine vermicular exfoliated graphite between 15 and 50 K. At least four phases of physisorbed oxygen are found. The monolayer δ phase consists of a centered parallelogram lattice, with the molecular axes parallel to the graphite surface. The data are consistent with a triple point at 26 K. The melting transition at a coverage of one monolayer appears to be first order. At higher coverages the molecules undergo a lying-down to standing-up transition; the higher coverage ζ phase froms an approximately triangular lattice with the molecular axes perpendicular to the graphite surface. Satellite peaks around the (1, 0) Bragg peak indicate, however, that this cannot be a simple triangular lattice; possible explanations include successively incommensurate layers or a sinusoidal density modulation. For coverages in the two-layer region the ζ phase modulation peaks disappear at 37 K, and at 40 K the adsorbed oxygen appears to undergo a first order melting transition into a fluid phase. With increasing coverage, the 2D X-ray diffraction profiles and phase boundaries do not connect smoothly onto those of the 3D α and β phases. At low temperatures (T < 30 K) the ζ phase always coexists with bulk crystallites; for temperatures near the 2D melting transition the 3D peaks are not observable. These data, taken together with the heat capacity results, suggest a wetting transition with only the bilayer lamellar phase or bulk O₂ being stable at low temperatures.