Search

We report on the various phases formed by sulfur adsorbed on Au(111), at less than 1 monolayer (ML) coverage, as monitored by low-energy electron diffraction (LEED). The phases transform from one to another via coexistence regions. The observation of ordered phases was matched to the coverage information obtained from Auger measurements, using as a calibration point the S coverage of 0.28 ML, corresponding to the sharpest (5 × 5) LEED pattern observed. The influence of emitting filaments (the Auger beam, ion gauge, or LEED electron beam) upon the adsorption itself is discussed. LEED structural investigations were performed on two particular S–Au phases. The analysis of the Au(1 × 1)–S structure provided the values for the first four gold interlayer spacings, showing a 1% expansion of the top layer, interpreted as a remnant of the expansion of the Au(111)–(22 × √3) reconstructed phase. The analysis of the Au(5 × 5)–7S phase confirmed the fcc adsorption site of the S atoms reported previously, and resulted in an average S–Au distance of 1.57 ± 0.10 A, and an S–Au bond length of 2.29 ± 0.07 A.

The surface chemical properties of Y, Ti, Zr and Hf relevant to exoelectron and ion emission have been characterised by a range of techniques (XPS, UPS, AES, LEED, TDS, area-average and threshold φ). The chemisorptive properties of chlorine, bromine, sodium and rubidium are described, and (alkali + halogen) coadsorption has been studied also. Evolution of the trihalide corrosion layer is examined as are its decomposition properties; these features are rationalised in thermochemical and kinetic terms. In certain cases (Y/Na, Y/Rb, Hf/Rb) alkali adsorption results in a system work function equal to that of the bulk alkali at one-half monolayer coverage. Coadsorption of alkali and halogen leads to some particularly interesting observations. Substantial discrepancies (~ 1 eV) are found between the area-average work function \gf and the photoelectric threshold work function (φ∗); over certain coverage regimes, φ∗ actually decreases as \gf increases. This implies the formation of very low work function (< 2 eV) “exopatches”. Desorption from such systems does not occur via the alkali halide — in marked contrast with the behaviour of less electropositive substrates; this can be accounted for. The presence of halogens substantially increases the adsorption strength of the alkali. Models for the various surface phases are advanced and discussed.

Graphitic nanofibers have been synthesized on Ni(1 0 0) substrates via DC plasma chemical vapor deposition using a gas mixture of hydrogen and methane. The diameter of the fibers is in the range 50–200 nm, controlled by the size of the nickel tip of each fiber. The substrate temperature during deposition was 350–400°C. No special treatment of the substrate was employed to promote the nucleation of nanofibers, though scratches on the substrate surface were found to be favorable for the nucleation of the fibers. Scanning electron microscopy, transmission electron microscopy, energy dispersive spectroscopy and Raman spectroscopy were used to characterize the fibers obtained. The effect of the plasma on the growth of the fibers low temperature is discussed.

Bias-assisted hot filament chemical vapour deposition (HFCVD) and ion-assisted pulsed laser deposition (IA-PLD) have been employed to deposit carbon nitride films. Crystalline C-N films composed of α- and βC 3 N 4 , as well as some unpredicted C-N solids have been synthesized on Ni(100) substrates via HFCVD using a gas mixture of nitrogen and methane. The crystal constants of the α- and β-C 3 N 4 phases, obtained via X-ray diffraction, coincide well with those predicted theoretically; additionally, cone-shaped crystals are observed on the Si(100) substrates. Similarly, high density cone-shaped (Si)-C-N crystals have been obtained on Si(100) substrates via ion-assisted pulsed laser ablation of a carbon (graphite) target intersecting a nitrogen ion beam. Amorphous C-N films were also produced using this method.

We have examined the properties of several isochromat detectors and selected a Galileo 4800 series channel electron multiplier/CaF 2 combination for use in IPES experiments on clean and CO covered Ru(0001). These experiments, performed at T = 300 K and at an off-normal angle of incidence, showed a bulk, a crystal-induced surface state and an image potential state for clean ruthenium with the image state being suppressed at intermediate CO coverages and re-emerging with the formation of ordered overlayers. Two new features emerge upon CO adsorption. They are attributed to a splitting of the 2π ∗ state into bonding and anti-bonding components.

We have performed scanning tunneling microscopy and spectroscopy on the GaAs(001)(2 × 4) surface. The measurements were performed on arsenic capped samples grown ex-situ by MBE. The n-doped epilayers were grown on flat GaAs(001) n-doped wafers and substrates cut 2° towards (010). The epilayers were terminated with 100 A of intrinsic material. Using STM we establish that capping does not compromise the surface quality of the (2 × 4) reconstruction. We find that provided the ex-situ MBE grown surfaces are of sufficient quality, the arsenic cap can be stripped off to yield surfaces equal to in-situ samples. Results from vicinal samples indicate that As from the cap adheres more strongly to step edges, which can be removed by subsequent annealing. The surfaces of all the samples studied were a mixture of (2 × 4) and c(2 × 8) domains. Tunneling I–V curves are interpreted as primarily due to tunneling out of the doubly occupied lone-pair states located on each surface As atom. Compared with GaAs(110), the spectra are much less sensitive to the distribution of empty states above EF. This behavior was independent of spatial location on the surface. The insensitivity to Ga derived empty states persists even when spectra are obtained above exposed Ga sites in the missing dimer rows where they may be expected to contribute. The measurements serve to locate the Fermi level on these samples pinned 0.6 ± 0.2 eV above the valence band maximum.

Various films have been deposited on gold-coated quartz crystals as part of a quartz crystal microbalance. In situ measurements of the effects of atomic oxygen on the net mass change in these materials have been determined by observing the frequency shift of the oscillator in conjunction with a mass-frequency calibration. Films were exposed to a neutral atomic oxygen beam produced by supersonic expansion of a lightly seeded helium plasma. Atomic oxygen fluxes as large as 1017 cm-2 s-1 and incident kinetic energies up to 2.5 eV were typically used. Based on an assumed reaction path between amorphous carbon (a-C) and atomic oxygen, the reaction probability is 0.01. Hydrogenated amorphous carbon (a-C:H) films show a three-fold increase in etch rate. Layered coatings of amorphous carbon films can be identified by their differing etch rates. Amorphous carbon films were used to show how reaction rates vary with changes in reaction parameters. A 25% reduction in reactivity is found for a 25% reduction in incident kinetic energy. Synergistic effects were identified by comparing mass loss rates during exposure of amorphous and hydrogenated amorphous carbon films to a neutral atomic oxygen beam and to an oxygen plasma. Plasma-deposited SiOx, SiNx, a-Si:H and plasma polymer hexamethyldisiloxane are shown to be quite inert and are useful as protective coatings. The protective nature of a coating may be ascertained by overcoating a film of reactive material as shown for plasma-deposited silicon oxide and silicon nitride over amorphous carbon films. Films of aluminium have a uniform and adherent oxide layer demonstrated to be stable in atomic oxygen environments. This is in contrast to silver films, which indicate that a protective oxide layer is not formed.

Deposition of diamond-like amorphous carbon films has been carried out primarily by high frequency plasma discharge and/or ion beam methods using hydrocarbon gases. Laser ablation has been used recently for film deposition of elemental materials1,2 and the very rapid energy input can potentially produce novel materials due to the non-equilibrium conditions created at the evaporation source and the possibility for photo-chemical processes.

An improved version of the technique of optical simulation of LEED patterns is presented. Examples of computer generated comensurate, incommensurate and statistical surface models and their optical diffraction patterns are shown.

The adsorption of cesium on the Ru(0001) surface at T = 300 K has been studied by LEED, Auger spectroscopy, work function measurements, thermal desorption and optical simulation techniques. Work function measurements indicate a dipole-dipole interaction ordering adsorbates in the low coverage regime which is in agreement with other data published for similar systems. In the high coverage region the work function curve reaches a local maximum at 1.9 eV before its value decreases to that characteristic of bulk cesium. Auger intensity measurements show an apparent change in the slope of the uptake curves for Cs(47 eV) and Ru(231 eV) for the same cesium dose at which the local maximum in the work function curve occurs. Three peaks are found in the TDS results at desorption temperatures 1130 K, 690 K and 380 K, respectively. The LEED patterns obtained show that some of the initially observed (2 × 2) diffraction spots at the coverage θ = 0.25 start splitting as the cesium dosing continues. Optical diffraction experiments lead to a model of the cesium overlayer near monolayer coverage that explains the spot splitting as due to the existence of separate (2 × 2) domains, out-of-phase with respect to each other. The analytical formula for electron diffraction spot positions in the presence of such anti-phase domains is developed. Detailed examination of the model shows that the borders between anti-phase domains must lie along three specific directions on the substrate, defined by the criterion of minimum energy of the attractive interaction between bordering domains. This also implies that there are open areas (not covered by cesium atoms) at domain boundaries near the completion of the first layer with the minimum spacing between adjacent domains equal to ( 7 2 )a (a is the 2 × 2 mesh size), which explains the local maximum in the work function at this coverage.

Experimental data for electron and photon emissions during the chemisorption of chlorine on pure and sodium-doped zirconium are presented, and a quantitative model is given which adequately describes the gross features of the electron emission process. The model is based on invoking a tunnelling mechanism and relaxation by a radiationless (Auger) process, and draws heavily on previous successful theories for Auger ion neutralization to predict the energy distribution functions for subsequent emissions. For a high electron affinity species such as chlorine, the observed chemisorptive emission yields are strongly dependent on work function varying from 10 −6 to 10 −2 electrons per incident atom, and following a cubic dependence on the excess energy available. Chemisorptive luminescence is also observed, at levels of ~10 −8 photons per incident atom in accordance with previous estimates. A self-consistent kinetic analysis is presented to describe the rapid decay of these emission processes following the onset of chemisorption.

The adsorption of oxygen on a polycrystalline tungsten surface at ~300 K has been studied by means of electron stimulated desorption (ESD) Although precision gas dosing was not employed, the initial sticking probability for dissociative adsorption appears to be essentially unity, while the variation with coverage suggests that a high degree of order exists and that precursor state kinetics are significant. A most noticeable and reproducible discontinuity in ESD parameters occurs at a fractional coverage θ ~ 0.8 (exposure ~ 1.4 × 1015moleculescm2 incident) which is interpreted as an order-disorder transition within a single (β1) chemisorption state, and results in an increase in the ionic desorption cross-section by a factor of ~ 1.26. A discussion of the adsorption kinetics and the disorder transition is given in terms of current models of dissociative adsorption which include the effects of nearest neighbour lateral interactions.

The adsorption of potassium on the Ru(0001) surface at 450 K has been studied by LEED, Auger spectroscopy, work function change measurements and neutral atom and ion thermal desorption. The results reveal the existence of a new, ionic desorption peak in the high temperature range and the effects of diffusion of potassium into the bulk at 450 K and above. Work function measurements indicate a dipole-dipole interaction ordering the adiayer over the low coverage range (from θ = 0 toθ = 0.15) and covalent bonding leading to the depolarization of the surface layer at higher coverages ( α = 28 A 3 at θ = 0.37 ). Thermal desorption of ions shows a slight increase in activation energy (from 334 to 343 kJ/mol) for the coverage range studied. LEED results show a series of ordered, commensurate structures formed on the surface starting with the (2 × 2) phase for θ = 0.25 and ( 3 × 3 )R30° for θ=0.33. For higher coverages, a mixed phase consisting of “islands” of ( 3 × 3 )R30° and ( 7 3 × 7 3 ) R±11° structures is observed until the maximum obtainable coverage of 0.37 is reached when the overlayer transforms entirely into a two-domain ( 7 3 × 7 3 ) R±11° pattern.

The adsorption of sodium on the Ru(0001) surface at 300 K and above has been studied by LEED, Auger spectroscopy, thermal desorption and work function measurements. A new high temperature peak observed in the thermal desorption spectra was identified as due to the desorption of ions occurring at very low coverages. In the high coverage range new LEED patterns were observed corresponding to the formation of ordered first and second layers of sodium. Models of “non-planar” commensurate structures (23 × 23)R30° for the first adlayer and (43 × 43) for the second are proposed. A reversible change in the LEED patterns was found to occur at T = 600 K. This effect is explained as a result of the existence of two types of binding states in the first layer due to the three- and six-fold symmetry sites on the substrate. The sodium atoms in those states may be considered as two separate components and, thermally activated, produce diffraction patterns as if they belonged to two different layers.

The chemisorption and desorption of K at the (1010) surface of Ru has been investigated by LEED, Auger spectroscopy, Δφ, neutral atom and ion desorption measurements. The results reveal that substantial K-enrichment of the immediate subsurface region occurs for small potassium coverages at 430 K. Desorption of neutral K atoms is characterised by a monotonic fall in activation energy from 280 kJ mol −1 at zero potassium coverage to 180 kJ mol −1 near the monolayer coverage point. Analysis of the K + desorption data yields a value for the activation energy to atomic K desorption which agrees well with that determined directly from the neutral atom desorption spectra. Work function measurements indicate a constant adatom dipole moment of ~ 3.9 D over the coverage regime of 0 ⩽ θ(K) ⩽ 0.18. At higher coverages, the alkali overlayer depolarises (α = 30 A 3 ), a near metallic overlayer being formed at completion of the first monolayer (θ(K) = 0.67). LEED shows the existence of a series of ordered phases for 0.36 ⩽ θ(K.) ⩽ 0.67. An examination of the splitting of the fractional-order beams indicates that the Ru(1010) surface can accommodate increasing K coverage by two distinct mechanisms: growth of c(2 × 2) domains when θ(K) ⩽ 0.5 and uni-axial compression along [1210] when θ(K) ⩾ 0.5.

Exoelectron and ion emission during chemisorption has been studied under UHV conditions for a variety of halogens (X = chlorine, bromine) and group IIIA (Y) and group IVA (Ti, Zr, Hf) surfaces, both clean and with alkali (M = Na, Rb) predosing. The experimental data include absolute yields and energy distribution functions, as well as the kinetics of the emission processes. The variation in electron yield over four orders of magnitude is acceptably described by a previous model invoking Auger relaxation, as are certain aspects of the energy distributions. Ionic species ejected during the reaction include M + , X − and MX 2 − , the latter exhibiting a kinetic behaviour strongly correlated to the formation of low work function “exopatches”, believed to consist of a polar halide overlayer.

The chemisorption of chlorine and sodium and their co-adsorption on polycrystalline foils of yttrium and scandium has been characterised by Auger electron spectroscopy, thermal desorption spectroscopy and work function measurements, preparatory to an investigation of exo-electron and ion emission during halogen chemisorption. Both monohalide and trihalide evolution is observed. An analysis of the thermochemistry indicates that a revised value of the bond strength of gaseous ScCl is warranted, to a value of at least 502 kJ mol −1 .

Chloride growth on zirconium is investigated by means of the novel thermal desorption spectroscopy of both neutral and negatively ionized products, and interpreted in terms of a single ionic adsorption model. The monohalide desorption energy decreases from a low-coverage value of 320 kJ to 255 kJ mol −1 at the monolayer level, at which point trihalide growth continues indefinitely, indicative of a corrosion layer. In contrast, the desorption energy for the chloride ion remains essentially independent of coverage, at 293 kJ mol −1

The preparation is reported of a series of metal(II) complexes of a quinquidentate ligand produced in situ or by complexation with the Schiff-base condensation-product of two moles of 2-acetylpyridine with 3,3′-iminobispropylamine. From the in situ synthesis of the Ni(II) compound only a quadridentate, mono-Schiff-base complex with coordinated acetyl-pyridine is isolable. The effect of varying the triamine used in this preparation is described and rationalised by consideration of ligand-strain, the effect of stability-constants on ligand-dissociation, and the influence of the acetyl(as opposed to a formyl-) group. Analytical data, infra-red studies, magnetic moments and solution-spectra of the complexes are described, and the interconversion of the two types of Ni(II) complex investigated.

The effects of energy broadening on the spontaneous emission of electrons during chemisorption (“chemisorptive emission”) are considered using an approach formerly used with success by Hagstrum for the related process of ion neutralization. Broadening is assumed to arise from two sources: electron tunnelling occurs over a range of distances from the surface, and the lifetime of the initial state is finite. Experimentally determined electron yields for electronegative gases adsorbing on highly electropositive metals are used to determine appropriate broadening parameters, which generally agree well with resonance widths used by other authors for similar systems.

A kinetic study of the rapid iodination of hexaphenyl ditin is reported. The reaction displays simple second-order kinetics and the reaction is radical-free. Other methods, including flash photolysis, electron spin resonance and reaction with 1,1-diphenyl-2-picrylhydrazyl have been studied in an effort to detect radicals derived from hexaphenylditin: these methods indicate that there is no detectable tendency to form radicals of the R 3 M· type.

We have used a titrimetric to study the rates of hydrolysis of a series of acetates R 3 MOAc (M = S or Ge; R = Ph, iPr, nPr, c-hex) in dioxane. The effects of the concentration of water, temperature and catalysis by potassium acetate have been studied and are discussed in terms of the proposed mechanism for the hydrolysis, in which bond-breaking processes appear to be important. High orders in the water concentration are observed and the germanium acetates are much more reactive than the silicon acetates.

The preparation and physical properties of a number of complexes formed between divalent nickel and aromatic diazo compounds of type (substituted phenyl)-azo-2-naphthol are described. The abnormal magnetic moment of these complexes in the solid state is explained in terms of incorporation of water to give six-coordinated molecules. Experimental evidence in support of this hypothesis has been obtained by the use of X-ray, magnetic, and infrared spectroscopic methods.

Trisacetylacetonato-germanium (IV) tetrachloroferrate (III) was prepared by the reaction of Ge(aca) 2 Cl 2 and anhydrous FeCl 3 : Ge(aca) 3 FeCl 4 can be prepared from GeCl 4 and acetylacetone in the presence of ferric chloride and Ge(aca) 2 Cl 2 is shown to be possible intermediate in this reaction. Conductance experiments indicate that Ge(aca) 3 FeCl 4 is a 1:1 electrolyte in acetonitrile, nitromethane and nitrobenzene. The ultra-violet and the visible spectra indicate the presence of the FeCl 4 − anion and the infra-red spectra confirm the presence of FeCl 4 − . The trisacetylacetonato-germanium compound may be formulated as [Ge(aca) 3 + ][FeCl 4 − ].

The series of compounds (n−Pr) 3 MO.CO.CH (3−n) Me n M = Si and Ge = O to 3) were prepared and the rates of hydrolysis measured in aqueous dioxane. The entropy of activation is shown to be the main factor determining the rate differences. pK a Values of the substituted acetic acids correlate well with log(k 1 ). A transition state similar to a solvated proton and a solvated carboxylate ion is proposed.

Previous work on exoelectron and ion emission during chemisorption under UHV conditions has been extended to include halogen (chlorine, bromine) adsorption on group IIIA (Y, Sc) and group IA (Zr) surfaces, both clean and with Na predosing. Ionic species observed include Na + Cl − and NaCl − 2 . By means of double dosing experiments, it is shown conclusively that the latter dichloride ion is formed by direct titration of Na by Cl − 2 , and that halogen dissociation is not a precursor step.