Leed studies of vicinal surfaces of silicon

Clean silicon surfaces inclined at small angles to (111), (100) and (110) planes were investigated by LEED. Surfaces oriented at low angles to the (111) plane contain steps with edges towards [2̄11] or [21̄1̄]. Steps with edges towards [2̄11] have a height of two interplanar distances d₁₁₁ at low temperatures. At 800°C the reversible reconstruction of this step array into the steps of monolayer height takes place. Steps with edges towards [21̄1̄] can be seen at low temperatures only. They are of monolayer height and disappear at annealing in vacuum. Surfaces oriented at low angles to the (100) plane contain steps with (100) terraces and have a height of about two interplanar distances d₁₀₀. Surfaces at low angles to (110) planes are facetted and contain facets of the (47 35 7) type. The information about surface self-diffusion of silicon may be obtained using the kinetic data of structural reconstructions on surfaces close to (111) at different temperatures.     

Iodine adsorption on a platinum stepped surface: Pt(s)[6(111) × (111)]

I₂ adsorption on Pt(s)[6(111) × (111)] surfaces under vacuum and atmospheric pressure conditions was studied by LEED, AES and thermal desorption. In contrast to smooth Pt(111), the surface structures were composed of multiple phase domains having (3 × 3) or ($\text{3}\text{×}\text{3}$)R30° local geometry and structural coincidence of the adjacent terraces. No special stability or instability of iodine adsorption at steps was observed.     

Interaction of CO,CO2 and O2 with nonpolar,stepped and polar Zn surfaces of ZnO

The nonpolar (10$\text{1}$0), stepped (40$\text{4}$1) and (50$\text{5}$1), and the polar (0001) surfaces of ZnO were prepared. Stable unreconstructed nonpolar and stepped surfaces were obtained. LEED analyses showed that the step height and the step width of the stepped surfaces were similar to the theoretical values. The polar surface showed a 1 × 1 LEED pattern of six-fold symmetry after annealing at 500°C, and evidence of a more complicated pattern at 300–400°C. Temperature programmed desorption of CO resulted in the desorption of CO from the stepped and the polar surfaces. However, desorption of CO₂ was observed from the stoichiometric nonpolar surface, and no desorption from the reduced nonpolar surface. CO₂ was also observed by interacting CO with all surfaces at elevated temperatures. A total of four temperature programmed desorption peaks of CO₂, α, β, γ, and δ were observed. The α and β peaks were observed on the nonpolar and the stepped surfaces, and the γ peak was observed on the polar surface. The α peak was assigned to adsorption on a surface ZnO pair, and the β peak was assigned to adsorption on an anion vacancy or a step. While adsorbed water enhanced the β, preadsorbed methanol reduced it. O₂ adsorption was similar on the nonpolar and the stepped surfaces, but was weak on the polar surface.     

Field-ion microscopy of GaAs and GaP

Clean surfaces of GaAs and GaP were studied by field-ion microscope (FIM). Field-ion images with ordered surfaces were first obtained in pure hydrogen, neon-50% hydrogen and pure neon gases at 78 K, by using channeltron electron multiplier arrays (CEMA). The field-ion images of GaAs were quite similar to those of GaP with respect to the surface structure and the image contrast. They showed the anisotropies of the ion emission and the surface structure between the [111] and $\text{[}\text{1}\text{1}\text{1}\text{]}$ orientations. Ring steps expected from a spherical surface were observed on the (111) and {100} planes, but not on the $\text{[}\text{1}\text{1}\text{1}\text{]}$ and {110} planes. The regional brightness of the FIM patterns was discussed in terms of the Knor and Müller model and the atomic and electronic structures of the surface. The image field of these crystals was much lower than that of metals usually used in FIM. For example, the image field strength for the hydrogen and GaAs system was about 1.1 V/Å. The reduction of the field necessary to image was also discussed in terms of the field penetration effect.     

A LEED study of UO2(∼100) vicinal surfaces

The ordering and faceting properties of UO₂(～100) vicinal surfaces have been studied via LEED and Auger measurements. The measurements have demonstrated a reduced tendency for step ordering on UO₂(～100) vicinal surfaces when compared to step ordering on UO₂(～111) vicinal surfaces. The UO₂(～100) vicinal surfaces were observed to decompose irreversibly into low-index facets, including prominent (100) facets, at temperatures below those needed for creation of lowest index faceting on UO₂(～111) vicinal surfaces. These properties suggest that (100) terraces, in contrast to (111) terraces, act as surface diffusion barriers that limit longrange surface communication while growing at the expense of intermediate faceting stages.     

A study of the (00) LEED beam intensity at normal incidence from CdS(0001), Cu(001), Cu(111), and Ni(111)

A Faraday cage apparatus is used for the measurement of the (00) LEED beam intensity, I₍₀₀₎, and the total secondary emission coefficient, δ(E_k), for angles of incidence from 0° ± 2° to 8° ± 2°, with an energy resolution of ± 0.037 of the incident beam energy, in the energy range 1 to 200 eV. The data are normalized and expressed as a fraction of the incident beam intensity. The basic principle of operation is the separation of the incident and specularly diffracted beams in a uniform magnetic field. Monolayer, or in-plane, resonances associated with the emergence of nonspecular beams, as well as beam threshold minima, are observed in I₍₀₀₎ at normal incidence from clean CdS(0001), Cu(111), and Ni(111). Some major differences are observed in the I₍₀₀₎ profiles for the clean (111) surfaces of nickel and copper. All secondary Bragg peaks, except the 2$\text{2}\text{3}$ order, have greater intensities for Ni(111) in the energy range 50–150 eV, thus indicating that the atomic scattering cross-section for electrons in this energy range is larger for nickel than for copper. For the (111) surface of nickel, the (11) resonance is missing, but the (10) resonance and all $\text{1}\text{3}$ order secondary Bragg peaks between the second and fifth orders are observed. For Cu(111) both the (10) and (11) resonances are observed, but the $\text{1}\text{3}$, $\text{2}\text{3}$, 1$\text{2}\text{3}$, and 3$\text{1}\text{3}$ order secondary Bragg peaks are missing in this energy range. These data indicate that multiple scattering with evanescent intermediate waves, or “shadowing”, is predominate on the (111) surfaces on nickel and copper for energies above 30 eV, and that below 30 eV multiple scattering with propagating intermediate waves is predominate on Cu(111). Correlation of the (00) beam intensity profiles from clean Ni(111) at 0°, 2°, and 6° with the intensity profiles of the (10). (1̄0), and (11) non-specular beams is nearly one-to-one from 30 eV to 100 eV, thus supporting the dynamical theories of LEED in which peaks in the (00) beam are expected to occur at nearly the same energies as peaks in the non-specular beams.     

Author index

The mean adsorption lifetimes of F, Cl and Br on (100) and (111) Mo surfaces have been obtained from the first order desorption kinetics observed in low converage conditions (θ < 10^?2 of a monolayer) using a pulsed ionic beam method. The mean adsorption lifetimes τ fit a general expression $\text{τ = τ}{}_0\text{exp}\text{(}\text{E}\text{kT}\text{)}$ in a large temperature range (1700–2400 K) allowing the determination of the binding energies E. The main results of this study are (1) the binding energies decrease from F through Br; (2) the binding energies on both (100) and (111) orientations are similar, E(F)～4.65 eV, E(Cl)～4.15 eV, and E(Br)～3.65 eV. These results are discussed and compared with those previously reported on (100) and (111) Nb and W surfaces. The close binding energies of F, Cl and Br on (100) and (111)^?Nb and Mo surfaces suggest that halogens have a different chemisorption behaviour with respect to O and N.     

The epitaxy of gold on (110) tungsten studied by leed

Growth of gold condensed on the (110) plane of tungsten has been studied using LEED and AES. Three ordered surface structures were observed when condensation takes place at or above 700 K, and no detectable order is seen below this temperature. Structure 1 is developed as the coverage approaches one monolayer and has gold atoms held in the W(110) array with a resultant 2% reduction in gold atom diameter. The second gold layer adopts the Au(111) structure with $\text{Au}\text{[1}\text{2}\text{1]}$ rotated by 2.5° from $\text{W}\text{[1}\text{1}\text{0]}$ and the first gold layer may also be constrained to adopt this structure. Deposition of more gold produces three dimensional crystallites with Au(111)∥W(110) which are double-positioned with their 121 directions parallel to the 121 directions of tungsten. Addition of half a monolayer of oxygen before condensation, completely prevents formation of structures 1 and 2. Instead, at coverages of 3 or more monolayers, three dimensional crystallites develop with Au(111) ∥ W(110) and $\text{Au}\text{[1}\text{2}\text{1] ∥}\text{W}\text{[1}\text{1}\text{0]}$. This behaviour is compared with the reported behaviour of copper and silver on W(110).     

The adsorption of CO,O2, and H2 on Pt: I. Thermal desorption spectroscopy studies

Thermal desorption spectroscopy (TDS) has been used to study the chemisorption of CO, O₂, and h₂ on Pt. It has been found that TDS is quite sensitive to local surface structure. Three single crystal and two polycrystalline Pt surfaces were studied. One single crystal was cut to expose the smooth, hexagonally close-packed plane of the fee Pt crystal (the (111) surface). The other two single crystals were cut to expose stepped surfaces consisting of smooth, hexagonally close-packed terraces six atoms wide separated by one atom high steps (the 6(111) × (100) and 6(111) × (111) surfaces). Only one predominant desorption state was observed for CO and H adsorbed on the smooth (111) single crystal surface, while two predominant desorption states were observed for these gases adsorbed on the stepped single crystal surfaces. The low temperature desorption states on the stepped surfaces are attributed to desorption from the terraces, while the high temperature desorption states are attributed to desorption from the steps. TDS of CO from the polycrystalline foils exhibited some desorption states which were similar to those observed on the stepped single crystal surfaces, indicating the presence of adsorption sites on the polycrystalline foils that were similar to the terrace and step sites on the stepped single crystals. In general, these results suggest a high density of defect sites on the polycrystalline foils which can not be attributed simply to adsorption at grain boundaries. Oxygen was found to adsorb well on the stepped single crystals and on the polycrystalline foils, but not on the smooth (111) single crystal, under the conditions of these experiments. This is attributed to a higher sticking probability for dissociative O₂ adsorption at steps or defects than on terraces.     

LEED studies of clean high Miller index surfaces of silicon

Clean high Miller index surfaces of silicon were studied by LEED at different temperatures. Three types of surfaces were observed depending on orientation: flat surfaces, stepped ones with the (100) and (111) terraces and with the step heights of one or two interplanar distances, and “hill and valley” or facetted surfaces. All clean surfaces of silicon are reconstructed and contain surface structures with the periodicities different from the periodicities in the bulk. At certain temperatures the reversible order-order and order-disorder transitions take place on clean surfaces of silicon.     

Adsorption and desorption of oxygen on stepped tungsten surfaces

The adsorption and desorption of oxygen on stepped tungsten surfaces with orientations close to the (110) orientation and steps parallel to the most densely packed crystal direction ([111]) is studied with low energy electron diffraction, Auger electron spectroscopy, work function measurements and thermal desorption spectroscopy. With increasing deviation from the (110) orientation, an increasing preference for the formation of the p(2 × 1) domain with the densely packed direction parallel to the steps is noted. The adsorption kinetics does not differ markedly from that on the flat (110) surface, however the desorption behaviour at low coverages (θ < 0.3) is quite different. The results are interpreted in terms of the dissociation of a mobile precursor at terrace and step sites, the competition between the two domains during their growth and a step-induced premature transition to the complex structure observed on flat (110) surfaces at $\text{θ ? 8}$. The steps are believed to play also a significant role in desorption.     

Magnetic energy states of shallow acceptors in silicon from spin-dependent luminescence

Quenching of the shallow donor-acceptor radiative transition probability in silicon as a function of the product of electron and hole spin-polarizations enables one to determine neutral acceptor static magnetic susceptibility functions. From these, acceptor g_A-factors for $\text{m}{}_{\mathrm{J}}\text{= ±}\text{3}\text{2}$ levels, with [100]∥H, denoted as g₀ [100] (the outer level g_A factor), have been determined for B, Al, Ga and In acceptors to be 1.10 ± 0.02,0.77 ± 0.02, 0.89 ± 0.02 and 0.56 ± 0.03. The method is insensitive to the inner level g_A factors, g_i, which correspond to the $\text{m}{}_{\mathrm{J}}\text{= ±}\text{1}\text{2}$ levels. The susceptibility measurements were anisotropic and for the [111]∥H orientation could be fit either with an independent set of g₀ [111] values, which turned out to be very different from g_o [100] for B and In, or by assuming g₀ [100] =g₀[111] and introducing a local strain term along 〈111〉 directions. In view of other reports indicating near equality of g₀ [100] and g_o[111], we believe the invoked strain energies are applicable. Their respective values for B, Al, Ga and In are ?0.10, ?0.03, 0.00 and 0.13 meV, consistent with internal strains along 〈111〉 which have as origin the size misfit of the respective acceptors in the silicon lattice. At very high value of H/T, where electron and hole spin-polarizations approach 1, the luminescence does not go to zero as would be expected from an s -envelope function for the acceptor, but approaches a residual luminescence intensity. This fractional residual luminescence is shown to arise from d-function mixing with the predominantly s-function acceptor envelope state, which mixing was first calculated by Schechter. The residual luminescence parameters, which are a measure of the amount of d-function admixture, are respectively for B, Al, Ga and In equal to 0.23, 0.18, 0.16 and 0.08. The progressive decrease of these values with increasing acceptor ionization energies are suggested to be related to s-envelope function stabilization with respect to increased core potentials. Spin-resonance of the donor electrons was optically detected via the unquenching of the luminescence signal, and donor spin-lattice relaxation rates, arising from a neutral donor-neutral acceptor interaction, were determined as a function of D⁰?A⁰ pair separation.     

Superlattices formed by interaction of hydrogen bromide and hydrogen chloride with Pt(111) and Pt(100) studied by LEED,Auger and thermal desorption mass spectroscopy

HBr and HCl react with Pt(111) and Pt(100) surfaces to form adsorbed layers consisting of specific mixtures of halogen atoms and hydrogen halide molecules. Exposure of Pt(111) to HBr yielded a (3×3) LEED pattern beginning at $\text{Θ}{}_{\mathrm{Br}}\text{=}\text{2}\text{9}$ and persisting at the maximum coverage which consisted of $\text{Θ}{}_{\mathrm{<mtext>Br</mtext>}}\text{=}\text{1}\text{3}$ plus $\text{Θ}{}_{\mathrm{<mtext>HBr</mtext>}}\text{=}\text{1}\text{9}$. The most probable structure at maximum coverage, Pt(111)[c(3 × 3)]-(3 Br + HBr), nas a rhombic unit cell encompassing nine surface Pt atoms, and containing three Br atoms and one HBr molecule. On Pt(100) the structure at maximum coverage appears to be Pt(100)[c(2√2 × √2)]R45°-(Br + HBr), $\text{Θ}{}_{\mathrm{<mtext>Br</mtext>}}\text{= Θ}{}_{\mathrm{<mtext>HBr</mtext>}}\text{=}\text{1}\text{4}$; the rectangular unit cell involves four Pt atoms, one Br atom and one HBr molecule. Each of these structures consists of an hexagonal array of adsorbed atoms or molecules, excepting slight distortion for best fit with the substrate in the case of Pt(100). Treatment of Pt(100) with HCl produced a diffuse Pt(100)(2 × 2)-(Cl + HCl) structure at the maximum coverage of Θ_Cl = 0.13, Θ_HCl = 0.11. Exposure of Pt(111) to HCl produced a disordered overlayer. Thermal desorption, Auger spectroscopy and mass spectroscopy provided coverage data. Thermal desorption data reveal prominent rate maxima associated with the structural transitions observed by LEED. Br and HBr, Cl and HCl were the predominant thermal desorption products.     

On the origin of pyramids and cones on ion-bombarded copper surfaces

It is shown that ion-bombardment-induced pyramids or cones on pure Cu surfaces originate mainly, at least in the absence of major surface impurities, from pre-existing asperities. Such features can in principle be formed due to mechanical treatment, to chemical treatment, to inclusions, to surface impurities, or to other causes, the precise details being unimportant. Once asperities are present, regions which have a convex-up curvature should, as argued by Carter, Colligon, and Nobes, evolve into facets having an angle near θ&#x0302; between the beam and the target normal, θ&#x0302; being the angle at which the sputtering coefficient, S(θ), maximizes. By considering the velocity of motion of the facets it follows that, after a critical bombardment dose given approximately by $\text{Nh}{}_1\text{[S(}\text{θ}\text{?}\text{)?S(0)]}$, the asperities will be pyramidal. (N = number density, h₁ = height of asperity.) It is also shown that the following play little or no direct role with Cu in causing pyramids or cones: grain boundaries, chemical etch pits, a Ti impurity on a smooth surface, minor surface impurities due to oxide or hydrocarbon films, annealing an otherwise stable surface. Major hydrocarbon contamination appeared to play a role but it was unclear whether it was a direct or indirect one, that is, whether or not pre-existing roughness was involved.     

Reactive epitaxy of cobalt disilicide on Si(100)

The growth of cobalt disilicide on the Si(100) surface by reactive epitaxy at T=350°C was studied within the 10–40 ML cobalt coverage range. A new method of mapping the atomic structure of the surface layer by inelastically scattered medium-energy electrons was employed. The films thus formed were shown to consist of CoSi₂(221) grains of four azimuthal orientations turned by 90° with respect to one another. This domain structure originates from substrate surface faceting by (111) planes, a process occurring during silicide formation. B-oriented CoSi₂(111) layers grow epitaxially on (111) facets.     

Elastic leed intensity-energy studies of clean (001), (110) and (111) nickel surfaces

Elastic low energy electron diffraction (LEED) intensity-energy (I-E) measurements for clean (001), (110), and (111) nickel surfaces were obtained at room temperature. Surface composition was monitored by Auger spectroscopy. I-E data from 15 to 220 eV were obtained at normal incidence for the non specular beams and for the specular beams at incidence angles from 4° to 20° on the 0° and 45° azimuths of (001), on the 0° and 90° azimuths of (110), and on the 0° azimuth of (111) nickel. Normalization of the data was performed electronically during data acquisition. Intensities were calibrated with the use of a shielded, biased Faraday collector. The effects of instrumental and experimental uncertainties were examined and minimized to obtain intensities accurate to ± 15 %, energy scales accurate to ± 0.35 eV, and incident and azimuthal angles accurate to ± 0.25° and ± 1.0° respectively.All nickel surfaces have I-E spectra which are characteristic of strong multiple scattering. Angular evolution features for (001) and (110) spectra may be correlated with intraplanar resonances associated with the onset of propagating beams. Only the (001) surfaces were found to have pronounced, sharp resonance features associated with surface barrier resonances and inelastic loss processes. Kinematic analysis of the Lorenzian-shaped I-E peaks on all surfaces in consistent with surface expansion using either an energy-dependent or a constant inner potential of 10.75 ± 0.5 eV. The widths of these same peaks on all surfaces were found to vary as $\text{E}{}^{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}$ above 40 eV and $\text{E}{}^{\mathrm{<mtext>1</mtext><mtext>3</mtext>}}$ below.     

Room temperature adsorption and growth of Ga and In on cleaved Si(111)

Low energy electron diffraction (LEED), Auger electron spectroscopy (AES) and photoemission yield spectroscopy (PYS) measurements have been performed on a set of ultrahigh vacuum cleaved Si(111) surfaces with different bulk dopings as a function of Ga or In coverage θ. The metal layers are obtained by evaporation on the unheated substrate and θ varies from zero to several monolayers (ML). First, the 2×1 reconstruction of the clean substrate is replaced by a $\text{3}\text{×}\text{3}$ R30° structure at $\text{1}\text{3}$ ML, meanwhile the dangling bond peak at 0.6 eV below the valence band edge E_vs is replaced by a peak at 0.1 eV for Ga or 0.3 eV for In, below E_vs. At the same time, the ionization energy decreases by 0.4 eV (Ga) or 0.6 eV (In), while the Fermi level pinning position gets closer to the valence band edge by about 0.1eV. Upon increasing θ, new LEED structures develop and the electronic properties keep on changing slightly before metallic islands start to grow beyond θ ～1 ML.     

The angular dependence of flux,mean energy and speed ratio for D2 molecules desorbing from a Ni(111) surface

Experimental data are presented for the angular dependence of the relative flux, the mean energy and the speed ratio of deuterium molecules desorbing from a Ni(111) crystal surface at a surface temperature of T_s = 1143 K and at sulphur coverages ranging between 30% and less than 2% of a monolayer.The angular flux distribution is sharply peaked in the forward direction (cos^dθwith 3 ? d ? 5) and the mean energy 〈E〉 of the desorbate depends strongly on the desorption angle θ. For normal desorption $\text{(θ = 0°)}\text{〈E〉}\text{2k}$ is about 700 K higher than T_s and for glancing angles (θ = 80°) it decreases to about 400 K below T_s The results obtained on sulphur free and sulphur covered Ni(111) surfaces are compared with our former data on polycrystalline nickel. The main differences in the kinetic features can be ascribed to the surface roughness. Accordingly, the angular distributions of flux, mean energy, and speed ratio, which deviate strongly from the Knudson and Maxwellian law, do not seem to depend considerably on sulphur coverage and surface structure. A qualitative explanation for these deviations is presented using the principle of detailed balancing.