Influence of interface structures on Sn thin film growth on Si(1 1 1) surface

Using coaxial impact collision ion scattering spectroscopy, we have investigated Sn thin film growth on Si(1 1 1)√3×√3-Sn and hydrogen-terminated Si(1 1 1) surfaces at room temperature. Sn formed crystalline film with β-Sn structure on Si(1 1 1)√3×√3-Sn surface, but on the hydrogen-terminated Si(1 1 1) surface, the epitaxial growth of Sn thin film was disrupted, and Sn grew as a polycrystalline film. The growth orientational relationship of the Sn film grown on Si(1 1 1)√3×√3-Sn surface was found to be

Full-size image

. In the works, we found that interface structure plays a decisive role for the growth mode, crystallinity, and growth orientation of the growth of thin film.     

Vanadium oxide nanostructures on Rh(1 1 1): Promotion effect of CO adsorption and oxidation

The adsorption of CO and the reaction of CO with pre-adsorbed oxygen at room temperature has been studied on the (2 × 1)ORh(1 1 1) surface and on vanadium oxideRh(1 1 1) “inverse model catalyst” surfaces using scanning tunnelling microscopy (STM) and core-level photoemission with synchrotron radiation. Two types of structurally well-defined model catalyst V₃O₉Rh(1 1 1) surfaces have been prepared, which consist of large (mean size of 50 nm, type I model catalyst) and small (mean size <15 nm, type II model catalyst) two-dimensional oxide islands and bare Rh areas in between; the latter are covered by chemisorbed oxygen. Adsorption of CO on the oxygen pre-covered (2 × 1)ORh(1 1 1) surface leads to fast CO uptake in on-top sites and to the removal of half (0.25 ML) of the initial oxygen coverage by an oxidation clean-off reaction and as a result to the formation of a coadsorbed (2 × 2)O + CO phase. Further removal of the adsorbed O with CO is kinetically hindered at room temperature. A similar kinetic behaviour has been found also for the CO adsorption and oxidation reaction on the type I “inverse model catalyst” surface. In contrast, on the type II inverse catalyst surface, containing small V-oxide islands, the rate of removal of the chemisorbed oxygen is significantly enhanced. In addition, a reduction of the V-oxide islands at their perimeter by CO has been observed, which is suggested to be the reason for the promotion of the CO oxidation reaction near the metal-oxide phase boundary.     

Strain due to nickel diffusion into hydrogen-terminated Si(1 1 1) surface

We studied the affection of thin (i.e., 0.2–0.8 nm) Ni films on hydrogen-terminated Si(1 1 1) substrate surface by using strain-sensitive X-ray diffraction. It was reported that Ni deposition onto hydrogen-terminated Si surface apparently does not cause film growth, but rather diffuses into the Si crystal, creating an “Ni diffusion layer” up to Ni deposition 0.8 nm thick. Measured rocking curves of the Si 1 1 3 reflection and integrated intensities of the rocking curves for the substrate provide information about the evolution of the strain field introduced near the substrate surface during Ni diffusion into the substrate. Comparing the measured and calculated rocking curves indicates that compression of the {1 1 1} spacing of the Si occurs gradually up to an Ni thickness of 0.6 nm, and that above this thickness, strain relaxation occurs.

We found that the slope of the integrated intensity of the rocking curve versus X-ray wavelength correlates to the strain field near the surface, in the same way that the shape of the rocking curves correlate to the strain field near the surface. Dynamical diffraction calculations indicate that measurement of the slope of the integrated intensity of the rocking curve versus X-ray wavelength is useful for strain analysis, because the dependence is not only sensitive to strain fields, but is also insensitive to the effect of absorption by the overlayer, which otherwise would cause deformation of the shape of the rocking curve.     

Surface structure evolution during Sb adsorption on Si(1 1 1)–In(4×1) reconstruction

The Sb adsorption process on the Si(1 1 1)–In(4×1) surface phase was studied in the temperature range 200–400 °C. The formation of a Si(1 1 1)–InSb (2×2) structure was observed between 0.5 and 0.7 ML of Sb. This reconstruction decomposes when the Sb coverage approaches 1 ML and Sb atoms rearrange to and (2×1) reconstructions; released In atoms agglomerate into islands of irregular shapes. During the phase transition process from InSb(2×2) to Sb (θ_Sb>0.7 ML), we observed the formation of a metastable (4×2) structure. Possible atomic arrangements of the InSb(2×2) and metastable (4×2) phases were discussed.     

A novel (8 × 4) superstructure as precursor to the c(4 × 4) phase during Sb/Si(0 0 1) desorption

The role of kinetics in the superstructure formation of the Sb/Si(0 0 1) system is studied using in situ surface sensitive techniques such as low energy electron diffraction, Auger electron spectroscopy and electron energy loss spectroscopy. Sb adsorbs epitaxially at room-temperature on a double-domain (DD) 2 × 1 reconstructed Si(0 0 1) surface at a flux rate of 0.06 ML/min. During desorption, multilayer Sb agglomerates on a stable Sb monolayer (ML) in a DD (2 × 1) phase before desorbing. The stable monolayer desorbs in the 600–850 °C temperature range, yielding DD (2 × 1), (8 × 4), c(4 × 4), DD (2 × 1) phases before retrieving the clean Si(0 0 1)-DD (2 × 1) surface. The stable 0.6-ML (8 × 4) phase here is a precursor phase to the recently reported 0.25-ML c(4 × 4) surface phase, and is reported for the first time.     

Highly stable passivation of a Si(1 1 1) surface using bilayer-GaSe

As a stable and ‘epitaxial’ passivation of a Si surface, we propose the bilayer-GaSe termination of a Si(1 1 1) surface. This surface is fabricated by depositing one monolayer of Ga on a clean Si(1 1 1) surface and subsequent annealing in a Se flux at around 520 °C, which results in unreconstructed 1×1 termination of the Si(1 1 1) surface by bilayer-GaSe. We found by scanning tunneling microscopy observation that slow cooling of the clean Si(1 1 1) surface from 850 to 520 °C with simultaneous deposition of a Ga flux results in better termination of the Si(1 1 1) surface. It was also found that this surface is stable against heating around 400 °C in O₂ atmosphere of 3×10⁻³ Pa. By utilizing these properties of the bilayer-GaSe terminated surface, we have succeeded in fabricating ZnO quantum dots on this substrate.     

STM and LEED observations of erbium silicide nanostructures grown on Si(1 0 0) surface: atomic-scale understandings

In this work, erbium silicide is grown on the Si(1 0 0) surface by depositing Er onto the substrate and annealing at 600–700 °C. Many nanowires of Er silicide are formed with lengths in the range 10–100 nm. The formation and evolution of this nanostructure are investigated at atomic scale directly with scanning tunneling microscopy and low-energy electron diffraction. The direction of these nanowires is found perpendicular to that of Si dimer rows. It is shown that Er coverage and annealing temperature have an effect on the formation of nanowires. On the surface between nanowires, new (5×2) and c(5×4) reconstructions are observed, giving an implication to understand the growth behaviors of Er silicide on Si(1 0 0) surface.     

Observation of triangular terraces and triangular craters of CaF2 film on Si(1 1 1) substrate

We have investigated the growth mode and surface morphology of CaF₂ film on Si(1 1 1)7×7 substrate by reflection high-energy electron diffraction (RHEED) using very weak electron beam and atomic force microscopy (AFM). It was found by RHEED intensity oscillation measurements and AFM observations that three-dimensional (3D) islands grow at RT; however, rather flat surface appears with two-dimensional (2D) islands around 300 °C. Especially, at high temperature of 700 °C, characteristic equilateral triangular terraces (or islands) with flat and wide shape grow with the tops directed toward [1 1 −2] of substrate Si(1 1 1). On the other hand, the desorption process of the CaF₂ film due to electron stimulated desorption (ESD) was also examined. It was found that the ESD process at 300 °C forms characteristic equilateral triangular craters on the film surface with the tops (or corners) directed toward [−1 −1 2] of substrate Si(1 1 1), provided that the film was grown at 700 °C.     

Oxygen induced reconstruction of (h11) and (100) faces of copper

A LEED and AES study on oxygen adsorption on Cu(100) and (h11) faces with 5 h 15 has been performed under various adsorption conditions (220 K T 670 K and 1 × 10⁻⁸ P 6 × 10⁻⁵ Torr of oxygen). The dependence of adsorption temp on the oxygen surface superstructures is pointed out. At least, three oxygen surface states exist on a Cu(100) face. For low temperature exposures to oxygen, under conditions of slow surface diffusion, on the (100) face, two oxygen surface phases exist: a “four spots” and a c(2 × 2) superstructure, both observed even at saturation coverage; on all the stepped faces, a c(2 × 2) appears and no faceting is observed. For high temperature exposures, on the (100) face, two oxygen superstructures are observed, a “four spots” followed by a (2√2 × √2)R45° at higher coverages; on all the stepped faces, surface diffusion is activated and oxygen induced faceting occurs. The appearance of faceting is associated with the onset of the formation of the (2√2 × √2)R45° structure on the (100) face. The oxygen induced faceting and the oxygen surface meshes are reversible with coverages. At saturation coverage, a non-reversible surface transition between the c(2 × 2) and (2√2 × √2)R45° superstructures is observed at 420 ± 20 K. The importance of impurity traces on the surface meshes is emphasized. Oxygen coverage at saturation is independent of the studied faces and adsorption temperature. Faceting occurs at a critical coverage value, whatever the stepped faces and adsorption temperature are. Models of the oxygen structure on the (h10) stepped faces are discussed.     

Adsorption of oxygen and carbon dioxide on cesium-reconstructed Ag(1 1 0) surface

The adsorption of oxygen and carbon dioxide on cesium-reconstructed Ag(1 1 0) surface has been studied with scanning tunneling microscopy (STM) and temperature programmed desorption (TPD). At 0.1 ML Cs coverage the whole surface exhibits a mixture of (1 × 2) and (1 × 3) reconstructed structures, indicating that Cs atoms exert a cooperative effect on the surface structures. Real-time STM observation shows that silver atoms on the Cs-covered surface are highly mobile on the nanometer scale at 300 K. The Cs-reconstructed Ag(1 1 0) surface alters the structure formed by dissociative adsorption of oxygen from p(2 × 1) or c(6 × 2) to a p(3 × 5) structure which incorporates 1/3 ML Ag atoms, resulting in the formation of nanometer-sized (10–20 nm) islands. The Cs-induced reconstruction facilitates the adsorption of CO₂, which does not adsorb on unreconstructed, clean Ag(1 1 0). CO₂ adsorption leads to the formation of locally ordered (2 × 1) structures and linear (2 × 2) structures distributed inhomogeneously on the surface. Adsorbed CO₂ desorbs from the Cs-covered surface without accompanied O₂ desorption, ruling out carbonate as an intermediate. As a possible alternative, an oxalate-type surface complex [OOC–COO] is suggested, supported by the occurrence of extensive isotope exchange between oxygen atoms among CO₂(a). Direct interaction between CO₂ and Cs may become significant at higher Cs coverage (>0.3 ML).     

Method for independent control of particle size and distance in rhodium epitaxy on TiO2(110)-(1×2) surface An STM study

A method for independent control of the particle size and distance is presented for rhodium epitaxy on TiO₂(110)-(1×2) surface. The real space imaging of the surface morphology was performed by scanning tunneling microscopy. The amount of the deposited rhodium was checked by Auger electron spectrometry. The method consists of two steps: (i) evaporation of 0.001–0.050 ML equivalent of rhodium at room temperature with a post-annealing at 1100 K (“seeding”); (ii) post-deposition of rhodium for growing of the Rh nanoparticles formed in step (i) (“growing”). The mechanism of this procedure is based on the large difference of the surface diffusion coefficient between Rh adatoms and Rh nanocrystallites larger than 1–2 nm. In the first step the average distance between the metal particles is controlled in the range 5–200 nm, the second step determines the particles size (2–50 nm). This work demonstrates that the diffusion processes of metal nanoparticles of different sizes and the growing modes of the crystallites can be studied in detail by application of seeded surfaces.     

On the microscopic origin of the kinetic step bunching instability on vicinal Si(0 0 1)

A scanning tunneling microscopy/atomic force microscopy study is presented of a kinetically driven growth instability, which leads to the formation of ripples during Si homoepitaxy on slightly vicinal Si(0 0 1) surfaces miscut in [1 1 0] direction. The instability is identified as step bunching, that occurs under step-flow growth conditions and vanishes both during low-temperature island growth and at high temperatures. We demonstrate, that the growth instability with the same characteristics is observed in two dimensional kinetic Monte Carlo simulation with included Si(0 0 1)-like diffusion anisotropy. The instability is mainly caused by the interplay between diffusion anisotropy and the attachment/detachment kinetics at the different step types on Si(0 0 1) surface. This new instability mechanism does not require any additional step edge barriers to diffusion of adatoms. In addition, the evolution of ripple height and periodicity was analyzed experimentally as a function of layer thickness. A lateral “ripple-zipper” mechanism is proposed for the coarsening of the ripples.     

LEED structure determination of the Ni(1 1 1)(√3×√3)R30°-Sn surface

Quantitative low energy electron diffraction has been used to determine the structure of the Ni(1 1 1)(√3×√3)R30°-Sn surface phase. The results confirm that the surface layer comprises a substitutional alloy of composition Ni₂Sn as previously found by low energy ion scattering (LEIS), and also shows that there is no stacking fault at the substrate/alloy interface as has been found in (√3×√3)R30°-Sb surface alloys on Ag and Cu(1 1 1). The surface alloy layer is rumpled with the Sn atoms 0.45 ± 0.03 Å higher above the substrate than the surrounding Ni atoms. This rumpling amplitude is almost identical to that previously reported on the basis of the LEIS study. Comparison with similar results for Sn-induced surface alloy phases on Ni(1 0 0) and Ni(1 1 0) shows a clear trend to reduced rumpling with reduced surface atomic layer density, an effect which can be rationalised in terms of the different effects of valence electron charge smoothing at the surface.     

An AFM study of the processing of hydrogen passivated silicon(1 1 1) of a low miscut angle

We present atomic force microscopy (AFM) measurements from a passivated silicon crystal miscut by 0.1° and show the etching regime to be significantly different from surfaces with a larger miscut angle. A simple kinetic model is developed to explain the results and is used to derive the optimal etching conditions for nominally flat Si(1 1 1)–(1×1)H. We show that small changes in miscut angle can alter the kinetic steady state and promote the formation of deep etch pits, even on the least stable, miscut surface. Collisions of steps with these pits result in arrays of stable, self-aligned ‘etch hillocks' over micron dimensions. Following preparation, we use AFM to observe the initial growth of native oxide on the Si(1 1 1)–(1×1)H surface, and demonstrate that AFM is a sensitive probe to surface oxidation in the sub-monolayer regime.     

Photoemission studies of the interactions of CdTe and Te with Si(100)

The interactions between CdTe, and in particular Te, and the (100) surface of Si have been probed using photoemission and low energy electron diffraction with a view to investigating the mechanisms responsible for (100) and (111) growth orientations for CdTe on Si(100). The interfacial reactions have been studied both on room temperature deposition followed by annealing and on depositions at typical epitaxial growth temperatures. In both cases the same precursor stage of an ordered submonolayer of Te on the Si(100) surface has been identified. Line shape analysis of the Si 2p core level has suggested a structural model in which Te adatoms make up an incomplete monolayer bound in bridge sites. This model is in excellent agreement both with the (1 × 1) LEED pattern and recent SEXAFS studies of this surface. The implications of the cubic symmetry of this surface in terms of the subsequent growth orientation of CdTe are discussed. Termination of the surface by Te was also seen to induce band bending suggestive of Fermi level pinning at around midgap, in contrast to the passivating behaviour of other group VI elements on this surface. The Si 2p core level line shape analysis on termination by Te has also provided evidence to support the “covalent dimer” interpretation of the clean dimerised Si(100) surface.     

Covalent binding of cyclohexene, 1,3-cyclohexadiene and 1,4-cyclohexadiene on Si(1 1 1)-7 × 7

The adsorption reactions and binding configurations of cyclohexene, 1,3-cyclohexadiene and 1,4-cyclohexadiene on Si(1 1 1)-7 × 7 were studied using high-resolution electron energy loss spectroscopy (HREELS), ultraviolet photoelectron spectroscopy (UPS), X-ray photoelectron spectroscopy (XPS) and DFT calculation. The covalent attachments of these unsaturated hydrocarbons to Si(1 1 1)-7 × 7 through the formation of Si–C linkages are clearly demonstrated by the observation of the Si–C stretching mode at 450–500 cm⁻¹ in their HREELS spectra. For chemisorbed cyclohexene, the involvement of π_C=C in binding is further supported by the absence of C=C stretching modes and the disappearance of the π_C=C photoemission. The chemisorption of both 1,3-cyclohexadiene and 1,4-cyclohexadiene leads to the formation of cyclohexene-like intermediates through di-σ bonding. The existence of one π_C=C bond in their chemisorbed states is confirmed by the observation of the C=C and (sp²)C---H stretching modes and the UPS and XPS results. DFT calculations show that [4 + 2]-like cycloaddition is thermodynamically preferred for 1,3-cyclohexadiene on Si(1 1 1)-7 × 7, but a [2 + 2]-like reaction mechanism is proposed for the covalent attachment of cyclohexene and 1,4-cyclohexadiene.     

Adsorption of hydrogen and carbon monoxide on Rh(1 1 1)/V surface alloys

A detailed study of the interaction of hydrogen and carbon monoxide with two different Rh(1 1 1)/V surface alloys (1/3 monolayer of V in the second atomic layer or 1/3 monolayer of V in form of islands on the surface) is presented in comparison to the clean Rh(1 1 1) surface. For hydrogen a decrease in the sticking coefficient is found for both alloy surfaces. The sticking coefficient of H₂ as a function of the translational energy is similar to the Rh(1 1 1) surface, showing a direct activated adsorption mechanism. For low translational energies hydrogen adsorption is dominated by dynamical steering on Rh(1 1 1) and by a precursor mechanism on the Rh(1 1 1)/V subsurface alloy. The H₂ TPD desorption peaks are shifted to lower temperatures on the alloy surfaces, caused by the downshift of the metal d-band due to V alloying. On all three surfaces the saturation coverage of hydrogen was measured, giving 1.2, 1.0 and 0.8 monolayer for Rh(1 1 1), the Rh(1 1 1)/V subsurface alloy and for the Rh(1 1 1)/V islands, respectively. For CO the sticking coefficients and the saturation coverages are basically the same on the Rh(1 1 1) and the alloy surfaces. There is an extrinsic precursor on the ordered CO (√3×√3) phase on the Rh(1 1 1) surface, but there is no evidence for such a precursor on the Rh(1 1 1)/V subsurface alloy. On the Rh(1 1 1)/V islands surface, the extrinsic precursor exists on the Rh(1 1 1) surface between the V islands. Apparently this precursor is only stable on the ordered CO layer on Rh(1 1 1).     

The structure of oxygen on Cu(1 0 0) at low and high coverages

The local adsorption structure of oxygen on Cu(1 0 0) has been studied using O 1s scanned-energy mode photoelectron diffraction. A detailed quantitative determination of the structure of the 0.5 ML (√2×2√2)R45°-O ordered phase confirms the missing-row character of this reconstruction and agrees well with earlier structural determinations of this phase by other methods, the adsorbed O atoms lying only approximately 0.1 Å above the outermost Cu layer. At much lower coverages, the results indicate that the O atoms adopt unreconstructed hollow sites at a significantly larger O–Cu layer spacing, but with some form of local disorder. The best fit to these data is achieved with a two-site model involving O atoms at Cu–O layer spacings of 0.41 and 0.70 Å in hollow sites; these two sites (also implied by an earlier electron-energy-loss study) are proposed to be associated with edge and centre positions in very small c(2×2) domains as seen in a recent scanning tunnelling microscopy investigation.     

STM studies on the growth of monolayers: Co on Cu(110) with one half monolayer of preadsorbed oxygen

The growth of the first cobalt monolayer (ML) on the Cu(110)-(2×1)O surface was studied by scanning tunneling microscopy. Extensive exchange of Cu and Co atoms takes place in the first stages of the deposition. The displaced Cu atoms form new Cu---O---Co mixed islands, with the same structure as those of the terrace surface. At 0.25 ML Co, a new structure nucleates, which contains three Cu atoms, four Co atoms and two O atoms per 2×2 cell. The structure consists of rows in the [ 10] direction with an internal periodicity of two lattice units. The rows are separated from one another by two lattice units along the [001] direction, and are found both in-phase and out-of-phase relative to one another. The result is a mixed p(2×2) and c(2×4) surface. The fraction of the surface covered by the new structure increases with Co coverage, and completely covers the surface at 1 ML Co.     

Nanoparticle-induced multi-functionalization of silicon: A plug and play approach

Here, we demonstrate a “plug and play” approach to achieve multi-functionalization of Si. In this approach, externally synthesized functional nanoparticles are introduced onto device quality Si wafers and the surface chemical bonds are manipulated. Sonochemically synthesized Fe₂O₃ nanoparticles are introduced onto Si from an alcohol suspension. On annealing this sample in ultra-high vacuum, the oxygen atoms change the bonding partner from Fe to Si and desorb as SiO at 750 °C. This results in the formation of nanoparticles of Fe on the surface and exhibits ferromagnetic behavior. Deposition of a thin layer (2 nm) of Si onto the sample containing the metallic Fe nanoparticles followed by annealing at 560 °C leads to optically active Si. Photoluminescence measurements show that this sample emits light at three different wavelengths, namely 1.57, 1.61 and 1.63 μm, when excited by He–Ne or Ar lasers. Oxidation of this material results in the formation of a selective capping layer of SiO₂. Thus we obtain multi-functional Si in an “all in one” form and we believe that this approach is universal.