Search

The molecular hopping of a lone 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA) molecule adsorbed on a Ge(001) surface is studied by density functional theory calculations and the climbing-ima...

We have investigated the initial pentacene growth on the Si(111)--In surface using scanning tunneling microscopy. At a low coverage, the molecules lie down flat and form locally ordered molecular layers that are nucleated at the Si step edges. At a higher coverage, two kinds of pentacene crystals with the so-called herring-bone arrangement are grown directly on the surface with the longer molecular axis (‘c’ axis) being directed parallel to the surface. The pentacene growth on the surface is considered to be a 3D island growth mode. Two minor crystal structures that are poly-crystal and striped layer are also observed. All experimental results suggest that the surface is rather moderate compared with a reactive surface, which allows for molecular diffusion and rearrangement.

Initial growth process of pentacene molecules on clean Si(001)-2 × 1 substrate was investigated by scanning tunneling microscopy. It is found that the wetting layer, which is not crystalline but disordered, forms before the growth of a first crystalline layer. The wetting layer consists of double layer of the flatly adsorbed pentacene molecules. The formation of the wetting double layer is discussed. The first crystalline layer that is grown on the disordered wetting layer with standing-up pentacene molecules consists of some domains that differ in their crystal structures. Among them, a new pentacene crystal structure is found that can form only on the first layer. In contrast, the second crystalline layer has only single domain. Moreover, we investigated the electronic properties of the pentacene layers by the current–voltage measurements. The pentacene layers are semiconducting with a band gap of about 4 eV.

Hydrogen etching of a 4H-SiC(0001) surface at a moderate temperature of 1200 °C with molecular hydrogen gas was investigated to obtain enough flat and clean surface for large-scale high-quality epitaxial graphene synthesis. We found after a prolonged hydrogen etching that micro scratches, large depressions, and contaminations produced on the wafer in the manufacturing process disappeared and that a periodic array of atomic steps appeared, maintaining initial flat surface morphology. One hour of etching with a flow of 1.0 l/min was the optimum condition to obtain a flat and clean SiC surface in the present study. Using such surfaces, we were able to synthesize the so-called zero layer graphene by thermal annealing in ultrahigh vacuum.

Cu atom penetration through a defect in zero-layer graphene (ZLG) epitaxially grown on an SiC substrate was theoretically investigated, using density functional theory calculations, as a possible mechanism for pure single-layer graphene formation by Cu intercalation on an SiC surface. Our model calculation predicted that a Cu monolayer formed by Cu intercalation causes a lift of the graphene surface of about 0.2 nm, which supports our previous experimental observation. Our calculations on Cu intercalation through a graphene defect implied the possibility that a transition of the defect shape from a 5–8–5 to a double-vacancy model causes the timing of the passage of the Cu atom through the ZLG surface to reduce the potential barrier for the penetration. In addition, it was found that the SiC substrate stabilizes the Cu atom after penetration via an Si–Cu interaction. Furthermore, a preceding intercalated Cu atom was found to be capable of facilitating subsequent Cu penetration by suppressing its inverse re...

Adsorption of 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA) on the Ge(001) surface was studied using scanning tunneling microscopy (STM), scanning tunneling spectroscopy (STS), and the density-functional theory (DFT) calculations. Only single adsorption configuration of the PTCDA molecule was observed at low coverages on the Ge(001) at room temperature, unlike on the Si(001) where several adsorption configurations were reported. This indicates that the PTCDA molecules on the Ge(001) were more mobile than those on the Si(001). Atomic structure of the adsorption configuration on the Ge(001) was determined by comparison between the STM experiments and the DFT calculations. Bias-dependent STM images, STS, and calculated projected density of state curves show nontrivial hybridization of molecular orbitals with surface states of the Ge substrate. Interactions of the PTCDA molecule with the Ge and the Si surfaces were in detail analyzed by the DFT calculation, considering five main competing contributions...

Very initial stages of the homoepitaxy on the $\mathrm{Si}(111)7\ifmmode\times\else\texttimes\fi{}7$ surface are studied at 310, 356, and $366{\phantom{\rule{0.16em}{0ex}}}^{\ensuremath{\circ}}\mathrm{C}$ by in situ continuous observation using high-temperature scanning tunneling microscopy (HTSTM) at the atomic scale during Si deposition under a slow rate of 0.02 bilayer $(\mathrm{BL})\phantom{\rule{0.16em}{0ex}}{min}^{\ensuremath{-}1}$. The substrate surface is reconstructed to the well-established dimer-adatom-stacking fault (DAS) structure with the $7\ifmmode\times\else\texttimes\fi{}7$ unit cell consisting of two triangular half unit cells (HUCs): a stacking faulted (F)-HUC and a normally stacked one. It is expected that the complex, large unit cell compels the initial homoepitaxy to proceed in a quite different manner from that on unreconstructed surfaces. Formation and growth of various adsorbed clusters are pursued by the continuous observation of the same narrow areas during the deposition, avoiding the tip-shadow effect. The most anomalous finding is a quasiliquid cluster (QLC) spreading to plural HUCs (spread QLC) on the $7\ifmmode\times\else\texttimes\fi{}7$ DAS substrate. This appears as a result of the difficulty of the F-HUC to be transformed into the normally stacked BL, being essential for the homoepitaxy. After the transformation, the spread QLC undergoes the following structural changes with an increase of the Si deposition: crystallization to a small epitaxial BL \ensuremath{\rightarrow} surface reconstruction to the DAS structure. Validity and reasons of the transitional formation of the spread QLC are discussed. The spread QLC mediated homoepitaxy mechanism is concluded and is a new mode in the crystal growth. Real in situ HTSTM observations of the same areas on the nanoscale during Si deposition are indispensable to explore for the dynamic atomistic mechanism of the homoepitaxial growth on the $\mathrm{Si}(111)7\ifmmode\times\else\texttimes\fi{}7$.

Silicon carbide (SiC) has the potential to serve as an extremely important semiconductor material in next-generation electronics. However, a major stumbling block for its practical application has been the preparation of high-quality interfaces with insulating materials. We have discovered a way to prepare a 0.6-nm thick silicon oxynitride (SiON) layer having an epitaxial interface with the SiC(0 0 0 1) surface. This review article focuses on the atomic and electronic structures of the SiON layer. Based on various experimental techniques and theoretical studies, we understand the SiON layer to be a complex but unique hetero-double-layered structure: a topmost Si 2 O 5 monolayer is connected to an interfacial Si 2 N 3 monolayer via Si–O–Si linear bridge bonds. The most striking feature of the SiON structure is that there is no dangling bond in the unit cell, rendering it remarkably robust to air exposure. Stability and processes for the formation of the SiON on SiC(0 0 0 1) are discussed on the basis of the structural features obtained. Scanning tunneling spectroscopy measurements of the SiON exhibit a bulk SiO 2 -like band gap of ∼9 eV as well as first-principles calculations. The remarkable band-gap opening of such a thin insulator film is investigated by the combination of element-specific soft x-ray absorption/emission spectroscopies and by first-principles calculations, revealing the Si 2 N 3 and Si 2 O 5 monolayers to have band gaps of corresponding bulk-like values. Promising applications of the SiON to electronic devices are discussed.

The epitaxial silicon oxynitride (SiON) layer grown on a 6H-SiC(0001) surface is studied with core level photoemission spectroscopy. Si 2p spectra show three spectral components other than the bulk one. Chemical shifts and emission angle dependence of these components are well explained within a framework of a determined structure model of the SiON layer.

Atomic and valence-band electronic structures of a recently discovered epitaxial silicon oxynitride (SiON) layer on a 6H-SiC(0001) surface were investigated with x-ray diffraction (XRD) and angle-resolved photoemission spectroscopy (ARPES). The atomic structure optimized by XRD analysis well agrees with a previous low-energy electron diffraction analysis and a first-principles calculation. Band dispersions of surface states observed by ARPES can be explained by the previous calculation. Interface states intrinsic to the SiON layer were not observed above the valence-band maximum of SiC, but a diffuse, non-dispersive state was found by ARPES. Its origin is suggested to be a by-product of graphite-like clusters formed on the SiON layer during heat treatment.

In response to a recent Surface Science Editorial (2010) [1] and accompanying Prospective article [2], we propose that the community of surface scientists and nanoscientists select a common data format for reporting experimental data and atomic coordinates, and have it accepted by relevant journals.

The atomic structures of the Bi/Si(111)-( 3 × 3 ) reconstructed surfaces obtained at different adsorbate coverages were studied by dynamical low-energy electron diffraction (LEED) I–V analysis. The experimentally obtained I–V curves of the β-Bi/Si(111)-( 3 × 3 ) surface, which is formed by a Bi coverage of 1 ML, showed good agreement with the curves calculated for the milk stool model, a model proposed in the literature. In contrast to this result, the I–V curves of the α-phase obtained at a coverage of 1/3 ML do not agree with those reported in early LEED studies. We found that the I–V curves reported in the former LEED studies resemble closely to the results obtained for the surface on which the β and α-phases coexist. Based on the obtained I–V curves, we propose a more reliable structural model for the α-phase, and present the way to determine the presence of a high quality α-phase by using LEED.

Low-energy electron diffraction (LEED) have been used to determine the Cu(0 0 1)–c(4 × 4)-Sn structure formed at 300 K. It is demonstrated that a structural model suggested by scanning tunneling microscopy observations is correct: The model consists of one substitutional Sn atom and four Sn adatoms in the unit cell. Optimum parameters of the determined c(4 × 4) structure reveal that Sn adatoms laterally are displaced by 0.30 A away from ideal fourfold-hollow sites along the 〈100〉 directions. It is proposed that such displacements of the Sn adatoms cause the formation of a network of octagonal rings on Cu(0 0 1). The substitutional Sn atom is located at each center of the octagonal rings. The formation conditions of the network are discussed.

The adsorption of calcium (Ca) atoms on a Cu(0 0 1) surface has been studied by low-energy electron diffraction (LEED) at 130, 300 and 400 K. It is found that a (4 × 4) was the only LEED pattern appeared at 400 K while a quasi-hexagonal structure was formed in a wide range of submonolayer coverage at 130 K. At 300 K, the (4 × 4) LEED spots were broad and weak. The (4 × 4) structure formed at 400 K was determined by a tensor LEED I–V analysis. It is a new-type of surface alloys consisting of five substitutional Ca atoms, nine surface Cu atoms, and two atomic vacancies in the unit cell. In spite of a quite large size-difference between Ca (3.94 A) and Cu (2.55 A) atoms, all Ca atoms are located at the substitutional sites. Among surface alloys so far reported, the atomic size ratio between Cu and Ca in the (4 × 4), 1.54, is the largest. Optimized structural parameters reveal that large lateral displacements of surface Cu atoms, being enabled by the appearance of the vacancies, allow the formation of the (4 × 4) structure.

Ordered nanofacet structures on vicinal 6H–SiC(0 0 0 1) surfaces, consisting of pairs of a (0 0 0 1) basal plane and a ( 1 1 2 ¯ n ) facet, are investigated in terms of stable surface stacking of the (0 0 0 1) basal planes. The surface termination of S3 (or S3*), i.e., ABC (or A*C*B*), was suggested by a structural model based on quantized step-bunching, which typically gives a one-unit-cell bunched step configuration at the ( 1 1 2 ¯ n ) facet. Here, we evaluate the surface termination at basal planes covered with a layer of silicon oxynitride by means of quantitative low-energy electron diffraction (LEED) analysis combined with scanning tunneling microscopy (STM), and show the validity of the structural model proposed.

Formation of a ( 2 × 1 ) local structure is found at submonolayer growth of Tl on the Si(1 1 1)- ( 7 × 7 ) surface by means of scanning tunneling microscopy (STM). At low coverages, the ( 7 × 7 ) grid remains and the ( 2 × 1 ) structure is formed inside of the half-unit cells of the ( 7 × 7 ) . The ( 2 × 1 ) coexists with other structures, such as of “magic” clusters observed in the case of all other group III metals. Based on our STM observations we propose an atomic arrangement of the ( 2 × 1 ) structure.

We report on the re-determination of the structure of p(2 × 2)Sb/Cu(0 0 1) from measured LEED I/V data. The structure solution is performed in two steps. First an approximate “image” of the surface structure model is found by an application to the measured data of the recently-proposed Phase and Amplitude Recovery and Diffraction Image Generation Method (PARADIGM). This reveals that the p(2 × 2) structure is formed by the protrusion of one atom out of four in the outermost surface layer. Since an Sb atom is larger than one of Cu, this suggests a possibility that the p(2 × 2) structure is formed by Sb substitution for one out of four Cu atoms in the outermost layer. Second, a tensor LEED refinement of the model suggested by the PARADIGM yields the same dominant top-layer substitutional model proposed in the earlier LEED study, but with a much lower reliability factor, making much less likely an admixture of other surface phases.

Thin MnSi films growing by units of a quadruple layer of a B20-type structure are observed on the Si(111) surface by scanning tunneling microscopy. Based on the observation, a model of growth controlled by an interplay between MnSi crystal growth and supply of Si atoms from substrate is proposed. A simple Monte Carlo model efficiently simulating significant experimentally observed features is presented. [DOI: 10.1380/ejssnt.2008.276]

The adsorption of K atoms on Cu(0 0 1) has been studied by low-energy electron diffraction (LEED) at room temperature (RT) and 130 K. At RT, a (3 × 2)- p2mg LEED pattern with single-domain was observed at coverage of 0.33, whereas the orthogonal two-domain was found at 130 K. At 130 K, a c(4 × 2) pattern with orthogonal two-domain was observed at coverage 0.25. Both the (3 × 2)- p2mg and c(4 × 2) structures have been determined by a tensor LEED analysis. It is demonstrated that K atoms are adsorbed on surface fourfold hollow sites in the c(4 × 2), while in the (3 × 2) structure two K atoms in the unit cell are located at an asymmetric site with a glide-reflection-symmetry. The asymmetric site is at near the midpoint between the exact hollow site and bridge-site but slightly close to the hollow site. A rumpling of 0.07 A in the first Cu layer was confirmed, which might stabilize K atoms at the asymmetric site. Surface structures appearing in a coverage range 0.25–0.33 are discussed in terms of the occupation of the asymmetric site with increase of coverage.

Adsorption structures of the 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA) molecule on the clean Si(001) − 2 × 1 surface were investigated using scanning tunneling microscopy (STM) experiments in conjunction with first principles theoretical calculations. Four dominant adsorption structures were observed in the STM experiments and their atomic coordinates on the Si(001) surface were determined by comparison between the experimental STM images and the theoretical simulations. Maximizing the number of the Si—O bonds is more crucial than that of the Si—C bonds in the PTCDA adsorption.

By means of low-energy electron diffraction (LEED), we found a reversible structural change of ( 1 × 1 ) ↔ ( 3 × 3 ) on thallium (Tl) adsorbed Si(1 1 1) surfaces by switching the polarity of applied DC voltage for heating the sample. It was shown in the literature that Tl adatoms are located on the T4 sites of the bulk-terminated surface both in the (1 × 1) and ( 3 × 3 ) . It is clarified that the structural change is caused by the electromigration of the Tl adatoms. Tl atoms migrate towards the cathode, being induced by the electric field (10–20 V/cm). We discussed an atomic process of the electromigration.

We considered the origin of the arc-shaped streaks connecting usual diffraction spots observed in an LEED pattern, for a Li system adsorbed on a Cu(001) surface at low coverage, Θ < 1/2. We noted a condition that the natural distance between adsorbed atoms, b nat , of less than Via-in particular, b nat ≅ 1.39a-is consistent with the formation of the arced streaks. Given this condition, adsorbed atoms fill the surface and form a c(52 x 2)R45° structure, which is one of the 'ladder structures', at Θ = 3/5. Even for Θ < 1/2, the atoms form the ladder structure locally. We could observe atomic pairs with second-neighbour distance having a c(2 x 2) structure unit in the ladder structure (where the distance is basically d = 2a), that were shrunk and tilted. These shrunk and tilted pairs produce straight streaks extending from the M(±1/2, ±1/2) points, with a tilted angle in the vicinity of M points. Considering the other arrangements of the second-neighbour pairs, which have a different inclination, a variety of straight streaks exists. The envelope function of these straight streaks is nothing but the arc shape observed in the experiment.

The c(4 × 2) structures in (0 0 1) surfaces of Si and Ge have been studied by low-energy electron diffraction (LEED). Using a proper cleaning method for the Si surface, we were able to observe clear c(4 × 2) LEED patterns up to incident energy of ∼400 eV as well as the Ge surface. Extensive experimental intensity–voltage curves allowed us to optimize the asymmetric dimer model up to the eighth layer (including the dimer layer) in depth in the dynamical LEED calculation. Optimized structural parameters are almost the same for the Si and Ge except for the height of the buckled-up atom of the asymmetric dimer. For the Ge surface, the structural parameters are in excellent agreement with those obtained by a previous theoretical calculation. The tilt angle and bond length of the dimer are 18 ± 1 (19 ± 1)° and 2.4 ± 0.1 (2.5 ± 0.1) A for the Si(0 0 1) (Ge(0 0 1)), respectively.

We have studied the individual adsorption of Mn and Bi, and their coadsorption on Cu(0 0 1) by low-energy electron diffraction (LEED). For Mn, we have determined the c(2 × 2) structure formed at 300 K, whose structure had been determined by several methods. We reconfirmed by a tensor LEED analysis that it is a substitutional structure and that a previously reported large corrugation (0.30 A) between substitutional Mn and remaining surface Cu atoms coincides perfectly with the present value. In the individual adsorption of Bi, we have found a c(4 × 2) structure, which is formed by cooling below ∼250 K a surface prepared by Bi deposition of ∼0.25 ML coverage at 300 K where streaky half-order LEED spots appear. The c(4 × 2) structure has been determined by the tensor LEED analysis at 130 K and it is a substitutional structure. In the coadsorption, we found a c(6 × 4) structure, which has been determined by the tensor LEED analysis. It is very similar to the previously determined structure of the c(6 × 4) formed by coadsorption of Mg and Bi, and embedded MnBi4 clusters are arranged in the top Cu layer instead of MgBi4. Large lateral displacements of Bi atoms in the c(6 × 4)-(Mn + Bi) suggest that the Mn atoms undergo the size-enhancement caused by their large magnetic moment.

The coverage dependence of arced streak intensities observed in LEED pattern, for Li on Cu(001) surface at low coverage, (0–0.50), is studied theoretically. The streaks essentially originate from t...

Two coadsorption systems of Bi + Mg and Sb + Mg on Cu(0 0 1) have been studied by means of low-energy electron diffraction (LEED). A (4 × 4) LEED pattern with glide planes was observed in either case at ∼400 K. The two (4 × 4) structures were determined by a tensor LEED analysis. Both (4 × 4) structures are equivalent, and consist of a grid network of Mg 2 Bi 2 (Mg 2 Sb 2 ) plane clusters on a partially Bi (Sb) substituted Cu(0 0 1) layer. Structural parameters show that not only the formation of the clusters but also the arrangement of the clusters into the grid network contributes to the stabilization of this ordered-mixed-surface-structure.

We have found that various ordered mixed surface structures are formed by coadsorption of two dissimilar metal atoms on Cu(0 0 1) at room tepmerature, using low-energy electron diffraction (LEED) I – V analysis. As coadsorbates, we employed Mg, Bi, Li and K, and surface structures formed by the coadsorption systems of (Mg, Li), (Mg, K) and (Mg, Bi) are presented. A tensor LEED analysis provided detailed geometries of the coadsorbates and the substrate surface. It was found that the surface structures in the above three coadsorption sytems exhibit the restructuring of the Cu(0 0 1) surface. The phase separation into individual adsorbates does not take place, implying that some additional stabilization arises. We demonstrate two origins for the stabilization of the ordered mixed surface structures on Cu(0 0 1). Structures and features formed by the individual adsorption of Mg, Bi, Li and K atoms on Cu(0 0 1) are described first, then those of (2√2×√2)R45°-Mg,Li, (√5×√5)R26.7°-Mg,K, c(2×2)-Mg,Bi, and c(6×4)-Mg,Bi structures formed by the coadsorption are presented. We consider on the basis of the determined structural parameters the question why ordered mixed surface structures are formed instead of the phase separation.

A c(6 × 4) structure formed on Cu(0 0 1) by the coadsorption of Mg and Bi atoms at room temperature has been determined by a tensor low energy electron diffraction analysis. It is an ordered surface ternary alloy with a thickness of single layer, in which Mg, Bi and Cu atoms are mixed in the top layer. In the primitive unit cell, there are one Mg, four Bi, six Cu atoms and one vacancy in the top layer, and substituted Mg and Bi atoms form MgBi 4 plane clusters being arranged in the c(6 × 4) order. Structural parameters show that Mg–Bi bond distances in the MgBi 4 cluster are 3.01 and 3.07 A, which are shorter than the summation of metallic radii of Mg and Bi. It is concluded that a direct, attractive interaction between Mg and Bi atoms plays critical role in the formation of the c(6 × 4) structure.

By using low-energy electron diffraction (LEED), we observed a ( 5 × 5 )R26.7° pattern in the coadsorption of Bi and K (or Cs) on Cu(0 0 1) at room temperature. Both ( 5 × 5 )R26.7°-Bi,K and -Bi,Cs structures were determined by a tensor LEED analysis. The two structures are ordered mixed phases and similar to each other: Bi atoms are located at surface fourfold hollow sites and K (or Cs) atoms occupy on-top sites of the substrate surface Cu atoms. Significant rumpling of 0.13 A (-Bi,K) or 0.19 A (-Bi,Cs) is found in the top Cu layer. It is suggested that the on-top site occupation of K or Cs atoms is stabilized by additional bonds with surrounding Bi atoms. The Hamilton-ratio tests suggest that K atoms displace laterally from the exact on-top position, but that Cs atoms are upright on the on-top sites.

The coadsorption of dissimilar metals, Mg and Bi, on Cu(0 0 1) at room temperature was studied using low energy electron diffraction (LEED). Five LEED patterns were observed at various coverages: c(2×2), c(4√2×2√2) R 45°, c(6×4), c(6×6) and p(4×4). The c(2×2) structure was determined by a tensor LEED analysis. Mg atoms substitute for every second Cu atom in the top layer, and Bi atoms are located at the on-top sites of the remaining surface Cu atoms. The structure parameters demonstrate that Bi atoms pull up Mg atoms located in the substitutional sites, indicating a direct, attractive interaction between coadsorbates. Reasons for the formation of the c(2×2) structure are discussed in comparison with the surface structures formed by the individual adsorption of Mg and Bi.

The formation mechanism of the 7×7 reconstruction on annealed Si(1 1 1) surfaces has been demonstrated at the atomic level. In situ observations of unreconstructed regions (‘1×1’) on terraces after rapid quenching to 380 °C were done using scanning tunneling microscopy (STM) with a scanning speed of 1.7 s per frame. In the narrow ‘1×1’ regions, we imaged isolated single-faulted (F) halves of the dimer–adatom–stacking-fault (DAS) structure from “birth” to “death”. During “life”, the isolated single F-halves frequently changed their size. The size changes between odd-sized F-halves always took place through even-sized F-halves of intermediate sizes: 5×5-F ′ ↔6×6-F↔7×7-F ′ ↔8×8-F↔9×9-F ′ ↔10×10-F↔11×11-F ′ ↔12×12-F↔13×13-F ′ , where the 5×5-F ′ , 7×7-F ′ and so forth are irregular-type structures of the odd-sized F-halves. Even-sized F-halves and the irregular-type structures are necessary in the size changes, whereas the regular-type structures have never been involved. Lifetimes of the 10×10-F, 8×8-F, and 6×6-F at 380 °C are about 10.5, 6, and 2–3 s, respectively, which are much shorter than those of the isolated irregular-type structures of the odd-sized F-halves. With the aid of room temperature STM images of a rapidly quenched surface, we determined the atomic structures of the even-sized F-halves. We have proposed a sequential size change (SSC) model, including undiscovered parts of the size changes ‘1×1’ ↔2×2-F↔3×3-F ′ ↔4×4-F↔5×5-F ′ , as the formation and decay mechanism of isolated single F-halves in the ‘1×1’ region. The SSC model has the following sequence: ‘1×1’ ↔ 2×2-F↔3×3-F ′ ↔4×4-F↔5×5-F ′ ↔6×6-F↔7×7-F ′ ↔8×8-F↔9×9-F ′ ↔10×10-F↔11×11-F ′ ↔12×12-F↔13×13-F ′ . It was found by collecting statistics of size-change directions that one of two equivalent sides of the irregular-type structures, which have a mirror symmetry, is involved in the size changes thus indicating that other parts of the F-halves remain unchanged. Based on such findings, we have proposed the atomic processes for bond-rearrangements in the SSC model. The bond-rearrangements proceed along one side of a triangular F-half by breaking the existing dimers and forming new dimers like a “zipper”. Proposed atomic processes of the zipper-like restructuring are illustrated by a ball-and-stick model. The reason for the appearance of the even-sized F-halves and the irregular-type structures of the odd-sized ones is discussed in terms of the energy barrier heights along a reaction path in the size change of single F-halves.

Recently we demonstrated a prototype instrument that employs a new technique for determining surface structures. The instrument consists of a scanning tunneling microscope tip as a field emission gun and a detector for projecting electron scattering patterns. Previous results showed a low efficiency in detecting scattered electrons, because the potential between the tip and the sample drew the electrons back toward the sample. To improve the efficiency, we attempted detection using a tip-shield that can extract the scattered electrons toward a vacuum region effectively. Our results confirmed that the tip-shield improves detection efficiency. Consequently, we observed an extra beam that might be scattered toward the surface-parallel direction. The results of our electron trajectory calculations agreed qualitatively with the experimental data. We have tentatively designated the two spots we obtained as (0 0) and (1 0) diffraction beams of the graphite surface.

The formation processes of a ( 2 2 × 2 )R45° structure formed by the coadsorption of Li and Mg on Cu(0 0 1) at room temperature, whose structure has been determined [Surf. Sci. 493 (2001) 91], are examined by low-energy electron diffraction (LEED). By depositing Mg atoms on the Li pre-adsorbed (2×1) surface, a sequence of LEED patterns (2×1)→ c (4×4)→(2 2 × 2 )R45° was observed. The c(4×4) LEED pattern is proved to be due to the coexistence of the (2×1)-Li and ( 2 2 × 2 )R45°-Li,Mg structures. On the other hand, when Li was adsorbed on the Mg pre-adsorbed surface, a disordered surface was formed until the ( 2 2 × 2 )R45°-Li,Mg structure appeared at Li coverage of about 0.22. Different formation processes of the ( 2 2 × 2 )R45° structure in the two adsorption sequences are discussed at the atomic level.

We have carried out in situ observation of homoepitaxial growth on Si(1 1 1) 7×7, whose structure is known as the dimer–adatom–stacking-fault (DAS) model, by scanning tunnelling microscopy at 356°C. In order to grow epitaxially, it is necessary to cancel the underlying DAS structure. At low Si coverage, we observed small clusters within half unit cells. Then, large clusters that have no atomic image are formed. After further Si deposition, these large clusters change into a two-dimensional (2D) island with atomic protrusions, which implies that the underlying reconstructed layer was cancelled. After that, the 2D island transforms into the DAS structure.

Pd-intercalated graphene grown on a SiC(0001) substrate was investigated using STM, angle-resolved photoemission spectroscopy, and XPS. Pd atoms deposited at room temperature on a zero layer graphene grown on a SiC(0001) substrate were intercalated between the zero layer graphene and the SiC substrate after the thermal annealing above 700 °C, forming a Pd-intercalated single layer graphene. No charge transfer occurred between the intercalated Pd layer and the graphene, which resulted in the formation of the electrically neutral graphene. The Pd-intercalated graphene remained electrically neutral throughout the annealing temperature range between 700 and 1100 °C. The charge transfer, however, occurred between the intercalated Pd layer and the SiC substrate, which caused a band bending confirmed in the core level spectra measured by XPS.

Coadsorption of Na and K on Ag(001) at room temperature has been studied by low energy electron diffraction (LEED). A 3 × 3 LEED pattern was observed irrespective of the order of adsorption. For this formation, it is necessary to deposit Na and K atoms with appropriate coverage. We have determined the 3 × 3 structure by a tensor LEED analysis. It is a restructured surface and is very similar to the previously determined 3 × 3 structure formed on Ag(001) by pure Na adsorption. In the coadsorption, Na and K atoms occupy preferable sites selectively, and construct an ordered mixed structure on Ag(001). That is, small Na atoms are located in the missing row sites, while large K atoms sit on the hollow sites of four-Ag-atom islands. The reason for the site selectivity of Na and K atoms in the mixed 3 × 3 structures is discussed.

The coadsorption of Li and Mg on Cu(0 0 1) at room temperature was studied by low-energy electron diffraction (LEED). A (2√2×√2) R 45° structure was observed and determined by a tensor LEED analysis. It is an ordered mixed phase, in which Mg atoms are located on fourfold substitutional sites, while Li atoms occupy surface hollow sites, instead of the separation into c(2×2)-Mg and (2×1)-Li phases formed in the individual adsorption. The (2√2×√2) R 45° structure is comprised of alternative atomic-rows of substitutional Mg atoms and hollow-site Li atoms aligned along the [0 1 0] direction. The stability of this structure is discussed in terms of adsorption energy of Li and Mg atoms and the displacement of surface Cu atoms.

We study structures of simple metal monolayer adatoms on fcc(0 0 1) metal surface, both on theoretical and experimental sides. (i) Theory: using a simple Hamiltonian, two types of theoretical phase diagram of structures of adatoms are obtained. Then, the following structures are observed in the phase diagrams; (A) c (n 2 × 2 )R45° structures in a coverage range, 5/9 ⩽ σ ⩽2/3; (B) square-like lattice-gas structures at σ =9/13,4/5 and 13/16; (B ′ ) hexagonal arrangement of adatoms at higher coverage or lower amplitude of substrate potential, λ , regions, which is neighboring to the structures (B). (ii) Experiment by low energy electron diffraction (LEED): On Mg on Cu(0 0 1) surface, four-spot patterns and quasi-hexagonal patterns appear with increasing coverage which correspond to structures above (B) and (B ′ ). These are in good agreement with a low λ part of the phase diagrams, while LEED does not exhibit the c(2×2) and (A) patterns. Their origin will be discussed also by using the phase diagrams.

The coadsorption of Mg and alkali-metals (Li, K or Cs) on Cu(0 0 1) at room temperature has been studied by low-energy electron diffraction (LEED). A (2√2×√2)R45° structure is formed by Mg and Li at coverages both about 0.25 ML, and which has been determined by a tensor LEED analysis. It is an ordered mixed phase, in which Mg atoms are located in fourfold substitutional sites and Li atoms occupy fourfold surface hollow sites, forming alternative one-dimension-like atomic rows along the [0 1 0] direction. On the other hand, a (√5×√5)R26.7° structure is formed by the coadsorption of Mg with either K or Cs at coverage about 0.2 ML each. These structures have also been determined: Mg atoms are in fourfold substitutional sites, while K (Cs) atoms are on surface hollow sites, forming square arrangements of Mg and K (Cs). It is found that the top layer Cu atoms in both ordered structures are displaced laterally, and that Mg atoms are located at deeper positions in fourfold substitutional sites than in a c(2×2) substitutional structure formed by individual Mg adsorption. The reason for the formation of the ordered mixed surface structures is discussed.

We have carried out scanning tunneling microscopy (STM) observations of unreconstructed regions on quenched Si(1 1 1) surfaces at 380°C at a scanning speed of 1.7 s per frame. In the regions, it is found that single faulted-halves of the dimer-adatom-stacking-fault (DAS) structure are formed isolatedly or at the edges of the surrounding DAS domains sharing one corner hole. In such “living” regions, we have succeeded to observe sudden structural changes of the faulted-halves during line scans in single frames of STM images.

We report on the scanning tunneling microscopy observation of small Si islands with a 1 × 1 termination on the Si(1 1 1) surface. The islands were prepared by thermal desorption of Tl from the Tl-terminated silicon sample by means of annealing to 400–600 °C. Structure of the islands is interpreted as the dimer-stacking-fault (DS) model. We propose that the otherwise unfavorable 1 × 1 termination is stabilized by subsurface dimers of DS.

We continuously observed the growth of the dimer–adatom–stacking-fault (DAS) domain, in unreconstructed regions remaining after quenching a Si(111) surface to 370–380°C, by using scanning tunneling microscopy. It was observed that a single faulted (F) half of the 9×9 unit cell of the DAS structure grows to a small 9×9 DAS domain. Continuous measurements showed that new F-halves are created sharing corner holes with existing F-halves. The creation of new isolated F-halves was very seldom at 370–380°C, and the region of the DAS structure was grown by expanding the area of the existing DAS domains.

Recent revelation on even-sized stacking-faulted halves as well as irregular odd-sized ones on the quenched Si(111) surface is theoretically studied. The employed model for the analysis is an extended version of the cell model which was proposed to interpret the phase transition of the 7×7 structure in terms of the formation energy of each element of the DAS structure. The probability distributions of the various sizes/types of stacking-faulted halves are calculated in corresponding situations of the experiment. The calculated probability distributions are qualitatively coincident with the experimental result.

The growth patterns of the Si(111)-(7×7) structure (DAS structure) on the quenched surface are calculated using Monte-Carlo simulation. The model Hamiltonian, which was constructed in terms of the states of each 7×7 triangular half, based on an assumption of one-by-one formation of faulted halves in the previous paper, is employed. In the simulation the out-of-phase effect for growth is taken into account. Calculated patterns of the DAS domain growth reproduce the characteristic features of the observed patterns for growths on the terrace, from the F- and U-steps, respectively.

We have observed peculiar faulted (F) halves of the dimer–adatom-stacking fault (DAS) structure by scanning tunneling microscopy on quenched Si(111) surfaces at 380°C, and denoted them irregular F-halves (F′). In the F′-halves, the arrangement of adatoms is partially different from that of corresponding regular F-halves and one adatom is missing. We observed single 5×5-F′, 7×7-F′, 9×9-F′, 11×11-F′ and 13×13-F′ in unreconstructed regions, and proposed their structures. Fluctuations in the size and structure of the faulted halves have been observed in real time by rapid imaging at 380°C. It is found that single odd-sized faulted-halves appearing in the size changes are always the irregular ones.

We have observed single isolated faulted-half (F) of DAS structure in unreconstructed regions of Si(111) remained after rapid quenching to 380°C, by using scanning tunneling microscopy. We have observed even-sized F of 6×6, 8×8, 10×10 and 12×12 during the size-changes of odd-sized F. The odd-sized F appearing in the size-changes is always irregular-type. We proposed structural models of the even-sized and irregular odd-sized F. As a whole, we observed the size-changes of single F: irregular 5×5-F↔6×6-F↔ irregular 7×7-F↔8×8-F↔irregular 9×9-F↔10×10-F↔irregular 11×11-F↔12×12-F↔irregular 13×13-F . From these results, we proposed the mechanism of DAS domain formation.

The phase transition of the Si(111)7×7 structure and its growth on a quenched surface are investigated theoretically. The surface structure is described in terms of states of faulted and unfaulted triangular halves of the 7×7 unit cell. A model Hamiltonian which satisfies peculiar features of the structure is derived. The Monte-Carlo simulation for this model shows that the 7×7 structure undergoes the first-order phase transition. Growth patterns of the 7×7 structure on the quenched surface are calculated with the same parameters and they reproduce characteristic features of experimental results. The calculation also qualitatively reproduces the temperature dependence of the critical nucleus size for the 7×7 domain growth on the quenched terrace.

The structure of a Si(001) surface with monatomic steps is investigated by means of scanning tunneling microscopy at 95 K and Monte Carlo simulations (MCS) at various temperatures. In particular, we pay attention to the effect of kinks at the step edge on the formation of c (4×2) domains, and both studies reveal that the kinks do not govern the growth of the domains. MCS demonstrates that the nucleation of c (4×2) domains takes place at central regions of the terrace as a result of thermal fluctuation, and that the domains propagate to the step edge with decreasing temperature.