Clean surfaces, reconstructions

Ibach and Rowe found the reconstructed surface structures to be very stable to hydrogen adsorption, in that on 111 7x7 and 100 2 x 1 surfaces, no change in LEED patterns occurred up to saturation coverage. The favoured interpretation was based on the model of clean surface reconstruction proposed by Lander [76] and extended by Phillips [202], whereby surface vacancies produce warped benzene ring structures in the first and second surface layers, and such distortions are too large to be removed by hydrogen adsorption. The 111 2 x 1 structure was found to be somewhat less stable, however, and the fractional order spots disappeared on hydrogen adsorption. This was related to the Haneman clean... 

So, clean surfaces tend to restructure to satisfy the unbalanced atomic forces. When foreign atoms adsorb onto such surfaces, further reconstruction is possible. For example, the chemisorption of contaminant atoms can destroy the clean surface reconstructions described above. Alternatively, new structures may form, as when carbon is chemisorbed on nickel (100) surfaces. If such carbon-coated surfaces were brought together in an adhesion experiment, the carbon would have to diffuse out before full Ni-Ni adhesion could be attained. Such diffusion and restracturing effects could explain the observed changes of adhesion with time. Also, hysteresis in adhesion values could then be accounted for. 

Reconstruction is a relative term here reconstruction is understood with respect to the ideal bulk-lattice termination, rather than with respect to the actual clean surface. Of particular interest in adsorption are several cases the induction of a new reconstruction, when none was present on the clean surface the removal of a clean-surface reconstruction (sometimes inelegantly called un-reconstruction or de-reconstruction) and the change from one reconstruction to another (which could be called rereconstruction). 

This entry indicates that the adsorbate (In) has removed the clean-surface reconstruction. The adatom is located at a height doi = 1.85 0.05 A above the first shell of outermost Si atoms. The adsorption occurs in a T4 site (as stated higher in the table and in the description included in this entry). And the spacing... 

Table 9 lists the adsorbate-induced changes on fcc(lOO). Adsorbates induce reconstructions of fcc(lOO) perhaps more frequently than for the hexagonal surfaces, but the statistics are poor, and the choice of metals studied probably not random. However, the adsorbates that cause reconstructions tend to be the same as on the hexagonal surfaces primarily alkali adatoms, some other metal adatoms that form a one-layer alloy with the substrate, and oxygen. While sulfur induces reconstructions on fcc(lll) and fcc(llO), none is evident on fcc(lOO). On the other hand, adsorbates can also remove a clean-surface reconstruction (few such cases are listed in Table 9, but examples are known qualitatively by the change in their LEED patterns). 

Adsorbates, however, can totally remove these complex clean-surface reconstructions, and restore a bulk-like termination, cf. Tables 31 and 33. Sulfur and oxygen can bond at top sites on a full-bilayer-terminated substrate, capping the dangling bonds, while the interlayer spacings in the substrate can relax. Or S can substitute for the outer half of the external bilayer. 

I.P.P.D and its relatives have become standard procedures for the characterization of the structure of both clean surfaces and those having an adsorbed layer. Somoijai and co-workers have tabulated thousands of LEED structures [75], for example. If an adsorbate is present, the substrate surface structure may be altered, or reconstructed, as illustrated in Fig. VIII-9 for the case of H atoms on a Ni(llO) surface. Beginning with the (experimentally) hypothetical case of (100) Ar surfaces. Burton and Jura [76] estimated theoretically the free energy for a surface transition from a (1 x 1) to a C(2x 1) structure as given by... 

Some fascinating effects occur in the case of CO on Pt(lOO). As illustrated in Fig. XVI-8, the clean surface is reconstructed naturally into a quasi-hexag-onal pattern, but on adsorption of CO, this reconstruction is lifted to give the bulk termination structure of (110) planes [56]. As discussed in Section XVIII-9E very complicated changes in surface structure occur on the oxidation of CO... 

When atoms, molecules, or molecular fragments adsorb onto a single-crystal surface, they often arrange themselves into an ordered pattern. Generally, the size of the adsorbate-induced two-dimensional surface unit cell is larger than that of the clean surface. The same nomenclature is used to describe the surface unit cell of an adsorbate system as is used to describe a reconstructed surface, i.e. the synmietry is given with respect to the bulk tenninated (unreconstructed) two-dimensional surface unit cell. 

The majority of deposits formed in this group have been on Au electrodes, as they are robust, easy to clean, have a well characterized electrochemical behavior, and reasonable quality films can be formed by a number of methodologies. However, Au is a soft metal, there is significant surface mobility for the atoms, which can lead to surface reconstructions, and alloying with depositing elements. In addition, Au it is not well lattice-matched to most of the compounds being formed by EC-ALE. 

The clean siuface of solids sustains not only surface relaxation but also surface reconstruction in which the displacement of surface atoms produces a two-dimensional superlattice overlapped with, but different from, the interior lattice structure. While the lattice planes in crystals are conventionally expressed in terms of Miller indices (e.g. (100) and (110) for low index planes in the face centered cubic lattice), but for the surface of solid crystals, we use an index of the form (1 X 1) to describe a two-dimensional surface lattice which is exactly the same as the interior lattice. An index (5 x 20) is used to express a surface plane in which a surface atom exactly overlaps an interior lattice atom at every five atomic distances in the x direction and at twenty atomic distances in the y direction. 

Surface reconstruction on metal crystals depends on the interior lattice as well as on the nature of the metal, such as Au (100)-(5 x20), Au (lll)-(l x 23) and Pt (llOHl X 2) [Kolb, 1993]. In general, the activation energy of surface reconstruction is relatively great ( 1 eV) on clean metal surfaces so that the reconstruction is frequently suppressed at room temperature. Usually, surface adsorption changes the activation energy that catalyzes or inhibits surface reconstruction. 

Some metal surfaces reconstruct either in the clean state or in the presence of adsorbed gases. Platinum, iridium, and gold (100) surfaces, which have square symmetry, all reconstruct to hexagonal close-packed (111) surfaces... 

Numerous studies with low-energy electron diffraction (LEED) revealed that most of the clean surfaces of the platinum group metals exhibit an atomic arrangement that is identical to that expected from an undistorted termination of the bulk. Variations of the vertical lattice spacings between the topmost atomic layers are very small, if present at all (66). Exceptions are, however, found with the (100) and (110) planes of Ir and Pt. The clean and thermodynamically stable structures of the Pt(100) (67-69) and Ir(100) (70, 71) surfaces were found to reconstruct and to exhibit 5 x 1-LEED patterns. A plausible explanation (72) is that in these cases the topmost atomic layer forms a hexagonal arrangement, similar to that within the (111)... 

We call this Pt(100) surface reconstructed. Surface reconstruction is defined as the state of the clean surface when its LEED pattern indicates the presence of a surface unit mesh different from the bulklike (1 x 1) unit mesh that is expected from the projection of the bulk X-ray unit cell. Conversely, an unreconstructed surface has a surface structure and a so-called (1 x 1) diffraction pattern that is expected from the projection of the X-ray unit cell for that particular surface. Such a definition of surface reconstruction does not tell us anything about possible changes in the interlayer distances between the first and the second layers of atoms at the surface. Contraction or expansion in the direction perpendicular to the surface can take place without changing the (1 x 1) two-dimensional surface unit cell size or orientation. Indeed, several low Miller index surfaces of clean monatomic and diatomic solids exhibit unreconstructed surfaces, but the surface structure also exhibits contraction or expansion perpendicular to the surface plane in the first layer of atoms (9b). 

In the first study of its kind, second harmonic generation has been used to study potential induced reconstruction on Au(lll) and Au(100) by Kolb and coworkers [156]. These surfaces have been known to reconstruct in UHY when they are clean [153, 157], Surface reconstruction occurs when the surface atoms of a solid rearrange themselves in a structure different from that expected from simple termination of the bulk lattice. Various studies by cyclic voltammetry, electroreflectance spectroscopy and ex situ electron diffraction have suggested that flame-treated crystals form stable reconstructions in solution. Unfortunately, due to the lack of in situ probes, very little direct evidence for this reconstruction has been available. 

Abstract. Clean metal surfaces often display an atomic arrangement at the surface that differs from the one in the bulk. Some of these surface reconstructions show mesoscopic order and are very adequate to act as a template for the ordered growth of arrays of atoms, molecules or clusters. The electronic states at some surfaces can be prototypes of highly dense 2D electron gases where a number of fundamental properties can be addressed in detail. Localized surface states, on the other hand, are relevant in chemical processes at surfaces. The recent developments in experimental and theoretical techniques allow the exploration of these issues with unprecedented precision. 

---