Surface reconstructions transition metals

It is clear that much work remains to be done to extend our understanding to polax surfaces of transition metal oxides in which the cations have partially filled d orbitals. An especially challenging issue is related to mixed valence metal oxides, such as Fe304, in which the cations exist under two oxidation states. In addition, considering the rapid development of ultra-thin film synthesis and characterization, a simultaneous effort should be performed on the theoretical side to settle the conditions of stability of polar films. More generally, on the experimental side, it seems that one of the present bottlenecks is in a quantitative determination of the surface stoichiometry, an information of prominent interest to interpret the presence or absence of reconstruction. 

Surface states exist on normal metals as well as on transition metals. Occupied and unoccupied surface states were recently studied experimentally and theoretically on single crystal surfaces of e.g. Ag (Reihl, 1985) Cu (Bartynski et al., 1985) and Ni (Borstel et al., 1985). Surfaces of transition metals are particularly interesting as they not only show a structural relaxation - an effect which is mostly weak on normal metals - but can also exhibit magnetic properties which differ from those of the bulk (Freeman, 1983). A surface enhanced magnetic order as well as magnetic surface reconstruction were observed on Gd(OOOl) (Weller et al., 1985). The magnetic hysteresis loop of the Fe 100) surface, very recently measured by means of the spin polarization of secondary electrons (Fig. 3), shows a softer behaviour within the outermost 5 8 due to reversed domain nucleation (Allens-pach et al., 1986). The structural and electronic differences of surfaces and bulk may manifest themselves as surface core level shifts (Eastman and Himpsel, 1982 Erbudak et al., 1983). In the case of rare earth metals and alloys they may even appear as surface valence transition (Netzer and Matthew, 1986 Kaindl et al., 1982). 

With respect to metallic substrates, the low-index surfaces of transition metals provide naturally narrow surface bands due to almost locahzed d orbitals. The surface reconstructions on W(OOl) and Mo(OOl) (Chapter 4), for example, have been explored with respect to the controversy whether the transiAon is due to CDW formation or local bond formation. It has been noted, however, that these... 

Some surfaces of bcc metals, such as W(100) or Mo(100), reconstruct [28], The reconstruction occurs reversibly upon cooling the crystals below 200-250 K, and it can be viewed as a continuous, temperature-driven, order-disorder phase transition. The room temperature phase is (1 x 1) and the low temperature phase has a c(2 x 2)p2mg unit cell involving only short-range atomic displacements [29]. Much has been speculated with respect to the role of surface states in driving this reconstructions, but no clear evidence has been presented so far [30]. 

The oxides containing cations in octahedral coordination, such as MgO and TiOz, seem to suffer little or no reconstruction of the crystal structure at the surface, but there are major changes in electronic structure. These effects have important implications for surface reactivity, especially when oxygen vacancy defects are considered and particularly in transition-metal oxides. The example of TiOj was discussed above more complex behavior is shown by species such as TijOj and NiO (Henrich, 1987). The latter, like other transition metals, shows increasing [Pg.415]

Fig. 1 shows the rocksalt lattice [15]. We will discuss MgO and NiO as limiting cases of oxides, one containing a simple metal ion and the other one a transition metal ion. The (100) surface of such a material represents a non-polar surface, the (111) surface represents a polar oxide surface. Since the lattice constants are very similar for both oxides (MgO 4.21 A, NiO 4.17 A) [15], we expect the surface structures to be similar. The non-polar surface exhibits a nearly bulk terminated surface as shown in Fig. 2a and it is very similar for both materials. We have put together information from FEED [16-21] and STM [22-25] analysis. There is very small interlayer relaxation and only a small rumpling of the surface atoms, whereby the larger anions move outwards and the small cations very slightly inward. A completely different situation is encountered for the polar (111) surfaces. Due to the divergent surface potential [13] on an ideally, bulk terminated polar surface, the surfaee reconstructs and exhibits a so... 

Surface Reconstructions in W. In earlier sections, we have noted that the pair functional formalism is not always appropriate as the basis for an energetic description of metals. One case where this is evident is in the context of the surface reconstructions seen on certain bcc transition metals. In particular, the (001) surface of both W and Mo exhibits fascinating reconstructions, and as in the case of the Au(OOl) surface, the symmetry breaking distortions leave few vestiges of their crystallographic ancestry. 

The (110) surfaces of the transition metals Au, Ir. Pt display both a stable (1x2) and a metastable (1x3) reconstruction. The structural details of these reconstructions are listed in table 2 and the (1x2) reconstructed surface is illustrated in fig. 2. Both the (1x2) and (1x3) reconstructions are of the missing-row type which involve the removal of every second (1x2) or third (1 x3) row of atoms from the top atomic plane of the bulk termination. The removal of this row is accompanied by significant atomic relaxations of at least the first three atomic planes perpendicular to the surface, see table 2. Both Pt(110)(lx2), Pt(110)(lx3) and Au(110)(lx2) have relaxations of a similar magnitude but the relaxations of Ir(110)(lx2) are significantly smaller. In addition to the planar relaxations, all of these surfaces exhibit lateral motions of the atoms within the second atomic plane out towards the valleys left by the missing rows. In addition, the removal of the atomic row causes a buckling of the third atomic layer which conforms with the hill and valley structure of the missing-row surface. 

The synthesis of hydrocarbons from CO hydrogenation over transition metals is a major source of organic synthetic chemicals and fuels. The Fischer-Ttopsch (FT) reaction, which is directed to the production of hydrocarbons from syngas, implies the polymerisation of-CHx entities and carbon-carbon bond formation is required. The historical achievements have been revised on several occasions see for instance, the reviews by Vannice,2 Schulz3 and the special issue of Catalysis TodayA devoted to FT. Also, a synopsis of the main recent industrial developments has been presented by Adesina.5 Furthermore, the recent edition of Topics in Catalysis6 should be mentioned, where different aspects of the reaction mechanism, surface reconstruction of active surfaces, improved reactors and optimisation of catalyst preparations have been treated by various specialists, scientists and engineers. 

---