Portal de Periódicos da CAPES

Scanning tunneling microscopy study of the reaction of Br2 with Cu(100)

1998; Elsevier BV; Volume: 398; Issue: 3 Linguagem: Inglês

10.1016/s0039-6028(98)80035-6

1879-2758

C.Y. Nakakura, Eric I. Altman,

Molecular Junctions and Nanostructures

The adsorption of Br2 on Cu(100) and the further reaction of Br2 with Cu(100) to form CuBr was studied using scanning tunneling microscopy (STM). Although low energy electron diffraction indicated the formation of a c(2 × 2) layer immediately upon the dissociative adsorption of Br2, thermal fluctuations in the adsorbed layer prevented this structure from being observed until the coverage exceeded roughly 80% of saturation. As the c(2 × 2) layer saturated, the substrate steps were observed to rotate to align parallel to the 〈100〉 close-packed directions of the adlayer. The rotated steps exhibited thermal fluctuations that STM movies showed were due to kink diffusion. Further Br2 exposure greatly decreased the step fluctuations and resulted in the formation of bulk CuBr. The former was due to a decreased kink density that allowed more Br atoms to be accommodated in the c(2 × 2) layer. The 〈100〉 steps were the predominant source of Cu atoms for the reaction to CuBr. As the reaction progressed the steps receded until they reached another step and formed a {110} facet. Defects such as kinks and domain boundaries were not observed to contribute significantly to the reactivity of the steps. The reaction at the steps was not uniform; certain segments of the steps were found to be far more reactive than others. Annealing the c(2 × 2) layer increased the lengths of the reactive segments, consistent with previous temperature programmed desorption data that showed that annealing increases the reactivity of the surface. Comparison of STM images of reactive and unreactive sections of the steps suggested that the variation in reactivity is due to a thermally driven relaxation of the atomic positions around the step. The CuBr formed at the steps was mobile and aggregated into clusters at locations that showed no obvious correlation with the locations of changes in the surface structure associated with consumption of the Cu surface. The implications of these results on oxidation and etching reactions of metal surfaces are discussed.