Open Collections

LEEDS studies of adsorbate-induced reconstruction of metal surfaces

University of British Columbia

Using the technique of low-energy electron diffraction (LEED), the structures of three systems have been investigated, where small, electronegative atoms chemisorb and induce reconstruction at a metal surface; these surfaces are designated Ni(111)-(2x2)-0,Cu(110)-(2x3)-N, and Pd(100)-(V5x45)R27°-0. In each case, experimental data consist of a set of intensity-versus-energy (I(E)) curves measured at normal incidence with a video LEED analyser. Multiple-scattering calculations were done for a number of proposed models of the surface structure, and comparisons were made between calculated and experimental 1(E) curves. The objective in a LEED analysis is to find the geometry in the calculations which leads to the best match between the two sets of intensity curves. Reliability indices are used to quantify the level of correspondence between experiment and theory; a lower R-value indicates better agreement, and consequently the model is more likely to be correct. Two basic reliability indices were used in this work, and they are the modified Zanazzi-Jona R-factor and the Pendry R-factor. In addition to the more conventional LEED analysis, the recently-developed tensor LEED/directed search (TLEED) method was also used in the analyses of the N/Cu and 0/Pd surfaces. As well as determining surface structure, this work has the further objective of using the details identified to develop some principles of surface structural chemistry and to relate these principles to the broader framework of structural chemistry. For the Ni(111)-(2x2)-0 structure, oxygen atoms chemisorb on 3-fold hollow sites which continue the fcc stacking of the nickel substrate. Top-layer Ni atoms which are bonding to 0 atoms are displaced both vertically toward (by 0.12 A) and laterally away from (by 0.07 A in a rotated manner) the 0 atoms, while those Ni atoms not bonding to 0are displaced vertically toward the bulk. The determined surface O-Ni bond length of 1.83A agrees closely with the predicted value of 1.82 A for 3-coordinate 0 on Ni, and the surface relaxations have been confirmed in a subsequent study with SEXAFS. Nitrogen, activated by an ion gun, chemisorbs on the Cu(110) surface to form Cu(110)-(2x3)-N. Many techniques have been applied to study this surface, but little consensus exists as to its structure. Even structural conclusions from the present LEED analysis have been revised as new information, as well as TLEED program codes, became available. Currently, TLEED appears to favour a reconstruction in which the topmost layer is a pseudo-(100)-c(2x2)-N overlayer with substantial corrugation in the top two copper layers. Large lateral displacements of both N atoms in the overlayer and Cu atoms in the topmost Cu(110) layer result in a total of three 5-coordinate adsorption sites per unit mesh (as compared to one when no lateral relaxations are allowed). Average N-Cu bond lengths for the 5-coordinate sites (1.85 A) agree well with prediction based on the structure of bulk Cu3N, while the bond lengths for the 4-coordinate sites (1.94 A) appear rather long. Very recently published STM images seem to suggest that nitrogen chemisorbs first on the 5-coordinate sites, before occupying the less favourable 4-coordinate sites, and that opens the possibility that the 4-coordinate sites may be only partly occupied under the experimental conditions of this work. The Pd(100)-(453015)R27°-0 surface is formed on extended 0-dosing with the sample temperature at greater than about 550 K. A tensor LEED analysis of 15 independent beams supports a surface oxide model, as first postulated in outline by Orent and Bader. The detailed model which gives the best correspondence with experimental intensity data has a PdO(001) overlayer stacked onto the Pd(100) surface such that rumpling is induced in both the oxide and top two Pd(100) layers. The average 0-Pdbond length for 2-coordinate 0 on the Pd surface (1.73 A) is close to the predicted value of 1.76 A based on the structure of bulk PdO. This analysis in particular highlights the potential advantages of the TLEED approach in opening up LEED crystallography for determining a wider range of surface structures than has typically been the case until now.

Thesis/Dissertation

application/pdf

2008-09-10

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/1792

Chemistry

University of British Columbia

UBCV

DSpace