Xenon adsorption on palladium. I. The homogeneous (110), (100), and (111) surfaces

Abstract

Systematic LEED, UPS, work function, and flash desorption investigations on the face specificity of xenon adsorption on macroscopic palladium single crystal surfaces are reported. The observed trends appear particularly suited to test current theoretical predictions about weak adsorption. With LEED, formation of hexagonally close-packed overlayers are observed on Pd(100) and Pd(110) and a 7/8 × 7/8 R 30° structure is detected on Pd(111), indicating complete registry on the (111) face, only one dimensional registry on the (110) face, and no registry on the (100) face. The latter two are in accord with earlier reports. UV-photoelectron spectroscopy reveals xenon induced face specific modifications of the Pd valence band emission which are regarded as fingerprints for electronic interaction between metal and absorbate. The electron binding energies of the Xe (5p3/2,1/2) photoemission peaks (in the limit of zero coverage) on the three Pd surfaces are found to obey the relationship EVB,0=EFB,0+φc, where EVB,0 and EFB,0 denote the binding energy with respect to the vacuum and Fermi level, respectively, and φc is the clean substrate work function. This substrate independence of EVB,0 , which is also observed on various surfaces of many other metals for both 5p-valence (UPS) and 3d-core levels (XPS) is explained as a pure surface electrostatic effect. The thermal desorption energies E0ad corresponding to the initial stage of adsorption decrease in the order Pd(110)>Pd(100)>Pd(111) and are estimated to be 10.2, 9.4, and 8.3 kcal/mol, respectively, using a preexponental of ν=1015 s−1. This ordering is in agreement with simple coordination arguments but not with an interpretation in terms of the charge-transfer-no-bond (CTNB) model for rare gas adsorption. Systematic results on the face specificity of the initial xenon dipole moment μ° are reported for the first time: μ°(111)=0.7D, μ°(100)=0.6D, and μ°(110)=0.44D. This ordering, which is reverse from that of E0ad, is incompatible with simple image potential arguments but, can instead be reconciled with a xenon–metal interaction model which assumes partial filling of an adsorption induces Xe (6s) resonance. The experimental data presented and their interpretation lay the basis for the understanding of the adsorption behavior of xenon on heterogeneous, i.e., polycrystalline, stepped, and sputter-roughened Pd surfaces which will be presented in part II of this work.