Photoelectron diffraction studies of Cu on Pd(111) random surface alloys

The study of surface alloys is motivated by their use in many applications of different segments of industry, such as in the search for new catalysts and sensors, in surface protection against corrosion, in lowering friction, and in testing electronic devices. An important aspect of surface alloys studies is that of the precise quantification of segregation and diffusion processes as well as the determination of surface structure. In this paper we report a combined low-energy electron diffraction and photoelectron diffraction (PED) (using synchrotron radiation) study of surface alloy formation when Cu ultrathin films are evaporated onto Pd(111) single-crystal surfaces. We present results for two different coverages (1 and 3 ML) and three annealing temperatures (300, 600, and $800\mathrm{K}$). For these preparation conditions, a random alloy phase with different concentrations seems to form in the first few layers. Through the analysis of PED data performed using a multiple scattering formalism and the average $T$-matrix approximation it was possible to determine the atomic structure and the atomic concentration of the first three layers.

Figure 1

LEED and photoelectron diffraction analysis for a clean Pd(111) surface. (a) LEED pattern for normally incident primary electrons of $162\mathrm{eV}$ energy. (b) Contour map of PED $R_a$ factor analysis for Pd $3d$ emission as function of first and second interlayer distances related to the bulk value. (c) PED experimental raw data excited with $650\mathrm{eV}$ photons from Pd $3d$, and (d) PED theoretical simulation using optimized parameters (details in the text).

Figure 2

Planar projection of PED patterns (raw experimental data) for 1 ML of Cu on Pd(111) surface evaporated and measured at room temperature. (a) Pd $3d$ emission and (b) Cu $3p$ emission excited with $530\mathrm{eV}$ photons.

Figure 3

Layer-by-layer determination of the surface alloy concentration and theory-experiment comparisons for PED data sets for 1 ML of Cu on Pd(111) annealed at $600\mathrm{K}$ and excited with photons of $530\mathrm{eV}$. (a) Experimental raw data for Pd $3d$ emission; (b) experimental raw data for Cu $3p$ emission; (c) and (d) correspond to optimized PED theoretical simulations respectively for Pd $3d$ emission and Cu $3p$ emission considering a PdCu random alloy in the first and second layers with the concentrations found by the $R_a$ factor contour map showed in (e) and (f). (e) Contour map of $R_a$ factor for Pd $3d$ emission as function of Cu concentration in the first and second layer. (f) Same for Cu $3p$ emission.

Figure 4

XPS spectra for Cu $3p$ for 1 and 3 ML of Cu on Pd(111) before (gray curves) and after annealing (black curves).

Figure 5

Layer-by-layer determination of the surface alloy concentration and theory-experiment comparisons for PED data sets for 3 ML of Cu on Pd(111) annealed at $800\mathrm{K}$ and excited with photons of $700\mathrm{eV}$. (a) Experimental raw data for Pd $3d$ emission; (b) experimental raw data for Cu $3p$ emission; (c) and (d) correspond to optimized PED theoretical simulations respectively for Pd $3d$ emission and Cu $3p$ emission considering a PdCu random alloy in the first an second layers with the concentrations found by the $R_a$ factor contours map showed in (e) and (f). (e) Contour map of $R_a$ factor for Pd $3d$ emission as function of Cu concentration in the first and second layer. (f) Same for Cu $3p$ emission.