Photoelectron diffraction of ${\mathrm{C}}_6{\mathrm{H}}_8∕\mathrm{Si}\left(001\right)$: A model case for photoemission study of organic molecules adsorbed on silicon surfaces

The core-level photoemission of 1,4-cyclohexadiene molecules on the quasi-single-domain $2\times 1\text{−}\mathrm{Si}\left(001\right)$ surface below the saturation condition was studied. Taking advantage of core-level-shift photoelectron diffraction of the C $1s$ core level, the details of the surface structure were found. The model was essentially constituted by an unbuckled (planar) molecule chemisorbed on the surface, inclined at an angle of about 18° with respect to the surface normal, keeping one double bond unsaturated. An accurate analysis of the line shape parameters of the strongly asymmetric C $1s$ core level was mandatory for the structural study. Conversely, the success of the structural determination represented a confirmation of the correct deconvolution procedure. Such a feedback loop was decisive in reducing the indeterminacy affecting the analysis of core-level photoemission in the case of complex molecules adsorbed on surfaces.

Figure 1

Adsorption of 0.4 ML 1,4-CHD on the quasi-single domain $2\times 1\text{−}\mathrm{Si}\left(001\right)$ surface. Core-level decomposition and diffraction effects in the asymmetric C $1s$ core-level components along three different electron detection directions ($\theta$ and $\varphi$) are shown from (a) to (c). $\varphi =\left[110\right]$ is the majority dimer bond direction.

Figure 2

Plan view along (001) crystal plane (a), $\left(1\overline{1}0\right)$ plane (b), and (110) plane (c) of the stick and ball sketch of 1,4-CHD on $\mathrm{Si}\left(001\right)\text{−}2\times 1$. Silicon is depicted by black spheres and carbon by gray spheres. Hydrogen atoms are not included. Angles with respect to the $\left(1\overline{1}0\right)$ plane describing the molecule inclination are defined in (d).

Figure 3

(Color online) Comparison of theoretical (left panel) vs experimental (right panel) PD anisotropy of the C $1s$ core level of 0.4 ML 1,4-CHD on quasi-single-domain $2\times 1\text{−}\mathrm{Si}\left(001\right)$ surface. From top to bottom the three C $1s$ PD patterns corresponding, respectively, to components C1 and C2 at $+0.42\mathrm{eV}$ BE and C3 at $-0.27\mathrm{eV}$ BE are reported together with the total C $1s$ core-level pattern. $x$ axis is the majority dimer bond direction.

Figure 4

Azimuthal cuts of theoretical (continuous line) vs experimental (triangles) PD anisotropy of the C $1s$ core level of 0.4 ML 1,4-CHD on quasi-single-domain $2\times 1\text{−}\mathrm{Si}\left(001\right)$ surface. From top to bottom the three C $1s$ PD patterns corresponding, respectively, to C1 and C2 at $+0.42\mathrm{eV}$ BE and C3 at $-0.27\mathrm{eV}$ BE are reported together with the total diffraction intensity. From left to right each cut at $\theta$ values in the range 30°–80° reports the normalized anisotropy for the values $\varphi$ in the range 0°–90°, with $\varphi =0°$ the majority Si dimer bond direction.

Figure 5

$R$-factor values in the $({\varphi }_1,{\varphi }_2)$ parameter space at the ${\varphi }_3$ value of model (c) of Table . The errors affecting the two parameters are indicated by bars according to the criterion of the 10% $R$-factor variation from the best fit.