Co/Au(111) is a prototypical system of self-organized growth : Co grows on the Au(111) 22×?3 herringbone reconstruction as a regular array of 2 atomic layer high dots with a rectangular lattice parameter of ?~7.6 nm and 2?~20 nm (Fig. 1), with 2 Co islands per unit cell. This peculiar growth process has been investigated in many STM and theoretical studies. However, several questions were left unanswered. GISAXS and anomalous-GISAXS have been used to address them.

 

STM picture of Co/Au(111)

Figure 1: a) STM picture of Co/Au(111) (after a deposit of 0.8A of Co). (b) Average super-cell.

 

The first issue is the degree of ordering of the Co dots at the different stages of growth. The advantage of GISAXS over STM is indeed that it probes a large area, providing statistical information. Figure 2(a) shows that the order along the <11-2> directions is very long-range. This is well understood since this periodicity is directly related to that of the 22×?3 reconstruction. In contrast, Fig. 2b shows a cumulative disorder along the <1-10> directions, which is related to a much more subtle energy balance of the “kinks” periodicity.

GISAXS CCD images after Co deposition

Figure 2: GISAXS CCD images after 7.3A of equivalent thickness Co deposition, for two azimuts: X-ray beam is along the <110> direction for (a) and along the <112> for (b)

 

The second unanswered question relates to the morphology of the Co film for thick deposits (continuous film) and of its interface with the Au substrate. Indeed, in a naïve way, we would expect that the GISAXS intensity should dramatically decrease beyond coalescence of the Co dots into a continuous film (i.e. around 1.6 Å equivalent thickness for coalescence along the <11-2> azimuths). By contrast, the integrated intensity of GISAXS satellites increases up to 5 Å, and remains sizeable up to 35 Å, while keeping a FWHM of 3%, attesting a degree of ordering that does not degrade. A possible origin could be a periodic strain-field in the Au substrate, induced by the 18% lattice parameter misfit with Co. Another possibility could be that some Au atoms have diffused in or around the Co dots. Anomalous GISAXS has been used to disentangle this question: after 4.2 Å of Co deposit, we recorded images at 9 energies around the Co-K edge (7.709 keV), with the X-ray beam along the <110> direction. Near an absorption edge, the atomic scattering factor of the concerned element has a complex expression (f(Q)=f0(Q)+f’(Q)+f”(Q)). The f’ and f” have been measured, by recording the transmission factor through a Co foil, as a function of energy. The Fa (scattering factor of Co, anomalous atoms), and Ft (scattering factor of all atoms) were then extracted by fitting the intensity of the diffraction rods over nine energies. The result is shown in Fig. 3. Fa is very similar to Ft, which means that the contribution of Au to the intensity is negligible. The two above hypotheses to explain the excess of intensity can then be ruled out. Other possibilities may be the presence of non-coherent interfaces at the junctions between neighbouring Co islands.

 

Scattering factors Ft and Fa

Figure 3: Scattering factors Ft (all atoms) and Fa (Co atoms)

 

The growth of superparamagnetic Co “pillars” (3) was also performed, using the method shown in Fig. 4a. GISAXS was used as a “crystallographic-like” technique, which we call “super-crystallography”, to determine the structure of these pillars within the super unit cell. GISAXS images were recorded every 1 degree over 180°. For each image, the intensity along the line corresponding to ?=0 has been extracted to build a map of the in-plane reciprocal space, which is shown in Fig. 4b. It exhibits a 6-fold symmetry arising from the contribution of 3 variants of the super lattice, rotated 120° from each other. The white arrows denote two crystallographic directions, corresponding to the same variant. The corresponding map is simulated (Fig. 4c) using the IsGISAXS software.

Schematic view of Co pillars, reconstructions and simulation.

Figure 4: (a) Schematic view of Co pillars. (b) Reconstructions of in-plane reciprocal space. (c) Simulation.