Former activities performed in SFB 290 - Co Films on Cu(001) Surfaces:
Non-Arrhenius Growth Behavior, Reentrant Layer Growth, and Floating Cu Layers
 zurück zur Übersicht
 

1. Experimental method: Grazing scattering of fast atoms

A beam of 25 keV He atoms is scattered under a polar angle of 1°-2° to the Cu(001) surface plane and an azimuthal angle of a few degrees to the [110] surface lattice direction. Growth-related irregularities like nucleated islands perturb the correlated scattering process (surface channeling), which results in a loss of the specular-beam intensity.




Interpretation is straightforward and based on classical mechanics computer simulations emulating trajectories of scattered atoms. The method works in real-time and real-space and yields quantitative, statistical information on the morphology of the surface of the growing film.

The specular-beam intensity is measured as function of Co coverage for constant temperature and deposition rate F=4,5±0,3·10-3 ML/s. The chemical composition of films is deduced from electron- and proton-induced Cu MVV (60eV) and Cu LMM (920eV) Auger-intensities.

 

2. N-shaped dependence of island density on growth temperature




Fig. 2.1  We observe a unique N-shaped dependence of the density nx of nucleated islands (circles). This is at variance with an Arrhenius-behavior, as expected from classical nucleation theory (nx% exp (E / kT). This complex behavior is also proposed by ab-initio DFT-kMC calculations (lines) and explained as follows (R. Pentcheva et al., Phys. Rev. Lett., submitted):

  • Below 320 K: Arrhenius-behavior. Hopping of Co atoms. Hopping barrier 0.59 eV (experiment: 0.54 eV). Small, rectangular Co islands.
  • 320 K: Activation of atomic exchange between Co and Cu (Fig. 2.2). Exchange barrier 1.00 eV. Enhanced island density due to pinning of migrating atoms at substitutional Co atoms
  • Above 320 K : About 50% of Co atoms exchanged into substrate. Combined growth of large, Co-decorated Cu islands and small Co islands




Fig. 2.2  Barrier for exchange of Co and Cu along [100] direction

 

3. Bilayer growth and reentrant layer growth



Fig. 3.1  Initial growth of Co on Cu(001) is strongly dependent on temperature. Above a coverage of 2 monolayers (ML) layer-by-layer growth is observed:

  • Above 450 K : Diffusion length of adatoms (mostly Cu) exceeds the mean terrace width of the substrate surface of about 600 Å.
    Þ Step-flow growth

  • 430 K - 350 K: enhanced concentration of Cu adatoms. Large diffusion length of Cu.
    Þ Layer-by-layer growth

  • Below 320 K: no Co/Cu exchange. Tendency to form bilayer Co islands (presumably owing to large surface free energy of Co compared to Cu).
    Þ Bilayer growth (Fig. 3.3)

  • Below 200 K: Island size comparable with Cu lattice constant.
    Þ
    Condensation energy of deposited Co adatoms enough to surmount Schwöbel-barrier(„transient mobility“).
    Þ Reentrant layer-by-layer growth (Fig. 3.2)



Fig. 3.2  Transition from bilayer to reentrant layer-by-layer growth with decreasing temperature




Fig. 3.3  Specular-beam intensity as function of temperature for nominal Co coverages as indicated





Fig. 3.4  Cu MVV Auger signals, induced by grazingly scattered protons, show that the amount of Cu at the film surface increases with increasing Co coverage at temperatures where bilayer growth is observed (250 K).

 

4. Floating Cu layers

Ab initio DFT calculations on the ground state of Co/Cu(001) propose a capping of Co films with Cu (R. Pentcheva and M. Scheffler, Phys. Rev. B 61(2000)2211). In order to deduce depth-dependent composition profiles, we measured Cu MVV (60 eV) and LMM (920 eV) Auger signals, induced by keV electrons at oblique incidence or 25 keV protons at grazing incidence (Fig. 4.2). Excitation by grazingly scattered protons results in a sensitivity which is restricted to the film surface (topmost layer), whereas the information depth in conventional, i.e., electron-induced Auger spectroscopy amounts to about 2ML for the MVV Auger line and 10ML for the LMM line. Assuming perfect layer-by-layer growth, we find that Co films grown at high temperatures (410 K) are capped by 2 layers of Cu (Fig. 4.3). Growth at low temperatures (140 K) results in pure Co films, i.e., interlayer mass transport between Co and Cu is suppressed.




Fig.4.1 Proton-induced Cu MVV Auger signal during deposition of 4ML Co on Cu(001) at 410 K. After cooling down to 210 K, deposition of additional 6ML Co onto this film.



Fig. 4.2  Electron- and proton induced MVV (60eV) and LMM (920eV) Auger signals of Cu during deposition of Co on Cu(001) at different temperatures.

Solid lines in Fig. 4.2 show calculated Auger signals for pure Co films (140 K), Co films capped with 2 layers of Cu (Fig. 4.3), and a hypothetical c(2x2) CoCu alloy throughout the film.




Fig. 4.3  Model for chemical composition of Co films on Cu(001) (Cu substrate not shown): 2 - 3 layers of Cu „float“ on top of a pure Co film (growth temperature 410 K).

 zurück zur Übersicht
 
© www.hu-pgd.de