Have We Found The Final Explanation For The Onset Of Rapid Interlayer Mass Transport?
- PDF / 275,359 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 15 Downloads / 166 Views
Have We Found The Final Explanation For The Onset Of Rapid Interlayer Mass Transport? M. Giesen; H. Ibach Institut für Schichten und Grenzflächen, ISG 3, Forschungszentrum Jülich, D 52425 Jülich, Germany ABSTRACT In recent STM experiments on multilayer island decay on Cu(111) in UHV a novel rapid interlayer mass transport was observed. This new transport mechanism may accelerate island stack decay by several orders of magnitude. Initial explanations for the onset of rapid decay events were based on the correlation between the occupation of surface states and the presence of a step edge barrier for adatom hopping over the island edges. We present new experimental data measured for Ag(111) which are in contradiction to the former interpretation of rapid interlayer mass transport. We show that all experimental results may be explained by considering an exchange mechanism when island edges in a stack are in close proximity. INTRODUCTION The coarsening of surface structures during and after homoepitaxial growth is determined by the kinetics of the intra- and the interlayer mass transport. While on (111)-surfaces of fcc metal the surface diffusion barrier is typically very low (see e.g. tables in [1]), the crossing of step and island edges is generally hindered by an Ehrlich-Schwoebel (ES-) barrier [2, 3]. Therefore, it was an important finding when a novel rapid interlayer mass transport mechanism was reported which seems to overcome the ES-barrier and enhances interlayer diffusion by two orders of magnitude [4]. The new mechanism was observed during multilayer island stack decay, where top layer islands decayed much faster than expected in the presence of an ESbarrier. The rapid decay sets in when on the average island edges in a multilayer stack come closer than a critical distance. This distance was 6.4 atomic rows on Cu(111) [4]. We could also show that the rapid decay mechanism plays a significant role during coalescence of multilayer islands [5]. In a following paper, the onset of the rapid island decay was correlated with the occupation of surface states on (111)-surfaces [6]: It was shown that the experimental data could be explained under the assumption that the ES-barrier vanishes if the island edges in a stack come closer than a critical distance wc. Furthermore it was shown that for Cu(111) wc=6.4 atomic rows coincides with the distance between island edges (assuming the edges to be infinite barriers for electrons) where surface state is pushed above the Fermi level and the surface state is emptied. This result seemed to corroborate the idea of Memmel and Bertel [7, 8] that the ESbarrier is correlated with the occupation of surface states. Since the surface state on Cu(111) is hardly temperature dependent between 300 and 400K [9] one would expect no temperature dependence of the critical distance between island edges when the rapid decay sets in, in agreement with the experimental data [4]. From photoemission data of the surface state on Ag(111) [10], on the other hand, one finds that wc should be about 22 at
Data Loading...
