Collective diffusion on a stepped substrate

Surface Science

Volumn 381, Issue 2-3, 1997, Pages

  Merikoski, J ^a,b   Ying, S C ^a  

^a BROWN UNIVERSITY   (United States)

^b UNIVERSITY OF JYVÄSKYLÄ   (Finland)

Author keywords

Models of surface kinetics; Surface diffusion

Indexed keywords

ATOMS; COMPUTER SIMULATION; DIFFUSION; GREEN'S FUNCTION; MATHEMATICAL MODELS; MOLECULES; MONTE CARLO METHODS;

LATTICE GAS MODEL; SURFACE DIFFUSION;

SURFACE PROPERTIES;

EID: 0031162513     PISSN: 00396028     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0039-6028(97)00123-4     Document Type: Article

References (22)

1. M.G. Lagally (ed.), Kinetics of Ordering and Growth at Surfaces, Plenum, New York, 1990.
2. R. Gomer, Rep. Prog. Phys. 53 (1990) 917.
3. R.L. Schwoebel, J. Appl. Phys. 44 (1969) 614.
4. X.D. Xiao, X.D. Zhu, W. Daum, Y.R. Shen, Phys. Rev. B 46 (1992) 9732.
5. X.D. Xiao, Y. Xie, Y.R. Shen, Phys. Rev. B 48 (1993) 17452.
6. X.D. Xiao, Y. Xie, C. Jakobsen, H. Galloway, M. Salmeron, Y.R. Shen, Phys. Rev. Lett. 74 (1995) 3860.
7. A. Natori, R.W. Godby, Phys. Rev. B 47 (1993) 15816.
8. R.A. Tahir-Kheli, R.J. Eliott, Phys. Rev. B 27 (1983) 844.
9. T. Ala-Nissila, J. Kjoll, S.C. Ying, R.A. Tahir-Kheli, Phys. Rev. B 44 (1991) 2122.
10. Our rate equations and the Green's functions as defined here correspond to collective diffusion, which can be defined solely in terms of density fluctuations, and thus the identity of individual particles can be neglected [9].
11. J. Merikoski, S.C. Ying, to be published.
12. K. Kehr, in: K. Binder (Ed.), Applications of the Monte Carlo Method in Statistical Physics, Springer, Berlin, 1984.
13. C. Uebing, R. Gomer, Surf. Sci. 306 (1994) 419; C. Uebing, R. Gomer, Surf. Sci. 317 (1994) 165, 427. In this work diffusion on a stepped substrate characterized by L = 4 was studied by extensive MC simulations.
14. C. Uebing, R. Gomer, Surf. Sci. 306 (1994) 419; C. Uebing, R. Gomer, Surf. Sci. 317 (1994) 165, 427. In this work diffusion on a stepped substrate characterized by L = 4 was studied by extensive MC simulations.
15. For c→0 and c→1, the result of Eq. (4) is exact, and in the single-particle limit c→0 it reduces to that given in Ref. [7].
16. For increasing terrace width the computational effort in the MC simulations increases rapidly and, therefore, a full numerical study of collective diffusion as a function of terrace width is not feasible.
17. yy will be discussed in detail in forthcoming work [11].
18. xx on terrace width and the resulting crossover criterion at finite coverage are substantially different from those of the single-particle case analyzed in Ref. [7]. For decreasing temperature the collective effects lead actually to an exponentially increasing deviation from the zero-coverage behavior already for c≈1/L [11].
19. 2 proposed in Ref. [4].
20. Our model is expected to describe reasonably well the physics at a coverage close to one, where the effect of the possible adsorbate interactions is mainly to renormalize the intrinsic activation barriers [11].
21. For this system the Schwoebel barrier is believed to be very small; P.J. Feibelman, private communication.
22. step are the activation barriers for diffusion on terraces and at step edges [11].