Adsorption-induced lattice relaxation and diffusion by concerted substitution

Physical Review B - Condensed Matter and Materials Physics

Volumn 59, Issue 8, 1999, Pages 5892-5897

  Feibelman, P J ^a   Stumpf, Roland ^a  

^a SANDIA NATIONAL LABORATORIES   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0000880715     PISSN: 10980121     EISSN: 1550235X     Source Type: Journal    
DOI: 10.1103/PhysRevB.59.5892     Document Type: Article

References (35)

1. P. J. Feibelman, Phys. Rev. Lett. 65, 729 (1990).
2. G. L. Kellogg and P. J. Feibelman, Phys. Rev. Lett. 64, 3143 (1990).
3. C. Chen and T. T. Tsong, Phys. Rev. Lett. 64, 3147 (1990).
4. For a recent review, see G. L. Kellogg, Surf. Sci. Rep. 21, 1 (1994).
5. For Rh/Rh(001), see G. Ayrault and G. Ehrlich, J. Chem. Phys. 60, 281 (1974). No experimental determination of the self-diffusion mechanism has been published for Pd/Pd(001), but the theoretical results presented here imply that hopping has the lower barrier in this case.
6. G. L. Kellogg, A. F. Wright, and M. S. Daw, J. Vac. Sci. Technol. A 9, 1757 (1991).
7. B. D. Yu and M. Scheffler, Phys. Rev. B 56, R15 569 (1997).
8. Specific results concerning diffusion barrier vs. surface stress and surface stress vs. strain are not provided in Ref. 7.
9. 20% if one uses a formulation of Density Functional Theory, 30% if one uses the non-local Generalized Gradient Approximation. See Ref. 7 for details.
10. Below, incidentally, we show that the symmetric dimer over a vacancy is not the transition geometry for Ir(100) and Pt(100), but a metastable state.
11. Actually, it is not experimentally certain that CS is the low-barrier self-diffusion process only on Pt(100) and Ir(100). Though hopping does prevail on Rh(100), Ni(100) is a doubtful case. The noble metals are too soft for Field Ion Microscopy. Pd(100) has not been investigated. See Ref. 4, Table 2.
12. For a continuum picture, however, see R. Cammerata, Surf. Sci. 279, 341 (1992).
13. P. Hohenberg and W. Kohn, Phys. Rev. B 136, 864 (1964);
14. W. Kohn and L. J. Sham, Phys. Rev. A 140, 1133 (1965).
15. G. Kresse and J. Hafner, Phys. Rev. B 47, 558 (1993);
16. Phys. Rev. BG. KresseJ. Hafner49, 14 251 (1994).
17. G. Kresse and J. Furthmüller, Comput. Mater. Sci. 6, 15 (1996).
18. G. Kresse and J. Furthmüller, Phys. Rev. B 54, 11 169 (1996).
19. G. Kresse and J. Hafner, J. Phys.: Condens. Matter 6, 8245 (1994).
20. J. P. Perdew, in Electronic Structure of Solids ’91, edited by P. Ziesche and H. Eschrig (Akademie Verlag, Berlin, 1991);J. P. Perdew and Y. Wang (unpublished).
21. M. Methfessel and A. T. Paxton, Phys. Rev. B 40, 3616 (1989).
22. O. H. Nielsen and R. M. Martin, Phys. Rev. B 32, 3792 (1985).
23. The Ir/Ir(100) barrier has been revised down to 0.74 eV from the higher value reported in Ref. 3, after a tip temperature recalibration. See T. Y. Fu, H. T. Wu, and T. T. Tsong, Phys. Rev. B 58, 2340 (1998).
24. Formation of a metastable dimer over a vacancy has been observed in Re CS on Ir(100). See C. Chen and T. T. Tsong, Phys. Rev. Lett. 72, 498 (1994);
25. T. T. Tsong and C. Chen, Nature (London) 355, 328 (1992).
26. R. Shuttleworth, Proc. Phys. Soc. London, Sect. A 63, 444 (1950).
27. D. Wolf, Phys. Rev. Lett. 70, 627 (1993).
28. V. Fiorentini, M. Methfessel, and M. Scheffler, Phys. Rev. Lett. 71, 1051 (1993).
29. The sharp-eyed will realize that our values of S/γ-1 are rather different from the stress-excess results presented in Table I of Ref. 25. There are three reasons [V. Fiorentini (private communication)].
30. The first is that the τ’s reported in that table are values of total surface stress per surface atom, and not of “excess” stress, despite what is stated in the main text and in the table-caption. The second reason is that, notwithstanding the implication of that caption, the values of τ reported were computed for unrelaxed not relaxed surfaces; in the case of Ir(100), this makes the value of τ almost 60% larger than what A. Filippetti and V. Fiorentini, Surf. Sci. 377-379, 113 (1997) find when the surface is relaxed. Third, the energies and stresses of Ref. 25 differ from ours because of the use of the Local Density rather than the Generalized Gradient Approximation in accounting for electronic exchange and correlation effects. In the end, in parallel with our own conclusions regarding CS diffusion, Fillippetti and Fiorentini find that reconstruction of Ir(100) cannot be attributed to “excess” tensile stress, since in LDA there is none. Importantly, this conclusion implies a need to revisit the discussion given in Ref. 7 of the low-barrier self-diffusion mechanisms on Ag(100) and Au(100). The argument there is directly based on the “tensile excess stress” results given in Ref. 25.
31. Adding further potential for confusion, stress-change components can change sign as CS proceeds. For example, in the CS transition geometry of Ir/Ir(100), which is asymmetric (cf. Table III), (Formula presented) and (Formula presented) Thus, stress is less tensile in this barrier configuration than in the starting four-fold hollow geometry, even though it is more tensile farther along the diffusion path, where the CS dimer is symmetric (cf. Table VII).
32. For a recent review, see H. Ibach, Surf. Sci. Rep. 29, 193 (1997).
33. C. E. Bach, M. Giesen, H. Ibach, and T. L. Einstein, Phys. Rev. Lett. 78, 4225 (1997).
34. M. Schmid, A. Biedermann, S. D. Bøhmig, P. Weigand, and P. Varga, Surf. Sci. 318, 289 (1994).
35. G. L. Kellogg, Phys. Rev. Lett. 79, 4417 (1997).