Migration-enhanced epitaxy of cubic BN: An ab initio study

Japanese Journal of Applied Physics, Part 1: Regular Papers and Short Notes and Review Papers

Volumn 43, Issue 7 A, 2004, Pages 4092-4100

  Koga, Hiroaki ^a,b   Miyazaki, Tsuyoshi ^b   Watanabe, Satoshi ^a   Ohno, Takahisa ^b  

^a UNIVERSITY OF TOKYO   (Japan)

^b NATIONAL INSTITUTE FOR MATERIALS SCIENCE   (Japan)

Author keywords

Adsorption; CBN(001); Cubic boron nitride; Diffusion; First principles calculation; Migration enhanced epitaxy

Indexed keywords

ADSORPTION; AMORPHIZATION; COMPUTER SIMULATION; DEPOSITION; DIFFUSION; DIMERS; EPITAXIAL GROWTH; MONOLAYERS; SEMICONDUCTOR QUANTUM DOTS; SEMICONDUCTOR QUANTUM WIRES; SURFACE REACTIONS;

CBN(001); FIRST-PRINCIPLES CALCULATION; MIGRATION-ENHANCED EPITAXY (MEE); POLYCRYSTALLINE FILMS;

CUBIC BORON NITRIDE;

EID: 4644302239     PISSN: 00214922     EISSN: None     Source Type: Journal    
DOI: 10.1143/JJAP.43.4092     Document Type: Article

References (36)

1. P. B. Mirkarimi, K. F. McCarty and D. L. Medlin: Mater. Sci. Eng. R 21(1997)47.
2. T. Taniguchi, K. Watanabe, S. Koizumi, I. Sakaguchi, T. Sekiguchi and S. Yamaoka: Appl. Phys. Lett. 81 (2002) 4145.
3. K. Nose, K. Tachibana and T. Yoshida: Appl. Phys. Lett. 83 (2003) 943.
4. H. Feldermann, C. Ronning and H. Hofsäss: J. Appl. Phys. 90 (2001) 3248.
5. X. W. Zhang, H.-G. Boyen, N. Deyneka, P. Ziemann, F. Banhart and M. Schreck: Nature Mater. 2 (2003) 312.
6. J. Furthmüller, J. Hafner and G. Kresse: Phys. Rev. B 50 (1994) 15606.
7. H.-G. Boyen, N. Deyneka, P. Ziemann and F. Banhart: Diamond Relat. Mater. 11 (2002) 38.
8. Y. Horikoshi, M. Kawashima and H. Yamaguchi: Jpn. J. Appl. Phys. 25(1986) L868.
9. Y. Horikoshi: J. Cryst. Growth 201/202 (1999) 150.
10. This is essentially equivalent to growing cBN as an artificial superlattice consisting of B and N atomic monolayers.
11. K. F. McCarty, M. J. Mills, D. L. Medlin and T. A. Friedmann: Phys. Rev. B 50 (1994) 8907.
12. For the migration of a B adatom on the N surface, and that of a N adatom on the B surface, the energy barriers are roughly estimated to be 4eV. Thus, we do not consider these migration processes in this study.
13. For the migration of a N adatom on the N surface, the energy barrier is roughly estimated to be 4eV. Thus, we do not consider this migration process in this study.
14. 26) suggested the use of MEE, but merely as a means of realizing monolayer coverage surfaces, not as a nonenergetic epitaxial growth method of cBN. They did not even mention that nonenergetic methods had been unsuccessfully used for depositing cBN films.
15. P. Hohenberg and W. Kohn: Phys. Rev. 136 (1964) B864.
16. J. P. Perdew, K. Burke and M. Ernzerhof: Phys. Rev. Lett. 77 (1996) 3865.
17. W. Kohn and L. J. Sham: Phys. Rev. 140 (1965) A1133.
18. Simulation Tool for Atom TEchnology (STATE) Ver. 5.01, Research Institute for Computational Sciences (RICS), National Institute of Advanced Industrial Science and Technology (AIST), 2002.
19. D. Vanderbilt: Phys. Rev. B 41 (1990) 7892.
20. K. Shiraishi: J. Phys. Soc. Jpn. 59 (1990) 3455.
21. J. Yamauchi, M. Tsukada, S. Watanabe and O. Sugino: Phys. Rev. B 54 (1996) 5586.
22. D. J. Chadi: J. Vac. Sci. & Technol. A 5 (1987) 834.
23. We briefly explain the electron counting (EC) model and the EC rule of III-V semiconductor surfaces. Models of surfaces where the number of electrons per supercell exactly fill all the group-V dangling bonds and empty all the group-III dangling bonds are called EC models. The EC rule states that stable surfaces of III-V semiconductors are EC models. The stability of EC models is due to two factors: (i) The electrons are transferred from the group-III dangling bond levels, which are higher in energy, to the group-V dangling bond levels, which are lower in energy; and (ii) The bandstructure becomes semiconducting.
24. K. P. Loh, I. Sakaguchi, M. N. Gamo, T. Taniguchi and T. Ando: Phys. Rev. B 57 (1998) 7266.
25. K. P. Loh, M. N. Gamo, I. Sakaguchi, T. Taniguchi and T. Ando: Diamond Relat. Mater. 8 (1999) 1296.
26. M.-H. Tsai and C. F. Liu: Phys. Rev. B 63 (2001) 073305.
27. 21) also for other structures, as can be seen in Table I.
28. a = -2.1 eV. Considering the rather large adsorption energy at HBu, it is reasonable to assume that the incident atom can overcome the barrier between HBu and HB, and be adsorbed at HB.
29. 21) relatively favors the dimerization compared to our GGA calculation: The energy changes due to the dimerization are +0.01 and -0.13 eV/1 × 1 for the N and B relaxed surfaces, respectively. This difference can be explained as follows. Since the lattice constant is underestimated in LDA, the distance between the nondimerized adatoms of the relaxed surface is also underestimated. This results in the overestimation of the electrostatic repulsion between these adatoms. The stability of the relaxed surfaces are thus underestimated in LDA.
30. The N dimerization [Fig. 7(d)] occurs to decrease the number of N dangling bonds, since the adsorption of a N atom at T4 increases the number of N dangling bonds and decreases the number of B dangling bonds, resulting in the shortage of electrons to fill all the N dangling bonds. The B dimerization [Fig. 7(j)] can be explained in a similar manner.
31. a at HH, the barrier for the migration from the B dimer-center site HH to the adjacent B sublattice site T4' is expected to be small.
32. It should be cautioned that spurious stable positions may be found near T3 and T4 in geometrical optimization calculation, because the potential energy'surfaces around these points are very fiat.
33. These values should be understood as upper limits to the true migration barriers, since there may be migration paths other than those studied in this work.
34. T. Ohno: Thin Solid Films 272 (1996) 331.
35. Generally, the reaction energy barrier is calculated by determining the minimum energy reaction path. However, since the present results imply that the B-N exchange reaction is in fact barrierless, we do not consider it necessary to determine the minimum energy reaction path within the scope of this work.
36. 2 in the chamber may reach the growth surface during B deposition. However, since NZ molecules are relatively nonreactive and cannot be adsorbed on the N surface, we do not expect this to be a serious problem.