The migration of Sn was studied by using molecular beam epitaxially grown layers into which Sn and Sb distributions (table 2) were introduced during growth. The Sn and Sb profiles were measured by using secondary-ion mass spectrometry. By comparing the diffusion of Sn and Sb, in an inert N ambient, to that in a nitriding NH3 ambient (where vacancies were injected), the fractional vacancy contribution to the diffusion of Sn was found to be equal to that of Sb; which, in turn, was known to be close to unity. It was therefore concluded that Sn (figure 7) diffused predominantly via a vacancy-mediated mechanism in Si. On the other hand, the activation energy for diffusion was found to be higher than expected for a vacancy-mediated diffusion mechanism. This was explained by assuming the existence of Sn-vacancy configurations which were different from the configuration in which a vacancy was trapped next to a Sn atom.
P.Kringhøj, A.N.Larsen: Physical Review B, 1997, 56[11], 6396-9
Table 2
Diffusivity of Sb and Sn in Si at 1000C in Various Ambients
and with Various Sb and/or Sn Distribution in the Sample
Distribution | Ambient | Diffusant | D (cm2/s) |
Sn and Sb | N2 | Sn | 2.55 x 10-16 |
Sn and Sb | NH3 | Sn | 1.40 x 10-15 |
Sn and Sb | N2 | Sb | 1.57 x 10-15 |
Sn and Sb | NH3 | Sb | 8.00 x 10-15 |
Sb | N2 | Sb | 1.58 x 10-15 |
Sb | NH3 | Sb | 7.45 x 10-15 |
Sn | N2 | Sn | 3.13 x 10-16 |
Sn | NH3 | Sn | 1.51 x 10-15 |
Figure 7
Diffusivity of Sn in Si