The diffusion of Ni into Si, and into a transition layer between these elements in a diffusion couple, was studied at 470 to 1070K. The samples used were plates which were oriented in the (111) plane, and had dislocation densities of the order of 1000/cm2. Layers (0.0003mm) of Ni which contained 63Ni were then deposited onto the Si. The distribution of Ni in the Si single crystals and in the transition layers was determined via the autoradiography of oblique sections. It was found that the results at above 870K could be described by:

transition layer:     D (cm2/s) = 2.3 x 103 exp[-3.25(eV)/kT]

Si:     D (cm2/s) = 2.3 x 10-3 exp[-1.7(eV)/kT]

When below 870K, the relationships were:

transition layer:     D (cm2/s) = 9.1 x 10-12 exp[-0.36(eV)/kT]

Si:     D (cm2/s) = 3.4 x 10-12 exp[-0.13(eV)/kT]

It was concluded that a principal feature of interdiffusion in the Si-Ni system was the presence of an amorphous Si layer.

V.A.Uskov, A.B.Fedotov, A.I.Rodionov, N.S.Dumarevskaya: Izvestiya Akademii Nauk SSSR - Neorganicheskie Materialy, 1984, 20[7], 1148-51. (Inorganic Materials, 1984, 20[7], 989-92)

The best linear fits to the solute diffusion data ([124] to [129], [133] to [144], [146] to [176], [188] to [192], [196] to [211], [215] to [223], [234] to [242], [252] to [283], [292] to [298], [306] to [314]) yield:

Al: Ln[Do] = 0.45E – 32.8 (R2 = 0.81); As: Ln[Do] = 0.29E – 23.2 (R2 = 0.87);

Au: Ln[Do] = 0.16E – 12.4 (R2 = 0.16); B: Ln[Do] = 0.29E – 22.6 (R2 = 0.79);

Cu: Ln[Do] = 0.22E (R2 = 0.86); Fe: Ln[Do] = 0.62E – 15.8 (R2 = 0.53);

Ga: Ln[Do] = 0.20E - 16.9 (R2 = 0.78); Ge: Ln[Do] = 0.29E – 23.2.8 (R2 = 0.98);

H: Ln[Do] = 0.17E - 9.9 (R2 = 0.07); Li: Ln[Do] = 0.25E – 9.6 (R2 = 0.48);

Ni: Ln[Do] = 0.29E - 19.4 (R2 = 0.66); O: Ln[Do] = 0.34E – 21.6 (R2 = 0.95);

P: Ln[Do] = 0.35E - 27 (R2 = 0.94); Sb: Ln[Do] = 0.35E – 29.3 (R2 = 0.96);

Si: Ln[Do] = 0.33E - 29 (R2 = 0.86)