“…This phenomenon is similar to the formation of NiSi 2 by annealing a Ni(Ti)/oxide/Si film stack at 300 C, where the oxide layer acts as a diffusion barrier. 39 In addition, the thermodynamic driving force (negative DG total ) for the nucleation of NiSi 2 on Si at a low temperature can be provided by the following factors: (1) the low interfacial energies 39,40 DG i at the NiSi 2 /Si(100) and NiSi 2 /Si(111) interfaces as compared to that of NiSi/Si(100) interface, (2) the small lattice mismatch (0.4%) between NiSi 2 and Si, 39,41 and (3) the small volume change 42 of only 2% arising from the difference in thickness between NiSi 2 and Si (volume change for the difference in thickness between NiSi and Si is 10%). Furthermore, due to the fact that (111) planes of NiSi 2 are parallel to Si (111) and the nucleation of NiSi 2 is easiest on (111) Si planes, the NiSi 2 crystallites therefore take the form of a pyramidal shape.…”