“…On the other hand, as pointed out in our previous paper, 2) because the sputtered atoms can migrate for a relatively long time on the substrate to arrive at suitable lattice sites under the low-deposition-rate condition, this effect is equivalent to an increase in T s . In addition, as described in the previous paper, 2) if we assume that the two-dimensional superlattice cell of (111)Rh and that of (11 " 2 20) sapphire consist of 14 Â 17 unit cells and 5 Â 11 unit cells, and that the twodimensional superlattice cell of (001)Rh and that of (11 " 2 20) sapphire consist of 24 Â 13 unit cells and 7 Â 6 unit cells, respectively, the lattice mismatches for the two combinations of two-dimensional superlattice cell are approximately the same. Therefore, because the difference in film/substrate interfacial energy between the systems of (111)Rh/(11 " 2 20) sapphire and (001)Rh/(11 " 2 20) sapphire is small, it is expected that the epitaxial growth plane of Rh films on (11 " 2 20) sapphire may be controlled by changing T s instead of P rf as discribed in our previous paper.…”