CCL: MOPAC2007 polarizability calculation

["Chern Chuang" <r96223127%ntu.edu.tw>]

 Sent to CCL by: "Chern  Chuang" [r96223127:_:ntu.edu.tw]
 Hi, I have downloaded MOPAC2007 package, both windows and linux version, and
 tried to calculate the polarizability and the first and second
 hyperpolarizabilities of some fullerenes, for which the simplest case is C60. I
 used keyword POLAR. And the program did run a successful geometry optimization
 process, but the calculated polarizabilities are seemingly very strange to me:
 all the components of gamma (2nd hyperpolarizability) are zero. Furthermore, for
 fullerenes with non-spherical shapes, for example D5h C70, the calculated
 polarizability alpha does not come out with the expected anisotropy, that is,
 the three diagonal elements of alpha are all the same and seemed to scale
 linearly with the size of the fullerene molecule I used for input. Both windows
 and linux version give the same result. I checked the manual from the official
 MOPAC website and couldn't find a trace of the solution to this kind of problem.
 Did I missed something?
 Great thanks to any help!
 Chern Chuang
 r96223127|,|ntu.edu.tw
 Chem. Dept., National Taiwan Univ., Taipei, Taiwan