AM1 parameter bug

[Letif Mones]

Dear CP2K users,

I found a bug related to an AM1 parameter of the phosphorus atom. It is in file semi_empirical_parameters.F line 1258: 
sep%fn3(2) = 0.03329000_dp
However, the corresponding correct parameter should be:
sep%fn3(2) = 1.77935400_dp

With this correction the energy and forces of a test system (some phosphate) are consistent with those obtained from other programs (sander/gaussian/mopac).

Kind regards,

Letif