Posted by Evgeniy on December 08, 2009 at 09:01:49:
Dear Developers of Molcas,
I have a question concerning geometry optimization with the constraint Ediff=0.000 used to search for the minimum energy intersections between two states (using the procedure given on page 51 of the manual; the example of singlet-triplet crossing in acrolein). Although I am interested in singlet-singlet crossing (in a large system) I guess the procedure should also work, shouldn't it? However as the optimization runs i observe that it leads to some completely bizarre structure where some hydrogens lie very close to each other and the energy always goes up. The optimization has not yet finished, but I am warring whether it is running correctly in view of the above. In particular, I am wondering, whether it is important which state is assigned to RunFile2. Should it be the lowest state or the next state (to be assigned to RunFile2)?
Many thanks!
Best regards,
Evgeniy