[zhk2611]

Hi guys,
I am new to the Gaussian program. Now I am using Gaussian 09W to optimize the geometry of some organic molecules. The problem I encountered- while using keyword Opt, the program often gives out results with different topological connections of carbon atoms (i.e. breaking rings of carbon atoms in the input structure and forming rings elsewhere). So is there a way to tell the program to keep the bonds not-to-be-eliminated?
Thanks. 

[ropy]

make sure the input molecule is close to the local energy minima . Dont give any strained molecule. As gaussian 09 finds closest "local" minima not the "global" ,SCF is converging to the minima closest to your input and stopping there . You may find many optimised structure according to your input if you play with it a little bit . 

[curiouscat]

Hi guys,
I am new to the Gaussian program. Now I am using Gaussian 09W to optimize the geometry of some organic molecules. The problem I encountered- while using keyword Opt, the program often gives out results with different topological connections of carbon atoms (i.e. breaking rings of carbon atoms in the input structure and forming rings elsewhere). So is there a way to tell the program to keep the bonds not-to-be-eliminated?
Thanks.

No. You can't. Not typically. There are ways to constrain bond lenths etc. in DFT codes but not typically used.