In Gausssian, I have tried to reoptimize that structure by freezing the bond which is breaking in excited state during optimization. Before freezing the bond i have already tried scf = xqc and scf =nothing i.e default value.
If you use GaussView, the absence of the chemical bond shown with Gaussview in output could be an alert in your calculation. When no bond is shown between atoms, which may mean that the optimized bond length is larger than the default distance defined in GaussView (the change in structure is sometimes due to the improper input that does not assign the right bond lengths and angles). Also, the absence of a bond in the output may signify an optimization problem caused for example by the used method or basis set. So, it's better to conduct a Frequency calculation (no imaginary frequency is sufficient to ensure that your structure is truly minimized) and you have to check the mode of that bond, the bond may be broken if you cannot find the bond translational mode in the output file.
On the other hand, assuming there is really some structure with bonds intact, do a constrained optimization first (opt=modredundant). Freeze that bond (using approximate but not too off bond lengths) you want to freeze using the redundant editor and perform a relaxation. After the calculation is over, remove the restriction and perform a completely unrestricted optimization with a max step that asks for 'small optimization steps' so that the system does not 'jump' to another local minimum, or even the global minimum. All that done, make sure to perform a frequency check with no constraints to ensure what you have obtained is indeed a minimum.
Dear Ravi
I think that you can use a new optimizer algorithm in the opt keyword, for instance opt=rfo. You are in a loop over the last cycles.
- Badges
- Science topic