Hi,
I've a problem to find the local / global minimum. I optimise the structure of a organic crown ammounium complex.
I used to TPSS(MARI-J)-D3(BJ)/def2-TZVP, $ scfconv 8, gridsize m4 and jobex -ri -c 100 -gcart 4 > xxx.out. Everything else is default.
There is no problem with BP86 (MARI-J) without D3(BJ =Becke Johnson) . With TPSS(MARI-J)-D3(BJ) I have problems. I have lound a few local minima. But I have found a structure which is energetically lower all local minma. This structure has two imaginary frequencies. So it should exist a deeper Minium. With screwer I could eliminate one imaginary frequency and reach a more energetically deeper structure. This structure still contains one imaginary frequency. I can not get rid of this imaginary frequency. I tried screwer with +2000 Kelvin and -2000 Kelvin . The structural optimization lands again in the same structure with an imaginary frequency.
Is there a way to use a different optimization mechanism, or to influence optimization in any way (with special commands)?
Or is it possible that because of the dispersion correction may this transition state is a ground state, because DFT and dispersion correction have a conflict (D3(BJ) try to bring the atoms as close as possible, DFT try to get a covalent distance)?