negative frequencies of optim. geometry

[eijck]

Hi,

I have done a geometry optimisation PREC=accurate and the residual forces are all below 10e-10 eV/A. When I do a freq. calculation a large part of my frequencies are imaginary/negative. I've tried a stepsize POTIM of 0.02 and 0.005, but there's always a large amount of imag.freq's.

The molecule is 4 thiophene rings (C,S,H) and I'm using paw/LDA potentials.

Does anyone have a suggestion what settings to test/check.

I'll send input/output files if needed.

cheers,
Lambert van Eijck

[matlgen]

In general, negative (imaginary) frequencies mean that you have found a transition state. Transition states should have (roughly) zero forces, so what you are seeing is consistent. Try moving the atoms along one of the imaginary frequency motions and reoptimizing.