My Community

Dear all,
We can optimize a specific excited state by using the FERWE and FERDO flags. Has anyone used this approach to calculate frequencies using DFPT? I calculated the excited-state frequencies with the FERDO and FERWE flags but encountered negative frequencies (-11 THz). Is this reasonable?

Hi,

We would need more detail about your calculation. Could you please provide the input files (but not POTCAR), as well as the OUTCAR file, as explained at forum/viewtopic.php?t=17928?

Dear abien_tran1,
Inupt and outcar files are attached.

Did you relax the positions of the atoms in the unit cell for this particular excited state? If not, then the forces on the atoms may be non-negligible, leading to negative phonons frequencies.

Besides, note that you are using DFTP, which means that only the phonons at the GAMMA point are calculated.

Dear fabien_tran1,
Yes. I optimized the geometry for the excited state.

I optimized your structure in the ground state (i.e., FERWE and FERDO are not used), and then applied DFPT. I also obtained negative frequencies and the largest is even 10 times larger in your case. Did you also calculate the phonons frequencies for the ground state and obtained negative frequencies?

Yes. I optimize the ground state and I didn't get any negative frequencies.

Ok, could you please provide the OUTCAR files of these calculation for the ground state (geometry optimization and frequencies)?