[CP2K-user] Imaginary frequencies upon vibrational analysis - Anyway to get overcome this issue?

[Anantha Venkatraman]

Hello,

Hope everyone is safe and healthy during this time. I have an geometrically 
optimized system (DFT, PBE functional, DZVP basis set) that consists of 110 
atoms. While performing a vibrational frequency calculation on this system, 
I am obtaining one or two imaginary modes (< 40 cm-1). I believe by 
increasing the force convergence criteria during geometry optimization I 
might be able to get better results. However, even with a force criteria of 
0.002 ev/A I am not able to overcome this issue of imaginary modes. My 
objective is to obtain vibrational modes that are all positive (which I 
believe corresponds to a minimum).

I have also tried a longer method where I visualize these particular 
imaginary modes, slightly change the geometry around the atoms that 
contribute to this imaginary mode, reoptimize the resulting system and then 
conduct a vibrational analysis on the newly optimized system. This also has 
not worked.

I am attaching my input and frequency files. Any input is greatly 
appreciated. 

Thank you for your time and patience. 

Sincerely,
Anantha Nagarajan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200409/903eaab5/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ag25_sch3-frequencies.mol-1.mol
Type: chemical/x-mdl-molfile
Size: 1740381 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200409/903eaab5/attachment.mol>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: no_constraint.inp
Type: chemical/x-gamess-input
Size: 3119 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200409/903eaab5/attachment.inp>