Hi
this depends on the coupling strength to the fixed atoms.
An alternative to get a partial set of vibrational frequencies is to use the
MODEL_SELECTIVE algorithm
Sent: Monday, May 8, 2023 5:59 PM
To: cp2k
Subject: Re: [CP2K:18780] VIBRATIONAL ANALYSIS
Hello Dr.Hutter,
Thanks for your response. In this case, would the rest of the frequencies (3N-6) still be reliable? or they are also impacted by the fixed atoms?
Thanks!
On Monday, May 8, 2023 at 2:39:22 AM UTC-5 Jürg Hutter wrote:
Hi
If you fix atomic coordinates, your system looses its free translations and rotations.
For such a system you expect 3N non-zero frequencies.
regards
JH
________________________________________
From:
cp...@googlegroups.com <
cp...@googlegroups.com> on behalf of Hana <
holi...@illinois.edu>
Sent: Friday, May 5, 2023 5:19 PM
To: cp2k
Subject: Re: [CP2K:18774] VIBRATIONAL ANALYSIS
Hello Lucas and Krack,
Thanks for your comments. I have attached my input file (input.txt) for your notice. My system consists of a single ion pair (
sys.xyz<
http://sys.xyz>) with no periodic boundary condition and I use the GAPW method with a 400Ry cutoff (Upon convergence test, this cutoff gave me 1e-4 convergence threshold for the total energy). For the normal mode analysis I fix the atoms of one ion and just study the vibrations of the other one. Please let me know what you think.
To view this discussion on the web visit
https://groups.google.com/d/msgid/cp2k/b9552901-a336-4e82-bbaa-5c3ecf75c851n%40googlegroups.com<
https://groups.google.com/d/msgid/cp2k/b9552901-a336-4e82-bbaa-5c3ecf75c851n%40googlegroups.com?utm_medium=email&utm_source=footer>.