CCL: [CCL] CCL: Negative frequencies with C1 symmetry (Orca)

[Peter Jarowski <peterjarowski!^!gmail.com>]

Hi All:

A few comments.

Do we have a reference for this 10 cm-1 rule-of-thumb? Has anyone published a comparison study using different softwares? I know NWchem ignores lateral frequencies (« imaginary frequencies »). So careful the outcome might vary even assuming the same implementation of the partition functions. I guess we are talking sub Kcal/mol (cal / mol K :298 K) variations in thermal corrections.

For solvatation, its tricky. The hard and fast rule is that the frequencies must be evaluated at the same level as the optimization. Thus, if you optimize in PCM implicit solvent you must do the thermal corrections in solvo as well.  It often makes no significant difference, but I imagine it would introduce some small lateral frequencies if you do it in gas phase. In NWChem it requires numerical integration and takes forever.

Best regards,

Peter

On Sun, 10 Jul 2022 at 07:57, Sachin Ramesh sachinadityaramesh() gmail.com <owner-chemistry:ccl.net> wrote:

Hi,

   Imaginary modes under '-10 cm-1' if not many can be safely ignored. You could try to get rid 

of the imaginary mode by distorting the structure along the bond, angle or dihedral which is 

causing the imaginary mode. 

   I should add that imaginary modes are tricky to get rid of when the solvent correction is used. 

The optimization could be done using solvent correction and frequency can be computed in gas phase. 

  'TightSCF TightOpt Grid7 Gridx9' seems to work for all of our systems upto 200 atoms, producing 

accurate results in good agreement with experimental observations.  

Hope this helps 

with best 

- Sachin 

On Sat, Jul 9, 2022 at 7:32 PM Grigoriy Zhurko reg_zhurko.:.chemcraftprog.com <owner-chemistry**ccl.net> wrote:

Sent to CCL by: "Grigoriy  Zhurko" [reg_zhurko()chemcraftprog.com]
Hello,
I compute some molecules with Orca 5.0.3, and the computation sometimes
produces small negative frequencies, despite the symmetry of the molecules is
C1. If I repeat the computation with another starting point, with some
probability all frequencies are positive.
I found that this problem arises when two additions to the model are added:
the solvent model (!CPCM(Water)) and additions of some explicit water
molecules to the whole model. I understand that these water molecules produce
very small frequencies along the h-bonds, and these frequencies can become
negative because of some problems with numerical integration. I tried the
combination of options !DefGrid3, !TightOpt, !VeryTightSCF, and still
sometimes the negative frequencies are produced. Maybe I need to further
increase the computational accuracy with options like "!VeryTightOpt",
"!DefGrid9", UltraTightSCF? I suppose you understand what I mean; currently
I didn't find the information in the manual, how to further decrease the
optimization convergence threshold or increase the accuracy of DFT grid.
Please suggest how to do that.
In my work I can avoid using the frequencies since I need mainly the
energies; however I need to explain this somehow in the paper. I suppose, it
is not good to compute the entropy of molecules with my keywords, because
small frequencies produce big errors with vibrational entropy. So, please
suggest me, what should I write in my papers, to explain that it is not good
to compute the Gibbs energy of my molecules, but it is correct to compute and
use the common energy.

Grigoriy Zhurko
https://chemcraftprog.com



-= This is automatically added to each message by the mailing script =-
E-mail to subscribers: CHEMISTRY**ccl.net or use:
      http://www.ccl.net/cgi-bin/ccl/send_ccl_message

E-mail to administrators: CHEMISTRY-REQUEST**ccl.net or use
      http://www.ccl.net/cgi-bin/ccl/send_ccl_message
      http://www.ccl.net/chemistry/sub_unsub.shtml

Before posting, check wait time at: http://www.ccl.net

Job: http://www.ccl.net/jobs
Conferences: http://server.ccl.net/chemistry/announcements/conferences/

Search Messages: http://www.ccl.net/chemistry/searchccl/index.shtml
      http://www.ccl.net/spammers.txt

RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/