VSCF frequencies look strange

[Matthew Graneri]

Hi there,

I'm trying to test out the VSCF output from NWChem 7.2.3, but I am having a bit of trouble. I want to calculate the anharmonic fundamental frequencies of ammonia and water, as well as the first overtones and combination bands, coupling three vibrations. The input for both of the calculations takes the form:

title "Water test"

echo

print low

geometry

 zmat

 O

 H 1 1.0

 H 1 1.0 2 100.0

 end

end

basis spherical

 * library cc-pvdz

end

mp2

  freeze atomic

end

driver

  tight

  GMAX 0.0000001

end

vscf

  coupling triplet

  iexcite 2

end

task mp2 optimize

task mp2 freq

task mp2 vscf

I cannot attach the NH3 output file here (it's >600MB), but the final results are:

  RESULTS OF VIBRATIONAL SCF CALCULATION: Frequencies, cm-1

  Mode      Harmonic     Diagonal      VSCF       PT2-VSCF

   1       10886.08     11016.32     10681.36     10452.07

   2       10886.08     11295.53     10630.50     10842.60

   3       10458.22     10712.71     10609.35     10437.15

   4        5063.60      5427.09      5198.71      5043.55

   5        5063.60      5432.63      5196.03      5042.98

   6        3482.56      3650.69      3091.60      2718.97

 ......Finished vibrational SCF......

  IR INTENSITIES CALCULATED USING DIPOLE MOMENT:

  mode   frequency, cm-1  intensity, km/mol

   1       10452.07          0.04

   2       10842.60          0.01

   3       10437.15          0.00

   4        5043.55          0.01

   5        5042.98          0.01

   6        2718.97          1.20

These frequencies are very high. The harmonic frequencies in the same output file are listed as 1160.862, 1687.876, 3486.094 and 3628.716 cm–1, which is far closer to where they should be.

The same thing happens in the water calculation (I have tried attaching the output below, but it is also too big):

  RESULTS OF VIBRATIONAL SCF CALCULATION: Frequencies, cm-1

  Mode      Harmonic     Diagonal      VSCF       PT2-VSCF

   1        7941.89      8168.85      8003.97      7917.62

   2        7703.43      7979.99      7927.74      7790.98

   3        3355.55      3577.93      3389.69      3342.26

 ......Finished vibrational SCF......

 

  IR INTENSITIES CALCULATED USING DIPOLE MOMENT:

  mode   frequency, cm-1  intensity, km/mol

   1        7917.62          0.01

   2        7790.98          0.82

   3        3342.26          0.05

Again, the harmonic frequency calculation gives a far more accurate answer, with frequencies of 1677.784, 3851.738 and 3970.972 cm–1.

I'm not quite sure what I might be doing wrong. I tried separating the optimisation and VSCF calculations, just in case it was using the input geometry as opposed to the optimised geometry, but that did not help.

Am I looking in the wrong spot to find the fundamental frequencies? Or have I written something incorrectly in the input?

Any help would be very much appreciated!

Regards,

Matthew

[Edoardo Aprà]

I am not familiar at all with the VSCF module.

The only suggestion I have is to have a look at the input and output files for the h2o_vscf QA test case (that uses Hartree-Fock)

https://github.com/nwchemgit/nwchem/tree/master/QA/tests/h2o_vscf

[Matthew Graneri]

Yeah, the test case was the first one I ran. I assume the results are correct: the frequencies are a little bit high, I assume that's because it is not using any electron correlation. They are nowhere near as high as the MP2 calculations, though.

[Daniel Mejia Rodriguez]

Hi Matthew,

It seems that you are comparing first overtones (iexcite=2) with the fundamental frequencies obtained from the freq task

The ammonia frequencies are even larger, do you by any chance used iexcite=3 in that case? They seem more like second overtones 

Daniel

--
You received this message because you are subscribed to the Google Groups "NWChem Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nwchem-forum...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/nwchem-forum/783e2d02-87d8-492c-a3ca-b2fc33d6bf01n%40googlegroups.com.

[Matthew Graneri]

Hi Daniel,

Ah, maybe that's my mistake—I set iexcite=2 for water and iexcite=3 for ammonia, assumed that would print the fundamentals, first overtones and combination bands for water and those plus the second overtones for ammonia.

Do you only get the overtones if you set iexcite? Do you actually have to recalculate all those grid points each time you want to get fundamentals and overtones? And how would you get combination bands?

Matthew