Differences in energy and orbitals in NBO3 and 6

Differences in energy and orbitals in NBO3 and 6

Postby Markus » Wed Jul 08, 2015 8:21 am

Hello,

I'm a little confused concerning differences in the results obtained by NBO3 and NBO6 via Gaussian 09 (Rev. D01). To the background: I'm interested in the second order pertubation energy and their orbitals visualization of a molecule with several ligands. At first I observed different energies for the same orbitals, it's not a big difference, but in prinicple I wonder why? The second issue is related to the visualization of the corresponding orbitals. I tried to visualize both in GaussView 5.0.8 and JMol (via NBO Visualization Helper 2.0) and observed that I can't find the 'right' orbitals in GaussView (calculated from the Fchk file with cubegen in GaussView).
So I set up a test system: water cation. I calculated the NBOs via Gaussian for NBO3 with

Code: Select all
# wb97xd/cc-pvtz opt pop=(full,nboread,savenbos) gfinput formcheck

H2O

1 2
 O             
 H                  1            B1
 H                  1            B2    2            A1

   B1             0.96000000
   B2             0.96000000
   A1           109.50000006

$GENNBO $END
$NBO file=H2O_NBO3 PLOT $END
$COORD



and

Code: Select all
# wb97xd/cc-pvtz opt pop=(full,nbo6read,savenbos) gfinput formcheck

H2O

1 2
 O             
 H                  1            B1
 H                  1            B2    2            A1

   B1             0.96000000
   B2             0.96000000
   A1           109.50000006

$GENNBO $END
$NBO file=H2O_NBO6 PLOT $END
$COORD


The results of the first five alpha orbitals are shown in the attached picture. Why is there such a big difference between NBO3 and 6 in GaussView? Even if you neglect the energetic order the orbitals, corresponding orbitals look completely different, e. g. 2. BD (1) O1 - H3 in NBO3 versus 5. BD (1) O1 - H3 in NBO6. NBO3 orbitals are plotted in reordered way. Secondly, why do the orbitals of NBO6 calculation differ in GaussView and JMol? In JMol the orbitals or their shape (again neglecting energetic order) correspond to the NBO3 calculation, although there are quite small differences (see LP (2) O1).

Thanks,
Markus
Attachments
NBOproblem.png
NBOproblem.png (184.93 KiB) Viewed 5061 times
Markus
 
Posts: 2
Joined: Wed Jul 08, 2015 7:45 am

Re: Differences in energy and orbitals in NBO3 and 6

Postby ericg » Wed Jul 08, 2015 9:55 pm

Markus,

The NBO3 and NBO6 orbitals differ slightly due to several small changes that were made recently to the NBO search algorithm. The principal change is the switch from symmetric orthogonalization (SO) to occupancy-weighted symmetric orthogonalization (OWSO) in the construction of the natural hybrids orbitals. The result is that there are small quantitative differences (as you see in the perturbative analysis) between the NBOs of NBO6 and NBO3.

I'm not a GaussView user, but it appears to me that GaussView correctly receives the NBOs from NBO3 via the formatted checkpoint file when requesting pop=savenbos. The orbitals that you show for GaussView from NBO6 are MOs, not NBOs. It is likely that pop=savenbos with NBO6 doesn't write the NBOs to the formatted checkpoint, so you see MOs instead. The Jmol/NBO6 orbitals are clearly NBOs. They differ in appearance, slightly, from GaussView/NBO3 in part due to the differences described in the preceding paragraph, and perhaps because the isosurface values (and algorithms for generating the isosurfaces) differ somewhat for GaussView and Jmol.

(Note that $GENNBO and $COORD input is not needed in your G09 input files. $GENNBO and $COORD are keylist identifiers for FILE47 that are only used for the standalone GenNBO program.)

Eric
ericg
 
Posts: 277
Joined: Sat Dec 29, 2012 9:31 am

Re: Differences in energy and orbitals in NBO3 and 6

Postby Markus » Thu Jul 09, 2015 1:48 am

Dear Eric,

thanks a lot for your explanations.

Markus
Markus
 
Posts: 2
Joined: Wed Jul 08, 2015 7:45 am


Return to General NBO Discussion

Who is online

Users browsing this forum: Bing [Bot] and 1 guest

cron