Are there any suggestions as to why a particular basis set is preventing Gaussian from saving NBO to the checkpoint file?

25 February 2015 8 348 Report

I have been approached by a colleague who is having difficulty executing OPT->FREQ->Pop(NBOs) over a number of structures.

So I performed the OPT and FREQ separate from the population analysis using the following (theory/basis-set, consistent with OPT and FREQ):

# b3lyp/6-31++g(d,p) scrf=(checkpoint) geom=allcheck pop=(nbo,savenbos)

I determined a number of things.

1) If I remove "savenbos", it would work. But I then lack the 'surfaces' of the orbitals in my checkpoint file. I am only left with reports in the .log file.

2) Replacing b3lyp with wb97xd, it fails!

3) Replacing 6-31++g(d,p) with lan2ldz coupled with either b3lyp of wb97xd, it works!

2) When I take out the diffuse functions (6-31g(d,p)) and couple with either b3lyp or wb97xd, it works!

For those that failed I have included the end of the log file at the bottom of this entry.

So it appears as though diffuse functions in my basis set are causing a problem for NBO population analysis coupled with "savenbos". Any ideas why?

Many thanks

Anthony

Labels of output orbitals: BD LP LP BD BD BD BD BD BD BD BD BD BD BD BD BD BD LP LP BD

Labels of output orbitals: BD BD LP BD LP BD BD LP LP LP BD BD* BD* BD* BD* BD* BD* BD* BD* BD*

Labels of output orbitals: BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD* BD*

Labels of output orbitals: BD* BD* BD* BD* BD* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY*

Labels of output orbitals: RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY* RY*

Failed in SchOr1 in NBStor.

Error termination via Lnk1e in /usr/local/g09_d01/g09/l607.exe at Wed Feb 25 09:29:32 2015.

Job cpu time: 0 days 0 hours 39 minutes 55.8 seconds.

File lengths (MBytes): RWF= 184 Int= 0 D2E= 0 Chk= 26 Scr= 1

Bartosz Trzaskowski

NBO implemted in Gaussian has some technical problems with handling large systems and using "savenbos" options and selected basis sets (particularly large ones). I have never encountered them, though I've been extensively using wb97xd wit 6-311++G** basis sets (but no savenbos option), but I remember that this problem was documented:

http://www.ccl.net/chemistry/resources/messages/2010/11/16.004-dir/index.html

http://joaquinbarroso.com/category/white-papers/nbo-white-papers/

and it looks like it was not addresses in the next gaussian/nbo versions.

Anthony Nash

Oh dear! That is a real shame it hasn't been resolved yet. Thanks so much for the information. Like you, I too have been working extensively with wb97xd/ 6-311++G** or Lan2ldz, and I've never had problems. Yet this is the first time where I've had to really look into NBOs.

Thanks for the information.

Kurt A Kistler

Anthony, there are other triple-zeta diffuse sets that might work, such as aug-cc-pVTZ. Or try removing one star and have diffuse only on heavy atoms. I've never had this issue with added polarization functions, but I could see there being an issue with diffuse functions in some systems, since they can extend out onto other atoms. Do you need diffuse functions for your system?

Anthony Nash

Thanks Kurt - This is actually someone elses system I am helping them with, so I will ask them whether they need (diffuse) or not, although I suspect they don't. Either way, I have tried with 6-31+g(d,p) and I get the same error. If I remove diffuse altogether then the error goes and the NBOs are saved to the checkpoint file.

Many thanks for the suggestion of aug-cc-pVTZ. I will do some background reading to see if it is appropriate for my colleague's system.

Thanks again

Martin Schwarzer

If you desire to just plot the orbitals, you can invoke the generation of several files via pop=nboread in the route section and in an additional section at the end of the file $nbo PLOT $end. This will generate a set of FILE.## with ##= 31..40. I personally use ChemCraft as a visualizer, so I am not certain which other programs are capable of displaying these files. (Find more information in the attached manual.)

I also suggest changing to a better basis. The Pople basis sets can be considered antiques, that are nice to look at and even nicer to talk about, but they were not optimised for today's computations. The previously mentioned Dunning family of correlation consistent or the Ahlrichs def2-family are both reasonable choices and much better performing.

http://www.ccl.net/cca/software/NT/mopac6/nbo.htm

Anthony Nash

Thanks Martin,

I've switched to aug-cc-pVTZ.

Ajay Khanna

Hey Hi,

I recently bumped into this error, I used b3lyp/6-31+g(d,p). However, as I removed the diffuse function everything worked fine. So If your system is not affected much by the addition or removal of this diffuse function "+" remove it and re run the calculations, it will work fine.

If not try switching between different basis sets like aug-cc-pVTZ to see which works best for your system.

Hope this Helps,

Regards

Ajay Khanna

Fatima Fadl

actually I am about to post the same question, I am using the same method with some structures, it was all okay but it gave an error with the last structure. there is no way to change the basis set or remove the diffuse function.. has anyone found the solution?

Does the pressure(-600 mmHg) affects pH?

After a reference from Int J Body Compos Res, can any body help?

Do anyone here can help us in the field of molecular docking?

Analysis of CRD two factor using R studio ?

What is the exact mechanism of the antimicrobial effect of local anesthetics?

How can I selectively dry etch Ti on a polymer substrate?

Resistivity of PVD TiNx thin films vs. N2 flow?

In what ways do fellow reaerchers think that mRNA vaccination may be under CRISPR systems technology application ?

What are your alternative data collection methods during the pandemic? What are the pros and cons?

Can bio-decomposer be a successful alternative to reduce residue burning?

error in running a model in Abaqus

error in modelling model in Abaqus

MD MOE RMSD

MD MOE RMSD

MGL tools

fatal error

fatal error

fatal error

fatal error

i have some problem with chtMultiRegionSimpleFoam