Trying to display Mulliken charges from Gaussian in Avogadro shows wrong numbers

General Discussion Support
1

Hi all,

I am new to using Avogadro and I am running into an annoying issue I can’t fin solution or explanation to. I ran a simple single-point calculation in Gaussian to find the atomic charges in my molecule and I opened the .log in Avogadro to see the numbers displayed on the atoms (by turning on Labels > Partial charges), only to see that the numbers don’t match the results on the .log.

I don’t really understand why this is happening or how to fix it. Maybe I am understanding wrong what Partial charges are in Avogadro? Is there a way I can display the Mulliken charges listed in the .log?

Environment Information

Avogadro version: 1.2.0
Operating system and version: Windows 11 Enterprise

Expected Behavior

When turning on Labels > Partial Charges I would expect to see the Mulliken charge from the .log file on the atom.

Actual Behavior

For example, in this case I have only one N atom (correctly identified by Avogadro, the element label names it correctly) that shows -0.36 on Avogadro, but the Mulliken charge from the .log is -0.49

2

We can’t support Avogadro 1.2 anymore. I would highly suggest using the most recent release of Avogadro2: Install — Avogadro 1.101.0 documentation

If, for some reason the Mulliken charges from your log file don’t show up as labels in Avogadro2, please post your file and I’d be happy to take a look.

3

Hi Geoff,

Thanks for your quick reply. I downloaded the recommended version, but it also does not behave as expected. Firstly, the .log cannot be opened (‘Error while reading file, whereas for version 1.2 it worked fine). The .fchk file does load though, but still shows different values than those shown in the .log (-0.9 instead of -0.49).

I am attaching both the .log and the .fchk here (since .fchk is not an accepted extension I’ll just share a link - available for 3 days). Looking forward to your response!

sp.log (52.5 KB)

1 Like
4

Yeah, for some reason my build is currently showing me MMFF94 charges, not the Mulliken charges - not entirely sure why that’s happening.

You mentioned the error from the .log file. You grabbed the Windows 1.101 release?

5

Yes, I did get that release.

Is there no workaround to display the correct numbers, maybe, while the solution gets crafted?

6

My impression is that for some reason, the Mulliken charges aren’t even getting imported.

If you need an image really fast, I’d suggest using the custom label tool in the toolbar to add the labels you want.

7

Okay, it’s obvious that the fchk code never imported Mulliken charges. Fixed.

There was also a bug when reading the charges from output files that I didn’t notice because we’re using ORCA more in the group now.

It’ll be about a day before this shows up in the nightly build, but it should work now from either the log file or the fchk.

8

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.