UT theoretical chemistry code forum

In addition to my previous post...

I have read on this forum about similar problems on small symmetric molecules like ethylene. In there, it was suggested to use the keywords "-vac 1e-5" to remove the density minima close to the nuclei. I did it, but the charges are still different! So no good solution...I also tried to calculate the Voronoi charges on the same systems, but they are different too.

I am using the 0.28a version of the algorithm (7/12/2012), with an FFT-grid of 56 x 115 x 79. Does anyone know how to increase the number of the grid points (this was another possible solution to the problem)?!?

Please help me!! Thank you.

Andrea.

Hi andrea,

From my limited experience of this program (so far!), it appears that the small dimensions of your grid are almost certainly giving rise to your problem.
Using the cubegen facility that ships with Gaussian, it is possible to specify a larger number of grid points. Check out the page on the 'cubegen' utility in the online g09 manual and it will talk you through exactly how to achieve this.

Basically all you need to do is execute the program with the following syntax (where all the terms apart from 'cubegen' are parameters that you are able to change).

cubegen memory kind fchkfile cubefile npts format

In order to change the number of points you simply specify a value of npts. The number you input will be cubed and this will be the number of points in your cube grid.

Hope this helps,

Cheers,
Chris

Hi Chris,

thank you for your suggestion, it seems to work! If I use a medium (-3) or fine (-4) grid with the cubegen utility the atomic charges tend to be identical! Also the total number of valence electrons is closer to the expected value than before.

Great help!

Ciao.

Andrea.