Problem with Gaussian cube-files

When I produce a Gaussian cube-file editing a molecule in MarvinSketch (example indigo), computing Tools-Geometry-Geometry and saving it as Gaussian cube file, I get the file indigo1.cub. Opening this file with GaussView I get the picture you see in the screenshot. What's wrong ?

Hi Hans-Ulrich,





this can be an Angstrom-Bohr representation issue.


Can you give us the generated cube file and also the version number of GaussView?


Is the saved structure correct if you just sketch and save it (without computing Geometry)?





Kind regards,


Judit

To discuss things detailed I will send two posts.





First the cube-file indigo1.cub produced with Marvin 4.1.13, calculating the geometry and saving as indigo1.cub.





Bildschirmphoto235: The file indigo1.cub





Bildschirmphoto236: Opening indigo1.cub with Marvin and showing the bond lengths. Looks good and seems to be Angstrom.





Bildschrimphoto237: Opening indigo1.cub with GaussView 3.0.7. To much bonds calculated from the Cartesian coordinates.





The next post comes immediately after this one.

Producing the molecule with Marvin, adding explicitly the H-Atoms and saving as indigo3.cub





Bildschirmphoto239: The file indigo3.cub





Bildschirmphoto238: Opening indigo3.cub with Marvin. Standard C-C etc. bonds, but C-H-bonds are very long (Coming of course from the standard bond length in drawing the molecules).





Bildschirmphoto240: Opening indigo3.cub in GaussView.

Hi Hans-Ulrich,





there was a problem in the cube file export in Marvin 4.1.13, but it is already corrected in 5.0.0.


The cube files generated with 5.0 seem to appear correctly, though we could test it with GaussView 4.0 only.





Kind regards,


Judit

Hi Judit,


this is a very good message. As a long year programmer, I hesitate to use x.0.0 versions. But in this case I will do it (in spite of the problem with the help button I've already posted). See Screenshot243, opening indigo4.cub with GaussView 3.0.7.


I have already had the suspicion that the failure is in line 4-6. Seems to be a 3x3 matrix for the xyz-Coordinates. In the correct file there are the conversions au / Angstrom, see Screenshot244.





And there is a new feature in Marvin: You can now produce Gaussian *.com files. We are using Gaussian for ab initio calculations and now we have two ways to produce input files: GaussView and Marvin.





Kind regards, Hans-Ulrich

Glad to hear this! :)





You might already know, that we also support displaying volumetric data with MarvinSpace, see the last row in the image gallery for an example: http://www.chemaxon.com/marvinspace/gallery.html





Kind regards,


Judit