I don't well understand your question. If you load a wavefunction into Multiwfn, you can accurately obtain properties at any specified point in main function 0, namely Multiwfn calculate properties at this point directly based on wavefunction and geometry. This function does not involve "grid".

I have a follow-uo question about how the grid is defined to get the property at each point, for example what is the shape of space and how fine is the grid, and can we modify the fineness of grid in multiwfn.

Thank you

Thank you! It works

No. This cube file records the grid data of the function used to define the isosurfaces, namely electron density grid data.

You can directly use main function 1 to calculate various properties at any given point, including ESP. In addition, if you input f12 in main function 1 first, and then input coordinate, gradient of ESP will also be directly printed.

Hi,

I use the wfn file generated from Gaussian 16 and do ESP post-analysis with the Multiwfn. I found there is an option " -2 Export the grid data to surf.cub in current folder". Attached is the surf.cube I generated. How to understand this file? Does it contains the ESP at each grid?

The reason I do this analysis is I want to quantify the local electric field between two atom from eletrostatic potential. One of this two atoms can be replace with other atoms. So we want to build a relation of a local electric field with the different atoms (bond length as well). If I have could have the electrostatic potential at these two atoms' coordinates, it will be easy to obtain the electric field by gradient.

If this export Surf.cube doesn't contain the electrostatic potential of each point in a 3d grid, any other options I can try?

Thank you!