Ah, thank you, Tian Lu! Now i get it why it would crash.

When a wavefunction file is used as the input file for this function, the atomic wavefunctions should employ exactly the same basis set as the loaded wavefunction. Alternatively, simply using a file only containing geometry information like .xyz as input file, then you do not need to worry about the consistency.

Another thing, we've just tested using a DFT-derived .wfn on Linux (same as I shared with you). There seem to be two separate issues at play because now it saves a wfn but topology and nuclear coordinates are still nonsensical.

By the way, generation of promolecule wfn is flawless with xyz as a starting point. Saves a wfn too. Did not realize multiwfn could read xyz's!

I think i was using the 3.8(dev) version, but for a good measure, I have just redownloaded it and reran everything with the same result.
I sent the files to the email shown here.

Due to network problem, I cannot access your files, please send them to my E-mail. Also, please make sure you are using the latest version of Multiwfn, version 3.8(dev) has been updated hundreds of times on Multiwfn website (3.8 formal version has not been released yet).

Dear Tian Lu,

I am using Multiwfn 3.8 on Windows 11. I have been trying to generate a promolecule .wfn file from a .wfn derived from DFT (not that it matters, but PBE0-D3BJ/def2-SVP). The heaviest atom in the molecule is Neon, so no ECPs are involved. When I open a DFT-derived .wfn and do topology analysis on it, it shows exactly what I see in other QTAIM software -- my non-covalent complex. Now, if i do:

-- 1000 -- 17 (with atomwfn folder at the same path as multiwfn executable and source .wfn) -- "yes" to "Do you want to make wavefunction information in memory correspond to the just generated promolecular wavefunction?"

and do topology again, the result looks nothing like my starting molecule: nuclear coordinates and attractors have been moved around beyond recognition. More so, if i try to save the file as a .wfn to check it externally (100 => 2 => 5 => some .wfn path), Multiwfn crashes. Do you have an idea of what this could be, or what I am doing wrong?

Thanks in advance and best regards,
Serhii

PS. The file that I am handling can be found here.