Multiwfn forum

johnnyy

Member

Registered: 2024-07-16

Posts: 1

Integration inaccuracy for heavy atoms

Hello,

I have a problem that I couldn't find an answer to in the manual for the outputs of integration inaccuracy for heavy atoms. Basically I wanted to do Basin analysis >  Generate basins and locate attractors > ELF > Lunatic quality grid > Integrate real space functions in the basins > Electron density (rho). For smaller atoms like carbon, oxygen, fluorine the integration seems to be alright (loss of 0.001 electron), but for atoms like chlorine, bromine, etc. the outputs are quite inaccurate (even up to 0.5-0.7) electrons. So my question is: is there a way to get more accurate results like maximum loss of 0.1-0.2 electrons?

I tried setting my own grid with Output and plot specific property within a spatial region (calc. grid data) > ELF and then I tried almost every option available there, mainly option 4, 5 and 8 using already made cube file to use it as a grid for integration. I also tried setting different parameters to see how the results change. I got quite confused cause for using option 8 (with cube file) BrCl (that had loss of 0.8 electrons normally) had 52.03 electrons (BrCl has 52 el.) and HBr had 35.01 electrons which seemed weird to because its lighter than BrCl. I got couple of results where BrCl got results of 100 electrons. So basically sometimes the results where better sometimes worse and I'm not sure why it happens.

I'm quite new to this topic and I wanted to know if there's a way to get better results.

Thank you in advance!
John

sobereva

Tian Lu (Multiwfn developer)

From: Beijing

Registered: 2017-09-11

Posts: 1,777

Website

Re: Integration inaccuracy for heavy atoms

Dear John,

Multiwfn employs uniform grid for integrating ELF basins. The integration accuracy for valence ELF basins should be fine if grid spacing is small, however it is not expected that integration accuracy for core basins is also satisfactory for heavy atoms, because electron density varies quite sharply around their nuclei. Usually only valence ELF basins are of interest, it is not a problem that core basins cannot be well integrated using uniform grids.

Best regards,

Tian