-
Notifications
You must be signed in to change notification settings - Fork 13
Description
I am having difficulties running std2 (version 2025-2-6 10:30:41.979) when the molecule contains sulfur atoms. This could be related to issue #8.
As the simplest example, i took water, and replaced the central oxygen with sulfur.
For water everything seemingly works fine. I run xtb (version 6.7.1), create the molden input file, with which i can run std2 without (technical) problems. I did not validate the resulting excitation energies yet though.
For H2S, running xtb goes fine and the molden file looks OK. However, running std2 gives an “internal MO norm error”. Running the std2 program with the -chk flag i get a discrepancy between the mulliken populations, and there is a crash. This is as in issue #8.
Optimization of the H2S prior to running the std2 -check does not alter the situation, although the number of electrons from the Mulliken population analysis is different.
I compiled both programs with the intel oneapi 2022.1.0 compilers and MKL library.
The problem seems related to the integral computation with the libcint, as I can run std2 on both water and H2S without it (-libcintOFF). Perhaps i should also report the issue there?