Hagen Neugebauer

Dear Andres, In principle, one can calculate the spin constants via numerical differentiation of atoms using DFT (see for example [Fe](https://pubs.acs.org/doi/abs/10.1021/jp068802p?casa_token=s47O5Uk59sAAAAAA:sO-UnPq-iVtgeqcXL-VhTVHdpf8hR0AYlzkDoxKsAza8qLzPADyuoNzrvmorEaQl82UXwRHffh1_aoOZ)). Also, we are currently working on generating and thoroughly...

We don't have support for NPT ensembles, but you can employ all the GFNn-xTB methods in [DFTBplus](https://dftbplus.org/about-dftb), which has support for NPT ensembles.

None of the GFNn-xTB methods has spin-dependent terms in their energy expression so there is no spin density available. We are working on the incorporation of spin polarization into the...

For what purpose do you want to generate this ensemble? Can you say if this is just a GFN artifact? Does the same protonation occur in a DFT optimization with...

This works fine for me, once I set 'ulimit -s unlimited' and 'export OMP_STACKSIZE=4G' xtb he834.xyz --namespcae test ``` * xtb version 6.5.1 (579679a) compiled by 'ehlert@majestix' on 2022-07-11 ......

This seems to be a specific problem for linear molecules with the way we calculate the numerical hessian.

We generated spin constants with PBE and tested them for spin splittings of 3d, 4d, and 5d transition metal complexes with GFN1 and GFN2-xTB. You can find them in the...

What would be the purpose of such a feature? The GFN methods were parameterized with a dispersion correction and would be incomplete without it.

Just set the spin constants to zero for the 2s. Best Hagen

Hi, SCF problems are expected for metal oxides, which have not been the focus of the GFN1-xTB development. You can try to increase the temperature of the Fermi-smearing : xtb...