HOWTO.txt

Dateiinhalt

FODplot with TURBOMOLE (6.7 or higher)

1. Prepare a normal control file, restricted as well as unrestricted
   WF are possible
2. call setfermi <Temperature> (just setfermi gives some advice)
3. Adjust the value of nue (difference of the number of alpha and
   beta electrons (2*S), e.g. 0 for a singlet, 1 for a doublet, 2=triplet, etc) 
   in the $fermi input line in file control
3. run ridft or dscf 
4. run foden 
5. the file <molden.input> contains the FOD. It can be plotted with
   molden and converted to cube format
6. As stated in the original paper one should check in borderline
   cases another contour plot value (0.002 instead of the 0.005 default)


Important release note: in the version before 1.7.2015, the code
used in case of UKS/UHF calculations an average alpha-beta Fermi energy
to derive the fractional occupations. This leads to a difference w.r.t. to
the published, ORCA derived data which are based on different Fermi
energies for alpha/beta spaces. In some sense the ORCA definition is 
spin-conserving while the TURBOMOLE one is not. Although the TURBOMOLE
definition may be reasonble in special situations, at this point we 
take as a default the more physical, spin-conserving one as published. 
The new version of the foden code implements this also for TURBOMOLE by reading 
the alpha/beta levels and re-calculating the Fermi distribution seperately.
Note that the different defintions lead to apparant differences only in cases where
the alpha and beta Fermi levels are very different (e.g. the trityl radical
shown in the original ACIE paper). The TURBOMOLE code is correct and the 
results are identical to ORCA if nue (see above) is properly set to the 
right S value.