![\fbox{
\parbox[h][\height][l]{12cm}{
\small
\noindent
{\bf Reference literature:...
...newblock {\em J.Chem.Phys.},
{\bf 105},\hspace{0.25em}2804, (1996).
\end{list}}}](img41.png)
The calculation of the rotational g tensor is invoked through the keyword .MOLGFA in the **PROPERTIES input module. A complete input file for the calculation of the molecular g tensor is thus:
**DALTON INPUT .RUN PROPERTIES **WAVE FUNCTIONS .HF **PROPERTIES .MOLGFA **END OF DALTON INPUT
The molecular g tensor consists of two terms: a nuclear term and a
term which may be interpreted as one definition of a paramagnetic part
of the magnetizability tensor as
described in Ref. [74].
By default the center of mass is chosen as
rotational origin, as this corresponds to the point about which the
molecule rotates. The use of
rotational London atomic
orbitals can be turned off through
the keyword .NOLOND.
For a SOPPA or SOPPA(CCSD) calculation of rotational g tensors the additional keywords .SOPPA or .SOPPA(CCSD) have to be specified in the **PROPERTIES input module. For SOPPA an MP2 calculation has to be requested by the keyword .MP2 in the **WAVE FUNCTIONS input module, whereas for SOPPA(CCSD) a CCSD calculation has to be requested by the keyword .CC in the **WAVE FUNCTIONS input module with the *CC INPUT option .SOPPA(CCSD). This will also disable automatically the use of London orbitals.
The basis set requirements for the rotational g tensors are more or
less equivalent with the ones for the molecular magnetizability,
that is, the augmented cc-pVDZ of Dunning and
Woon [65,67], available from the basis set library
as aug-cc-pVDZ.
If more close control of the different parts of the calculation of the rotational g tensor is wanted we refer the reader to the section describing the options available. For the calculation of the molecular g tensor, these are the same as listed above for magnetizability calculations.