Alex Granovsky
gran@classic.chem.msu.su
In other words, you need to prepare your input as if it were a regular
MCSCF surface scan, set MPLEVL to 2, provide desired XMCQDPT input
group, and specify runtyp=rsurfx.
Hope this helps,
Alex
On Wed Jul 5 '17 11:07pm, Alex Granovsky wrote
----------------------------------------------
>Dear Bernhard,
>runtyp=rsurfx works similarly to regular runtyp=rsurface with a few exceptions.
>First, CITYP is forsibly set to NONE, and MPLEVL is set to 0;
>and a constrained geometry optimization is performed.
>When done, CITYP and MPLEVL are restored to their initial values
>and a single point energy is computed.
>These steps are repeated in the scan loop.
>Hope this helps.
>Kind regards,
>Alex
>
>
>
>
>
>
>On Tue Jul 4 '17 6:15pm, Bernhard Dick wrote
>--------------------------------------------
>>I wonder whether it is possible to make a relaxed surface scan with a MCSCF wavefunction, and do a many body perturbation calculation for each point after convergence of the geometry optimization. I though I had seen that somewhere, but can not find it.
>>Also, in the manual a RUNTYP=RSURFX is mentioned, but nowhere explained. The option is accepted, but what are the properties of this mode?
>>best regards,
>>Bernhard
>>