Molcas Forum

cymantren

Member

Registered: 2017-09-18

Posts: 37

Optimization of H3 (with GUGA)

Hello everybody,

I try to do an optimization of dublet/quartet H3 /6-31G**.
---------

&GATEWAY
  coord=H3_optvib_dubl_1.Opt.xyz
  basis=6-31G**

>>> EXPORT MOLCAS_MAXITER=150
>>> DO WHILE

 &SEWARD

 &GUGA
SPIN  = 2
ELEC  = 3
NACT  = 3 0
INACT = 0 0
CIALL = 1
PRINT = 5
 &ALASKA
 &SLAPAF 

>>> END DO

 &McKinley
 &GRID_IT
ALL

---------
After the detection of the Cs-Symmetriy, the output gives....
---------

..
..
              ORBITALS PER SYMMETRY
                  1    2
      INACTIVE    0    0
      ACTIVE      0    0
      VALENCE     0    0
      CORE        0    0
      ONEOCC      0    0
--- Stop Module: guga at Fri Mar 15 09:22:49 2019 /rc=-1 ---
*** files: xmldump
..
...

---------

To be honest, the input is not totally clear to me, because I am not sure, wheter to use a symmetry adapted or an Openmolcas specific input. Here I used the former input, because there are 3 elect. in A' and 0 in A''. This have been fond by a former scf calculation.
The output shows that there are 0 active -WHY ???

It would be very kind, if someone shows me the mistake in my input, or whatever goes wrong.

Thank you in advance !

Best greetings !

Alfred

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: Optimization of H3 (with GUGA)

I think you want to use ACTIve (number of active orbitals) instead of NACT (number of active electrons, alternative to ELEC).

I also believe GUGA will not give you any energy, you have to run e.g. MRCI after... and I'm not sure gradients are supported (not even numerical).