Molcas Forum

dommett110

Member

Registered: 2017-03-01

Posts: 16

CASSCF crossing points of differing multiplicity

Hi all,

I am looking for some help in optimising conical intersection geometries in Molcas. I am looking to optimise the geometry of the crossing point between the ground state singlet (S0) and the first triplet state (T1). Would the following input file calculate optimise to this point on the potential energy surface?

&GATEWAY
 Title = m4 t1/s0 CI
 coord  = m4.xyz
 basis = 6-31G*
 group = C1
Constraints
   a = Ediff
  Value
   a = 0.001
End of Constraints
>>> EXPORT MOLCAS_MAXITER=300
>>> Do while

&SEWARD


>>> If ( Iter = 1 ) <<<


 &SCF

>>> EndIf <<<

&RASSCF
LumOrb
Spin = 1; Nactel = 8 0 0; Inactive = 45; Ras2 = 8
CiRoot
2 2; 1 2; 1 1
Rlxroot=1
&ALASKA
 PNEW
>>COPY $WorkDir/$Project.RunFile $WorkDir/RUNFILE2

&RASSCF
   LumOrb 
   Spin = 3; Nactel = 8 0 0; Inactive = 45; Ras2 = 8
   CiRoot 
   2 2; 1 2; 1 1
&ALASKA
 PNEW
 &SLAPAF


>>> EndDo

I think that for my second RASSCF, I am not specifying the correct roots - however I am not exactly sure what they should be in this case, where a triplet is involved.

Does anyone have any experience running these calculations and could offer some assistance? Note that at this time all I am looking for is the crossing geometry, and will calculate properties such as spin-orbit coupling at a later point.

Thanks

Michael

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: CASSCF crossing points of differing multiplicity

In the second RASSCF you are requesting two roots, presumably T1 and T2 (if the ground state is S0). But since you don't specify RlxRoot, it will default to the highest root, i.e. T2. So that file will try to optimize a crossing between S0 and T2. You should add RlxRoot=1 to the second RASSCF too.

dommett110

Member

Registered: 2017-03-01

Posts: 16

Re: CASSCF crossing points of differing multiplicity

Ah I see, thank you!

Further, is is possible to optimise crossing points of different multiplicities using State-Averaging, where the same state-averaged orbitals are used for each of the singlet and triplet calculation: ie the triplet state is an excited state from the ground state singlet state (the reference state).  My understanding of the method above is that different orbitals will be used so states being compared are not equivalent?

Thanks

michael

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: CASSCF crossing points of differing multiplicity

It is currently not possible to "mix" different symmetries (spatial or spin) in a single state-average calculation.

dommett110

Member

Registered: 2017-03-01

Posts: 16

Re: CASSCF crossing points of differing multiplicity

I see. Thank you for your help, it is much appreciated