Molcas Forum

josejc

Member

Registered: 2016-09-18

Posts: 9

state tracking in a minimum energy path

According to "J. Comput. Chem., 2015, 36, 320", it's possible perfom state tracking of a particular electronic state along a minimun energy path. The above it is already implemented in Molcas 8 (via the RASSI module) but there is little information in the documentation. I include an example:

&GATEWAY
coord=example.xyz
basis=AUG-CC-PVDZ
Group=Nosym

>>  FOREACH  ITER  in  (1  ..  1000)
      &SEWARD

   >>  IF  (  $ITER  =  1  )
      &RASSCF
      LumOrb
      FileOrb=/home/XXXX/........
      ORBListing=NOTHing
      symmetry=1;
      spin=1;
      INACtive=28;
      nactel=10 0 0;
      Ras2=11;
      CIRoot= 3 3 1;
      RLXRoot=3
   >>  ENDIF

   &RASSCF
   ORBListing=NOTHing
   symmetry=1;
   spin=1;
   INACtive=28;
   nactel=10 0 0;
   Ras2=11;
   CIRoot= 3 3 1;
   &RASSI
   HOP

   &SLAPAF
   MEP-search
   MEPStep=0.02

>  EndDo

where the state 3 becomes the state 2 at some point in the MEP. In that point, the HOP option of the slapaf module should make the change of states (as I understand, it should change the RLXRoot option):

-----output of RASSI module
 Checking for a hop to a root lower in energy.
      The scalar products of the CI-vectors:
      CIpr(state1) * CI(state1) =   0.06
      CIpr(state1) * CI(state2) =   1.00
      CIpr(state2) * CI(state1) =   1.00
      CIpr(state2) * CI(state2) =  -0.06

      +------------------------------------------------------------------------------+
      |                                                                              |
      |                           A HOP event is detected!                           |
      |                                                                              |
      |                        From state:  3    To state:  2                        |
      |                                                                              |
      +------------------------------------------------------------------------------+

however, the slapaf module still take the information from the state 3 for the MEP analysis; moreover, in the next iteration the gradient it is still evaluated for the state 3, and the HOP option currently doesn't work.

My question is, what is the right way of use the HOP option to perfom state tracking?
Am I missing something in the Molcas input?

I would greatly appreciate any help you could provide

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: state tracking in a minimum energy path

According to "J. Comput. Chem., 2015, 36, 320", it's possible perfom state tracking of a particular electronic state along a minimun energy path. The above it is already implemented in Molcas 8 (via the RASSI module) but there is little information in the documentation.

and this part of the code is still changing, expect new input syntax and structure in future versions...

where the state 3 becomes the state 2 at some point in the MEP. In that point, the HOP option of the slapaf module should make the change of states (as I understand, it should change the RLXRoot option):

(Remember that HOP is in RASSI, not SLAPAF.)
Indeed, it should change RlxRoot, but RlxRoot is "fixed". If I remember correctly, the way to do this in 8.0 is using a new keyword: MDRlxRoot. This is specified at the beginning, but in a way that RASSI can modify it:

&GATEWAY
  ...

>> Export MOLCAS_MAXITER = 1000
>> DoWhile

  &SEWARD
    ...

  &RASSCF
    ...
    MDRlxRoot = 3

  &RASSI
    HOP

  &SLAPAF
    MEP-search
    MEPStep=0.02

>> EndDo