state-average MCSCF calculation

[Yuan Li]

Dear Molpro experts,

I have one question about the state-average MCSCF calculation. In my case, there are two different ways:

1). perform state-average optimization over more than one spin multiplicity, for example:

{ MCSCF
  OCC, 50; CLOSED,38; frozen,38, 2144.2
  START,2144.2
  WF,  95,1,  5  !!
  state,      196
  WF,  95,1,  3  !!
  state,     1449
  MAXITER,40
  pspace, 9999
  PRINT, REF,REF1,CIVECTOR
  natorb,,ci;
  ORBITAL,2260.2
}

2). preform state-average optimization only for the same spin multiplicity, for example:

{ MCSCF
  OCC, 50; CLOSED,38; frozen,38, 2144.2
  START,2144.2
  WF,  95,1,  5  !!
  state,      196
  MAXITER,40
  pspace, 9999
  PRINT, REF,REF1,CIVECTOR
  natorb,,ci;
  ORBITAL,2260.2
}

{ MCSCF
  OCC, 50; CLOSED,38; frozen,38, 2144.2
  START,2144.2
  WF,  95,1,  3  !!
  state,     1449
  MAXITER,40
  pspace, 9999
  PRINT, REF,REF1,CIVECTOR
  natorb,,ci;
  ORBITAL,2261.2
}

After the MCSCF calculations, if we carried out the initial MRCI then spin-orbit coupling calculations with these different spin multiplicities, I want to understand what is the essential differences in terms of methodology between such two different MCSCF optimizations. Any comments would be highly appreciated.

Best regards,

Yuan Li