RASSI after MS-CASPT2

Posted by Nicolas Ferre' on October 17, 2002 at 17:45:02:

I'm a little bit confused with the output of Rassi when I ran it after a MS-CASPT2, in order to calculate the oscillator strengths.
The second part of the Rassi's output expresses the Hamiltonian and the overlap matrices of input states, i.e. the MS-CASPT2 states.
What I don't undestand is that I get a Hamiltonian with non-zero off-diagonal elements while the overlap matrix is the unity one, that means that states are othogonal therefore do not interact.
In the third part of the Rassi's output, the previous Hamiltonian is diagonalized, then the corresponding eigenvalues are the SA-CASSCF energies.
This is because (if I understood well the theory) the MS-CASPT2 states are first order wavefunctions.
Finally I wan't to be sure: to calculate oscillator strengths of MS-CASPT2 states (according to the 2/3.(TDM)^2.(Delta E) relation), I should consider the MS-CASPT2 one-electron properties (multipole) of this Rassi's output with the MS-CASPT2 energies (in the caspt2's output).
Is it really that ?

Many thanks for any explanation ! I think I certainly miss something in the MS-CASPT2 theory !

Follow Ups:

Post a Followup

Name:
E-Mail:

Subject:

Comments:
: I'm a little bit confused with the output of Rassi when I ran it after a MS-CASPT2, in order to calculate the oscillator strengths. : The second part of the Rassi's output expresses the Hamiltonian and the overlap matrices of input states, i.e. the MS-CASPT2 states. : What I don't undestand is that I get a Hamiltonian with non-zero off-diagonal elements while the overlap matrix is the unity one, that means that states are othogonal therefore do not interact. : In the third part of the Rassi's output, the previous Hamiltonian is diagonalized, then the corresponding eigenvalues are the SA-CASSCF energies. : This is because (if I understood well the theory) the MS-CASPT2 states are first order wavefunctions. : Finally I wan't to be sure: to calculate oscillator strengths of MS-CASPT2 states (according to the 2/3.(TDM)^2.(Delta E) relation), I should consider the MS-CASPT2 one-electron properties (multipole) of this Rassi's output with the MS-CASPT2 energies (in the caspt2's output). : Is it really that ? : Many thanks for any explanation ! I think I certainly miss something in the MS-CASPT2 theory !