Molcas Forum

moabe

Member

Registered: 2018-06-03

Posts: 23

Character of Ground State

Dear All,

I am trying to perform relaxed PE scan along a dihedral angle for the trans-to-cis isomerization of azobenzene. I am working at the SA(3)-CASSCF(10,8)/cc-pvdz level. At the FC geometry, the character of S1 and S2 states are n1(π*)1, and n0(π*)2, respectively. At the conical intersection both S0 and S1 are a mixture of n2(π*)0, n1(π*)1, and n0(π*)2 configurations. My problem rises after the CI: the character of GS is n0(π*)2 and S2 has the character of GS, they switched their character. As I realized, the only way to fix this is using a new set of SCF orbital after the CI, which leads to discontinuity of the potential. I greatly appreciate if someone can help me to understand and fix this issue.

Thanks.

Last edited by moabe (2019-10-24 20:33:07)

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: Character of Ground State

When you say that "S2 has the character of GS", I guess you mean that it is mainly a closed shell configuration (all orbitals doubly occupied or empty). I don't see a problem with what you describe, the ground state changes as you change the geometry, and away from the minimum the closed shell configuration is not as favoured (e.g., think of the H2 dissociation). Especially if you have a conical intersection. Then you should expect the "character" of the two states to change.

But have analyzed the CI coefficients with the average orbitals, or the occupations (and shape) of the state-specific natural orbitals? The latter will probably be more helpful to interpret the GS.

moabe

Member

Registered: 2018-06-03

Posts: 23

Re: Character of Ground State

Thank you Ignacio,

Actually the main problem is when the systems reaches to the minimum (cis isomer) the character of GS is still n0(π*)2. I agree that away from the minimum the closed shell configuration is not as favored, but what about the minimum. I need to ensure that having this character for the GS even for the minimum is not a problem and more importantly have no effect on the potential. I will try to analyze the CI coefficients.

There is another thing regarding these calculations that I like to ask. I am trying to see the effects of ortho halogen (fluorine and chlorine atoms) substitutions of azobenzene on the PESs compared to the parent structure. The repulsive interactions between halogen atoms lead to distortion of the structure and we speculate it can affact the shape of potential and particularly the S1/S0 conical intersection. I wonder whether I need to include some orbitals of halogen atoms like lone pairs in the active space to describe these interactions. In the current active space, I cannot see any orbital on those atoms.

Thanks.