[molpro-user] Problems in State-average CASSCF opt with different basis sets

[貓...]

Dear Molpro users,

 

I’m working on a singlet diradical molecule with excited-stategeometry optimization

.

The molecule I study is similar to 1,3-diphenylcyclopentane with onephenyl ring on each side.

The conformation of cyclopentane is planar, which has diradical oncarbon 1 and 3 (singlet diradical). Therefore, the conformation of wholestructure can be planar and symmetric with Cs point group (preliminarilyoptimized in first singlet excited state by CIS).

*singlet diradical

*one phenyl ring on each side

*symmetric conformation

 

The selected active space is 10 electrons and 10 orbitals. 5 virtualorbitals includes 1 orbital for singlet diradical, and 4 orbitals forpi-orbital (with one node) of phenyl ring.

 

Although the input structure has Cs symmetry, the setting in inputfile is “nosym”. Thus, the symmetry in calculation is C1.

 
I use CASSCF to do the geometry optimization (OPT) at the firstsinglet excited state (S1) and start with 3-21G basis set. The below is my setting and the results (description of optimizedstructure)

(1) 3-21G, opt S1: the optimized structure is asymmetric – one ofthe rings maintains phenyl ring and another is like the chemical structure ofcyclohexa-1,4-diene.

----input----

print,basis,orbitals  

memory,512,m

file,2, 321G-optS1.wfu

symmetry,nosym

geomtyp=xyz

angstrom

geometry={……}

 basis=3-21G

hf

 {casscf

closed,73

occ,83

wf,156,1,0

state,2                    

WEIGHT,0.0,1.0             

CPMCSCF,GRAD,2.1           

}

 optg

 ----input end----

 (2) 3-21G, opt S1 with state-average, an average of 0.0/0.5/0.5between the S0/S1/S2 states (The energies of S1 and S2 states are preliminarilychecked by CIS calculation. The S1 and S2 states are almost degenerate and highwith respect to the ground state.).

The optimized structure is symmetric and has two phenyl rings witharomaticity. Because the input structure is very symmetric, we think that theoptimized structure should also be symmetric. This weight factors and optimizedstructure are both we expected.

----input----

basis=3-21G

hf

 {casscf

closed,73

occ,83

wf,156,1,0

state,3 

WEIGHT,0.0,0.5,0.5 

CPMCSCF,GRAD,2.1    

}

 optg

 ----input end----

 

Then we change the basis set to 6-31G(d).

(3) All the settings are the same as using 3-21G basis set, opt S1with state-average. 

However, the optimized structure is asymmetric, same as (1) 3-21G,opt S1.

  The problem is:

For the same settings but only different from basis set, whychanging the basis set let the optimized structures quite different? One issymmetric structure (by 3-21G) but another is asymmetric (by 6-31G(d)).

 

We struggle with the calculation and cannot understand what’s goingon.

If you have any comments, ideas and suggest literatures, we’ll bevery appreciative.

Extra opinions are always welcome.

 

Thankyou!

Sincerely, Cherry, Chun-Jui Tan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20160907/4f586529/attachment-0001.html>