The problem of discontinuity in the potential energy curve

[Mhmd Al ahmad]

Dear DIRAC Users, 

I have noted that insertion of  spin-orbit coupling into the Hamiltonian, in Dirac program (at the CCSD(T) level with the Dirac-Coulomb Hamiltonian and the dyall.4zp basis set), led to kinks (discontinuities) in the potential energy curve of one molecule under study! Is this a generic feature that appears quite often, and if so, what is the best way to fix it while taking into account this coupling ? 

My output files and the potential energy curves of Spin-orbit and Spin-Free cases are attached.

Best regards, 

Mohamed

[peterso...@gmail.com]

Dear Mohamed,

 

If you look closely at your converged orbitals/spinors, you’ll see that in the spin-free case the two open-shell orbitals are the Au 6s and a Au 6d.  In the SO case, the calculation converged to 6d3/2 and a 6d5/2 spinors.  You can either do some REORDER MOS on the latter converged spinors or I would recommend using the MJSELEC option to specify the mj of the occupied and open shell spinors.  I would also recommend adding:

 

**ANALYZE

.MULPOP

*MULPOP

 

So that you get a little more detail on the character of the spinors in the Mulliken analysis.

 

Regards,  -Kirk

--
You received this message because you are subscribed to the Google Groups "dirac-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dirac-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/dirac-users/8b714ec0-ca7c-4cd3-b2ed-29cb97e53eban%40googlegroups.com.

[Miroslav Iliaš]

Hello,

These "jumps" of the curves are because for (or from) certain interatomic distances the SCF electronic occupation changes to some excited state.

As Kirk wrote, you have to keep the lowest electronic state orbitals occupation for the whole potential curve. Either via REORDER or MJSELEC. Check the curve, with SO effects, first at the SCF level, and if you keep proper occupations, add CCSD(T) correlation.

In general, when drifting away from the optimized geometry (here r_e), systems are always prone to change SCF electronic state. One has to watch the (ground) SCF state in all geometries.

Best, 

Miro

--