differences in energies after jobex and (after single point) ridft

[yasmeen]

Hi,
I’m using Turbomole 6.3 version.
I’m trying to optimize geometry of manganese porphyrins with different ligands using BP-DFT  method (def-TZVP basis, Ri approx.) and I’ve found a little bit strange problem with energy level. When after the geometry is optimized, I make a ridft calculation, the energy level is slightly decreased. It happens in both cases – when a single point calculation is made using only obtained geometry with a new control file, and when the control file and all additional files (including alpha and beta) are taken from geometry optimization.
E.g. when E=-2215, 882357, after ridft E=-2215,885735 in both cases. For another molecule : E=-2216,245884, after ridft E=-2216,249449 (using only converged geometry) or E=-2216,248802.

I would appreciate any suggestions where the differences could come from and how to solve this problem.

[uwe]

Hi,

if you run an energy calculation after a geometry optimization the total energy indeed must not change. Check your ridft output file, it should be only 3 SCF iterations in there until convergence, because you are not just using the optimized coordinates but also the converged orbitals.

You should send your files to the Turbomole support, I guess without looking at your input in detail it is not possible to find out where the problem comes from.

Regards,

Uwe

[yasmeen]

Hi,
the problem is fixed and the reason is the simplest in the world. I was looking at the wrong value (still I don't know how I did it). So everything is ok - sorry for making trouble.
Best regards