Hello!

I have one important question. Usually, the NACMEs have 1/bohr units. What is about in Turbomole?

I can see in egrad.out:

cartesian gradient of the energy (hartree/bohr)

dE/dx.....

dE/dy...

dE/dz...

----------

After I can see

cartesian nonadiabatic coupling matrix elements

dE/dx.....

dE/dy...

dE/dz...

So I assume that NACMEs have the hartree/bohr, too. Therefore I can (dE/dx)/E and have d/dx, for example.

However, there is a very interested situation.

the values of NACME (1/bohr) in GAMESS-US at CASSCF level of theory (only the static correlation) have the values which are similar (one order of magnitude) to NACME of Turbomole (with dynamic correlation). I was confused and therefore ask this question.

Thank you very much for the answer!