1e Integrals

[barbatti]

Dear Turbomole developers,

I would like to inquire about the intsdebug keyword to print 1e integrals in the dscf program.

Although it is not documented since version 5.9, it still worked up to version 6.3.
It was not working in version 6.4 and it came back in version 6.5, but printing the integrals in SAO basis, not in CAO anymore.

I am worried because these integrals are used during the computation of nonadiabatic couplings by Newton-X when running dynamics with Turbomole.

What are the plans for the next versions of Turbomole regarding this keyword?
Is it going to be discontinued?
Are there other ways to access the 1e integrals in the Turbomole output?

To simplify your answer, only C1 symmetry cases are relevant for me.

By the way, if I understood correctly, in C1, SAO and CAO 1e integrals are equal. Am I right?

Thank you in advance,

Mario

 

[uwe]

Dear Mario,

the non-documented keyword for printing one-electron integrals has been changed long time ago, and the format was SAO also in (many) previous versions even if dscf printed CAO in the header, because the integrals are being computed in this basis and not in CAO any more.

SAO in C1 symmetry is not CAO but AO (5d, 7f, etc. functions and not 6d,... as in CAO).

If you want to make sure that this feature will be available in future versions too, I'd recommend to contact the Turbomole GmbH directly or try it via the Turbomole support (turbomole@cosmologic.de). A test case has to be added to the developers version of the test suite, a properly defined interface to your program should be defined and checked, and so on...

Regards,

Uwe