mos file: AO basis function order and norm?

[Jonas]

Dear all,

I'm currently playing around with initial guess orbitals from an EHT calculation with TURBOMOLE v. 5.10. In the end I want to read these -- more precise: the density matrix from these -- into my own SCF program, and look at the SCF convergence behavior depending on the initial guess data. Thus, I need to know:

-- in which order do the MO coefficients come in the mos file?
From playing around, I feel that the MO coefficient order corresponds to the AO basis function order as given in the basis file ($basis), but for angular momenta > 0 things are tricky ...
-- what is the normalization of the AO basis functions?

It would be really great if someone could help me with that, provided that this information is not a developer's secret  ...

Thanks in advance,
Regards, Jonas

[Jonas]

Dear all,

apparently it's much simpler to solve than I thought. I found a way to cast the EHT MOs into the well-known MOLDEN format using the tm2molden script. Everything is fine.

Thank you,
Regards, Jonas

[uwe]

Hi,

the format of the MOs is described in detail if you use the 'infsao' option in the molecular orbital menu of define.

Regards,

Uwe

[JakubV]

Dear all,

apparently it's much simpler to solve than I thought. I found a way to cast the EHT MOs into the well-known MOLDEN format using the tm2molden script. Everything is fine.

Thank you,
Regards, Jonas

Hello,

excuse my ignorance, please - the script tm2molden works in this way:  "tm2molden" and then I will be asked about the input file and there will be written all info from files mos, infsao_file, basis and coord from which will some version of molden (not my current one, please, do you know from which version higher I need?) be able to plot molecular orbitals?

I don't want to use the computing of MOs values on grid by turbomole since I suppose it will be time-consuming (is it possible to run just this on grid computation of already computed MOs (without re-running scf/dft, please?))

Thank you very much,
Jakub

[uwe]

Jakub,

tm2molden writes the molecular orbitals coefficients and the atomic basis set to the molden input file. This is just a conversion of the format. It is molden itself which calculates the values of the orbitals in space to visualize them.

But you can also generate 3D data of the values of the orbitals by adding

$pointval mo <list-of-orbitals>

to the control file of a finished job. Then call

dscf -proper

or

ridft -proper

to generate the orbitals without doing an SCF calculation first.

This will generate plt files which can either be visualized in TmoleX, in gOpenMol or VMD.

Regards,

Uwe