DFT Structure Optimization for Open-Shell System possible?

[paul.j...@gmail.com]

Dear DIRAC Users and Developers,

Do you know if it is simply not possible to optimize an open-shell system using DFT with DIRAC-25? Or am I simply doing something wrong?

I have an uranium complex containing three unpaired electrons (multiplicity = 4 using other quantum-chemical software). After the first SCF iteration, get the error message " error in molecular gradient: not implemented for open-shell DFT".

Using the .NUMGRAD keyword in the *OPTIMIZE section does not change this, unfortunately.

Here is my input file:

And here are parts of my output:

Thank you very much in advance!

Best wishes,

-- Paul

[Ayaki Sunaga]

Dear Paul,

> " error in molecular gradient: not implemented for open-shell DFT".  

In the DIRAC code, the optimization is not implemented for open-shell DFT.
If your complex is small (diatomic or triatomic), you can get an optimized geometry by running many one-point calculations and doing the fitting.

For more efficient optimization, I recommend using other codes for open-shell systems.

Best regards,
Ayaki

2025年5月27日(火) 16:55 paul.j...@gmail.com <paul.j...@gmail.com>:

--
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/def63bef-45cf-467d-b7ed-1e4ce6d331ddn%40googlegroups.com.

[Miroslav Iliaš]

Dear Paul,

I checked similar cases (DC-SF, open-shell, PBE, numgradient) with up-to-date DIRAC. 

Here is H2O(+), https://github.com/miroi/open-collection/tree/master/theoretical_chemistry/software/dirac/runs/H2O/H2O_geom_optim/H2O_1p   

This is fine, numerical gradient is kept.

Here is CN, https://github.com/miroi/open-collection/tree/master/theoretical_chemistry/software/dirac/runs/CN/pbe_4c 

Here it is getting the same error as you "error in molecular gradient: not implemented for open-shell DFT"

Miro

--

[paul.j...@gmail.com]

Thank you for the prompt responses!

@Ayaki: Unfortunately, the complex is a bit larger so I'd have to rely on an optimization algorithm.

@Miro: Interesting oberservation! Thank you for sharing! :) Is anyone able to look "under the hood" what might cause this?

Many thanks and best wishes,

   Paul

[Hans Jørgen Aagaard Jensen]

Dear Paul,

 

I have investigated a little. Your calculation should have worked with .NUMGRAD, however, by mistake a quit was inserted in the code for DFT and calculation of gradient in January 2025, which is active for both numerical and analytical gradients; it should only have been active for analytical gradient evaluation.

 

If you remove these lines in …/src/main/dirac.F after ca. line 170:

 

      if (dirac_cfg_dft_calculation .and. OPTIMI) then

        if (nopen > 0) then

          call quit('error in molecular gradient:' //

     &         ' not implemented for open-shell DFT')

        end if

        if (nbsym > 1) then

          call quit('error in molecular gradient:' //

     &    ' not implemented for DFT with symmetry')

        end if

      end if

 

and recompile, then your geometry optimization should work (no guarantee, haven’t tested – but it ought to). Let us know if it worked.

 

All the best, Hans Jørgen.

 

 

 

Fra: <dirac...@googlegroups.com> på vegne af "paul.j...@gmail.com" <paul.j...@gmail.com>
Svar til: "dirac...@googlegroups.com" <dirac...@googlegroups.com>
Dato: onsdag den 28. maj 2025 kl. 08.51
Til: dirac-users <dirac...@googlegroups.com>
Emne: Re: [dirac-users] DFT Structure Optimization for Open-Shell System possible?

 

Du får ikke ofte mails fra paul.j...@gmail.com. Få mere at vide om, hvorfor dette er vigtigt

Thank you for the prompt responses!

 

@Ayaki: Unfortunately, the complex is a bit larger so I'd have to rely on an optimization algorithm.

 

@Miro: Interesting oberservation! Thank you for sharing! :) Is anyone able to look "under the hood" what might cause this?

 

 

Many thanks and best wishes,

 

   Paul

 

 

miro....@gmail.com schrieb am Dienstag, 27. Mai 2025 um 20:35:06 UTC+2:

Dear Paul,

 

I checked similar cases (DC-SF, open-shell, PBE, numgradient) with up-to-date DIRAC. 

 

Here is H2O(+), https://github.com/miroi/open-collection/tree/master/theoretical_chemistry/software/dirac/runs/H2O/H2O_geom_optim/H2O_1p   

This is fine, numerical gradient is kept.

 

Here is CN, https://github.com/miroi/open-collection/tree/master/theoretical_chemistry/software/dirac/runs/CN/pbe_4c 

Here it is getting the same error as you "error in molecular gradient: not implemented for open-shell DFT"

 

Miro

 

 

On Tue, 27 May 2025 at 17:55, paul.j...@gmail.com <paul.j...@gmail.com> wrote:

Dear DIRAC Users and Developers,

 

Do you know if it is simply not possible to optimize an open-shell system using DFT with DIRAC-25? Or am I simply doing something wrong?

 

I have an uranium complex containing three unpaired electrons (multiplicity = 4 using other quantum-chemical software). After the first SCF iteration, get the error message " error in molecular gradient: not implemented for open-shell DFT".

 

Using the .NUMGRAD keyword in the *OPTIMIZE section does not change this, unfortunately.

 

Here is my input file:

 

And here are parts of my output:

 

 

Thank you very much in advance!

 

Best wishes,

-- Paul

--

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/def63bef-45cf-467d-b7ed-1e4ce6d331ddn%40googlegroups.com.

--
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/9458d9e5-33ee-4248-ab0a-4774edc3606cn%40googlegroups.com.

[paul.j...@gmail.com]

Dear Hans Jørgen,

That indeed solved the issue! Thank you very, very much!!

Best wishes,

   Paul

[Trond SAUE]

Hi Paul,

just a little remark: If you run an open-shell system with DFT in DIRAC, the code will use fractional occupation.

All the best,

   Trond

--

Trond Saue
Laboratoire de Chimie et Physique Quantiques
UMR 5626 CNRS --- Université Toulouse III-Paul Sabatier
118 route de Narbonne, F-31062 Toulouse, France

Phone : +33/561556361 Fax: +33/561556065
Mail : trond...@irsamc.ups-tlse.fr
Web : https://dirac.ups-tlse.fr/saue
DIRAC : http://www.diracprogram.org/
ESQC : http://www.esqc.org/
ERC Advanced Grant: HAMP-vQED

Book: Principles and Practices of Molecular Properties: Theory, Modeling, and Simulations