Problems optimising a heme (Fe2+) group

[Sara Llap]

Dear all, 

I am having problems on a geometry optimisation of a heme group with Fe2+ using PBE in CP2K. My calculations run and the total energy is decreasing. However, the job has been running for almost a month and it doesn't seem to converge. 

I am not an expert on CP2K nor on transition metals. I would really appreciate it if you could help me sort this out. I have  attached the restart file and the output for clarity. 

Thanks a lot, 

Sara

[Ivan Gladich]

Dear Sara

Not an expert of Heme but I think one problem could be the multiplicity (and charge?) of your system

Fe2+ in an aqueous solution should be high spin (Multiplicity 5)

https://doi.org/10.1021/acs.jpclett.2c00380

In your input

    BASIS_SET_FILE_NAME BASIS_MOLOPT
     POTENTIAL_FILE_NAME GTH_POTENTIALS
     MULTIPLICITY  1
     CHARGE  -2

Check also for HEME

https://doi.org/10.1021/acs.inorgchem.6b01360

Hope it helps

Best

Ivan

[Victor Volkov]

Good morning Sara.

Several months ago there was a paper devoted to

electronic and structural properties of different hemes

to explain the color of blood

“Mapping blood biochemistry by Raman Spectroscopy at the cellular level.” Chem. Sci. 2022, 13, 133.

This may be relevant.

Victor

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/0b3eff04-56c9-42de-90c5-f6ca8cce446an%40googlegroups.com.

[Marcella Iannuzzi]

Dear Sara

Before starting a geometry optimisation it is  good practice to get a converged single point calculation. 

If I understand correctly from the output you posted, this was not the case. 

I played a bit with the settings for your system and I got a converged single point calculation at the PBE level of theory 

for the molecule charged --2 (see the attached file).

Kind regards

Marcella

[Sara Llap]

Dear Ivan, Victor and Marcella, 

Thank you so much for your help. 

I have been reading the bibliography you suggested and using your inputs and I almost optimised the heme group. I got to a point where the optimisation does not seem to converge but it is really close. 

--------  Informations at step =   114 ------------

  Optimization Method        =                   SD

  Total Energy               =      -517.5364879764

  Real energy change         =         0.0000000050

  Decrease in energy         =                   NO

  Used time                  =              460.425

  Convergence check :

  Max. step size             =         0.0000000000

  Conv. limit for step size  =         0.0010000000

  Convergence in step size   =                  YES

  RMS step size              =         0.0000000000

  Conv. limit for RMS step   =         0.0010000000

  Convergence in RMS step    =                  YES

  Max. gradient              =         0.0012644252

  Conv. limit for gradients  =         0.0010000000

  Conv. for gradients        =                   NO

  RMS gradient               =         0.0002927175

  Conv. limit for RMS grad.  =         0.0010000000

  Conv. in RMS gradients     =                  YES

 ---------------------------------------------------

I have been playing a bit with the parameters but I haven't managed to make it work. 

Thanks again, 

Kind regards, 

Sara

--
You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/C7w_GP_CIxw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/bf21e618-f8d1-43d4-9e58-463e1b671947n%40googlegroups.com.

[Marcella Iannuzzi]

Dear Sara 

Starting around step 50 the coordinates do not change anymore

Max. step size             =         0.0000000000

The gradient  above tolerance is only one. It is the x component of the force on atom 47, a hydrogen. 

It remains practically constant for many steps.

It seems that the optimisation algorithm is not in the conditions of performing any CG step, and the SD gets stuck. 

So since step 50 nothing changes in the structure.

I would try and change the optimiser to BFGS.

Kind regards

Marcella

[Sara Llap]

Dear Marcella, 

Thank you for your help. It took me a lot of attempts but I managed to converge the calculation. 

Thank you again!

Sara

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/0b749181-a26b-4863-b3d1-be7e321f02dan%40googlegroups.com.

[Sara Llap]

Dear Marcella, 

I used your advice on several other calculations and it worked but I struggle with one that I am unable to converge the SCF. Could you help me or give some advice on how to solve this issue?

The initial structure comes from a PM6 calculation. 

Thank you very much in advance, 

Sara