[CP2K-user] [CP2K:20011] Inquiry on ROKS Calculation Schemes within CP2K

[Jürg Hutter]

Hi
I haven't worked on this for 20 years. The implementation was also not done by me.
The same is true for the MOM method. Maybe, somebody with more recent
experience with these methods can comment and give advice?
regards
JH

________________________________________
From: cp2k at googlegroups.com <cp2k at googlegroups.com> on behalf of Lukas Razinkovas <lukasrazinkovas at gmail.com>
Sent: Wednesday, March 6, 2024 10:03 AM
To: cp2k at googlegroups.com
Subject: Re: [CP2K:20004] Inquiry on ROKS Calculation Schemes within CP2K

Dear Prof. Jurg,


Thank you for your response. I am currently exploring ROKS schemes to model excited state potential energy surfaces and have several questions to improve my understanding:

  1.  Does the LOW_SPIN_ROKS approach derive from your collaboration with Imgrad in the study "Molecular dynamics in low-spin excited states" [J. Chem. Phys. 108, 10 (1998)], and its subsequent extension to accommodate any number of unpaired electrons [Chem. Phys. 373, 283 (2010)]?
  2.  In my investigation, I calculated the quartet to doublet excitation energy using the von Barth/spin projection method combined with UKS, yielding results remarkably close to experimental data. However, LOW_SPIN_ROKS calculations yielded energies significantly higher than expected. Considering I am examining localized defect states (d orbitals of Chromium) in a crystal environment, localized orbitals might shift above the conduction band minimum. This scenario could misrepresent the configuration as (a b CBM) rather than the intended (a b c). Is there a way to investigate single-particle states within LOW_SPIN_ROKS calculations to address this concern?
  3.  Upon reviewing Filatov's publication (Chem. Phys. Letter 288, 689 (1998)), I understand it lays the foundation for a formal multi-determinant state approach within ROKS. This approach appears to be activated through the "ROKS_SCHEME GENERAL" setting. However, looking at the code, specifically "SUBROUTINE combine_ks_matrices_2," suggests its current implementation is limited to single-determinant solutions due to the absence of iterations over energy scaling parameters. Could you confirm if my observation is accurate?
  4.  Lastly, I'd like to know whether the MOM method can be applied to HIGH-SPIN or GENERAL schemes. I aim to model spin-contaminated configurations such as |a b \bar{c}|, which seem unattainable with the MULTIPLICITY 2 keyword since it strictly enforces a |a \bar{a} b| configuration.


I appreciate your time and look forward to your replies.


Best regards,

Lukas

On Mon, Mar 4, 2024 at 3:07 PM Jürg Hutter <hutter at chem.uzh.ch<mailto:hutter at chem.uzh.ch>> wrote:
Hi

these methods / implementations are based on the work by Irmgard Frank.
You can find earlier work e.g. by Claude Daul and also M. Filatov.
There has been recently some new work by other groups that is closely related.

1) this is for single determinant only (all excess electrons in one spin channel)
2) I think it is correct, but I would have to look more closely
3) ROKS was implemented before MGGAs. It seems the ROKS branch was not updated for these type of functionals.
4) See the work of Daul and Filatov

regards
JH

________________________________________
From: cp2k at googlegroups.com<mailto:cp2k at googlegroups.com> <cp2k at googlegroups.com<mailto:cp2k at googlegroups.com>> on behalf of Lukas Razinkovas <lukasrazinkovas at gmail.com<mailto:lukasrazinkovas at gmail.com>>
Sent: Sunday, March 3, 2024 8:19 AM
To: cp2k
Subject: [CP2K:19983] Inquiry on ROKS Calculation Schemes within CP2K

Dear CP2K Community,


As a newcomer to CP2K (my experience is with VASP, GPAW, and Quantum Espresso), I have been exploring CP2K's capabilities, particularly for modeling low-spin bi- and tri-radical states of crystal defects through ROKS (Restricted Open-Shell) calculations for potential energy surface analysis. I am pleasantly surprised by this code's performance and capabilities.


However, I am slightly confused with ROKS calculations within CP2K, especially regarding the implementation and application of its varying schemes, namely:


  *   High-Spin ROKS (default ROKS_SCHEME value)
  *   GENERAL (an alternative ROKS_SCHEME value)
  *   Low-Spin ROKS


I would greatly appreciate any directed literature or resources that elaborate on these schemes' implementation and practical use.


Additionally, I have a few specific inquiries that I hope could be clarified:


  1.  Is the High-Spin ROKS scheme tailored for single-determinant maximal spin states, where all unpaired electrons align in one spin direction? Can it also model mixed (spin-contaminated) single-determinant states for the von Barth/approximate spin projection method?
  2.  For modeling the doublet state in a triradical scenario, I am considering the following configuration:
&LOW_SPIN_ROKS
  # DOBLET: E(D) = E(Q) - 3/2*E(Q) + 3/2*E(M)
  ENERGY_SCALING -1.5 1.5
  SPIN_CONFIGURATION 1 1 1 # Psi(Q) |a a a>
  SPIN_CONFIGURATION 1 1 2 # Psi(M) |a a b>
&END LOW_SPIN_ROKS
Is this the correct approach?
  3.  I encountered difficulties running LOW_SPIN_ROKS with the R2SCAN functional (however, it worked with PBE). Are there compatibility issues with meta-GGA functionals?
  4.  Can the energy outcomes from maximal-spin and low-spin calculations be compared to deduce exchange splitting (excitation) energy?


I look forward to any recommendations or advice you could provide.



Best wishes,

Lukas

--
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+unsubscribe at googlegroups.com<mailto:cp2k%2Bunsubscribe at googlegroups.com><mailto:cp2k+unsubscribe at googlegroups.com<mailto:cp2k%2Bunsubscribe at googlegroups.com>>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/025e1887-09f0-4460-a1ae-24662777baf9n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/025e1887-09f0-4460-a1ae-24662777baf9n%40googlegroups.com?utm_medium=email&utm_source=footer>.

--
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/xPbCQHw3caA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+unsubscribe at googlegroups.com<mailto:cp2k%2Bunsubscribe at googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZR0P278MB0759EFE14E0AC16C9A0570A39F232%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.

--
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+unsubscribe at googlegroups.com<mailto:cp2k+unsubscribe at googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAFKUn1oRRs%3DNy%2BDh0%3DwEhLtF3kRtu1sLq%3DcZOkWBfzpxkK55wA%40mail.gmail.com<https://groups.google.com/d/msgid/cp2k/CAFKUn1oRRs%3DNy%2BDh0%3DwEhLtF3kRtu1sLq%3DcZOkWBfzpxkK55wA%40mail.gmail.com?utm_medium=email&utm_source=footer>.

-- 
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+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/ZR0P278MB07598EA611EC01B7440DBCE69F202%40ZR0P278MB0759.CHEP278.PROD.OUTLOOK.COM.