Off-diagonal terms of the self-energy matrix

[Haowei Xu]

Dear admin,

I tried to use DMFTwDFT code recently and it is a great code!

I find that in the output only the diagonal terms of the self-energy matrices are printed. Is it possible to get the off-diagonal terms?

Also, I could not find any explanations on the meanings of the "CIX paramters" in the example INPUT.py. Could you please explain a bit or point me to some documentations? Thank you very much!

Best wishes,

Haowei

[Singh, Vijay Ramdin]

Dear User,

Thank you for using DMFTwDFT. The following are the answers to your question:

       1. One can also compute the off-diagonal terms but ctqmc will suffer from the minus-sign problem, so we tested for diagonal cases.

       2. CIX parameters: 

para - is this paramagnetic or magnetic run - we support paramagnetic (1)

l - anular momentum (l = 0, l =1, l=2 supported)

n = a list of occupancies used in oca output 

OCA_G =  We do not support OCA. This is an impurity solvers based on hybridization expansion. Therefore, it has been kept ".FALSE.".

CoulombF =  Ising, Full, Fulls, and FullK options - these are Slater parametrization - for other options please see atom_d.py 

Ewindow - energy window for the states kept (used in ctqmc only)

Please see the atom_d.py in the source code. Also, if you would like to get more information about the impurity solver. I suggest to visit http://hauleweb.rutgers.edu/tutorials/Tutorial0.html.

I hope this helps. All the best.

Have a nice day.

Sincerely,

Vijay

---

From: dmft...@googlegroups.com <dmft...@googlegroups.com> on behalf of Haowei Xu <xuhaow...@gmail.com>
Sent: Wednesday, September 29, 2021 10:45 AM
To: DMFTwDFT <dmft...@googlegroups.com>
Subject: Off-diagonal terms of the self-energy matrix

 

--
You received this message because you are subscribed to the Google Groups "DMFTwDFT" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dmftwdft+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dmftwdft/2000e2f7-fc6a-45ac-924c-be23eacd8513n%40googlegroups.com.

[Haowei Xu]

Dear Admin,

Thanks for the prompt responses!

So is it possible to turn on off-diagonal elements calculation in DMFTwDFT? If we ignore the off-diagonal terms, would it make a big difference in the final results? 

Best wishes,

Haowei

[Hyowon Park]

Dear Haowei,

The off-diagonal term would be necessary if you have a system with the finite hybridization between local correlated orbitals.

Usually, the off-diagonal hybridization is zero if you choose the proper local axis for orbitals but this could happen if you have a lower symmetry of the structure.

If you want to include the off-diagonal hybridization term, you should change the CTQMC cix file accordingly.

I recommend you follow the manual of the CTQMC code.

http://hauleweb.rutgers.edu/tutorials/Tutorial0.html.

Best Regards

To view this discussion on the web visit https://groups.google.com/d/msgid/dmftwdft/5299959e-a69d-4985-bdd9-181e94a7368bn%40googlegroups.com.

--

Hyowon Park
Assistant Professor,

Department of Physics,                 Materials Science Division,

University of Illinois at Chicago.     Argonne National Lab.

Email: hyo...@uic.edu   hp...@anl.gov

Phone: 1-312-996-8913

Fax: 312-996-9016

[Haowei Xu]

Dear Admin,

Thank you so much for all these explanations!

I briefly read the CTQMC manual just now. My understanding is that if the system has high symmetry, and we turn on "L_rot=1" in DMFTwDFT, then the off-diagonal terms would be minimized and can often be ignored. Is this understanding correct?

Best wishes,

Haowei

[Singh, Vijay Ramdin]

Yes, Haowei. You're right. We have mentioned these details in our DMFTwDFT paper. Please have a look: https://www.sciencedirect.com/science/article/pii/S001046552030388X

All the best.

Sincerely,

Vijay

---

From: dmft...@googlegroups.com <dmft...@googlegroups.com> on behalf of Haowei Xu <xuhaow...@gmail.com>

Sent: Wednesday, September 29, 2021 2:43 PM
To: DMFTwDFT <dmft...@googlegroups.com>
Subject: Re: Off-diagonal terms of the self-energy matrix

 

To view this discussion on the web visit https://groups.google.com/d/msgid/dmftwdft/fdf3d3be-6dae-4ce2-8cc4-036c4dca47c7n%40googlegroups.com.

[Haowei Xu]

Now I understand, thank you so much!

On Sep 30, 2021, at 12:02, Singh, Vijay Ramdin <vsin...@uic.edu> wrote: