geometry optimization completed

[Tahere Kheshti]

Hello everyone

I am attaching the input and output files. As you see none of the criteria for optimized geometry has been reached. Do I need to increase the number of optimization steps? how many is enough? (my structure has 114 atoms)

Are the number of inner and outer SCF steps appropriate?

Can you please tell me your suggestions?

Thanks

 

[Thomas Kühne]

Dear Tahere, 

your geometry optimization is far from being converged. Instead of limiting the number 

of optimization steps to 30, better simply rely on the convergence criteria. 

SCF convergence, however, is much better when OT switches into the SD fallback mode 

is always bad news. Try to reduce the OT stepsize …

Best, 

Thomas

--
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/CAJwLP85QLZ%2BPDgx7g3kzzPYLoFunGWfUuEfVcaG54zoL%3D2UW7w%40mail.gmail.com.
<In2Cl10C20H12Br4N4F8pbc.inp><In2Cl10C20H12Br4N4F8pbc.out>

==============================

Thomas D. Kühne

Dynamics of Condensed Matter

Chair of Theoretical Chemistry

University of Paderborn

Warburger Str. 100

D-33098 Paderborn

Germany

thomas...@upb.de

+49/(0)5251/60-5726

[fabia...@gmail.com]

Hi

Your initial guess seems to be pretty bad: The structure at step 0 has a pressure of 5'546'592 bar and the first BFGS iteration reduces the energy by more than 50 Hartree, which is enormous. The volume of your cell increases by a factor of 2 over the 30 BFGS iterations. Maybe you could try to improve the initial configuration. This would certainly help the convergence of the SCF and BFGS.

Cheers Fabian

[Tahere Kheshti]

Dear Thomas

Thanks for your helping! I am a new user and I do not know some points.

Do you mean to use the default value for the number of optimization steps?

Yes, unfortunately in the convergence of SCF, DIIS switched to SD, you suggested that I reduce the size. What value is good?

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/7cd7f77f-10fe-48ea-bb14-c46c64ebca6bn%40googlegroups.com.

[Tahere Kheshti]

Dear Fabian

I am interested in studying the structural, electronic, and optical properties of periodic systems. I used to calculate the gas phase using Gaussian software. Now for the first time, I want to calculate in solid-state by cp2k. I have got some basic questions about periodic calculations. Please forgive me if I am asking very general questions, and so many....!

I want to start from the crystal X-ray structure...you can see it in the attached file!

1) Can I replace some of its atoms with arbitrary atoms? for example F atom instead of H atom... If I can, I must do calculations by cell-opt run type??? or geo-opt is also good?! 

 2) As you see in the attached file, in my x-ray unit cell, there are some atoms outside of the box.  Should such atoms be deleted before converting the CIF file to XYZ format? or not?

3) Can this structure be investigated in 2D?

4) If I want to investigate the electronic and optical properties of this structure, can I use the single unit cell or should it be in the form of a supercell? Are the properties of the unit cell and supercell different?

5) There is an important physical concept in the solid-state called k point! Is it necessary to enter its keyword ?! or it is set by default...I don't know?  Do you think I should enter it in the input file? whether unit-cell or supercell!

  

I am very sincerely eager to get an answer? please.

Thanks a lot for helping