Hi,
Recently I was running VASP to calculate the spontaneous polarization of my alloy structure.
I followed the instructions in this PDF wrote by Dr. Nabin (Nabaraj) Pokhrel, below is my INCAR setting:
KPAR = 1
NPAR = 8
NCORE = 1
PREC = Accurate
EDIFF = 1.0E-7
GGA = PE
ENMAX = 400
ISMEAR = -5
LCALCPOL = .TRUE.
DIPOL = 0.30 0.30 0.30
LWAVE = .FALSE.
LCHARG = .FALSE.
After I run the calculation, one of the intermediate images got the warning which is as follows:
The calculation of the macroscopic polarization by means of the Berry-phase expressions (LCALCPOL=.TRUE.) requires your system to be insulating. This does not seem to be the case. VASP will skip this part of your job, sorry...
Does anyone know how to fix this? Please let me know if I need to provide anymore information, thanks!
Hi Yen-Chun Huang
The warning message reads: [...] This does not seem to be the case [...]. It is not clear from your message whether your system is metallic or insulating. Regardless, you can check the density of states (DOS) or the band structure to confirm if there is a gap. Metallic systems have a finite DOS (i.e., non-zero g(E)) at the Fermi level. If your structure is indeed metallic, you would need to take another approach for calculating polarization, as the Berry-phase method requires an insulating system, according to VASP output. Yet, if the system is supposed to be insulating, you might want to double check your set parameters to understand why your calculation is triggering that VASP warning, as it shouldn't.
Rgds
Hi Gabriel Vinicius
Thank you for your response!
I believed that the system is supposed to be insulating, and I believe that VASP might have mistakenly identified my system as metallic since it has non-zero g(E) at the Fermi level.
My set parameters are shown as above for INCAR.
The version of VASP that I am using is [email protected].
So, what I did is I create 5 intermediate images between the initial and final state (7 images in total), and the warning only appear in the 3rd intermediate image.
I also tried with 8 intermediate images, but then the warning appears in the 4th and 5th intermediate images.
Is there any way can fix this?
Hey Yen-Chun Huang
The fact that the insulating system is identified as metallic in certain intermediate images may suggest that during the transition between the initial and final states, this/these configuration/s (image/s) momentarily appear/s metallic to VASP. This could be due to structural deformation in the particular intermediate image/s leading to band gap closure - weak guess since your INCAR does not have the ISIF flag, and assuming the default is in use, so no deformation should be happening. It is also possible that your k-point sampling is not dense enough, causing VASP to misidentify the electronic structure. Also, try adjusting the EDIFF and ENMAX parameters, as the precision of the calculations might not be high enough to resolve the band gap accurately- I would suggest increasing ENCUT and decreasing EDIFF.
This is all I can think of for now! Try these troubleshooting steps and see how it goes!
Rgds
Hi Gabriel Vinicius
Thank you for your response!
Just want to clarify, the suggestion that you said: "I would suggest increasing ENCUT and decreasing EDIFF."
Are you referring to ENMAX or ENCUT?
Unfortunately, the problem is still there even after I increased the ENMAX and decreased the EDIFF.
I also increase the k-point from 2x2x2 to 4x4x4 and 6x6x6 but the result is the same.
Now I am trying to increase the k-point, increase the ENMAX and decrease the EDIFF at the same time.
Is there any other reason you think might cause the error?
Heey Yen-Chun Huang
I assume you used NEB to create the pathway. Before using NEB, ensure that your initial and final states are well-converged so that the entire pathway, including the intermediate images, is accurately represented. I would suggest double checking the convergence criteria for the SCF calculations regarding the initial and final states. Perhaps your criteria are not strict enough, and these states are not well-relaxed, leading to a possibly unphysical metallic intermediate state. Another possibility to explore would be finding different local minima to use as the initial and final states, so NEB can define a different pathway that might be more feasible.
These are general guesses, and if I knew exactly which alloy you are working with and which functional you are using, I might be able to come up with something more specific.
Cheers!
- Badges
- Science topic
- Similar topics
- Filing