Hi Nhut Nguyen,

The INCAR is fine, I was able to run 1 iteration using you input files. I can help you if you can send me the error VASP gives you.

Note that you k-point mesh is not set correctly, your system is a surface and it is significantly longer in a-direction compared to b-direction, so you should correct the k-point mesh before continuing your calculations. Also, you have specified to do spin-polarized calculations (ISPIN = 2 in INCAR), but you did not specify the magnetic moments on the atoms.

/Vancho

Vancho Kocevski

Thank you for your help. Actually i always receive a complete work notification from VASP and i attach my results below. My question is that how i can optimize my modelling now (try to reduce the dE from the result to minimum). You said the k-point is not correct and which way i can improve it? ( i just know increase if current valued does not work). Then What " but you did not specify the magnetic moments on the atoms" men ?. Also i always get this error even the system said my work is " complete" but i do not know understand the last sentence meaning then whether i can reply on my current result. I mean the line

"1 F= -.70118388E+02 E0= -.69688339E+02 d E =-.701184E+02 mag= 11.4108'. my apology if my problem mind you so much!!!

You are looking at the d E from the first step of optimization. That step went through 60 SCF cycles and did not converge. That is why d E is so large. However, let it run. The geometry may get adjusted a bit in Step 2, but the SCF cycles may converge to a more acceptable level. As the calculation progresses, you should see d E get smaller. My method for looking at the results is to check the OSZICAR file. If you are in a Linux environment, try this at the command prompt:

grep ' d E' OSZICAR

Nhut Nguyen

B. R. Ramachandran gave you a great answer what the last line means and what you should expect from the VASP calculations. However, your calculations seems to crash with the error: "EDDDAV: Call to ZHEGV failed. Returncode = 13 2 16". To overcome this error you can try setting up ALGO = Normal, and decrease POTIM to < 0.1 (see more details here https://abinitio21stp.wordpress.com/2011/06/07/warning-in-eddrmm-call-to-zhegv-failed-returncode-6-3-9/). When specifying ISPIN = 2 it is good to explicitly set the magnetic moments on the atoms using the MAGMOM tag (https://www.vasp.at/wiki/index.php/MAGMOM). About the k-point mesh, I don't want to go in details what it means, the rule of thumb is that longer lattice parameters require smaller k-points, and if in some direction there are no bonds (in your case c-direction, perpendicular to the surface), the k-point is 1; you can check these discussions (or look into solid state theory books):

https://www.researchgate.net/post/What_is_the_kpoints_in_electronic_structure_calculations

https://www.quora.com/What-is-the-meaning-of-the-k-points-in-solid-state-physics-1

Keep in mind that you should always test the convergence of the k-point mesh before using it for all of your calculations. Also, note that VASP has automatic k-point generator (https://www.vasp.at/wiki/index.php/KPOINTS), which is great if you want to compare the energy systems with different sizes using equivalent k-point meshes. /Vancho

Just for the minimum energy calculation or relaxation please edit the INCAR file and change the following tags.

ISIF = 2 #for ionic relaxation

ISPIN = 2

NSW = 100

ALGO = Normal # it helps with electronic convergence

IVDW = 12

LWAVE = T #you would like to further run calculation, saves time

#avoid using AMIX BMIX. use it only if there is any electronic convergence issues.

after convergence, copy CONTCAR to POSCAR and re run the calculation with more tighter EDIFF and EDIFFG. let me know if you face any problem.

To calculate reaction mechanism you must use NEB method. for calculating reaction mechanism you can read the following paper.

https://doi.org/10.1016/j.mtcomm.2020.101602

Hi Sadiq Shahriyar Nishat,

Thank you for your rely. Would you mid if i ask you some question about my problem. I follow your guideline and try to run a convergence test on my Bi2O3 with nitrogen atom. But it does not work when i command grep reach OUTCAR it said "aborting loop because EDIFF is reached' but nothing more. i attached all my input file and could you show me what i need to improve. i have optimized Bi2O3 atom and my idea is to transform it to slag mode (surface area) and attach a nitrogen atom on the top to calculate the absorbed energy.