During each iteration of my calculation process, when the electron steps converge, there is a sudden change in energy.

[yu_meng]

Dear forum experts，
When I was calculating the Co2C/Co calculation model, during the structural optimization or static calculation process, I found that after each electronic step convergence, the energy obtained by nF would suddenly change (the energy of nF differed by several eV from the energy obtained after the final electronic convergence step as shown in the vasp.log image attached below). What could be the reason for this? You can find my calculation-related files in the attachment. Thank you for all the experts' answers.

[ferenc_karsai]

The last eletronic step ("RMM: 10") and the final total energy ("F=") are different because you have IDIPOL=3 activated. I have tried the calculation without the tag and both lines give the same. IDIPOL>0 adds a term to the total energy after the write out of the ionic step and before the final total energy. This is why you see the differences.

[yu_meng]

Dear ferenc_karsai,
Thank you very much!

[yu_meng]

Dear ferenc_karsai,
I saw the following description about "IDIPOL" in the VASP manual:
For 1 to 3, the dipole moment will be calculated only parallel to the direction of the first, second or third lattice vector.
The corrections for the total energy are calculated as the energy difference between a monopole/dipole and quadrupole
in the current supercell and the same dipole placed in a super cell with the corresponding lattice vector approaching
infinity. This flag should be used for slab calculations.
So, for the surface model I calculated, how should I set the parameters of IDIPOL to be more appropriate?
We sincerely hope to receive your guidance and answers. Thank you very much!

[ferenc_karsai]

The VASP wiki page for IDIPOL writes the following:

This flag should be used for slab calculations, with the surface normal being the direction in which the IDIPOL is set, and optionally specifying the center of mass of the slab with the DIPOL tag.