Hello Maximo Ramirez,

By your INCAR file, I see you are doing volume relaxations to get the cell at zero pressure (ISIF = 3). Are you following the suggestions at section 7.6 of VASP manual?

When you perform this kind of calculations you have problems related to Pulay stress. This may be causing this difference in your results. Follow the suggestions of the VASP manual and see if you get better results.

In my opinion, the best way to get optimized zero pressure structures is to perform calculations in several volumes and then fit an equation of state (EoS). In your case, this is straightforward, since your cell is cubic.  The EoS is what gives you the right pressure for your system. The difference between the EoS pressure and the pressure given by VASP (that is calculated with the stress tensor) is small, but can lead to the kind of differences you are getting in your calculations. If you have already followed the steps in the manual, try to calculate the lattice parameter using the EoS method.

Hi Maximo Ramirez,

As for as I know you will not get same lattice parameter as of experimental data since there will be unavoidable stress/strain present always in the experimental system & here in your DFT calculation you are doing full relaxation (ISIF=3), where you will not be getting any forces and stress in your OUTCAR file. 

Hope this helps !

best

Rohit

Hi Maximo Ramirez,

I think you may select "spin-polarization calculation"  which can be operated by add "ISPIN = 2" in the INCAR. And also the MAGMOM should be assigned in the INCAR. I also suggest make PREC = Accurate.

I know that I didn't get the exact experimental value, but in some paper report that with this procedure have got lattice constant 8.1, 8.20 and 8.218 without include AFM or FM configurations.

Any other suggestion?

Hello Maximo Ramirez,

By your INCAR file, I see you are doing volume relaxations to get the cell at zero pressure (ISIF = 3). Are you following the suggestions at section 7.6 of VASP manual?

When you perform this kind of calculations you have problems related to Pulay stress. This may be causing this difference in your results. Follow the suggestions of the VASP manual and see if you get better results.

In my opinion, the best way to get optimized zero pressure structures is to perform calculations in several volumes and then fit an equation of state (EoS). In your case, this is straightforward, since your cell is cubic.  The EoS is what gives you the right pressure for your system. The difference between the EoS pressure and the pressure given by VASP (that is calculated with the stress tensor) is small, but can lead to the kind of differences you are getting in your calculations. If you have already followed the steps in the manual, try to calculate the lattice parameter using the EoS method.

Dear Michel,

Your are correct I'm using 7.6 Volume vs. energy, volume relaxations, Pulay Stress VASP section.

I added to my INCAR PREC = Accurate and ISYM=2 and know I get acceptable lattice constant is 8.1574, less than 1% of porcentual error.

About EoS, I have not clear How can I solve one of this equations, Do you know any good resource?

Hi,

To learn more about equations of state, I suggest this book of Jean Paul Poirer.

To fit EoS, you can used the code EoSFit. Or, if you have quantum espresso installed, you can used the ev.x code (quantum espresso is a free package to perform plane wave calculations, if you don't have it, you can download and install).

The International Union of Crystallography also has a good resource page, I am sending the link.

http://www.cambridge.org/us/academic/subjects/earth-and-environmental-science/solid-earth-geophysics/introduction-physics-earths-interior-2nd-edition

http://www.quantum-espresso.org/

http://www.iucr.org/resources/commissions/crystallographic-computing/newsletters/1/fitting-equations-of-state

I think you must use FP-LAPW method and apply Spin polarization calculation because you have many Mn atoms in your compound.   

Thanks Michel for the EoS resources.

Dear Arash, I appreciate your suggestion, but this system has been studied and reported with consistent results with PAW and Ultra-Soft methods. I believe that I'm leaving out details that senior researchers have considered.