Jyoti Kapil It is hard to say in general. But for spin-orbit coupling like relativistic parameters, you have to use pseudopotential having relativistic effect. Likely we can say for other things... If your materials is novel, then it is better to move with all electron calculation instead of using pseudopotentials.

Then for your calculation...It also depends upon the ecut, kpoints and other values. So, check with different values and find out whether your total energy become stable. Then calculate the desired properties.

Also, if you get correct magnetic moment(comparing experimental results), then you can't conclude that the structure is correct.

Moreover, the mulliken population itself is not accurate(if you calculated magentic moment from Mulliken population).

Anyway have a look for pseudopotential basics

http://www.tcm.phy.cam.ac.uk/~jry20/gipaw/tutorial_pp.pdf

Thanks Aadhityan Sir.

I have checked the convergence with respect to ecut and kpoints and using the converged values to find out the relaxed structure (vc-relax). So how does one decide and at what point that we are getting correct result (if only getting magnetic moment doesn't help) as I am getting the band structure as well as the DOS correct. Please Guide. The lattice constant that I am getting is 5.611 angstrom ; the experimental value is 5.645 Angs.

And one more thing I want to ask is how to decide in vc-relax which option do we have to go for as in do we allow the shape to be modified during relaxation etc.

Thanks a lot in advance.

Please excuse if my questions seem a little trivial.

Jyoti Kapil If you get correct DOS means...Almost That's enough.

Also the difference in lattice constant is very small. So, it is not at all a problem. So, you may proceed with this result.

Vc relax option in quantum espresso is depend upon the information you have about your material(to optimize). If you need more info... Let me know...

Thanks Sir . It really helped.

Yes I would like to more about vc-relax.

Thanks in advance.

Jyoti Kapil when you do vc-relax, the cell and atomic position get optimized(changed). If you know the unit cell shape, you can fix using the keyword ibrav as mentioned in the input description.

https://www.quantum-espresso.org/Doc/INPUT_PW.html#idm200

The same applies to the symmetry. If you know the symmetry and want to fix means you can using nosym=.FALSE

https://www.quantum-espresso.org/Doc/INPUT_PW.html#idm303

Thanks...

Thanks a lot Sir for your help.