I've performed groundstate and electronic structure calculations using EXCITING_carbon.

The material that I'm currently studying is perovskite BaSnO3 (BSO) (cubic structure).

However, I've found that there are several problems in results of density of state (DOS) calculation.

First, I performed a groundstate calculation for BSO 1 u.c., and also BSO 2 u.c. (I need to calculate a structure with a bigger (5x4x1) supercell, so I wanted to check whether the bigger supercells (BSO 2 u.c.) generates the same result with that of BSO 1 u.c.).

Then, I calculated total DOS of them. Here, the calculated total DOS of BSO 1 u.c. and BSO 2 u.c. were slightly different…

Is there any parameters in an input file that I should modify when the supercell size increase??

Second, I checked the partial DOS of BSO 1 u.c.

In BaSnO3, three oxygen atoms are symmetrically equivalent.

But, their partial DOSs look very different, which is not correct.

Could I get any advise or suggestion, please?

Thanks a lot.

Hwanhui Yun