interstitial contribution to DOS

Message

jan · #1 Post by **jan** » Sat Jun 07, 2008 4:32 am

Hello !
Is there any way to find a contribution of interstitial regions to total DOS?
I found that the sum of all s, p, d, f of all atoms per unit cell is not equal to the total DOS. I also conjecture that the sum depends on RWIGS keyword.
Then, is it true that the total DOS - the sum of all projected DOS's = interstitial DOS, based on the RWIGS setting? Or VASP just prints out the interstitial DOS somewhere?
Can anyone explain this?
I found some impurity states missing in the sum, forcing me to think that actually impurity states arise from the interstitial region (somewhat nonsense?).

Thanks,
Jay

#2 Post by **admin** » Thu Jun 12, 2008 1:44 pm

you are correct, the interstital DOS would be the total dos - the sum of all projected DOS. It is not explicitely calculated and written by vasp.
Please however note that the projected DOS (and therefore the interstitial DOS, as well) are somewhat arbitrary, depending on the choice of the atomic radius (RWIGS), in all PW- and mixed-basis set codes.

jan · #3 Post by **jan** » Fri Jun 13, 2008 8:51 am

thanks for the clarification.

Jan