Page 1 of 1

Question about the CHGCAR file

Posted: Wed Dec 02, 2020 3:52 am
by Shenlan.Yang@monash.edu
Dear all,

I have a question about the CHGCAR file and I hope you could give me some help. Though I noticed that there are already some discussions on CHGCAR unit, I'd like to double check for that. Based on the VASP guide, the charge values inside the CHGCAR file indicate the charge density ρ(r) multiplied by the grid volume defined by NGX*NGY*NGZ, however, I'm not quite sure about the physical meaning of this charge density ρ(r), does it represents the charge for one grid point or it shows the charge density defined by the real supercell volume? Thanks for the help!

Regards,
Elina

Re: Question about the CHGCAR file

Posted: Wed Dec 02, 2020 8:21 am
by martin.schlipf
The CHGCAR file is setup in such a way that the typical integral
N = int d^3r rho(r),
where N is the number of electrons, is replaced by a discrete sum
N = sum_xyz rho(x,y,z) / (N_x N_y N_z),
where N_x N_y N_z, are the dimensions of the grid along the 3 lattice vectors. The rho(x,y,z) are the one you find in the CHGCAR file.

Re: Question about the CHGCAR file

Posted: Thu Dec 03, 2020 3:08 am
by Shenlan.Yang@monash.edu
Thank you for the help. And there is one thing I'm not quite sure, based on what you said, rho(x,y,z) should be the charge at (x,y,z) multiplied by (N_x N_y N_z), am I right?

Re: Question about the CHGCAR file

Posted: Thu Dec 03, 2020 8:11 am
by martin.schlipf
Yes.