Meaning of electron charge distribution (CHGCAR) on the nucleus
Posted: Mon May 02, 2011 9:06 am
Dear all,
I have a question concerning the meaning of electron charge distribution given by CHGCAR file.
From the VASP Manual it is clear, that this file should contain information about valence electron only (determined by selected potential).
Summation of all points from CHGCAR and dividing it by the volume of primitive cell gives exact number of used valence electron
For example, sum=28 for primitive cell Na Co O2 where used potenitals are:
O: s2p4 * 2 atom = 12
Co: d8 s1 * 1 atom = 9
Na: p6s1 * 1 atom = 7
This electron "concentration" is given on 3D mesh, equidistant in each direction. Some point of this mesh lies inside "core spheres" of atoms (sphere centered at nucleus with radius RCORE for given atom type).
My question is how valence electron can be inside the "core sphere" and even in the center of the nucleus (f.e. in Na Co O2, Na has fractional coordinates (0;0;0), whereas CHGCAR lattice also "starts" from (0;0;0) point ?
May be one has to "screen" this charge distribution function rho(R) by multiplying on damped function with F(R_nucleus) = 0 and F(R_core) = 1 ?
Thanks in advance
I have a question concerning the meaning of electron charge distribution given by CHGCAR file.
From the VASP Manual it is clear, that this file should contain information about valence electron only (determined by selected potential).
Summation of all points from CHGCAR and dividing it by the volume of primitive cell gives exact number of used valence electron
For example, sum=28 for primitive cell Na Co O2 where used potenitals are:
O: s2p4 * 2 atom = 12
Co: d8 s1 * 1 atom = 9
Na: p6s1 * 1 atom = 7
This electron "concentration" is given on 3D mesh, equidistant in each direction. Some point of this mesh lies inside "core spheres" of atoms (sphere centered at nucleus with radius RCORE for given atom type).
My question is how valence electron can be inside the "core sphere" and even in the center of the nucleus (f.e. in Na Co O2, Na has fractional coordinates (0;0;0), whereas CHGCAR lattice also "starts" from (0;0;0) point ?
May be one has to "screen" this charge distribution function rho(R) by multiplying on damped function with F(R_nucleus) = 0 and F(R_core) = 1 ?
Thanks in advance