DOS/Spin/ELF

[Dipankar Saha]

1) Is there any tutorial dealing with....ELF? How do you implement it... (Is it the same... as it was derived by Becke and Edgecombe?) ....?? Then what does it signify --> "ELF ranges from 0 to 2" .....when we take.... (up_spin + dwn_spin) ??

2) Besides,  for "Spin: Unpolarized" calculations.... I find that the DOS considers "all spin" ....!! For such materials, where there  is no magnetic moment is expected..., should it be divided into two parts....just for referring those to the spin-up and spin-down components?

[Umberto Martinez]

1) you can rad the corresponding entry in the Reference Manual where reference is included:
http://quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/ref.electronlocalizationfunction.html
Yeu will be ables, you can specify Spin.Up and Spin.Down which will be the same for unpolarized solutions and they will both range from 0 to 1.

2) For unpolarized calculations, spin up and down will be exactly the same. Still, you can plot both components if you wish.

[Dipankar Saha]

What that is confusing to me.... is the value  "2" !! It's said that....ELF is a measure of Pauli Repulsion (meaning the probability of finding a charge carrier with same spin in its neighborhood) ...Thus it lies within 0 to 1 ('1' -->when no other electron with the same spin may be found in the vicinity of one) ....Is it not?? / Then..while we find in the contour plot... a region showing ELF=2...what does that physically signify??

[Umberto Martinez]

Ok I think I see where the confusion comes from.
If you plot your ELF as isosurface or contour plot you will be asked which grid to plot, i.e. up-down, up+down, up, or down.
Indeed if you choose up+down your data will range from zero to two since VNL is putting the two grids together.
However, the point is that for this analysis it does not make a lot of sense to plot up-down or up+down.
This is exactly for the reason that you underline.

So, you should make two plots, for up and down components. Both will range from 0 to 1.

Does this clarify your issue?

[Dipankar Saha]

Now this is perfectly alright...!!! I was also feeling  that...,  there is no point...in taking spin_up+spin_down together...while plotting  ELF.....!!!
Thanks a lot Umberto.....