[Wien] a question about the lattice relaxation and the magneton number

[Stefaan Cottenier]

> I do some calculation about Zn1-xCoxO (x=1/8). I want to ask a question:
Because the electronic structure of Co is very different from that of Zn, I
should consider the lattice relaxation. How can I deal with it?

In principle by simulataneously relaxing the 6 crystal cell parameters
(a,b,c,alpha,beta,gamma) and the internal coordinates. In practice:

* determine the best a,b,c (using 'x optimize'), often the angles are not
too much affected.
* possibility 1: when doing the previous step, in your optimize.job file do
always a relaxation of internal coordinates by min_lapw. This is a rather
accurate but time-consuming procedure.
* possibility 2: relax the internal coordinates (min_lapw) only for the best
a,b,c. Then freeze them, again search the best a,b,c, then relax internal
coordinates again, and so on (but very often 1-2 iterations are fine)

Information about how to use 'optimize' or 'min_lapw' is found in the
usersguide.

>Another question is how I could get the magneton number.

Do you mean magnetic moments expressed in mu_B? They are given by the
:MMIxxx labels in case.scf.

Stefaan