Full magnetization density from two-component calculations

[antti_karttunen]

Hi all,
I’m currently investigating small metal clusters using the two-component formalism implemented in Turbomole. I’d be really glad if someone could help me with two small details.

1) I’m interested in plotting magnetization densities as 3D vector plots (see for example Figures 1 and 2 in http://www.mariecurie.org/annals/volume4/phy3.pdf). I did not find any tips from the manual section “Keywords for wave function analysis and generation of plotting data”, but then I tested just simple
$pointval fmt=xyz
and I’m indeed getting a sd.xyz with the following data:
# cartesian coordinates x,y,z and f(x,y,z)
-0.4000D+01 -0.4000D+01 -0.4000D+01 0.2415D-05 0.2357D-05 0.9666D-05 0.1024D-04
In the case of a scalar calculation, the sd.xyz only contains x,y,z and the spin density (scalar). So, it would seem that one really gets the magnetization vector components and the norm of the vector with this approach. Am I correct? This would be really good news.

2) My second question concerns the contents of spinvec.txt. The file contains the atomic coordinates and spin vector at each atomic position (so actually this is also related to the question above). Now, the file also states that “x,y,z are arbitrary chosen” for the spin vector. I’m a bit confused about this. What does “arbitrary” mean here? The spin vectors are not given in the same coordinate system as the molecule? I’ve read several papers on the fundamentals of the two-component formalism, but somehow I cannot quite figure out what this means. Is this related to the rotational invariance of the non-collinear approach?

Thanks,
Antti

[antti_karttunen]

OK, I should have been more careful with the manual, my first question was actually answered on page http://www.cosmologic.de/data/DOK_HTML/node261.html
So, $pointval fmt=txt indeed prints out the vectors, great. I'm still confused about the "arbitrary chosen" x,y,z, though 

Antti