What does a magnetic moment of an atom mean?

[schrodingerscat11]

I am using VASP and it recommended to use the experimental magnetic moment for MAGMOM tag. The problem is I can't find a standard table of magnetic moments. I see different values from different sources. E.g In Web Elements, the magnetic moment of Fe is 0.096 μ/μ_N. But in another source, it is 2.22 μ_N per atom. Is there an "official" table of magnetic moments in standard units (whatever that unit may be)? Thanks.

 

[Simon Bridge]

2.2 bohr magnetons seems about right for the atomic magnetic moment of atomic Iron... I have not found a source disagreeing with that.

I found this:
http://www.webelements.com/iron/
... that what you saw?

About half way down the page, under the heading "iron isotopes" there is a table of properties, listing the value you quoted under the column "nuclear magnetic moment".
0.09 bohr seems about right for the nuclear magnetic moment.

I cannot check the other reference because you did not cite it.

What counts as a good reference depends on what you want the information for.
A standard desk reference is usually good enough for proof-of-concept type stuff, i.e. the Web Elements page used:
Mills, I. et al. Quantities, Units and Symbols in Physical Chemistry, Blackwell Scientific Publications, Oxford, UK, 1988.

Bleeding edge development may be sensitive to the precision of numbers so it should either be measured in-house or the best recent measurement from the literature for your field should be used.

Standard accepted values can usually be found published by the national standards body for your jurisdiction.

 

[schrodingerscat11]

Thanks! I found an example and they used the one in Bohr magnetons, (2.22). The source is Kittel's book.. :)

 

[Simon Bridge]

Well done.
In general, when looking up standard references, you have to be sure to read the surrounding text carefully - that will explain what, exactly, is being measured and, sometimes, how. People get caught out all the time. A good source should also have an error estimate with measurements.

 

[schrodingerscat11]

Thanks! I will follow your advice. :)