Ask A Scientist© General Science Archive

Please provide information as to the unique properties of iron, cobalt
and nickel that make them magnetic...I am a chem teacher with ca. 20 yrs
experience (I suppose I really SHOULD know this answer!).
I am asking specifically for the electronic properties that may
relate to hybridized orbitals, para/di magnetism, etc. that are the
cause of the formation of domains in these metals...
------------------------------------------------
Ferromagnetism is actually a very complicated phenomenon.  I will not be
able to answer your question to your satisfaction, but I will take a crack
at it.

I assume that you can rationalize that isolated atoms of iron, cobalt, and
nickel are high-spin.  This is a consequence of having d-orbitals close
enough in energy that electrons prefer to singly occupy all of the
d-orbitals before pairing up in any one of them.  To review, electrons have
a spin of 1/2, which means that an individual electron can have a "spin
quantum number" of -1/2 or +1/2.  When electrons pair up in orbitals, one
has a spin  Q.N. of 1/2 and th other -1/2.  When electrons have several
orbitals close in energy to choose from, they avoid pairing up in the same
orbital in order to minimize the Coulombic repulsion that would result from
being constrained to the same region of space.

Usually, discussions of the paramagnetism of atoms stop here.  However, if
you think about it, there is no obvious reason why electrons in different
orbitals should all have the same spin Q.N..  If you have four
singly-occupied orbitals, why should the magnetic moments of all the
electrons point in the same direction?  Just because they don't HAVE to
alternate (as would happen if they paired up) doesn't mean that they HAVE
to reinforce each other.  In fact, you might expect them to cancel each
other to the greatest extent possible, just as bar magnets are happiest
when they are aligned north-to-south, which does not reinforce the magnetic
field.  The reason they do reinforce each other is beyond me, but it's
called Hund's Rule, and it has something to do with the fact that these
orbitals (all on the same atom) occupy much of the same space.
(Mathematically, the orbital wavefunctions have zero overlap, because of
their symmetry properties, but their squares have considerable overlap.)
So, the individual magnetic dipoles of the unpaired electrons reinforce to
make a much larger atomic magnetic dipole.

Now, in a ferromagnetic substance, the atomic dipoles further align so that
they all point in the same direction.  The region in which all the atomic
dipoles are aligned is called a domain, as you know.  If a sample of iron
is not "magnetized," then the magnetic fields from all its domains all
point in every which way, and on a distance scale much larger that the
domain size, they cancel.  If the iron is magnetized, then the magnetic
fields from its domains STILL point in every which way, but they don't
quite all cancel.  The domains that point partly or largely in the
direction of the magnetism will have slightly stronger magnetic fields than
the domains that point partly or largely opposite the direction of
magnetism.  So, when you add the fields from all the domains together, the
sample has an overall magnetic field.

This begs the question:  Why, in a ferromagnetic substance, do domains form
at all?  Why do the atoms with their own magnetic moments align so that
their magnetic fields reinforce those of their neighbors?  As far as I can
understand, it has something to do with the overlap of atomic orbitals in
the larger sample.  Sorry, I don't know why this occurs in iron but not
molybdenum, for instance.


Richard Barrans Jr., Ph.D.
Chemical Separations Group
================================================