Magnetic multipole moments

We characterized the scalar potential V for a localized

charge distribution at a point far from the charges with a
multipole expansion of V:

We can do the same with a vector potential due to a

localized current distribution.
The monopole moment turns out to vanish. The electric
monopole moment was just the total charge of the
distribution. Since we have no magnetic monopoles, we might
expect the monopole moment to vanish. It is clear that the
first term vanishes in the case of filamentary currents. That
integral is just the vector displacement about a closed loop,
which is zero. The same scenario will hold for a closed surface
current. The proof that the monopole term vanishes for the
case of a volume current density is given in Wangsness
Section 19-1-1. It relies on a vector identity and use of the
divergence theorem.
The magnetic monopole moment always vanishes.

The magnetic dipole term will be the dominant term in the

multipole expansion. It is the only term we will consider. It
will also turn out to be independent of the choice of origin
of coordinates. This might also be expected since we found
that the electric dipole moment was independent of the
choice of origin when the total charge (electric monopole)
vanished.

The nice thing about the electric multipole expansion was

that we could use it to characterize a charge distribution.
So we would like to manipulate our magnetic dipole term
to separate the source description from the position of the
point at which we calculate the vector potential. We will
find the following result, which defines the magnetic
dipole moment.
The proof of the relation given above for the case of a
continuous volume current distribution is given in Section
19-1-2 of Wangsness. It is quite involved but straightforward
so you can read it there. Here, we will go through the
equivalent proof of the relation for the case of a filamentary
circuit.

We want to cleanly separate all dependence on the field

point from the integration over the circuit. We can
manipulate this form into an ordinary line integral form and
apply Stoke’s theorem to accomplish this. Consider Stoke’s
theorem in general first.
So the desired separation is achieved here.
So we can summarize the dominant dipole term in a common
form for each type of current distribution.

Given the vector potential, one can calculate the magnetic

field associated with the idealized dipole contribution by
taking the curl. In the electrostatic case, we had a simple
picture of a physical dipole and a straightforward result for
the leading dipole term of the electric field.
The above would hold for a loop lying in the x-y plane at the
origin. For a loop of radius a, small relative to r,
 Magnetism in the Presence of Matter
Now we will go beyond the case of currents in vacuum. The
concept of magnetic dipole will play the role that was taken
by electric dipole when we discussed dielectric materials.

Electric dipoles, as embodied by the charge separation of

particles in atoms and molecules, felt the influence of
external electric fields with the result characterized by the
macroscopic polarization, and then described by bound
volume and surface charge densities.

With magnetic materials, it is ultimately the motion of

charged particles (primarily electrons) that leads to the
equivalent property of magnetization. Both orbital motions
and intrinsic angular momentum (spin – quantum
mechanical property) contribute to the magnetic properties
of materials. The microscopic situation is complex and
includes cases of paramagnetism (alignment of intrinsic
magnetic moment by external field), diamagnetism (all
materials, external field induces magnetic moment) and
ferromagnetism (domains of aligned magnetic moments
arising from electron spin interactions).

We will describe magnetization in terms of volume and

surface current densities.
Take a small circular current loop as our prototype of a
magnetic dipole. We can consider the influence on the loop
(force and torque) of an external magnetic field.

If the applied magnetic field is constant, the force on the loop

vanishes. In the case of a position dependent magnetic field,
the simplest way to derive an expression for the force is to
approximate the value of the field at each point on the small
loop via a Taylor expansion relative to the value of the field at
the center of the loop (at the origin). The full details are given
in Wangsness Section 19-4. But we will just quote the result
because it is just like the situation we had in electrostatics:

So this just serves to demonstrate that our magnetic

dipoles can experience a force due to an applied field.
 Likewise, a magnetic dipole experiences a torque due to
an applied magnetic field – even for a constant field.
Here, consider that the applied field is constant.

So we can see the possibility of alignment of magnetic

dipoles within a material at the microscopic level.
With polarized dielectrics, we used the polarization to find
bound volume and surface charge densities. With
magnetized material, we will find volume and surface current
densities from the magnetization. The easiest way to extract
those current densities is by consideration of the vector
potential for the material. We know how to write the
contribution to the vector potential of a loop of current due
to its dipole moment.

This is the dominant term for distances large relative to the

loop size.
We will leave the second volume integral intact but we will
manipulate the first one into a surface integral. The volume
over which we are integrating is the volume of the
magnetized material.
Use this expression to rewrite our result for the vector
potential.
This is the form we want. Recall we have seen the vector
potential expressed in terms of current densities. This
corresponds to precisely those forms, for equivalent surface
and volume current densities.

These surface and volume current densities, arising as a

result of the magnetization of a material are referred to as
equivalent, Amperian, magnetization, or bound current
densities. We’ll use the term in Wangsness – magnetization
current densities.

Just as with the bound charge densities in dielectrics,these

magnetization current densities are real currents. When the
magnetization is uniform (curl M =0), there are only surface
currents. By thinking of the aligned magnetic dipoles as small
current loops, it is easy to see how the surface current
density arises.
For uniform magnetization, in the volume of the material,
currents from adjacent “dipole loops” cancel. If the
magnetization has a spatial dependence such that its curl
does not vanish, then this cancellation is incomplete and the
volume current density is nonvanishing.
So we have a continuity statement again, but for the case of
a current density associated with magnetization, we find that
it cannot contribute to any charge accumulation.

With dielectric materials, we progressed from the concept of

polarized material as an assembly of dipoles described the
the polarization, to a description in terms of equivalent
volume and surface charge densities.

With magnetic material, we go from the concept of

magnetized material as an assembly of magnetic dipoles
(think of as small current loops) described by the
magnetization, to a description in terms of equivalent
volume and surface current densities.
The Biot-Savart law for the magnetic field is modified to
include magnetization current densities:

The subscript “outside” is a reminder that we developed this

description of magnetized material starting with the
multipole expansion of the vector potential. However, as long
as we do not try to describe magnetic material at the
microscopic level, the result also holds inside matter, just as
we found for the case of dielectrics.

Now we can do a bit of a summary by revisiting our results

on magnetism for the vacuum case, with our new
understanding of the addition of magnetic materials. We will
extend our results for magnetism in this way.
 Magnetism so far, Introduction of H, and Some
Boundary Conditions

Adding magnetic materials, the magnetic field due to

magnetization current densities arises due to the motion of
charges, just as for free currents. So the form of the magnetic
field is identical. Thus our result remains intact – after all, we
interpret it as implying no static magnetic charges.
So this result is modified when we include magnetic materials.

It is important to note that this result was derived for the case
of static fields. It will be further modified .

We called the above result Ampere’s law in differential form.

Like Gauss’ law in diff form, it has limits to its application. Both
laws only work for situations where the fields are
differentiable. So we saw that the electric field was
discontinuous at an interface containing a surface charge. In
that case, we had to use the integral form of Gauss’ law.

Likewise, the diff form of Ampere’s law will not hold at the
surface of a magnetized material where we have a surface
current density. An obvious example is an infinite plane sheet
of current, where the magnetic field is clearly discontinuous.
Let’s be guided by what we did with the introduction of
dielectrics in further manipulating Ampere’s law in diff and
integral forms. We separated free and bound charge density.
 Just as we found a form of G’s law for displacement, we
can find a form of Ampere’s law for H. Consider the flux of
curl H through an open surface S.

It is important to remember that these results were derived

for steady currents – that is, using