University Physics Notes: Electricity and Electromagnetism - Energy in Electric Fields in the Presence of Dielectrics
We can illustrate how energy is stored in an electric field using
the example of a parallel plate capacitor. The capacitor is initially
uncharged, and a potential difference
is
built up between the plates by transferring a charge
from
one plate to the other. Energy – supplied for example by a battery
– is needed to move the charge, and at the moment when the
potential difference is
an additional amount of energy
is
needed to transfer a further charge![]()
(1)
If a dielectric is placed between the plates then
will
be changed since the dielectric changes the value of
but
only
the
free charge on the capacitor plate, appears in (1) which contains no
reference to polarization charges. This is because it is only the
free charge which is moved across the potential difference by
external forces.
In a parallel plate capacitor of area
and
plate separation
the
energy stored when it is filled with dielectric is
The
volume of the capacitor is
and
therefore the energy density is![]()
We can regard the energy density
as
residing in the field. We can imagine field is generated by a large
number of parallel plate capacitors by thin conductors placed along
closely placed equipotentials. The total electrostatic energy stored
in a volume
is
then![]()
The same equation applies to the potential energy stored by an
arbitrary distribution of charges. We start with an unpolarised
dielectric with no free charges and assemble the charge distribution
by bringing the charges from infinity. Work equal to
is
done in assembling the charge distribution where
is
the potential at the position of
If
the free charges are distributed with surface charge density
on
a number of conducting surfaces, and with volume charge density
in
the region
bounded
by the conductors, the sum is replaced by an integral and
(2)
On the conducting surfaces the outward normal to
is
the inward normal to the conductor so
and
together with
we
have from (2)
![]()
We can use Gauss's Divergence Theorem on the second term to give
![]()
Now we can use the vector identity
and put
to obtain![]()