W3007 Electricity and Magnetism 1

The electric field of a dipole

We know that the dipole contribution to the electrostatic potential is
1 r p
1
r p

Vdip (
r) = = , (1)
40 r 2 40 r 3
where
p is the charge distributions dipole moment, which is an intrinsic (vector) property of the
source and does not depend on
r. What is the corresponding electric field? It is

E
dip .

dip = V (2)

To take the gradient, it is convenient to use the cartesian tensor notation, with Einsteins
convention that whenever an index appears twice in a product, summation over all possible values
that index can take is understood. So, for instance, the potential is
1 ri pi
Vdip = , (3)
40 r 3

3
where ri pi i=1 ri pi =
r p
. With this notation, the j-th component of the electric field is

1  ri 
Ejdip = Vdip = pi . (4)
rj 40 rj r 3

To compute the partial derivative, we first decompose it into

 ri  ri 1 1
= + r i . (5)
rj r 3 rj r 3 rj r 3

As to the first term, we have

ri
= ij , (6)
rj
where ij is the Kronecker-delta: it is one for i = j, and zero otherwise. Eq. (6) is just a fancy
way of writing
x x
=1, =0, etc. (7)
x y
in a compact notation. As to the second term in eq. (5), we have

1  1  r  
r  rj
=

= 3 = 3 = 3 5 . (8)
rj r 3 r j
3 r j
4 r j5 r

Eq. (5) thus reduces to

 ri  1 ri rj 
= 3 ij 3 2 . (9)
rj r 3 r r
2 The electric field of a dipole

As a final step, to compute the electric field according to eq. (4), we have to multiply eq. (9) by
pi and, following the summation convention, sum over i = 1, 2, 3. We have

1 1 (pi ri )rj 
Ejdip = p i ij 3 . (10)
40 r 3 r2
Now, pi ij = pj , because the Kronecker-delta combined with the summation over all possible is
selects the term with i = j. Also pi ri = p

r. We finally get

1 1  (
p
r)rj 
Ejdip = 3 p j , (11)
40 r 3 r2
or, going back to vector notation:

dip = 1 1  (
p
r)
r 
E 3 p

(12)
40 r 3 r2
1 1 
= p
3(
r)r p

(13)
40 r 3