Chapter 5

Electrostatic Fields in Mediums

This chapter is about electrostatics in materials (or mediums). First, we describe the interaction between

an electric field and a medium. Then, we consider the situation where the source of the field is inside the

medium. We then discuss energy in a medium, boundary conditions, conductors, and capacitors.

5.1 Mediums and Perturbation Electric Fields

5.1.1 An Atom

An atom can be thought of as a cloud of electrons surrounding a nucleus. You may think of this cloud of

electrons as an effective negative point charge and the nucleus as an effective positive charge. Assuming

both particles having the same charge, an atom can be considered as a dipole with dipole moment p.

Figure 5.1: An atom.

5.1.2 A Medium

A medium can be thought of as a group of atoms as shown below.

1
ELEC 325 Hamad M. Alkhoori Fall 2024

Figure 5.2: A medium.

A medium is not described by p, but rather by a polarization vector P. Suppose that inside a medium,

there exists a region ∆V with a total dipole moment ∆p. Then, the polarization vector is defined as

∆p dp
P = lim = . (5.1)
∆→0 ∆V dV

Figure 5.3: Polarization vector.

5.1.3 Interaction between a Field and a Medium

In many mediums, and in the absence of an electric field, the dipoles are randomly oriented, such that

P = 0 (as s simple case). When an incident electric field Einc is applied, the dipoles align themselves

2
ELEC 325 Hamad M. Alkhoori Fall 2024

either parallel, antiparallel, or in an arbitrary configuration with respect to Einc . These dipole are said to

be polarized. Due to their new arrangement, there will be a net P, which gives rise to a perturbation

electric field Epert outside the medium, and an internal electric field Eint inside the medium. Those

fields, which result because of P, must depend not only on Einc , but also on the composition of the medium.

These include (i) the force due to the mass of the electron, (ii) the binding force since the electron is attached

to the proton, and (iii) the damping force.

5.1.4 Computing the Perturbation Electric Field

In many applications, it is the perturbation electric field that is being calculated. Given a material occupying

a region V ′ with a polarization P, how to compute the perturbation electric field? This can be done more

easily by deriving an expression for the perturbation potential Vpert , and then find the perturbation electric

field from Epert = −∇Vpert .

The differential perturbation potential due to one dipole moment in V ′ is given by

1 dp • R
dVpert = . (5.2)
4πε0 R3

Using Eq. (5.1), the perturbation potential becomes

P(r′ ) • R
Z
1
Vpert = dV ′ . (5.3)
4πε0 V′ R3

Then, Epert = −∇Vpert .

Another Point of View

As another point of view, using some vector identities, the perturbation potential can be rewritten as

P(r′ ) • ân′ ′
I Z
1 1 1
Vpert = dS + [−∇′ • P(r′ )]dV ′ . (5.4)
4πε0 S′ R 4πε0 V′ R

If we define

ρbs = P • ân , (5.5)

as the bound surface charge density, and

ρbv = −∇ • P, (5.6)

3
ELEC 325 Hamad M. Alkhoori Fall 2024

as the bound volume charge density, Eq. (5.4) can be written as

ρbs (r′ ) ′ ρbv (r′ ) ′

I Z
1 1
Vpert = dS + dV . (5.7)
4πε0 S′ R 4πε0 V′ R

Then, Epert = −∇Vpert . Equation (5.7) shows that the potential (and hence the field) due to a polarized

object is equivalent to a potential due to a surface charge density ρbs , and to a potential due to a volume

charge density ρbv .

Hence, if we expose a medium to a field, either we say, a polarization vector P is formed, which gives

Epert , or we say induced charges ρbs (on the surface of the medium) and ρbv (inside the medium) are formed,

which give Epert .

Example 1: A cube centered at the origin with a side length L is composed of a medium characterized by

P = P0 (xâx + yây + zâz ). Find the bound charges.

Ans:
ρbv = −∇ • P = −3P0
 
 



 P âx ,
• x = L/2 


 1, x = L/2

 


 




 P • (−â x ), x = −L/2 


 1, x = −L/2

 


 

 P • ây ,
 y = L/2 
 1, y = L/2
b
ρs = = P0 L/2
 



 P • (−ây ), y = −L/2 


 1, y = −L/2

 


 




 P • â z , z = L/2 


 1, z = L/2

 


 

 P • (−âz ), z = −L/2
 
 1, z = −L/2
Further Readings:

ˆ Section 4.1 in ’Introduction to Electrodynamics by David’.

ˆ Section 4.2 in ’Introduction to Electrodynamics by David’.

ˆ Section 4.3 in ’Introduction to Electrodynamics by David’.

ˆ Section 9.4.3 in ’Introduction to Electrodynamics by David’. (Although the derivation in that section

is carried out for an electromagnetic field, it accommodates an electrostatic field by simply setting the

operating frequency of the electromagnetic field to zero.)

ˆ Sections 5.5 and 5.6 in ’Elements of Electromagnetics by Matthew’.

4
ELEC 325 Hamad M. Alkhoori Fall 2024

5.2 Sources Inside Mediums

In this section, we learn how to compute the electric field when the source itself lies inside the medium.

Our starting point is Maxwell equations


∇×E=0 


, (5.8)
ρv 
∇•E=


ε0
in particular,
ρv
∇•E= . (5.9)
ε0

In free space, ρv is just the free (impressed) charge ρfv . In a medium, ρv is the sum of the free charge ρfv , as

well as the bound charge ρbv . Hence,

ρfv ρb ρb ∇•P
∇•E= + v = v − . (5.10)
ε0 ε0 ε0 ε0

On moving ∇ • P to the left side, and multiplying both sides by ε0 we get

∇ • (ε0 E + P) = ρfv . (5.11)

It is customary to define the quantity inside the brackets as an auxiliary field D. That is

D = ε0 E + P. (5.12)

Therefore, Gauss’s law becomes

∇ • D = ρfv , (5.13)

or in integral form,
I
D • dS = Qfenc . (5.14)
S

With the D field, Maxwell equations for electrostatics become


∇×E=0 


. (5.15)

f
∇ D=ρ
• 
v

It remains to specify D in terms of E to eliminate D. Such a relation is called a constitutive relation.

That is,

D ≡ f (E). (5.16)

5
ELEC 325 Hamad M. Alkhoori Fall 2024

So, to determine the electric field in any medium, we solve


∇×E=0 


, (5.17)

f
∇•D=ρ


v

subject to

D ≡ f (E). (5.18)

Example 2: An infinite polarized cylinder of radius a and an unknown polarization vector P carries a uniform

ρfv . Find D and E everywhere.

Ans: I
D • dS = 2πρLD
S
Z ZL Z2πZρ
ρ<a: Qfenc = ρfv dV = ρfv ρdρdϕdz = ρfv πρ2 L
V
0 0 0
ρf ρ ρf ρ
2πρLD = ρfv πρ2 L ⇒ D = v ⇒ D = v âρ , ρ < a.
2 2
Z ZL Z2πZa ZL Z2πZρ
f f f
ρ > a : Qenc = ρv dV = ρv ρdρdϕdz + 0 ρdρdϕdz = ρfv πa2 L
V
0 0 0 0 0 a
ρfv a2 ρfv a2
2πρLD = ρfv πa2 L ⇒ D = ⇒D= âρ , ρ > a.
2ρ 2ρ
 
f
 ρv ρ â , 0 < ρ < a
 
 unknown, 0 < ρ < a

 ρ
⇒D= 2 ⇒ E =
ρf a2 ρf a2
 v âρ , a < ρ < ∞  v âρ , a<ρ<∞

 

2ρ 2ρε0
E is unknown inside because P is unknown.

5.2.1 Isotropic Dielectric Medium

Now, we consider the simplest medium. When the polarization P is either parallel or antiparallel to the

electric field E, without any dependence on the locations of the dipoles, such a medium is called isotropic.

This can be described mathematically as

P = ε0 χe E, (5.19)

where χe is the electric susceptibility scalar. On substituting P = ε0 χe E into D = ε0 E + P, one obtains

D = εE, (5.20)

6
ELEC 325 Hamad M. Alkhoori Fall 2024

where ε = ε0 (1 + χe ) = εr ε0 is the permittivity of the medium. The quantity εr is called the relative

permittivity of the medium

Hence, in isotropic mediums, we find D first from

I
D • dS = Qfenc , (5.21)
S

where Qfenc contains only free charges, and then we find E from

D = εE. (5.22)

Example 3: An infinite cylinder of radius a is composed of an isotropic dielectric medium with permittivity

ε, and it is carrying a uniform ρfv . (a) Find E everywhere. (b) Locate the bound charges.

Ans: (a)  
 ρfv ρ  ρfv ρ

 âρ , 0 < ρ < a 
 âρ , 0 < ρ < a
D= 2 ⇒ E = 2ε
f 2 f 2
ρ a ρv a
 v âρ , a < ρ < ∞ âρ , a < ρ < ∞

 


2ρ 2ρε0
(b)
εr − 1 ρfv ρ
P = ε0 χe E = âρ
εr 2
εr − 1 f
ρbv = −∇ • P = − ρv
εr
εr − 1 ρfv a
ρbs = P • âρ =
ρ=a εr 2
Example 4: A sphere of radius a carries a uniform surface charge density ρfs is immersed in a sphere of

radius b (b > a) composed of an isotropic dielectric medium with permittivity ε. (a) Find E everywhere. (b)

Locate the bound charges.

7
ELEC 325 Hamad M. Alkhoori Fall 2024

Ans: (a)
I
D • dS = 4πr2 D
S
Z
f
r < a : Qenc = ρfs dS = 0 ⇒ D = 0 ⇒ E = 0
S
ρf a2 ρf a2
Z
f
a < r < b : Qenc = ρfs dS = ρfs 4πa2 ⇒ D = s 2 âr ⇒ E = s 2 âr
S r εr
Z f 2 f 2
ρ a ρ a
r > b : Qfenc = ρfs dS = ρfs 4πa2 ⇒ D = s 2 âr ⇒ E = s 2 âr
S r ε0 r







 0, 0<r<a
 0,
 0<r<a 
 f 2
ρs a
⇒D= f 2
⇒E= âr , a < r < b
 ρ s a  εr2

 â r , a < r < ∞ 

ρf a 2
r2

 s âr , b < r < ∞


ε0 r 2
(b)
εr − 1 ρfs a2
P = ε0 χe E = âr
εr r2

ρbv = −∇ • P = 0
 
ε −1 f
 − r
 
 P • (−âr ),
 r=a  ρs , r = a
b
ρs = = εr
f 2

 P • âr ,
 r=b  εr − 1 ρs a , r = b


εr b2

5.3 Energy in a Medium

Recall that the electrostatic energy in free space is

Z
ε0
W = E 2 dV, (5.23)
2 V∞

where this integration is carried out over the entire space V∞ . In an isotropic dielectric medium of permit-

tivity ε, Eq. (5.23) generalizes to

Z
ε
W = E 2 dV, (5.24)
2 V∞

which can be rewritten as

Z
1
W = D • E dV. (5.25)
2 V∞

Although Eq. (5.25) was derived for an isotropic dielectric medium, it is valid for any medium. Note that,

Eq. (5.25) represents the total energy (i.e., the energy of the free charges W f and the energy of the bound

charges W b ). To obtain the energy of the bound charges only,

Z Z Z
1 ε0 1
Wb = W − Wf = D • E dV − E • E dV = P • E dV (5.26)
2 V∞ 2 V∞ 2 V∞

8
ELEC 325 Hamad M. Alkhoori Fall 2024

Example 5: A sphere of radius a composed of an isotropic dielectric medium with permittivity ε carries a

uniform volume charge density ρfv . Find (a) the total energy W , (b) the free energy W f , and (c) the bound

energy W b .

Ans:  
f ρf r
 ρv r ,  v ,
 
 0<r<a  0<r<a
D= 3 E= 3ε
f 3 ρf a3
 ρv a âr , a < r < ∞  v âr , a < r < ∞

 

3r2 3ε0 r2
(a)
Z∞
 a 
Z Z
1
W = D • E dV = 2π  D • E r2 dr + D • E r2 dr
2 V∞
0 a
Z∞  f 3 2
 a 
 f 2
2π(ρfv )2 a5 1
Z  
1 ρ v r 1 ρ v a 1
= 2π  r2 dr + r 2 
dr = + .
ε 3 ε0 3r2 9 5ε ε0
0 a

(b)
2π(ρfv )2 a5 4π(ρfv )2 a5
 
f 1 1
W = + =
9 5ε0 ε0 15ε0

(c)
2π(ρfv )2 a5
 
b f 1 1
W =W −W = − .
45 ε ε0

5.4 Boundary Conditions

When two or more mediums are adjacent to each other, the electric fields of these mediums can be related

to each other at the boundaries of these mediums. Such relations can be obtained from the boundary

conditions among these mediums.

Figure 5.4: Boundary separating two regions.

9
ELEC 325 Hamad M. Alkhoori Fall 2024

Consider region 1 to be composed of an isotropic dielectric medium of ε1 and region 2 to be composed

of an isotropic dielectric medium of ε2 . A surface S (a boundary) is separating the two mediums (see Fig.
∥
5.4). The field in region 1, E1 , can be decomposed into a tangential component E1 and a normal component

E⊥
1 by

∥
E1 = E1 + E⊥
1. (5.27)

Likewise, the field in region 2, E2 , can be decomposed similarly by

∥
E2 = E2 + E⊥
2. (5.28)

∥ ∥
On the surface S, we want to find a relation between E1 and E2 , and between E⊥ ⊥
1 and E2 . This becomes

useful if one field is known and the other is unknown.

∥ ∥
Tangential components: To find a relation between E1 and E2 on the surface S, we apply
I
E • dl = 0, (5.29)
L

where L is shown in Fig. 5.5.

Figure 5.5: Illustration for boundary conditions.

With the assumption that the dimensions ∆w and ∆h are very small, we can take the integrand out of

the integral. Therefore, Eq. (5.29) becomes

∥ ∆h ∆h ∥ ∆h ∆h
E1 ∆w − E1⊥ − E2⊥ − E2 ∆w + E2⊥ + E1⊥ = 0, (5.30)
2 2 2 2

which reduces to
∥ ∥
(E1 − E2 )∆w = 0. (5.31)

10
ELEC 325 Hamad M. Alkhoori Fall 2024

To make L intersect with S, we must let ∆h → 0. Therefore,

∥ ∥
E1 = E2 , (5.32)

which can be written as

ân × (E1 − E2 ) = 0, (5.33)

where ân is a unit normal on S that points into region 1. Therefore, the tangential component of the electric

field is said to be continuous across S. Keep in your mind that this boundary condition is valid only on the

surface S. That is,

∥ ∥
E1 = E2 , r ∈ S, (5.34)

or

ân × [E1 − E2 ] = 0, r ∈ S. (5.35)

Normal components: To find a relation between E⊥ ⊥

1 and E2 on the surface S, we apply

I
D • dS = Qfenc , (5.36)
Ω

where Ω is a cylindrical surface of radius a as shown in Fig. 5.6.

Figure 5.6: Illustration for boundary conditions.

With the assumption that the dimensions ∆S and ∆h are very small, Eq. (5.36) becomes

∥ ∥
D1⊥ ∆S + D1 (2πa∆h) − D2⊥ ∆S + D2 (2πa∆h) = ρfs ∆S. (5.37)

Upon letting ∆h → 0, Eq. (5.37) becomes

D1⊥ − D2⊥ = ρfs , (5.38)

11
ELEC 325 Hamad M. Alkhoori Fall 2024

or

ân • (D1 − D2 ) = ρfs . (5.39)

The electric fields of both mediums can be found as

ε1 E1⊥ − ε2 E2⊥ = ρfs , (5.40)

or

ân • (ε1 E1 − ε2 E2 ) = ρfs . (5.41)

Again, this boundary condition is valid only on the surface S. That is,

D1⊥ − D2⊥ = ρfs , r ∈ S, (5.42)

or

ân • (D1 − D2 ) = ρfs , r ∈ S. (5.43)

A summary: The boundary conditions are


∥ ∥
E1 = E2



. (5.44)

D1⊥ − D2⊥ = ρfs 


Therefore, given E1 , the field E2 is given by

∥ 1
ε1 E1⊥ − ρfs ân .


E2 = E1 + (5.45)
ε2

Example 6: Free space ε1 = ε0 and a medium with ε2 = 4ε0 are separated by a plane x = 0. Free space

occupies x > 0 and the medium occupies x < 0. Given E1 = 12âx − 10ây + 4âz , find E2 .

Ans:
ân = âx

ε1 ⊥
E⊥
1 = (E1
• ân )ân = 12âx ⇒ E⊥
2 = E = 3âx
ε2 1
∥ ∥ ∥
E1 = E1 − E⊥
1 = −10ây + 4âz ⇒ E2 = E1 = −10ây + 4âz

⇒ E2 = 3âx − 10ây + 4âz .

Example 7: Two mediums with ε1 = 4ε0 and ε2 = 3ε0 are separated by a plane z = 1. Medium 1 occupies

z > 1 and medium 2 occupies z < 1. Given E1 = 5âx − 2ây + 3âz , find E2 .

12
ELEC 325 Hamad M. Alkhoori Fall 2024

Ans:
ân = âz

ε1 ⊥
E⊥
1 = (E1
• ân )ân = 3âz ⇒ E⊥
2 = E = 4âz
ε2 1
∥ ∥ ∥
E1 = E1 − E⊥
1 = 5âx − 2ây ⇒ E2 = E1 = 5âx − 2ây

⇒ E2 = 5âx − 2ây + 4âz .

Example 8: Repeat the previous example when a surface charge ρfs = ε0 exists on the surface.

Ans:
∥ ∥
E2 = E1 = 5âx − 2ây

1 1 11
E⊥
2 = (ε1 E1⊥ − ρfs )âz = (12 − 1)âz = âz
ε2 3 3
11
⇒ E2 = 5âx − 2ây + âz .
3
Further Readings:

ˆ Section 5.9 in “Elements of Electromagnetics by Matthew”.

5.5 Conductors

In a dielectric medium, each atom contains electrons and protons that are attached to each other (see the

spring model in Fig. 5.1). In a conductor, by contrast, one or more electrons are free to move (these are

called free electrons). One consequence of this freedom is that a conductor can conduct a current (to be

discussed in Section 6.1). The mathematical consequence of this freedom is apparent in the form of the

electric susceptibility χe of a conductor.

In a perfect conductor, the number of these free electrons is infinite. The electric susceptibility χe is

directly proportional to the number of electrons. Because the number of electrons if infinite, it follows that

χe → ∞. From

P = ε0 χe E, (5.46)

because P has to be finite, we conclude that E = 0 inside a conductor. Physically, when an incident field

Einc exists in a conductor, the electrons respond by generating a perturbation field Epert that opposes the

incident field. Because the number of electrons is infinite, the cancellation of the incident field is perfect

(i.e., Epert = −Einc ), and the total field (E = Einc + Epert ) becomes zero.

13
ELEC 325 Hamad M. Alkhoori Fall 2024

A perfect conductor may not exist in real life, but metals display characters very similar to that of

a perfect conductor. Because E = 0 inside a perfect conductor, a perfect conductor has the following

properties:

1. ρv = 0, which follows from Gauss’s law.

2. The net charge resides on the surface of the conductor.

3. The potential difference between any two points on the surface is zero, hence, the perfect conductor is

an equipotential surface.

4. On the surface of the conductor from outside, the electric field is perpendicular (normal) to the surface.

Further Readings:

ˆ Section 2.5.1 in “Introduction to Electrodynamics by David”.

ˆ Section 5.4 in “Elements of Electromagnetics by Matthew”.

5.6 Capacitors

Suppose we have two conductors, one with a positive charge Q and the other with a negative charge −Q as

shown in Fig. 5.7.

Figure 5.7: Capacitor configuration.

14
ELEC 325 Hamad M. Alkhoori Fall 2024

The potential difference between these two conductors is given by

Z+
V0 = V+ − V− = − E • dl, (5.47)
−

where E is the electric field that exists in the space surrounding the two conductors. Note that V0 ∝ E.

From
Z
1 ρs R ′
E= dS , (5.48)
4πε S′ R3

and
Z
Q= ρs dS ′ , (5.49)
S′

one notices that E ∝ ρs and ρs ∝ Q. Hence, it follows that E ∝ Q, and thus, V0 ∝ Q. This proportionality

is described by the capacitance through

Q = CV0 , (5.50)

or
Q
C= . (5.51)
V0

Capacitance is a quantity that depends on the shapes of the conductors, as well as on the medium surrounding

them. We assume that the shapes of the two conductors are the same. Hence, a capacitor is merely two

conductors of the same shape, and the same magnitude of charge but with an opposite sign. It is used to

store charges. Higher capacitance means that more charges can be stored into the capacitor, and vice versa.

The unit of capacitance is Farad F .

Q
Although we can find capacitance using C = , we will do it in a different way. The energy is given by
Z V0
1 1
W = D • E dV , which we can write as W = V0 Q. Hence,
2 V∞ 2

Q2
C= . (5.52)
2W

Example 9: Consider the plane z = 0 and the plane z = d. Plane z = 0 carries ρs = −Q/A and plane z = d

carries ρs = Q/A. The region in between is filled by an isotropic dielectric medium of permittivity ε. Find

the capacitance for an area A of this configuration.

15
ELEC 325 Hamad M. Alkhoori Fall 2024

Ans:  
 



 0, −∞ < z < 0 


 0, −∞ < z < 0

 

D= Q ⇒E= Q
− âz , 0 < z < d − âz , 0 < z < d


 A 

 εA

 

 0,
 d<z<∞  0,
 d<z<∞

Q2 d
Z
ε
W = E 2 dV = .
2 V∞ 2εA
εA
⇒C=
d
Example 10: Two coaxial infinite cylinders of radii a and b (b > a) carry ρs = Q/2πda and ρs = −Q/2πdb,

respectively. The region in between is filled by an isotropic dielectric medium of permittivity ε. Find the

capacitance for a length d of this configuration.

Ans:  
 


 0, 0 < ρ < a 

 0, 0<ρ<a

 

Q Q
 
D= âρ , a < ρ < b ⇒ E = âρ , a < ρ < b


 2πdρ 

 2πdερ

 

 0, b<ρ<∞  0, b<ρ<∞
 

Zb  2
Q2
Z
ε 2 Q
W = E dV = επd ρdρ = ln(b/a)
2 V∞ 2πdερ 4πdε
a

2πεd
⇒C=
ln(b/a)

16
ELEC 325 Hamad M. Alkhoori Fall 2024

Example 11: Two concentric spheres of radii a and b (b > a) carry ρs = Q/4πa2 and ρs = −Q/4πb2 ,

respectively. The region in between is filled by an isotropic dielectric medium of permittivity ε. Find the

capacitance of this configuration.

Ans:  
 



 0, 0<r<a 


 0, 0<r<a

 

Q Q
D= â , a < r < b ⇒ E = 
2 r
â , a < r < b
2 r


 4πr 
 4επr

 

 0,
 b<r<∞  0,
 b<r<∞
Zb  2
Q2
Z  
ε 2 Q 2 1 1
W = E dV = ε2π r dr = −
2 V∞ 4επr2 8πε a b
a

4πε
⇒C=
1 1
−
a b
Further Readings:

ˆ Section 2.5.4 in “Introduction to Electrodynamics by David”.

ˆ Section 6.5 in “Elements of Electromagnetics by Matthew”.

17