The University of Queensland Department of Physics 2008

The University of Queensland
Department of Physics
2008
Lecture notes and tutorial problems of the
undergraduate course
PHYS3050/PHYS7051
ELECTROMAGNETIC THEORY III
Lecturer: Zbigniew Ficek
Physics Annexe(6): Rm 436
Ph: 3365 2331
email: cek@physics.uq.edu.au
http://www.physics.uq.edu.au/people/cek/
Consultation Hours: Wednesday 2pm 4pm
Preface
This lecture notes covers the principal elements of classical electromagnetic
theory embodying Maxwells equations with applications mainly to situations
where electric charge can be treated as a continuous uid. The intention is
to introduce students to the background of classical eld theory and the
applications of the electromagnetic theory to solid state physics, classical
optics, radiation theory and telecommunication.
The goal of this course is to provide a compact logical exposition of the
fundamentals of the electromagnetic theory and the applications to various
areas of physics and engineering. The treatment is quantitative throughout
and an attempt has been made to imbue students with a sound understanding
of the Maxwells equations and with the ability to apply them to modern
problems in physics.
The organization of the lectures is fairly standard and includes vector
analysis, electrostatic, magnetostatic, mathematical techniques in the solu-
tion of the Maxwells equations and the Laplaces equation, time varying
elds and applications of the solution of the Maxwells equations. The ma-
terial on vector analysis gives greater emphasis to the relationship between
elds and their sources.
A number of revision questions have been included at the end of each
chapter. These questions have been designed not only to point out to the
student the essential material of the chapter but also to test the students
understanding of the material presented in the chapter. In addition, tutorial
problems have been included at the end of the chapters that contain the
most important elements of the electromagnetic theory. These problems
will be discussed and solved in details during the tutorial sessions, but it
is hoped that the student will attack these problems before attending the
tutorial session.
2
Contents
1 The Classical Theory of the Electromagnetic Field 13
1.1 Elementary Aspects of Electromagnetism . . . . . . . . . . . . 13
1.2 Macroscopic Charges and Currents . . . . . . . . . . . . . . . 16
1.2.1 Charge Density . . . . . . . . . . . . . . . . . . . . . . 16
1.2.2 Current density . . . . . . . . . . . . . . . . . . . . . . 18
Revision questions . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2 Mathematical Description of Vector Fields 20
2.1 Gradient of a Scalar Function . . . . . . . . . . . . . . . . . . 20
2.2 Divergence Function . . . . . . . . . . . . . . . . . . . . . . . 22
2.3 The Flux of a Vector Field . . . . . . . . . . . . . . . . . . . . 24
2.4 Gausss Divergence Theorem . . . . . . . . . . . . . . . . . . . 24
2.5 The Continuity Equation for Electric Current . . . . . . . . . 25
2.6 Curl (Rotation) Function . . . . . . . . . . . . . . . . . . . . . 26
2.7 Stokess Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.8 Successive Application of . . . . . . . . . . . . . . . . . . . 28
Tutorial problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3 Vectorial Equations in Electromagnetism: Scalar and Vector
Potentials 34
3.1 Maxwells Equations, an Example of Vectorial Equations . . . 34
3.2 Scalar Potential . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.3 Vector Potential . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4 The Experimental Basis of the Development of Electromag-
netic Theory 41
4.1 Coulombs Law Force between Static Charges . . . . . . . . 41
4.2 Gausss Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.3 Biot-Savart Law Force between Static Currents . . . . . . . 49
4.4 Current Element and Charge Element . . . . . . . . . . . . . . 51
4.5 The Lorentz Force . . . . . . . . . . . . . . . . . . . . . . . . 52
4.6 Amperes Circuit Law . . . . . . . . . . . . . . . . . . . . . . . 53
4.7 Faradays Law of Electromagnetic Induction . . . . . . . . . . 55
3
5 Dierential Equations for the EM Field and Maxwells The-
ory 61
5.1 Dierential Equations for the EM Field . . . . . . . . . . . . . 61
5.1.1 Divergence of

E . . . . . . . . . . . . . . . . . . . . . . 62
5.1.2 Curl of

E . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.1.3 Divergence of

B . . . . . . . . . . . . . . . . . . . . . . 63
5.1.4 Curl of

B . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.2 The Maxwells Theory . . . . . . . . . . . . . . . . . . . . . . 64
6 Maxwells Equations and Prediction
of Electromagnetic Waves 69
6.1 The Wave Equation for EM Waves in Vacuum . . . . . . . . . 70
6.2 Plane Wave Solution to the Wave Equation . . . . . . . . . . . 71
6.3 Harmonic Waves . . . . . . . . . . . . . . . . . . . . . . . . . 73
6.4 The Transverse Nature of Plane Waves in Vacuum . . . . . . . 74
7 EM Theory and Einsteins Special Theory of Relativity 80
7.1 Lorentz Transformation Equations for Space and Time . . . . 81
7.2 Force Transformation Equations . . . . . . . . . . . . . . . . . 82
7.2.1 The Force between Two Charges Moving with Con-
stant Velocities . . . . . . . . . . . . . . . . . . . . . . 83
7.3 Electric Field Lines of a Moving Charge . . . . . . . . . . . . 87
7.4 Magnetic Field Lines of a Moving Charge . . . . . . . . . . . . 88
7.5 Invariance of the Maxwell equations under the Lorentz trans-
formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
7.5.1 Remark on the Electromagnetic Induction . . . . . . . 95
8 Energy in the Electromagnetic Field:
Poyntings Theorem 99
4
8.1 Rate of Doing Work by the Field on the Current Ohmic
Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
8.2 Electrostatic Field Energy Density . . . . . . . . . . . . . . . 102
8.3 Magnetostatic Field Energy Density . . . . . . . . . . . . . . . 103
8.4 Poynting Vector . . . . . . . . . . . . . . . . . . . . . . . . . . 103
8.5 Phase Relationships in EM Waves . . . . . . . . . . . . . . . . 105
8.6 Momentum Flux . . . . . . . . . . . . . . . . . . . . . . . . . 105
8.7 EM Energy Flow: Circuit versus Field Theory . . . . . . . . . 106
8.7.1 Energy Flow in a Resistive Wire . . . . . . . . . . . . . 106
8.7.2 Energy Flow out of Battery . . . . . . . . . . . . . . . 108
8.7.3 Propagation of an Electromagnetic Wave along a Wire 109
9 Solution of Laplaces Equation and Boundary Value Problem112
9.1 Uniqueness of the Solution of Laplaces Equation . . . . . . . 113
9.1.1 Dirichlet Theorem . . . . . . . . . . . . . . . . . . . . 113
9.2 Solutions of Laplaces Equation . . . . . . . . . . . . . . . . . 114
9.2.1 Method of Separation of Variables . . . . . . . . . . . . 115
9.2.2 Solution of the Laplace Equation in Spherical Coordi-
nates . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
10 General Solution of the Maxwells Equations 138
10.1 Diculty of the Direct Solution of Maxwells Equations . . . . 138
10.2 Scalar and Vector Potentials . . . . . . . . . . . . . . . . . . . 140
10.2.1 Lorenz Gauge . . . . . . . . . . . . . . . . . . . . . . . 142
10.2.2 Coulomb Gauge . . . . . . . . . . . . . . . . . . . . . . 143
10.3 Solution of the Inhomogeneous Wave Equations . . . . . . . . 145
10.4 Rigorous Solution: Green Functions Method . . . . . . . . . . 148
11 Electromagnetic Antennas: Hertzian Dipole 152
11.1 Generation of electromagnetic waves . . . . . . . . . . . . . . 154
11.1.1 Field of an Element of Alternating Current . . . . . . . 154
11.2 Power Radiated from the Current Element . . . . . . . . . . . 161
11.3 Gain of the Dipole Antenna . . . . . . . . . . . . . . . . . . . 163
5
12 Electromagnetic Theory of Polarizable
Materials 166
12.1 Potential and Electric Field of a Single Dipole . . . . . . . . . 167
12.2 Polarization Vector . . . . . . . . . . . . . . . . . . . . . . . . 168
12.3 Maxwells Equation for

E in a Dielectric . . . . . . . . . . 171
12.4 Macroscopic Eects of the Polarizability . . . . . . . . . . . . 173
12.5 Dense Dielectrics: The Clausius-Mossotti Relation . . . . . . . 176
12.6 Dielectric in a Time Dependent Field . . . . . . . . . . . . . . 178
12.7 The Complex Susceptibility and Permitivity . . . . . . . . . . 182
12.8 The Loss Tangent . . . . . . . . . . . . . . . . . . . . . . . . . 185
13 Electromagnetic Theory of Magnetizable Materials 190
13.1 Magnetic Polarization Currents . . . . . . . . . . . . . . . . . 190
13.2 The Magnetic Intensity Vector

H . . . . . . . . . . . . . . . . 195
13.2.1 Static Magnetic Fields . . . . . . . . . . . . . . . . . . 196
13.3 Linear Magnetic Materials . . . . . . . . . . . . . . . . . . . . 197
13.4 Non-Linear Magnetic Materials . . . . . . . . . . . . . . . . . 200
13.4.1 Work done in magnetization of iron . . . . . . . . . . . 201
13.5 Permanent Magnetic Materials: Ferromagnets . . . . . . . . . 202
13.5.1

B and

H elds of a ferromagnet . . . . . . . . . . . . . 203
13.5.2 Magnetic Poles . . . . . . . . . . . . . . . . . . . . . . 205
13.6 Time Dependent Magnetic Fields and Energy Loss . . . . . . 207
13.7 The complex magnetic susceptibility . . . . . . . . . . . . . . 208
13.8 Maxwells Equations in Dielectric and Magnetic Materials . . 209
14 Poyntings Theorem Revisited 214
14.1 Poynting Vector in Terms of

E and

H . . . . . . . . . . . . . 214
14.2 Poynting Vector for Complex Sinusoidal Fields . . . . . . . . . 216
6
15 Plane Wave Propagation in Dielectric and Magnetic Media 220
15.1 Dispersive Equation and Complex Propagation Number . . . . 222
15.2 Wave Refraction and Attenuation . . . . . . . . . . . . . . . . 225
15.2.1 Propagation of an EM wave in a low loss dielectric . . 225
15.2.2 Propagation of an EM Wave in a Good Conductor . . 227
15.2.3 Skin Eect . . . . . . . . . . . . . . . . . . . . . . . . . 229
16 Transitions Across Boundaries for Electromagnetic Fields 234
16.1 Normal Components of the Fields . . . . . . . . . . . . . . . . 234
16.1.1 Normal Component of

B . . . . . . . . . . . . . . . . . 234

H . . . . . . . . . . . . . . . . . 236

D and

E . . . . . . . . . . . . . 237
16.2 Tangential Components of the Fields . . . . . . . . . . . . . . 237
16.2.1 Tangential Component of

E . . . . . . . . . . . . . . . 238

H . . . . . . . . . . . . . . . 240
16.2.3 Tangential component of

B . . . . . . . . . . . . . . . 240
17 Propagation of an EM Wave Across a Boundary 244
17.1 Representation of Plane Waves in Dierent Directions . . . . . 245
17.1.1 Representation of

B in terms of

E . . . . . . . . . . . . 246
17.2 Directions of Reected and Transmitted Waves . . . . . . . . 247
17.3 Angle of Reection and Snells Law of Refraction . . . . . . . 252
18 Fresnels Equations 254
18.1

E
i
normal to plane of incidence . . . . . . . . . . . . . . . . . 255
18.2

E
i
in the plane of incidence . . . . . . . . . . . . . . . . . . . 257
18.3 Fresnel Equations for dielectric media . . . . . . . . . . . . . . 258
19 Applications of the Boundary Conditions and the Fresnel
Equations 262
19.1 Applications in dielectrics . . . . . . . . . . . . . . . . . . . . 262
7
19.1.1 Polarization by reection . . . . . . . . . . . . . . . . . 262
19.1.2 Total internal reection . . . . . . . . . . . . . . . . . . 264
19.2 Transmission and Reection at a Conducting
Surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
19.2.1 Field vectors at normal incidence . . . . . . . . . . . . 269
20 Propagation of an EM Wave in a Rectangular Waveguide 273
20.1 Transverse Electric (TE) Modes . . . . . . . . . . . . . . . . . 275
20.2 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . 278
20.3 TE Modes in a Lossless Waveguide . . . . . . . . . . . . . . . 280
20.4 Phase and Group Velocities of Mode Propagation . . . . . . . 283
21 Relativistic Transformation of the Electromagnetic Field 287
21.1 The Principle of Relativity . . . . . . . . . . . . . . . . . . . . 287
21.2 Transformation of Electric and Magnetic Field Components . 293
21.3 Transformation Rules in Terms of Parallel and Normal Com-
ponents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
21.3.1 Rules for Parallel Components . . . . . . . . . . . . . . 295
21.3.2 Rules for Normal Components . . . . . . . . . . . . . . 295
21.4 Transformation of the Components
of a Plane EM Wave . . . . . . . . . . . . . . . . . . . . . . . 297
21.5 Doppler Eect . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
21.6 Transformation of Energy of a Plane EM Wave . . . . . . . . 301
Revision Questions for the Final Examination 306
Appendix A: Proof of the Amperes law 312
Appendix B: Proof of the vector theorem, Eq. (73) 314
Appendix C: PHYS3050 Facts and Formulae 315
8
Literature
1. J.D. Jackson, Classical Electrodynamics, 3rd ed. Wiley 1999.
The course is aimed at this level of treatment.
2. R.K. Wangsness, Electromagnetic Fields, 2nd ed. Wiley 1986.
This is an alternative textbook for the course.
3. R. Plonsey and R.E. Collin, Principles and Applications of Elec-
tromagnetic Fields, McGraw Hill 1961.
This is a good general text on electromagnetic elds.
4. B.I. Bleaney and B. Bleaney, Electricity and Magnetism, 3rd ed.
Oxford U.P. 1983.
A very readable book ranging over circuit theory, electronics, electric
properties of matter and eld theory.
5. J.A. Stratton, Electromagnetic Theory, McGraw Hill, 1941.
This is a very comprehensive reference on electromagnetic theory.
6. W.K.H. Panofsky and M. Phillips, Classical Electricity and Mag-
netism, 2nd ed. Addison-Wesley 1962.
This is a good text on the relativistic representation of electromag-
netism.
7. R.P. Feynman, R.B. Leighton, and M. Sands, The Feynaman
Lectures on Physics, Vol. 2, Addison Wesley, 1964.
A good reference to the physics of the emission of radiation.
9
To anyone who is motivated by anything
beyond the most narrowly practical,
it is worthwhile to understand Maxwells equations
simply for the good of his soul
J.R. Pierce
11
1 The Classical Theory of the Electromag-
netic Field
Classical theory of the electromagnetic eld or Classical Electrodynamics,
formulated by Maxwell more than 100 years ago, is now a well established
theory with many applications in dierent areas of physics, chemistry and en-
gineering. In this context classical means non-quantum, but we would
like to point out that the basic equations of electromagnetism, the Maxwells
equations, hold equally in quantum and classical eld theory.
We begin the study of the electromagnetic theory with a brief review of the
elementary aspects of electromagnetism that the student learnt in the rst
year PHYS1002 course. These aspects are in fact about the microscopic
nature of the electromagnetic theory. To indicate the classical approach to
the electromagnetic theory and the macroscopic nature of electromagnetic
phenomena we are going to explore in this lecture notes, we rst introduce
the concept of macroscopic distributions of charges and currents, i.e. we
dene volume, surface and linear charge densities. Next, we will introduce
the concept of the current density and a total (macroscopic) current.
1.1 Elementary Aspects of Electromagnetism
In many textbooks on electromagnetism, the elementary aspects of electro-
magnetism are introduced as postulates. In fact, they are based on experi-
mental observations, and we will explore this fact in full details in one of the
lectures. At this introductory lecture, let us review briey the elementary
aspects of electromagnetism that are:
1. Electromagnetic interactions are ONE of FOUR fundamental types:
1
Type of interaction Relative Strength
Strong interaction (nuclear) 1
Electromagnetic 10
2
Weak interaction (e.g. decay) 10
12
Gravitation 10
40
1
At our level of discussion there is no relation between these four types of interaction,
i.e. they cannot be considered as dierent manifestations of a single FORCE.
13
2. Electromagnetic (EM) forces or interactions are due to ELECTRIC
CHARGE, which is NOT in turn explicable in terms of anything else.
3. Charges are of two kinds called positive and negative. In the static
limit like charges repel and unlike attract.
4. Charges are quantized in units of e 1.6 10
19
[Coulombs].
5. In the static limit the inverse square (Coulomb) law of force holds:
F
q
2
=
1
4
0
q
2
q
1
r
2
r ,
r
q
1
q
2
,
that the charge q
1
acts on the charge q
2
with the force

F
q
2
. The force
acts along the distance r and the parameter
0
determines the property
of the medium and is called the electric permittivity.
The Coulombs law holds for microscopic charges, i.e. it holds
only for charges whose the spatial dimensions are small com-
pared with the distance separating them.
The Coulombs law does not tell us how the rst charge knows the
other one is present. One usually assumes that the charge produces an
electric eld which then interacts with the other charges. To express
this explicitly, the Coulomb force is often written as
F
q
2
= q
2
E
q
1
,
where

E
q
1
is the electric eld produced by the charge q
1
.
The Coulomb law can be tested to great accuracy indirectly by showing
that no charge rests on the inside of a statically charged hollow con-
ductor.
2
2
S.J. Plimpton and W.E. Lawton, Phys. Rev. 50, 1066 (1936).
14
If the exponent in the Coulombs law for point charges
E
1
r
n
is not n = 2 but n = 2(1), the potential inside the hollow conductor
would be large. Since the potential inside the hollow conductor found
by the experiments was less in magnitude than a small detectable po-
tential, then 10
9
, the level of sensitivity of the detector.
6. Electric charge is conserved (algebraically)
whole universe
q = constant
7. In the NON-STATIC case, i.e. the case of moving charges, the force is
no longer given by Coulombs Law. It is given by the Lorentz equation
F
q
2
= q
2
(
E +v

B) ,
where

E and

B are the electric and magnetic elds due to q
1
, and v is
the velocity of the charge q
2
.
8. In the electromagnetic theory, we assume that the elds

E and

B de-
pend on the frame of reference of the observer, that

E,

B and the
force

F must follow the required relativistic transformation law.
However, we do not think that q depends on the frame of reference.
i.e. q does not depend on its velocity with respect to an observer. In
terms of the relativistic theory, the charge is invariant under the Lorentz
transformation.
3
The student has noticed that the Lorentz force involves both, the electric and
magnetic elds. Why there must be a

B and how

E and

B are computed for
3
This is because in ordinary matter electrons move much faster than ions, their speeds
depend on temperature, and electric elds are not observed to arise from changes in
temperature.
15
arbitrary motion of charges is the substance of electromagnetic theory.
We now have the following picture of the basis of the electromagnetic theory:
1. CHARGE 1 =
ELECTROMAGNETIC
FIELD =
FORCE ON
CHARGE 2
2. Fields are generated by charges - NOT by other elds.
1.2 Macroscopic Charges and Currents
We know from the Millikan experiment that electric charge is quantized. The
electron is a point charge on the smallest scale measurable. We may then
speak, on a subatomic scale, of a microscopic theory of electromagnetism. On
a subatomic scale there must be very strong and rapidly varying electric and
magnetic elds on spatial scales 10
8
m and temporal scales 10
10
s.
When we measure the elds around a macroscopic circuit, clearly we are
not looking at these elds. We are measuring elds on distance scales much
larger than 10
8
m and time scales much longer than 10
10
s.
In the macroscopic context, we do not distinguish individual charges. It is
convenient and justiable to regard the charge as a continuous uid dis-
tributed over the volume or surface of a charged material.
Thus, we may introduce the concept of macroscopic charges and currents to
indicate the macroscopic nature of electromagnetic phenomena.
1.2.1 Charge Density
Charges may be distributed throughout the volume of a material, or on the
surface of the material, or may be distributed along one dimensional wires.
Thus, there are dierent spatial distributions of charges considered in the
electromagnetic theory that result in three types of charge densities: volume
charge density, surface charge density and linear charge density.
16
When we encounter a large number of point charges in a nite volume, it
is convenient to describe the source in terms of a volume charge density,
dened as
= lim
q
V
,
where q is the algebraic sum of the charge in the volume V .
The limit is not to zero, but to a V much larger than atomic scale size,
which is still very small on the laboratory scale.
If the volume charge density is represented by a continuous function (r),
the total charge Q in a volume V is given by
Q =
V
(r) dV .
Note that the charge density is a function of the position which varies smoothly
inside a charged material. An exception is a boundary between two charged
materials where may change discontinuously due to the presence of surface
charges of a non-zero density. Thus, we may introduce the concept of surface
charge density.
A surface charge density is dened, analogously to the volume charge den-
sity, as
= lim
q
S
,
where q is the algebraic sum of the charge on the surface S.
Then, the total charge continously distributed on a surface S is
Q
S
=
S
(r) dS .
In the same fasion, we may dene a linear charge density on a length L
= lim
q
L
,
and then the total linear charge continously distributed on a length L is
Q
L
=
L
dL .
17
1.2.2 Current density
For many purposes in the electromagnetic theory, it will be necessary to
introduce the concept of current density.
Figure 1: Illustration of the current
ow through an area A and the evalua-
tion of the current density through the
area A.
It measures the amount of current owing
through an area normal to the direction
of the current.
We dene the current density in the fol-
lowing way. Let I
A
= q/t is a current
through the area A, as shown in Figure 1.
Then the current density, normal to the
area A is dened by
J = lim
I
A
n = lim
q
t A
n
= lim
Av t
t A
n = v ,
where v = v n, and n is the unit vector
normal to the area A, and the limit is
taken in the same sense as for .
Total current through an arbitrary surface area
For a surface of an arbitrary shape, the current may not be normal to the
surface at all points on the surface.
Figure 2: Illustration of evaluation of the
current density through an area A. The
vector n is a unit vector normal to the sur-
face element dA.
However, if the current density

J is
known at all very small areas, that
can be approximated by at surfaces of
a macroscopic surface A, we still can
obtain the total current through the
area A.
Consider a current density

J through
a macroscopic area A, as shown in
Figure 2. To nd the total current
through the area, we rst divide the
area A into a lot of small at areas dA,
18
and nd the current through dA as
4
I =

J ndA = J cos()dA .
Then the current through the total area A is the sum of the contributions
from all the small elements of the area:
I
A
=
I =
J n dA =
J d
A ,
where d
A = dA n is a vector representing the element dA of the surface A.

Note: In vector analysis it is common to represent a surface by a vector
whose length corresponds to the magnitude of the surface area and whose
the direction is specied by the unit vector n normal to the surface.
In summary, Electromagnetic theory formulated in terms of space depen-
dent quantities, charge densities , or , and current density

J is regarded
as a MACROSCOPIC THEORY.
Revision questions
Question 1. What are the elementary aspects of electromagnetism?
Question 2. How do we dene volume, surface and linear charge densities?
Question 3. How do we dene and evaluate the current density through an area?
Question 4. If the current density through an area is known, how does the
current through the area is evaluated?
4
In literature, very often small areas that a macroscopic area is divided to are called
elementary areas or elementary surfaces.
19
2 Mathematical Description of Vector Fields
The study of electromagnetic theory requires considerable knowledge of vec-
tor analysis. In this lecture, we will introduce vector operations such as
gradient, divergence and curl that we will need for our study of electromag-
netic theory. As we shall see, these vector operations are very convenient to
determine the properties of electromagnetic eld, also considerably simplify
the formulation of electromagnetic theory and allow get a better inside into
electromagnetic phenomena.
2.1 Gradient of a Scalar Function
Let us rst dene a vector operation: Gradient of a scalar function. Assume
that a function represents a scalar eld and that is a single valued, con-
tinuous, and dierentiable function of position. Let d represents a change
of with a distance ds.
Figure 3: Illustration of the
evaluation of gradient of a scalar
function .
The gradient of the scalar function is
dened as:
grad =

s
n ,
where n is a unit vector in the direction
the rate /s has its maximum value. In
other words, gradient tells us in which
direction the change in is maxi-
mal.
For some other direction d
X, a change of can
be found by projection of gradient of on d
X:
d = d
X =

s
n d
X =

s
cos dX .
We know from the vector analysis that it is convenient to represent a vec-
tor in a reference (coordinate) frame. Commonly used are the rectangular
(cartesian) coordinates, in which we can easily nd that the x component of
20
the gradient is
()
x
=
i =

s
n
i =

s
cos
= lim

(s/ cos )
= lim

x
=

x
.
Similarly, the y and z components of the gradient are
()
y
=

y
, ()
z
=

z
.
Hence, in cartesian coordinates, the gradient of a scalar function can be
written as
5
=

x
i +

y
j +

z
k .
The gradient is analogous to multiplication of a vector by a scalar. The re-
sult, of course, is a vector.
6
Remember: Mathematically, gradient of an arbitrary scalar function is a
vector that is pointing in the direction of maximal changes of the function.
Exercise in class:
Consider a scalar function = x. Calculate the gradient
of = x.
Solution: Since
x
= 1 and

y
=

z
= 0 ,
we nd that
=
x
x
i +
x
y
j +
x
z
k =
i .
This exercise explicitly shows that gradient is a vector pointing in the direc-
tion of maximal changes in .
5
Expressions for the gradient in other coordinate systems are given in the Appendix C.
6
We do not usually take a gradient of a vector, the result would be a tensor.
21
The student should be able now to give a quick answer to simple questions
as Is gradient a scalar or a vector? or In which direction does gradient of
a given scalar function point? A practical test: Try to give the answer to
the following question without any calculations: What is the direction of the
gradient of a scalar function = x + y?
2.2 Divergence Function
We have already dened the nabla operator and showed how it evaluates a
certain property of a scalar function. We now dene the divergence function,
which involves the nabla operator and, as we shall see, tells us about the ow
of a vector eld and its sources.
Denition of divergence:
The divergence is the scalar function which results from operation
of upon a vector

F in a fashion analogous to the dot product of
two vectors. The result is a scalar function.
In Cartesian coordinates, the divergence function is written as
7
div

F

F =
F
x
x
+
F
y
y
+
F
z
z
.
Properties: A positive divergence means that there is a source of a vector
eld, and a negative divergence means the presence of a sink of the eld.
When

F = 0 everywhere, the eld

F is called solenoidal, since no start-
ing points or sources can be assigned to the lines describing the eld. In
other words, it has no sources or sinks.
Things to remember: Mathematically, divergence of an arbitrary vector
is a scalar. Physically, a non-zero

F implies a source (if positive) or a
sink (if negative), and if

F = 0, there is no source or sink the eld lines
have no beginnings or ends.
7
The forms that the divergence takes in other coordinate systems are given in the
Appendix C.
22
Exercise in class:
Consider two vectorial elds

F
1
= x
i and

F
2
= y
i. Which of
the elds is solenoidal?
Solution: A vector eld is solenoidal when

F = 0. Thus,
we calculate the divergence of the elds
div

F
1
=
x
x
(
i) +
x
y
(
i) +
x
z
(
i) = 1 .
and
div

F
2
=
y
x
(
i) +
y
y
(
i) +
y
z
(
i) = 0 ,
from which we see that the eld

F
2
= y
i is solenoidal.
This exercise shows that divergence of a given vector eld is dierent from
zero only if the eld amplitude changes in the direction of the eld. So now it
is easy to visualize the divergence of a vector eld. The divergence is related
to how the eld changes as you move in the direction of the eld.
One more example for a better understanding of the meaning of divergence.
Consider a vector r = x
i +y
j +z
k that represents the position of an arbitrary

point in the Cartesian coordinates. The student can easily nd that divr = 3.
What does the number 3 mean? It means that the eld changes by one unit in
each direction. However, this statement may not be generally true. For example,
divergence of a vector

F = 2x
i + y
j is also equal to 3, but in this case the eld

changes by two units in the x direction and one unit in the y direction.
23
2.3 The Flux of a Vector Field
A vector eld propagating in space may cross some surfaces not necessary
normal to the eld direction.
Figure 4: A vector eld

F lines
crossing a surface S.
In this case, we may speak about a ux of the
eld through the surface. The ux is measured
by the number of eld lines crossing the sur-
face.
Mathematically, the ux of a eld

F through
a surface S, not necessary a closed S, as shown
in Figure 4, is dened as:
=
F d
S =
F ndS ,
where n is the unit vector normal to the surface at the point where the ele-
ment of area dS is located. Thus, we calculate the ux through a surface S of
arbitrary shape by dividing it up into a lot of small surfaces dS and calculate
the ux through each of these small surfaces. next, we add (integrate) the
uxes to obtain the total ux through the surface S.
2.4 Gausss Divergence Theorem
We now introduce an important identity in vector analysis, the Gausss di-
vergence theorem or Gausss law, which we will use frequently in our study
of the theory of the electromagnetic elds.
Consider a vector eld

F crossing a closed surface bounding a volume V , as
shown in Figure 5.
Gausss law, or sometimes called as the Gausss divergence theorem, says that

F dV =
F ndS . (1)
The Gausss law states that the volume integral of the divergence
of a vector eld over a volume V is equal to the closed surface in-
tegral of the vector over the surface bounding the volume V .
24
Figure 5: Illustration of the eval-
uation of the Gausss law for a vec-
tor eld

F crossing a closed sur-
face bounding a volume V .
Mathematically, the Gausss divergence the-
orem converts a volume integral of the di-
vergence of a vector to a closed surface inte-
gral of the vector, and vice versa. Physically,
the Gausss divergence theorem says that the
number of the eld lines owing through
the surface S is equal to the strength
of the eld source contained inside the vol-
ume V .
Remember that n is the unit outward
normal over S, as shown in Figure 5,
and S is a closed surface bounding a vol-
ume V .
Associated with this is the pictorial representation of elds by lines of force,
the direction of

F given by the tangent to a line and the strength of

F given
by the line density per unit area.
Note that from the divergence theorem

F = lim
dV 0
1
dV

F ndS
= lim
dV 0

dV
,
i.e. the divergence of a vector eld is the emanating ux per unit volume.
2.5 The Continuity Equation for Electric Current
The Gausss divergence theorem allows us to derive the continuity equation
for the electric current known as the equation of conservation of the charge.
Suppose we have some charge of density in a volume V enclosed by a sur-
face S, as shown in Figure 6. Let v is a macroscopic velocity of the charge.
Then, the rate of decrease of the total charge in the volume V is equal to the
rate of transport of the charge out through the surface S.
Mathematically, we can express this statement as
V
dV =
S
v ndS .
25
Figure 6: Illustration of a ow of
a charge through a surface S closing
a volume V .
Using the Gauss divergence theorem, we
can write
S
v ndS =
V
(v) dV .
Thus
t
dV =
V
(v) dV .
However, this relation holds for arbitrary V ,
so we can drop the integrals and obtain
t
= (v) .
Since
v =

J ,
we nally obtain
t
+

J = 0 ,
which is well known as the continuity equation, or the equation of conser-
vation of charge.
When stationary currents are involved, then /t = 0. In this case
J = 0,
that is for stationary currents the current density is solenoidal.
2.6 Curl (Rotation) Function
Another very important property of vector elds is curl or rotation or cir-
culation. Curl is the operation of operator upon a vector in a fashion
analogous to the cross product of two vectors. The result is a vector that in
Cartesian coordinates is written as
8
curl

F

F =
F
x
+

j

F
y
+

k

F
z
,
8
Expressions for the curl in other coordinate systems are given in the Appendix C.
26
or

F =
F
z
y

F
y
z
i +
F
x
z

F
z
x
j +
F
y
x

F
x
y
k
=
i

j

k
z
F
x
F
y
F
z
.
Properties: Curl is nonzero when the eld increases (or decreases) in a dif-
ferent direction that the eld pointed. If the eld is pointed in the same
direction as that in which is increased, the curl is zero. So the curl is related
to how the eld changes as you move across the eld.
When

F = 0 everywhere, the eld

F is called irrotational.
Remember: Mathematically, curl of an arbitrary vector is a vector.
Exercise in class:
Find which of the following vectorial elds is irrotational.
F
1
=
r
r
2
,

F
2
= 2r ,

F
3
= r
3
,
where r = [r[ = 0 and r is the unit vector in the direction
of r.
2.7 Stokess Theorem
The curl function allows us to introduce an another important identity in
the theory of vectorial elds, the Stokess theorem that is formulated as
F d
l =

F ndS ,
where

dl is the innitesimal vector length tangent to a closed path of length l
bounding a surface area S.
27
Figure 7: Illustration of the evaluation of
the Stokess Theorem.
The Stokess theorem states that
the closed line integral of a
vector eld

F along the con-
tour bounding an open sur-
face S is equal to the sur-
face integral of the curl of
the vector eld over the sur-
face.
We may write the Stokess theorem in
the form

F n = lim
dS0
1
dS

F

dl
.
This gives an intuitive meaning to any component of

F in terms of the
line integral around a small element of surface, as shown in Figure 7. Simply,

F is a measure of the vorticity of the eld.
2.8 Successive Application of
We can introduce scalar and vector products in which the operator ap-
pears more than once. For example, since the gradient of an arbitrary scalar
function is a vector, we can take the divergence of the gradient
=
i +

y
j +

z
=

2
x
2
+

2
y
2
+

2
z
2
.
The same result is obtained if we take as a new operator
2
with
properties
2
=

2
x
2
+

2
y
2
+

2
z
2
.
The operator
2
is called the Laplacian and is a scalar.
The Laplacian may also be applied to a vector, with the result
F =

2
F
x
2
+

2
F
y
2
+

2
F
z
2
.
28
It is also possible to form the curl of the gradient, which is identically zero
= 0 . (2)
The divergence of the curl of a vector is also identically zero
9

F = 0 . (3)
These two properties are very useful in the vector eld theory, in particular in
the electromagnetic theory. The relation (2) shows that an irrotational eld
can always be expressed as gradient of an arbitrary scalar eld. Similarly,
relation (3) shows that any solenoidal eld can always be expressed as a curl
of an arbitrary vector eld. Further discussion of the successful application
of is left to the tutorial problems.
9
The proof of this and the above property is left to the student as a tutorial problem.
Note that each of these properties involves two vector operations, but in writing them
we usually do not put brackets that would indicate in which order the two operations
should be applied. Is it clear to the student in which order the vector operations should
be applied?
29
Revision questions
Question 1. What is the physical meaning of the gradient of a scalar function?
Question 2. Derive the explicit form of in the Cartesian coordinates.
Question 3. What is the physical meaning of the divergence of a vector?
Question 4. What is the physical meaning of the curl of a vector?
Question 5. Write the curl of an arbitrary vector in the Cartesian coordinates.
Question 6. Explain, without using any mathematics, why

F = 0.
Tutorial problems
Problem 2.1 Direction of a vector
Given a vector

A =
i + 2
j 2
k in Cartesian coordinates, nd
the expression for the unit vector

A in the direction of

A.
Problem 2.2 Relation between two vectors
Show that, if

A

B =

A

C and

A

B =

A

C, where

A is
not a null vector (
A =
0), then

B =

C.
Problem 2.3 Multiple applications of and
Consider arbitrary vectors

A,

B,

C and

D.
(a) Is

A(
B

C) equal to (
C

B)

A? Explain.
30
(b) Is

A (
B

C) equal to (
C

B)

A? Explain.
(c) Does

A

B =

A

C implies

B =

C? Explain.
(d) Does

A =

B implies

A =

B? Explain.
(e) Which of the following expressions do not make sense? Ex-
plain.
A

B/[
B[,

C

D/(
A

B),

A

B/(
C

D),
A

B

C, (
A),
2
(
B

B),
2
(
A

B), (
A

B) .
(f ) Which of the following expressions are vectors and which are
scalars?
(

A), (
A )
B, (
A

B),
2
(
A

B),
A (
B

C),

A[
B (
C

D)],
2
(
A

B),
(
A

B).
Problem 2.4 Vector components and directions
(a) Find the relative position vector

R of the point P(2, 2, 3)
with respect to Q(3, 1, 4). What are the direction angles of

R?
(The direction angles are the angles between

R and the x, y and z
axes repectively).
(b) Given the two vectors

A =
i +2
j +3
k and

B = 4
i 5
j +6
k,
nd the angle between them. Find the component of

A in the
direction of

B.
(c) Given the vectors

A = 2
i +3
j 4
k and

B = 6
i 4
j +
k nd
the component of

A

B along the direction of

C =
j +

k.
31
Problem 2.5 Flux of a vector eld
Figure 8: Contour for evaluation of the ux
of a vector eld.
(a) Given the vector eld

A =
xy
i + yz
j + zx
k, evaluate directly
from the denition the ux of

A
through the surface of a rectan-
gular parallelepiped of sides a, b, c
with the origin at one corner and
edges along the positive directions
of the rectangular axes, see Fig-
ure 8.
(b) Also evaluate

AdV directly
and show that it is equal to the cal-
culated ux as predicted by Gausss
Theorem.
Problem 2.6 Component in the direction of a gradient
(a) A family of hyperbolas in the xy plane is given by u = xy.
Find u.
(b) Given the vector

A = 3
i + 2
j + 4
k, nd the component of

A
in the direction of u at the point on the curve for which u = 3
and x = 2.
Problem 2.7 Unit normal to ellipsoids
The equation giving a family of ellipsoids is:
u =
x
2
a
2
+
y
2
b
2
+
z
2
c
2
.
Find the unit vector normal to each point of the surface of these
ellipsoids.
Problem 2.8 Divergence of a radial eld
For elds of the form r
n
r, (r = 0), nd for which values of n
the divergence is zero.
32
Problem 2.9 Field of a cylindrical form
For elds of the cylindrical form
n
, ( = 0), nd for which

values of n the curl is zero.
Problem 2.10 Line integral of a vector
The vector eld
A = x
2
y
i +xy
2
j +a
3
e
y
cos x
k ,
where a, , are constants.
Figure 9: Contour for evaluation of the curl
of a vector.
Evaluate directly the line integral of

A
around the closed path in the xy plane
as shown in the Figure 9. The straight
portions are parallel to the axes and
the curved portion is the parabola
y
2
= kx, where k = constant. Eval-
uate the surface integral of
A over
the area S enclosed by C. Show that
the two are related as expected by
Stokes Theorem.
33
3 Vectorial Equations in Electromagnetism:
Scalar and Vector Potentials
In this lecture, we will illustrate the application of the Gausss and Stockess
theorems to vectorial equations involving and operations. These type
of operations are involved in the Maxwells equations, the basic equations for
the electromagnetic theory. We shall discuss only the vectorial nature of
the equations, deferring a detailed physical interpretation of the equations
to the next chapter when we consider the experimental basis for electromag-
netism. Furthermore, we illustrate the Helmholtz Theorem for the unique
determination of a given vector and discuss some properties of the successive
application of , which will allow us to introduce the concept of vector and
scalar potentials to the electromagnetic eld theory.
3.1 Maxwells Equations, an Example of Vectorial Equa-
tions
Consider the Maxwells equations in the dierential form that are an example
of dierential vectorial equations involving and operations. These
dierential equations are for the vector elds

E and

B and are of the form

E =

0
, (4)

B = 0 , (5)

E =
B
t
, (6)

B =
0
J +
1
c
2
E
t
, (7)
These four equations determine and of the two vectors

E and

B,
and we need both operations for each of the EM eld vectors to completely
determine

E and

B.
An obvious question arises: Why do we need both and equations for
each of the vectors to completely determine

E and

B?
34
The answer is in the Helmholtz Theorem, which says that an arbitrary
vector

F can always be written as
F =
1
4

F
r
dV +
1
4

F
r
dV
=

F
l
+

F
t
.
In other words, if the divergence and curl of a vector eld are known every-
where in a nite region, then the vector eld can be found uniquely. Thus,
specication of

F and

F is necessary and sucient to determine

F.
Hence, we need four equations of and type to determine

E and

B.
Since

F can be found from the knowledge of and , the divergence and
curl are often called the sources of the eld.
3.2 Scalar Potential
The electromagnetic eld equations, the Maxwells equations, are coupled dif-
ferential equations whose the solution is not in general simple and straight-
forward. Even the calculation of electric eld produced by a macroscopic
charge from the Coulombs law is usually a dicult task. Often the solutions
are aided by the use of potentials.
What do we mean by a potential and how helpful it is in the calculation of
the elds?
The idea is that the expression of the elds in terms of potentials simpli-
es the equations to a form that can be readily solved. In particular, the
Maxwells equations are simplied to a form that can be solved in a reason-
ably simple way.
A potential is a quantity from which a vector eld can be derived by some
process of dierentiation. To show this, consider a eld with zero rotation
or curl,

F = 0. Since 0, then the equation

F = 0 will
have an integral of the form:
F = .
35
Thus, the eld with zero curl may be derived from the gradient of the scalar
function . In the eld theory, the function is called a scalar potential.
The scalar potential function is very often used in the electromagnetic eld
theory. For example, the electrostatic (Coulomb) eld has zero curl,
E =
0. Hence, we can always write the electrostatic eld as
10
E = .
The minus sign is inserted in the denition of to agree with the denition
of as a potential ENERGY. The negative sign can also be understood phys-
ically from the fact that

E is in the direction that a positive charge moves,
hence in the direction of decreasing potential.
Figure 10: Illustration of a path
along which a unit charge is moved
from a point A to a point B in an
electric eld

E.
The scalar potential is useful in the calcu-
lations of work done on charges moving in
an electric eld.
The work done by the eld in moving the
unit charge q from a point A to a point B
in an electric eld

E, as illustrated in Fig-
ure 10, is given by
W
q
=
B
E d
l =
B
A
d
l
=
B
r
r d
l =
B
A
d
=
B
A
=
AB
.
We see that the work done on moving the
charge from the point A to the point B
is equal to the potential dierence between
the two points.
The work is independent of the path, it depends only on the position of the
start and end points. If the charge gains energy in moving from A to B, we
10
As one can see from the Maxwells equations,
E = 0 in general in electromagnetism,
so it is not in general possible to write

E = .
36
say that the point B is at higher potential than the point A, and vice versa.
Note an another interesting property of the eld with

E = 0.
Figure 11: Two arbitrary paths leading
from a point A to a point B.
From the Stokess law we have that

E d
l =

E ndS = 0 .
It says that no work is done by the eld
on a test charge if it is moved along
an arbitrary closed path in the eld.
Consequently, the work along two ar-
bitrary paths from A to B, as illus-
trated in Figure 11, is
(AB)
1
=
(BA)
2
=
(AB)
2
.
In summary, the work done by the eld is independent of the path chosen.
In other words, if a charge is moved around any closed path, no net energy
is requred. Thus, a eld

F with

F = 0 is a conservative eld of force.
3.3 Vector Potential
An another useful potential in the electromagnetic theory is the vector po-
tential. To illustrate the idea of the vector potential and how it is dened,
consider a eld with zero divergence,

F = 0. Since for an arbitrary
vector

A:

A 0 ,
we have that the vector

F can always be written as
F =

A .
Thus, the eld with zero divergence may be derived from the curl of the
vector eld

A. In the eld theory, the vector

A is called a vector potential.
We shall see that

B = 0 always in electromagnetism so there will always
be a vector eld

A such that

B =

A and such an

A is referred to as
37
the vector potential in electromagnetism (though there may be other elec-
tromagnetic eld functions with zero divergence).
Note also that just writing

A =

B does not completely specify

A even
if

B is known everywhere. According to the Helmholtz Theorem, one needs
to specify

A as well to completely determine

A. Equivalently, we can say
that dening

A =

B still leaves us free to dene

A.
Exercise in class: Applications of and .
(a) For a given scalar function and vectors

A,

B,

C, in-
dicate successive steps you would follow in the calculation
of the following expressions:

A ;

A
B ;

A

B ;

A

B ;
A

B

C ;

A

B ;

A

B ;

A

B

C ;

A ;

A
B

C .
(b) Which of the expressions in part (a) are vectors?
38
Revision questions
Question 1. Explain why in electromagnetism we need two divergence and two
curl equations for the electric

E and magnetic

B elds to com-
pletely determine the elds

E and

B?
Question 2. What is the Helmholtzs theorem telling us about a vector

F?
Question 3. Why are the scalar and vector potentials useful in electromagnetism?
Question 4. Explain, which vectors can be expressed in terms of a scalar poten-
tial and which can be expresses in terms of a vector potential.
Question 5. Explain, when a eld

F is a conservative eld of force?
39
Tutorial problems
Problem 3.1 If the potential satises the Laplace equation,
2
= 0, show
that the vector is both solenoidal and irrotational.
Problem 3.2 Show from the form of Maxwells equations that it should be
possible to dene electromagnetic potentials

A and such that
the elds can be calculated from
E =

A
t
,
B =

A .
Problem 3.3 In an application of the electromagnetic theory to nuclear physics,
the so-called vectorial mesons are described by three vectorial elds

E,

B,

A
and a real scalar eld . The elds satisfy the following equations

E =
2
,

B =

E
t

2

A ,
E =

A
t
,
where
2
is a positive constant.
Show that the above equations can be converted into two dieren-
tial equations
(i)

A+

t
= 0 ,
(ii)
2

2

2
t
2
= 0 .
40
4 The Experimental Basis of the Develop-
ment of Electromagnetic Theory
In this lecture, we will analyse experiments that led to the development of
the electromagnetic theory. We will show that the Coulomb Law for the elec-
trostatic force between two point charges is the experimental basis for the
development of the electric theory and the Biot-Savart Law for the magnetic
force between two static currents is the experimental basis for the develop-
ment of the magnetic theory. We also illustrate two useful laws of electro-
magnetism, the Gausss and Amperes Circuit laws, that are very helpful in
calculations of the electric and magnetic elds produced by some symmetric
distributions of charges and currents.
4.1 Coulombs Law Force between Static Charges
In 1785, Coulomb investigated the nature of the force between charged bod-
ies, and the results of his experiments can be formulated mathematically in
what is known as Coulombs law
F
2
=
1
4
0
q q
r
2
r .
We can write Coulombs law in terms of the electric eld produced by the
charge q that acts on the charge q
F
q
= q
E ,
where
E =
1
4
0
q
r
2
r
is the electric eld produced by the charge q. The electric eld is an example
of a vector eld. In principle, we can always calculate an electric eld using
Coulombs law. However, there is an alternative way we can nd the electric
eld. In particular, the eld may be represented by means of the ux con-
cept. The total ux of

E from a point charge q may be readily calculated by
integrating

E d
S over a surface enclosing q.

41
4.2 Gausss Law
Consider a macroscopic charge q closed by a surface S, as shown in Figure 12.
We will show that the ux of the electric eld produced by the charge q is
proportional to the charge q, and is independent of the shape of the surface
closing the charge.
calculation of the Gausss law.
Consider rst the ux through a spherical sur-
face, for which we nd after a simple algebra,
that the ux is
S1
=
S1
E ndS =
S1
q
4
0
r
2
0
r ndS
=
q
4
0
r
2
0
S1
dS =
q
4
0
r
2
0
4r
2
0
=
q
0
.
Thus, we see that the ux through the
spherical surface is proportional to the
charge enclosed by the surface. An obvi-
ous question arises: Is this relation valid
only for the case of spherical surfaces or
maybe it is valid for an arbitrary sur-
face?
In order to answer this question, consider the ux through an arbitrary sur-
face
S2
=
S2
d
S2
=
S2
q
4
0
r
2
r ndS ,
where S2 is an arbitrary surface enclosing the charge q.
However, from the inverse square law and some geometry, we nd
d
S2
=
q
4
0
r
2
r ndS =
q
4
0
dS cos
r
2
=
q
4
0
d ,
where d is the solid angle subtended by dS at q.
We see from Figure 12 that the element of solid angle d is independent of
where we cut the bundle of electric lines of force. Thus
d
S2
= d
S1
and then
d
S2
=
d
S1
=
q
0
.
42
Hence
S2
=
d
S2
=
q
0
. (8)
Equation (8) is the statement of the Gausss Law. It says that the ux
of the elcetric eld

E through a closed surface equals 1/
0
times the total
charge contained within the surface.
Gausss law for a number of point
charges enclosed by a surface S.
Furthermore, the Gausss Law applies not
only to a single charge contained within
a surface S, but also for some arbitrary
number of point charges q
1
, q
2
, q
3
, , q
n
or
a continuous distribution of charges within
the surface S.
We can write this statement in the following
way
d =
E
1
+

E
2
+

E
3
+ ) n dS
=
1
+
2
+
3
+
=
q
1
0
+
q
2
0
+
q
3
0
+ =
1
q ,
where

q is the algebraic sum of all charges
within the surface S. For a contiuous distribution of charges inside the
surface S
q =
V
dV ,
where is the charge density in the volume V enclosed by the surface S.
This general property is what one could expect, the Gausss law applies to an
arbitrary number of charges due to the additive nature of the elds produced
by each charge separately.
Gausss Law for a source eld charge outside S
43
If the source charge q is outside S, as illustrated in Figure 14, the surface in-
tegral vanishes since the total solid angle subtended at q by the surface is zero.
Proof:
Consider rst the ux of the electric eld through a surface element dS
1
seen
from the charge q under a solid angle d:
d
1
=
q
4
0
r n
1
dS
1
r
2
1
.
However
r n
1
dS
1
r
2
1
=
dS
1
r
2
1
= d ,
where minus sign is from the fact that at the surface dS
1
the angle between r
and n
1
is larger than 90
.
Figure 14: Illustration of the calcula-
tion of the Gausss law for the source
charge outside the surface S.
Similarly, the ux of the electric eld
through a surface element dS
2
seen from
the position of the charge q under the
same solid angle d is
d
2
=
q
4
0
r n
2
dS
2
r
2
2
=
q
4
0
d ,
where now we have positive sign since at
the surface dS
2
the angle between r and
n
2
is smaller than 90
. Thus, we nd that
d
1
+ d
2
= 0 ,
and integrating over all S, we obtain that
the total ux through the surface S is
=
S
d = 0 .
The physical interpretation of this result is that eld lines originating from
an external charge and entering the surface S must also leave this surface.
44
In summary, the Gausss law says that the total electric ux through a
surface S enclosing a charge q is:
=
q
0
, where q =
charges INSIDE S .
Using the denition of the ux, we often write the Gausss law as
E d
S =
q
0
.
The power of the Gausss law lies in the fact that we are free to apply it to any
closed surface whose shape can be chosen arbitrary such that the evaluation
of the surface integral becomes a simple straightforward task. The Gausss
law is particularly useful in the calculation of the electric eld produced by
certain symmetrical charge distributions. If the distribution of the charges
does not correspond to any simple symmetry, the Gausss law is not much
helpful in the calculations. We illustrate this in the following example.
Example of an application of the Gauss Law
An innitely long line is positively and uniformly charged with a constant
linear charge density
l
. Use (a) Coulomb law, (b) Gausss law to nd the
electric eld about the line.
(a) If we have to calculate the eld due to a static macroscopic distribution
of charge using the Coulomb law, we divide the macroscopic charge into
innitesimal (point) charges dq which produce an electric eld d
E. The eld
d
E of the point charge dq is given by the Coulomb eld

d
E =
1
4
0
dq
r
2
r .
Then the total electric eld is found by vector addition
E =
E .
Since we are adding vectors, a caution must be employed. We use the fol-
lowing procedure, which is general and can be employed to any system:
45
1. Write the expression for the electric eld d
E produced by the innites-

imal (point) charge dq.
2. Choose a reference frame and resolve the vector d
E into components
dE
x
, dE
y
, and dE
z
.
3. Calculate each component of E separately by integration, e.g. calculate
E
x
=

dE
x
.
4. Find the resultant

E from its components
E = E
x
i + E
y
j + E
z
k .
Return now to our example of the charged innitely long line and lets try to
solve the example problem following the above procedure.
Figure 15: Illustration of an ap-
plication of the Coulomb law to the
calculation of the electric eld of an
innitelt long line charge.
Take a small element dl of the line contain-
ing a point charge dq. Electric eld pro-
duced by the point charge at A distance r
from dl is given by the Coulomb eld
d
E =
1
4
0
dq
r
2
r .
We see from the gure that
r =
h
sin
, l = hcot .
Since the density of the charges on the line is
constant, the charge on the line element dl is
dq =
l
dl , where dl =
h
sin
2
d .
Then
d
E =

l
4
0
h
sin
2
sin
2
h
2
d
cos
i + sin
=

l
4
0
h
cos
i + sin
d ,
where we have decomposed the unit vector r into two (x, y) components
r = cos
i + sin
j .
46
Integrating over from = to = 0, as the line is innite and we go in the
direction of increasing l (from x = , = to x = , = 0), we obtain
E =

l
4
0
h
(sin 0 + sin )
i + (cos 0 cos )
=

l
2
0
h
j .
Thus, the electric eld produced by the charged line depends inversely on the
distance from the line and points in the direction perpendicular to the line.
(b) Let us now calculate the eld by the direct application of the Gausss law.
The electric eld near the uniformly charged line must be radially directed
because of the symmetry of the problem. The eld must have cylindrical
symmetry because the problem is unchanged by rotating the line about its
axis. The eld must also be independent of position along the line because
the distance to either end is innite.
Figure 16: Application of the Gausss
law to an innitely long line charge.
This is the ideal situation for the
application of Gausss Law. Due
to the cylindrical symmetry of the
eld, we can apply a cylinder sur-
face of radius h and length L cen-
tered about the line of charge, as shown
in Figure 16. Such a surface is of-
ten referred to as a Gaussian sur-
face.
According to the Gausss law
E d
S =
q
0
,
where q =
l
L is the charge on the line closed by the cylinder surface.
The ux through the cylinder surface splits into three uxes
E

dS =
E

dS
A
+
E

dS
B
+
E

dS
C
.
47
Since

E

dS
A
,

E

dS
C
,

E |

dS
B
, and the magnitude of

E is constant
along the surface B, the ux through the cylinder reduces to that over the
surface B only
E

dS =
B
EdS
B
= 2hLE ,
and since the cylinder is symmetrically positioned about the line of charge,
the magnitude of E is constant over the surface B. Then, according to the
Gausss law
2hLE =

l
L
0
,
which gives
E =

l
2
0
h
.
Note how simple are the calculations of the electric eld using the Gausss
law. However, we were able to solve this problem because we knew the
direction of the eld at any point around the line.
48
4.3 Biot-Savart Law Force between Static Currents
There is a force not only between electric charges but also between electric
currents. This force has a dierent nature than that one due to electric
charges, it is due to magnetic elds produced by the currents.
Figure 17: Example of an experiment
illustrating the existance of the magnetic
force between two current carrying wires.
It has been rst noticed by Oer-
sted in 1819, and few years later,
in 1827 by Ampere who showed that
quantitatively the magnetic forces
in macroscopic circuits can be ac-
counted for by what has come
to be known as the Biot-Savart
Law.
To illustrate the presence of mag-
netic forces between electric currents,
consider an experiment involving two
long parallel wires carrying currents I
1
and I
2
and separated by a distance d,
as shown in the Figure 17.
Experimental observations:
If I
1
|I
2
then the force F is attractive.
If I
1
anti| I
2
then F is repulsive.
If one of the wires is rotated through 90
then F = 0.
The force is proportional to the currents I
1
and I
2
.
All the above observations can be conbined into a single equation for the
force acting on the current I
2
:
d
F
2
=

0
4
I
2
d
2
(I
1
d
1
r)
r
2
,
where
0
= 4 10
7
[H/m] in SI units, is the permeability of the vacuum.
49
We can write the force as
d
F
2
= I
2
d
2
d
B ,
where
d
B =

0
4
I
1
d
1
r
r
2
,
which is known as the Biot-Savart law for magnetic eld produced by the
current element I
1
d
1
.
The Biot-Savart law allows to compute magnetic eld produced by an arbi-
trary current distribution Id
B =

0
I
4
l
d
r
r
2
. (9)
The method requires integration of small current elements. Note, all three
quantities appearing under the integral change during the integration, which
complicates the evaluation of the integral.
We can simplify the calculations of

B by using the following procedure.
If we replace r/r
2
by (1/r), the integrand becomes
d
(1/r) .
Next, using a vector identity we nd that
=
1
r
d
1
r
= d
1
r
,
since d
= 0. Hence, we can write
B =

0
I
4
l
d
r
. (10)
Thus, the magnetic eld can be expressed as
B =

A .
50
We see that

B = 0 always. Thus, in this case we can rst calculate

A:
A =

0
I
4
l
d
r
, (11)
which involves only two variables d
and r, and then using (10), we nd

B.
The integral for

A is easier to calculate than the original expression (9) for

B.
Since the curl operation is readily performed, we may use (11) as an inter-
mediate step for nding

B in a simpler way.
As we have already mentioned, the vector

A is called a vector potential,
and will see later in the course many useful applications of

A in electromag-
netic theory.
4.4 Current Element and Charge Element
The Biot-Savart law shows that the magnetic eld is produced by a current
element. Here, we will show that in fact the magnetic eld is produced by
moving charges.
Figure 18: Current element of a charge
moving with a velocity v.
A charge dq moving with a velocity v
is equivalent to an element of cur-
rent Id:
dq = Idt = I
d
v
,
where dt is the time for all the charge
in d to pass out of the volume, as il-
lustrated in Figure 18.
Hence
dqv = Id
.
The force between the two current elements can then be written as
d
F
2
=

0
4
dq
2
v
2
(dq
1
v
1
r)
r
2
,
51
or
d
F
2
= dq
2
v
2
d
B
1
,
with
d
B
1
=

0
4
dq
1
v
1
r
r
2
.
This shows that magnetic eld is produced by a moving electric charge.
Moreover, it shows that the magnetic eld can act with a force only on
moving charges. No force of a magnetic eld on stationary charges.
4.5 The Lorentz Force
When a charge is moving in electric and magnetic elds, the force acting on
the charge is no longer the Coulomb force. It is the Lorentz force that is
obtained putting Coulombs Law

F
E
= q
E and the Biot-Savart Law

F
M
=
qv

B together:
F
EM
=

F
E
+

F
M
= q(
E +v

B) .
Thus, a motion of electric charges is modied by both the electric and mag-
netic forces. If the charge is stationary, the force depends only on

E, if it
moves, there is an additional force proportional to v.
52
4.6 Amperes Circuit Law
Amperes circuit law is a useful relation between currents and magnetic elds.
This law allows us to calculate magnetic eld produced by certain current
distributions in a very eective way.
Figure 19: Contour for evaluation of
the Amperes circuit law.
The Amperes law says that for an arbi-
trary closed path around a current car-
rying conductor, as shown in Figure 19,
the component of magnetic eld tangent
to the path is proportional to the net cur-
rent passing through the surface bounded
by the path

B d
=
0
I ,
where the integration is over a closed loop
and I is the total current through the
loop.
11
The Amperes law can be applied in highly symmetric situations to nd the
magnetic eld more easily than by computing with the Biot-Savart law. In
either case, the result is the same. In case that lack the proper symmetry,
the Amperes law is not easily applied. The following example illustrates the
use of Amperes law.
Worked Example: An application of the Amperes Law
An innitely long wire caries a constant current I. Use (a)
Biot-Savart law, (b) Amperes law to nd the magnetic eld
about the wire.
(a) First, we divide the current into the small current el-
ements Id
. The element of magnetic eld d
B due to an
element I d
at a point A distance r = r r is found from the

Biot-Savart formula:
d
B =

0
4
I d

r
r
2
,
11
The prove of the Amperes law is complicated and will not be presented at the lecture.
The student, if interested in details of the prove is referred to the Appendix A.
53
where we note that all d
B are in the same

direction normal
to the direction of the current. So we see from this symme-
try that the eld lines are circles concentric with the current.
Furthermore, along any such circular path the eld is con-
stant in magnitude.
Let us calculate the magnitude of the magnetic eld. Since
d
r = dl sin
, r =
h
sin
, l = hcot ,
we have
dl =
h
sin
2
d
and then
d
B =

0
I
4
sin
2
h
2
h
sin
2
sin d

=
0
I
4h
sin d
.
Integrating the above equation over the length of the wire,
we obtain
B =
0
I
4h
sin d
=

0
I
4h
(cos 0 cos )

=

0
I
2h
.
(b) Let us now calculate the eld using the Amperes law.
Since the eld lines are circles concentric with the current,
54
and along any such circular path the eld is constant in mag-
nitude, this is the ideal situation for the application of Am-
peres law:

B d
=
0
I 2hB =
0
I B =

0
I
2h
.
Note how much simpler are the calculations of the magnetic
eld using the Amperes law.
4.7 Faradays Law of Electromagnetic Induction
Faraday discovered electromagnetic induction by changing magnetic eld. If
we consider a closed stationary circuit located in a varying magnetic eld,
as shown in Figure 20, the induced electromotive force around this circuit is
equal to the negative time rate of change of the magnetic ux through the
circuit
c =
d
dt
,
where is the total magnetic ux through the circuit.
12
Figure 20: Illustration of a closed cir-
cuit located in a magnetic eld.
From the denition of the ux
=
B

dS ,
and from the fact that the emf force c is
equal to the work done per unit charge,
we have
c
q
=
E +v

B) d
.
In a stationary circuit v = 0 (and any-
way v

B d
since v | d
). Thus
12
The electromotive force is induced by varying the ux of the magnetic eld that is
proportional to the magnetic eld and the area of the loop. Thus, the electromotive force
can also be induced in a constant magnetic eld by varying the area S of the loop.
55
v

B d
= 0.
Hence
c =
E d
and nally
E d
B d
S ,
or
E d
B
t
d
S .
This is the Faradays law written in the integral form.
The student can easily show, using the Stokess theorem, that the Faradays
law can be written as

E =
B
t
,
which is called the dierential form of the Faradays law.
In summary: The Faradays law tells us that time-varying magnetic elds
give rise to electric elds. This shows that the elds are related to each other,
and we then must speak of electromagnetic elds, rather than separate
electric and magnetic elds.
56
Revision questions
Question 1. State the Gausss Law and explain under what conditions the
Gausss law is useful in calculating the electric eld produced by a
macroscopic charge.
Question 2. State the Biot-Savart Law and show that magnetic eld is produced
by moving charges.
Question 3. State the Amperes Law and explain under what conditions the
Amperes law can be successfully applied for the calculation of the
magnetic eld produced by a macroscopic currents.
Question 4. What is the Faradays law of electromagnetic induction and derive
its dierential form.
57
Tutorial problems
Problem 4.1 Field of a capacitor
(a) An innite at sheet has a uniform charge density of [Cm
2
].
Use Gauss Theorem plus an argument from symmetry to nd the
electric eld strength everywhere.
(b) A capacitor that has plates (area A) of diameter very much
greater than the spacing d between them has charge densities +
and on the plates, respectively. Use the result in (a) to nd
the total electric eld strength due to the charges on the plates,
within and outside the capacitor.
(c) Calculate the force per unit area on one plate due to its at-
traction by charges of opposite sign on the other plate.
Problem 4.2 Properties of a radial vector eld
If

F is the vector r
n
r, (r = 0) show that:

F = 0 ,

F = (n + 2)r
n1
.
Problem 4.3 The Poisson equation
Note from the above tutorial problem 4.2 that the divergence of
the radial eld is zero for n = 2. The student may immedi-
ately comment that this result means that the Coulomb eld of
this symmetry has no point source. However, we know that the
Coulomb eld is produced by a point charge. How then the diver-
gence is zero if there is a point source of the eld?
To resolve this dilemma, show that for all r including r = 0
r
r
2
= 4(r) ,
where (r) is the three-dimensional Dirac delta function.
58
Problem 4.4 Field of a non-uniformly charged sphere
A sphere of radius a has a charge density increasing linearly with
radius from zero at its centre to
0
at r = a. Use the Gausss law
to nd the electrostatic eld inside as well as outside the sphere.
Use the results to conrm that

E =

0
and

E = 0 ,
everywhere.
Problem 4.5 Field of charged coaxial cylinders
Figure 21:
Two innitely long coaxial cylinders have
radii a and b with b > a, as shown in
Figure 21. The region between them is
lled with charge of volume density given
in cylindrical coordinates as
= Ar
n
,
where A and n are constants. The charge
density is zero everywhere else. Find the
electric eld

E everywhere, i.e. inside,
between and outside the cylinders.
Problem 4.6 Field of a uniformly charged cylinder
An innitely long cylinder has a circular cross section of radius a.
It is lled with a charge of constant volume density . Find the
electric eld of this charge for all points both inside and outside
the cylinder.
Problem 4.7 Charge density for a cylindrically symmetric electric eld
A certain electric eld is given in cylindrical coordinates by

E =
E
0
(r/a)
3
r for 0 < r < a and

E = 0 otherwise. Find the volume
charge density.
59
Problem 4.8 Charge in a uniform magnetic eld
(a) A charge q enters a region of uniform magnetic eld

B moving
at right angles to the eld. Show that the charge will undergo cir-
cular motion in the eld and nd an expression for the frequency
of the circular motion.
(b) Describe the motion of the charge if it is not moving at right
angles to the magnetic eld.
Problem 4.9 Magnetic eld of an innite current sheet
Consider an innite plane sheet containing an electric current
which is in the same direction everywhere (the current can be con-
sidered to close at innity if we want to justify use of closed circuit
theorems). The strength of the current would be described by a
current per unit length (with the length drawn at right angles
to the direction of the current). So the units of surface current
density would be amps/metre. Note that if the current is entirely
conned to a plane the ordinary current density J in amps/metre
2
must be considered as innite.
(a) Use an argument from symmetry plus Amp` eres circuital law
to nd the magnetic eld due to an innite plane with surface
current density J
s
amps/metre. It may be helpful to think of the
surface current as the limit of a large number of parallel wires.
(b) Using the result of part (a), nd the force per unit area be-
tween two parallel current sheets containing surface current den-
sities I
1
and I
2
if the currents in the two sheets are both in the
same direction.
60
5 Dierential Equations for the EM Field and
Maxwells Theory
We know now that electromagnetic forces are carried by electromagnetic
elds that propagate at speed c 3 10
8
ms
1
. Because of the nite propa-
gation speed we are forced to assign energy and momentum to the elds i.e.
we must think of them as real physical entities as against mere mathematical
conveniences (as is the case for static elds). An electromagnetic system then
qualies as static only if all the charges have been at rest longer than the
time taken to traverse the system at speed c.
In the 1830s Michael Faraday carried out experiments to measure a nite
electromagnetic propagation speed. He was unsuccessful due to lack of time
resolution in his apparatus. Faraday would have had no reason to think that
the electromagnetic speed was the same as the speed of light (then known).
In the 1860s, James Clerk Maxwell, seeking to advance Faradays ideas about
electromagnetic elds, by a brilliant process of intuition worked out how to
generalize certain dierential equations deduced from static experiments. He
produced a set of eld equations known by his name today. Maxwell also
had a theory i.e. a set of qualitative ideas underpinning his equations. The
theory, unlike the equations, has not stood the test of time.
5.1 Dierential Equations for the EM Field
Let us take as the source of the electromagnetic eld a continuous macro-
scopic charge and current distribution represented in terms of the macro-
scopic source quantities, charge and current densities
q =
dV , I =

J ndS .
Using the Gausss and Stokess Theorems, we will nd dierential vector
equations for the elds

E and

B from the integral forms of observational
results discussed in the previous chapter.
61
5.1.1 Divergence of

E
Consider the Coulombs Law that can be written in the form of the Gausss
law as
E d
S =
q
0
=
0
dV .
Applying the Gausss Theorem, Eq. (1), to the left-hand side of the above
equation, we obtain

E dV =
0
dV .
Since this relation must hold for arbitrary V , no matter how small, we can
drop the integrals, and obtain

E =

0
.
This equation is called the dierential form of the Coulombs (Gausss) Law.
From this equation it follows that the divergence of the electric eld is zero
in all regions where there is no electric charge.
5.1.2 Curl of

E
Consider the Faradays ux cutting rule
E d
B
t
ndS .
Applying Stokess Theorem to the left-hand side of the above equation, we get

E ndS =
B
t
ndS .
Since this relation must hold for arbitrary S and n, it implies that

E =
B
t
.
This equation is called the dierential form of the Faradays law.
62
5.1.3 Divergence of

B
We calculate the divergence of

B from the Biot-Savart law
B =

0
4
I d

r
r
2
.
By taking of both sides of the Biot-Savart law and applying the vector
identity
(
A

B) =

B

A

A

B ,
we obtain

B =

0
I
4
r
r
2
d

r
r
2
. (12)
However, d
= 0, since d
is a constant vector with respect to the

dierentiation over x, y and z, and
r
r
n
0 for any n .
Hence, the right-hand side of Eq. (12) is always zero, so that always

B = 0 .
This equation, sometimes called Gausss law for magnetism, states that the
ux of the magnetic eld

B is zero through any closed surface.
5.1.4 Curl of

B
We nd curl of

B from the Amperes circuit law
B d
=
0
I =
0
J ndS .
Simply, by applying Stokess Theorem to the Amperes law, we nd that the
following relation

B ndS =
0
J ndS
63
holds for arbitrary S. Thus

B =
0
J . (13)
This equation is called the dierential form of the Amperes circuit law.
However, Maxwell realized that unlike the previous three dierential equa-
tions, this one, Eq. (13), could not be generally true. To see this, take its
divergence and remember that

F 0 for any vector function

F, so
that

B = 0 =
0

J .
Thus

J 0 ,
which means that the Amperes law does not include source currents. This
result is in contradiction with the continuity equation of

J varying with time.
We have already seen that conservation of electric charge requires

J =
t
.
Thus

J 0 implies that
t
0 i.e. = const.
Hence, the Ampere law may be applied only for constant currents, i.e. ac-
cepting the Ampere law as general, we could never charge or discharge a
capacitor.
No wonder Maxwell was confused!
5.2 The Maxwells Theory
Maxwell guessed the right form of

B as follows. Since

J +

t
= 0
64
and using for the rst Maxwell equation
=
0

E ,
we have

J +

0

E
t
= 0 , or
J +
0
E
t
= 0 .
If then we, after Maxwell, write:

B =
0
J +
0
E
t
,
instead of

B =
0
J ,
we obtain

B 0
J +
0
E
t
= 0 ,
which is in accordance with conservation and motion of charge.
Note from above that the term that Maxwell added
E
t
,
has the dimensions of current density. Maxwell called it the displacement
current density.
Note that the displacement current density does not explicitly involve charges,
they do not appear explicitly in the denition of the displacement current
density. Why then we call it as a current? Maxwell had a theory under-
pinning his equations in which the displacement current was a real physical
current - due to polarization of the electromagnetic ether. This theory has
not survived. Nevertheless the above term is still referred to as the displace-
ment current.
65
With the modication of the current density, we then write the fourth Maxwells
equation as

B =
0
J +
0
E
t
,
which contains all the physical processes involved in the generation of mag-
netic elds.
In summary:
The dierential form of the Maxwells equations is easier to inter-
pret physically and is also useful in deriving the boundary condi-
tions that the eld vectors must satisfy.
The Maxwells equations are self-consistent and no experimental
evidence for requiring any further modications has been found.
The equations are used to explain and predict all physical phenom-
ena that involve charges and/or currents.
Exercise in class: Fields within a capacitor
A plane parallel capacitor is being charged with a current I.
Show that the displacement current between the plates of
the capacitor is equal to the conduction current I in the ex-
ternal charging circuit. Remember that the displacement
current density is, by denition,
0
E
t
so the displacement
current through a surface S is:
I
D
=
E
t
ndS .
Assume that the external wires are perfect conductors so
that

E is zero in them. Assume the space between the
plates is a perfect insulator so no conduction current ows
within the capacitor.
66
Can you see any curious consequence in this case if the
displacement current is assumed to be a real physical cur-
rent (ow of charges)?
67
Revision questions
Question 1. Derive the dierential forms for the Coulomb, Ampere and Faraday
laws.
Question 2. Derive the so-called Gausss law for magnetism.
Question 3. Explain, how Maxwell resolved the diculty with general applica-
tions of the Amperes law.
Question 4. What is the displacement current and explain its signicance in
the EM theory.
Question 5. Are the Maxwells equations independet of each other? Explain.
Tutorial problems
Problem 5.1 Coulomb eld
Demonstrate that the Coulomb eld for stationary point charge
E =
1
4
o
q
r
2
r
follows from Maxwells equations.
Problem 5.2 Magnetic eld of an innite straight wire
Magnetic eld at distance r from an innite straight wire carrying
constant current I in the z direction is given by
B =

o
I
2r
,
where

is the unit vector in the direction of

B in the xy plane.
Show, using only the circuit, surface and volume integrations, and
the Gauss and Stokes Theorems that the magnetic eld satises
Maxwells equations.
68
6 Maxwells Equations and Prediction
of Electromagnetic Waves
In the previous lecture, we have derived from the experimental laws vectorial
expressions for both the divergence and curl of the two basic eld vectors

E
and

B. These expressions put together completely determine the electro-
magnetic elds and are termed the Maxwells equations
I.

E =

0
,
II.

B = 0 ,
III.

E =
B
t
,
IV.

B =
0
J +
0
E
t
.
As we have shown in the previous lecture, the rst equation, which we shall
call the Maxwells equation number I, is the dierential form of the Gausss
law, the second equation, which we shall call the Maxwells equation num-
ber II, tells us about the non-existence of magnetic charges, the third equa-
tion, which we shall call the Maxwells equation number III, is the dieren-
tial form of the Faradays law, and the nal equation, which we shall call
the Maxwells equation number IV, is the dierential form of the modied
Amperes law.
Maxwells immediate triumph with the modication of the Amperes law
was to predict the existence of electromagnetic waves and their propagation
speed. The calculated speed came (within experimental error) to be equal
to the measured speed of light. This prediction obviously led to the con-
clusion that light was electromagnetic in nature. Thus arose a synthesis of
electromagnetism and optics. In this lecture, we will show how the Maxwells
equations predict the existence of the electromagnetic waves and will analyse
their properties when the elds propagate in vacuum. Much of our discus-
sion, however, will be about how to solve the Maxwells equations for elds
propagating in vacuum.
69
6.1 The Wave Equation for EM Waves in Vacuum
In a vacuum there are no sources, i.e. = 0 and

J = 0. Hence, the Maxwells
equations reduce to the following dierential equations

E = 0 , (14)

B = 0 , (15)

E =
B
t
, (16)

B =
0
E
t
. (17)
These equations are readily to solve. The procedure is to eliminate

E or

B
between equations (16) and (17) to obtain dierential equations for

E or

B
alone, using where required Eqs. (14) and (15).
Method:
Think of and /t as linear (dierential) operators. By analogy with
methods of solving linear algebraic equations, applying into (16) and
/t into (17), we obtain

B
t
,

B =
0
E
t
2
.
Hence
=
0
E
t
2
,
and using a vector identity
=

E
2
E ,
we obtain

E
2
E =
0
E
t
2
.
70
Because

E = 0 in the vacuum, it follows at once that
E =
1
c
2
E
t
2
, (18)
where
c
2
=
1
0
.
The parameter c has the dimensions of velocity and is numerically equal
to 3 10
8
ms
1
.
Equation (18) is the standard form of a three-dimensional vector wave equa-
tion. Following the same procedure, we can easily show that the magnetic
eld

B satises the same equation. The proof is left to the student.
6.2 Plane Wave Solution to the Wave Equation
The wave equation in a vacuum is
2

X =
1
c
2
2

X
t
2
for

X

E,

B.
We will solve the wave equation in one dimension assuming that the wave
propagates in the z direction.
In this case, /x = /y 0, and then
2
=

2
x
2
+

2
y
2
+

2
z
2
=

2
z
2
.
The dierential equations for

E and

B both have the same form:
2

X
z
2
=
1
c
2
2

X
t
2
.
Such an equation has solutions of the form
X = f(z ct) , (19)
71
where f is an arbitrary function with the argument z ct.
This solution represents a signal propagating with speed c as can be seen
from the following discussion:
Let X
0
= f(z
0
ct
0
) i.e. it represents the solution X at t = t
0
and z = z
0
.
Now examine X at time t later and distance z further along in z. Since
a harmonic wave does not change in vacuum, we have
X
1
= f(z
0
+ z c(t
0
+ t))
= f(z
0
ct
0
) = X
0
when z = ct ,
i.e. the signal propagates a distance z = ct in time t i.e. it propagates
with speed c.
Proof of solution (19):
Let f represent f(z ct). Then
f
z
=
f
(z ct)
(z ct)
z
=
f
(z ct)
= f
,
where
means dierentiation wrt z ct. Similarly
2
f
z
2
=
f
z
=
f
(z ct)
(z ct)
z
= f
,
and
f
t
=
f
(z ct)
(z ct)
t
= cf
.
Next
2
f
t
2
= (c)(c)f
,
and consequently
2
f
z
2
=
1
c
2
2
f
t
2
.
72
Thus
c
2
=
1
0
c =
1
0
,
which with the numerical values of the parameters
0
8.85 10
12
Fm
1
,
0
= 4 10
7
Hm
1
gives
c 3 10
8
ms
1
.
The fact that the numerical value of c is equal to the velocity of light in
vacuum led Maxwell to propose an electromagnetic theory of light, one of
the brilliant contributions to physics in the nineteenth century.
6.3 Harmonic Waves
In vacuum, we chose a plane wave representation for the propagating EM
wave. For a harmonic wave
c = =

k
, = 2, k =
2
,
where = frequency (Hz), = wavelength (m), = radian frequency (ra-
dians s
1
), and k = propagation constant (m
1
).
f(z ct) = f(z

k
t) = f
1
(t kz) .
A plane wave is represented by the electric eld
E =

E
0
e
i(tkz)
, (20)
whose the propagation is characterized by the frequency and the wave
vector

k. Since the phase of the wave described by Eq. (20) is by denition
the argument of the exponent, we see that the surfaces of the constant phase
are planes whose normal is the z-axis:
i(t kz) = i
t
z
c
73
given at any denite time by z = const. These surfaces of constant phase
travel with a constant velocity c often referred as the phase velocity of the
wave.
13
Therefore, the waves described by Eq. (20) are called plane waves.
6.4 The Transverse Nature of Plane Waves in Vacuum
We now investigate the relations between the amplitudes and phases of the
electric and magnetic elds in a plane harmonic wave. While it is true that
the magnetic eld satises the same wave equation as the electric eld, it
is not independent of the latter, since one must satisfy the Maxwell equa-
tions III and IV.
Since

B = 0 always in electromagnetism, and

B =
B
x
x
+
B
y
y
+
B
z
z
= 0 + 0 +
B
z
z
for a plane wave propagating in the z direction, we have
B
z
z
= 0 .
However, for a plane wave
B
z
z
= ikB
z
.
Hence, the rhs must be zero, which means that either k = 0 (zero frequency)
or B
z
= 0 (transverse wave).
For a propagating wave k = 0, so the wave is transverse in

B.
In a vacuum

E = 0 and then by the same argument we conclude that
the plane wave is also transverse in

E. In other cases in electromagnetism
13
Note that some textbooks on electromagnetism, for engineers in particular, use the
letter j instead of i for the imaginary number. We will use i throughout this lecture notes.
74
(e.g. for a plasma in a magnetic eld) a plane wave may not be purely trans-
verse in

E.
In addition:

E

B for a plane EM wave in a vacuum.
Proof:
Consider a harmonic wave propagating parallel to the z axis. In this case
the eld components are of the forms
E =

E
0
e
i(tkz)
,

B =

B
0
e
i(tkz)
.
In order to prove the statement, we will use the Maxwells equation III

E =
B
t
,
and expand it in Cartesian coordinates remembering that /t i
i

j

k
z
E
x
E
y
E
z
= i(B
x
i + B
y
j + B
z
k) . (21)
For electromagnetic plane waves propagating along the z axis in a vacuum:
E
z
= B
z
= 0,

x
=

y
= 0,

z
= ik .
Hence, Eq. (21) reduces to
i

j

k
0 0 ik
E
x
E
y
0
= i(B
x
i + B
y
j) .
Expanding the determinant, and comparing the left and right-hand sides,
we nd
x component : ikE
y
= iB
x
B
x
=
k
E
y
,
y component : ikE
x
= iB
y
B
y
=
k
E
x
.
75
What does it tell us about the relation between

E and

B?
If we consider a scalar product

E

B, and use the above result, we nd
E

B = (
iE
x
+

jE
y
) (
iB
x
+

jB
y
)
= (
iE
x
+

jE
y
) (
i
k
E
y
+

j
k
E
x
)
=
k
E
x
E
y
+
k
E
x
E
y
= 0 .
This means that

E

B.
In addition, note that
E
B
=

k
= c .
Thus, we may conclude that in the electromagnetic theory when E and B
are related, their ratio is always a velocity characteristic of the problem in
hand.
This is as far as Maxwell took the subject. It was for others like Heinrich
Hertz 1884 to show how to solve Maxwells equations with source terms ,

J
included (i.e. the generation of electromagnetic waves). We will consider
this later.
You should be aware that we have not derived Maxwells equations from the
static limits like Coulombs Law and the Biot-Savart Law. The solutions to
Maxwells equations include the static limits as special cases but many more.
Maxwells equations have the status of postulates suggested by experimental
results.
In summary, we have the following important results for related electric
and magnetic elds propagating in vacuum:
1. The electric and magnetic elds propagate in a form of plane waves,
so-called electromagnetic (EM) waves.
2. The plane EM wave is transverse in

E and

B, i.e. both elds are
perpendicular to the direction of propagation.
76
3. The electric and magnetic elds are perpendicular to each other.
4. The ratio E/B is constant and equal to the velocity of the wave, that
is equal to the speed of light in vacuum.
Exercise in class: Electric and magnetic elds of a plane wave
For a given electric eld polarized in x direction and prop-
agating in z direction
E = E
0
sin (t kz)
i ,
calculate

B from III and show that the elds represent
a plane electromagnetic wave propagating in vacuum, i.e.
they satisfy the Maxwells equations (14)(17).
77
Revision questions
Question 1. Show that the Maxwells equations for the EM elds propagating
in vacuum can be reduced to two independent second order dier-
ential equations, the so-called the wave equations.
Question 2. State the properties of an EM wave propagating in vacuum.
Question 3. How do we dene phase velocity of an EM wave?
Question 4. The wave equations for the electric and magnetic elds propagating
in vacuum are independent of each other. How then the electric
and magnetic elds are related to each other? Explain.
Tutorial problems
Problem 6.1 Fields of a plane wave
A plane electromagnetic wave propagates in the z direction in
a vacuum. Its electric eld is polarized in the x direction and so
has the form:
E = E
0
sin(t kz)
i
(a) If the wave amplitude is 10 volts m
1
and its frequency is 1
MHz, nd numerical values for E
0
, , k and the wavelength.
(b) Calculate

E directly from the above expression for

E.
Plot

E and

E as functions of z for t = 0 (on the same plot).
(c) Hence calculate the magnetic eld of the wave using the Maxwell
equation:
B
t
=

E .
78
(i) Show that

B varies in phase with

E.
(ii) Calculate the ratio E/B.
(d) Show that in the stated circumstances, the remaining Maxwells
equations should be:

E = 0,

B = 0, and

B =
0
E
t
.
(e) Write expressions for

E and

B of a right-circularly polarized
plane wave having the same amplitude and frequency as the above
one.
Problem 6.2 EM wave in free space
An EM wave traveling in free space ( = 0,

J = 0) in the pos-
itive z direction is described by
E = E
0
cos (t kz)

i,
B = B
0
cos (t kz)

j,
where k, E
0
, and B
0
are constants. Under what circumstances do
these

E and

B elds satisfy all of Maxwells equations?
79
7 EM Theory and Einsteins Special Theory
of Relativity
The purpose of this lecture is to test the laws of electromagnetism against
the principle of special relativity. We will demonstrate how one might infer
the law of Biot-Savart and in general relate the magnetic eld to electric
eld from application of special relativity to Coulomb law. Special relativ-
ity, formulated in 1905, grew out of Einsteins meditation on electromagnetic
theory and the properties of space and time. Historically, the insights of
Einsteins theory follow after electromagnetism. Logically however, special
relativity contains more general statements about nature than electromag-
netism. Electromagnetic eld theory is just one of a possible set of eld
theories that are compatible with the Einstein theory of space and time. It
is evident that relativistic eects are important if we would have to calculate
the eld of a charge moving with a speed comparable to that of light. What
is not so obvious is that special relativity oers insights into aspects of elec-
tromagnetic theory even in the case of the low speed charges we consider in
this course.
Two such aspects are:
(1) The unity of the electromagnetic eld i.e. the eld is a single entity
with 6 components (represented by two vectors

E and

B, each with three
components).
(2) Understanding the nature of causal relationships in electromagnetic the-
ory, e.g.

E =
B
t
.
Does this mean that a time changing

B causes a spatially changing

E?
80
7.1 Lorentz Transformation Equations for Space and
Time
We will analyse how the fundamental laws of electromagnetism change when
charges move with a constant velocity. In particular, we will show that the
Maxwells equations are derived from the Coulombs law. Thus, the student
will learnt that the whole electromagnetism follows naturally from electro-
statics. In our calculations, we will follow the principle of special relativity.
The principle of special relativity
1. The laws of physics are the same in all inertial reference frames.
2. The speed of light in vacuum is independent of the uniform motion of the
observer or source.
Figure 22: Two inertial reference
frames in relative motion. A stationary
frame S and a frame S
moving with a
velocity v parallel to the x axis.
The constancy of the velocity of light, in-
dependent of the motion of the source,
gives rise to the relations between space
and time coordinates in dierent inertial
reference frames known as Lorentz trans-
formations.
Consider a stationary reference frame S
and a inertial frame S
moving with a ve-

locity v parallel to the x axis, as shown in
Figure 22. Let x, y, z are coordinates in
the S frame and x
, y
, z
are the coordi-

nates in the S
frame. The time and space

coordinates in S
are related to those in S

by the 1D Lorentz transformations:
x
= (x vt) ,
y
= y ,
z
= z ,
t
t
vx
c
2
,
81
where
=
1
v
2
c
2
1
2
is the Lorentz factor.
The above transformation corresponds to a situation of v parallel to the x
axis. Later in the course, Chap. 21, we will consider the general case of the
velocity v of the frame S
in an arbitrary direction.
7.2 Force Transformation Equations
Let us investigate the force between two charges as viewed by observers in S
and S
. We will evaluate the force from the point of view of the observer in S
.
Consider a particle which is moving with velocity u = u
x
i + u
y
j + u
z
k in
the S frame and is acted on by a force with components F
x
, F
y
and F
z
. Then,
according to the special theory of relativity, the force in the S
frame is
F
x
= F
x
vu
y
c
2
(1
vux
c
2
)
F
y
vu
z
c
2
(1
vux
c
2
)
F
z
,
F
y
=
(1
u
2
c
2
)
1
2
1
vux
c
2
F
y
=
F
y
(1
vux
c
2
)
,
F
z
=
(1
v
2
c
2
)
1
2
1
vux
c
2
F
z
=
F
z
(1
vux
c
2
)
.
Suppose F
x
, F
y
, F
z
represents the velocity independent Coulomb force. Then
in the S
frame (source of the eld now moving) the force is no longer veloc-
ity independent. Thus, the observers disagree about the magnitude of the
force. They, in fact, disagree not only about the magnitude but also about
the origin of the force. In electromagnetic theory we say that there is now a
magnetic force and we dene a magnetic eld

B that determines the mag-
netic force.
We now present a detailed calculation that illustrates how the form of the
Maxwells equations is determined by nature obeying Einsteins special the-
ory of relativity.
82
7.2.1 The Force between Two Charges Moving with Constant Ve-
locities
Invariance of electric charge
In ordinary matter, electrons move with much greater speeds than protons
yet there is no associated electric eld. This implies that electric charge is
independent of velocity unless electromagnetic laws modied in some compli-
cated way (see discussion by King in Physical Review Letters 5, 562 (1960)).
Figure 23: Two moving charges seen in two
dierent reference frames.
Consider charges q
1
and q
2
moving
with velocities u and v in an iner-
tial frame S. No loss of generality
occurs if v is taken parallel to the x
axis.
Now consider a frame S
moving
with velocity v along x axis, that q
2
is stationary in S
. Assume that at
time t = 0, the frames S and S
overlap.
From the principle of relativity, in
the S
frame Coulombs law holds.

The force on q
1
seen in S
is therefore
F

=
1
4
0
q
1
q
2
r
3
r

.
We shall transform this expression to nd the force observed in the S frame
in which the source of the eld (q
2
) is moving.
We shall see that what we normally call the MAGNETIC FIELD arises
as a natural consequence of relativistic invariance with no extra assumptions.
We start with the x component of the force, that is
F
x
=
1
4
0
q
1
q
2
r
3
x
,
83
and similarly for y and z.
We need transformations of x
to x and r
to r. The transformation of x
to x
is found from the Lorentz transformations that at t = 0 are
x
= x ,
y
= y ,
z
= z ,
t
=
vx
c
2
.
To do the complete transformation of the force component, we also need
transformation from r
to r. It can be found as follows. It is seen from

Figure 23 that there is an axial symmetry, so that the transformation
could be expressed in terms of the angle . Thus, we can write
r
2
= x
2
+ y
2
+ z
2
=
2
x
2
+ y
2
+ z
2
=
2
x
2
+
y
2
+ z
2
=
2
x
2
+
1
v
2
c
2
y
2
+ z
2
=
2
x
2
+ y
2
+ z
2
v
2
c
2
y
2
+ z
2
=
2
r
2
v
2
c
2
r
2
sin
2
=
2
r
2
1
v
2
c
2
sin
2
,
where sin =
(y
2
+ z
2
)/r. Hence
r
= r
1
v
2
c
2
sin
2
1
2
.
Thus, substituting the transformations into F
x
, we obtain
F
x
=
1
4
0
q
1
q
2
x
3
r
3
1
v
2
c
2
sin
2
3
2
= q
1
gx ,
where
g =
1
4
0
q
2
2
r
3
1
v
2
c
2
sin
2
3
2
.
84
Similarly for the remaining two components, we nd
F
y
=
q
1
gy
, F
z
=
q
1
gz
.
In summary, the force transformations are
a) x component
F
x
= F
x
vu
y
c
2
1
vux
c
2
F
y
vu
z
c
2
1
vux
c
2
F
z
.
Thus
q
1
gx = F
x
vu
y
c
2
1
vux
c
2
F
y
vu
z
c
2
1
vux
c
2
F
z
.
b) y component
F
y
=
F
y
1
vux
c
2
,
from which we nd
F
y
=
1
vu
x
c
2
y
=
1
vu
x
c
2
q
1
gy
,
which we can write as
F
y
= q
1
gy
1
vu
x
c
2
.
c) z component
F
z
= q
1
gz
1
vu
x
c
2
.
Substituting for F
y,z
in x equation
F
x
= q
1
gx + q
1
gy
vu
y
c
2
+ q
1
gz
vu
z
c
2
.
Note: Here is the germ of the magnetic eld. The last two terms are typical
second order relativistic terms v
2
/c
2
. In a nonrelativistic calculation we
85
would have F
x
= F
x
.
We now combine results a), b) and c) into a single vector equation for the
force

F. First, note that
vu
x
= u v .
Next, we can write the x component as
F
x
= q
1
gx
1
vu
x
c
2
+ q
1
gx
vu
x
c
2
+ q
1
gy
vu
y
c
2
+ q
1
gz
vu
z
c
2
= q
1
g
1
vu
x
c
2
x + q
1
g
v
c
2
(u r)
and with the y and z components
F
y
= q
1
g
1
vu
x
c
2
y ,
F
z
= q
1
g
1
vu
x
c
2
z ,
these three components combine into
F = q
1
g
1
v u
c
2
r + q
1
g
v
c
2
(u r)
= q
1
gr +
q
1
g
c
2
[v (u r) r (u v)] ,
which can be written as
F = q
1
gr +
q
1
g
c
2
u (v r) .
We can write this equation in the form of the Lorentz force
F = q
1
E +u

B
,
where
E = gr =
1
4
0
q
2
2
r
3
1
v
2
c
2
sin
2
3
2
r ,
and
B =
v gr
c
2
=
v

E
c
2
. (22)
86
Thus, a force which we see in the moving frame S
as the Coulomb force

appears as the Lorentz force in the stationary frame S.
The relation (22) shows that

B and

E have no independent existence. A
pure electric eld in one coordinate frame appears as a mixture of electric
and magnetic elds in another coordinate frame.
Note that as v 0, 1, and then
E
1
4
0
q
2
r
3
r .
Moreover, the ration of magnitudes of magnetic to electric term in the force
equation is uv/c
2
, i.e. magnetic forces are second order relativistic eects.
7.3 Electric Field Lines of a Moving Charge
Let = v/c. Then, we can write the electric eld as
E =
1
4
0
q(1
2
)
r
2
1
2
sin
2
3
2
r .
Figure 24: Electric eld lines of the sta-
tionary charge and the charge moving with
uniform velocity v.
For a given , the electric eld E
still varies as 1/r
2
, but the eld
lines are crowded in the direction per-
pendicular to v, as shown in Fig-
ure 24.
In the forward direction = 0, and
then
E =
1
4
0
q(1
2
)
r
2
< E
s
,
where E
s
is the electric eld of the sta-
tionary charge, i.e. the static electric
eld (at v = 0). Thus, the eld ampli-
tude is lower in the direction of motion
relative to the static eld amplitude.
87
In the perpendicular direction = /2, and then
E =
1
4
0
q
r
2
1
2
1
2
> E
s
.
Thus, the eld amplitude is larger in the transverse directions relative to the
to the static eld amplitude.
14
In summary, the electric eld lines radiate from the present position of the
charge and are crowded in the direction perpendicular to the direction of
motion of the charge.
7.4 Magnetic Field Lines of a Moving Charge
From the relation between electric and magnetic elds, we nd
B =
v

E
c
2
=
1
4
0
q(1
2
) v r
c
2
r
3
1
2
sin
2
3
2
.
Figure 25: Magnetic eld lines of the
charge moving with uniform velocity v.
In order to nd the direction of the
magnetic eld, we refer to the spheri-
cal polar coordinates, and nd that
B = B
r
r + B
+ B
= B

,
since

B v r, and r,

,

are unit
vectors.
Thus, the magnetic eld lines form
concentric rings about v, and there is
symmetry about the plane = /2. In
the non-relativistic case of v < 1 the
factor 0, and then the magnetic
eld reduces to
B =
1
4
0
q v r
c
2
r
3
,
14
An interesting question: What is the electric eld distribution of a charge moving with
velocity v = c?
88
which is the Biot-Savart law. Applied to a continuous line current I:
B =
1
4
0
I

dl r
c
2
r
3
.
The constant 1/(
0
c
2
) is normally written
0
- the magnetic permeability of
free space.
7.5 Invariance of the Maxwell equations under the Lorentz
transformation
According to the principles of relativity, the laws of physics are invariant un-
der the Lorentz transformation. In this lecture, we will show that the basic
equations for the electromagnetic theory, the Maxwells equations, are invari-
ant under the Lorentz transformation, i.e. they represent laws of physics.
(1) Equation for the total electric ux
Consider the total electric ux through a surface S closing a moving charge q:
E
=
E n dS .
Figure 26: Evaluation of the total electric
ux through a surface S closing a moving
charge.
We will use the axial symmetry and
break sphere up into rings, as shown
in Figure 26, lying between and +
d. The radius of a ring seen
from the center under the angle
is r sin , and the width of the ring
is rd.
Since

E has a radial symmetry,

E | n
everywhere, and then the ux d
E
is
d
E
=

E n dS = EdS . (23)
However, the area of the ring (stripe)
between and + d is equal to
dS = 2(r sin ) rd = 2r
2
sin d .
89
Hence, substituting this result into Eq. (23) and the explicit form of E, we
nd that the ux through the ring is
d
E
=
1
4
0
q(1
2
)
r
2
1
2
sin
2
3
2
2r
2
sin d
=
q(1
2
)
2
0
sin d
1
2
sin
2
3
2
.
This gives the total ux through the surface S:
E
=
d
E
=
q(1
2
)
2
0
=0
sin d
1
2
sin
2
3
2
.
To calculate the integral, put cos = x, so that sin d = dx, and then
I =
dx
(1
2
+
2
x
2
)
3/2
=
1
dx
1
2
2
+ x
2
3/2
=
1
dx
(a
2
+ x
2
)
3/2
,
where a =
1
2
/. Performing the integration, we obtain
I =
1
a
2
3
x
(a
2
+ x
2
)
1/2
,
and nally including the limits of the integral, we get
I =
1
1
dx
(1
2
+
2
x
2
)
3/2
=
1
a
2
3
1
a
2
+ 1
+
1
a
2
3
1
a
2
+ 1
=
2
1
2
.
Thus, the total ux through the surface is
E
=
q(1
2
)
2
0
2
1
2
=
q
0
,
90
which is the same as for a stationary charge.
Hence, the electric eld produced by a moving charge satises the Gausss law.
If we apply the Gauss theorem to
E
, we get

E =

0
.
Thus, we conclude that the Gauss law (Maxwells equation I) is invariant
under the Lorentz transformation.
(2) Magnetic ux through a closed surface
Consider the magnetic ux through a closed surface
M
=

B n dS .
If we choose a sphere centered on q to calculate the integral, we nd that

B is
perpendicular to n everywhere on the sphere, i.e. is tangential to the surface
of the sphere. Thus, the ux through the sphere
M
= 0.
Similarly, we can show that

B = 0 always. It is easy to see that

B =
1
c
2
v

E
=
1
c
2
E v v

E
= 0 ,
since v = 0 as v is constant, and v is perpendicular to

E.
Thus, we conclude that the Maxwells equation II is invariant under the
Lorentz transformation.
(3) Spatial

E derivative related to temporal

B derivative.
We shall show that for a point charge under uniform velocity

E =
B
t
.
91
To do it, we write

E in spherical polar coordinates

E =
r
r sin
(sin E
1
sin
E
r

(rE
)
r
(rE
)
r

E
r
.
Since the electric eld has a radial symmetry, we have
E
= E
= 0 ,
and because the electric eld amplitude dependes only on r and , we get

E =
1
r
E
r
,
where
E
r
=
1
4
0
q(1
2
)
r
2
1
2
sin
2
3
2
=
K
1
2
sin
2
3
2
.
Calculate E
r
/:
E
r
3
2
1
2
sin
2
5
2
2
2
sin cos
=
3K
2
sin 2
2
1
2
sin
2
5
2
.
Hence

E =
1
4
0
3q(1
2
)
2r
3
2
sin 2
1
2
sin
2
5
2
.
To calculate
B/t, we use the theorem of partial derivatives. If

y = f(x, t)
92
then from the maximum change of y
dy =
y
x
dx +
y
t
dt = 0 ,
we obtain
y
t
=
y
x
x
t
.
Thus
B
t
= v
B
x
.
Alternatively, to see this physically, remember that the eld pattern moves
with constant velocity v. Let a stationary observer measure the change in the
eld

B in a time interval dt. This change is the same as he would observed
at a xed time by moving a distance dx = vdt, i.e.
d
B in dt d
B in dx = vdt .
Hence
B
x
=
B
vt
,
and then
B
t
= v
B
x
.
Now
B =
1
4
0
q(1
2
) v sin
c
2
r
2
1
2
sin
2
3
2
and
B
x
=
B
=
B
r
r
x
.
Since
sin =
y
2
+ z
2
r
=
a
r
,
93
we can write B
as
B
=
Ka
r
3
1

2
a
2
r
2
3/2
=
Ka
(r
2
2
a
2
)
3/2
,
where
K =
1
4
0
q(1
2
)v
c
2
and a =
y
2
+ z
2
.
Next
r
x
=
(x
2
+ y
2
+ z
2
)
1/2
x
=
1
2
x
2
+ y
2
+ z
2
1/2
2x =
x
r
= cos ,
B
r
= Ka
3
2
r
2
2
a
2
5/2
2r =
3Ka r
(r
2
2
a
2
)
5/2
.
Hence
B
x
=
1
4
0
3q(1
2
) vr
2
sin cos
c
2
r
5
1
2
sin
2
5
2
=
1
4
0
3q(1
2
) v sin 2
2c
2
r
3
1
2
sin
2
5
2
,
and then
B
t
= v
B
=
1
4
0
3q(1
2
) v
2
sin 2
2c
2
r
3
1
2
sin
2
5
2
=
1
4
0
3q(1
2
)
2
sin 2
2r
3
1
2
sin
2
5
2
.
Comparing with

E, we see that

E =
B
t
.
Thus, we conclude that the Faradays Law (Maxwells equation III) is invari-
ant under the Lorentz transformation.
94
7.5.1 Remark on the Electromagnetic Induction
It has been known since about 1831 when Faraday rst waved a magnet
near an electric circuit and played with transformers that when the magnetic
ux through a circuit changes, an electromotive force (emf) c appears in it.
Faraday gave the rule
c =
M
t
.
The nature of this phenomenon is, however often misinterpreted.
Figure 27: Contour for evaluation of an
electromotive force.
Consider a point charge q moving
near a closed circuit as shown in
Figure 27. Because of the de-
pendence of E, the electric eld
on side (1) is larger than that
on side (2). Thus, there is a
net driving force round the cir-
cuit.
Calculate the resulting electromotive
force in the circuit, which is equal to
the work done on a charge q
in the
circuit
c = W
q
=

F

dl
q
E +u

B)

dl ,
where u is the velocity of the charge q
in the circuit.
Now, since u

B is perpendicular to both u and

B, it is also perpendicular
to

dl, and then

u

B

dl = 0 .
Hence
c =

E

dl =

E n dS =

E

dS .
95
If we now take the conclusion that where is a spatially varying electric eld
there is also a time varying magnetic eld, we obtain
c =
B
t

dS =

B

dS =
M
t
.
However, it is obvious in this example that the changing magnetic ux is not
the CAUSE of the emf. The changing magnetic eld and the electric eld
have a common CAUSE through the charge q.
We can conclude that: Electric and magnetic elds do not produce
each other - they are both due to electric charges.
It is often thought in the textbooks, however that e.g. in a transformer the
changing
M
produces c. It happens because the ux cutting rule is an
extremely powerful one for calculating the integrated electric eld of electric
currents.
The Faradays rule c =
M
/t, which arises from

E =
B/t
should not be thought of as a casual relationship. What it means is that if
a charge moving with a constant velocity produces a time varying magnetic
eld then that charge also produces a spatially varying electric eld.
(4) Relation between spatial variation of

B and temporal variation
of

E.
Since
B =
v

E
c
2
,
and
v

E
v (v )

E +v

E ( v) ,
with v constant, we obtain
v = 0 ,
v = 0 .
96
Then

B =
1
c
2
v

E
=
1
c
2
(v )

E +v
.
Thus,
(v )

E =
v
x
x
+ v
y
y
+ v
z

E = v
E
x
=
E
t
as
v
x
= v v
y
= v
z
= 0 and

t
= v

x
.
Hence

B =
1
c
2
v

E +

E
t
.
If we recognize that

E =

0
and next that v =

J and 1/c
2
0
=
0
, we get the Maxwells equation IV.
Thus, we may conclude that the Maxwells equation IV is invariant under
the Lorentz transformation.
In summary:
Maxwells equations for a point charge moving with uniform velocity are

E =

0
,

B = 0 ,

E =
B
t
,

B =
0
J +
1
c
2
E
t
.
97
These equations arise from the necessity for the correct relativistic trans-
formations between frames in uniform relative motion. This shows that
the Maxwells equations are invariant under the Lorentz transfor-
mation. If the postulates of relativity are correct and Coulombs law gives
the eld of a stationary charge, these equations follow, and the force on a
charge is
F = q
E +v

B
.
Revision questions
Question 1. State the principle of special relativity.
Question 2. Is the Coulombs law for moving charges equivalent to the Lorentz
force?
Question 3. Are the electric and magnetic elds invariant under the Lorentz
transformation? Explain.
Question 4. Are the Maxwells equations invariant under the Lorentz transfor-
mation?
Question 5. Why the Faradays law is often misinterpreted? What is the correct
interpretation of the the Faradays law?
98
8 Energy in the Electromagnetic Field:
Poyntings Theorem
Energy may be transported through space by means of electromagnetic waves.
We expect energy ow in the direction of propagation of the wave,

E
B, as
illustrated in Figure 28. In this lecture, we will derive an expression for the
rate of the energy ow with a traveling EM wave. We will use the Maxwell
equations to derive this expression, which at the same time will show that
the Maxwell equations are consistent with the law of conservation of energy.
Figure 28: Direction of the

E

B vec-
tor of the eld components of a plane
wave.
We know from the circuit theory that
the power (energy) ow is related to the
product of voltage and current
P = V I .
We will show that in the eld theory,
where everything is expressed in terms
of the electric and magnetic eld ampli-
tudes, the power ow across an element
of area d
S is given by
c
2
E

B d
S ,
where S is a closed surface bounding a volume V containing a source of
an EM eld.
If the eld propagates in the direction determined by the cross product

E
B,
consider the source of the eld, i.e. consider the expression
(
E

B) . (24)
If we employ the vector identity
(
E

B) =

B

E

E

B
and use the Maxwell equations III and IV:

E =
B
t
and

B =
0
J +
0
E
t
,
99
we then may write Eq. (24) as
(
E

B) =
B
t

0
E

J
0
E
t
,
or in the form
1
0
(
E

B) =
E

J

t
1
2
0
E
2
+
1
2
B
2
.
On the left-hand side, we put 1/
0
=
0
c
2
and integrate the equation over
some closed surface S enclosing a volume V . Then, we obtain
0
c
2
(
E

B) dV =
E

J dV

t
1
2
0
E
2
+
1
2
B
2
dV . (25)
Now we apply Gausss theorem to the left-hand side of the above equation,
to convert the volume integral into the closed surface integral, and nd
0
c
2
(
E

B) ndS Energy ux
=
E

J dV Rate of doing work by eld on the current
1
2
0
E
2
+
1
2
B
2
dV Field energy. (26)

In words, this equation states that the instantaneous ow of power
across a closed surface S is equal to the time rate of decrease of the
energy stored in the eld in the interior of S plus the power loss due
to the work on current

J, conducting charges, within the volume V .
The above interpretation of the three terms in this equation can be justied
in the following way.
100
8.1 Rate of Doing Work by the Field on the Current
Ohmic Heating
Consider rst the second term:
E

J dV . We will prove that this term
is equal to the rate of doing work by the electric eld on the current that,
in other words, it is the ohmic heating. Simply, it shows that the power
produced by a surce contained inside the volume can be converted into the
ohmic power dissipated inside the volume.
We start our calculations from the circuit theory where we express all the
circuit variables in terms of the eld variables.
Figure 29: Illustration of a current I
owing through a resistive medium of a
length and cross section A.
From the circuit theory, we know that
in a resistive medium, as shown in Fig-
ure 29, the Ohms Law is: V = IR and
R = 1/A, where 1 is the resistivity
of the medium, I is the current owing
through the medium and A is the area
through which the current is owing.
Thus, we can write the current as
I = V/R = (V A)/(1) .
Since

E = V , so that E = V/ and
then we can nd the current density
through the area A, and write it in terms of the eld variable E as
J = I/A =
1
1
E .
Dening the conductivity = 1/1, the current density can be written as
I/A = J = E .
Having the current density expressed in terms of the eld variable E, we
can calculate the rate of conversion (dissipation) of the electromagnetic eld
energy into heat
H = IV =
V
2
R
=
E
2
2
1
A
= E
2
A = E
2
1 ,
101
where 1 = A is the volume of the resistive medium. Hence, the energy
dissipated per unit volume is
H
1
= E
2
= E J =

E

J .
Thus,

E

J is the rate of heating per unit volume in this case.
8.2 Electrostatic Field Energy Density
We now show that the term

V
1
2
0
E
2
dV represents energy contained in the
electric eld E, so that
V
1
2
0
E
2
dV
is the time rate of decrease of the energy stored in the electric eld.
Again, we start from the circuit theory from which we know that capacitors
are devices where electric eld energy can be stored. Thus, consider the work
required to charge a capacitor of a capacitance C to a voltage V
W =
1
2
CV
2
.
According to the eld theory of electromagnetism this work done corresponds
to conversion from some other form of energy into electrostatic eld energy
c = W =
1
2
CV
2
and C =

0
A
d
.
Since the electric eld in the capacitor is given by E = V/d, we obtain
c =
1
2
0
A
d
E
2
d
2
=
1
2
0
E
2
1 ,
where 1 = Ad is the volume of the capacitor. Hence
c
1
=
1
2
0
E
2
is the energy per unit volume contained in or carried out by the electric
eld E.
102
8.3 Magnetostatic Field Energy Density
We now show that the term

V
B
2
/(2
0
)dV represents energy contained in
the magnetic eld B, so that
V
1
2
B
2
0
dV
is the time rate of decrease of the energy stored in the magnetic eld.
As before, we refer to the circuit theory from which it is well known that a
solenoid is a device where the magnetic eld energy can be stored. Thus,
the work required to energize an inductor of inductance L to a current I is
c =
1
2
LI
2
and that the magnetic eld within a long solenoid of self-inductance
L =
0
n
2
A is B =
0
nI. According to the eld theory of electromagnetism
this work done corresponds to transformation of some other form of energy
to magnetic eld energy.
Thus, for a solenoid of length and area of cross section A we have energy
c = W =
1
2
LI
2
=
1
2
0
n
2
AI
2
=
1
2
(
0
nI)
2
A
1
0
=
1
2
B
2
0
1 ,
oand then energy per unit volume is
c
1
=
1
2
B
2
0
.
This is the energy per unit volume contained in or carried out by the mag-
netic eld B.
8.4 Poynting Vector
Since the right-hand side of Eq. (26) represents the rate of increase of the
electric and magnetic energies stored in the EM eld and substracted by
the ohmic power dissipated as heat, the left-hand side must be equal (in
103
consistency with the law of conservation of energy) to the power owing into
the volume through its surface. Thus, the expression
N =
0
c
2
(
E

B)
is a vector representing the power ow per unit area, and is referred to as
the Poynting vector.
It represents the energy ux in the electromagnetic eld, i.e. the energy ow
per unit area (measured normal to the ow) per unit time.
The quantity N has dimension of power per square meter. It is easy to see:
E has the dimension of volts per meter,
0
c
2
B has dimension of amper per
meter. Thus, the Poynting vector has dimension volts amper/(square me-
ter) = Watts/(square meter).
The energy ow equation can be converted into the form of a dierential
continuity equation or energy conservation law. From Eq. (25), we have
U
t
+

N =
J

E ,
where
U =
1
2
0
E
2
+
1
2
B
2
0
(27)
is the energy density of the EM eld.
The physical meaning of the dierential continuity equation is that the time
rate of change of electromagnetic energy within a certain volume, plus the
energy owing out through the boundary surface of the volume is equal to
the negative of the total work done by the elds on the source inside the vol-
ume. Thus.

J

E is a conversion of electromagnetic energy into heat energy.
Equation (27) shows that we consider the energy stored in electric and mag-
netic elds as distributed throughout the region of space where these elds
are present with densities
0
E
2
/2 and B
2
/2
0
, respectively.
104
8.5 Phase Relationships in EM Waves
Here, we will show that only the in-phase components of

E and

B contribute
to net energy ow averaged over a whole cycle of the radiation. If somehow,
the elds would oscillate with dierent phases, no energy can be transported.
Consider orthogonal elds

E and

B oscillating in phase. Thus, if

E =
E
0
cos(t)
i and

B = B
0
cos(t)
j then:
N =
0
c
2
E

B =
0
c
2
E
0
B
0
cos
2
(t)

k .
Since, cos
2
(t) =
1
2
, we obtain
N =
1
2
0
c
2
E
0
B
0
=
0
c
2
E
rms
B
rms
,
where E
rms
= E
0
/
2 and B
rms
= B
0
/
2.
Hence, the average Pointing vector is dierent from zero indicating that the
orthogonal elds

E and

B oscillating in phase can transport energy.
If, however,

E = E
0
cos(t)
i and

B = B
0
sin(t)
j then:
N =
0
c
2
E

B =
0
c
2
E
0
B
0
cos(t) sin(t)

k .
Since, cos(t) sin(t) = 0, we have N = 0.
Hence, the average Pointing vector is zero indicating that the orthogonal
elds

E and

B oscillating out of phase phase cannot transport energy.
In summary, electromagnetic waves in which the

E and

B elds oscillate
in phase can transport energy.
8.6 Momentum Flux
It should be noted here, that EM waves can transport not only energy but
also momentum.
105
To obtain an expression for the momentum carried by the electromagnetic
eld, we may employ the relativistic energy-momentum relationship
c
2
= p
2
c
2
+ m
2
0
c
4
.
Since for electromagnetic radiation m
0
= 0, we obtain
p =
c
c
.
Thus the momentum ux of the electromagnetic eld is:
=

0
c
2
E

B
c
=
0
c
E

B ,
which leads to the conclusion that electromagnetic waves transport not only
energy, whose the ow is given by the Poynting vector, but also momentum.
8.7 EM Energy Flow: Circuit versus Field Theory
We now consider simple examples illustrating applications of the Poynting
vector to some familiar circuit problems to show how the eld theory provides
an alternative (surprising!) way of viewing the energy transfer through the
circuits.
8.7.1 Energy Flow in a Resistive Wire
As a rst example, consider a wire (resistor) of length , carrying a current I,
as shown in Figure 30.
Figure 30: A resistive wire of length
carrying a current I.
Let V is a potential dierence ap-
plied along a resistive wire. We cal-
culate the power dissipated in the
wire using the circuit theory, and next
will compare the formalism with that
used in the eld theory. We are
particularly interested in the predic-
tion of the two theories of the direc-
tion of energy ow through the cir-
cuit.
106
Circuit theory calculation:
According to circuit theory the power dissipated in the wire is
P = V I = I
2
R ,
where I is the current owing through the wire and R is the resistance of
the wire. Since R is proportional to the length of the wire, a more power
(energy) is dissipated for a longer wire. Thus, according to the circuit theory,
energy ows along the wire.
Field theory calculation:
Let us now look at the same problem from the point of view of the eld
theory. According to electromagnetic eld theory, the direction of the ow
of energy is described by the Poynting vector

N =
0
c
2
E

B.
Figure 31: Directions of the electric and
magnetic elds produced by a wire and the
direction of the resulting Poynting vector.
In order to nd the direction of the
Poynting vector, we calculate the di-
rections of propagation of the electric
and magnetic elds produced by the
current. The electric eld propagates
along the wire, as shown in Figure 31,
and is given by
E = V (z) =
V (z)
z .
From the Amperes line integral the-
orem, we nd that the the magnetic
eld propagates around the wire, and at points distant a from the wire is
B =

0
I
2a
.
Hence, the Poynting vector is
N =
0
c
2
E

B =
0
c
2
V
0
I
2a
r =
V I
2a
r ,
where we have used the relations z

= r and
0
0
c
2
= 1.
107
Thus, the eld theory predicts that energy ows into the wire from the air
not along the wire. The energy is in the elds, the wire provides boundary
conditions and guides the elds.
The student may argue that the EM eld carries only a small part of the
energy dissipated in the wire. However, the total rate at which eld energy
ows into the wire by crossing the surface of the wire is given by
N d
S =

N r dS =
N dS
side
= N
dS
side
=
V I
2a
2a = V I ,
(
N d
S on the ends of the cylinder), which is in agreement with the re-

sult of the circuit theory. This result demonstrates quantitatively that all the
power which heats the resistor enters through the sides not through the wires.
If the energy is in the elds, it means that the electromagnetic energy goes
out of a battery into the air, and then goes into the wire from the air. This
is exactly the case we will show in the following example.
8.7.2 Energy Flow out of Battery
In the above example, we have shown that according to the eld theory, the
energy enters the resistor from the air.
Figure 32: Directions of the current

J
and the electric eld

E in a battery.
Then, a question arises: If the en-
ergy enters the resistor from the air,
how does the energy get out to the
air from a source of energy (bat-
tery)?
To answer this question, consider a bat-
tery which provides energy to the resis-
tor. As illustrated in Figure 32, inside
the battery

J and

E are in opposite di-
rections. The magnetic eld circulates
around the battery, so we see that the
Poynting vector

N points out into the air,
not along the wire.
108
Thus, the battery sends energy out into the air, not along the wire.
8.7.3 Propagation of an Electromagnetic Wave along a Wire
One of the above examples showed that eld energy ows into a wire so that
it can be dissipated as heat. Consider now a dierent situation that one
would like to transmit an electromagnetic wave through a resistive wire.
Figure 33: Directions of the electric and
magnetic elds around a wire transmitting
an EM wave of wavelength .
It is well known from experiments,
that an electromagnetic wave can be
transmitted along the wire with very
little loss of its energy. Why does it
happen?
This eect has a simple expla-
nation in terms of the EM the-
ory.
If we consider the case of a perfect
conductor we nd that there will be a
current wave along the wire with sur-
face charges induced producing a ra-
dial electric eld. Then

E

B is parallel to the wire and the eld descrip-
tion of energy transmission is that it is transmitted in the space around the
wire. In the space around the wire

E and

B are in phase and

N is always in
the same direction. Within a perfect conductor we will show later in Chap-
ter 12.6 that

E and

B are /2 out of phase so the mean N averaged over a
cycle is zero. Thus, there is no net energy transmission within the perfectly
conducting wire.
Exercise in class: Energizing of a capacitor
Consider the energizing of a plane parallel capacitor with
circular plates. Show that circuit and eld calculations
agree as to the rate of energizing of the capacitor, i.e.
P
c
= P
f
where:
109
P
c
= V I = rate of doing work (by current I and voltage V
between the plates) in charging the capacitor according to
circuit theory.
P
f
=

S
0
c
2
B d
S = rate of energy ow into the surface

of the capacitor according to eld theory.
From which direction does the energy ow in to the ca-
pacitor according to eld theory?
We will keep the calculation simple assuming the plates
of the capacitor to be uniformly charged. Under what con-
ditions is this assumption likely to be true?
Exercise in class: No Fluxes from Static Fields
Consider a source of electrostatic eld

E and magnetostatic
eld

B. If it is arranged so that

E

B, one may ask:
Should we expect to see an energy ux of
0
c
2
E

B?
110
Revision questions
Question 1. Derive formula for the electrostatic eld energy density.
Question 2. Derive formula for the magnetostatic eld energy density.
Question 3. State the denition of the Poynting vector.
Question 4. In a circle containing a resistor, what is the direction of the energy
ow according to the circuit theory and the eld theory?
Tutorial problems
Problem 8.1 Poynting vector on the surface of a long wire
Find the direction and the magnitude of the Poynting vector on
the surface of an innitely long, straight conducting wire of ra-
dius r and conductivity , carrying a current I.
Problem 8.2 Energy ow in a plane wave
A plane electromagnetic wave propagates in the z direction in
a vacuum. Its electric eld is polarized in the x direction and so
has the form:
E = E
0
sin(t kz)
i
(a) Find an expression for the vector energy ux in this wave.
(b) On the same time scale plot the electric eld strength and
energy ux magnitude versus time at position z = 0 for 2 cycles
of the wave.
(c) What is the mean energy ux averaged over cycles?
111
9 Solution of Laplaces Equation and Bound-
ary Value Problem
In one of the previous lectures, we discussed an advantage of using the scalar
and vector potentials in the calculation of the electric and magnetic elds. In
this lecture, we will illustrate applications of the scalar potential to physical
problems involving bounded elds.
There is a class of problems in electromagnetism in which a eld can be
derived without involvement of the complete set of the Maxwells equations.
This can be done with the help of the gradient of a scalar potential which
satises Laplaces equation
2
= 0 .
Proof:
The condition for this to happen is that a vector eld

F has vanishing diver-
gence and curl

F = 0 and

F = 0 . (28)
Since,
F = 0, we can always write

F as

F = , where is an arbitrary
scalar function. Then,

F = 0 means that
() =
2
= 0 .
Thus, the scalar potential contains all the necessary information to com-
pletely specify the eld of the properties (28).
Example 1: Electrostatic problems involving Laplaces equation
Since in general
E = /
0
and
E =
B we see that the requirement

for Laplaces equation to be relevant is that = 0 and /t = 0, i.e. a source-
free region and static conditions. Of course there must be a source of charges
somewhere or there would be no eld anywhere. The typical situation where
solution of Laplaces equation is relevant in electrostatic is where we have
source-free non-conducting regions between statically charged conductors.
112
Example 2: Magnetostatic problems involving Laplaces equation
Since in general
B = 0 and
B =
0
J +
1
c
2
E, we see that the require-

ment for Laplaces equation to be relevant is that

J = 0 and /t = 0, i.e.
a current free region and static conditions. Again, there must be currents
somewhere or there would be no elds anywhere. The typical situation is
to be calculating the magnetic eld in the non-conducting region between
constant currents.
9.1 Uniqueness of the Solution of Laplaces Equation
As we shall see below, general solutions of Laplaces equation are in terms of
some constants which are usually found from boundary (initial) conditions
for a given problem. A question arises:
What boundary conditions are appropriate for the Laplace equa-
tion to ensure that a unique and well-behaved (physically reason-
able) solution will exist inside the bounded region?
Our experience leads to believe that specication of the potential on a closed
surface denes a unique potential problem. This is called Dirichlet theorem
or Dirichlet boundary conditions.
9.1.1 Dirichlet Theorem
Consider a volume V completely bounded by a closed surface S. Within S
there is a potential satisfying
2
= 0. The Dirichlet theorem says that
the value of inside the volume is uniquely determined if the value of the
potential is specied everywhere on the whole boundary.
Proof:
Suppose, to the contrary, that there exist two solutions
1
and
2
satisfying
the same boundary condition, i.e.
2
1
= 0 and
2
2
= 0 within S, but
1
=
2
on S.
113
Let U =
1

2
is the dierence between the solutions. Since
1
and
2
are known to be solutions of the Laplace equation, then from the linearity
of the
2
operator
2
U = 0, i.e. U is also a solution of the Laplace equation.
We will prove that U = 0 inside the volume. To show this, we introduce a
vector

F = UU. Then using the vector property that

F = UU = U (U) +U U ,
and the Gauss Divergence Theorem, we get

FdV =
V
U UdV +
V
U UdV
=
V
U
2
UdV +
V
(U)
2
dV
=
S
UU d
S .
Now, the right-hand side of the above equation is equal to zero because U = 0
over S. Also, the integral
V
U
2
UdV = 0 ,
because
1
and
2
both satisfy the Laplace equation throughout V . Hence
V
(U)
2
dV = 0 .
Since the integral of a positive function is always positive, U must be zero
for the integral to be zero. Thus U = 0 and consequently, inside the vol-
ume V , the function U is constant. Since U = 0 on S, so that inside V , we
have then that
1
=
2
everywhere, as required.
9.2 Solutions of Laplaces Equation
There are dierent methods of solving the Laplace equation
Method of Images
114
Green functions method
Variational method
Method of lattices
Numerical Monte-Carlo simulations method
Method of separation of variables
Solution in spherical coordinates
We will illustrate last two methods which can be applied to a large class of
problems in electromagnetism. The other methods can be applied to spe-
cic problems. For these methods it is necessary that the boundaries over
which the potential is specied coincide with the constant bounding surfaces.
9.2.1 Method of Separation of Variables
In cartesian coordinates the Laplace equation for the scalar potential can be
written as
2
=

2
x
2
+

2
y
2
+

2
z
2
= 0 . (29)
Since x, y, z are independent variables, the solution of the Laplace equation
is of the form
(x, y, z) = X(x)Y (y)Z(z) .
Substituting this into the Laplace equation and dividing both sides by XY Z,
we obtain
1
X
d
2
X
dx
2
+
1
Y
d
2
Y
dy
2
+
1
Z
d
2
Z
dz
2
= 0 .
This equation can be separated into three independent equations. To show
this, we write this equation as
1
X
d
2
X
dx
2
=
1
Y
d
2
Y
dy
2

1
Z
d
2
Z
dz
2
.
115
Both sides of the above equation depend on dierent (independent) variables,
thus are equal to a constant, say
2
:
1
X
d
2
X
dx
2
=
2
,
1
Y
d
2
Y
dy
2

1
Z
d
2
Z
dz
2
=
2
.
The second equation can be written as
1
Y
d
2
Y
dy
2
=
2
1
Z
d
2
Z
dz
2
.
Again, both sides depend on dierent variables, thus are equal to a constant,
say
2
:
1
Y
d
2
Y
dy
2
=
2
,
1
Z
d
2
Z
dz
2
=
2
.
Hence, after the separation of the variables, we get three independent ordi-
nary dierential equations
1
X
d
2
X
dx
2
+
2
= 0 ,
1
Y
d
2
Y
dy
2
+
2
= 0 ,
1
Z
d
2
Z
dz
2
(
2
+
2
) = 0 .
The solutions of these equations depend on whether
2
and
2
are positive
or negative constants. If we choose
2
and
2
to be positive, the solutions
of the dierential equations for X and Y are in the form of trigonometric
functions whereas the solution for Z is in the form of a hyperbolic function
X(x) =
A
k
e
ix
+ B
k
e
ix
,
Y (y) =
C
l
e
iy
+ D
l
e
iy
,
Z(z) =
E
p
e
2
+
2
z
+ F
p
e
2
+
2
z
.
116
If
2
and
2
had been chosen as negative constants, the hyperbolic and
trigonometric solutions would be interchanged.
The solutions can be equally well written in the form
X(x) =
k
[A
k
sin(x) + B
k
cos(x)] ,
Y (y) =
l
[C
l
sin(y) + D
l
cos(y)] ,
Z(z) =
E
p
sinh
2
+
2
z
+ F
p
cosh
2
+
2
z
.
The above solutions of the Laplaces equation are in a general form valid for
an arbitrary problem. The constants , , A
k
, B
k
, C
l
, D
l
, E
p
and F
p
are found
from specic boundary conditions.
Consider two examples of an application of the general solutions:
1. We have a solution with known boundary conditions, nd the problem.
2. We have a problem with specic boundary conditions, nd the solution.
Example 1.
Consider the following two-dimensional solution of the Laplaces equation
(x, z) = X(x)Z(z) = V
0
sin(x) sinh(z) (30)
with the lower boundary
min
= 0 and the upper boundary
max
= V
0
.
In what circumstance is the above the solution?
Consider (x, z) in some limits. Note that X = 0 for x = 0 or x = , i.e.
x = /.
The solution thus satises the boundary conditions along the vertical lines
for = /b.
Since sinh(z) = 0 for z = 0, the boundary condition along the lower bound-
ary is satised.
117
Figure 34:
For the solution to satisfy the upper
boundary condition, the shape of the up-
per boundary must be such that
V
0
sin(x) sinh(z) = V
0
,
for all points x, z on the line, i.e.
sin
x
b
sinh
z
b
= 1 .
Since sin
x
b
0 at the edges, it is
greatest (= 1) at the center x = b/2.
Hence, sinh
z
b
must be equal to one at x = b/2. This happens when

z
b
= arc sinh(1) 0.885 ,
from which we nd
z =
0.885b
= 0.282b .
Figure 34 shows the shape of the box inside which the potential is of the
form (30).
Example 2.
We usually have reverse problems to the above that we have a set of elec-
trodes which constitute equipotential lines or surfaces, and need to nd the
appropriate solution of the Laplace equation. In this example we will illus-
trate this situation and we will try to nd potential inside a rectangular box
whose three sides have potential equal to zero, and the remaining side has a
potential V
0
.
Consider a two-dimensional problem with boundary conditions shown in Fig-
ure 35. This two-dimensional problem has a general solution
(x, z) =
n
[A
n
sin(x) + B
n
cos(x)]
[E
n
sinh(z) + F
n
cosh(z)] .
118
Figure 35:
The boundary condition of = 0
at the side x = 0 can be sat-
ised by B
n
= 0. The bound-
ary condition of = 0 at the
side z = 0 can be satised by F
n
=
0. In order to have = 0
at x = b, we must have b =
n.
Hence with the boundary conditions
along three sides of = 0, the general
solution reduces to
(x, z) =
n=1
K
n
sin
nx
b
sinh
nz
b
, (31)
where K
n
= A
n
E
n
.
To nd K
n
we apply the remaining boundary condition that = V
0
at z = a:
(x, a) = V
0
=
n=1
K
n
sin
nx
b
sinh
na
b
. (32)
This is a Fourier series in x and in the usual way we use the orthogonality
properties of sine functions to calculate K
n
:
2
0
sin(m) sin(n) d =
0 for m = n
for m = n
2
0
cos(m) cos(n) d =
0 for m = n
for m = n
2
0
sin(m) cos(n) d = 0 for all m and n .
Thus, multiplying both sides of Eq. (32) by sin(mx/b) and integrating over
x = 0 b, we get
b
0
V
0
sin
mx
b
dx =
n=1
K
n
sinh
na
b
b
0
sin
mx
b
sin
nx
b
dx .
119
From the orthogonality of the sine functions, we nd that all integrals on the
right-hand side are equal to zero except for m = n. Thus, after performing
the integration on the left-hand side, we are left with the term
V
0
cos
nx
b
b
0
b
n
= K
n
sinh
na
b
b
0
sin
2
nx
b
dx ,
which we can write as
V
0
b
n
[1 cos(n)] = K
n
sinh
na
b
b
0
1
2
1 cos
2nx
b
dx .
The cos (2nx/b) integrates to zero over the range 0 b, so that the above
equation becomes
V
0
b
n
[1 cos(n)] = K
n
sinh
na
b
b
2
,
from which we nd the constant K
n
:
K
n
=
2V
0
n
1 cos(n)
sinh
na
b
.
The solution for K
n
can be further simplied:
Namely, if n is an even number, cos(n) = 1 and then K
n
= 0.
If n is an odd number, cos(n) = 1, and then 1 cos(n) = 2. Hence, we
nally arrive at
K
n
=
4V
0
n
1
sinh
na
b
, for odd n
and K
n
= 0 for even n.
Substituting K
n
into Eq. (31), we nd that the potential inside the box has
the form
(x, z) =
odd n
4V
0
n
sinh
nz
b
sinh
na
b
sin
nx
b
.
120
As an exercise, check the correctness of this solution by showing that it
satises the Laplaces equation.
Exercise in class: Reduction of the solution of a three dimesional problem
to a two dimensional problem
Why for a two dimensional problem we simplify the general
three-dimensional solution of the Laplace equations into
= X(x)Z(z) ,
or
= Y (y)Z(z) ,
but never
= X(x)Y (y) .
121
9.2.2 Solution of the Laplace Equation in Spherical Coordinates
In this lecture, we continue the discussion of boundary-value problems and
will illustrate solution of the Laplace equation for general problems of spher-
ical symmetry. In this case, we shall work in spherical coordinates (r, , ),
in which the Laplace equation takes the form
2
=
1
r
2
r
2
+
1
r
2
sin
sin
+
1
r
2
sin
2
2
= 0 .
The procedure we follow in the solution of the above equation is as follows:
Multiplying both sides by r
2
, the Laplace equation can be written as a sum
of two separate parts
r
2
+
1
sin
sin
+
1
sin
2
2
= 0 .
The rst part depends solely on r, whereas the second part depends solely on
the angles , . Thus, the solution of this equation is of the separable form
= R(r)Y (, ) .
Hence, substituting = R(r)Y (, ) and dividing both sides by R(r)Y (, ),
we obtain
1
R
d
dr
r
2
dR
dr
=
1
Y
1
sin
sin
Y
+
1
sin
2
2
Y
.
Note, both sides of the above equation depend on dierent variables. It
follows that the both sides must be equal to the same constant, say :
d
dr
r
2
dR
dr
+ R = 0 ,
1
sin
sin
Y
+
1
sin
2
2
Y
2
Y = 0 .
Thus, the Laplace equation splits into two independent dierential equations.
We will call them the radial part and angular part, respectively.
122
Angular part: Equation for Y .
Multiplying both sides by sin
2
, we get
sin

sin
Y
sin
2
Y +

2
Y
2
= 0 .
This equation contains two separate parts, one dependent only on and the
other dependent only on . Therefore, the solution will be of the form
Y (, ) = X()() .
Hence, we get
1
X
sin
d
d
sin
dX
d
sin
2
=
1
d
2
d
2
.
As before, both sides must be equal to a constant, say m
2
:
1
X
sin
d
d
sin
dX
d
sin
2
= m
2
,
1
d
2
d
2
= m
2
.
First, we will solve the equation for , which we can write as
d
2
d
2
= m
2
.
It is the familiar dierential equation for a harmonic motion. The solution
of this equation is of simple exponent form:
() = Aexp(im) ,
where A is a constant.
To determine the constant m note that in rotation, and +2 correspond
to the same position in space: () = ( + 2), which is satised when
exp(im) = exp[im( + 2)] .
123
This leads to
exp(i2m) = 1 ,
that is satised when m = 0, 1, 2, . . ..
Hence, the constant m
2
is not an arbitrary number, is an integer.
Now, we will nd X() to complete the solution for Y .
Using the solution for , the dierential equation for X() can be written as
1
sin
d
d
sin
dX
d
+
m
2
sin
2
X = 0 .
Introducing a new variable z = cos , we can rewrite this equation as
1 z
2
d
2
X
dz
2
2z
dX
dz

+
m
2
1 z
2
X = 0 ,
or
d
dz
1 z
2
dX
dz
+
m
2
1 z
2
X = 0 . (33)
The above equation is known as the generalized Legendry dierential equa-
tion, and its solutions are the associated Legendry polynomials. For m = 0,
the equation is called the ordinary Legendry dierential equation whose so-
lution is given by the Legendry polynomials.
Lets look into the solution procedure of the above equation. This will allow
us to nd and X().
We assume that the whole range of z (cos ), including the north and south
poles (z = 1), is in the region of interest. The desired solution should be
single valued, nite, and continuous on the interval 1 z 1 in order to
represent a physical potential.
124
The dierential equation for X has poles at z = 1. In order to nd the
solution of this equation, we rst check what solution could be continuous
near the poles.
Lets check a possible solution near z = 1. Substituting x = 1 z, then
dx = dz and in terms of x the equation takes a form
d
dx
x (2 x)
dX
dx
+
m
2
x(2 x)
X = 0 .
We look for a solution in the trial form of power series in x
X(x) = x
s
n=0
a
n
x
n
.
Substituting this into the dierential equation for X, we get
2s
2
a
0
x
s1
+ (s + 1)(2sa
1
sa
0
+ 2a
1
)x
s
+ . . .
+
m
2
x(2 x)
(a
0
+ a
1
x + . . .)x
s
= 0 .
Near x 0, we can replace x(2 x) by 2x, and obtain
2s
2
a
0
m
2
2
a
0
x
s1
+ (. . .)x
s
. . . = 0 .
This equation is satised for all x only if the coecients at x
s
, x
s1
, . . . are
zero. From this, we nd that
s =
1
2
[m[ .
We take only s = +
1
2
[m[ as for s =
1
2
[m[ the solution for X(x) at x = 0
would go to innity. We require the solution to be nite at any point x.
Thus, the solution that is continuous near x = 0 is of the form
X(x) = x
1
2
|m|
n=0
a
n
x
n
,
125
or in terms of z
X(z) = (1 z)
1
2
|m|
n=0
a
n
z
n
.
Using the same procedure, we can show that near the pole z = 1, the
continuous solution is
X(z) = (1 + z)
1
2
|m|
n=0
a
n
z
n
.
Hence, we will try to nd the solution in the form
X(z) =
1 z
2
1
2
|m|

n=0
b
n
z
n
. (34)
What left is to determine the coeciets b
n
.
Substituting Eq. (34) into the dierential equation for X(z), Eq. (33), and
collecting all terms at the same powers of z
n
, we obtain
n
(n + 1)(n + 2)b
n+2
n(n 1)b
n
2([m[ + 1)nb
n
( +[m[ + m
2
)b
n
z
n
= 0 ,
from which we nd a recurrence relation for the coecients b
n
b
n+2
=
(n +[m[)(n +[m[ + 1) +
(n + 1)(n + 2)
b
n
.
We have two separate solutions for even and odd n. For b
0
= 0, we put
b
1
= 0, and the solution is given in terms of even n. For b
0
= 0, we put
b
1
= 0, and the solution is given in terms of odd n.
We cannot accept both the even and odd solutions at the same time, because
in this case the solution X(z) would not be a single valued function that is
would not be accepted as a physical potential.
For example, for b
0
= 0, we have = [m[ m
2
, but for b
1
= 0, we have
= 2 3[m[ m
2
. If we would accept both of the solutions at the same
time, the potential would have two dierent values.
126
We check now whether the series is converting when n which would
ensure that the potential is nite.
Since b
n+2
> b
n
, the series diverges for z = 1. Therefore, in order to get
the potential nite everywhere in the space, we have to terminate the series
at some n = n
0
. In other words, we assume that b
n
0
+1
= b
n
0
+2
= . . . = 0.
The series terminating at n = n
0
indicates that
(n
0
+[m[)(n
0
+[m[ + 1) + = 0 .
Introducing
l = n
0
+[m[ ,
we see that l [m[, and
= l(l + 1) , l = 0, 1, 2, . . .
Thus, the solution for X(z) is
X
lm
(z) =
1 z
2
1
2
|m|
l|m|
n
b
n
z
n
,
where the sum is over even n when l [m[ is an even number, and over odd n
when l [m[ is an odd number.
First few solutions
X
00
(z) = b
0
= b
0
P
0
0
(z) ,
X
10
(z) = b
1
z = b
1
P
0
1
(z) ,
X
11
(z) = b
0
1 z
2
= b
0
P
1
1
(z) ,
where P
0
0
(z) = 1, P
0
1
(z) = z, P
1
1
(z) =
1 z
2
, . . . are the associate Legendry
polynomials of the order l.
Useful examples of the associate Legendry polynomials, written in terms
of , (z = cos ):
P
0
0
(cos ) = 1 ,
127
P
0
1
(cos ) = cos ,
P
1
1
(cos ) = sin ,
P
0
2
(cos ) =
1
4
[3 cos(2) + 1] ,
P
1
2
(cos ) =
3
2
sin(2) ,
P
2
2
(cos ) =
3
2
[1 cos(2)] .
An important property of the Legendry polynomials is orthogonality that
1
1
P
m
l
(cos )P
n
k
(cos ) d(cos ) = 0 for m = n and l = k ,
1
1
[P
m
l
(cos )]
2
d(cos ) =
2
2l + 1
(l + m)!
(l m)!
for m = n and l = k .
Finally, with the above notation, the solution for the angular part of the
Laplace equation Y (, ) is of the form
Y (, ) =
m
A
lm
P
m
l
(cos )e
im
.
Radial part: Equation for R(r).
With = l(l + 1), the dierential equation for R takes the form
d
dr
r
2
dR
dr
l(l + 1)R = 0 .
Dividing by r and introducing a new function U(r) = rR(r), we obtain
d
2
U
dr
2

l(l + 1)
r
2
U = 0 .
Lets rst check the asymptotic solution for r 1. In this case we can ignore
the second term in the dierential equation, and nd that the asymptotic
equation has a solution U(r 1) = Cr, where C is a constant.
Following this asymptotic behavior, we will try the general solution of a form
U(r) = r
s
.
128
Substituting this into the dierential equation, we obtain
[s(s 1) l(l + 1)] r
s2
= 0 .
This equation is satised for any r when
s = (l + 1) or s = l .
Thus, the general solution for U that satises the asymptotic solution is of
the form
U(r) = C
1
r
l+1
+ C
2
r
l
,
and then
R(r) = C
1
r
l
+ C
2
r
(l+1)
.
Hence, the general solution of the Laplace equation in spherical polar coor-
dinates is of the form:
(r, , ) =
C
1l
r
l
+ C
2l
r
(l+1)
A
lm
P
m
l
(cos )e
im
.
The solution can be written as
(r, , ) =
C
1l
r
l
+ C
2l
r
(l+1)
[a
lm
cos(m) + b
lm
sin(m)] P
m
l
(cos ) .
This is the nal solution for the potential of an arbitrary spherically sym-
metric problem. The potential is nite, determined and is continuous at each
point (r, , ).
Example 1: Potential outside a conducting sphere
Consider an example of boundary-value problem with azimuthal symmetry:
A conducting sphere of a radius a in an uniform electric eld, as shown
in Figure 36.
129
In order to nd the potential, we rst have to determine boundary conditions
for the potential.
There are two boundaries: One of the boundaries is the surface of the sphere
and the other is the region at r .
The conducting sphere is an equipotential volume (else there would be electric
elds driving currents till it became so). We take, with no loss of generality,
the zero potential on the sphere.
Figure 36: A dielectric sphere in a
uniform electric eld.
Thus, the boundary conditions to be sat-
ised are:
1. The potential on the surface of the
sphere (a, , ) = 0.
2. The potential at innity is the uni-
form eld potential (no eect of the
sphere at innity), so =
E r =
Er cos at innity.
Since the applied potential is indepen-
dent of the angle , the induced poten-
tial will also be independent of . This
is a general property of the potential in-
side a bounded area that we shall ex-
plain in details at one of the tutorial se-
sions.
Thus, using this property, we can set m = 0 in the general solution and get
(r, ) =
C
1l
r
l
+ C
2l
r
(l+1)
P
0
l
(cos ) .
At innity (r ), the boundary condition is satised for all constants
C
1l
= 0 except for l = 1 (remember P
1
= cos ). Thus
(r, ) = C
11
rP
1
(cos ) +
l
C
2l
P
l
(cos )
r
l+1
.
130
We rst nd the explicit form of C
11
. As r , the potential (r, )
C
11
r cos = Er cos . Therefore C
11
= E.
The remaining coecients C
2l
we nd from the other boundary condition
that (a, ) = 0 on the surface of the sphere. Thus, at r = a
(a, ) = 0 = EaP
1
(cos ) +
l
C
2l
P
l
(cos )
a
l+1
.
We can determine the coecients C
2l
using the orthogonality properties of
the Legendry polynomials. Multiplying the above equation by P
k
(cos ) sin
and integrating over sin d = d(cos ), we obtain
0 = aE
1
1
P
1
(cos )P
k
(cos ) d(cos )
+
l
C
2l
a
l+1
1
1
P
l
(cos )P
k
(cos ) d(cos ) .
For k = 1 the rst term vanishes by orthogonality of the Legendry polyno-
mials. Moreover, all terms in the summation vanish except that for l = k.
Thus, for k = 1
0 =
C
2k
a
k+1
1
1
P
k
(cos )P
k
(cos ) d(cos ) .
Since the integral is nonzero, then C
2k
= 0.
For k = 1 and using the orthogonality property of the Legendry polynomials
1
1
P
m
l
(cos )P
m
l
(cos ) d(cos ) =
2
2l + 1
(l + m)!
(l m)!
,
we nd that for l = 1 and m = 0, the integral is equal to 2/3. Thus
0 = aE
2
3
+
C
21
a
2
2
3
,
from which, we nd C
21
= Ea
3
. Hence
(r, ) = Er cos + Ea
3
cos
r
2
.
131
The rst term is just the potential of a uniform eld E. The second term
is the potential due to the induced surface charges or, equivalently, is the
potential of the induced dipole moment p = 4
0
Ea
3
dip
=
p cos
4
0
r
2
.
This result shows that the sphere placed in an uniform electric eld behaves
as a dipole because of the eect of the charge distribution induced on its
surface.
Exercise in class: Potential inside a sphere
Let us suppose that the student would try to nd in-
side a sphere using the general solution for . We have
found before, using the Gausss law, that = const. in-
side the sphere.
The student immediately notice that the solutions appear
to be mathematically dierent since that found by using the
Gausss law is in a closed form while the general one is in
the form of a series. The question to be settled is whether
the two solutions are identical or whether, perhaps, only
one of the solutions is correct?
132
Revision questions
Question 1. What conditions are imposed on the eld that is completely deter-
mined by a scalar potential?
Question 2. State the Dirichlet theorem.
Question 3. Explain the method of separation of variables on example of the
Laplace equation in cartesian coordinates.
Question 4. Explain the method of solving the generalized Legendry dierential
equation that satises the conditions imposed on the scalar poten-
tial.
133
Tutorial problems
Problem 9.1 Electrostatic potential inside a rectangular box with symmetrical
boundary conditions
Find a series solution in rectangular harmonics for the electrostatic
potential inside a 2-dimensional box in which the sides at y = 0
and y = b are at zero potential and the sides at z = a and z = a
are at potential V
0
.
Problem 9.2 Potential inside a rectangular box
Consider a two-dimensional region with boundaries at x = 0, b
and z = 0, a, as shown in Figure 37. The boundary conditions are
Figure 37:
z
= 0 at z = 0 ,
= 0 at x = 0, b ,
= V
0
at z = a .
Find the potential at any point inside the two-dimensional region.
134
Problem 9.3 Potential inside an open rectangular box
Find the potential everywhere inside a two dimensional region
bounded from three sides by a nite plane at z = 0 and by semi-
innite planes located at x = 0 and x = b, as shown in Figure 38.
The boundary conditions are
= 0 at x = 0 and x = b ,
= V
0
at z = 0 ,
= 0 at z .
Figure 38:
Problem 9.4 Potential inside a 3-dimensional box
A 3-dimensional conducting box is dened by x = 0 to a, y =
0 to b and z = 0 to c. All the sides of the box are connected
together and grounded except that side dened by z = c, which
is insulated from the rest and held at potential V
0
. Show that the
potential within the hollow box can be written as the series
(x, y, z) =

n
odd
m
odd
16 V
0
nm
2
sin
mx
a
sin
ny
b
sinh
m
a
2
+
n
b
sinh
m
a
2
+
n
b
.
135
Problem 9.5 Potential due to a thin circular ring
(a) Show that for a system with axial symmetry such as a ring
of charge, the electrostatic potential can be written as a series in
spherical polar coordinates:
(r, ) =

+B
r
(+1)
(cos ) .
To determine the coecients A
and B
consider the solution along

the z-axis, where = 0. It is useful to know that P
(1) = 1 for
any value of . Thus, along the axis:
(z) =

+B
z
(+1)
.
(b) Calculate (z) directly from the Coulomb potential. Expand
this expression as a power series and hence nd the A
and B
by
comparison.
Hence, show that for a ring of radius a and charge per unit length
the potential everywhere can be written as the series:
(r, ) =

2
0
1 +

even
(1)
2
1.3.5 . . . ( 1)
2
! a
(cos )
,
where = 2, 4, 6, . . . (even).
The Legendre functions can be computed from the recurrence re-
lation:
( + 1) P
+1
(x) = (2 + 1) xP
(x) P
1
(x)
and knowing that P
0
(x) = 1 and P
1
(x) = x.
However, the functions become rather tedious for hand calcula-
tions for large , e.g.
P
10
(x) =
1
1024
252 + 13860x
2
120120x
4
+ 360360x
6
437580x
8
+ 184756x
10
.
136
However, there is Matlab etc.
(c) Calculate also the electric eld components of the circular
ring using the relation

E = .
(There is a useful recurrence relation for the derivative of the Leg-
endre function:
(x
2
1)
d
dx
P
(x) = xP
(x) P
1
(x) .
See e.g. Abramowitz & Stegun Handbook of Mathematical Func-
tions).
137
10 General Solution of the Maxwells Equa-
tions
In Chap. 6, we presented a method of solving the Maxwells equations for
the EM elds propagating in vacuum. We were able to reduce the Maxwells
equations into two dierential equations for

E and

B alone. The equations
were in the form of the wave equations whose the solutions are in the form
of plane and transverse waves. In this lecture, we will present a method for
general solution of the Maxwells equations in the presence of the sources,
charges and currents that depend on position r and time t. The concept of
vector and scalar potentials and gauges will be useful in the general solution.
10.1 Diculty of the Direct Solution of Maxwells Equa-
tions
Consider the Maxwells equations for the elds in the presence of charges and
currents
I.

E = /
0
, (35)
II.

B = 0 , (36)
III.

E =
B , (37)
IV.

B =
0
J +
1
c
2
E . (38)
In the Maxwells equations, the elds

E and

B depend on (r, t), the charge
and current densities also depend on (r, t). It is not explicitly stated in the
above equations, but we shall remember about this dependence in the fol-
lowing calculations.
Let us try to solve the Maxwells equations to nd the elds

E and

B pro-
duced by the source charges and currents

J.
The Maxwells equations involve two elds that satisfy coupled dierential
equations. First, we will try to separate the Maxwells equations into a dif-
ferential equation for

E alone or

B alone.
138
Assuming in the usual way that space and time operators commute, we act
with
1
c
2
t
on III and on IV, and obtain from III:
1
c
2

E +
1
c
2
2
t
2
B = 0 ,
and from IV:
1
c
2

E (

B) =
0

J .
Eliminating

E by subtraction of the two above equations, we get
(

B) +
1
c
2
2
t
2
B =
0

J .
Using the vector identity for double product, and from II:
(

B) =
2
B +(

B) =
2
B ,
we arrive to a wave equation
B
1
c
2
2
t
2
B =
0

J . (39)
Similarly, elimination of

B gives
E
1
c
2
2
t
2
E =
1
0
+
0
J . (40)
Equations (39) and (40) are in the form of coupled wave equations known as
inhomogeneous Helmholtz equations. We see that the current density

J en-
ters into both of these equations and enters in a relatively complicated way.
These equations are coupled through

J, and for this reason these equations
and are not readily soluble in general.
In the absence of currents and charges,

J = 0, = 0, and then the above
equations describe a free EM eld, and can be solved separately, as it was
illustrated in Chap. 6. The general solution of the wave equations, in the
presence of space and time varying currents and charges is complicated and
is more readily attained via the electromagnetic potentials.
139
10.2 Scalar and Vector Potentials
Generally, we do not nd elds

E and

B directly by integration of Eqs. (39)
and (40). We rather rst compute scalar and vector potentials from which
the elds may be found. We will illustrate here the advantage of working
with the potentials, i.e. with the the scalar and vector potentials.
Introduce the vector potential

A dened such that the Maxwells equation II
remains unchanged. The eld

B always has zero divergence,

B = 0, and
hence we can always write
B =

A , (41)
since

A is identically zero. Substitute this relation to the Maxwells
equation III, we obtain

E =

A
t
. (42)
The two curls are equal, but it does not mean that the vectors under the
curls are equal. Since always = 0, where is an arbitrary scalar
function, the two vectors are equal with the accuracy to :
E =
A . (43)
Scalar potential is a quantity from which a eld can be derived by a process
of dierentiation, e.g. in electrostatics
E = ,
where is the electrostatic potential.
In the static limit of

A/t = 0, the scalar function reduces to the familiar
electrostatic potential.
Equation (43) shows that the electric eld depends on the specic choice of
the potentials. We can change

A and and still get the same

E. One can
object that Eq. (41) ensure a xed value for

A, so Eqs. (41) and (43) are
140
satised for xed

A and . However, we can dene new potentials without
changing

E and

B
=

A + ,
=

t
. (44)
Proof:
From Eq. (43), we nd that
A +
A =

E .
Similarly, fro Eq.(41), we nd that
=

A
=

A +() =

A =

B .
as required.
The transformation (44) is called a gauge transformation, and the invariance
of the elds under such transformations is called gauge invariance.
In this case, how do we completely determine

A?
From the Helmholtz theorem we know that the denition

B =

A does
not completely dene

A despite the fact that

B is completely dened. The
vector potential

A is arbitrary to the extent that the gradient of some scalar
function can be added. Thus, innite set of possible potentials corresponds
to an innite set of possible vector potentials.
Recall the Helmholtz Theorem which says that any vector eld can be
written as a sum two terms
F =
1
4

F
r
dV +
1
4

F
r
dV
=

F
l
+

F
t
,
141
where

F
l
is called the longitudinal part of the eld and it has

F
l
= 0,
while

F
t
is called the transverse part as it has

F
t
= 0.
We see that

F and

F together determine

F but neither do alone.
Thus, if we dene

A, we complete the denition of

A. This is called
choosing the gauge of the potential. The above is an excellent illustration
of the power of the Helmholtz theorem. This theorem enables us to recog-
nize basic common properties of vector elds independent of their individual
physical properties.
10.2.1 Lorenz Gauge
How do we dene

A?
It is done as follows. We take of Eq. (43) and obtain

E =

A
2
. (45)
Thus, the electric eld

E will satisfy the Maxwells equation I when

A
2
= /
0
. (46)
From the Maxwells equation IV and

B =

A, we have
(

A) =
0
J +
1
c
2
,
2

A +(

A) =
0
J
1
c
2
2
t
2
A
1
c
2
t
,
or
2

A
1
c
2
2
t
2
A =
0
J +

A +
1
c
2
. (47)
The freedom of choosing

A and means that we can choose a set of potentials
to satisfy the condition

A +
1
c
2
t
= 0 .
142
This is called the Lorenz gauge and denes
A. This equation is sometimes

called the Lorenz equation.
Under the Lorenz gauge, Eq. (47) reduces to
2

A
1
c
2
2
t
2
A =
0
J ,
and applying the Lorenz gauge to Eq. (46), we get
2

1
c
2
t
2
= /
0
.
We see the advantage of using the vector and scalar potentials. In terms
of the potentials, the the Maxwell equations reduce to two uncoupled wave
equations that can be solved separately. The equations also show that the
Lorenz gauge is consistent with our experience on the sources of the EM
elds. The source of the vector potential that is related to the magnetic eld
is a current density, and the source of the scalar potential that is related to
charges is a charge density.
10.2.2 Coulomb Gauge
Another useful gauge of the potentials is the Coulomb gauge or transverse
gauge

A = 0 .
The origin of the name Coulomb gauge is in equation (45) that under the
condition

A = 0 reduces to the Poisson equation
2
= /
0
,
that determines the Coulomb potential due to the charge density .
Where the name transverse gauge came from?
Before we give the answer to this question, we rst show that the solution of
the Poisson equation is of the form
(r) =
1
4
0

r
dV .
143
It can be proved in the following way.
From the Coulombs law
E =
1
4
0
r
r
2
dV .
and using the relation
1
r
=
r
r
2
,
we can write the electric eld as
E =
1
4
0

r
dV = ,
where
=
1
4
0

r
dV . (48)
Since the electric eld satises the Maxwells equation I, we nd

E =
2
=

0
,
as required.
Now, we can nd the wave equation for

A under the Coulomb gauge
2

A
1
c
2
2
t
2
A =
0
J +
1
c
2
t
. (49)
In this equation, the term involving the scalar potential is called longitudi-
nal as it has vanishing . This suggests that it may cancel the longitudinal
part of the current density

J.
Let us check if the longitudinal part of the current density can be expressed
in terms of the scalar potential. According to the Helmholtz Theorem, the
current density can be written as
J =
1
4

J
r
dV +
1
4

J
r
dV
=

J
l
+

J
t
.
144
Using the continuity equation
t
+

J = 0
and the solution of the Poisson equation, we nd the longitudinal part of the
current density
J
l
=
1
4

J
r
dV =
1
4
t
r
dV
=
1
4
r
dV =
0
t
.
Then
J
l
=
0
t
=
1
c
2
t
,
which is equal to the second term on the right hand side of Eq. (49).
Hence, the inhomogeneous term in the wave equation (49) can be expressed
entirely in terms of the transverse current and then the wave equation for

A
reduces to
2

A
1
c
2
2
t
2
A =
0
J
t
.
This explains the origin of the name transverse gauge.
15
10.3 Solution of the Inhomogeneous Wave Equations
We have shown that the Maxwell equations can be reduced to two indepen-
dent wave equations for the potentials

A and . In fact, we have four scalar
equations for (A
x
, A
y
, A
z
, ). Each of these equations has the same form.
Therefore, it is enough to solve one of the four equations.
15
The transverse gauge is often used in atomic physics to calculate the EM elds pro-
duced by orbiting electrons. In this case, the orbiting electrons produce a current that is
solenoidal.
145
We will illustrate the solution on the equation for :
2

1
c
2
t
2
= /
0
. (50)
A general solution of the above equation may be found by considering two
limiting cases:
(a) Electrostatic limit: /t 0
In this limit the wave equation for reduces to the Poisson equation whose
the solution is
(r) =
1
4
0
(r
)
r
dV
.
(b) Source free limit: = 0
In this case, the wave equation (50) reduces to the homogeneous equation
2

1
c
2
t
2
= 0 .
This equation has a spherically symmetric solution of the form
(r, t) =
f(t r/c)
r
,
where f(t r/c) is an arbitrary function of the retarded time t r/c. The
retardation r/c is equal to the time needed for the electromagnetic wave to
pass the distance from the source to a given point in space.
Proof:
If there are no charged (boundary) surfaces in the space, the potential can
depend only on r, and must in fact be spherically symmetric. Thus, in
spherical coordinates only the radial part of the Laplacian will contribute to
the wave equation
2
=
1
r
2
r
2
.
146
Since
f
r
=
f
t
r
t
r
r
=
1
c
f
,
where t
r
= t r/c and f
= f/t
r
, we have
2
=
1
r
2
r
2
cr
+
f
r
2
=
1
r
2
rf
c
+ f
=
1
r
2
c
+
r
c
1
c
1
c
=
1
rc
2
f
,
where f
=
2
f/t
2
r
.
Moreover, /t = /t
r
, and then
1
c
2
t
2
=
1
c
2
r
f
.
Consequently, we obtain
2
=
1
c
2
t
2
=
1
c
2
r
f
1
c
2
r
f
= 0 ,
as required.
Summarizing the above analysis, we can construct a general solution of the
wave equation by noting that it must represent a spherical wave outside the
source and reduce to the appropriate static limit. This solution is
(r, t) =
1
4
0
V
(t r/c)
r
dV ,
where r is the distance coordinate from the source (from the charge dV ) at
the time when the potential wave left it. This exhibits the causal behavior
associated with the wave disturbance. The argument of shows that an
eect observed at the point r at time t is caused by the action of the source
a distant r away at an earlier or retarded time t
r
= t r/c. The time r/c is
the time of propagation of the disturbance from the source to the point r.
Thus, the Maxwells equations satisfy the causality principle.
147
10.4 Rigorous Solution: Green Functions Method
The wave equations all have the basic structure
2
(r, t)
1
c
2
2
(r, t)
t
2
= 4f(r, t) ,
where f(r, t) is a known (source distribution) function.
To solve this equation, we will introduce the Green function of the equation
and solve it as an inhomogeneous Helmholtz equation.
Suppose that (r, t) and f(r, t) have the Fourier integrals
(r, ) =
(r, t)e
it
dt ,
f(r, ) =
f(r, t)e
it
dt .
When we insert it into the wave equation, we nd that the Fourier transform
(r, ) satises the inhomogeneous Helmholtz wave equation
2
+ k
2
(r, ) = 4f(r, ) ,
where k = /c is the wave number.
The advantage of working in Fourier components is to remove the derivative
over time, and consequently to reduce the wave equation to a dierential
equation involving the space variables only.
Green function
For a unit point source the potential satises the Poisson equation
2
1
r
= 4(r) .
The function 1/r = G(r) is called a Green function of the above dierential
equation.
In analogy, we can dene the Green function of the wave equation
1
c
2
2
t
2
G(r, t t
0
) = 4(r)(t t
0
) .
148
The Fourier transform gives
2
+ k
2
G
k
= 4(r)e
it
0
.
where G
k
is the Fourier transform of the Green function G(r, t t
0
), which
we are trying to nd.
If there are no boundary surfaces, the Green function depends only on r,
and then the Laplacian operator in spherical coordinates depends only on r
giving
1
r
d
2
dr
2
rG
k
e
it
0
+ k
2
G
k
e
it
0
= 4(r) .
Everywhere except r = 0, the function rG
k
e
it
0
satises the homogeneous
equation
d
2
dr
2
rG
k
e
it
0
+ k
2
(rG
k
e
it
0
) = 0 ,
whose the solution is
rG
k
e
it
0
= Ae
ikr
+ Be
ikr
.
In this general solution for the Green function we could equally well choose
the exponential form
G
k
=
e
ikr
e
it
0
r
.
Using the inverse Fourier transform, we nd
G(r, ) =
1
2
e
ikr
r
e
i
d ,
where = t t
0
.
The integral
1
2
e
i(r/c)
d
149
is the delta function ( r/c). Thus
G(r, ) =
1
r
( r/c) .
The Green function is a casual response function, and has the same property
as the scalar potential of a point source.
In summary of the general solution of the Maxwells equations
The general (retarded) solutions of the Maxwells equations are given in terms
of the vector and scalar potentials
B =

A ,

E =

A
t
, (51)
with
(r, t) =
1
4
0
(t r/c)
r
dV , (52)
A(r, t) =
1
4
0
c
2

J(t r/c)
r
dV , (53)
and the potentials satisfy the Lorenz gauge.
Thus, we do not nd the elds by a direct integration of the Maxwells equa-
tions. If the charge and current distributions are known, we rst calculate
the scalar and vector potentials from Eqs. (52) and (53), and then nd the
electric and magnetic elds from Eqs. (51).
Revision questions
Question 1. Do we nd the elds by a direct integration of the Maxwells equa-
tions? Explain.
150
Question 2. Explain the usefulness of the scalar and vector potentials in the
solution of the Maxwells equations.
Question 3. Why do we use gauges in the solution of the Maxwells equations?
Question 4. Explain how do we solve inhomogeneous wave equations.
Question 5. What is the Green function of a given dierential equation?
151
11 Electromagnetic Antennas: Hertzian Dipole
The electromagnetic elds of charges in uniform motion are eectively bound
to the charges. The elds of accelerated (oscillating) charges on the other
hand can propagate as electromagnetic (EM) waves at the speed c and can
have a life of their own (until absorbed by some other charges).
In this lecture we will show how electromagnetic waves are generated by oscil-
lating charges. This will also illustrate an application of the general solution
of the Maxwells equations in calculations of the electric and magnetic elds
produced by a source system containing time varying charges and currents.
First, we will show that this problem can be solved with the help of only
the vector potential

A. Next, we will apply this concept to the problem of
generation of electromagnetic waves.
Consider the retarded solutions of the Maxwells equations
B =

A ,

E =

A
t
, (54)
with
(r, t) =
1
4
0
(t r/c)
r
dV ,
A

A(r, t) =
1
4
0
c
2

J(t r/c)
r
dV .
The above solution holds for the Lorenz gauge in which

A =
1
c
2
t
.
Assume that the position and time variations of the charges and currents can
be separated, so that the charges and currents vary sinusoidally in time
(r, t) = (r)e
it
and

J(r, t) =

J(r)e
it
.
In this case, the Lorenz gauge takes the form

A =
i
c
2
,
152
which gives
=
c
2
i

A . (55)
Thus, the scalar potential can be expressed in terms of the vector potential

A.
In other words, the scalar potential can be eliminated from the eld equations
leaving only the dependence on

A. As a result, we can express both

E and

B
in terms of the vector potential

A alone. Substituting Eq. (55) into Eq. (54),
we obtain
E =
c
2
i
(

A)

A
t
,

B =

A . (56)
Hence, both the electric and magnetic elds can be found from the vector
potential

A.
This result may seem rather strange at rst, since normally we should expect
to need both the scalar and vector potentials in order to completely determine
the EM elds. The explanation and in fact an another way of saying the same
thing is that time varying charge must satisfy the continuity equation

J =
t
= i ,
so that
=

J
i
.
We see that the time varying charge density can be expressed in terms of the
current density. Then the scalar potential becomes
(r, t) =
1
4
0
(t r/c)
r
dV =
1
4
0
i

J
r
dV
=
1
4
0
i

J
r
dV =
c
2
i

A .
Thus, specication of

J alone is sucient to completely determine all sources
of moving (oscillating) charges, and hence a solution for

A in terms of

J con-
tains all the necessary information to completely specify the time-varying
EM elds.
153
11.1 Generation of electromagnetic waves
We have learnt that the introduction of the concept of displacement current
by Maxwell led to the prediction of electromagnetic waves in vacuum. Now,
we inquire into the sources of electromagnetic waves, i.e. how to generate EM
waves of dierent wavelengths and dierent properties, how to control these
properties and how to propagate these waves in desired directions.
11.1.1 Field of an Element of Alternating Current
Consider a linear element l carrying an alternating current I = I
0
exp(it),
as shown in Figure 39. The current element may be viewed as two charges Q
and Q oscillating back and forth and can be served as an antenna, i.e. a
source of electromagnetic waves.
Figure 39: A linear alternating cur-
rent element l oriented in the z di-
rection. The generated elds are calcu-
lated in the r direction.
Assume that l is much smaller than
the wavelength = 2c/ correspond-
ing to the frequency of the oscillation.
In this case, we can ignore the phase
variation of the current along l. Of
course, this is an approximation that
may not hold for many realistic an-
tennas. However, an understanding
of the properties of such an antenna
is of great interest since, in principle,
all radiating structures can be consid-
ered as a sum of small radiating ele-
ments. Moreover, many practical an-
tennas working at low frequencies are
very short compared with the wave-
length.
For the elds around a small current element, l <, there are three spatial
regions (zones) of interest:
The near eld (static) zone: l <r <
The intermediate eld (induction) zone: l <r
The far eld (radiation) zone: l < <r
154
We will see that the elds have dierent properties in the dierent zones.
In the near zone the elds have the character of static elds, with a strong
dependence on the properties of the source. In the far eld zone, the elds
are transverse to the radius vector and fall of as r
1
, typical of radiation
elds, and are independent of the source.
Let us calculate the

E and

B elds around the current element. We start by
considering the retarded current element that cab be written as
J(t r/c)dV =

I(t r/c)dl = I
0
e
i(tkr)
dl ,
where k = /c.
Since the current is the same at any point of the antenna (l < ), it
appears as a constant for the integartion in

A, and then the vector potential
becomes
A =
I
0
4
0
c
2
e
i(tkr)
r
l ,
where

l =

dl.
In the near zone, where r < (or kr <1), the exponent exp(ikr) can be
replaced by unity. In the far eld zone kr 1, the exponential oscillates
rapidly, and in this region it is sucient to approximate exp(ikr) 1ikr.
In the intermediate zone, all powers of kr must be retained.
We now calculate directions and magnitudes of the electric and magnetic
elds produced by the current element. The elds are most easily evaluated
in spherical polar coordinates, if we choose the direction of the current ele-
ment

l along the z-axis, i.e.

l = l
k. We then have that in the spherical

coordinate system, the vector potential

A has components
A
r
= A
z
cos =
I
0
l cos
4
0
c
2
e
i(tkr)
r
,
A
= A
z
sin =
I
0
l sin
4
0
c
2
e
i(tkr)
r
, (57)
A
= 0 .
155
According to Eq. (56), to nd the elds

E and

B, we have to calculate

A
and

A, that in spherical polar coordinates are given by

A =
1
r
2
(r
2
A
r
)
r
+
1
r sin
(A
sin )
+
1
r sin
A
,
and

A =
r
r sin
(A
sin )
1
sin
A
r

(rA
)
r
(rA
)
r

A
r
.
Since A
= 0 and there is no dependence of A

r
and A
, i.e. A
r,
/ = 0,
the above equations reduce to

A =
1
r
2
(r
2
A
r
)
r
+
1
r sin
(A
sin )
, (58)

A =
(rA
)
r

A
r
r
. (59)
Magnetic eld

B
Calculate rst direction and magnitude of the magnetic eld produced by the
antenna. Since,

B =

A, we easily nd from Eq. (59) that the magnetic
eld of the current element is
B =

A = B
,
where
B
=
1
r
(rA
)
r

A
r
, (60)
and B
r
= B
= 0.
Conclusion: The magnetic eld produced by the antenna is perpen-
dicular to the radius vector at all distances.
156
Calculate the magnitude B
. Substituting Eq. (57) into Eq. (60), we obtain

B
=
I
0
l
4
0
c
2
r
sin e
i(tkr)
r
+
e
i(tkr)
r
cos
=
I
0
l
4
0
c
2
r
ik sin e
i(tkr)
+ sin
e
i(tkr)
r
.
Hence
B
=
I
0
l
4
0
c
2
ik
r
+
1
r
2
sin e
i(tkr)
.
A number of interesting general conclusions follow from this equation. In the
rst place, we note that the magnetic eld is composed of two terms: the
near zone term 1/r
2
and the far zone term 1/r. Secondly, we note that
in the limit of 0, the magnetic eld is composed of only the near zone
term that is the familiar Biot-Savart formula. Finally, the most important
is that the far zone term is only present for an oscillating eld ( = 0) and
therefore it represents a radiation eld arising from accelerated (oscillating)
charge.
Electric eld

E
We now calculate direction and magnitude of the electric eld produced by
the antenna, which as we have shown before can be found from the vector
potential
E =
c
2
i
(

A)

A
t
,
where

A
t
= i
A = i
A
r
r + A
.
We rst calculate

A. Substituting Eq. (57) into Eq. (58), we get

A =
1
r
2
(r
2
A
r
)
r
+
1
r sin
(sin A
157
=
I
0
l
4
0
c
2
1
r
2
r cos e
i(tkr)
r

1
r sin
sin
2
e
i(tkr)
r
=
I
0
l
4
0
c
2
cos
r
2
e
i(tkr)
ikre
i(tkr)
2 sin cos e
i(tkr)
r
2
sin
=
I
0
l
4
0
c
2
1
r
2
+
ik
r
cos e
i(tkr)
.
Next, we take gradient of

A. Since in spherical coordinates
= r

r
+

1
r
r sin
,
we obtain for the components of the gradient:
r
=
I
0
l cos
4
0
c
2
2
r
3

ik
r
2
+
k
2
r

ik
r
2
e
i(tkr)
=
I
0
l cos
4
0
c
2
2
r
3
+
2ik
r
2

k
2
r
e
i(tkr)
,
=
I
0
l
4
0
c
2
1
r
3
+
ik
r
2
sin e
i(tkr)
,
= 0 .
Hence, the radial part of the electric eld, E
r
, is
E
r
=
c
2
i
(

A)

A
t
r
=
I
0
l cos
4
0
i
2
r
3
+
2ik
r
2

k
2
r
e
i(tkr)
i
I
0
l cos
4
0
c
2
e
i(tkr)
r
=
I
0
l cos
4
0
c
2
ikr
3
+
2
r
2
+
ik
r

i
cr
e
i(tkr)
.
Since k = /c, the 1/r terms cancel and then E
r
simplies to
E
r
=
I
0
l cos
4
0
c
2
ikr
3
+
2
r
2
e
i(tkr)
. (61)
158
Similarly, we nd the polar component of E as
E
=
c
2
i
(

A)

A
t
=
I
0
l sin
4
0
i
1
r
3
+
ik
r
2
e
i(tkr)
+ i
I
0
l sin
4
0
c
2
e
i(tkr)
r
=
I
0
l sin
4
0
c
1
ikr
3
+
1
r
2
+
ik
r
e
i(tkr)
. (62)
The azimuthal component of the electric eld E
= 0.
Note that the radial part of the electric eld, Eq. (61), contributes only to
the near and intermediate zones, whereas the angular polar part, Eq. (62),
contributes to all of the zones.
Theorem:
The near-zone 1/r
3
part is the Coulomb type contribution. It is similar in
nature to a static eld surrounding a small linear-current element and an
electric dipole.
Proof:
The Coulomb or static eld is for 0. In this limit the 1/r
3
contribution is
E
=
I
0
l sin
4
0
c
1
ikr
3
e
i(tkr)
=
I
0
l sin
4
0
c
1
ikr
3
1 ikr +
1
2
(ikr)
2
+
=
I
0
l sin
4
0
c
1
r
2
,
where we have taken only the real (physical) part of the eld.
Since I
0
= q/t and l/t = c, we get
E
=
q sin
4
0
1
r
2
,
159
as required.
Electric and magnetic elds in near and far eld zones
Consider rst the near eld zone (r < ). In this limit the magnetic and
electric elds are
B
near
=
I
0
l
4
0
c
2
1
r
2
sin e
i(tkr)

,
E
near
= i
I
0
l
4
0
c
2
kr
3
e
i(tkr)
cos r + sin
.
Since the magnetic eld is real and the electric eld is imaginary, the Pointing
vector involving the near-zone eld components is a pure imaginary quantity.
It does not represent any ow of energy. This imaginary quantity represents
energy that oscillates back and forth between the source and the region of
space surrounding the source.
Consider now the far zone or radiation components of the magnetic and
electric elds.
E
rad
= E
R
, E
R
=
I
0
l sin
4
0
c
ik
r
e
i(tkr)
,
B
rad
= B
R
, B
R
=
I
0
l sin
4
0
c
2
ik
r
e
i(tkr)
. (63)
Note the following properties of the radiation components:
1. The electric and magnetic elds oscillate in phase.
2. The electric and magnetic elds are orthogonal to each other.
3. The ratio
E
R
B
R
= c, the value for plane waves in free space.

4. The electric and magnetic elds are transverse to the radius vector at
all distances.
5. The Poynting vector

N = c
2
E
rad

B
rad
is a real quantity and is in
the direction of the radius vector, indicating that the energy of the eld
propagates away from the current element.
These properties show that in the far zone the EM eld is in a form of plane
waves propagating with the speed of light c.
160
11.2 Power Radiated from the Current Element
Consider now the radiation power emitted by the antenna. By the radiation
power we mean that part of energy which is carried by the radiation com-
ponents of the eld. We will show that the radiation power is equal to the
radiation losses, i.e. energy carried by the plane electromagnetic wave that
propagates on its own independent of the source.
The power ux at any point is given by the Poynting vector
N = c
2
E

B .
Then, the total power radiated across a sphere of radius r is
W =
N

dS ,
where
dS = r
2
sin dd .
Only those partial products in

E

B which vary as 1/r
2
will have net radi-
ated power. The other partial products are small as they fall o more rapidly
than 1/r
2
. Thus, the only part of the elds entering into the expression for
the radiated power is the far eld zone part (radiation component) consisting
of the terms varying as 1/r.
The radiation components of

E and

B oscillate in phase as sine or cosine
functions. Thus, an average of their product over time is
E
R
B
R
=
1
2
E
R
B
R
0
,
where
E
R
0
and
B
R
0
are amplitudes of E
R
and B
R
.
On substituting from Eq. (63), we nd that the time averaged Poynting
vector is
N =
1
2
c
2
0
I
2
0
l
2
16
2
2
0
c
3
k
2
r
2
sin
2
=
I
2
0
l
2
32
2
0
c
4
2
2
r
2
sin
2
=
I
2
0
8
0
c
2
sin
2
r
2
.
161
Hence, integrating over all directions, we get the total power emitted by the
antenna
W =
2
0
r
2

N sin dd
=
2
0
I
2
0
8
0
c
2
d

0
sin
3
d .
Performing the integration over that is equal to 2, and over that is equal
to 4/3, we get
W = 2
I
2
0
8
0
c
2
4
3
=
I
2
0
3
0
c
2
.
We can write the total power radiated in terms of the power absorbed in an
equivalent resistance, called the radiation resistance, as
W =
1
2
2
3
0
c
2
I
2
0
=
1
2
RI
2
0
.
where
R =
2
3
0
c
2
is the radiation resistance.
Since 1/(
0
c) =
0
/
0
= 377, or 120, we obtain for R:
R = 80
2
2
.
For example, if l/ 0.1, then R = 0.8
2
8 ohms.
This example shows that for a current element which is 10% of the wave-
length long, the resistance is very small.
Thus, if l/ <1, the radiation losses are negligible, that the radiated power
is very small. In terms of the emitted radiation, the emitted EM waves are
very weak in power.
162
With a short linear current element, an appreciable power would be radiated
only if the current amplitude I
0
were very large. A large current, on the other
hand, would lead to large amounts of power dissipation in the conductor, and
hence a very low eciency.
We can conclude that current carrying systems that have linear dimensions
small compared with the wavelength radiate negligible power. An ecient
antenna should have dimensions comparable to or greater than the wave-
length.
11.3 Gain of the Dipole Antenna
A further property of the dipole antenna that is worthy of consideration is
the directional property of power radiated in dierent directions.
The gain or directivity function of a transmitting antenna is the ratio of the
Poynting ux to the ux due to an isotropic radiator emitting the same total
power W:
g
T
(, ) =
N(, )
N
iso
,
where
N
iso
=
W
4r
2
is the energy ux uniform in all directions.
Since for the innitesimal dipole:
N(, ) =
I
2
0
8
0
c
2
sin
2
r
2
,
and
W =
I
2
0
3
0
c
2
,
163
we obtain
g
T
(, ) = g
T
() =
3
2
sin
2
.
The directivity function g
T
(, ) denes a three-dimensional surface called
the polar radiation pattern of the antenna.
Figure 40: The polar radiation pattern of
the dipole antenna.
The function varies as sin
2
, and
hence the radiation is most intense
in the = /2 direction (perpen-
dicular to the axis of the dipole)
and zero in the directions = 0,
(on the axis of the dipole). The
maximum gain then is 1.5 for di-
rections dened by = /2, in
the equatorial plane of the dipole.
The gain function is independent
of .
The spatial distribution of the radiated power can be shown on a polar di-
agram, such as in Figure 40, which gives the relative values of the radiated
power at dierent positions on the surface of a sphere centered on the dipole.
We can conclude, that the directivity function g
T
(, ) is a measure of how
eective the antenna is in concentrating the radiated power in a given direc-
tion.
164
Revision questions
Question 1. Under what condition the elds can be expressed solely by the vec-
tor potential?
Question 2. What do we mean by near zone and far eld zone?
Question 3. What are the properties of the EM eld in the far eld zone?
Question 4. Show that the power radiated by an antenna is equal to the radia-
tion losses.
Question 5. Why emitted waves of a short antenna are very weak?
Question 6. What is the spatial distribution of the power emitted by a short
antenna?
165
12 Electromagnetic Theory of Polarizable
Materials
Condense matter physics and electromagnetism theories used to be regarded
as almost separate subjects. However, in next few lectures, we will illustrate
how matter and electromagnetism are integrated and how the electromag-
netic theory may be applied to condense matter physics to explain the prop-
erties of conducting materials and insulators, or dielectrics as they are often
called. Materials such as conductors or insulators are composed of atoms
that contain free and bounded electric charges. The charges can be redis-
tributed by an application of external elds. We will focus on new elds
induced by redistributed electric charges and will show how these problems
can be solved with the help of only the scalar potential .
We proceed our analysis by introducing the microscopic (atomic) model of
materials and for this purpose simplied models of the atom will be used.
In this simplied model, the atom consists of a positively charged nucleus
surrounded by a spherically symmetric cloud of electrons. In the absence of
an external electric eld, the electrons and nucleus are in a stable equilib-
rium. What happens if the atom is placed in an external electric eld? If
an uncharged dielectric (insulator) is placed in an electric eld, the eld will
redistribute the charges within the dielectric. An eective non-zero charge is
induced by rearrangement of bound charges within the atoms or molecules
of the dielectric. In dielectrics these charges are a set of molecular dipoles
that, in turn, will set up a secondary (induced) eld so that the net eld will
be modied from its original (external) value. We will show how the modi-
cation of the eld led to the concept of an another eld vector; the dielectric
displacement vector

D. The advantage of using the dielectric displacement
vector and its physical properties will be explored.
In materials, electric dipoles may exist permanently or may be induced by
the external eld. Dielectrics with permanent dipole moments are usually
electrically neutral due to random orientation, in the absence of electric elds,
of the dipole moments. An example is H
2
O. The application of an external
electric eld causes all molecules composing the dielectric to align themselves
with the external eld. This produces a macroscopic dipole moment. We will
166
study electromagnetic theory of polarizable materials in terms of microscopic
objects, electric dipoles, which we will treat as building blocks of dielectric
materials. As we shall see, despite this simplicity, the model satisfactorily
predicts the macroscopic behavior of dielectric materials.
12.1 Potential and Electric Field of a Single Dipole
Mathematically, for the calculations of the eld of distributed charges in-
side a dielectric material, it is convenient to deal with a separate object, the
dipole, that is composed of two bounded charges of oposite signs separated
by a small distance, not as just a pair of individual plus and minus charges.
Figure 41: A schematic model for
the calculation of the potential pro-
duced by an electric dipole.
Suppose that two opposite sign point
charges q are separated by a dis-
tance d. We will nd the potential
and the electric eld

E at a point A dis-
tance r and angle under the assump-
tion that the distance r is much larger
than the separation between the charges,
i.e. r d. The potential pro-
duced by two separated charges of op-
posite sign is called the dipole poten-
tial.
We dene electric dipole moment as the
product of the charge times the separation.
It is a vector that points from the negative charge to positive charge
p = q
d .
Since r
2
r
1
d cos at distances r d and d cos = p r, we get for the
potential at the point A:
=
q
4
0
r
1
q
4
0
r
2
=
q
4
0
r
2
r
1
r
1
r
2
=
q
4
0
d cos
r
1
r
2
1
4
0
p r
r
2
. (64)
Hence, we can nd electric eld of the dipole using the relation

E = .
Since the potential of the dipole depends on r and , it is convenient to work
167
in the spherical coordinates in which the electric eld is given by
E = E
r
r + E
+ E
,
Figure 42: The electric eld lines of a
dipole moment.
where the components are
E
r
=
r
=
1
4
0
2p cos
r
3
,
E
=
1
r
=
1
4
0
p sin
r
3
,
E
=
1
r sin
= 0 .
Hence
E =
p
4
0
r
3
2 cos r + sin

.
One can note that the eld has az-
imuthal symmetry. The eld is pro-
portional to (1/r)
3
so that it falls
rapidly with distance. Moreover, in the direction perpendicular to the dipole
moment, = 90
, the eld points in the opposite direction to the dipole

moment at all distances. Figure 42 shows a sketch of the electric eld lines
of an electric dipole moment.
12.2 Polarization Vector
If there are N dipoles per unit volume of a dielectric, the total dipole mo-
ment is:
P =
N
i=1
p
i
,
and is called the polarization.
The electric potential set up in the space by a dielectric material with an
arbitrary volume distribution of electric dipoles can be calculated by using
the potential produced by a single dipole, Eq. (64), and the above denition
168
of

P. The electric potential at an arbitrary point A distance r from a volume
element dV containing such dipoles is
d =
1
4
0
P r
r
2
dV ,
where r is the unit vector from dV towards A, and we have assumed that r
is much larger than the extent of the volume element dV .
Figure 43: Schematic diagram for cal-
culations of the scalar potential pro-
duced at point A by a macroscopic ma-
terial of the polarization

P.
Let r be the unit vector from A to-
wards dV . (We want to integrate over V
with the position of A xed). In this case,
we change r r, and obtain
d =
1
4
0
r
r
2
dV ,
and then the potential measured in the
direction towards the volume V is
=
1
4
0
r
r
2
dV .
It is dicult to interprete this form of
the potential as it is not in the form of
a potential known from the electrostat-
ics.
However, the result can be transformed into a form that can be interpreted
with the help of the general form of the static potentials.
First, noting that
r
r
2
=
1
r
,
we have that
r
r
2
=

P
1
r
.
169
Next, applying a vector identity (
A) =

A +

A , we can write
the above expression as
1
r
P
r

P
r
.
Hence, we can write the potential as a sum of two terms both involving
the 1/r function
=
1
4
0
P
r
dV +
1
4
0

P
r
dV .
Note that the second term varies with r explicitly as the 1/r function, but
in the rst term, the 1/r function is under the nabla operator. However, we
can use the Gausss divergence theorem to transform the rst term into the
surface integration. In this case, the rst term becomes a simple function
of 1/r, so that the potential takes the form
=
1
4
0
P n
r
dS +
1
4
0

P
r
dV , (65)
where n is the unit vector normal to the surface of the material.
Figure 44: Surface (left picture) and volume charges (right picture).
On comparing with the general form of the static potential, Eq. (48), the ex-
pression (65) can be interpreted as follows. The rst term on the right-hand
170
side, a surface integral, is a potential equivalent to that of a surface charge
density
s
=

P n. The second term is a potential equivalent to that of a
volume charge density
V
=

P.
Where did this interpretation came from? It can be explained by using a
simple picture shown in Figure 44. Surface charges exist because there are
no neighboring charges at the end surfaces of the material to cancel them
out. Volume charges exist because the number of dipoles per unit volume
changes, that there is an incomplete cancellation of charge density from ad-
jacent dipoles.
Materials which have a non-zero volume charge density are called inhomo-
geneous materials. Thus, a sucient condition for a material to be homoge-
neous is that the polarization of the material have a zero divergence.
12.3 Maxwells Equation for

E in a Dielectric
We have just seen that a non-zero potential outside a dielectric material is
produced by surface and volume charges. A question then arises: What is
the electric eld inside and outside the dielectric when extra polarization
charges are present?
In general, the electric eld in the dielectric can be found from the Maxwells
equation I:

E =

0
,
where is the total volume charge density inside the dielectric.
16
In the dielectric it is convenient to (mentally) separate the polarization
charges from whatever other charges might be there also. The other charges
are usually referred to as the free charges or conducting charges to distin-
guish them from the bound charges in the dielectric.
16
Why the surface charge density is not taken into account in the Maxwells equation I?
171
We may write the Maxwells equation I as

E =
f
+
p
,
where
f
is the free charge density and
p
is the polarization charge density
throughout the volume. If we express
p
in terms of

P, i.e. if we write
p
=

P, we obtain

E =
f

P .
Hence
(
0
E +

P) =
f
. (66)
Now it is common practice to drop the subscript f, but one must remember
that now stands for the charge density not counting the polarization charges.
Next, we can dene a new vector
D =
0
E +

P , (67)
called the dielectric displacement eld,
17
and write Eq. (66) as

D = .
We can read this equation that the source of the eld

D inside a dielectric is
the free charge density .
17
The reason for the name dielectric displacement can be easily understood if we refer
to the Maxwells theory. Take time derivative of both sides of Eq. (67):
D
t
=
0
E
t
+

P
t
.
We know from the Maxwells theory that the rst term on the rhs of the above equation
represents displacement current density, and the second term is the polarization current
density. Therefore,
D/t can be called a generalization of the displacement current

density, and then

D can be regarded as the dielectric displacement.
172
12.4 Macroscopic Eects of the Polarizability
We turn now to a consideration of the macroscopic eects of the polarizabil-
ity of dielectric materials. We will consider only ideal dielectrics.
Ideal dielectrics can be divided into following categories:
1. Homogeneous properties independent of the position.
2. Isotropic properties independent of direction.
3. Linear polarization proportional to

E.
4. Stationary properties independent of time.
Case of simple isotropic and linear dielectrics
Ordinary dielectrics (glass, teon, plastics etc.) are linear in polarization for
elds not strong enough to cause dielectric breakdown i.e.

P

E.
For these materials, we can write
P = N
E =
0
E ,
where is the polarizability of a single atom (molecule), and is the dielec-
tric susceptibility. Then, we can write the dielectric displacement as
D =
0
E +

P =
0
(1 + )
E =
0
E =
E ,
where
r
is the relative permittivity or dielectric constant, and is the per-
mittivity. Hence

D = (
E) =
or

E =

,
if is independent of position i.e. if is a permittivity of a homogeneous
dielectric.
173
The same result would be obtained by replacing
0
in Coulombs law by .
F =
1
4
q
1
q
2
r
2
r .
The ratio /
0
then represents the relative shielding of q
1
from q
2
by the
polarization charges induced in the medium.
How do we determine the permittivity of a given material?
The theory of the molecular structure of a material will yield an estimate
of and . The dipole moment p of a molecule will be proportional to
the local electric eld

E so we can dene a molecular susceptibility
m
such
that p =
m
0
E. Then
P =
i
p
i
= N
m
E ,
where N is the number of molecules per unit volume. Thus
=
0
r
=
0
(1 + N
m
)
and is a quantity directly measurable from the measurement of capacitance
C =
A
d
=
r
0
A
d
=
r
C
0
,
where C
0
is the capacitance without the dielectric.
We can summarize: Filling a capacitor with dielectric multiplies its ca-
pacitance by
r
.
Exercise in class: Capacitor lled with a homogeneous dielectric
A plane parallel capacitor has charges + and per unit
area on its plates. The capacitor is lled with a homoge-
neous and linear dielectric of dielectric constant
r
= 1+.
Show that:
174
(a) The electric eld within the dielectric is:
E =

0
.
(b) The polarization charge per unit area on the surface
of the dielectric adjacent to the surface of the negatively
charged plate is:
s
=

1 +
.
(c) The capacitance of the capacitor is C =
r
C
0
where C
0
is the capacitance of the same capacitor without the dielec-
tric (i.e. a vacuum or air between the plates).
175
12.5 Dense Dielectrics: The Clausius-Mossotti Rela-
tion
In the standard calculations of the polarization of macroscopic materials, it is
often assumed that the eld

E is the same at any point of the material. How-
ever, for an extended dense materials the eld

E may vary with the position
and the calculation of the dielectric constant of the material may not agree
with an experimental measurement. This is what really happens. Therefore
the approach of a constant eld through the whole area of the material must
be modied.
It was Lorentz, who proposed an approach that resolved this problem. The
Lorentz theory of polarizability of dense dielectric materials distinguishes be-
tween the mean electric eld

E and the local electric eld

E
loc
as seen by a
typical dipole. The typical dipole is considered to be at the center of a small
sphere that has been excavated from the dielectric.

E
loc
is thus the mean
eld

E minus

E
plug
where

E
plug
is the eld of the spherical volume excavated.
Let us see what would be the eld at this area if the dipole is not there.
Let

E be the mean eld throughout the dielectric.
Let

E
plug
be the eld due to the spherical plug alone.
Let

E
loc
be the eld in the spherical hole.
E =

E
loc
+

E
plug
and

E
plug
=
P
3
0
.
Thus, the eld acting to polarize the molecule is equal to

E minus the con-
tribution to the total average eld from the molecule itself. Hence
E
loc
=

E +
P
3
0
.
The argument now is that each molecule is at the centre of a small hole
and the eld acting on the molecule is thus

E
loc
. If is the molecular
polarizability, its induced dipole is thus:
p =
E
loc
176
If there are N molecules per unit volume then:
P = N p = N
E
loc
= N
E +
P
3
0
.
By denition:

P =
0
E = (
r
1)
0
E. Substituting for

P in the above
equation, we obtain
(
r
1)
0
E = N
E +
(
r
1)
0
3
0
.
Hence
(
r
1)
0
= N
1 +

r
1
3
=
N
3
(
r
+ 2) ,
and nally
r
1
r
+ 2
=
N
3
0
.
This is known as the Clausius-Mossotti relation for a dense dielectric. Assum-
ing that is known, we can compute
r
- dielectric constant of the material.
The Clausius-Mossotti relation works pretty well for most of dense materials.
However, it has a weak point. To illustrate this, we solve the Clausius-
Mossotti equation for
r
. Since
r
1 = (
r
+ 2)
N
3
0
,
we obtain for
r
:
r
=
1 +
2N
3
0
1
N
3
0
.
Note that
r
=
1 +
2N
3
0
1
N
3
0
as
N
3
0
1 .
We see that one can adjust the number of the dipoles to get an innite di-
electric constant of a nite material. This is called the Clausius-Mossotti
177
catastrophe.
This eect can be explained as follows: Removal of the plug leaves polariza-
tion charges, whose eld tends to line up the dipole parallel to the eld. The
system is self-polarizing (Clausius-Mossotti catastrophe) if
N
3
0
1, i.e. the
system induces

P without an external eld.
12.6 Dielectric in a Time Dependent Field
Let us consider what will happen if a dielectric is introduced into an alter-
nating electric eld. The student immediately conclude that the alternating
electric eld will create oscillating dipole moments inside the material, or
equivalently, an oscillating polarization of the material. This is true, but
then an another question arises: How does the polarization follow the oscil-
lating external eld?
Figure 45: A plane plate capacitor lled
with a dielectric and connected to an alter-
nating current.
To answer this question, let us
consider an experiment, illustrated
in Figure 45, involving a plane
plate capacitor lled with a di-
electric and connected to an os-
cillating AC current. The alter-
nating current will create an oscil-
lating electric eld inside the ca-
pacitor which will cause the elec-
tric dipoles of the dielectric to os-
cillate in time. On the other
hand, the oscillating electric eld
will induce oscillating polarization
charges.
What are the oscillating polarization charges equivalent to? The induced
polarization charges do not produce any currents inside the dielectric. There
is no DC current in response to a DC electric eld, but if

P is changing with
178
time (because

E is changing with time) there will be an AC current density:
J =

s
t
n =

P
t
.
Thus,
P/t plays the role of polarization current density.

With the alternating electric eld present, the polarization

P may lag in
phase behind the driving eld

E. This means there is internal friction and
heat dissipation. If it happen, the capacitor will exhibit resistive as well as
capacitive properties.
Since resistivity leads to a dissipation of the energy, we proceed by considering
the work done in charging the capacitor. It is dened as
dW
dt
= V I = Ed
dQ
dt
,
where V is the voltage.
Thus, we have to nd how quickly the charge on the plates is changing in
time.
Since we are interested in the time variation of the electric eld and the
polarization, we express the charge in terms of the eld quantities
Q = CV = CEd =
A
d
Ed = AE = AD ,
from which, we obtain
Q = A(
0
E + P) .
Then work in terms of the eld variables is
dW
dt
= Ed
d
dt
A(
0
E + P) = E dA
d
dt
(
0
E + P) .
Since dA = 1 is the volume of the capacitor, we can write
dW
dt
=
E
d
dt
(
0
E) + E
dP
dt
d1 ,
179
or
dW
dt
=
V
d
dt
1
2
0
E
2
d1 +
V
E
dP
dt
d1 ,
where we took into account a possibility that the electric eld and polariza-
tion can vary across the capacitors plates.
The rst term in the above equation is the rate of doing work building up

E
eld.
The second term is the rate of doing work on the dipoles by

E.
Thus, the supplied energy to the capacitor is used to build up the electric
eld inside the capacitor and to polarize the dielectric. Consider separately
both terms.
First term:
If E = E
0
cos(t), the rst term takes the form
V
(E
0
cos(t)
0
E
0
sin(t)) d1 .
The element of work done per unit volume and unit time is
dW
d1
=
0
E
2
0
cos(t) sin(t) .
Averaging over a cycle, we get
dW
d1
=
0
E
2
0
2/
t=0
cos(t) sin(t) dt = 0 .
No energy has been used to build up the electric eld inside the capacitor.
Work is done building up the eld in one part of the cycle but the stored
energy is given back in another part.
Second term:
180
If P =
0
E =
0
E
0
cos(t) the same zero net energy conversion averaged
over a cycle will happen with this term. If there is internal friction there will
be a phase dierence between P and E, that P does not follow the changes
in E. If we write
P =
0
E
0
cos(t + ) ,
where represents a phase dierence between P and E, we get for the
polarization
P =
0
E
0
cos cos(t)
0
E
0
sin sin(t) .
Taking the time derivative, we get
dP
dt
=
0
E
0
cos sin(t)
0
E
0
sin cos(t) .
Hence, the work done per unit volume per cycle will be
dW
d1
=
2/
0
E
2
0
cos cos(t) sin(t) dt
2/
0
E
2
0
sin cos
2
(t) dt .
Since the rst integral is zero, we get
dW
d1
=
0
E
2
0
sin
2/
0
cos
2
(t) dt . (68)
The integral on the rhs of the above equation is positive and dW/d1 must
be positive corresponding to energy dissipation (or the dielectric would keep
getting energy from its interior and building up the eld with it).
Thus, sin must be negative, so < < 0. This means that the polariza-
tion lags in phase the electric eld.
P =
0
E
0
cos(t ) .
In summary: Equation (68) shows that an energy is lost in each cycle of the
oscillations. In practice, it is dissipated as a heat in the material. The energy
loss is caused by the work required to change the polarization of the material.
181
12.7 The Complex Susceptibility and Permitivity
We have learnt that the polarization of a linear dielectric is proportional to
the eld and the proportionality is determined by the dielectric constant or
susceptibility. In the previous lecture, we have modeled the fact that in real
dielectrics, the polarization may lag in phase behind the driving eld

E by
introducing the phase dierence between

P and

E. We may now ask: If

P
lags in phase behind

E, how does then the relation between

P and

E is af-
fected?
To answer this question, it is convenient to use complex exponentials to
represent amplitude and phase of an oscillating quantity. We write
E =

E
0
e
it
and

P =

P
0
e
i(t)
, ( is positive) .
We can write the complex polarization in dierent forms
P =

P
0
e
i
e
it
= (
P
0
cos i
P
0
sin )e
it
,
P =
P
0
E
0
cos i
P
0
E
0
sin
E
0
e
it
,
P =
0
(
E
0
e
it
=
0
c
E
0
e
it
=
0
c

E .
The relation between

P and

E is the same as before for a constant external
eld. However, now
c
=
is a complex quantity called a complex

susceptibility.
Thus, the phase dierence between the polarization and the exter-
nal eld leads to a complex susceptibility of a dielectric material.
In other words, the internal friction of the material results in a
complex susceptibility.
With the complex polarization, the dielectric displacement takes the form
D =
0
E +

P =
0
E +
0
E =
0
(1 +
c
)
E ,
or
D =
0
(1 +
E =
0
E =
c
E ,
182
where
c
is a complex permittivity, and
r
is a complex relative permittivity
or dielectric constant
c
=
0
(1 +
) i
0
.
How do we in practice estimate losses in a given material? In other words,
how do we calculate
for a given material?

The imaginary part of
r
,
, can be determined from the following experi-

ment involving a capacitor lled with a dielectric.
Figure 46:
As we have just shown, the imag-
inary part of the dielectric suscep-
tibility represents losses, i.e. cor-
responds to net energy dissipation.
In the circuit theory language the
imaginary component of the dielec-
tric susceptibility adds a resistive
component to the capacitor. The
material lling the capacitor could
also have some ordinary ohmic con-
ductivity (due to the presence of
free charges as well as bound charges in the material). Let us calculate
the magnitude of the total resistance of the dielectric. Simply, we will con-
sider the Ohms law for the capacitor and will nd the relation between an
external current supplied to the capacitor and voltage. The relation will give
us an information about the resistance of the capacitor.
Let I be the oscillating current in the external circuit, and
I
p
=

J
p
d
A = A
dP
dt
be the polarization current in the dielectric. Let
I
c
=

J
c
d
A = AE = A
V
d
be the conduction current in the dielectric due to its nite conductivity .
183
If
s
is the charge density on the plates, then the eective charge on the
plates is
Q =
s
A = q
I
q
p
q
c
,
where q
I
is the charge supplied by I, q
p
is the charge removed by the polar-
ization current I
p
, and q
c
is the charge removed by the conduction current I
c
.
We can write the eective charge as
Q =
I dt
I
p
dt
I
c
dt .
Since the electric eld inside the capacitor is
E =

s
0
=
V
d
,
we can nd the surface charge density
s
=
0
V
d
.
and write the eective charge as
Q =
s
A =
0
V
d
A =
Idt
I
p
dt
I
c
dt .
By taking a derivative in time of the both sides of the above equation, and
substituting for P the relation, P =
c
0
E =
c
0
V
d
, we get
0
A
d
dV
dt
= I I
p
I
c
= I A
0
d
dV
dt
A
V
d
. (69)
Putting
dV
dt
= iV ,
c
=
,
and solving Eq. (69) for I, we get
I =

0
A
d
(1 +
)i +

V .
Separating real and imaginary parts and putting C
0
=
0
A/d (the capaci-
tance there would be if the dielectric were lossless), we nd
I =
C
0
0
+
+ iC
0
(1 +
V .
184
Figure 47:
Since the capacitor transmits some
charges through the internal dielec-
tric, in the circuit theory this system
is equivalent to a parallel circuit, as
shown in Figure 47. For the parallel
circuit
I =
1
R
+ iC
V .
Comparing with the above result for
current ow in the lossy capacitor
we see that the eective capacitance
is C
0
(1 +
) and the eective resistance is

R =
1
C
0
0
+
,
where we remember that C
0
is the capacitance in the absence of losses.
12.8 The Loss Tangent
The properties of a dielectric material are usually specied by giving its di-
electric constant K, and its loss tangent tan .
We can write the general Ohms law as
I = iC
0
(1 +
) i
V .
The quantity in [ ] brackets can be dened as a complex relative permittiv-
ity
r
:
r
= 1 +
.
This is a generalization on the previous denition of complex relative per-
mittivity to include the eects of ohmic conductivity. Then we can dene a
185
generalized permittivity and dielectric constant =
0
r
.
Now if we write
r
= K
e
i
= K
(cos i sin )
= K
cos (1 i tan ) = K(1 i tan ) .

Then the standard form for permittivity is
=
0
K(1 i tan ) ,
where tan is the loss tangent.
18
Since K
cos = (1 +
) and K
sin = (
) it follows that:
tan =
[1 +
]
.
In this equation, tan includes the eects of nite conductivity and the eects
of polarization damping force.
18
In practice, it is read out on some AC bridges as an alternative to reading out the
resistive property of a lossy capacitor.
186
Revision questions
Question 1. How do we calculate the electric eld of a microscopic dipole mo-
ment?
Question 2. What are the surface and volume charge densities?
Question 3. Explain the advantage of introducing the concept of dielectric dis-
placement eld.
Question 4. How do we calculate the polarization vector of a dense dielectric?
Question 5. In real dielectrics, does the polarization vector follow the changes
of an external electric eld?
Question 6. How are losses in a dielectric represented in the permitivity of the
dielectric?
Question 7. What it meant by the loss tangent of a medium?
Tutorial problems
Problem 12.1 Capacitor with 2 dierent dielectrics
Consider a capacitor in which there are two dierent dielectrics
of thickness d
1
and d
2
having dielectric constants
1
and
2
. The
separation of the plates is then d
1
+d
2
.
(a) Calculate the polarization charge distributions everywhere.
Show that at the junction of the two dielectrics there is a surface
charge density:
s
=

1
2
.
(b) Show that the capacity of the capacitor, if the area of the
187
plates is A, is:
C =

1
0
A
2
d
1
+
1
d
2
.
Problem 12.2 Capacitor lled with an inhomogeneous dielectric
A parallel plate capacitor has charges + and per unit area
on its plates which are separated by a distance d. The volume be-
tween the plates is lled with an inhomogeneous dielectric. The
dielectric susceptibility is zero at the positively charged plate and
increases linearly with distance, reaching its maximum value of
unity at the negatively charged plate.
If x is the coordinate measured from the positive plate toward
the negative plate and is the gradient of dielectric susceptibil-
ity, i.e. = x, show that:
(a) The electric eld through the dielectric varies like:
E =

0
(1 +x)
0 x d .
(b) There is a volume charge distribution throughout the dielec-
tric given by:
=

(1 +x)
2
,
together with a surface charge distribution on the dielectric adja-
cent to the negative plate given by:
1
=

2
.
(c) The capacitance of the capacitor is:
C =

0
A
ln 2
.
Hint :
You can see there will be a volume charge distribution within the
dielectric because:
=

P = (
0
E) = 0 , because = x .
188
Also since this is the only charge in the region the polarization
charge density also satises the general Maxwell equation:

E =

0
.
From these two equations you can develope a dierential equation
describing the variation of the electric eld across the capacitor.
Notice that it is
0
we use in the equations (not ) because we are
treating the polarization charges explicitly.
After solving for the electric eld E(x) one can calculate the
charge distribution from:
=

P = (
0
E) .
Problem 12.3 Equivalent circuit of a lossy capacitor
So-called electrolytic capacitors are made by electrolytic deposit
of a thin layer of dielectric on aluminium. Consider a 1000 F
capacitor that might be used in smoothing the ripple in a DC
voltage supply generated by rectifying the mains (50 Hz) supply.
The dielectric has a complex dielectric constant:
r
= 1 +
= 5 i 10
4
and a conductivity of = 2 10
13
[mho m
1
].
What is the equivalent parallel resistance of the capacitor at fre-
quency 50 Hz? and at 1 kHz?
Problem 12.4 Field at the centre of a uniformly polarized sphere
Consider a dielectric sphere with uniform polarization

P through-
out. The equivalent polarization charges will be a volume charge
density

P which will be zero because of the uniform charac-
ter of

P together with a surface charge density

P n. Show then
that the polarization charges will give rise to an electric eld at
the centre of the sphere given by:
E =
1
3
0
.
189
13 Electromagnetic Theory of Magnetizable
Materials
In rst lecture on electromagnetic theory of polarizable materials, we have
discussed how the polarization of dielectrics by an externally applied electric
eld is equivalent to creation of volume and surface distributions of charges.
Analogously, a magnetic eld can act on molecular scale current loops exist-
ing in the building blocks of materials, atoms, to produce macroscopic eects.
It was Amp` ere who rst suggested that the magnetism of matter was due to
the cooperative eects of currents circulating in atoms and not, as previously
thought, due to a separate magnetic charge called poles. Thus, the magnetic
properties of materials can be considered from an atomic viewpoint of elec-
trons currents, in which case a fundamental understanding of sources of the
magnetic eld can be developed.
13.1 Magnetic Polarization Currents
Let us illustrate the microscopic (atomic) theory of magetism from which we
will then formulate the macroscopic theory.
Figure 48: A schematic diagram of
a current loop created by orbiting elec-
tron.
According to the atomic theory, the elec-
trons rotating in orbital paths are equiva-
lent to circulating currents on an atomic
scale. We can dene the magnetic mo-
ment of an electron current loop as
= IA n ,
which is equal to the product of the area
of the plane loop closed by the electrons
orbit and the magnitude of the circulat-
ing current. The vector direction n of the
moment is perpendicular to the plane of
the loop and along the direction set by the right-hand rule, as shown in
the Figure 48.
190
A macroscopic material body contains a lot of current loops, so we can de-
ne a macroscopic dipole moment per unit volume of the material, called
magnetization
M =
i

i
.
The theory of magnetism is based on the above denition of magnetization,
and is formulated by a simple theorem:
Theorem: If a volume V enclosed by surface S has a magnetic dipole mo-
ment per unit volume

M, which may be a function of position,
the macroscopic magnetic elds so produced are equivalent to
those of:
A volume current density

J
V
=

M.
A surface current density

J
S
=

M n.
Proof of the Theorem:
The theorem is proved by showing that the vector potential due to the dipole
distribution in a volume V closed by a surface S can be written in the form
A =
1
4
0
c
2

M
R
dV +
1
4
0
c
2
M n
R
dS ,
where R is a distance from the centre of the volume dV closed by a surface S.
Figure 49: A geometry for the cal-
culations of the vector potential pro-
duced at point X by a current loop
of radius a.
Consider a current loop of radius a, as shown
in Figure 49. We will nd the vector po-
tential produced by the current loop at the
point X.
We start from the solution of the Maxwells
equations, which gives the (static) vector
potential of a current loop
A =
1
4
0
c
2

J
r
dV =
I
4
0
c
2

dl
r
. (70)
Due to the radial symmetry of the loop, it
is convenient to work in the polar spherical
191
coordinates, in which the current element

dl
can be written as
dl = ad

= a sin d

i + a cos d

j ,
and the distance from dl to X is given by
r =
(x a cos )
2
+ (y a sin )
2
+ z
2
1/2
,
which for a <R can be written as
r =
x
2
+ y
2
+ z
2
2ax cos + a
2
2ay sin
1/2
R
2
2ax cos 2ay sin
1/2
R
1
ax cos
R
2

ay sin
R
2
.
Hence
r
1
= R
1
1 +
ax cos
R
2
+
ay sin
R
2
,
where we have used the Taylor expansion of 1/(1 x) = 1 + x + . . ..
Thus, the vector potential (70) takes the form
A =
I
4
0
c
2
R
2
0
1 +
ax cos
R
2
+
ay sin
R
2
a sin d

i + a cos d

j
.
Since
2
0
sin d =
2
0
cos d =
2
0
sin cos d = 0 ,
the formula for

A simplies to
A =
I
4
0
c
2
R
3
2
0
a
2
y sin
2

i + a
2
x cos
2
d ,
A =
Ia
2
4
0
c
2
R
3
2
0
sin
2
d + x
2
0
cos
2
d
.
192
Next, since
2
0
sin
2
d =
2
0
cos
2
d = ,
we obtain
A =
Ia
2
4
0
c
2
R
3
i + x
. (71)
Using the relation
k

R =

k
x
R
i +
y
R
j +
z
R
=
x
R
j
y
R
i ,
we can write Eq. (71) as
A =
Ia
2
4
0
c
2
R
2
k

R =

4
0
c
2
R
R
2
=
1
4
0
c
2

1
R
,
where we have used the result
R
R
2
=
1
R
,
and the fact that is in the direction normal to the loop, i.e. in the z direc-
tion.
Summarizing what we have obtained so far: We have obtained a formula for
the vector potential produced by a single magnetic dipole moment .
Figure 50:
If we consider a set of dipole moments
and change, for a convenience, the direc-
tion of

R into
R, as shown in Figure 50,

we can write that the vector potential d
A
produced by a set of dipole moments con-
tained in a volume element dV is
d
A =
1
4
0
c
2
M
1
R
dV ,
193
where

M is the magnetic dipole moment
per unit volume.
Then, the vector potential produced by
the dipole moments contained in the total volume V is given by
A =
1
4
0
c
2

M
1
R
dV .
Using a vector identity
(
A) =

A +

A ,
we can write the integral function as
M
1
R
=
1
R

M =

M
R

M
R
,
and then the vector potential takes the form
A =
1
4
0
c
2

M
R
dV
1
4
0
c
2
M
R
dV . (72)
The rst term is in the form easy to interpret. Refering to the denition of
the vector potential, Eq. (53), we see that

M can be interpreted as the
volume current density. Thus, the rst term represents the vector potential
produced by the volume currents existing inside the material.
We still have to do something with the second term as it is in an unfamiliar
form.
In order to transform the second term into a familiar form, we apply a the-
orem that:
19
M
R
dV =
M n
R
dS . (73)
Thus, the application of the theorem to the second term in Eq. (72) leads to
the vector potential of the form
A =
1
4
0
c
2

M
R
dV +
1
4
0
c
2
M n
R
dS ,
19
Proof of the theorem is given in Appendix B.
194
or
A =
1
4
0
c
2
J
V
R
dV +
1
4
0
c
2
J
S
R
dS .
We see by refering to the denition of the vector potential that the eective
(Ampere) currents associated with a macroscopic dipole moment

M per unit
volume are:
(i) A current density

J
V
=

M throughout the volume.
(ii) A surface currents

J
S
=

M n.
Figure 51 illustrates the source of the macroscopic surface current.
Figure 51: Source of the macroscopic surface current.
13.2 The Magnetic Intensity Vector

H
When we were dealing with dielectric materials in the presence of electric
elds, it was convenient to introduce a new vector, the displacement vector
D, in order to eliminate the necessity of taking the electric dipole polariza-

tion

P of the material into account explicitly.
A similar procedure is used for the magnetic materials, where a new magnetic
intensity vector

H is introduced to eliminate the magnetization

M. We will
195
illustrate this idea for both static and time-varying elds.
13.2.1 Static Magnetic Fields
For a static current distribution, the Maxwells equation IV reduces to

B =
0
J .
In a medium where there are magnetic polarization currents as well as con-
duction currents we can write
J =

J
c
+

J
m
,
where

J
c
is the conducting current, and

J
m
is the magnetization current.
Thus
J =

J
c
+

M ,
and then

B =
0
(
J
c
+

M) .
This equation can be written as
(
B
0

M) =
0
J
c
,
or in the form
=

J
c
.
This shows that the vector

B/
0

M has as its source only the conduction
current

J
c
. Therefore, to eliminate the necessity of dealing directly with the
magnetization current

J
m
, we can dene a new vector
H =

M ,
which is called the magnetic (eld) intensity vector.
196
In terms of

H, the Maxwells equation IV for static elds in magnetic mate-
rials takes the form

H =

J
c
.
In dealing with magnetic materials we often know

J
c
but not

M (well not
directly anyway.) Think e.g. of an inductor lled with some magnetizable
material like iron. Then

H becomes useful. It is a way of avoiding a detailed
calculation of the polarization currents. The magnetic intensity

H is the
magnetic analogue of the dielectric displacement

D in the electric case. We
may drop the subscript c and write

H =

J ,
but we should remember that

J is now not the total electric current density
everywhere.
In summary: The use of the vector

H enables us to write the Maxwells
equation IV in terms of only the conducting current density in any magnetic
material. There is no need to deal with the magnetization

M.
13.3 Linear Magnetic Materials
For most materials (excluding ferromagnetics) the magnetization

M is lin-
early proportional to the applied external eld

B, and then because of the
linear relation
H =

M ,
the magnetization is also proportional to

H. Thus, for linear magnetic ma-
terials we write
M =
m
H ,
and then at any point the vectors

B,

M, and

H will be in the same direction,
and we get the following relation between

B and

H
B =
0
H +
0

M =
0
(1 +
m
)
H =
0
r

H =
H .
197
The parameter
r
= (1 +
m
) is called the relative permeability,
m
is the
magnetic susceptibility, and =
0
(1+
m
) is called the magnetic permeabil-
ity.
Since
B =
0
(
H +

M) ,
and we have dened

M such that
0

M =
0
H =

0
=

m
B
1 +
m
,
we get
H =
0
(1 +
m
)
.
Thus, if we know the material we use, we can nd

H.
Example: A solenoid lled with magnetizable material
In this example, we illustrate some interesting properties and re-
lations between magnetic eld vectors produced by a solenoid lled
with a magnetizable material. We will show how the eld

B is mod-
ied due to the presence of the material and what is the source of
the

H vector.
Consider the expression for the magnetic eld produced by a long
solenoid
B =
0
NI .
In the presence of the material, we should include the Ampere surface
currents as well as the conduction currents in the wire. Since
B =
0
I
,
where I
is the total current per unit length, we have

B =
0
(NI +M) =
0
(NI +
m
H) =
0
NI +

m
B
,
198
B
1

0
=
0
NI ,
or
B =

0
NI
1

0
m
=

0
NI
1
m
1+m
.
Note that if
m
is positive then B is greater than it would have
been in the absence of the magnetizable material. Evidently in this
case the macroscopic Ampere current is in the same sense as the
conduction current in the solenoid.
Since
1

m
1 +
m
=
1
1 +
m
=
1
r
,
we have for the magnetic eld
B =
r
0
NI = NI . (74)
Thus, the introduction of a magnetizable material into the solenoid
is equivalent to replace
0
by .
We may go further and consider the following question: What is the
source of the

H vector?
It is easy to answer this question. Since
B = H ,
we then have from Eq. (74) that
H = NI .
Thus,

H depends only on the parameters of the solenoid.
We see that the eect of lling the solenoid with a material of rela-
tive permeability
r
is to multiply B by a factor
r
(assuming the
current in the wire remains the same).
Example:
To illustrate further that

H depends only on the parameters of
the solenoid, consider an another example which shows that in a
solenoid, the magnetic intensity H is independent of the presence
199
or absence of the magnetic material.
As in the preceding example, we consider a solenoid lled with
a magnetic material. By the denition
H =
B
0
M .
When we remove the magnetic material, M = 0, and then
H =
B
0
=

0
NI
0
= NI .
With the material present
H =
B
0
M =
NI
m
H =
r
NI
m
H .
Hence
H = (1 +
m
)NI
m
H ,
H(1 +
m
) = (1 +
m
)NI .
Thus
H = NI ,
as before.
This gives rise to the notion of H as an inducing eld and B as a resultant
eld. This concept is much used in the study of magnetic properties of ma-
terials. The reason is that

H is the eld introduced to hide the properties of
the material. Thus,

H is the same independent of the presence or absence of
the material.
13.4 Non-Linear Magnetic Materials
There are materials, i.e. iron, whose the magnetic properties are not de-
scribed by the linear formula

B =
H. They are called ferromagnetic mate-

rials and have a specic property that placed in an external magnetic eld
200
their magnetization undergoes a saturation. This is because all the internal
current loops are lined up, which breaks the linear property. The reason for
this unusual behavior is that ferromagnetic materials do not have a unique
value of magnetic susceptibility because of strong magnetic nonlinearities.
For this reason, it is dicult to provide a relation between

B and

H, and
it is usually presented graphically in terms of the so-called hysteresis. An
example of the hysteresis is shown in Figure 52.
Figure 52: Hysteresis loop of a ferromagnetic material.
13.4.1 Work done in magnetization of iron
Think of the case of the solenoid of length , cross-section area A, and lled
with magnetic material. The applied voltage V is:
V = c =
d
dt
= N
d
dt
BA ,
where N is the number of turns.
The rate of doing work to power the solenoid is
P = V I = ANI
dB
dt
= 1 H
dB
dt
,
where 1 = Al is the volume of the solenoid.
201
In the time dt that it takes to change B by dB, then the work done is
dW = Pdt = 1 H dB. Thus, the work done per unit volume is
dW = H dB .
Hence, the work done per unit volume in one cycle of the hysteresis loop of
a non-linear material is given by the area of the loop.
Figure 53: An example of hysteresis loops of a hard (iron) and a soft (ferrite) fer-
romagnetic materials. Hard ferromagnetics retain some magnetization in the absence of
external elds. This property of hard ferromagnetics makes them useful for permanent
magnets.
It is sometimes useful to write
H dB =
0
H d(H + M) =
0
H dH +
0
H dM
=
1
2
0
H
2
+
0
H dM ,
where the rst term on the rhs is the work to establish magnetic eld, and
the second terms is the work by the eld H to establish magnetization dM.
Which part really uses the energy:
1
2
0
H
2
or
0
H dM ? To check it, con-

sider properties of magnetic materials in a time varying magnetic eld.
13.5 Permanent Magnetic Materials: Ferromagnets
We have so far considered magnetic materials that are called diamagnetics
and paramagnetics. In these materials the magnetization exists only in the
202
presence of an external magnetic eld, i.e.

M is a function of the external
eld, (

M

B). However, there is a class of materials, called ferromagnetics
or permanent magnets in which macroscopic magnetization exists even in the
absence of the external eld.
Consider rst an exercise, which will illustarte several points that are of in-
terest in designing of ferromagnetic materials.
Exercise in class. Magnetic eld of a solenoid
Consider a cylindrical current sheet on the surface of a cylinder of radius a,
called a solenoid. The lines of current ow are circles round the surface of
the cylinder.
Figure 54: Amp` eres loop to calcu-
late magnetic eld inside and outside a
solenoid.
As in the case of the plane sheet we
measure the strength of the current
by a current per unit length with the
length measured along the surface of
the cylinder parallel to the axis and
at right angles to the direction of cur-
rent ow.
(a) Assuming the magnetic eld within
the cylinder is uniform and parallel to the
axis and that the magnetic eld is zero
outside the cylinder show that within the
cylinder the eld is

B =
0
I where

I is
the current per unit length.
(b) Can you think of the arguments from Amp` eres theorem for the eld
being uniform within the cylinder and zero outside?
13.5.1

B and

H elds of a ferromagnet
We can now easily extend the arguments from the above exercise to analyse
the magnetic eld a long homogeneous ferromagnetic material. In this case

B
203
inside the material is due solely to the

M:
J
s
=

M n ,
or in terms of the magnitudes J
s
= M.
Let us nd

B and

H inside and outside the material.
Using Amp` eres law

B d
=
0
I, where I = J
s
= M, we obtain
B =
0
M i.e. B =
0
M ,
where we have used the fact that B is zero outside the magnet.
Thus
H =
B
0
M =

0
M
0
M = 0 ,
in the region where M = 0, i.e. inside the magnet.
Outside the magnet
H =
B
0
0 =
B
0
,
i.e.

H is just a scaled replica of

B.
204
13.5.2 Magnetic Poles
Since

B = 0 always and the lines of

B form closed loops, we have

H =
=

M .
Thus,

H has a source (eld lines start and stop) where

M varies i.e. at the
ends of the magnet (magnetic poles).
For a ferromagnet, we can write

H =

M =
m
.
Thus, we can think of
m
as a volume density of magnetic charge giving
rise to the

H eld. It must be stressed that this equivalence is purely math-
ematical, and does not prove a physical existence of magnetic charges.
Moreover, for the ferromagnet

H =

J
c
= 0 .
Note the similar mathematical properties of

H here to those of the

E eld
in electrostatics

E =

0
and

E = 0 .
Historically, magnetostatics of permanently magnetized materials (ferromag-
nets) developed via use of the

H eld and magnetic charges or poles. One
obtains a law analogous to Coulombs Law for the force between magnetic
poles.
205
Exercise in class: Plane magnetized material
An innite plane surface divides the universe into a vac-
uum on one side and a magnetic material on the other.
Within the magnetic material there exists a uniform
magnetic moment per unit volume

M which is parallel
to the surface.
(a) Show that while the direction of the magnetic in-
duction vector

B is dierent on the two sides of the
surface, its magnitude is given everywhere by:
B =
M
2
0
c
2
.
(b) Find the magnitude and direction of the magnetic
eld

B everywhere due to an innite plane parallel slab
of material of thickness d which is permanently uni-
formly magnetized with dipole moment per unit vol-
ume

M lying parallel to the bounding surfaces.
(c) Find the magnetic intensity

H everywhere.
206
13.6 Time Dependent Magnetic Fields and Energy Loss
In a time-dependent magnetic eld the magnetization

M may not stay in
phase with the driving eld

H. This corresponds to internal friction and
heat dissipation.
Let
H = H
0
cos(t) .
Then the magnetisation in a lossy material varies as
M = M
0
cos(t + ) ,
where M
0
and represent the amplitude and phase of the magnetization
response to the magnetizing eld. Thus
M = M
0
cos cos(t) M
0
sin sin(t) .
Now the work done in magnetization per cycle of the AC current producing
the magnetizing eld is
W =
0
H dM =
0
2/
t=0
H
dM
dt
dt .
Since
dM
dt
= M
0
cos sin(t) M
0
sin cos(t) ,
we have W = W
1
+ W
2
. Consider the term W
1
:
W
1
=
0
H
0
M
0
cos
2/
t=0
cos(t) sin(t) dt .
This term is zero - it represents reversible energy conversion to and from H.
Consider now the term W
2
:
W
2
=
0
H
0
M
0
sin
2/
0
cos
2
(t) dt .
207
This term represents work done against internal friction during magnetizing
and demagnetizing the material.
Since

cos
2
(t) dt is positive, sin must be negative so that work is done on
the material i.e. is negative.
13.7 The complex magnetic susceptibility
Let
H = H
0
e
it
,
which for a physical eld can be written as
H = H
0
cos(t) = Re
H
0
e
it
.
Then
M = M
0
e
i(t)
= M
0
e
i
e
it
or
M = (M
0
cos iM
0
sin )e
it
=
M
0
H
0
cos i
M
0
H
0
sin
H
0
e
it
.
This result can be written in terms of real and imaginary susceptibility
M = (
)H
0
e
it
= (
)H .
Using this complex number notation, we nd
B =
0
(H + M) =
0
H +
0
(
)H
= [
0
(1 +
) i
0
]H
and then
B = (
)H = H ,
where
=
0
(1 +
),
=
0
, and is the complex permeability.

208
13.8 Maxwells Equations in Dielectric and Magnetic
Materials
Let us complete the lectures on dielectric and magnetic properties of mate-
rials by incorporating our ndings into the Maxwells equations, to see how
the basic equations of electromagnetism are modied in the presence of the
material.
The Maxwell equation IV contains a current density term

B =
0
J +
0
E
t
.
We think this is always true provided

J is the total electric current den-
sity. Applying this in a region where there may be electric and magnetic
polarization eects we can write the current density as
J =

J
c
+

J
E
+

J
M
=

J +

P
t
+

M .
Thus, the Maxwells equation IV takes the form

B =
0
J
c
+
0
P
t
+
0

M +
0
E
t
,
=

J
c
+

t
(
0
E +

P) .
With the introduction of the new vectorial elds

D and

H, this equation
takes the form

H =

J +

D
t
,
where we must remember that

J represents the conduction current only.
209
In summary: The Maxwells equations for the electromagnetic eld prop-
agating in a material are of the following form

D = ,

B =

H = 0 ,

E =

H
t
,

H =

J +

D
t
,
but in general

H =

M .
These equations are supplemented by appropriate constitutive relations, which
connect the electric eld

E and the magnetic induction

B with the displace-
ment eld

D and the magnetic eld

H
D =
E and

B =
H .
These relations carry information about the material.
Revision questions
Question 1. State the denition of magnetization.
Question 2. How do we dene a volume current density and a surface current
density?
Question 3. Explain why it is useful to use the

H instead of

B for a magnetic
eld in a magnetic material.
Question 4. What is the form of the Maxwells equation IV for static elds in
magnetic materials?
210
Question 5. What it is a hysteresis loop, and what does it correspond to?
Question 6. What is meant by a homogeneous and isotropic material?
Question 7. How do we distinguish between soft and hard ferromagnetics?
Question 8. What are the modications to the Maxwells equations of an elec-
tromagnetic eld propagating in a material?
211
Tutorial problems
Problem 13.1 Field of a uniformly magnetized sphere
A magnetized sphere has a uniform dipole moment

M per unit
volume.
(a) Find and sketch a diagram of the magnetic polarization cur-
rents in and on the sphere.
(b) Apply the Biot-Savart law to this current distribution to nd
the magnetic ux density

B at the centre of the sphere. Show
that:
B =
2

M
3
0
c
2
.
Problem 13.2 Fields in an inhomogeneous magnetic material
A long solenoid of radius a and having N turns per metre car-
ries current I. It is lled with an inhomogeneous material with a
gradient in magnetic susceptibility. The susceptibility is zero at
the centre and increases linearly with radial distance r from the
centre of the solenoid, attaining a value of 2 at r = a.
(a) Find the vectors

B,

M,

H throughout the solenoid.
(Hint: Study

H rst by integrating the Maxwell equation

H =

J ,
and use the symmetries of the situation. Show that

H is un-
changed by the presence of the magnetizable material).
(b) Show that the self-inductance of the solenoid is the same
as would be obtained by replacing the above-mentioned magnatic
material with a homogeneous material having a magnetic suscep-
tibility of
4
3
.
(c) What are the magnetic polarization currents on the surface
and within the solenoid? Calculate them in terms of I.
212
Problem 13.3 Fields in a dense magnetic material
Using a method similar to that used in the case of electric elds
within dense dielectrics, show that if a spherical cavity is exca-
vated in a uniformly magnetized material, the magnetic eld

B
and the magnetic intensity

H in the hole and in the material are
related by
B
hole
=

B
material
2

M
3
0
c
2
,
H
hole
=

H
material
+

M
3
.
213
14 Poyntings Theorem Revisited
We have seen in Chapter 8 how energy of the electromagnetic eld may
be transported through vacuum (empty space) by means of electromagnetic
waves. We have shown that the direction of propagation of energy is de-
termined by the Poynting vector. In this lecture, we will reconsider the
Poynting theorem taking into account propagation of the electromagnetic
eld in magnetizable materials. A question we will try to answer: How the
energy is propagated inside a magnetizable material?
14.1 Poynting Vector in Terms of

E and

H
We have shown that it is useful to represent the magnetic eld in terms of

H
vector rather then

B vector when the eld propagates inside a magnetizable
material.
Using the concept of the

H vector for the eld in vacuum
H =
0
=
0
c
2
B ,
the Poynting vector of the EM eld in vacuum takes the form
N =
0
c
2
E

B =

E

H .
The cross product

E

H also turns out to be the correct expression for the
Poynting vector when magnetizable materials are involved and is the expres-
sion most commonly quoted for it.
In order to show this, we consider, as before in Section 8, a ow of the energy
through a closed surface S. Using the Gausss theorem and a vector identity
(
A

B) =

B (

A)

A (

B) ,
we can write
S
(
E

H) d
S =
V
(
E

H) dV
=
H (

E) dV
E (

H) dV .
214
Now, substitute from the Maxwell equations

E =

t
B ,

H =

J +

t
D ,
and obtain
S
(
E

H) d
S =
B
t
dV
D
t
dV
E

J dV .
The individual terms in this equation have the following interpre-
tation:
1. The lhs is the rate of ow of eld energy out of the volume V enclosed
by the surface S.
2. First term on the rhs is the rate of work in establishing the magnetic
eld in V .
3. Second term is the rate of doing work in establishing the electric eld
in V .
4. Third term is the rate of doing work on the currents in V .
We go further with the Poyntings Theorem.
If we substitute for

B =
H, we then get
B
t
dV =
1
2
H
2
dV =

t
1
2
B
2
dV ,
and
D
t
dV =
1
2
E
2
dV =

t
1
2
E
2
dV .
Hence
S
(
E

H) d
S =

1
2
E
2
+
1
2
B
2
dV

E

J dV .
215
Thus, the rate of ow of eld energy out of volume V is equal to the rate of
changing energy of the EM eld plus the arte of doing work on the currents
in V .
14.2 Poynting Vector for Complex Sinusoidal Fields
It is well known that the electromagnetic eld (e.g. light) is a real physical
quantity (observable). However, in the electromagnetic theory it is advanta-
geous to represent the real electromagnetic eld by complex sinusoidal quan-
tities because of its mathematical simplicity. In addition, what we usually
measure is the average intensity of the eld, '
E`, which is a real quantity.

For time varying elds, we usually write
E =

E
0
e
it
and

H =

H
0
e
it
,
where

E
0
and

H
0
are complex quantities including both amplitude and phase
information. We understand that the electric and magnetic elds are given
by the REAL PARTS of

E and

H.
The power of the complex exponential scheme lies in the fact that for oper-
ations such as summation, subtraction, integration etc., we take real parts
AFTER the operation. For example:
Re
E
1
+ Re
E
2
= Re
E
1
+

E
2
,
Re
d
E
1
dt
+ Re
d
E
2
dt
= Re
d
dt
E
1
+

E
2
.
However, some care has to be taken in evaluating the Poynting vector.
The Poynting vector is given by

N =

E

H, but if we write complex expo-
nential expressions for

E

H, we note that
N = Re
E
c
Re
H
c
= Re
E
c

H
c
,
where we put a subscript c to indicate that we are writing the elds using
complex exponentials.
Proof:
216
We can write the complex electric eld as
E
c
=

E
0
e
it
=
E
r
+ i
E
i
(cos t + i sin t)
=
E
r
cos t

E
i
sin t
+ i
E
r
sin t +

E
i
cos t
.
Similarly, for the magnetic eld we can write
H
c
=

H
0
e
it
=
H
r
+ i
H
i
(cos t + i sin t)
=
H
r
cos t

H
i
sin t
+ i
H
r
sin t +

H
i
cos t
,
where

E
r
,

E
i
,

H
r
and

H
i
are real vectors.
However, from the above, we see that
Re
E
c
=

E
r
cos t

E
i
sin t ,
Re
H
c
=

H
r
cos t

H
i
sin t .
Clearly, if we calculate Re(
E
c

H
c
) we get extra terms in addition to those
in the expression for Re
E
c
Re
H
c
, and then
Re
E
c
Re
H
c
= Re(
E
c

H
c
) ,
as required.
In experiments, we do not measure

N, but rather

N, the mean Poynting

vector averaged over the cycles of oscillations.
There is a useful expression for the MEAN POYNTING VECTOR in
terms of the complex exponential

E
c
and

H
c
:
N =
1
2
Re
E
c
=
1
2
Re
c

H
c
=

Re
E
c
Re
H
c
,
where the bar over

N indicates an average over the whole cycle of the sinu-
soidal eld to eliminate the rapid temporal oscillations at frequency .
Proof:
217
Calculate the Poynting vector involving measurable (real) parts of the com-
plex elds
N = Re
E
c
Re
H
c
=
E
r
cos t

E
i
sin t
H
r
cos t

H
i
sin t
=

E
r

H
r
cos
2
t +

E
i

H
i
sin
2
t (
E
r

H
i
+

E
i

H
r
) cos t sin t .
Since
1
T
T
0
cos
2
t dt =
1
T
T
0
sin
2
t dt =
1
2
,
and
1
T
T
0
cos t sin t dt = 0 ,
we obtain for the average Poynting vector
N =
1
2
(
E
r

H
r
+

E
i

H
i
) .
On the other hand, take
c
=
H
r
i
H
i
e
it
,
E
c
=
E
r
+ i
E
i
e
it
,
and then
(
E
c
c
) = (
E
r

H
r
+

E
i

H
i
) + i(
E
i

H
r

E
r

H
i
) .
Hence
1
2
Re
E
c
=
1
2
(
E
r

H
r
+

E
i

H
i
) ,
which is equal to

N, as required.
In summary: The average Poynting vector

N of complex exponential elds
satises the relation
N =

Re
E
c
Re
H
c
=
1
2
Re
E
c
,
and this is the expression for calculation of the Poynting vector of complex
elds.
218
Revision questions
Question 1. What is the form of the Poynting vector in terms of the

E and

H
vectors?
Question 2. How do we write the electric and magnetic eld vectors for oscillat-
ing elds?
Question 3. What is the useful form of the Poynting vector for complex sinusoidal
elds?
219
15 Plane Wave Propagation in Dielectric and
Magnetic Media
In this lecture, we shall examine in some details how existing radiation eld is
modied by the material it passes through. We will consider the propagation
in a linear and homogeneous material and as we shall see, the conductivity
is the most signicant parameter for modications, rather than the dielectric
and magnetic constants.
We have learnt that the properties of a lossy linear and homogeneous dielec-
tric material can be described using a complex permittivity and similarly,
the properties of a lossy magnetic material are described by a complex per-
meability. Thus, for a lossy linear and homogeneous material the Maxwells
equations describing the EM eld inside the material are of the form

E = / ,

H = 0 ,

E =

H
t
,

H =

J +
E
t
, (75)
where , are complex quantities that characterize the material, is a free
charge in the material, and

J is conduction current only, i.e.

J =
E.
Note from Eq. (75) that an EM eld propagating in a linear and homoge-
neous material is solely described by the

E and

H elds, and its properties
depend only on the material constants and . Thus, the only factors that
can modify the wave from that propagating in vacuum are the constants
and .
We remember that in the vacuum the EM eld propagates in the form of
plane transverse waves. Questions we will try to answer are: What are the
properties of the EM waves propagating inside a material? Does an EM wave
retain its transverse nature?
220
To answer these questions, consider a plane wave propagating in one dimen-
sion, the z direction
E =

E
0
e
i(tkz)
. (76)
First, we shall nd

H to check whether the wave retaines its transverse prop-
erties when propagating through the lossy material. Thus, for the electric
eld (76), the Maxwells equation IV takes the form

H =
E + i
E = ( + i)
E .
Since for a plane wave propagating in the z direction the derivatives /x
and /y of

E and

H are zero, we obtain
i

j

k
0 0

z
H
x
H
y
H
z
= ( + i)
E .
From the LHS, we see that
(

H)
z
= 0 ( + i)E
z
= 0 .
Hence E
z
= 0 unless ( + i) = 0. Thus,

E
k.
Since E
z
= 0, we have that

E = 0. Also

E =

H
t
can be used to show that

E

H.
In addition, we can nd from the above equation that H
z
= 0, which means
that

H
k.
We can summarize our thusfar ndings that the wave propagating
in a lossy material retains its transverse nature.
However, the losses are expected to modify somehow the wave. What kind
of modications are made by the losess?
221
Since

E = 0, the quantity occurs only in the equation for

H, and
it is common to proceed as follows:

H =

J +
E
t
=
E + i
E ,

H = i

i
+

E = i

E .
Now
i
= ,
which gives

H = i
E .
Physically what has been done is to lump together the conduction and the
lossy dielectric constant term. To an external observer they are inseparable.
Only using some theory of the internal structure of the dielectric, they can
be separated.
We can summarize that the electric and magnetic elds of the propagating
wave in a lossy conducting material satisfy the following equations

E = 0 , (77)

B = 0 , (78)

E =

H
t
= i
H , (79)

H = i
E , (80)
where we have left the bar o .
15.1 Dispersive Equation and Complex Propagation
Number
We now proceed to solve the propagation equations (77)(80). We will see
that the fact that and are complex numbers, the solution leads to a dis-
persive equation.
222
Thus, we look for plane wave solutions, and will try to nd how the propa-
gation number k behaves.
The procedure is as follows: Taking of (79) and using (80), we obtain
(

E) =
2
E ,
which, after applying the double vector identity, simplies to
E =
2
E .
Since

E of the transverse EM wave is independent of x and y, we have
E
z
2
=
2
E .
On the other hand, the second derivative of

E, given in Eq. (76), is
E
z
2
= k
2
E .
Hence, we obtain a dispersion equation
k
2
=
2
.
This dispersion equation is not as simple as it looks. We cannot just say that
phase velocity is
v
p
=

k
=
1
,
as we usually do for EM waves propagating in vacuum, because and are
complex quantities and then k is a complex number.
What does an imaginary value of k mean?
To give the answer to this question, we write the complex number k as
k = i ,
where and are real. Then
E = E
0
e
i(tkz)
= E
0
e
i(tz+iz)
= E
0
e
z
e
i(tz)
.
223
Clearly from this, the phase velocity is
v
p
=

,
i.e. the real part of k plays the role of the propagation number, and =
Im(k) is the attenuation (losses) coecient.
In practice, the losses come from (1) Conduction currents, (2) Lossy dielec-
tric, and (3) Lossy magnetic material, which are lumped together in and .
We can nd the explicit forms of and .
To do this, we write the complex number k as
k = ()
1
2
=
)
1
2
,
k =
1
2
.
Let us introduce a short notation for the real and imaginary parts of k:
p =
,
q =
.
Then
k = (p iq)
1
2
,
k =
p
2
+ q
2
1
2
e
iarctan(q/p)
1
2
,
or
k =
p
2
+ q
2
1
4
e
i
,
224
where =
1
2
arctan (q/p). Hence
= Re(k) =
p
2
+ q
2
1
4
cos ,
= Im(k) =
p
2
+ q
2
1
4
sin ,
which in general are quite complicated and dicult to interpret. The inter-
pretation becomes more transparent when we make some simplications that
are discussed in the following two examples.
15.2 Wave Refraction and Attenuation
We now consider two examples of the propagation of an EM wave in dierent
materials: Low loss dielectrics and conducting materials.
15.2.1 Propagation of an EM wave in a low loss dielectric
In the rst example, we illustrate the propagation of an EM wave in a low
loss dielectric with no magnetic eects.
For dielectrics, is negligible and also we disregard magnetic properties.
Thus, for a low loss dielectric
<
,
and
=
0
=
1
0
c
2
.
Hence, the propagation number simplies to
k =

0
c
1
2
=

0
c
)
2
+
2
1
2
e
i
1
2
,
225
where
= arctan
,
as the losses are small. Then
k =

c
0
e
i/2
=

c
0
e
i
+/
2
.
Hence, the real part of k takes the form
=

c
0
cos
+ /
2
0
=

c
r
.
We can then nd phase velocity of the wave and refractive index of the
dielectric material
v
p
=

= c
< c for
>
0
,
n =
c
v
p
=
0
=
r
> 1 .
Thus, we may conclude that inside the dielectric, the EM wave
propagates with a phase velocity v
p
< c and can be refracted from
the original direction on the entry to the dielectric.
Similarly, we can nd
= Im(k) =

c
0
sin
+ /
2
.
Since for small , sin , we get

c
+ /
2
=

c
+ /
2
0
.
Thus, losses (absorption of the wave) are small.
226
We can summarize, that the theory predicts that the refractive index for a
lossless dielectric is given by
n =
r
.
Table below compares theoretical values of n with that obtained experimen-
tally. An excellent agreement is observed, except for polar molecules. For
example, water has a refractive index of 1.33, whereas

r
= 9. The reason
for this discrepancy lies, not in the inadequacy of the laws of electromag-
netism, but rather in the tacit assumption that the dielectric constant is a
constant independent of the frequency. The refractive index of the water
exhibits a strong dependence on as it is composed of polar molecules that
the polarization depends strongly on .
r
n (experimental)
Air 1.00029 1.00029
Argon 1.00028 1.00028
CO
2
gas 1.00047 1.00045
Benzene 1.49 1.48
Ethanol 5.3 1.36
NaCl 2.47 1.54
Water 9.0 1.33
15.2.2 Propagation of an EM Wave in a Good Conductor
As a second example consider a propagation of the EM eld in a good con-
ductor with no dielectric and magnetic losses:
= 0.
Examples of good conductors: Cu, Ag.
Consider the Maxwell equation IV, which for an EM wave propagating in a
conducting material can be written as

H =

J +
E
t
=
E + i
E . (81)
227
Denition of a Good Conductor
Good conductor is when the conduction term on the right-hand
side of the Maxwell equation (81) dominates over the displace-
ment current, i.e. when . On physical grounds, for a good
conductor the conductivity current is much larger than the polar-
ization current.
For example, for copper at = 1 MHz = 2 10
6
rad
0
=
5.8 10
7
2 10
6
8.85 10
12
= 1.0 10
12
.
Consider general expression for k:
k =
1
2
,
that for a conductor with no dielectric losess reduces to
k =
0) i
+ 0
1
2
,
and nally
k =
1
2
.
Now we might as well drop the dashes on , understanding that they are
real quantities, and obtain
k =
1
2
=
1 i

1
2
,
that can be written as
k =
1 +
2
1
2
e
i arctan

1
2
.
228
Remembering that for a good conductor

1. Thus, we can drop 1 in

the [ ] brackets, and get
k =
e
i
2
1
2
=
e
i
4
.
Since cos(/4) = sin(/4) = 1/
2, we obtain
k =
2
i
1

2
(1 i) .
Then
= Re(k) =
2
= = Im(k) .
Knowing , we can nd the phase velocity of the wave
v
p
=
,
which shows that the velocity of propagation in a good conductor depends
on the frequency, even if and are assumed independent of frequency.
15.2.3 Skin Eect
Since the propagation number is a complex number and for a good conduc-
tor = , there is a very heavy attenuation of a propagating wave. For a
complex propagation number, we can write the electric eld as
E = E
0
e
z
e
i(tz)
.
We see, that the amplitude of the electric (and also magnetic) eld decreases
exponentially as z increases.
Since we can write
= =
2
,
229
the distance for attenuation e fold (i.e. amplitude falls to a factor
1
e
of its
original value in a distance
1
) is
z =
1
=

2
,
i.e.
=
1
=

2

6
!!
where is called the skin depth in the conductor. It is the distance the wave
must propagate in order to decay by an amount e
1
. This eect is sometimes
called skin eect as with an increasing the current ows in a narrower
and narrower layer, until in the limit of a true current exists only on
the surface of the conductor.
230
Revision questions
Question 1. Show that the propagation number of an electromagnetic eld in
dielectric and magnetic materials is a complex number.
Question 2. What are the physical consequences of the propagation number of
an electromagnetic eld being a complex number?
Question 3. What is the denition of a good conductor?
Question 4. Why the theoretical and experimental values of the refractive index
of water are signicantly dierent?
Question 5. What is meant by the skin eect of a propagating wave?
Tutorial problems
Problem 15.1 Relations between the electric and magnetic eld vectors in dielec-
tric and conductor
(a) Show that for a plane wave propagating in a dielectric along
the x-direction, the electric and magnetic eld vectors satisfy the
relations
E
y
x
=
B
z
t
,
E
z
x
=
B
y
t
,
H
z
x
=
D
y
t
,
H
y
x
=
D
z
t
.
(b) Show that for the propagation in a conductor
E
y
x
=
B
z
t
,
H
z
x
=
D
y
t
E
y
,
where is the conductivity of the conductor.
231
(c) Show that E
y
and H
z
each satises a second order dierential
equation of the form
2
E
y
x
2
=
2
E
y
t
2
+
E
y
t
.
Problem 15.2 Energy ow in a conductor
Using the relations of Problem 15.1, part (b), show that in a
conductor the rate of decrease of electromagnetic eld energy in a
volume element dV is equal to the net rate at which energy ows
out of this volume element plus the rate of Joule heating in the
element.
Problem 15.3 Practical example on propagation and attenuation
If a eld propagates as a plane wave in the form:
E = E
0
e
i(tkz)
,
where is the (real) radian frequency and k is a complex propa-
gation constant that can include the eects of losses due to con-
duction, dielectric loss and magnetic loss, in lectures we derived
a dispersion equation
k =
1
2
.
Consider propagation at a frequency of 1 MHz in a medium with
the following properties:
1. Conductivity = 10
9
mho m
1
.
2. Dielectric susceptibility
e
= 3 i10
4
.
3. Magnetic susceptibility
m
= 2 i10
3
.
(a) Calculate the phase velocity and wavelength at 1 MHz.
(b) Compare these values with those for free space propagation
at 1 MHz.
232
(c) Calculate the attenuation coecient at 1 MHz. How far will
the wave propagate before its amplitude falls to one tenth of its
initial value?
(d) Is this material a good conductor at 1 MHz? Over what
frequency range would it be classied as a good conductor?
233
16 Transitions Across Boundaries for Elec-
tromagnetic Fields
In the lectures on electric and magnetic properties of materials we have
conned our attention principally to the case of a single material medium
completely occupying the spatial region where electric and magnetic elds
existed. We now investigate the relations between the elds which hold at
a boundary between two dierent materials. They are of great help in the
calculations of propagation problems in optics, where light can propagate
between dierent material media.
Across boundaries between dierent materials there are sharp changes in
electrical properties of , , . On a macroscopic scale the elds may have to
be regarded as varying discontinuously across such boundaries. The source
of such discontinuities will be the surface polarization charges

P n and cur-
rents

M n discussed previously.
The question we will address in this lecture is: How the electromagnetic elds
transfer through a boundary between two materials? In the analysis of the
transfer properties of the electromagnetic eld across a boundary between
two dierent materials, it is convenient to decompose the elds into two
components, normal and tangential to the boundary between the materials
and study how each of the components transfers through the boundary.
16.1 Normal Components of the Fields
We will use the Maxwells divergence equations I and II to investigate the
transition of normal eld components at a boundary between two materials.

B
Consider two materials with dierent constants and , as shown in the
Figure 55. Suppose that a magnetic eld

B propagates from material (1)
to (2). We expect that

B will be dierent in dierent materials due to the
presence of dierent surface currents.
234
To check this, we rst apply the Maxwells equation II,

B = 0, to nd
how the normal component of

B transfers through the boundary between
the two materials.
Figure 55: A Gausss surface for the deriva-
tion of the boundary condition on the normal
component of

B.
Using the Gauss divergence theo-
rem, the Maxwells equation II can
be written in the integral form as
B ndS = 0 , (82)
where S is an arbitrary surface clos-
ing some area on the boundary
plane.
In order to evaluate the integral,
we consider a thin cylinder of
area S and thickness including
the boundary, as shown in Figure 55. In this case, we can distinguish three
surfaces that we represent by unit vectors n
1
, n
2
and n
3
, and therefore the
surface integral in Eq. (82) splits into three terms
(

B
2
n
1
) S + (

B
1
n
2
) S +
sides
B n
s
dS = 0 ,
where the rst term is the ux of the magnetic eld through the top surface,
the second term is the ux through the bottom surface, and the third term
is the ux through the side surface of the cylinder.
Since

B is nite everywhere and we are interested in the transformation of
the eld at the boundary, we make the height tend to zero, 0. Then,
the integral
sides
B ndS 0 as 0 .
Thus, we nd that the total ux of the magnetic eld

B over the surface has
contributions from the top and bottom ends only
B
2
n
1
+

B
1
n
2
= 0 . (83)
235
Now, if we introduce the notation
n
1
= n then n
2
= n ,
and we nd that Eq. (83) becomes
n (

B
2

B
1
) = 0 ,
or equivalently, we may write that
B
2
= B
1
,
where B
1
is the component of the eld

B in material (1) normal to the
boundary and B
2
is the component in material (2) normal to the boundary.
Thus, the normal component of

B is continuous at all points across
a boundary separating two dierent materials.
On physical grounds, we can understand this result by noting that the mag-
netic elds of polarization currents

M
1
n
1
and

M
2
n
2
are parallel to
the boundary and so do not aect the normal component of

B, as shown
in Figure 56.

H
Since

B =
H, we have
B
1
=
1
H
1
=
2
H
2
= B
2
.
Figure 56: Direction of magnetizations and re-
sulting polarization currents at the boundary be-
tween two dierent materials.
Thus, we obtain that
H
1
=

2
1
H
2
,
which shows that the nor-
mal component of

H is not
continuous across a boundary
between two dierent materi-
als since
1
and
2
are not
equal.
On physical grounds, this result
for the normal component of

H
is the consequence of dierent magnetizations of the materials, when
1
=
2
.
236

D and

E
Since in nonconducting dielectrics

D = 0, an identical argument to that
applied the above for

B will show that the normal component of

D is
continuous across a boundary.
In the case of dielectrics we write

D =
E. Since the normal component of

D
is continuous across a boundary, we have
1
E
1
=
2
E
2
,
which shows that the normal component of

E is not continuous across the
boundary.
Figure 57: Polarization charges at the
boundary between two dierent materials.
On the physical grounds, we can
understand it as follows. The
electric eld of the dielectric sur-
face charge

P n is normal to
the boundary, as shown in Fig-
ure 57. Since the polarizations
of the dielectrics are dierent, so
the elds of the dielectrics are
dierent. This dierence causes
the discontinuity in the incident

E
eld.
The discontinuity of the normal
component of the electric eld leads to a change of the direction of prop-
agation of the eld, as it is illustrated in Figure 58. The change of the
direction depends on the ratio
2
/
1
. When
2
/
1
> 1,
2
>
1
.
16.2 Tangential Components of the Fields
Dierent boundary conditions apply to the components of the elds parallel
(tangential) to a boundary between two dierent materials. We will use the
Maxwells curl equations III and IV to investigate the transition of tangential
eld components at a boundary between dierent materials.
237
Figure 58: Boundry between two dielectrics showing change of the direction of propa-
gation of an electric eld.

E
For the tangential component of

E, we will apply the Faraday induction law,
the Maxwells equation III, to a closed path such as shown in the Figure 59.
Figure 59: A closed loop for the derivation
of boundary conditions for the tangential com-
ponent of

E.
We choose a closed loop in which
the sides perpendicular to the
boundary are made innitely short
compared to the parallel sides L.
Consider the Maxwells equation III

E =
B
t
.
Integrating both sides of this equa-
tion over the surface a and applying
the Stokess theorem, we obtain
E d
l =
B
t
n
0
da .
Hence
E
2
t
2
+

E
1
t
1
L +
ends
E d
l =
B
t
n
0
da , (84)
238
where n
o
is the unit vector normal to the surface a, and

t
1
and

t
2
are unit
vectors along the paths L on the side (1) and (2), respectively.
20
As dl 0 the right-hand side and the second term on left-hand side of
Eq. (84) go to zero, since

E and

B are nite everywhere. In this limit, the
area enclosed by the path approaches zero.
Since,

t
2
=
t
1
=

t, we then have
E
2

E
1
t = 0 .
But

E
t is the component of

E tangential to the surface. Since this is true
for any

t, we obtain E
1
= E
2
, where E
1
and E
2
are the components of

E
1
and

E
2
parallel to the boundary. Thus, we may conclude that:
The tangential component of

E is continuous across a boundary
between two dierent dielectric materials.
Figure 60: Illustration of the direction of the
vector n

E.
The continuity of a tangential com-
ponent can be written in an equiv-
alent form as
n
E
2

E
1
= 0 .
Explanation
From Figure 60, we see that the
cross product n

E can be writ-
ten as
n

E
= E sin(/2 ) = E cos .
In other words, n

E is equal to the projection of the vector

E on the
boundry. Thus, n

E is tangential to the boundary, i.e. it is the tangential
component of

E.
20
Do not mix the unit vector n
o
normal to the surface a with the unit vector n normal
to the boundary. Actually, n
o
n.
239

H
Consider the Maxwells equation IV:

H =

J
c
+
E
t
,
where

J
c
is the conduction current, i.e. not counting polarization currents.
The application of the Stokess theorem then gives
H d
l =
J
c
nda +
E
t
nda .
Both terms on the right-hand side go to zero as l 0 because

J
c
and
E/t
are nite.
21
Hence
H
2

H
1
t = 0 ,
or
n
H
2

H
1
= 0 .
Thus, we conclude that the tangential component of

H is continuous
at all points across a boundary between two dierent materials.
16.2.3 Tangential component of

B
We may easily deduce that the tangential component of

B is not in general
continuous across a boundary because of the presence of the magnetic polar-
ization surface currents

M n, which do not have a nite current density as
they ow in an innitely thin surface layer.
Thus, if we examine the corresponding Maxwell equation for
B, the term
in the integral involving

J may stay nite as j 0.
21
Innite current density can occur in a material which has innite electrical conduc-
tivity, but here we ignore this special case.
240
This conclusion can be justied as follows. Since
H =
0
c
2
B

M ,
we have
H
2

H
1
=
0
c
2
B
2

B
1

M
2

M
1
.
Figure 61: Directions of magnetizations
and resulting magnetic elds produced at
a boundary between two materials.
Take scalar product of both sides
with

t, and using the fact that the tan-
gential component of

H is continuous
across a boundary
H
2

H
1
t = 0 ,
we obtain
B
2

B
1
t =
1
0
c
2

M
2

M
1
t .
It is seen that if

M
1
=

M
2
, the dier-
ence between the two magnetizations
generates a discontinuity in the

B
eld, as shown in Figure 61.
Summarizing:
Field components that are continuous across a boundary between
two materials:
The normal component of

D.

E.

B.

H.
241
Revision questions
Question 1. Which of the Maxwells equations are used to analyse the transmis-
sion of the normal components of the EM eld through a boundary
between two materials?
Question 2. Which of the Maxwells equations are used to analyse the trans-
mission of the tangential components of the EM eld through a
boundary between two materials?
Question 3. Explain, using physical grounds, why the normal component of the
magnetic eld

B is continuous across a boundary between two ma-
terials.
Question 4. Explain, using physical grounds, why the tangential component of
the electric eld

E is continuous across a boundary between two
materials.
Tutorial problems
Problem 16.1 Change in direction of a eld line
A (static) electric eld line makes an angle
1
to the normal to
the boundary between two dielectrics. The dielectric constant on
the incident side is
1
and on the other side is
2
. There are no
local charges other than the polarization charges in and on the
dielectrics. Show that the angle the electric eld makes to the
normal to the boundary in the second dielectric is
2
= arctan
1
tan
1
.
Problem 16.2 Angles and polarization charges
An innite slab having a dielectric constant of 4 is placed in a
vacuum which is permeated by an electric eld

E
0
.
242
(a) Find the angle that

E
0
should make to the boundary normal
so that

E inside the slab makes an angle /4 with the boundary
normal.
(b) Find the surface density of polarization charge on the slab
in terms of E
0
.
Problem 16.3 Wave reection from a dielectric at normal incidence
Consider waves normally incident on a boundary between two di-
electrics having dielectric constants 4 (on the incident side) and 9.
(a) Draw a diagram showing the directions and magnitudes of

E
and

B in the incident, reected and transmitted waves. Apply
the general boundary conditions for

E and

B directly. Remem-
ber that B = E/v
p
.
(b) Show that the magnitude of the amplitude reection coef-
cient is 0.2.
243
17 Propagation of an EM Wave Across a Bound-
ary
In the preceding lecture, we have derived relations for transmission of electric
and magnetic elds through a boundary between two non-conducting mate-
rials. We have treated each eld separately. We now proceed to consider a
propagation of an EM wave across a boundary between two materials. We
shall nd that for the propagation of an EM wave the situation is dierent
that we cannot treat the elds separately. The boundary conditions must
be satised simultaneously for both the electric and magnetic elds. This
leads to a modication that at a boundary between two dierent materials,
the general boundary conditions cannot be satised by a transmitted wave
alone. There has to be a reected wave also. What happens is that the inci-
dent radiation is absorbed by charges in the boundary and reradiated in all
directions. The waves interfere destructively except in two directions along
those of the reected and transmitted waves. The directions, amplitudes
and phases of the reected and transmitted waves can be derived from the
general boundary conditions already obtained. This is an explanation often
presented in texbooks on classical optics.
In this lecture, we shall attack these problems from the standpoint of the elec-
tromagnetic theory of light and will show how the propagation phenomenon
can be understood with the help of the Maxwells equations. More precisely,
in this and the following lecture, we consider a number of well-known op-
tical phenomena and will show that they can be explained in terms of the
electromagnetic theory of light.
244
17.1 Representation of Plane Waves in Dierent Di-
rections
Let us rst set up the formalism we shall use in the study of propagation
of an EM wave through a boundary between two materials. The question
we will try to answer is: What is the convenient method to represent plane
waves propagating in dierent directions?
Figure 62: Illustration of the wave front mon-
itored by an observed O distant r from the
point A.
Suppose that a plane wave prop-
agates in the z direction. Then
E =

E
0
exp[i (t kz)] .
Let n
p
is the unit normal to
the phasefront (wavefront) of the
propagating wave, and r is a po-
sition vector that is independent
of n
p
, as shown in Figure 62. The
vector r determines position of
a wave-front as seen by an ob-
server O.
We see from Figure 62 that the
distance z the wave propagated in time t is
z = n
p
r .
Hence, we can write
E =

E
0
exp[i (t n
p
r k)] .
Thus, if at the point A appear few plane waves propagating in dierent di-
rections, the observer can distinguish them by dierent ns. In other words,
the vector r sets a reference frame for the observation of dierent waves at
the point A.
Remember, that an EM wave is composed of the electric and magnetic elds,
so we have to consider the propagation of both elds. However, there is
a useful relation between

B and

E of a plane wave that we can nd the
magnetic eld from the knowledge of the electric eld.
245
17.1.1 Representation of

B in terms of

E
The magnetic eld of a plane wave propagating in the direction determined
by the unit vector n
p
can be found from the relation
B =
k
n
p

E . (85)
Proof:
We shall show that the relation (85) arises from the Maxwell equation III:

E =
B
t
.
To prove it, we expand both sides of this equation, and obtain
i

j

k
0 0

z
E
x
E
y
0
B
t
=
i
B
x
t

j
B
y
t
,
where we have used the fact that the wave is transverse and propagates in
the z direction.
Comparing the coecients standing at the same unit vectors, we nd that
the x component gives
E
y
z
=
B
x
t
ikE
y
= iB
x
or
E
y
B
x
=
k
.
The y component gives:
E
x
z
=
B
y
t
ikE
x
= iB
y
or
E
x
B
y
=

k
.
246
Thus
E =
E
2
x
+ E
2
y
1
2
=

k
B
2
x
+ B
2
y
1
2
=

k
B .
Since

E,

B, n
p
, are mutually orthogonal,

E

B gives the direction of n
p
(Poynting result) of the propagation direction. Then n
p

E gives the direc-
tion of

B. Hence, written vectorially
H =
=
k
n
p

E . (86)
If and are real, k =
, and then we have
H =
n
p

E . (87)
In the next few lectures, we will use the continuity conditions for

E and

H to
analyse dierent optical phenomena. With the relations (86) or (87), we will
be able to limit the analysis to the electric eld alone, as knowing properties
of

E, we can infer from Eq. (86) properties of

H.
17.2 Directions of Reected and Transmitted Waves
Armed with the formalism to study propagation of an EM wave between
two materials, we now consider some familiar elementary results of classical
optics on reection and transmission to show that they can be derived from
the Maxwells equations.
Figure 63: Propagation vectors for the in-
cident, reected and refracted beams at a
boundary between two materials.
The most useful results in this
connection concern the continu-
ity of tangential components of
the

E and

H elds. How-
ever, most of the results can
be derived from the fact that
the tangential component of

E
is continuous across a bound-
ary.
247
Consider properties of an EM wave
(radiation beam) at a boundary be-
tween two materials, as illustrated
in Figure 63.
Let the origin of position coordinates r be located on the boundary S. Then
the electric elds for incident, reected and transmitted waves are
E
i
=

E
0
exp [i (t n
i
r k
1
)] ,
E
r
=

E
1
exp [i (t n
r
r k
1
)] ,
E
t
=

E
2
exp [i (t n
t
r k
2
)] .
These equations determine directions of the three electric elds relative to
the direction of observation r. Note that we have assumed the presence of
the reected wave without any particular reasons.
Why do we need the presence of the reected wave in the propa-
gation of an incident wave through the boundary?
The answer to this question is provided by the requirement that the tangen-
tial components of

E and

H must be continuous through the boundary.
Suppose that

E is polarized in the plane of incidence, i.e plane containing n
and n
i
. Then, the following two equations result from the continuity condi-
tions
E
i
cos
i
E
r
cos
r
= E
t
cos
t
, (88)
and
H
i
+ H
r
= H
t
. (89)
Since
H =
k
E ,
the relation (89) takes the form
E
i
+ E
r
=
k
2
k
1
E
t
, (90)
248
as for a dielectric
1
=
2
=
0
.
Now we see that without E
r
, Eq. (88) gives
E
i
=
cos
t
cos
i
E
t
,
while Eq. (90) gives
E
i
=
k
2
k
1
E
t
.
Thus, without E
r
we would get two dierent values for E
i
or E
t
, which we
cannot accept as both continuity conditions Eqs. (88) and (89) must be satis-
ed at the same moment. Hence, we conclude that the continuity conditions
for

E and

H will be satised only if E
r
= 0.
The same argument applies to the need of the transmitted wave. Without E
t
,
we get two equations
E
r
=
cos
i
cos
r
E
i
and E
r
= E
i
,
which evidently cannot be in general satised simultaneously.
Now, we will try to answer the following question:
What are the relative directions of the three waves represented
here by the unit vectors n
i
, n
r
and n
t
?
To answer this question, we use the boundary condition for the tangential
component of

E:
n
E
i
+

E
r
= n

E
t
,
i.e.
n
E
0
exp [i (t n
i
r k
1
)] +

E
1
exp [i (t n
r
r k
1
)]
= n
E
2
exp [i (t n
t
r k
2
)]
.
249
This relation must hold over the whole surface S for all r (subject to nr = 0).
Thus, the exponential phase factors must all be the same. Otherwise, if it
was true for one r it would not be true for other rs, but we have a freedom
of choosing r. Hence
k
1
n
i
r = k
1
n
r
r = k
2
n
t
r ,
from which we have
n
i
r = n
r
r , (91)
and
n
i
r =
k
2
k
1
n
t
r .
These relations will help us to prove that:
Incident, reected and transmitted waves are coplanar.
In other words, n
i
, n
r
, n
t
are coplanar, the property observed in experiments.
To show this, we rst dene planes determined by pairs of the unit vectors
( n
i
, n), ( n
r
, n) and ( n
t
, n), and next will show that all the pairs of the unit
vectors form the same plane.
It is well known from the vector analysis that a cross product between two
vectors determines a plane in which these two vectors are. Thus, we will
determine the cross products n
i
n, n
r
n and n
t
n, and then will show
that the cross products are equal.
In order to nd the cross products, we use the following relation valid for
arbitrary r and n, such that n r = 0.
r = n ( n r) . (92)
Proof:
Since
n ( n r) = ( n r) n ( n n)r ,
250
and n r = 0 as the vector r lies on the plane S, we obtain
n ( n r) = r ,
as required.
Figure 64: Illustration that the inci-
dent and reected beams are coplanar.
Using Eq. (92), we can then write
Eq. (91) as
n
i
[ n ( n r)] = n
r
[ n ( n r)] .
Interchanging () and () products, we
nd
( n
i
n) ( n r) = ( n
r
n) ( n r) .
This must be true for all r in
plane S. Thus:
n
i
n = n
r
n .
This implies that n
r
is in the plane of incidence, i.e. the plane containing n
and n
i
, as shown in Figure 64.
Similarly, the relation
k
1
n
i
r = k
2
n
t
r
implies that
k
1
n
i
n ( n r) = k
2
n
t
n ( n r) .
As before, by interchanging () and () products, we nd that
k
1
( n
i
n) ( n r) = k
2
( n
t
n) ( n r) .
This relation is valid for all r in S. Thus
k
1
n
i
n = k
2
n
t
n .
This means that n
t
is in the plane of incidence. Thus, we may conclude that
n
i
, n, and n
t
are coplanar.
The coplanar property of the waves is observed in any experiment. Note that
what we have just shown is an another example of a remarkable triumph of
the Maxwells electromagnetic theory.
251
17.3 Angle of Reection and Snells Law of Refraction
We now illustrate that other familiar laws of elementary optics can also be
derived from the Maxwells equations.
We have already shown that
n
i
n = n
r
n ,
from which and from the denition of the cross product, we have that
sin
i
= sin
r
.
Hence
i
=
r
.
Thus, the angle of incidence equals the angle of reection, another
familiar law of elementary optics.
Moreover, we have shown that
k
1
n
i
n = k
2
n
t
n ,
from which we have
k
1
sin
i
= k
2
sin
t
,
and then
sin
i
sin
t
=
k
2
k
1
= n
12
=
n
2
n
1
.
This is the well-known law of refraction in optics, called the Snells law.
In the case where k
1
and k
2
are purely real (e.g. in dielectrics), the refractive
index has a simple physical interpretation
n
12
=
k
2
k
1
=
/k
1
/k
2
=
v
1
v
2
,
i.e. the refractive index in equal to the ratio of phase velocities.
252
Revision questions
Question 1. Why there are three waves, incident, reected and refracted in a
propagation of an EM wave between two dielectric materials?
Question 2. What is a plane of incidence?
Question 3. Incident, reected and transmitted waves are coplanar: True or
false?
Question 4. State the Snells law, and briey explain how it is derived from the
Maxwells equations.
Question 5. How does the refractive index depend on phase velocities of an EM
wave propagating between two dielectric materials?
253
18 Fresnels Equations
In the propagation of an EM wave between two dierent materials, it is im-
portant to know how much of the energy of the incident beam is reected
and transmitted. In this lecture, we will nd relations between the ampli-
tudes of the incident, reected and refracted beams. More precisely, we will
express the amplitudes of the reected and transmitted beams in terms of
the amplitude of the incident beam. We shall see that the relations depend
on the angle of propagation of the incident beam and the material constants.
The boundary condition on tangential

E does not give sucient information
to calculate

E
r
and

E
t
in terms of

E
i
. For a given

E
i
there are still two
unknowns in the equation for continuity of tangential

E viz

E
r
and

E
t
. We
need a second relation between

E
i
,

E
r
and

E
t
. This can be obtained from
the continuity of the magnetic eld.
We know that tangential component of

B is not continuous across a boundary
because of the presence of

M n surface currents in magnetized materials.
To allow for such possibilities we can use the more general condition that
tangential component of

H is continuous across a boundary, i.e.
n
H
i
+

H
r
= n

H
t
.
Since
H =
=
k
n
p

E ,
where n
p
is a unit ray vector in the direction of propagation, we have an
equation
n
k
1
1
n
i

E
i
+
k
1
1
n
r

E
r
= n
k
2
2
n
t

E
t
, (93)
which together with
n
E
i
+

E
r
= n

E
t
, (94)
contains sucient information to determine

E
r
and

E
t
in terms of

E
i
.
254
Since these two equations involve vectors of an arbitrary polarization, the
solution of these two equations is greatly facilitated by decomposing each
of the vectors into two components: electric eld components parallel and
normal to the plane of incidence. Then, we can solve these two cases sep-
arately that are necessary and sucient to determine the relations between
the amplitudes valid for an arbitrary polarization since a superposition of
these two cases gives the solution for an arbitrary polarized incident wave.
Equations (93) and (94) also provide a simple explanation of why we need
reected and transmitted elds at the boundary to obtain the correct results
for the eld amplitudes.
18.1

E
i
normal to plane of incidence
Suppose that the electric eld

E
i
of an EM wave acting on a boundary is
normal to the plane of incidence, as shown in Figure 65.
Figure 65: Incident beam with the elec-
tric eld normal to the plane of incidence.
In this case, the incident electric
eld

E
i
is purely tangential to the
boundary. Since the materials are
isotropic, the induced elds

E
r
and

E
t
will also be tangential to the the
boundary. Thus the condition
n
E
i
+

E
r
= n

E
t
gives
E
0
+

E
1
=

E
2
. (95)
Note then that
n

E
0
= n

E
1
= n

E
2
= 0 .
If we want to express E
1
and E
2
in terms of E
0
, we need an another equation
for

E
0
,

E
1
and

E
2
, which comes from the continuity condition of

H through
the relation
n
k
1
1
n
i

E
i
+
k
1
1
n
r

E
r
= n
k
2
2
n
t

E
t
.
255
Since the phase factors in

E
exp i(t n
r k) are the same, we obtain

k
1
n
i

E
0
+ n
n
r

E
1
=
k
2
2
n
n
t

E
2
.
Using a vector identity

A (
B

C) = (
A

C)
B (
A

B)
C, and the fact

that n

E
0
= n

E
1
= n

E
2
= 0, we obtain
k
1
n n
i
E
0
+ n n
r
E
1
=
k
2
2
n n
t
E
2
.
However
n n
i
= cos(
i
) = cos
i
,
n n
r
= cos
r
= cos
i
,
n n
t
= cos(
t
) = cos
t
,
and then, we obtain
k
1
E
0
cos
i

E
1
cos
i
=
k
2
E
2
cos
t
. (96)
Since

Es are all in the same direction, we might as well drop the vector
signs. Thus, Eqs. (95) and (96) take the form
E
0
+ E
1
= E
2
, (97)
E
0
cos
i
E
1
cos
i
=
k
2
1
k
1
2
E
2
cos
t
. (98)
Eliminating E
1
using Eq. (97), that E
1
= E
2
E
0
, we get
E
0
cos
i
(E
2
E
0
) cos
i
=
k
2
1
k
1
2
E
2
cos
t
,
2E
0
cos
i
=
cos
i
+
k
2
1
k
1
2
cos
t
E
2
,
or
2k
1
2
E
0
cos
i
= (k
1
2
cos
i
+ k
2
1
cos
t
) E
2
.
256
Thus
E
2
=
2k
1
2
cos
i
k
1
2
cos
i
+ k
2
1
cos
t
E
0
. (99)
Using the Snells law
k
1
sin
i
= k
2
sin
t
,
we can eliminate
t
. This will allow us to predict the amplitude of the
reected wave from only knowing materials properties (k
1
, k
2
,
1
,
2
) and the
angle of incidence. Thus, from the Snells law, we have
k
2
cos
t
=
k
2
2
k
2
1
sin
2
i
and then substituting to Eq. (99), we get
E
2
=
2k
1
2
cos
i
k
1
2
cos
i
+
1
k
2
2
k
2
1
sin
2
i
E
0
. (100)
Similarly, eliminating E
2
= E
0
+ E
1
from Eq. (98) above, we obtain
E
1
=
k
1
2
cos
i
k
2
2
k
2
1
sin
2
i
k
1
2
cos
i
+
1
k
2
2
k
2
1
sin
2
i
E
0
. (101)
Equations (100) and (101) are called Fresnel equations for the electric eld
amplitudes.
The corresponding

H elds are not parallel to each other, but their relative
magnitudes can be deducted from equations of the form
H =
k
n

E i.e. H =
kE
.
18.2

E
i
in the plane of incidence
Suppose now that

E
i
is in the plane of incidence. In this case

H
i
is tangential
to the boundary plane and then

H
r
and

H
t
are tangential too. Thus
n
H
i
+

H
r
= n

H
t
257
becomes
H
0
+

H
1
=

H
2
.
The continuity of tangential

E is given by
n
E
0
+

E
1
= n

E
2
.
We have two simultaneous equations. In contrast to the previous case, it is
now simpler to work in terms of

H.
Thus, we express

E in terms of

H
E =
k
n

B =
k
n

H .
Hence
n
1
k
1
n
i

H
0
+

1
k
1
n
r

H
1
= n
2
k
2
n
t

H
2
.
Continuing with a procedure similar to the case of

E
i
normal to plane of
incidence, we obtain
H
1
=
k
2
2
2
cos
i
1
k
1
k
2
2
k
2
1
sin
2
i
k
2
2
1
cos
i
+
2
k
1
k
2
2
k
2
1
sin
2
i
H
0
, (102)
H
2
=
2k
2
2
2
cos
i
k
2
2
1
cos
i
+
2
k
1
k
2
2
k
2
1
sin
2
i
H
0
. (103)
Equations (102) and (103) are called Fresnel equations for the magnetic
eld amplitudes.
This time the

Es are not parallel, but their relative amplitudes can be de-
ducted from the relation E = (/k)H.
18.3 Fresnel Equations for dielectric media
Consider now two specic examples of propagation of an EM wave between
two dielectric materials. We shall show that in this case, the Fresnel equa-
tions can be simplied to forms containing only geometrical factors.
258
In a dielectric: conductivities
1
=
2
= 0,
1
=
2
=
0
, k = 2/ =
real, and
v
p
=

k
=
1
0
.
Since k 1/ 1/v
p

, the Snells law (k

1
sin
i
= k
2
sin
t
) becomes
1
sin
i
=
2
sin
t
,
and then
sin
i
sin
t
=
1
=
v
1
v
2
= n
12
.
Consider two examples:
Example 1: In the rst example we assume that E is normal to the plane
of incidence. From Eq. (100), we have that in general
E
2
E
0
=
2k
1
2
cos
i
k
1
2
cos
i
+ k
2
1
cos
t
.
Since for a dielectric
1
=
2
=
0
and
k
2
k
1
=
sin
i
sin
t
,
we obtain
E
2
E
0
=
2 cos
i
cos
i
+
k
2
k
1
cos
t
=
2 cos
i
cos
i
+
sin
i
sin t
cos
t
=
2 cos
i
sin
t
cos
i
sin
t
+ sin
i
cos
t
.
Hence
E
2
E
0
=
2 cos
i
sin
t
sin (
t
+
i
)
.
259
Similarly, we can readily show that
E
1
E
0
=
sin(
t
i
)
sin (
t
+
i
)
. (104)
Example 2: Consider now the case of E in the plane of incidence. Following
the same procedure as above, we can show that
E
2
E
0
=
2 cos
i
cos
t
sin (
t
+
i
) cos (
i
t
)
,
and similarly
E
1
E
0
=
tan(
i
t
)
tan (
t
+
i
)
.
It is seen from the above examples that the reection and refraction ampli-
tudes are dierent for linearly polarized waves with the electric eld vector
oscillation in the direction normal to the plane of incidence and for waves
with the electric eld vector oscillating in the plane of incidence. However,
the above examples show an interesting feature that for normal incidence,
(
i
= 0), the reected and refracted amplitudes are independent of the po-
larization.
Revision questions
Question 1. Explain in general, what are the Fresnels equations?
Question 2. Under which conditions the ratios of the reected to the incident
and transmitted to the incident beams amplitudes depend solely on
the angles involved?
Question 3. Does the amplitude of the reected beam depend on polarization of
the incident beam?
260
Tutorial problems
Problem 18.1 Reected and transmitted elds at a boundary between two non-
conducting materials
Consider a propagation of an electromagnetic wave between two
dierent non-conducting materials. Prove the following:
(a) For

E
i
in the plane of incidence
E
1
E
0
=
sin 2
t
sin 2
i
sin 2
t
+ sin 2
i
and
E
2
E
0
=
4 sin
t
cos
i
sin 2
t
+ sin 2
i
.
(b) For

E
i
normal to the plane of incidence
E
1
E
0
=
sin(
t
i
)
sin(
t
+
i
)
and
E
2
E
0
=
2 sin
t
cos
i
sin(
t
+
i
)
.
261
19 Applications of the Boundary Conditions
and the Fresnel Equations
In this lecture, we examine some of the consequences of the Snells law. There
are two cases possible: n
2
> n
1
and n
1
> n
2
. In the rst case, an optical
wave travels from an optically rarer to optically denser medium. In the
second case, we have the inverse situation. We will consider these two cases
separately for dielectrics and for conductors.
19.1 Applications in dielectrics
In propagation between two dielectrics there are interesting polarization
dependent eects that the two components of the EM wave (parallel and
perpendicular to the plane of incidence) are not transmitted and reected
equally.
19.1.1 Polarization by reection
Consider the case of

E in the plane of incidence, as shown in Figure 66.
Figure 66: Geometry of reection and
refraction for the incident

E eld polarized
in the plane of incidence.
From the Fresnel equations, we have
that the ratio of reected to incident
amplitude is
E
1
E
0
=
tan(
i
t
)
tan(
i
+
t
)
.
Note from the ratio, if
i
+
t
=

2
then
tan(
i
+
t
) = and consequently
E
1
= 0 .
However, according to Eq. (104),
at the same time E
1
will not be
zero for the electric eld compo-
nent normal to the plane of inci-
dence.
262
Thus, if

E
i
has arbitrary polarization then

E
r
will be plane polarized with

E
r
normal to the plane of incidence.
If
i
+
t
=

2
then
t
=

2

i
. Thus
sin
i
sin
t
= n
21
=
1
=
sin
i
sin(
2

i
)
= tan
i
.
Hence, the angle of incidence for total linear polarization of the reected
wave is
i
= arctan
1
.
It is known in the literature as the Brewsters angle.
There is an equivalent alternative proof that if

E is in the plane of incidence,
then E
1
= 0 (no reected eld polarized in the plane of incidence).
Alternatively, it can be proved using the continuity conditions for the tan-
gential components of

E and

H, from which we have
E
0
cos E
1
cos = E
2
cos
t
, (105)
and
H
0
+ H
1
= H
2
. (106)
Since
H =
k
E , (107)
and
t
=

2
, we get
E
0
E
1
= E
2
tan ,
E
0
+ E
1
= n
12
E
2
, (108)
where n
12
= k
2
/k
1
.
263
However, from the Snells law we have that
sin
sin
t
=
sin
sin
=
sin
cos
= tan = n
12
.
Thus, Eqs. (108) will be satised simultaneously only if E
1
= 0, i.e. there is
no reected eld in the plane of incidence.
We now prove that the above conclusion is not true for

E
i
polarized normal
to the plane of incidence. In this case
H
0
cos H
1
cos = H
2
cos
t
, (109)
and
E
0
+ E
1
= E
2
. (110)
From Eq. (109) we nd, after applying Eq. (107), that
E
0
E
1
=
cos
t
cos
E
2
= n
12
E
2
.
Thus, E
1
must be present, otherwise E
0
or E
2
would have two dierent values.
19.1.2 Total internal reection
We now illustrate an another interesting eect that has many practical ap-
plications: Total internal reection at a boundary between two dielectrics.
It is well known from experiments that in the case of propagation from an
optically more dense to optically less dense medium, e.g. from water in to
air, the incident beam can be completely reected at the boundary with no
transmission. Does it mean that there is no refracted beam? If this is the
case, one can easily nd from the conditions of continuity of the tangental
components of

E and

H that the refracted beam has to be present to satisfy
those two conditions.
It looks that the EM theory is in a trouble, but no panic, lets see how we
can handle this problem in terms of the EM theory.
264
Consider the Snells law
sin
t
=
sin
i
n
21
, where n
21
=
1
.
One can see that in the case of n
21
< 1, that happens when
2
<
1
, and
when the wave is going from an optically more dense to optically less dense
medium, real angles
t
are obtained only for sin
i
n
21
.
Since
t
increases with
i
, an interesting situation arises when
t
= /2,
at which angle the refracted wave glaze along the boundary. The angle of
incidence at which
t
= /2 or equivalently at which
sin
i
= n
21
is called a critical angle for propagation.
For greater
i
, sin
t
> 1, and then the angle of refraction
t
becomes imagi-
nary. In this case, there is no real refracted wave, only a reected wave.
What does it mean no real refracted wave?
The answer to this question is provided by calculating the cosine of
t
, which
in the case of the total internal reection is an imaginary number
cos
t
=
1 sin
2
t
=
1
sin
2
i
n
2
21
= i
sin
2
i
n
2
21
1 = i ,
We see that although sin
t
is still real, cos
t
becomes imaginary when
sin
t
> 1.
What are the consequences of this property?
Consider the propagation of the transmitted wave in the less optically dense
medium, as illustrated in Figure 67, for which
E
t
= E
2
e
i(t ntrk)
,
265
with the propagation distance
n
t
r = z cos
t
+ x sin
t
= iz + x .
Then
E
t
= E
2
e
i(ti
x)
= E
2
e
z
e
i(t
x)
.
Here
= k and
= k. One can see that there is attenuation of the eld

amplitude in the z direction but no phase propagation. Phase propagation
occurs in the x direction along the boundary.
Figure 67: A schematic diagram for
calculation of the propagation distance
of the transmitted wave under the total
reection.
Thus, for sin
t
> 1, an evanes-
cent wave exists along the bound-
ary (in the x direction) which is at-
tenuated exponentially in the second
medium in the normal z direction. We
may say that the eld enters into the
second material, but the wave does
not.
This illustrates a general method of ap-
plying the Fresnel equations. For only a
limited range of circumstances will all the
angles
i
,
r
,
t
be real. We can how-
ever always apply a generalized Snells
Law k
2
sin
t
= k
1
sin
i
to nd sin
t
and
cos
t
and proceed as above.
Note that the planes of constant phase
are normal to the boundary (i.e. they have their normals tangential to the
boundary). The phase of the transmitted wave below the boundary must
match the incident wave above. The wavelength in the second medium is
=
2
=
2
k
=
2
2
0
sin
i
n
21
=

0
n
21
sin
i
,
where
0
is the wavelength of freely propagating waves in this medium.
The planes of constant amplitude in the transmitted medium are parallel to
the boundary.
266
19.2 Transmission and Reection at a Conducting
Surface
Propagation of EM elds in conductors (metals) is more complicated phe-
nomenon than in dielectrics. Consider a propagation of an EM wave in a
medium composed of a dielectric and a conductor, and assume that the in-
cident wave originates in the dielectric.
From previous work, we know that in a dielectric the propagation number is
k
2
1
=
1
2
and in a conductor
k
2
2
=
2
2
i
2
=
2
1
i
.
A characteristic property of propagation in the conductor is that in the limit
of of a good conductor, /(
2
) 1, the propagation number k
2
. As
we shall see this is a major factor for the propagation in the conductor to be
completely dierent than in the dielectric.
Figure 68: Propagation of an EM
wave from a dielectric to a conductor.
Consider the Snells law which can be
written as
sin
t
=
k
1
k
2
sin
i
.
Assume that the metal is a good con-
ductor. Since in a good conductor k
2

, we see from the above equation that
in the propagation from a dielectric to
a good conductor, sin
t
0 indepen-
dent of
i
. Thus, the direction of the
transmitted wave is normal to the sur-
face independent of the angle of inci-
dence
i
.
As a consequence, the eld vectors

E and

H in the conductor lie tangential
to the boundary and so the normal components of these vectors on the con-
ductor side of the boundary are zero.
267
It follows that:
Since the normal component of

B (or

H) is continuous across the
boundary, the normal component of

H or

B is zero on the dielectric
side also.
Thus the normal component of

B of the reected wave must be equal
and opposite to that of the incident wave.
In a good conductor
k
2
2
k
2
1
= K
1
i
as

1 ,
and then we have from the Fresnel equations
E
2
0 and E
1
E
0
.
This means that the electric eld in the conductor (which is tangential to
the boundary) 0.
Since E
is continuous across the boundary, we have that E
is zero also in
the dielectric at the boundary.
Thus, the tangential component of

E of the reected wave must be equal and
opposite to that of the incident wave.
In summary, we have two useful special boundary conditions at the surface
between a dielectric and a perfect conductor:
1. The tangential component of

E = 0.
2. The normal component of

B or

H = 0.
268
19.2.1 Field vectors at normal incidence
We now consider the special case of normal incidence at a boundary, i.e. when
the wave propagation vector coincides with the normal to the boundary.
Figure 69: The eld vectors in the
plane of the surface between dielectric
and conductor.
Figure 69 shows how the eld vectors
must look in the plane of the surface be-
tween dielectric and conductor. Since we
already know that for a good conductor
for which
,
the tangential component of the trans-
mitted electric eld E
t
= 0, and the tan-
gential component of the reected mag-
netic eld H
r
= H
i
, we nd that the re-
lations between the eld components take
the form
E
t
= E
i
E
r
0 ,
H
t
= H
i
+ H
r
2H
i
.
In this case, the power reection coecient becomes
p
=
E
r
H
r
E
i
H
i
=
E
2
r
E
2
i
1 as

.
Thus, we have the total reection of the energy carried by the EM wave.
We see an advantage of using dielectric-conductor boundary to propagate
an EM wave in a material.
1. The energy of the wave is completely reected, and in fact the total
reection is independent of the angle of incidence.
2. The polarization remains constant in the propagation.
Before concluding the lecture, let us clarify a problem regarding the reec-
tion at the normal incidence. Namely, one could think that under the normal
incidence, there is only the incident and transmitted wave with no reected
269
wave. Here, we prove the necessity of assuming the existence of the reected
wave.
Assume that

E is normal to the plane of incidence. Then, from the continuity
of the tangential components at the boundary, we have
E
i
E
r
= E
t
, (111)
H
i
+ H
r
= H
t
. (112)
However
H =
0
E =
0
E = n
0
E .
Thus, Eq. (112) can be written as
n
1
E
i
+ n
1
E
r
= n
2
E
t
,
and then we get two equations for the amplitudes of the electric eld
E
i
E
r
= E
t
,
E
i
+ E
r
=
n
2
n
1
E
t
.
If E
r
is missing, we could not simultaneously satisfy both equations. Thus,
there always is a reected wave in the normal incidence.
270
Revision questions
Question 1. What is a Brewsters angle and why one could call it a polarizing
angle?
Question 2. Under what conditions will the reected and transmitted amplitudes
for perpendicular polarization be the same as those for parallel po-
larization?
Question 3. Under which conditions an incident wave will be completely reected
at a boundary between two dielectrics?
Question 4. What is meant by an evanescent wave?
Question 5. What are the boundary conditions at a surface between a dielectric
and a perfect conductor?
Question 6. In the propagation of an EM wave between a dielectric and a perfect
conductor, the direction of the transmitted wave is normal to the
boundary surface independent of the angle of incidence: True or
false?
Question 7. In the propagation between a dielectric and a perfect conductor
what happens to the electric eld at the boundary?
Question 8. Is there a reected beam at the normal incidence to a boundary
between two materials?
271
Tutorial problems
Problem 19.1 Brewsters angle for dierent polarizations
Consider a propagation of an electromagnetic wave between two
non-conducting materials.
(a) For the polarization of the

E
i
in the plane of incidence, nd
the relation between the critical angle of incidence and the Brew-
sters angle. Note, at the critical angle of incidence
t
= /2, and
at the Brewsters angle of incidence
i
+
t
= /2.
(b) Show that under the condition of no reection at a boundary
between two non-conducting materials, the sum of the Brewsters
angle and the angle of refraction is /2 for
(i)

E
i
normal to the plane of incidence,
1
=
2
=
0
and
1
=
2
.
(ii)

E
i
in the plane of incidence,
1
=
2
=
0
and
1
=
2
.
Problem 19.2 Total internal reection independent of polarization
Show, using the Fresnels equations that under the total inter-
nal reection, energy of an incident wave is completely reected
independent of polarization of the wave.
272
20 Propagation of an EM Wave in a Rectan-
gular Waveguide
In many practical problems we want to send a signal (EM wave) to dier-
ent places in an ecient way. Sending it in an uncontrolled way through
the free space in air is not a good choice. Due to scattering of the wave on
dierent objects there will be signicant losses and uncontrolled changes of
the direction of propagation. Therefore, we need some device which would
help us to send the signal in more controlled way and where it would be pro-
tected from degradation. Conducting rectangular waveguides provide this:
the waves can be propagated and delivered to the specic points (receivers)
without signicany losses.
In this lecture, we discuss the propagation of bounded EM waves by consid-
ering the propagation of radiation through a waveguide where the radiation
is fully conned in the transverse plane. We will consider the case where
the bounding walls are made of a conductor, are planar and cross section is
rectangular. Figure 70 shows an example of the rectangular waveguide.
Figure 70: A schematic diagram of a rect-
angular waveguide.
An EM wave propagating along
the z direction may undergo a
multiple reection from the walls.
Thus, we are likely to get stand-
ing waves in the x direction due
to reections between the walls
x = 0 and x = a. We also
get standing waves in the y di-
rection due to reections between
the walls y = 0 and y =
b.
The principles behind the propaga-
tion of an EM wave along the rect-
angular waveguide are analysed by solving the Maxwells equations for the
elds inside the waveguide subject to the good conductor boundary condi-
tions being satised at x = 0, x = a, y = 0 and y = b. We write the z
273
dependence of any eld component in the form
e
z
Thus

z
,
where describes the propagation conditions, e.g. purely imaginary de-
scribes a wave propagating without loss.
We describe the electromagnetic eld by the vector pair

E,

H, and we use
the Maxwells equations in the form

E =

= 0 ,

B = 0 or

H = 0 ,

E =
B
t
=

H
t
,

H =

J +
E
t
=
E +
E
t
,
where is the conductivity of the interior of the waveguide, not the bounding
metallic surfaces. We assume the interior if lled with a dielectric, but allow
a possible non-zero conductivity of the dielectric.
As we shall see, certain characteristic modes of propagation are found.
In general a characteristic mode is one which propagates with constant po-
larization.
In a rectangular waveguide, one may have TE (transverse electric) waves, or
TM (transverse magnetic) wave, but not TEM wave. If both E and B elds
are transverse, the wave would be going straight down the guide. However,
such a wave would not satisfy various boundary conditions and could not be
transmitted through the waveguide.
For a TE wave, the electric eld

E is transverse to the direction in which the
wave is propagating that is, down the waveguide. The magnetic eld

B is
perpendicular to

E and therefore there is a substantial component of

B that
points in the direction of the waveguide.
274
20.1 Transverse Electric (TE) Modes
As we have already mentioned is possible to propagate a wave with the elec-
tric eld polarized in the xy plane, a TE wave with

E transverse to the
waveguide axis, by lifting the restriction that the magnetic eld is transverse
to the direction of propagation.
To show this, we will look for the solution of the Maxwells equations with
E
z
= 0 that satises the good conductor boundary conditions.
The outline of the procedure is as follows:
First, using the Maxwells equations, we nd relations between the compo-
nents of the electric and magnetic elds. Next, we convert these relations
into a single dierential (wave) equation for H
z
, the longitudinal component
of

H. Then we nd the solution of the wave equation that satises the good
conductor boundary conditions.
We proceed as follows: Since the waveguide is of a rectangular shape, we
consider the Maxwells equations in Cartesian coordinates.
From III,

E +

H
t
= 0, we have
i

j

k
y

E
x
E
y
0
+ i
H = 0 .
Hence
x component : E
y
+ iH
x
= 0 , (113)
y component : E
x
+ iH
y
= 0 , (114)
z component :
E
y
x

E
x
y
+ iH
z
= 0 . (115)
From II,

H = 0, we have
H
x
x
+
H
y
y
+
H
z
z
= 0 ,
275
and from the fact that /z = , we get
H
x
x
+
H
y
y
H
z
= 0 . (116)
From IV,

H (
E +
E
t
) = 0, we have
i

j

k
y

H
x
H
y
H
z
( + j)
E = 0 ,
which in terms of the components gives
H
z
y
+ H
y
( + i)E
x
= 0 , (117)
H
x
H
z
x
( + i)E
y
= 0 , (118)
H
y
x

H
x
y
= 0 (E
z
= 0) . (119)
From I,

E = / = 0, we have
E
x
x
+
E
y
y
= 0 (E
z
= 0) , (120)
where we have used the TE condition E
z
= 0.
We shall now express E
x
, E
y
, H
x
and H
y
in terms of H
z
to get a wave equa-
tion for H
z
.
From Eqs. (113) and (114), we get
E
x
H
y
=
E
y
H
x
=
i
. (121)
Using this relation, we can then express E
x
and E
y
in terms of H
y
and H
x
and then show that Eq. (115) becomes identical to Eq. (116).
276
Thus, we use Eq. (121) in Eqs. (117) and (118). In Eq. (117) substitute
for E
x
:
H
z
y
+ H
y
( + i)
i
H
y
= 0 .
Hence
H
y
=
1
( + i)
i
H
z
y
=

2
i( + i)
H
z
y
=
k
2
H
z
y
, (122)
where k
2
=
2
i( + i).
In Eq. (118) substitute for E
y
and nd
H
x
=

2
i( + i)
H
z
x
=
k
2
H
z
x
, (123)
Using Eqs. (121), (122), and (123), we nd that Eqs. (119) and (120) are
automatically satised.
Substituting Eqs. (122) and (123) into Eq. (116), we obtain two equations
2
i( + i)
2
H
z
x
2
,
2
i( + i)
2
H
z
y
2
H
z
= 0 ,
2
H
z
x
2
+

2
H
z
y
2
+ k
2
H
z
= 0 , (124)
where k
2
=
2
i( + i).
We have obtained the wave equation for H
z
which we will solve assuming the
good conductor boundary conditions.
277
Now, we proceed to solve Eq. (124) for H
z
with the boundary conditions for
the eld components at the surface of a good conductor. Having H
z
, then we
can solve Eqs. (122) and (123) for H
x
and H
y
. With these solutions, we then
will be able to solve Eq. (121) for E
x
and E
y
. After that, we will know all
the eld components. In other words, we will know how the elds propagate
through the rectangular waveguide.
20.2 Boundary Conditions
Let us proceed with the steps mentioned above. First, we solve the wave
equation (124) with the known boundary conditions at the surface of a good
conductor:
1.

H normal to boundary (in xy plane) = 0.
2.

E tangential to boundary (in xy plane) = 0.
According to Figure 71: We must have at the boundaries along the x-axis,
x = 0 and x = a, the eld components H
x
= 0 and E
y
= 0. Looking at
Eq. (118), we see that this means that the derivative H
z
/x must be equal
to zero at x = 0 and x = a.
Figure 71: Electric and magnetic eld
components at the surface of the waveguide.
Similarly, at the boundaries along
the y-axis, at y = 0 and y =
b, the eld components H
y
and E
x
must vanish, H
y
= 0 and E
x
=
0. Looking at Eq. (117), we see
that this means that the derivative
H
z
/y must be equal to zero at
the boundaries y = 0 and y =
b.
Summarizing, we see that the solu-
tion of Eq. (124), which satises these
boundary conditions is of the form
H
z
= H
0
cos(k
x
x) cos(k
y
y) e
itz
,
278
with k
x
x and k
y
y satisfying the standing wave conditions:
k
x
a = m and k
y
b = n .
Hence, possible values for H
z
inside the waveguide are
H
z
= H
0
cos
mx
a
cos
ny
b
e
itz
, (125)
where m, n = 0, 1, 2 . . ..
An (m, n) combination of the integer numbers represents a possible TE mode
of propagation. The modes are designated in the form TE
mn
.
What left is to determine the coecient . With the solution (125), the wave
equation (124) gives a condition for k:
m
a
n
b
2
+ k
2
H
z
= 0 .
For a non-trivial solution, H
z
= 0, and then
k
2
=
m
a
2
+
n
b
2
.
Since that on the other hand
k
2
=
2
i( + i) ,
we nd that
2
=
m
a
2
+
n
b
2
+ i( + i) . (126)
We see that in general the propagation constant is a complex number. We
can write = + i. Then the solution (125) takes the form
H
z
= H
0
cos
mx
a
cos
ny
b
e
z
e
i(tz)
.
Thus, = 2/
g
, where
g
is the wavelength in the waveguide at frequency .
The phase velocity in the waveguide is v
p
= /.
The parameter is the attenuation coecient describing losses in the waveg-
uide. Energy loss may be due to:
279
Ohmic resistivity ( of the medium nite).
Dielectric losses (described by having an imaginary component)
Magnetic losses (described by having an imaginary component)
20.3 TE Modes in a Lossless Waveguide
Suppose that the interior of the waveguide is lled with a perfect dielectric,
i.e. we assume a lossless propagation for which we have = 0 and and
both purely real. In this case, we nd from Eq. (126) that
2
=
2
+
m
a
2
+
n
b
2
,
or introducing the phase velocity, we obtain
2
=
2
v
2
0
+
m
a
2
+
n
b
2
where v
0
= 1/
is the phase velocity of propagation of waves in an

innite (unbounded) medium of the type lling the waveguide.
It is convenient to express
2
in terms of the wavelength of the input wave.
Since
v
0
=
2f
f
0
=
2
0
,
where
0
is wavelength of the input wave, or equivalently, the innite size
medium wavelength, we can write the expression for
2
as
2
=
2
+
m
a
2
+
n
b
2
.
One can see that the parameter
2
can be negative or positive even after
all the assumptions of a lossless propagation. The nature of the propaga-
tion depends on the size of the cross section of the waveguide, and changes
according as:
280
1. Assume that
2
negative. In this case, is purely imaginary, and we
have a propagating wave
H
z
= H
0
cos
mx
a
cos
ny
b
e
i(tkgz)
,
where we have put = ik
g
, with k
g
the guide propagation constant.
The wave then propagates with a guide wavelength
g
given by
2
g
= k
g
=
m
a
n
b
2
.
There is a maximum wavelength
0
(=
c
say) such that k
g
is real
1
2
c
=
m
2a
2
+
n
2b
2
.
Thus, there is a minimum frequency f
mn
such that the TE
mn
mode will
propagate down the waveguide
f
mn
= v
m
m
2a
2
+
n
2b
2
.
This can be derived from the cut-o condition f
mn
c
= v
m
.
Thus, the waveguide acts as a high-pass lter for any (m, n) mode.
2. Consider
2
positive. Then = say is purely real and
H
z
= H
0
cos
mx
a
cos
ny
b
e
z
e
it
.
There is no phase propagation but amplitude attenuation. Note there
is no energy loss mechanism available. This is an evanescent mode
at frequencies less that f
mn
analogous to the case of total internal
reection.
Field components in the TE modes.
Since
H
z
cos
mx
a
cos
ny
b
e
itz
,
281
we nd from Eq. (122) the H
y
component of the eld
H
y

H
z
y
cos
mx
a
sin
ny
b
e
itz
.
Then from Eq. (123), we nd the H
x
component
H
x

H
z
x
sin
mx
a
cos
ny
b
e
itz
.
Next, from Eq. (121), we nd the components of the electric eld
E
x
H
y
cos
mx
a
sin
ny
b
e
itz
,
and
E
y
H
x
sin
mx
a
cos
ny
b
e
itz
.
We see from the above equations that the mode m = n = 0 is not allowed to
propagate through the waveguide, since E
x
, E
y
, H
x
and H
y
all contain sine
terms so all the eld components vanish for m = n = 0. All other TE
mn
modes are allowed.
22
TM (transverse magnetic) modes
Put H
z
= 0 and go through the whole procedure again. E
z
= 0 now. Equa-
tions analogous to Eqs. (122), (123), and (124) appear for components of
the

E vector this time. Consequently the previous discussion about modes
and their cut-o frequencies for TE modes is also true for TM modes. The
only dierence is that more modes are not allowed, i.e. in contrast to the TE
propagation, TM modes with either m = 0 or n = 0 are not allowed.
23
22
The mode m = n = 0 is never possible in a transmission line consisting of a single
closed conductor like a rectangular waveguide. It is possible in 2-conductor lines e.g. the
coaxial line or the twin wire transmission line.
23
This is an interesting dierence between properties the TE and TM modes, and the
student is encourage to analyze, as a tutorial problem, where the dierence is coming from.
282
Special properties of the TE
10
mode
We now proceed to discuss interesting properties of some of the propagating
modes. For example, If the transverse dimensions of the rectangular waveg-
uide are dierent (a = b) there is a nite range of frequencies over which the
TE
10
mode is the only allowed mode.
Figure 72: Electric and magnetic elds
of the TE
10
mode.
This means that a waveguide can be
designed to allow propagation in one
mode only. In other words, it can work
as a frequency lter transmitting non-
degenerate modes. To show this, con-
sider the frequency of the TE
mn
mode
f
mn
= v
m
m
2a
2
+
n
2b
2
,
from which we nd that frequencies
of two neighbouring modes TE
10
and
TE
01
are
f
10
=
v
m
2a
< f
01
=
v
m
2b
,
where we adopt the convention that a > b. Thus, in the frequency range
from f
10
f
01
, the TE
10
mode is the only mode allowed.
20.4 Phase and Group Velocities of Mode Propagation
Consider the velocity with which the wave propagates inside the waveguide.
We look into the phase velocity, the velocity the wave front propagates
v
p
= f
g
=

k
g
=

m
a
n
b
2
.
Thus
v
p
=
v
0
m
0
2a
n
0
2b
2
,
283
where v
0
is the phase velocity in the unbounded medium.
We see that the phase velocity of the wave inside the waveguide is greater
than in an unbounded medium, and so may be greater than the speed of
light in vacuum.
Alternately, we could write
m
a
2
+
n
b
2
=
2
= k
2
c
,
where
c
is the innite medium wavelength at the cut-o frequency for
the m, n mode. Then
v
p
=

k
2
0
k
2
c
=

k
0
kc
k
0
2
=
v
0
2
, (127)
provided >
c
i.e. in the pass-band for that mode.
Note that for
c
, the phase velocity v
p
. Moreover, if a vacuum or
air lls the waveguide then v
0
= c and v
p
> c.
The group velocity
Since v
p
> c typically, we see that the phase velocity v
p
is not the velocity
of propagation of energy or information down the waveguide. It is prop-
agated with the group velocity which is smaller than the phase velocity if
the medium is dispersive. When dispersion is not present, phase and group
velocities are equal. According to Eq. (127), the waveguide is a dispersive
medium, since the phase velocity depends on frequency, so the energy or
information is propagated with the group velocity v
g
= d/dk
g
that diers
from the phase velocity v
p
= /k
g
.
Just a brief explanation how do we dene the group velocity. The group
velocity is the velocity of propagation of some modulation of multi-frequency
wave that carries information. A single frequency harmonic wave carries no
information. It is just there. A nite bandwidth is required to carry infor-
mation.
284
We can illustrate the concept of group velocity by considering a sum of two
cosine waves of slightly dierent frequencies, and + d:
cos(t kz) + cos[( + d)t (k + dk)z]
= 2 cos
d
2
t
dk
2
z
cos
2 + d
2
t
2k + dk
2
z
= 2 cos
d
2
t
dk
2
z
cos(t kz) ,
where we have assumed that d <2 and dk <2k.
The velocity of the amplitude modulation is
v
g
=
d
2
dk
2
=
d
dk
.
In the rectangular waveguide
k
g
=
2
g
=
k
2
0
k
2
c
=
2
c
v
m
,
where k
0
= /v
m
and k
c
=
c
/v
m
.
Remember, v
m
is the innite medium phase velocity and
c
is the cut-o
(angular) frequency of the (m, n) mode. Then dierentiating
dk
g
d
=
1
v
m
1
2
(
2
2
c
)
1
2
2 =

v
m
2
c
.
Finally
v
g
=
d
dk
g
=
v
m
2
c
= v
m
2
< v
m
.
Thus, the group velocity of a wave propagating inside the waveguide is smaller
than the phase velocity. Moreover
v
g
v
p
= v
m
2
v
m
2
= v
2
m
.
285
In a vacuum-lled waveguide v
g
v
p
= c
2
. Thus, relativity is still all right.
Revision questions
Question 1. Explain what is meant by TE and TM waves.
Question 2. Explain briey the procedure of nding the components of the elec-
tric and magnetic elds transmitted through a waveguide.
Question 3. Why a TEM wave cannot be transmitted through a rectangular
waveguide?
Question 4. Explain the notation, what does it mean TE
mn
?
Question 5. Explain, how a waveguide can work as a frequency lter.
Question 6. Dene the phase and group velocities and the relation between them.
Question 7. In a vacuum waveguide the phase velocity of an EM wave is greater
than speed of light: True or false?
286
21 Relativistic Transformation of the Elec-
tromagnetic Field
The nal part of the course is devoted to relativistic eects in the EM theory.
Using the argument that the Maxwells equations, as physical observables,
are invariant under the Lorentz transformation, we shall show how the elds
transform according to the relativistic rules. Then, we will illustrate on few
examples, how the elds change when one goes between dierent inertial
frames and how this aects propagation of an EM wave. Finally, we show
how frequency and energy transform according to the transformation rules
and point out the evidence of the dependence of the energy on frequency.
21.1 The Principle of Relativity
We begin from the recollection of the principles of relativity.
1. The laws of physics are the same in all inertial reference frames.
2. The speed of light in vacuum is independent of the uniform motion of
the observer or source.
The constancy of the velocity of light, independent of the motion of the
source, gives rise to the relations between space and time coordinates in dif-
ferent inertial reference frames known as Lorentz transformations.
Consider a stationary reference frame S and a inertial frame S
moving with a
velocity u parallel to the x axis, i.e. u = u
i. The time and space coordinates

in S
are related to those in S by the Lorentz transformations

x
= (x ct) ,
y
= y ,
z
= z ,
ct
= (ct x) ,
where = (1
2
)
1/2
is the Lorentz factor, and = u/c.
287
The above transformation corresponds to a situation of u parallel to the x
axis. If the axis in S and S
remain parallel, but the velocity u of the frame S
is in an arbitrary direction, the generalization of the above transformations is

r
= r + ( 1)
(r
ct ,
ct
ct
,
where

= u/c.
Proof:
Decompose the vector r into two components: parallel and normal to u
r = r
+r
.
Then, using the one dimensional Lorentz transformations, we have
r
ct
,
r
= r
.
However, we can write the parallel and normal components as
r
=
(r
2
r
= r r
.
Hence
r
= r
+r
ct
+r r
= r + ( 1)
(r
ct .
The transformation of time can be proved in the similar way. Since
ct
ct r
,
and
r
=

r
=
(r
= r
,
288
we obtain
ct
ct
,
as required.
We will need the inverse Lorentz transformations, which are
r = r
+ ( 1)
(r
2
+
ct
,
ct =
ct
+

r
.
Note that r is a function of r
and t
.
The principle of relativity indicates, and we have showed it explicitly in
Chap. 7, that the Maxwell equations are invariant under the Lorentz
transformation. The same rule applies to the continuity equation.
Thus, the Maxwells equations together with the continuity equation have
the same form in two dierent inertial frames.
If in the frame S:

D = ,

B = 0 ,

E =
B
t
,

H =

J +

D
t
,

J =
t
,
then in the frame S
= 0 ,
=

J
,
where the prime variables are functions of the transformed variables, t
and r
.
289
Of course, the EM elds together with the current and charge densities in
the S
frame must be dierent from that in the S frame in order to match the
same Maxwell equations. From this, an interesting question arises: What are
the relations between the EM elds, charge and current densities in the S
and S frames?
In order to answer this question, we have to nd the transformation of a
time derivative of an arbitrary scalar function , and the transformation of
divergence

F.
Consider the transformation of a time derivative /(ct)
(ct)
=

(ct
)
ct
(ct)
+

x
(ct)
+

y
(ct)
+

z
(ct)
=

(ct
)
ct
(ct)
+
(ct)
.
However
ct
(ct)
= ,
r
(ct)
=
,
which gives
(ct)
=

(ct
.
Consider now the divergence

F =
F
x
x
+
F
y
y
+
F
z
z
.
Since
F
x
x
=
F
x
(ct
)
ct
x
+
F
x
x
x
+
F
x
y
x
+
F
x
z
x
,
and
ct
x
=
x
,
290
x
x
= 1 + ( 1)
2
x
2
,
y
x
=
z
x
= 0 ,
we obtain
F
x
x
=
x
F
x
(ct
)
+
x
F
x
x
,
where
x
= 1 + ( 1)
2
x
/
2
.
Similarly, for F
y
and F
z
, and nally we get

F = (

F)
F
(ct
)
,
where is a 3 3 diagonal matrix
=
1 + ( 1)
2
x
2
0 0
0 1 + ( 1)
2
y
2
0
0 0 1 + ( 1)
2
z
,
as required.
Using the above transformations, we can derive transformations for the cur-
rent density

J and the charge density .
In order to do it, we consider the continuity equation, that can be written as

J =
c
(ct)
.
Hence
(

J)
J
(ct
)
=

(ct
c ,
or
(

J)
(c) =

(ct
.
291
Since
(c) =
(c
) ,
we obtain

J c
=

(ct
.
Thus, the continuity equation will be invariant under the Lorentz transfor-
mation if
c
=

J c
. (128)
In order to understand the physical meaning of these equations, consider the
following example.
Example
Assume that in the S frame there is a stationary volume charge of density
= 0. Since is stationary, there are no currents in the S frame (
J = 0).
What are the charge and current densities as seen in the S
frame?
In the S frame
J = 0 , = 0 .
According to Eq. (128), in the S
frame
= c
= .
Thus, there is a current in the S
frame. As seen from S
a given part of the

charge is length contracted in the direction of motion so the charge density
is correspondingly increased by the factor > 1. The length contracted
charge density appears from S
to move in the opposite direction. We can

understand this result: The stationary charge in the S frame moves with
velocity u in the S
frame. This eect is predicted by the Galileo transfor-

mation.
292
Less obvious and more interesting is the following situation. Lets in the S
frame one observes
J = 0 , = 0 .
Then, according to Eq. (128), someone will see a non-zero charge density
= 0 in the S
frame.
This is a pure relativistic eect, which cannot be predicted by the Galileo
transformation. Note, the charge
is proportional to u/c
2
.
21.2 Transformation of Electric and Magnetic Field
Components
To nd the transformation rules for electric and magnetic eld components
we will use the transformations of the time and space derivatives derived
above.
Consider two of the Maxwell equations that in the S frame are

D = ,

H =

J +

D
t
.
These equations should be equivalent of two equations
=

J
,
in the S
frame.
Using the transformations of the time and space derivatives, we have

(ct

(ct
D =

J ,

(ct
D = c .
293
Substituting the transformations of

J and , we nd that the

D and

H
vectors transform as
c

1
c
D +

H
D +
1
, (129)
where
1
is the inverse of the matrix

1
=
1 + (
1
1)
2
x
2
0 0
0 1 + (
1
1)
2
y
2
0
0 0 1 + (
1
1)
2
z
.
From the Maxwell equations

B = 0 ,

E =
B
t
,
we nd that the

E and

B vectors transform as

E +

c

E +
1
c
. (130)
21.3 Transformation Rules in Terms of Parallel and
Normal Components
Suppose that the frame S
is moving with speed u in the direction paral-

lel to the z axis. In this case,
x
=
y
= 0,
z
= = 0, and then the
transformations take the form
c
= cD
x
i + cD
y
j + cD
z
k +
k

H ,
= c
k

D + H
x
i + H
y
j + H
z
k ,
= E
x
i + E
y
j + E
z
k + c
k

B ,
c
k

E + cB
x
i + cB
y
j + cB
z
k . (131)
294
It is useful to rephrase the transformation rules in terms of components
parallel and normal to u. The parallel components are the z components
and the normal components lie in the xy plane. For example
E =

E
+

E
E
x
i + E
y
+ E
z
k ,
and the same for

D,

H and

B.
21.3.1 Rules for Parallel Components
From Eq. (131), we readily nd that
=

E
k =

E
z
k =

E
,
and similarly
=

B
,

H
=

H
,

D
=

D
.
Thus, the components parallel to the direction of u are invariant under the
transformations (129) and (130).
21.3.2 Rules for Normal Components
Consider now the components perpendicular to u:
c
= cD
i + cD
j
= cD
x
i + cD
y
j + H
x
j H
y
i
= (cD
x
H
y
)
i + (cD
y
+ H
x
)
j ,
= H
i + H
j
= (H
x
+ cD
y
)
i + (H
y
cD
x
)
j ,
= E
i + E
j
= (E
x
cB
y
)
i + (E
y
+ cB
x
)
j ,
295
c
= cB
i + cB
j
= (cB
x
+ E
y
)
i + (cB
y
E
x
)
j ,
c
+

H
+

c
. (132)
We shall illustrate the transformation rules on the following examples:
Example 1 - purely electric eld in S
Suppose that in S, one observes

E = 0 but

B = 0.
Then from the transformation rules, in S
=

E
,

E
= 0 ,

B
c
2
u

E
.
Thus
=

B
=
u

E
c
2
=
u

E
c
2
,
since u

E
= 0.
Thus what appears to be purely an electric eld to one observer is seen as
both an electric and a magnetic eld to a second observer moving with re-
spect to the rst.
Example 2 - purely magnetic eld in S
296
Now suppose that in S, one observes

E = 0, but

B = 0.
Then using the transformation rules, in S
=

B
,

B
= 0 ,

E
= u

B
.
Thus
=

E
= u

B
= u

B .
We see that what appears to be a purely magnetic eld for one observer will
appear to be both an electric and a magnetic eld to a relatively moving ob-
server.
This result could be used to calculate the emf in an electric dynamo from
the point of view of an observer watching the conductor move in a magnetic
eld or from the point of view of an observer moving with the conductor.
21.4 Transformation of the Components
of a Plane EM Wave
In this nal lecture of the course, we illustrate the transformation rules of
the components of a plane EM wave. We illustrate this on two examples.
Example 1
Suppose that a plane wave propagates in vacuum along the z axis. Then the
electric and magnetic elds of the wave are
E =

iEe
i(tkz)
=
iE
0
=
iE
x
,
B =

jBe
i(tkz)
=

jB
0
=

jB
y
.
Hence from the transformation rules (132), in the S
frame moving in the

same direction:
E
0
i cB
0
= (E
0
uB
0
)
i ,
c
cB
0
j E
0
= (cB
0
E
0
)
j .
297
Since in vacuum
cB
0
= E
0
,
we obtain
1
u
c
E
0
i =
1
u
c
u
c
2
E
0
i =
c u
c + u
E
0
i ,
and
c
1
u
c
cB
0
j =
c u
c + u
cB
0
j .
Thus, we observe that the amplitudes of the elds are reduced, but the ratio
[
[/[
[ = [E
0
[/[B
0
[ is constant and independent of u, as it should be, since
speed of light is the same in all inertial frames.
This is consistent with the principle of relativity that speed of light is inde-
pendent of the motion of the observer.
Example 2
Suppose that an observer S
moves in the direction of the electric eld of

an EM wave, i.e. u = u
i, as shown in Figure 73.

Figure 73: Text
In this case
= E
0
i + cB
0
k =
i +
u
c
E
0
,
and
c
= cB
0
j .
Thus, the magnetic eld remains un-
changed, but the electric eld turns to-
wards the direction of propagation of the
wave.
298
An interesting situation occurs when the
velocity u c. In this case, we have
u
c
=
1
u
c
2
u
c
,
and then
= E
k ,
i.e. the direction of

E
becomes perpendicular to u.
However, the Poynting vector of the wave is still in the direction of u:

H =
u
c
E
0
B
0
j =
2
u
c
E
0
B
0
i .
21.5 Doppler Eect
Consider a plane wave propagating in vacuum
E(t) =

E
0
e
i(t
kr)
.
In the moving frame S
this wave will have a dierent frequency
and the
wave vector

k
, but the phase of the wave will remain unchanged as it is

invariant under the transformation, i.e.
= t
k r =
.
Using the inverse Lorentz transformation
r = r
ct
,
ct =
ct
+

r
,
we will nd the relations between
, and k
, k that ensure the invariance

of the phase of the wave.
299
Thus, with the above transformations, we nd
= t
k r
=

c
ct
+

r
k r
ct
= t
k u +
k u
.
Hence
k u
k

c
2
u .
Consider two special cases. In the rst case, assume that the wave propagates
in vacuum,

k = (/c)
k. Then
1
u
c
cos
,
where is the angle between the direction of propagation of the wave and
the direction of the motion u.
Figure 74: Text
For the propagation direction = 0
c u
c + u
.
The frequency in S
is smaller than that

in S, and
0 as u c. On the
other hand, the frequency
when
u c.
If the wave propagates in a material body,
where v
f
< c, we have
k = (/v
f
)
k ,
and then the frequency in S
is
1
u
v
f
cos
.
300
For the case of = 0
=
1
u
v
f
u
c
2
.
When u = v
f
, the frequency
= 0, but for c > u > v

f
, we obtain that the
frequency
< 0.
21.6 Transformation of Energy of a Plane EM Wave
Consider an EM wave propagating in the
k direction and an observer moving

in the z direction, as shown in the Figure 75.
Let

E = E

i. Then, in the frame S
E +
E + cos

c
B .
Since

B = B
j + B

k (the wave propagates in the plane yz), and cB = E,
we obtain
E cos cB

i
= (1 cos )

E .
Figure 75: An EM wave is propagat-
ing in the

k direction and an observer
is moving in the z direction.
Consider now energy of an electric eld
of a plane EM wave conned in a vol-
ume V . In the S frame
W
e
=
1
4
0
E
2
0
V .
In the S
frame
W
e
=
1
4
0
(E
0
)
2
V
.
However
V
= x
.
301
Since, we have assumed that the observer (S
frame) moves in the direction

of the z axis, and the wave propagates in the direction

k u = ucos , we
obtain
V
= xyz
,
where
z
=
z
1
u
c
cos
.
Hence
W
e
=
1
4
0
E
2
0
1
u
c
cos
2
V
1
u
c
cos
= W
e
1
u
c
cos
.
It is interesting to compare the transformation of energy with the transfor-
mation of frequency. Since
W
e
= W
e
1
u
c
cos
,
and
1
u
c
cos
,
we see that the energy and frequency transform in the similar way!
This indicates that W
e
. Note, that a similar proportionality was pre-
dicted in quantum physics, W
e
= h.
In the electromagnetic theory, the proportionality forms backgrounds of the
so called quantum electrodynamics.
302
Revision questions
Question 1. Are the Maxwells equations invariant under the Lorentz transfor-
mation? Explain.
Question 2. What are the rules for transformation of the electric and magnetic
eld components?
Question 3. What are the rules for transformation of the components of an EM
wave?
Question 4. What is Doppler eect?
Question 5. How does the energy of an EM wave transform in the relativistic
case?
303
Tutorial problems
Problem 21.1 A relativistic linear dynamo
Figure 76:
A linear dynamo is constructed using a hor-
izontal U-shaped or rectangular section of
wire with one side moving at speed v =
dx
dt
in a vertical magnetic eld

B as shown in
Figure 76.
The rectangle has width . The dynamo is
relativistic in that v may be comparable to c.
The general denition of emf round a cir-
cuit is:
c = W =

F
q
d
E +v

B) d
.
(a) Calculate the emf induced along the moving arm:
(i) In the frame of reference in which the U-shaped section is
at rest.
(ii) In the frame of reference in which the slide-wire is at rest.
Do you arrive at the same value of emf in each frame? If not,
does that bother you?
(b) In practice, emf s are usually calculated using Faradays Flux
Cutting Rule because it is the simplest way mathematically. The
ux cutting rule c =
d
dt
may be thought of as an integral of the
Maxwell equation

E =

B
t
. Now Maxwells equations are
supposed be relativistically correct, i.e. they take the same form
in all inertial frames of reference. Show that the ux cutting rule
holds in both reference frames mentioned above.
Problem 21.2 Plane wave propagating in a material
Suppose that a plane wave propagates in a material with =
0
and =
0
. Show that in the frame S
, moving with velocity u

304
in the direction of propagation of the wave (z-axis), the ratio
[
[/[
[ depends on the velocity u.

Problem 21.3 Electric and magnetic elds in a moving frame
Suppose that in a stationary frame S one has measured an uniform
electric eld

E = E
0
i and an uniform magnetic eld

B = B
0
j,
such that cB
0
> E
0
.
Is it possible to nd a frame S
moving along the z-axis in which

one could not measure either

E
or

B
?
If yes, nd the amplitude of the measured eld and the veloc-
ity u of the frame S
.
305
Revision Questions for the Final Examination
Question 4.1 Prove that the total electric ux through a closed surface S is
proportional to the total charge inside the surface.
Question 4.2 Prove the Amperes circuit law.
Question 4.3 Derive the integral form of the Faradays law and then transform
it into the dierential form

E =
B
t
.
Question 5.1 Starting from the Maxwells equations derive the continuity equa-
tion, i.e. show that conservation of charge is built into the Maxwells
equations.
Question 6.1 Using the Maxwells equations show that

B satises the same
wave equation as

E.
Question 6.2 Show, using the proof of solution of the wave equation, that
f(z +ct) represents a signal propagating in the negative z direc-
tion with speed c.
Question 7.1 Show that magnetic and electric elds of a charge moving with a
constant velocity v are related by
B =
v

E
c
2
.
Question 7.2 Show that the magnetic eld lines produced by a charge moving
with a constant velocity v form concentric rings about v.
Question 7.3 Show that the electric eld produced by a moving charge satises
the Gausss law.
Question 7.4 Explain the statement: Electric and magnetic elds do not pro-
duce each other - they are both due to electric charges.
306
Question 8.1 Using the Maxwells equations derive the continuity equation for
the Poynting vector.
Question 8.2 Show, using the eld theory calculation, that the power dissipated
along a resistive wire is P = V I, the same predicted by the circuit
theory.
Question 9.1 (a) Obtain a series solution of the Laplace equation in the cartesian
coordinates for the electrostatic potential inside a closed three di-
mensional area.
(b) What would be the solution of the Laplace equation for a
closed two dimensional problem.
Question 9.2 Consider a conducting sphere of radius R. The surface of the
sphere is kept at a potential
(R, , ) = V
0
sin sin .
Using the above as a boundary condition, nd the potential at
any point inside the sphere.
Question 10.1 Show that in general the Maxwells equations cannot be simpli-
ed to two separate dierential equations for the

E and

B elds.
Then show that by introducing the concept of vector and scalar
potentials one can arrive to dierential equations for

A and
that can be separated from each other under the Lorenz gauge.
Question 10.2 Explain, why the Coulomb gauge is often called Transverse
gauge.
Question 10.3 Prove that the homogeneous wave equation
2

1
c
2
t
2
= 0
has the solution of the form
(r, t) =
f(t r/c)
r
,
307
where f(t r/c) is an arbitrary function of the retarded time
t r/c.
Question 11.1 Show that in spherical polar coordinates, the magnetic eld of a
short current element I

l =

lI
0
exp(it) has only an azimuthal
component of the form
B =
I
0
l
4
0
c
2
ik
r
+
1
r
2
sin e
i(tkr)

.
Question 11.2 Show that in the far eld zone of a radiating short current ele-
ment, the electric and magnetic elds oscillate in phase and are
orthogonal to each other.
Question 11.3 (a) Given the expressions for the EM eld of a Hertzian dipole,
show that the total radiated power from the dipole is
W =
I
2
0
3
0
c
2
.
(b) Show that the emitted power is equivalent to the power lost
on a resistor of resistance
R = 80
2
2
.
Question 11.4 Show that the time averaged Poynting vector of the EM eld
emitted by a short current element is maximal in the equatorial
plane of the element.
Question 12.1 Show that the electrostatic potential due to a distribution of elec-
tric dipoles of moment per unit volume

P throughout a volume V
enclosed by a surface S is that of a volume charge density
P
together with a surface charge density

P n.
Question 12.2 Illustrate the Lorentz theory of polarizability of dense dielec-
tric materials and derive the Caussius-Mossotti relationship for a
dense dielectric.
Question 12.3 Show that the polarization of a dielectric driven by a time varying
electric eld lags in phase the driving eld.
308
Then show that the phase dierence between the polarization and
a time-varying electric eld results in a complex permittivity of
the dielectric.
Question 13.1 Explain the concept and advantage of introducing the magnetic
intensity vector

H.
Question 13.2 Show that inside a ferromagnet H = 0 and explain the physical
meaning of this result.
Question 13.3 Derive the Maxwells equations for the EM elds in electric and
magnetic materials.
Question 14.1 (a) Derive the special form of Poyntings Theorem applicable in
certain material media
S
(
E

H) d
S =
B
t
dV
D
t
dV

E

J dV .
(b) Interpret the above equation in terms of energy storage and
energy ow etc.
(c) State qualitative meaning of each term in the equation.
Question 14.2 Prove that the useful formula for the mean Poynting vector for
sinusoidal elds is
N =
1
2
Re
E
c
,
where

E
c
=

E
0
exp(it) and

H
c
=

H
0
exp(it).
Question 15.1 Show that the amplitude of a plane wave propagating in a non-
conducting material is damped with the rate which arises from
the imaginary parts of the complex permittivity and permeability.
Question 15.2 Derive the expressions for the attenuation coecient and the
phase velocity of an EM wave propagating in a conducting medium.
What are the values of the quantities for a propagation inside a
309
good conductor?
Question 15.3 Show that in a good conductor an EM wave propagates on the
surface of the conductor.
Question 16.1 Prove the following boundary conditions at a bounding surface
between two dielectrics:
(a) The normal component of

B is continuous across the bound-
ary.
(b) The tangential component of

E is continuous across the bound-
ary.
(c) The tangential component of

H is continuous across the bound-
ary.
Question 17.1 Show, using the Maxwells equations that the electric and mag-
netic vectors of an EM wave are related by
H =

B
=
k
n
p

E ,
where n
p
is the unit vector in the direction of propagation of the
wave.
Question 17.2 Prove that in the reection and refraction at a bounding surface,
the direction of incident, reected and refracted waves are copla-
nar.
Question 17.3 Derive the familiar laws of elementary optics:
(a) Angle of reection equals to the angle of incidence.
(b) Snells law of refraction.
Question 17.4 Show, using the continuity conditions for

E and

H that both
reection and refraction takes place in the incidence of light on a
boundary between two dielectrics.
310
Question 19.1 Show that under the Brewsters angle of incidence there is no
reected electric eld in the plane of the incidence.
Question 20.2 Show that in a vacuum-lled rectangular waveguide v
p
> c, and
v
g
v
p
= c
2
.
Question 21.1 Find the condition under which the continuity equation for
and

J is invariant under the Lorentz transformation.
Question 21.2 Show that electric charge Q is invariant under the Lorentz trans-
formation.
311
Appendix A: Proof of the Amperes Law
Consider a long wire of radius a carrying current I. Let P is a point on the
integration path, as shown in Figure 77.
Figure 77: The source circuit and the integration path to prove the Ampere law.
The magnetic eld at P is
B =

0
I
4
s
ds ( r)
r
2
.
Moving P by d
is equivalent to moving the current circuit by d
.
The solid angle subtended by d
, ds at P is:
(d
ds) r
r
2
=
d
ds r
r
2
= d

ds r
r
2
.
(The element of area normal to r is d
s r)
Thus due to the path element d
, the change in solid angle subtended at P

by the circuit is:
d = d
s
ds ( r)
r
2
.
312
Hence
d = d
0
I
s
d
B =
4
0
I
B d
,
where integration is around the circuit s giving the magnetic eld

B at some
point P as shown in the diagram.
Now integrating round the closed path:
B d
=

0
4
I
d .
If P moves round a closed path (returning to its original position but not
circulating through the current loop:
d = 0 .
But if P circulates through the loop:
d = 4 ,
and then
B d
=

0
4
I 4 =
0
I .
We conclude that the line integral of the magnetic eld round a closed loop
path is equal to
0
I, where I is the current passing through the loop.
313
Appendix B
Proof of the vector theorem, Eq. (73):
M
R
dV =
M n
R
dS .
This is an application of the more general theorem

F dV =
F n dS .
Let

C be a constant vector. Then
(
F

C) = (

F)

C (

C)

F =

C (

F) . (133)
We will prove the general theorem by using the divergence theorem
V
(
F

C) dV =
S
(
F

C) n dS .
For an arbitrary constant vector

C, and using (133), we get

F dV =
C

F dV
=
V
(
F

C) dV =
F

C n dS .
Hence

F dV =
C

F n dS .
However
C

F n =

C

F n
and then we obtain

F dV =
F n dS .
Since this is true for arbitrary

C, we nally have

F dV =
F n dS ,
as required.
314
Appendix C: PHYS3050 Facts and Formulae
Gauss
Divergence Theorem :
F ndS =

F dV
Stokes
s Theorem :
F d

F ndS
Numerical values in SI units:
0
= 8.85 10
12
,
0
= 4 10
7
, c = 3 10
8
[ms
1
] .
For the electron: e = 1.6 10
19
[C] , m = 9.11 10
31
[kg]
The Lorentz force law :

F = q(
E +v

B)
Coulomb
s Law :

F =
1
4
0
q
1
q
2
r
2
r
Biot Savart Law : d
B =

0
4
I d
r
r
2
Gauss
Law :

E ndS =
Q
0
Amp`ere
s Circuital Law :

B d
=
0
I
Maxwells Equations in vacuum:

E =

0
,

B = 0 ,

E =
B
t
,

B =
0
J +
1
c
2
E
t
.
Maxwells Equations in material bodies:

D = ,

H = 0 ,

E =
B
t
,

H =

J +

D
t
.
315
Poynting vector:
N =
0
c
2
E

B
Poyntings Theorem:
0
c
2
(
E

B) ndS =
E

J dV

t
1
2
0
E
2
+
1
2
B
2
dV
but in polarizable materials where it is convenient to dene

D =
0
E +

P
and

H =

B

M:
E

H ndS =
B
t
dV
D
t
dV
E

J dV
A theorem on the calculation of the mean Poynting vector from complex
elds:
N =
1
2
Re
E
c
The rate of doing work in magnetization

dW
dt
= H
dB
dt
and B =
0
(H + M).
Hemholtz Theorem: An arbitrary vector

F can be written as:
F =
1
4

F
r
dV +
1
4

F
r
dV
=

F
l
+

F
t
Fields and potentials:
E =

A
t
,

B =

A
316
Dierential equation for the vector potential:
2

A
1
c
2
2

A
t
2
=
0
J +

A +
1
c
2
In the Lorentz gauge:

A =
1
c
2
t
the dierential equations for the electromagnetic potentials are:
2

1
c
2
t
2
=
2

A
1
c
2
2

A
t
2
=
0
J
and these have solutions of the form:
=
1
4
0

r
dV
A =
1
4
0
c
2

J
r
dV
The eld of a Hertzian dipole:
E
r
=
I
0
cos
4
0
c
2
ikr
3
+
2
r
2
e
i(tkr)
E
=
I
0
sin
4
0
c
1
ikr
3
+
1
r
2
+
ik
r
e
i(tkr)
B
=
I
0
sin
4
0
c
2
ik
r
+
1
r
2
e
i(tkr)
The mean energy ux from the Hertzian dipole:
N =
I
2
0
8
0
c
2
sin
2
r
2
317
A series solution in 2 dimensions to Laplaces equation in Cartesian coordi-
nates:
(x, z) =
k
[A
k
sin(x) + B
k
cos(x)][C
k
sinh(z) + D
k
cosh(z)]
A series solution in 3 dimensions to Laplaces equation in spherical polar
coordinates:
(r, , ) =
C
1
r
+ C
2
r
(+1)
m
[a
lm
cos(m) + b
lm
sin(m)] P
m
(cos )
Useful properties of trigonometrical functions:

sin( ) = sin cos sin cos
cos( ) = cos cos sin sin
sin
2
=
1
2
(1 cos 2)
cos
2
=
1
2
(1 + cos 2)

0
sin
3
d =
4
3
2
0
sin(m) sin(n) d =
0 for m = n
for m = n
2
0
cos(m) cos(n) d =
0 for m = n
for m = n
2
0
sin(m) cos(n) d = 0 for all m and n
318
Properties of Legendry polynomials:
1
1
P
m
l
(cos ) P
n
k
(cos ) d(cos ) = 0 unless m = n and l = k
1
1
[P
m
l
(cos )]
2
d(cos ) =
2
2l + 1
(l + m)!
(l m)!
P
0
= 1 , P
0
1
= cos , P
1
1
= sin , P
0
2
=
1
4
(3 cos(2) + 1) ,
P
1
2
=
3
2
sin(2) , P
2
2
=
3
2
(1 cos(2)) ,
P
l
(1) = 1 , for all l .
A theorem on the electrostatic potential due to a distribution of electric
dipoles of moment per unit volume

P:
=
1
4
0
P n
r
dS +
1
4
0

P
r
dV
A theorem on the vector potential due to a distribution of magnetic dipoles
of moment per unit volume

M:
A =
1
4
0
c
2

M
r
dV +
1
4
0
c
2
M n
r
dS
A dispersion equation:
k =
1
2
319
The skin depth in a good conductor:
=
General boundary conditions:


B is continuous across a boundary.

D is continuous across a boundary.

E is continuous across a boundary.

H is continuous across a boundary.
Special boundary conditions at the surface between a dielectric and
a perfect conductor:

E = 0.

B or

H = 0.
The Fresnel equations:
Case 1:

E normal to the plane of incidence.
Reection:
E
1
=
k
1
2
cos
i
k
2
2
k
2
1
sin
2
i
k
1
2
cos
i
+
1
k
2
2
k
2
1
sin
2
i
E
0
Transmission:
E
2
=
2k
1
2
cos
i
k
1
2
cos
i
+
1
k
2
2
k
2
1
sin
2
i
E
0
Case 2:

E in the plane of incidence.
Reection :
H
1
=
1
k
2
2
cos
i
2
k
1
k
2
2
k
2
1
sin
2
1
k
2
2
cos
i
+
2
k
1
k
2
2
k
2
1
sin
2
i
H
0
320
Transmission:
H
2
=
2
1
k
2
2
cos
i
1
k
2
2
cos
i
+
2
k
1
k
2
2
k
2
1
sin
2
i
H
0
In dielectric media the Fresnel equations become:
Case 1:

E normal to the plane of incidence:
E
1
E
0
=
sin(
i
t
)
sin(
i
+
t
)
,
E
2
E
0
=
2 cos
i
sin
t
sin(
i
+
t
)
Case 2:

E in the plane of incidence:
E
1
E
0
=
tan(
i
t
)
tan(
i
+
t
)
,
E
2
E
0
=
2 cos
i
cos
t
sin(
i
+
t
) cos(
i
t
)
Rectangular waveguides:
In the propagation direction z, the elds vary as e
z
.
For TE modes, the longitudinal component of

H satises:
2
H
z
x
2
+

2
H
z
y
2
+ k
2
H
z
= 0 ,
where
k
2
=
2
i( + i)
Then satisfying the boundary conditions (assuming the walls are perfect con-
ductors) requires:
2
=
m
a
2
+
n
b
2
+ i( + i)
For the lossless waveguide:
321
Cut-o frequency for the mn mode:
f
mn
= v
m
m
2a
2
+
n
2b
2
Phase velocity:
v
p
=
v
m
fc
f
2
Group velocity:
v
g
= v
m
f
c
f
2
VECTOR FORMULAS
( + ) = +
(
A +

B) =

A +

B
(
A +

B) =

A +

B
() = +
(
A) =

A +

A
(
A

B) =

B (

A)

A (

B)
(
A) =

A +

A
(
A

B) =

A

B

B

A + (
B )
A (
A )
B
=
2
(

A) = 0
= 0
(

A) = (

A)
2

A
A (
B

C) =

B (
C

A) =

C (
A

B)
A (
B

C) =

B(
A

C)

C(
A

B)
322
FORMS OF VECTOR OPERATIONS IN CYLINDRICAL
COORDINATES
=
+ z

A =
1
(A
+
1
+
A
z
z

A =
A
z
z

A
z
+ z
1
(A
2
=
1
+
1
2
+

2
z
2
VECTOR AND DIFFERENTIAL OPERATIONS IN
SPHERICAL COORDINATES
A = (A
x
sin cos + A
y
sin sin + A
z
cos ) r
+(A
x
cos cos + A
y
cos sin A
z
sin )

+(A
x
sin + A
y
cos )

= A
r
r + A

+ A
=

r
r +
1
r
+
1
r sin
323

A =
1
r
2
(r
2
A
r
)
r
+
1
r sin
(sin A
+
1
r sin
A

A =
r
r sin
(sin A
1
sin
A
r

(rA
)
r
(rA
)
r

A
r
2
=
1
r
2
r
2
+
1
r
2
sin
sin
+
1
r
2
sin
2
2
324

The University of Queensland Department of Physics 2008

Transféré par

Informations du document

Description originale:

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

The University of Queensland Department of Physics 2008

Transféré par

Droits d'auteur :

Formats disponibles

The University of Queensland

A = dA n is a vector representing the element dA of the surface A.

k that represents the position of an arbitrary

j is also equal to 3, but in this case the eld

, ( = 0), nd for which

S over a surface enclosing q.

E of the point charge dq is given by the Coulomb eld

E produced by the innites-

= 0. Hence, we can write

and r, and then using (10), we nd

E and the Biot-Savart Law

. The element of magnetic eld d

at a point A distance r = r r is found from the

B are in the same

is a constant vector with respect to the

means dierentiation wrt z ct. Similarly

moving with a ve-

are the coordi-

frame. The time and space

are related to those in S

frame Coulombs law holds.

to r. It can be found as follows. It is seen from

as the Coulomb force

B/t, we use the theorem of partial derivatives. If

dV Field energy. (26)

S on the ends of the cylinder), which is in agreement with the re-

S = rate of energy ow into the surface

F = 0, we can always write

B we see that the requirement

E, we see that the require-

must be equal to one at x = b/2. This happens when

consider the solution along

A. This equation is sometimes

k. We then have that in the spherical

= 0 and there is no dependence of A

. Substituting Eq. (57) into Eq. (60), we obtain

= c, the value for plane waves in free space.

, the eld points in the opposite direction to the dipole

D/t can be called a generalization of the displacement current

P/t plays the role of polarization current density.

is a complex quantity called a complex

for a given material?

, can be determined from the following experi-

) and the eective resistance is

cos (1 i tan ) = K(1 i tan ) .

R, as shown in Figure 50,

D, in order to eliminate the necessity of taking the electric dipole polariza-

is the total current per unit length, we have

H. They are called ferromagnetic mate-

H dM ? To check it, con-

, and is the complex permeability.

E`, which is a real quantity.

N, the mean Poynting

1. Thus, we can drop 1 in

E. Since the normal component of

, and then we have

r k) are the same, we obtain

C, and the fact

, the Snells law (k

= k. One can see that there is attenuation of the eld

is continuous across the boundary, we have that E

is the phase velocity of propagation of waves in an