Undergraduate Texts in Physics

Hiqmet Kamberaj

Electromagnetism
With Solved Problems
Undergraduate Texts in Physics

Series Editors
Kurt H. Becker, NYU Polytechnic School of Engineering, Brooklyn, NY, USA
Jean-Marc Di Meglio, Matière et Systèmes Complexes, Université Paris Diderot,
Bâtiment Condorcet, Paris, France
Sadri D. Hassani, Department of Physics, Loomis Laboratory, University of Illinois
at Urbana-Champaign, Urbana, IL, USA
Morten Hjorth-Jensen, Department of Physics, Blindern, University of Oslo, Oslo,
Norway
Michael Inglis, Patchogue, NY, USA
Bill Munro, NTT Basic Research Laboratories, Optical Science Laboratories,
Atsugi, Kanagawa, Japan
Susan Scott, Department of Quantum Science, Australian National University,
Acton, ACT, Australia
Martin Stutzmann, Walter Schottky Institute, Technical University of Munich,
Garching, Bayern, Germany
Undergraduate Texts in Physics (UTP) publishes authoritative texts covering topics
encountered in a physics undergraduate syllabus. Each title in the series is suitable as
an adopted text for undergraduate courses, typically containing practice problems,
worked examples, chapter summaries, and suggestions for further reading. UTP
titles should provide an exceptionally clear and concise treatment of a subject at
undergraduate level, usually based on a successful lecture course. Core and elective
subjects are considered for inclusion in UTP.
UTP books will be ideal candidates for course adoption, providing lecturers with a
firm basis for development of lecture series, and students with an essential reference
for their studies and beyond.

More information about this series at https://link.springer.com/bookseries/15593

Hiqmet Kamberaj

Electromagnetism
With Solved Problems
Hiqmet Kamberaj
Faculty of Engineering
International Balkan University
Skopje, North Macedonia

ISSN 2510-411X ISSN 2510-4128 (electronic)

Undergraduate Texts in Physics
ISBN 978-3-030-96779-6 ISBN 978-3-030-96780-2 (eBook)
https://doi.org/10.1007/978-3-030-96780-2

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To the memory of my Mother and Father
Preface

Physics is part of any curriculum in science and engineering. The main objective of
this course is to help students of engineering and other sciences in more advanced
courses in these fields. The textbook will introduce the students to the fundamental
concepts of physics and how different theories developed from physical observations
and phenomena.
It starts with electrostatics in free space, introducing basic concepts, such as
electrical charge Coulomb’s law, and ideas of electric field and electric field lines
(Chap. 1). Chapter 2 introduces the electric flux and Gauss’s law. Electrostatic poten-
tial and electrostatic potential energy are introduced in Chap. 3. Chapter 4 presents
the concepts of capacitance and dielectrics. Also, the electrostatics of a macroscopic
medium and Maxwell’s equations of the electrostatic field are discussed.
Chapter 5 introduces the concepts of electric current and Ohm’s law. Chapter 6
continues with the magnetic field and its interactions with charges and currents. Then,
Chap. 7 introduces the concept of magnetic field sources, where Biot-Savart law
and Ampère’s law are introduced. Chapter 8 describes the magnetism in medium,
Faraday’s law, and Maxwell’s magnetic field equations. Then, Chap. 9 describes
Maxwell’s equations of electromagnetic fields. In particular, this chapter focuses on
the potential vector and scalar of the electromagnetic field, electromagnetic field
energy, and conservation laws. Then, the dynamics of charged particles in the elec-
tromagnetic field and averaging of microscopic properties to obtain the macroscopic
Maxwell’s equations are introduced. Chapter 10 describes some advanced topics on
the induction law and alternating current circuit systems, aiming at understanding
electromagnetism applications to wireless charging. Chapter 11 introduces some
applications of the theory of electromagnetism in macromolecular solutions and
wireless charging technology.
Chapter 12 introduces electromagnetic waves in vacuum and medium, coherence
of electromagnetic waves, the polarization of electromagnetic waves, reflection and
refraction of electromagnetic waves, and Fresnel’s equations. Chapter 13 introduces
electromagnetic wave equations in dispersive media. This chapter also describes
the absorption, Lorentz’s oscillator model of a dielectric, the wave equation of a

vii
viii Preface

conductor, the wave equation of a dilute plasma, and the magnetized plasma or
dielectric.
Besides, Appendix introduces some mathematical background in vector analysis
and vector differential operators.
The textbook is geared more towards examples and problem-solving techniques.
The students will get a firsthand experience of how the theories in physics are applied
to problems in engineering and science. The textbook is mainly aimed at undergrad-
uate students in engineering and science. However, some chapters and sections are
aimed at senior undergraduate students working in the final year thesis in theoretical
and computational biophysics, physics, electrical and electronic engineering, and
chemistry.

Skopje, North Macedonia Hiqmet Kamberaj

January 2022

Acknowledgements I thank my family for their continuous support: Nera (my wife), Jon (my son)
and Lina (my daughter).
Contents

1 Electrostatics in Free Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Electrical Charges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Coulomb’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Coulomb’s Law for a System of Charges . . . . . . . . . . . . . . . . . . . . . 4
1.4 Electric Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.1 Force Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4.2 Superposition Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.5 Electric Field Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.6 Motion in Uniform Electric Field . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2 Gauss’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1 Electric Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1.1 Uniform Electric Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1.2 General Electric Field Flux . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2 Gauss’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.1 Gauss’s Law for a System of Charges . . . . . . . . . . . . . . . . 29
2.3 Applications of Gauss’s Law to Insulators . . . . . . . . . . . . . . . . . . . 32
2.4 Conductors in Electrostatic Equilibrium . . . . . . . . . . . . . . . . . . . . . 32
2.4.1 Property 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.4.2 Property 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.4.3 Property 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.4.4 Property 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3 Electrostatic Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.1 Electrostatic Potential Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2 Electric Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.3 Potential Difference in a Uniform Electric Field . . . . . . . . . . . . . . 50
3.4 Equipotential Surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

ix
x Contents

3.5 Electric Potential of a Point Charge . . . . . . . . . . . . . . . . . . . . . . . . . 52

3.6 Electric Potential of a System of Point Charges . . . . . . . . . . . . . . . 54
3.7 Electric Potential of a Continuous Charge Distribution . . . . . . . . . 57
3.8 Differential form of Electric Potential . . . . . . . . . . . . . . . . . . . . . . . 58
3.9 Multipole Expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.10 Electric Potential of a Charged Conductor . . . . . . . . . . . . . . . . . . . 63
3.10.1 Cavity Within a Conductor . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.11 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4 Capacitance and Dielectrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.1 Capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.2 Calculating Capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.2.1 Spherical Conductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.2.2 Parallel-Plate Capacitors . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.3 Combination of Capacitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.3.1 Parallel Combination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.3.2 Series Combination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.4 Energy Storage in the Electric Field . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.5 Electrostatics of Macroscopic Media and Dielectrics . . . . . . . . . . 87
4.5.1 Dielectrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.5.2 Comparison Between Dielectric Materials
and Conductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.5.3 Molecular Theory of Dielectrics . . . . . . . . . . . . . . . . . . . . 89
4.5.4 Energy Stored in Capacitor . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.6 Electric Polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.7 Set of Maxwell Equations for Electrostatic Field . . . . . . . . . . . . . . 95
4.7.1 Maxwell Equations for Free Space Electrostatic
Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
4.7.2 Maxwell Equations for Dielectric Media
Electrostatic Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
4.8 Potential Energy of Electrostatic Field . . . . . . . . . . . . . . . . . . . . . . . 97
4.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5 Electric Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
5.1 Electric Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
5.1.1 Direction of Electric Current . . . . . . . . . . . . . . . . . . . . . . . 110
5.1.2 Charge Carrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
5.2 Microscopic Model of Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
5.3 Resistance and Ohm’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5.3.1 Ohm’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.3.2 Classical Model for Electrical Conduction . . . . . . . . . . . . 115
5.3.3 Resistance and Temperature . . . . . . . . . . . . . . . . . . . . . . . . 118
5.4 Superconductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
5.5 Electric Energy and Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Contents xi

5.6 Electromotive Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

5.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
6 Magnetic Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
6.1 Magnetic Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
6.2 Magnetic Force Acting on a Current-Carrying Conductor . . . . . . 134
6.3 Torque on a Current Loop in a Uniform Magnetic Field . . . . . . . . 137
6.4 Motion of a Charged Particle in a Uniform Magnetic Field . . . . . 141
6.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
7 Sources of Magnetic Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
7.1 Biot-Savart Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
7.2 Magnetic Force Between Two Parallel Conductors . . . . . . . . . . . . 156
7.3 Ampére’s Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
7.4 Magnetic Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
7.5 Gauss’s Law in Magnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
7.6 Displacement Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
7.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
8 Magnetism in Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
8.1 Magnetic Moments of Atoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
8.2 Magnetization Vector and Magnetic Field Strength . . . . . . . . . . . . 179
8.3 Classification of Magnetic Substances . . . . . . . . . . . . . . . . . . . . . . . 181
8.3.1 Ferromagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
8.3.2 Paramagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
8.3.3 Diamagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
8.4 The Magnetic Field of the Earth . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
8.5 Faraday’s Law of Induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
8.6 Rowland Ring Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
8.7 Maxwell’s Equations of Magnetism . . . . . . . . . . . . . . . . . . . . . . . . . 192
8.8 Vector Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
8.9 Multipole Expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
8.10 Energy of the Magnetic Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
8.11 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
9 Maxwell’s Equations of Electromagnetism . . . . . . . . . . . . . . . . . . . . . . . 207
9.1 Maxwell’s Equations of Electromagnetism . . . . . . . . . . . . . . . . . . . 207
9.2 Vector and Scalar Potentials of Electromagnetic Field . . . . . . . . . 210
9.3 Electromagnetic Field Energy and Conservation Law . . . . . . . . . . 212
9.4 Conservation Law of Momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
9.5 Dynamics of Charged Particles in Electromagnetic Fields . . . . . . 218
9.6 Macroscopic Maxwell Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
9.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
xii Contents

10 More About Faraday’s Law of Induction . . . . . . . . . . . . . . . . . . . . . . . . 233

10.1 Moving Conductor in a Closed Circuit . . . . . . . . . . . . . . . . . . . . . . 233
10.1.1 Induced Electric Potential and Electric Field . . . . . . . . . . 235
10.1.2 Generators and Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
10.2 Inductance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
10.2.1 Self-inductance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
10.2.2 Mutual Inductance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
10.3 Oscillations in an LC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
10.4 The RL Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
10.5 The RLC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
10.5.1 Case 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
10.5.2 Case 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
10.5.3 Case 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
10.6 Alternating Current Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
10.6.1 AC Sources and Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
10.6.2 Resistors in an AC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . 256
10.6.3 Inductors in an AC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . 260
10.6.4 Capacitors in an AC Circuit . . . . . . . . . . . . . . . . . . . . . . . . 262
10.6.5 The RLC Series in an AC Circuit . . . . . . . . . . . . . . . . . . . . 263
10.7 Power in the AC Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
10.8 Resonance in the RLC Series Circuit . . . . . . . . . . . . . . . . . . . . . . . . 268
10.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
11 Some Applications of Electromagnetic Theory . . . . . . . . . . . . . . . . . . . 283
11.1 Electrostatic Properties of Macromolecular Solutions . . . . . . . . . . 283
11.1.1 The pH and Equilibrium Constant . . . . . . . . . . . . . . . . . . . 283
11.1.2 Charge on DNA and Proteins . . . . . . . . . . . . . . . . . . . . . . . 285
11.1.3 Charge States of Amino Acids . . . . . . . . . . . . . . . . . . . . . . 286
11.1.4 Salt Binding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
11.1.5 Energy Cost of Assembling a Collection of Charges . . . 292
11.1.6 The Poisson-Boltzmann Equation . . . . . . . . . . . . . . . . . . . 295
11.1.7 Calculation of pKa of Amino Acids
in Macromolecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
11.2 Wireless Charging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
11.2.1 Tightly Coupled Wireless Power Systems . . . . . . . . . . . . 304
11.2.2 Loosely Coupled Highly Resonant Systems . . . . . . . . . . . 307
11.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
12 Electromagnetic Waves in Vacuum and Linear Medium . . . . . . . . . . 319
12.1 Electromagnetic Wave Equations in Vacuum . . . . . . . . . . . . . . . . . 319
12.2 Relationships Between k, E, B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
12.3 Electromagnetic Waves Equations in Linear Medium . . . . . . . . . . 323
12.4 Energy and Momentum of Electromagnetic Waves . . . . . . . . . . . . 325
12.5 Coherence of Electromagnetic Waves . . . . . . . . . . . . . . . . . . . . . . . 327
Contents xiii

12.6 Polarization of Electromagnetic Waves . . . . . . . . . . . . . . . . . . . . . . 330

12.6.1 Linear Polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
12.6.2 Circular and Elliptical Polarization . . . . . . . . . . . . . . . . . . 331
12.7 Reflection and Refraction of Electromagnetic Waves . . . . . . . . . . 333
12.7.1 Laws of Reflection and Refraction . . . . . . . . . . . . . . . . . . . 334
12.8 Fresnel Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
12.8.1 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
12.8.2 Perpendicular Polarization . . . . . . . . . . . . . . . . . . . . . . . . . 337
12.8.3 Parallel Polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
12.8.4 External and Internal Reflection . . . . . . . . . . . . . . . . . . . . . 341
12.8.5 Normal Incidence of Electromagnetic Waves . . . . . . . . . 345
12.8.6 Reflectance and Transmittance . . . . . . . . . . . . . . . . . . . . . . 346
12.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
13 Electromagnetic Waves in Dispersive Media . . . . . . . . . . . . . . . . . . . . . 359
13.1 Dispersion and Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
13.1.1 Lorentz’s Model of Oscillations in Dielectrics . . . . . . . . . 360
13.2 Dispersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
13.2.1 Wave Packets and Group Velocity . . . . . . . . . . . . . . . . . . . 366
13.2.2 Normal and Anomalous Dispersion . . . . . . . . . . . . . . . . . . 368
13.3 Refractive Index of a Conductor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
13.4 Wave Propagation in a Dilute Plasma . . . . . . . . . . . . . . . . . . . . . . . . 372
13.4.1 Electromagnetic Waves in a Dilute Plasma . . . . . . . . . . . 373
13.4.2 Phase and Group Velocity in a Dilute Plasma . . . . . . . . . 374
13.4.3 Plasma and Dielectric at High Frequency . . . . . . . . . . . . . 375
13.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378

Appendix: Vectorial Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379

Chapter 1
Electrostatics in Free Space

The chapter aims to introduce some basic concepts of the

electrostatics, such as the concept of charge, interactions
between the charged particles, and the electric field.

In this chapter, we will introduce the electrostatics in free space. First, we will intro-
duce the concept of the charges, and then present Coulomb’s law of the interactions
between the charges. Next, we discuss the concept of electric field and electric field
lines. Also, we will describe the motion of a charged particle in the presence of an
electric field. The reader can also consider other literature (Holliday et al. 2011) for
further reading.

1.1 Electrical Charges

There exist several simple experiments to demonstrate the existence of electrical

charges and forces. For example,
1. When we comb our hair on a dry day, we find that the comb attracts pieces of
paper.
2. The same effect of attracting pieces of paper occurs when materials such as glass
or rubber are rubbed with silk or fur.
As a general rule, for every material behaving in that way, we can say that it is
electrified, or it becomes electrically charged.
Benjamin Franklin (1706–1790) found that there exist two types of electric
charges, namely positive and negative. The following experiment can be used to
demonstrate his finding. Suppose that we rubber with fur a hard rubber rod. In addi-
tion, we rub a glass rod with silk material. Then, if the glass rod is brought near

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 1

H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_1
 2 1 Electrostatics in Free Space

the rubber rod, we will observe that the two attract each other. However, if we bring
near each other two charged rubber rods or two charged glass rods, then the two
repel each other. This experiment indicates the existence of two different states of
electrification for the rubber and glass. Furthermore, it finds that like charges repel
Sign each other and unlike charges attract each other.
con-
vention By convention, the electric charge on the glass rod is positive, and that on the
of rubber rod is negative. Based on that convention, any charged object repelled by
charges
another charged object must have the same sign of charge with it, and any charged
object attracted by another charged object must have an opposite sign of charge. It is
important to note that the electricity model of Franklin implies that electric charge
is always conserved. That is, an electrified state (positive or negative) is due to the
charge transfer from one object to the other. In other words, when an object gains
some amount of positive/negative charge, then the other gains an equal amount of
Quanti- the electric charge of the opposite sign.
zation
of Robert Millikan (1868–1953), in 1909, discovered that electric charge always
charges appears as a multiple integer of a fundamental amount of charge, called e such that
the electric charge q, which is a standard symbol for the charge, is quantized as

q = Ne (1.1)

Here, N is an integer number, N = 0, ±1, ±2, . . ..

1.2 Coulomb’s Law

Based on an experiment performed by Coulomb, the electric force between two

charged particles at rest is proportional to the inverse of the square of distance r
between them and directed along the line joining the two particles. In addition, the
electric force is proportional to the charges q1 and q2 on each particle. Also, the
electric force is attractive if the charges are of opposite sign and repulsive if the
charges have the same sign. That is known as Coulomb’s Law.
Coulomb’s
Law Definition 1.1 Force is proportional to the product of the magnitudes of the charges
and inversely proportional to the square of the distance between them. Mathemati-
cally, the law may be written as

| q1 || q2 |
F = ke (1.2)
r2
In Eq. (1.2), ke is the Coulomb constant. Note that, in SI, the unit of charge is the
coulomb (C). Therefore, the Coulomb constant ke in SI units has the value

ke = 8.9875 × 109 N · m2 /C2 (1.3)

Often, the constant is written as

1.2 Coulomb’s Law 3

Fig. 1.1 Electric force

vectors of the interactions
between two charges
(positive or negative)

1
ke = (1.4)
4π0

where 0 is the permittivity of free space given by

0 = 8.8542 × 10−12 C2 /N · m2 (1.5)

Coulomb’s force is a vector; hence it has a magnitude expressed by Eq. (1.2) and a Coulomb’s
force
direction. Therefore, the Coulomb’s law can be expressed in vector form concerning vector
the electric force, F12 , exerted by the charge q1 (positive or negative) on another
charge q2 (positive or negative) as
q1 q2
F12 = ke r̂ (1.6)
r2

In Eq. (1.6), r̂ denotes a unit vector pointing from q1 to q2 . Note that based on
the Newton’s third law, the electric force, F21 , exerted by a charge q2 (positive or
negative) on a second charge q2 (positive or negative) is

F21 = −F12 (1.7)

Figure 1.1 illustrates graphically the direction of Coulomb’s force vectors for
different combinations of the pairs of positive and negative charges, namely negative-
negative, positive-positive, and negative-positive charge-charge interactions.
 4 1 Electrostatics in Free Space

Fig. 1.2 A system of N

interacting charges

1.3 Coulomb’s Law for a System of Charges

Consider a system of N charges (positive and negative), q1 , q2 , . . . , q N , as shown in

Fig. 1.2. The force exerted on any charge i from all other charges is


N 
N
qi q j
Fi = F ji = ke r̂ ji (1.8)
j=1=i j=1=i
r 2ji

where r̂ ji is a unit vector directed from q j to qi . In Eq. (1.8), F ji is the force exerted on
i charge particle from j charge particle, and the some runs over all charges excluding
the charge i.

1.4 Electric Field

1.4.1 Force Fields

The field forces act through space, producing an effect even when no physical contact
between the objects occurs. As an example, we can mention the gravitational field.
Michael Faraday developed a similar approach to electric forces. That is, an electric
field exists in the region of space around any charged body, and when another charged
The body is inside this region of the electric field, an electric force acts on it.
electric
field
Definition 1.2 The electric field E at a point in space is defined as the electric force
Fe acting on a positive test charge q0 placed at that point divided by the magnitude
of the test charge:
Fe
E= (1.9)
q0
1.4 Electric Field 5

Fig. 1.3 A positive test

charge q0 placed in the
electric field, E, of another
charge Q distributed
uniformly in a sphere

The vector E has the SI units of newtons per coulomb (N/C). Figure 1.3 illustrates
the electric field E created by a positively charged sphere with total charge Q at
the positive test charge q0 . Here, we have assumed that the test charge q0 is small
enough that it does not disturb the charge distribution of the sphere responsible for
the electric field.
Note that E is the field produced by some charge external to the test charge, and
it is not the field produced by the test charge itself. Also, note that the existence
of an electric field is a property of its source. For example, every electron comes
with its electric field. An electric field exists at a point if a test charge at rest at that
point experiences an electric force. The electric field direction is the direction of
the force on a positive test charge placed in the field. Once we know the magnitude
and direction of the electric field at some point, the electric force exerted on any
charged particle (either positive or negative) placed at that point can be calculated.
The electric field exists at some point space, including the free space, independent
of the existence of another test charge at that point. Direc-
tion of
To determine the direction of electric field, consider a point charge q located some electric
distance r from a test positive charge q0 located at a point P, as shown in Fig. 1.4. field

Coulomb’s law defines the force exerted by q on q0 as

qq0
Fe = k e r̂ (1.10)
r2

where r̂ represents the usual unit vector directed from q toward q0 (see Fig. 1.4).
Electric field created by q (positive or negative) is
 6 1 Electrostatics in Free Space

Fig. 1.4 A positive test

charge q0 placed at distance
r from another charge q
(positive or negative)

Fe q
E= = ke 2 r̂ (1.11)
q0 r

From Eq. (1.11), when q < 0, then E is pointing opposite to vector r̂, and hence
the electric field of a negative charge is pointing toward that charge, see Fig. 1.4a.
On the other hand, when q > 0, E and r̂ are parallel, and hence the electric field of
a positive charge is pointing away from that charge, as shown in Fig. 1.4b.

1.4.2 Superposition Principle

Super-
posi- According to superposition principle, at any point P, the total electric field due to a
tion
princi- set of discrete point charges, q1 , q2 , . . . , q N , positive and negative charges, is equal to
ple of
dis- the sum of the individual charge electric field vectors (see Fig. 1.5). Mathematically,
crete we can write
charges
N N
qi
E(r) = Ei = ke r̂i (1.12)
i=1 i=1
| r − ri |2

In Eq. (1.12), | r − ri | is the distance from qi to the point P (the location of a test
charge), where r is the position vector of the point P with respect to some reference
frame, as indicated in Fig. 1.5, and ri is the position vector of the charge i in that
reference frame. Furthermore, r̂i is a unit vector directed from qi toward P.
Note that in Eq. (1.12) the dependence of E on only position vector of point
P, r, assumes a static configuration of the charges in space. That is, for some other
configuration distribution of charges in space, E at the same point P may be different.
Note that often for convenience, Eq. (1.12) is also written as
1.4 Electric Field 7

Fig. 1.5 Superposition of

the electric field created by
set of charges at the point P


N
qi (r − ri )
E(r) = ke (1.13)
i=1
| r − ri |3

where
r − ri
r̂i = (1.14)
| r − ri |
Elec-
If the distances between charges in a set of charges are much smaller, compare tric
field of
with the distance of the set from a point where the electric field is to be calculated, a
then charge distribution is continuous. contin-
uous
To calculate the net electric field created by a continuous charge distribution charge
distri-
in some volume V , we follow these steps. First, we divide the charge distribution bution
into macroscopically small elements with small charge Δqi , as shown in Fig. 1.6a.
Δqi = ρi ΔV , where ρi is seen from a microscopic viewpoint as a uniform charge
density within the volume element i, which represents one of the possible configura-
tions of microscopic description. It is important to note that with “macroscopically
small” we should understand a small volume in space with a characteristic micro-
scopic configuration of the charges inside it that can, on average, macroscopically
be represented as a point-like charge, Δqi . Then, we calculate the electric field due
to one of these macroscopically point charges, Δqi , at some point P at distance
| r − ri | from the charge element, Δqi , as

Δqi
ΔE(r, ri ) = ke rˆi (1.15)
| r − ri |2

where rˆi is a unit vector directed from the charge element Δqi toward P. Here, r is
position vector of point P in some reference frame, and ri is the position vector of
the macroscopically point charge Δqi .
To evaluate the total electric field at P due to all charge elements distributed in
the volume V , we apply the superposition principle:
8 1 Electrostatics in Free Space

Fig. 1.6 Superposition of the electric field created by a continuous charge distribution at the point
P. a Macroscopically small volume elements ΔV with charge Δqi = ρi ΔV , where ρi is the micro-
scopic uniform charge distribution of a small volume ΔV , and b a continuous charge distribution in
the limit when ΔV → 0 with a position dependent charge density ρ at the position r with respect
to some frame

 Δqi
E(r) ≈ ke r̂i (1.16)
i
| r − ri |2

Using the limit of small element volume ΔV → 0,

 ρi ΔV
E(r) = lim ke r̂i (1.17)
ΔV →0
i
| r − ri |2
  
ρ(r ) r − r
= ke dr
V | r − r  |3

In Eq. (1.17), r denotes the position vector of a point in the volume V with respect
to reference frame, as shown in Fig. 1.6b, and the integration is performed over all
volume extended by the charge distribution. Note that r̂i (or r̂ when ΔV → 0)
depends on the position of the volume element (or position vector r when ΔV → 0,
r̂ = (r − r )/ | r − r |), and hence, it cannot be put outside the sign of integral.
Furthermore, d V = dr .
If we assume that ρ is constant, i.e., uniform charge distribution, then
  
r − r
E(r) = ke ρ dr (1.18)
V | r − r  |3

Note that if the charge Q is uniformly distributed in a volume V , then the

charge density can be expressed as
1.4 Electric Field 9
 
Q C
ρ= (1.19)
V m3

For a charge Q uniformly distributed on a surface area A, a surface charge

density σ is defined as  
Q C
σ= (1.20)
A m2

If a charge Q is uniformly distributed along any line of length L, a linear

charge density μ is as follows:
 
Q C
μ= (1.21)
L m

In general, for a charge Q nonuniformly distributed over a volume, surface,

or line, we can define the charge densities as

dQ
ρ= (1.22)
dV
dQ
σ=
dA
dQ
μ=
dL

1.5 Electric Field Lines

Elec-
tric
By definition, electric field lines are drawn to follow the same direction as the electric field
field vector at any point. Furthermore, the electric field vector is tangent to the line lines

at every point along the field line.

The electric field lines are such that E is tangent to the electric field line at each
point. The number of lines per unit surface area passing a surface perpendicular
to the lines is proportional to the magnitude | E | in that region. Furthermore,
the lines are directed radially away from the positive point charge. Moreover,
the lines are directed radially toward the negative point charge.

In Fig. 1.7, we show the electric field lines of a negative and positive point charge.
It can be seen that for a negative point charge, −q, the electric field lines are drawn
toward the charge (see Fig. 1.7a). On the other hand, for a positive point charge, +q,
electric field lines are leaving the charge, as shown in Fig. 1.7b.
10 1 Electrostatics in Free Space

Fig. 1.7 a Electric field lines

of a negative point charge
−q; b Electric field lines of a
positive point charge +q

Fig. 1.8 a Electric field lines

of a negative point charge
−q; b Electric field lines of a
positive point charge +q; c
Electric field lines of the
resultant electric field. The
positive charge is located at
the point (0, 3, 0) and the
negative charge at (0, −3, 0)

The following general rules for drawing electric field lines apply:

The lines start from a positive charge and end on a negative charge. Also, the
number of lines drawn, leaving a positive charge, or approaching a negative
charge is proportional to the magnitude of the charge. Moreover, no two field
lines can cross.

In Fig. 1.8, we show the electric field vector for a positive point charge +q located
at the point (0, 3, 0) (Fig. 1.8b) and a negative point charge −q located at (0, −3, 0)
(Fig. 1.8a), colored according to the magnitude of the electric field E using a color
scaling, as depicted in Fig. 1.8. Besides, the electric field lines of the resultant electric
field are shown in Fig. 1.8c.

1.6 Motion in Uniform Electric Field

Suppose a charge particle of mass m and charge q is moving in a uniform electric

field E. Electric field E exerts on a particle placed in it the force

F = qE (1.23)
1.6 Motion in Uniform Electric Field 11

If that force is equal to the resultant force exerted on the particle, it causes the particle
to accelerate, based on Newton’s second law:

ma = qE (1.24)

The acceleration gained by the charge is given as

q
a= E (1.25)
m
Therefore, if E is uniform (that is, constant in magnitude and direction), then a is
constant. Furthermore, if the particle has a positive charge, then its acceleration is
in the direction of the electric field. On the other hand, if the particle has a negative
charge, then its acceleration is in the direction opposite the electric field.

1.7 Exercises

Exercise 1.1 Suppose the electron and proton are in a hydrogen atom. Assume
their average separation is approximately 5.3 × 10−11 m. What are the mag-
nitudes of the electric and gravitational force between the two particles?

Solution 1.1 Using the Coulomb’s law,

| −e || +e |
Fe = ke (1.26)
r2
  (1.60 × 10−19 C)2
= 8.9875 × 109 N · m2 /C2
(5.3 × 10−11 m)2
−8
= 8.2 × 10 N

Using the Newton’s law of gravitation,

mem p
Fg = G 2 (1.27)
r
 
N · m2 (9.11 × 10−31 kg)(1.67 × 10−27 kg)
= 6.70 × 10−11
kg2 (5.3 × 10−11 m)2
= 3.6 × 10−47 N

The ratio is
Fe 8.2 × 10−8 N
= ≈ 2 × 1039 (1.28)
Fg 3.6 × 10−47 N
12 1 Electrostatics in Free Space

Exercise 1.2 Consider a configuration in space of three point charges at the

corners of a right triangle. Their charges are q1 = q3 = 5.0 µC, q2 = −2.0 µC,
and a = 0.10 m, as shown in Fig. 1.9. Find the resultant force exerted on q3 .

Solution 1.2 Using the Coulomb’s law, F3 = F13 + F23 , and

F3 = F3x i + F3y j (1.29)

where

F3x = F13x + F23x (1.30)

F3y = F13y + F23y

a √
F13x = F13 cos θ = F13 √ = F13 / 2 (1.31)
2a
a √
F13y = F13 sin θ = F13 √ = F13 / 2
2a

and
| q1 || q3 |
F13 = ke √ (1.32)
( 2a)2
 
N · m2 (5.0 × 10−6 C)(5.0 × 10−6 C)
= 8.9875 × 109
C2 2(0.10 m)2
= 11 N

Thus, √ √
F13x = 11 N/ 2 = 7.9 N; F13y = 11 N/ 2 = 7.9 N (1.33)

Furthermore,

F23x = −F23 (1.34)

F23y = 0 N

and
1.7 Exercises 13

Fig. 1.9 A system of three

point charges

| q2 || q3 |
F23 = ke (1.35)
a2
 2
9 N·m (−2.0 × 10−6 C)(5.0 × 10−6 C)
= 8.9875 × 10
C2 (0.10 m)2
=9N

Thus,
F23x = −9 N (1.36)

Then,

F3x = F13x + F23x = −1.1 N (1.37)

F3y = F13y = 7.9 N

and
F3 = (−1.1i + 7.9j) N (1.38)

Exercise 1.3 It can be seen that the point charges lie along the x-axis, see
also Fig. 1.10. The positive charge q1 = 15.0 µC is at x = 2.00 m, the positive
charge q2 = 6.00 µC is at the origin, and the resultant force acting on q3 is
zero. What is the x coordinate of q3 ?

Solution 1.3 We know F3 = F13 + F23 = 0, or projecting along the x-axis:

F3x = F13x + F23x = 0 (1.39)

Therefore,
14 1 Electrostatics in Free Space

Fig. 1.10 A system of three

point charges along the
x-axis

F13x = −F23x (1.40)

Furthermore,

| q1 || q3 |
F13x = +F13 = ke (1.41)
(2.0 − x)2
| q2 || q3 |
F23x = −F23 = −ke
x2
Then,
| q1 || q3 | | q2 || q3 |
ke = ke (1.42)
(2.0 − x)2 x2
or
(2.0 − x)2 | q2 |= x 2 | q1 | (1.43)

After replacing the numerical values, we get

(4.0 − 4.0x + x 2 )(6.0 × 10−6 C) = x 2 (15.0 × 10−6 C) (1.44)

Solving for x we get x = 0.775 m.

Exercise 1.4 A charge q1 = 7.0 µC is located at the origin, and a second

charge q2 = −5.0 µC is located on the x-axis, 0.30 m from the origin (see also
Fig. 1.11). What is the electric field at the point P with coordinates (0, 0.40) m?
1.7 Exercises 15

Fig. 1.11 A system of two

point charges along the
x-axis

Solution 1.4 Using superposition principle, E = E1 + E2 , where

q1
E1 = k e r̂1 (1.45)
r12
 2
9 N·m 7.0 × 10−6 C
= 8.9875 × 10 j
C2 (0.40 m)2
 
N
= 3.9 × 105 j
C

and
q2
E2 = k e r̂2 (1.46)
r22
 
N · m2 −5.0 × 10−6 C
= 8.9875 × 109 (− cos θi + sin θj)
C2 (0.50 m)2
= (−1.8 × 105 N/C) (−0.6i + 0.8j)

Then,  
E ≈ 1.1 × 105 i + 2.5 × 105 j N/C (1.47)

Exercise 1.5 Consider an electric dipole formed by a positive charge q and

a negative charge −q separated by some distance (see Fig. 1.12). Find the
electric field E at P due to the charges, where P is a distance y  a from the
origin.
16 1 Electrostatics in Free Space

Fig. 1.12 An electric dipole

formed by two point charges
along the x axis

Solution 1.5 Using superposition principle,

E = E1 + E2 (1.48)

q
E1 = k e r̂1 (1.49)
r2
q
= ke 2 (cos θi + sin θj)
r 

q a y
= ke 2 i+ j
a + y2 a2 + y2 a2 + y2
q
= ke 2 (ai + yj)
(a + y 2 )3/2

Similarly,

−q
E2 = k e r̂2 (1.50)
r2
−q
= ke 2 (− cos θi + sin θj)
r 

−q a y
= ke 2 − i+ j
a + y2 a2 + y2 a2 + y2
q
= −ke 2 (−ai + yj)
(a + y 2 )3/2

Then,
2qa 2qa
E = ke i ≈ ke 3 i (1.51)
(a 2 +y )
2 3/2 y
1.7 Exercises 17

Fig. 1.13 Electric field

created by a rod of length 
having a uniform charge
distribution Q

Exercise 1.6 A rod of length  has a uniform positive charge per unit length
μ and a total charge Q (see Fig. 1.13). What is the electric field at P located
along the long axis of the rod and at distance a from one end?

Solution 1.6 We can chose a reference frame as shown in Fig. 1.13. Then the charge
in a small element d x at distance x from P is dq = μd x > 0 and its electric field at
P is
dq
dE = −ke 2 i (1.52)
x
and the total electric field is
 a+
dq
E = −ke i (1.53)
a x2
 a+ 
dx
= −ke μ i
a x2
 
1 a+
= −ke μ − i
x a
 
1 1
= −ke μ − + i
a+ a

= −ke μ i
a(a + )
1
= −ke Q i
a(a + )

If a  , then
1
E = −ke Q i (1.54)
a2
which is the electric field of a point charge Q at a distance a from its position.
18 1 Electrostatics in Free Space

Fig. 1.14 A configuration of

three charges with equal
magnitude:
| q1 |=| q2 |=| q3 |,
distributed along the line of a
circle with radius r

Exercise 1.7 Consider the configuration of three charges shown in Fig. 1.14.
They have equal magnitude, q, and are placed along the line of a circle separated
by an angle of 120◦ . Find the resultant electric field at the center of the circle,
if it has a radius of r .

Solution 1.7 The electric field E1 of the negative charge q1 at the center of the circle
O is q
E1 = −ke 2 r̂1 (1.55)
r
Similarly, the electric field E2 of the negative charge q2 at the center of the circle O
is q
E2 = −ke 2 r̂2 (1.56)
r
and of the positive charge q3 is E3 given as
q
E3 = k e r̂3 (1.57)
r2
Choosing a vertical y-axis, as shown in Fig. 1.14, we can express the following unit
vectors in terms of the vertical unit axis vector r̂0 as

1
r̂1 = cos 60◦ r̂0 = r̂0 (1.58)
2
1
r̂2 = cos 60◦ r̂0 = r̂0 (1.59)
2
r̂3 = −r̂0 (1.60)

Therefore, the resultant electric field at the center of the circle is

1.7 Exercises 19

E R = E1 + E2 + E3 (1.61)
 
q 1 1
= −ke 2 + + 1 r̂0
r 2 2
q
= −2ke 2 r̂0 (1.62)
r

Exercise 1.8 Consider again the configuration of three charges shown in the
previous exercise, having equal magnitude, q, and placed along the line of a
circle separated by an angle of 120◦ . Determine the sign of the charges for this
configuration for which the resultant electric field at the center of the circle
vanishes, if it has a radius of r .

Solution 1.8 In Fig. 1.15a and b, we show two possible configurations of charges
that produce a zero resultant electric field at O.
For the first configuration, shown in Fig. 1.15a, we have
q q
E1 = −ke 2
r̂1 = −ke 2 r̂0 (1.63)
r 2r
q q
E2 = −ke 2 r̂2 = −ke 2 r̂0 (1.64)
r 2r
q q
E3 = −ke 2 r̂3 = ke 2 r̂0 (1.65)
r r
Hence, the resultant field at the center of circle is

E R = E1 + E2 + E3 = 0 (1.66)

On the other hand, for the configuration, shown in Fig. 1.15b, we have

Fig. 1.15 A configuration of three charges with equal magnitude: | q1 |=| q2 |=| q3 |, distributed
along the line of a circle with radius r : a negative charges and b positive charges
20 1 Electrostatics in Free Space

q q
E1 = k e r̂1 = ke 2 r̂0 (1.67)
r2 2r
q q
E2 = ke 2 r̂2 = ke 2 r̂0 (1.68)
r 2r
q q
E3 = ke 2 r̂3 = −ke 2 r̂0 (1.69)
r r
Then, the resultant field at the center of circle is

E R = E1 + E2 + E3 = 0 (1.70)

In our calculations, r̂1 , r̂2 , r̂3 , r̂0 are unit vectors as indicated in Fig. 1.15.

Exercise 1.9 A positive point charge q of mass m is released from rest in a

uniform electric field E directed along the x-axis. Describe its motion.

Solution 1.9 The acceleration is q

a= E (1.71)
m
since E is constant, then a is constant and different from zero only along x-axis:
q
ax = Ex (1.72)
m
ay = 0
az = 0

and the motion is along x-axis.

From kinematics,

x f = xi + vxi t + ax t 2 /2 (1.73)
vx f = vxi + ax t
vx2 f − vxi
2
= 2ax (x f − xi )

For xi = 0 and vxi = 0, we get

q Et 2
x f = ax t 2 /2 = (1.74)
2m
q Et
vx f = ax t = (1.75)
m
and
1.7 Exercises 21

vx2 f = 2ax x f (1.76)

qE
=2 xf
m
 
qE 2 2
= t
m

The final kinetic energy

K f = mvx2 f /2 (1.77)
 
m qE 2 2
= t
2 m
(q E)2 2
= t
2m
Work done by electric force is

W = ΔK = K f − K i = K f (1.78)
(q E) 22
= t
2m
= q E x f = Fx x f

Exercise 1.10 Consider two oppositely charged flat metallic plates creating
an electric field in the region between them approximately uniform. Suppose
an electron of charge −e is projected horizontally into this field with an initial
velocity vi i. Describe the motion of the electron.

Solution 1.10 Because the electric field E is in the positive y direction, the accel-
eration is
eE
a=− j (1.79)
m
and vxi = vi and v yi = 0, thus

x f = vi t (1.80)
eE 2
y f = a y t 2 /2 = − t (1.81)
2m
vx f = vi (1.82)
eE
vy f = ay t = − t (1.83)
m
22 1 Electrostatics in Free Space

From here, we get

 2
eE xf
yf = − (1.84)
2m vi

Hence, the trajectory is a parabola.

Reference

Holliday D, Resnick R, Walker J (2011) Fundamentals of physics. John Wiley and Sons, New York
Chapter 2
Gauss’s Law

This chapter aims to introduce Gauss’s law and its applications

to electrostatics.

In this chapter, we introduce the electric flux and Gauss’s law. Also, the application of
Gauss’s law to insulators and conductors will be discussed. As extra reading material,
the reader can also consider other literature (Holliday et al. 2011).

2.1 Electric Flux

2.1.1 Uniform Electric Field

The electric flux concept describes quantitatively the electric lines. The number of Uni-
form
field lines per unit area (also called line density) going through a rectangular surface of electric
area A, which is perpendicular to the field, is proportional to the magnitude of electric field
flux
field, E, as shown in Fig. 2.1. Furthermore, the total number of lines penetrating
the surface is proportional to the product | E | A. By definition, the product of the
magnitude of electric field | E | and surface area A perpendicular to the field is called
the electric flux:
Φ E =| E | A (2.1)

Using Eq. (2.1), from the SI units of E and A, we derive the SI units of the electric
flux:  
N
[E] = , [A] = [m2 ] (2.2)
C

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 23

H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_2
24 2 Gauss’s Law

Fig. 2.1 A uniform electric

field perpendicular to a
rectangular surface area A

Fig. 2.2 A uniform electric

field penetrating a
rectangular surface area A
with unit vector n
perpendicular to surface
forming an angle θ with E

Thus, we obtain SI units of Φ E :

 
N · m2
[Φ E ] = (2.3)
C

Note that the electric flux is proportional to the number of electric field lines
penetrating some surface.
Moreover, consider the electric flux on any surface with an arbitrary orientation
with respect to electric field E, as shown in Fig. 2.2. Electric flux going through the
surface (with area A) not perpendicular to E is smaller than the product | E | A. That
is, the number of lines that cross this area A is equal to the number of lines that cross
the area A = A cos θ, which is a projection of A aligned perpendicular to the field.
Mathematically, the electric flux is given by (Fig. 2.2)

Φ E =| E | A =| E | A cos θ (2.4)

From the definition, Eq. (2.4), we can say that the maximum electric flux is
achieved when θ = 0◦ ; that is, the surface is perpendicular to E: Φ Emax =| E | A
(see also Eq. (2.1)). Or, equivalently, when normal vector n to the surface is parallel
2.1 Electric Flux 25

Fig. 2.3 A nonuniform

electric field penetrating an
arbitrary element surface
area ΔAi with unit vector ni
perpendicular to surface
forming an angle θi with Ei
at that point

to E. On the other hand, the minimum electric flux is achieved when θ = 90◦ , that
is, the surface is parallel to E: Φ Emin = 0. In this case, normal vector n to the surface
is perpendicular to E. In general, denoting the vector A = An, we can write

ΦE = E · A (2.5)

2.1.2 General Electric Field Flux

Let us consider the case of a general surface with a nonuniform electric field Ei ,
as shown in Fig. 2.3. We can partition the surface on small infinitesimal elements
ΔAi , such that electric field is constant on every point of ΔAi , then the electric flux
through ΔAi is
ΔΦ E,i = Ei · ΔAi (2.6)

The total flux can be approximated as


ΦE ≈ Ei · ΔAi (2.7)
i

Taking the limit when ΔAi → 0 on both sides of Eq. (2.7), we obtain the exact
electric flux through general surface (Fig. 2.3): Elec-
tric
  flux
through
Φ E = lim Ei · ΔAi = E · dA (2.8) general
ΔAi →0 surface
i sur f ace
 26 2 Gauss’s Law

Fig. 2.4 A nonuniform

electric field penetrating a
closed surface

Note that Eq. (2.8) gives the electric flux through any surface. Often, the electric
Closed flux is calculated through a closed surface.
surface
A closed surface defines the surface that divides space into the inside and outside
regions, and to move from one region to another, one has to cross the surface. For
Elec- example, the surface of a sphere, ellipsoid, etc., are all closed surfaces.
tric
flux Consider a nonuniform electric field penetrating a closed surface, for example, a
through cylinder, as shown in Fig. 2.4. Every electric field line that passes through a surface
a
closed element ΔAi is going to leave the closed surface at some other surface element ΔA j .
surface
The electric flux going through the surface element ΔAi is

ΔΦ E,i = Ei ΔAi =| Ei | ΔAi cos θi (2.9)

In Eq. (2.9), θi is the angle between electric field Ei and unit vector normal to surface
element ni : ⎧
⎨ > 0, 0 ≤ θi < 90◦
cos θi = = 0, θi = 90◦ (2.10)
⎩
< 0, 90◦ < θi ≤ 180◦

Similarly, the electric flux going through the surface element ΔA j is

ΔΦ E, j = E j ΔA j =| E j | ΔA j cos θ j (2.11)

Here, θ j is the angle between E j and unit vector normal to surface element n j .
If by convention, we define unit vector ni as outward to the surface element,
then ΔΦ E, j > 0 indicates that electric field lines leave the closed surface through
ΔA j , and ΔΦ E,i < 0 indicates that electric field lines are entering the closed surface
through ΔAi .
Therefore, if the electric field lines are entering the closed surface, then their
electric flux contribution is negative. If the electric field lines are leaving the closed
surface, their contribution to electric flux is positive.
Since the total flux is approximated as

ΦE ≈ Ei · ΔAi (2.12)
i
2.1 Electric Flux 27

therefore, the net electric flux through the surface is proportional to the net number
of lines leaving the surface, or the number leaving the surface minus the number
entering the surface.
As a consequence, if more lines are leaving than entering, the net flux is positive.
On the other hand, if more lines are entering than leaving, the net flux is negative.
Mathematically, we can write the net flux Φ E through a closed surface as

ΦE = E · dA (2.13)
S

= E ·ndA
S

= En d A
S

2.2 Gauss’s Law

Gaus-
sian
By definition, a Gaussian surface is called a closed surface. Gauss’s law relates the surface
electric flux through the closed surface and the charge enclosed by the surface.
For illustration consider a nonconducting sphere of radius r with an elementary
positive charge +q at the center of sphere. Note that the sphere’s surface is equal to
Gaussian surface, as indicated in Fig. 2.5. Electric field created by the charge +q at
any point at distance r from the charge is
q
E = ke r̂ (2.14)
r2

Fig. 2.5 A nonconducting

sphere of radius r containing
an elementary positive
charge at the center, +q
 28 2 Gauss’s Law

Fig. 2.6 Electric flux of an

elementary charge, q

The electric flux through the sphere surface is

q
ΦE = E·ndA = 2
r̂ · n d A
ke (2.15)
r
S S

q q q
= ke 2 d A = ke 2 d A = ke 2 (4πr 2 )
S r r S r
= 4πke q

or
1 q
Φ E = q4π = (2.16)
4π0 0

The result given by Eq. (2.16) indicates that Φ E is independent of radius r . Further-
more, since both q and 0 are constants, then the electric flux is constant. Therefore,
the same electric flux is passing through the surface of any other sphere with radius
R > r , which has an elementary charge at the center.
Now, consider several closed surfaces surrounding a charge q, as shown in Fig. 2.6,
S1 (spherical), S2 and S3 (nonspherical). The flux that passes through S1 is given by
Eq. (2.16). Since flux is proportional to the number of electric field lines passing
Flux of through a surface, then the flux through S2 and S3 also is constant (see Eq. (2.16)).
a point
charge By definition, the net flux through any closed surface is independent of the shape
of that surface. The net flux through an arbitrary closed surface surrounding a point
charge q is given by Eq. (2.16).
Now, suppose a charge +q is outside a closed surface (any shape), as shown in
Fig. 2.7. In that case, an electric field line that enters the surface leaves the surface
at another point. The number of electric field lines entering the surface equals the
number leaving the surface, thus
ΦE = 0 (2.17)

We can conclude that the net electric flux through a closed surface that surrounds no
charge is zero (Fig. 2.7).
2.2 Gauss’s Law 29

Fig. 2.7 Electric flux of an

elementary charge, q outside
a closed surface area

2.2.1 Gauss’s Law for a System of Charges

Consider a system of N discrete point charges, q1 , q2 , . . . , q N , positive or negative,

as shown in Fig. 2.8. Using the superposition principle of electric field, discussed in
Chap. 1, we can write
 N
E= Ei (2.18)
i=1

Then, the net electric flux through a closed surface is

ΦE = E · dA (2.19)
S

 N

= Ei · dA
S i=1

Fig. 2.8 Electric flux of a

set of discrete charges
through a Gaussian surface
 30 2 Gauss’s Law

N

= Ei · dA
i=1 S

where Ei is electric field of charge qi .

The electric flux for the charge qi through the closed surface is

Φ E,i = Ei · dA (2.20)
S
 qi
, if qi inside closed surface
= 0
0, if qi outside closed surface
Elec-
tric Therefore, the net electric flux becomes
flux of
a dis-
crete 
N
system
of ΦE = Φ E,i (2.21)
charges i=1


Nin
qi Q in
= =
i=1
0 0

where Q in is the net charge inside the closed surface and Nin denotes the number of
charges inside the closed surface.


Nin
Q= qi (2.22)
i=1

Gauss’s law for a system of charges q1 , q1 , . . . , q N says that the net electric flux
through any close surface is given by Eq. (2.21). When using this equation, we should
note that
1. The charge Q in is the net charge inside the closed surface.
2. E represents the total electric field, which includes contributions from charges
both inside and outside the surface.
For a continuous system, see also Fig. 2.9, we partition it into macroscopically
elementary charges Δqi , and then using the superposition principle of electric field,

E≈ ΔEi (2.23)
i

The net electric flux through a closed surface is approximately

2.2 Gauss’s Law 31

Fig. 2.9 Electric flux of a

continuous charge
distribution with density ρ
through a Gaussian surface.
Δqi and Δq j are two
macroscopically small
elementary charges

ΦE = E · dA (2.24)

 
S

≈ ΔEi · dA
S i

= ΔEi · dA
i S

where ΔEi is electric field of charge Δqi . The electric flux for a charge Δqi through
the closed surface is

ΔΦ E,i = ΔEi · dA (2.25)
⎧S
⎨ Δqi
, if Δqi inside closed surface
= 
⎩ 0, 0 if Δq outside closed surface
i

Therefore, the net electric flux is


ΦE ≈ ΔΦ E,i (2.26)
i


Nin
Δqi
=
i=1
0
Elec-
Taking the limit when Δqi → 0 on both sides of Eq. (2.26), we get the net electric tric
flux of
flux for a continuous system of charges: a
contin-
uous
system
of
charges
32 2 Gauss’s Law


Nin
Δqi
Φ E = lim (2.27)
Δqi →0
i=1
0

dq Q in
= =
Vin 0 0

where Vin is the volume of charges inside the Gaussian surface, Q in is the total charge
inside the closed surface, and Nin is the number of macroscopically elementary
charges inside the closed surface.

2.3 Applications of Gauss’s Law to Insulators

In the following, we are summarizing some tips for solving problems using Gauss’s
law. First, Gauss’s law is useful in determining electric fields when the charge dis-
tribution is characterized by a high degree of symmetry. We should pay attention
to ways of choosing the Gaussian surface over which the surface integral given by
either Eq. (2.21) or Eq. (2.27) can be simplified and the electric field is determined.
In choosing the surface, we should always take advantage of the symmetry of the
charge distribution so that we can remove E from the integral and solve it. Using this
calculation, we can determine a surface that satisfies one or more of the following
conditions:
1. The value of the electric field can be argued by symmetry to be constant over the
surface.
2. The dot product can be expressed as a simple algebraic product Ed A because E
and dA are parallel.
3. The dot product is zero because E and dA are perpendicular.
4. The field is zero over the surface.

2.4 Conductors in Electrostatic Equilibrium

A good electrical conductor contains free electrons (with charge −e) moving inside
the material. When there is no net motion of charge within a conductor, the conductor
is in an electrostatic equilibrium, having the following properties:
1. There is no electric field inside the conductor; that is, the electric field is zero.
2. Any excess charge in an isolated conductor resides on its surface.
3. The electric field just outside a charged conductor is perpendicular to the surface
of the conductor and has a magnitude σ/0 , where σ is the surface charge density
at that point.
4. For an irregularly shaped conductor, the surface charge density is greatest at
locations where the radius of curvature of the surface is the smallest.
2.4 Conductors in Electrostatic Equilibrium 33

Fig. 2.10 A conductor in a

uniform electric field E

2.4.1 Property 1

Consider a conducting slab placed in an external field E, as shown in Fig. 2.10. If the
field were not zero inside the conductor, then free charges in the conductor would
accelerate under the action of the field. This motion of electrons would mean that the
conductor is not in electrostatic equilibrium, which is not true. Thus, a zero electric
field inside the conductor is a necessary condition for the existence of electrostatic
equilibrium.
In the absence of external field, free electrons distribute uniformly throughout
the conductor and are free to move about the conductor. On the other hand, if we
apply an external E, as shown in Fig. 2.10, the free electrons drift to the left. Thus,
a plane of negative charge will be seen on the left surface. As the electrons move to
the left, a plane of positive charge creates on the right surface. These charges create
an additional electric field inside the conductor that opposes the external field. As
the electrons move, the surface charge density increases until the magnitude of the
internal field is equal to that of the external field, and hence there is a net field of zero
inside the conductor. The time it takes for an excellent conductor to reach equilibrium
is of the order of 10−16 s, which for most purposes, can be considered instantaneous.

2.4.2 Property 2

Let’s apply the Gauss’s law to a conductor of arbitrary shape. First, the Gaussian
surface is a closed surface inside the conductor, as indicated in Fig. 2.11. Then, using
the Gauss’s law, we write

Φ E = E · dA (2.28)
S
= 0 · dA = 0
S
Q in
=
0
34 2 Gauss’s Law

Fig. 2.11 A Gaussian

surface inside the conductor

Combining Eqs. (2.27) and (2.28), we get Q in = 0. Because the net charge inside
the arbitrary Gaussian surface is zero, any net charge on the conductor must reside
on its surface.

2.4.3 Property 3

First, a Gaussian surface in the shape of a small cylinder is drawn with end faces
parallel to the surface of the conductor, as indicated in Fig. 2.12. It can be seen
that a part of the cylinder is outside the conductor, and only a portion is inside the
conductor. Furthermore, the field vector is perpendicular to the conductor’s surface
from the condition of electrostatic equilibrium. If E had a component parallel to the
conductor’s surface, the free charges would move along the surface, and therefore,
the conductor would not be in equilibrium.
Therefore, we have
1. For the curved part of the cylindrical Gaussian surface there is no flux through
this part of the Gaussian surface because E is parallel to the surface.
2. There is no flux through the flat face of the cylinder inside the conductor because
here E = 0.
3. The net flux through the Gaussian surface is only through the flat face outside the
conductor, where the field is perpendicular to the Gaussian surface.

Fig. 2.12 A cylinder

Gaussian surface
perpendicular to the
conductor
2.4 Conductors in Electrostatic Equilibrium 35

Applying Gauss’s law to this surface, we obtain

ΦE = E ·ndA (2.29)
 S

= E dA = EA
Base

Using Eqs. (2.27) and (2.29), we obtain

Q in σA
EA = = (2.30)
0 0

where σ is the surface charge density such that Q in = σ A. Therefore, we get

σ
E= (2.31)
0

2.4.4 Property 4

Consider an irregularly shaped conductor. Then, the surface charge density is highest
at locations where the radius of curvature of the surface is smallest. To demonstrate
that, suppose we partition irregular surface into small elements d Ai of equal angle
at the center of the conductor.
Then, properties (1) and (3) require that σi d Ai = constant, and since d Ai
depends on the radius of curvature, the smaller the radius of curvature, the smaller
the d Ai , thus property (4) follows.

2.5 Exercises

Exercise 2.1 Determine the electric flux through a sphere that has a radius of
1.00 m and carries a charge of +1.00 µC at its center.

Solution 2.1 Magnitude of electric field at distance r = 1.00 m from the charge is
q
E = ke (2.32)
r2
1.00 × 10−6 C
= (8.99 × 109 N · m2 /C2 )
(1.00 m)2
= 8.99 × 103 N/C
36 2 Gauss’s Law

The direction of E is radially and outward from the charge and perpendicular to the
surface of sphere; thus, the electric flux penetrating the surface of sphere is

ΦE = E A (2.33)
= (8.99 × 10 N/C)(4πr )
3 2

= 1.13 × 105 N · m2 /C

Exercise 2.2 Find the electric flux through a sphere that has a radius of 0.50
m and carries a charge of +1.00 µC at its center.

Solution 2.2 Magnitude of electric field at distance r = 0.50 m from the charge is
q
E = ke (2.34)
r2
1.00 × 10−6 C
= (8.99 × 109 N · m2 /C2 )
(0.50 m)2
= 3.60 × 104 N/C

The direction of E is radially and outward from the charge and perpendicular to the
surface of sphere; thus, the electric flux penetrating the surface of sphere is

ΦE = E A (2.35)
= (3.60 × 10 N/C)(4πr )
4 2

= 1.13 × 105 N · m2 /C

Exercise 2.3 Consider a uniform electric field E along the x-axis. What is the
net electric flux through the surface of a cube of edges a, oriented as shown in
Fig. 2.13?

Solution 2.3 We can distinguish the flux through six faces of the cube: two along
x-axis, two along y-axis, and two along z-axis. The total electric flux is
  
ΦE = E · dA + E · dA + E · dA (2.36)
1 2 3
+ E · dA + E · dA + E · dA
4 5 6
2.5 Exercises 37

Fig. 2.13 A uniform electric

field E oriented in the x
direction passing through the
surface of a cube of edges a

Since E is along x-axis

 
E · dA = E · dA (2.37)
1
4
= E · dA
5
= E · dA = 0
6

The total electric flux is

 
ΦE = E · dA + E · dA (2.38)
3  2

= E · n1 d A + E · n2 d A
3 2

Since E is constant
 
ΦE = − Ed A + Ed A = −Ea 2 + Ea 2 = 0 (2.39)
3 2

Exercise 2.4 Consider a spherical Gaussian surface enclosing a point charge

q. Determine what happened to the net flux (a) if the charge is tripled; (b) if
the radius of the sphere is doubled; (c) if the surface is changed to a cube; and
(d) if the charge is moved to another location inside the surface.

Solution 2.4 1. The flux through the surface is tripled because the flux is propor-
tional to the amount of charge inside the surface.
38 2 Gauss’s Law

2. The flux does not change because electric field lines from the charge pass through
the sphere, regardless of its radius.
3. The flux does not change when the shape of the Gaussian surface changes because
electric field lines from the charge pass through the surface, regardless of its shape.
4. The flux does not change when the charge is moved to another location inside
that surface because Gauss’s law refers to the total charge enclosed, regardless of
where the charge is located inside the surface.

Exercise 2.5 Calculate the electric field of an isolated point charge q using
Gauss’s law.

Solution 2.5 Using the spherical symmetry of the problem, we choose a spherical
Gaussian surface of radius r centered on the point charge, as shown in Fig. 2.14.
Since the charge is positive, then the electric field is directed radially outward, and
hence it is normal to the surface at every point.

ΦE = E · dA (2.40)
S
= (E · n)d A = En d A
S S

= E n d A = E n (4πr 2 )
S
Q in q
= =
0 0
or
q
En ≡ E = (2.41)
4π0 r 2
q
= ke 2
r

Fig. 2.14 The Gaussian

surface and an isolated point
charge q
2.5 Exercises 39

Fig. 2.15 Gaussian surfaces

for calculation of the electric
field created by an insulating
solid sphere of radius a has a
uniform volume charge
density ρ and carries a total
positive charge Q

Exercise 2.6 Consider an insulating solid sphere of radius a, which has a uni-
form volume charge density ρ and carries a total positive charge Q. Determine
the magnitude of the electric field at a point (a) outside the sphere, and (b)
inside the sphere.

Solution 2.6 (a) Because the charge distribution is spherically symmetric, we again
select a spherical Gaussian surface of radius r , concentric with the sphere (see
Fig. 2.15), and

ΦE = E · dA (2.42)
S
= (E · n)d A
S

= En d A = En dA
S S
Q in
= E n (4πr 2 ) =
0
Q
=
0
or
Q
En ≡ E = (2.43)
4π0 r 2
Q
= ke
r2
(b) We select a spherical Gaussian surface with r < a, concentric with the insu-
lated sphere, as shown in Fig. 2.15. Let us denote the volume of this smaller sphere
by V0 . To calculate Q in , we use
40 2 Gauss’s Law

4
Q in = ρV0 = ρ πr 3 (2.44)
3
Q 4 3
= πr
4 33
πa
3
r3
= Q 3 (< Q)
a
Using the Gauss’s law,

ΦE = E · dA (2.45)
S
= (E · n)d A
S
= En d A = En dA
S S
Q in
= E n (4πr 2 ) =
0
1 r3
= Q
0 a 3

Or

Q r3
En ≡ E = (2.46)
4πr 2 0 a 3
Qr
= ke 3
a

Exercise 2.7 Suppose a thin spherical shell of radius a has a total charge Q
distributed uniformly over its surface (see Fig. 2.16). What is the electric field
at points (a) outside and (b) inside the shell?

Solution 2.7 (a) For finding E outside the spherical shell, we select a spherical
Gaussian surface of radius r > a, concentric with the spherical shell, as shown in
Fig. 2.16, and
2.5 Exercises 41

Fig. 2.16 A thin spherical

shell of radius a with a total
charge Q distributed
uniformly over its surface,
and the Gaussian surfaces

ΦE = E · dA (2.47)
S
= (E · n)d A = En d A
S S
Q in
= E n d A = E n (4πr 2 ) =
S 0
Q
=
0
or
Q
En ≡ E = (2.48)
4π0 r 2
Q
= ke
r2
(b) For finding E inside the spherical shell, we select a spherical Gaussian surface
of radius r (r < a), concentric with the spherical shell (see also Fig. 2.16), and

ΦE = E · dA (2.49)
S
= (E · n)d A = En d A
S S

= En d A
S
= E n (4πr 2 )
Q in
= =0
0
or
En ≡ E = 0 (2.50)
42 2 Gauss’s Law

Fig. 2.17 A line of positive

charge of infinite length and
constant charge per unit
length λ. Gaussian surface is
a cylinder of length  and a
radius of the base r , where
dA is a small surface
element vector

Exercise 2.8 What is the electric field at a distance r from a line of positive
charge of infinite length and constant charge per unit length λ (see Fig. 2.17)?

Solution 2.8 Because of the symmetry of the charge distribution, the vector E is
perpendicular to the line charge and directed outward, as shown in Fig. 2.17. Based
on symmetry of the charge distribution, a cylindrical Gaussian surface of radius r
and length  can be selected that is coaxial with the line charge. From Gauss’s law:

ΦE = E · dA (2.51)
S
= (E · n)d A
S
= En d A
S

= E n d A = E n (2πr )
S
Q in λ
= =
0 0
or
λ
En ≡ E = (2.52)
2π0 r 
λ
= 2ke
r

Exercise 2.9 Determine the electric field of a nonconducting, infinite plane of

positive charge with uniform surface charge density σ, as shown in Fig. 2.18.
2.5 Exercises 43

Fig. 2.18 The infinite plane

of positive charge with
uniform surface charge
density σ and a Gaussian
cylinder surface

Solution 2.9 By symmetry, E must be perpendicular to the plane and must have the
same magnitude at all points equidistant from the plane. The fact that the direction
of E is away from positive charges indicates that the direction of E on one side of
the plane must be opposite to its direction on the other side, as shown in the figure.
We chose a Gaussian surface, which because of the symmetry is a small cylinder
with an axis perpendicular to the plane and with both ends have an area A and are
equidistant from the plane. From Gauss’s law,

ΦE = E · dA (2.53)
A 
= (E · n)d A + (E · n)d A
Base1 Base2
Q in σA
= 2E A = =
0 0
or σ
E= (2.54)
20

Exercise 2.10 Let us consider a solid conducting sphere of radius a carrying

a net positive charge 2Q (see Fig. 2.19). Suppose a conducting spherical shell
of inner radius b and outer radius c is concentric with the solid sphere and it
carries a net negative charge −Q. Find the electric field in the regions labeled
1, 2, 3, and 4. What is the charge distribution on the shell when the entire
system is in electrostatic equilibrium?

Solution 2.10 The charge distributions on the sphere and shell have spherical sym-
metry around their common center. To determine the electric field at different dis-
44 2 Gauss’s Law

Fig. 2.19 The charge

distributions on both the
sphere and the shell are
characterized by spherical
symmetry around their
common center

tances r from the center, we construct a spherical Gaussian surface for each of the
four regions of interest.
1. Region 1: From Gauss’s law,

ΦE = E · dA (2.55)
S
Q in
=
0
= 0 (Q in = 0)

This implies that E = 0.

2. Region 2: From Gauss’s law,

ΦE = E · dA (2.56)
S
= Ed A = E(4πr 2 )
S
Q in
=
0
2Q
=
0

Thus, we obtain the electric field as

1 2Q
E= (2.57)
4π0 r 2
2Q
= ke 2
r
2.5 Exercises 45

3. Region 3: The electric field must be zero because the spherical shell is also a
conductor in equilibrium. From Gauss’s law,

ΦE = E · dA (2.58)
S
=0
Q in
=
0

Thus, we get Q in = 0.
Therefore, the charge on the inner surface of the spherical shell is −2Q to cancel
the charge +2Q on the solid sphere.
4. Region 4: From Gauss’s law

ΦE = E · dA (2.59)
S
Q in
= Ed A = E(4πr 2 ) =
S 0
2Q − Q
=
0

We obtain the electric field as

1 Q
E= (2.60)
4π0 r 2
Q
= ke 2
r

Reference

Holliday D, Resnick R, Walker J (2011) Fundamentals of physics. John Wiley and Sons, New York
Chapter 3
Electrostatic Potential

This chapter aims to introduce electrostatic potential and the

methods used to calculate it.

In this chapter, we will introduce electrostatic potential and discuss the methods
used to calculate the electrostatic potential. Furthermore, we present the potential
difference and equipotential surfaces. Moreover, we will discuss the calculation of
the electrostatic potential for a single point charge and system of charges. As extra
reading material, the reader can also consider other literature (Holliday et al. 2011;
Jackson 1999; Griffiths 1999).

3.1 Electrostatic Potential Energy

Suppose a test charge q0 is placed in the electric field E of some other charged
object. Then, electric force on q0 is F = q0 E. If E is created by a system of charges
q1 , . . . , q N , then from superposition principle


N
ER = Ei (3.1)
i=1

The electric force acting on q0 is


N
F= q0 Ei = q0 E R (3.2)
i=1

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 47

H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_3
 48 3 Electrostatic Potential

Therefore, if the test charge q0 moves in the electric field E, the electrostatic forces
(see also Eq. (3.2)) do work on q0 .
Suppose that q0 moves in the field E by some external agent, then the work done by
electric field is negative of the work done by external agent. Let ds be an infinitesimal
displacement in the electric field, then work done by the field on test charge q0 is
calculated as
dWe = F · ds = q0 E · ds = −dU (3.3)
Elec-
tro-
static
In Eq. (3.3), −dU is the decrease of the potential energy of charge-field system.
poten- Therefore,
tial
energy dU = −dWe = −q0 E · ds (3.4)

That is, the electric field’s work E decreases electrostatic potential energy of the
charge moving in the field.

3.2 Electric Potential

For a finite displacement of charge from A to some B, the change in potential energy
of the system charge-field ΔU is
 B  B
ΔU = U B − U A = dU = −q0 E · ds (3.5)
A A

The integral in Eq. (3.5) is called path integral or line integral, and it is performed
along the path that q0 follows as it moves from A to B. Since electric force is
conservative force, then the integral does not depend on the path taken from A to B.
U
Elec- The quantity is independent of q0 , but it depends only on E.
tric q0
poten- U
tial By definition, the ratio is called electric potential φ:
q0

U
φ= (3.6)
q0

Elec- Equation (3.6) implies that electric potential, φ, is a scalar quantity.

tric
poten- Potential difference is the difference of electric potential between two points A
tial and B:
differ-
ence
3.2 Electric Potential 49

Δφ = φ B − φ A (3.7)
ΔU
=
q0
 B
=− E · ds
A

Note that potential difference, Δφ, is different physical quantity than change in
potential energy, ΔU :
ΔU = q0 Δφ (3.8)

Electric potential is a scalar quantity characterizing the electric field. It is

independent of the test charge placed in the field. On the other hand, the
potential energy refers to a charge-field system, and hence it is not independent
of the test charge placed in the field.
Moreover, the change in potential energy, ΔU , of a test charge q0 is minus
the work done by the electric field on the charge (or it is equal to the work done
on the charge by the external agent). The potential difference, Δφ = φ B − φ A ,
is equal to the work per unit charge that an external agent does to move a test
charge from A to B without accelerating it (or without changing the kinetic
energy of the test charge).

Similar to potential energy, only differences in electric potential are meaningful.

Often, it is avoided to work with potential differences; therefore, by convention,
the value of the electric potential is considered zero at some convenient point in
an electric field. Practically, it is established that the electric potential is zero at a
location that is at an infinite distance from the system of charges producing the field.
Then, the electric potential at an arbitrary point P in an electric field is equal to
the work needed to move a positive unit test charge from a point at the infinity to that
point. Thus, the electric potential at any point P is
 P
φ P = φ P − φ∞ = − E · ds (3.9)
∞

From Eq. (3.9), SI units of electric potential is N · m/C = J/C, which is defined as
Volt (V):
J
1V ≡ 1 (3.10)
C
SI units for electric field could also be
N V
1 =1 (3.11)
C m
 50 3 Electrostatic Potential

Elec- Often, in atomic and nuclear physics, the units of energy are given in electron
tron
volt volt (eV). By definition, 1 eV is the energy an electron (or proton) gains or loses by
moving it through a potential difference of 1 V:

1 eV = 1.60 × 10−19 C · (1 V) = 1.60 × 10−19 J (3.12)

where the charge of 1.60 × 10−19 C is a fundamental charge equal to the magnitude
of the charge of an electron.

3.3 Potential Difference in a Uniform Electric Field

Consider the motion of a point charge q in a uniform electric field E, as shown in

Fig. 3.1. As a result of interaction with the electric field, the charge moves from A to
B, assuming that it is a positive test charge.
First, we express the electric potential difference between points A and B:
 B
Δφ = φ B − φ A = − E · ds (3.13)
A

We also define E = −Ej, where j is a unit vector along y-axis. Moreover, ds = −dsj.
Then, Eq. (3.14) can also be written as
 B  B
Δφ = − E(j · j)ds = −E ds = −Ed (3.14)
A A

where (−) sign indicates that φ B < φ A . Therefore, the electric field lines always
point in the direction of decreasing electric potential.
Since the electric potential φ is a characteristic of the electric field E, for the same
field E as in Fig. 3.1, and an arbitrary test charge q0 (positive or negative) moving

Fig. 3.1 Movement of a

charge q in a uniform
electric field E
3.3 Potential Difference in a Uniform Electric Field 51

Fig. 3.2 Movement of a

charge q in a uniform electric
field E along a diagonal
direction from A to B

from A to B, the change in potential energy of the charge-field system is

ΔU = q0 Δφ = −q0 Ed (3.15)

If q0 > 0, then ΔU < 0; that is, a positive charge loses its potential energy when
it moves in the direction of an electric field, and the field does work on the charge.
Therefore, when released from the rest in an external electric field, a positive charge
particle gains acceleration in the direction of the external electric field.
On the other hand, if q0 < 0, then ΔU > 0; that is, a negative charge increases
its potential energy when it displaces along the electric field direction. Therefore,
if a negative charge releases from rest in the field E, it accelerates in the opposite
direction to the external electric field.
In the following, we consider a more general case. Assume a charged particle
moves freely between any two points in a uniform electric field directed along the
x-axis, as indicated in Fig. 3.2. Let s be the displacement vector between A and B,
then

Δφ = φ B − φ A (3.16)
 B
=− E·s
A
 B
= −E · ds
A
= −E · s

The change in the potential energy of a charge q0 is

ΔU = q0 Δφ = −q0 E · s (3.17)
 52 3 Electrostatic Potential

Fig. 3.3 Movement of a

charge q in a uniform electric
field E along a diagonal
direction from A to B

3.4 Equipotential Surface

Consider again the potential difference between the points A and B, see Fig. 3.3, as

Δφ = φ B − φ A = −E · s = −Es cos θ = −Ed (3.18)

On the other hand, the potential difference between the points C and A is

Δφ = φC − φ A = −Ed (3.19)

Combining Eqs. (3.18) and (3.19), we obtain the following potential difference rela-
tionship:
φ B − φ A = φC − φ A (3.20)

Thus, φ B = φC .
Equipo-
tential By definition, the equipotential surface is called any surface consisting of a con-
surface tinuous distribution of points having the same electric potential.
Using Eq. (3.8), since φ B = φC for any two points B and C in an equipotential
surface
ΔU = q0 (φC − φ B ) = 0 (3.21)

Hence, no work is done in moving a test charge between any two points on an
equipotential surface.
The equipotential surfaces of a uniform electric field consist of a family of planes
that are all perpendicular to the field.

3.5 Electric Potential of a Point Charge

Consider an isolated positive point charge q, as indicated in Fig. 3.4. Note that such
a charge produces an electric field that is directed radially outward from the charge,
as discussed in Chap. 1:
3.5 Electric Potential of a Point Charge 53

Fig. 3.4 Electric potential of

a point charge q

q
E = ke r̂ (3.22)
r2
Consider an arbitrary path from some point A to the point B, and a small displacement
vector along that path ds at the position r relative to the charge q, as shown in Fig. 3.4.
Using Eq. (3.7), the potential difference is
 B
φB − φ A = − E · ds (3.23)
A
 B
q
=− r̂ · ds
ke
A r2
 B
q
=− ke 2 cos θds
A r
 B
q
=− ke 2 dr
r

A

1 1
= ke q −
rB rA

where r A and r B are the distances of A and B relative to q, and r̂ is a unit vector
along the radial direction. In Eq. (3.23), θ is angle between the small displacement
vector ds along the path between A and B and the radial small displacement vector
dr, as in Fig. 3.4.
Equation (3.23) implies that the electric potential of any arbitrary charge q at a
distance r from the charge is given as
q
φ(r ) = ke (3.24)
r
which is a function of the distance r from the charge q.
Equation (3.24) indicates that if q is a positive charge, then the work done to
bring a positively charged object from infinite toward the charge q (where r = 0)
increases to infinite and it is positive. That is, the region around a positive charge
q is analogous to a hill for a positively charged object approaching q. In contrast,
54 3 Electrostatic Potential

for a negative charge q, the work is done to bring a positively charged object from
infinite toward the charge q decreases, and it is negative. Hence, the region around a
negative charge q is analogous to a hole for a positively charged object approaching
q. Furthermore, when a charged object is infinitely distant from another charge, the
electric potential surface is flat and has a value of zero.

3.6 Electric Potential of a System of Point Charges

Let q1 , q2 , . . . , q N be a set of N static discrete charges, positive or negative, as

shown in Fig. 3.5. Based on superposition principle, the electric potential resulting
from those point charges at some point P, with position vector r with respect to the
origin of the reference frame, is


N
qi
φ(r) = ke (3.25)
i=1
| r − ri |

where ri is the position vector of the ith charge with respect to the origin O, as
indicated in Fig. 3.5.
Equation (3.25) indicates that the total potential at any point P of a set of N point
charges is the sum of the potentials due to the individual charges.
In particular, for a system of two charged particles, we denote by φ1 the electric
potential created by the charge q1 at a point P at distance r from the charge q1 , which
is taken to be at the origin O of a coordinate system:
q1
φ1 = k e (3.26)
r
The work done by an external agent to move the second charge q2 from infinity to
P without accelerating it (i.e., the kinetic energy remains constant) is

Fig. 3.5 A set of N static

discrete charges
q1 , q2 , . . . , q N . P is an
observation point at position
vector r from the origin O of
a coordinate system. The
position vectors of the
charges with respect to the
origin O are denoted by ri ,
i = 1, 2, . . . , N , and r − ri
gives the relative position
vector of P with charge i
3.6 Electric Potential of a System of Point Charges 55

W = q 2 φ1 (3.27)
The
Therefore, from Eq. (3.27), the work is equal to the interaction potential energy work

U12 of the particles, when they are separated by a distance r12 :

q1 q2
U12 = ke (3.28)
r12

For two particles with the same sign of charges, U12 is positive, which is in agreement
with what we know from the previous discussion that positive work has to be done
by an external agent on the system to move the two charges near one another. That
is also in agreement with the view that charges with the same sign repel each other
(see Chap. 1). On the other hand, if the charges have opposite sign, U12 is negative;
that is, the external agent does a negative work to move the charges near each other
against the attractive force between charges (in this case, it is the field which does
the positive work).
When the system is composed of more than two charged particles, we can similarly
obtain the total potential energy of the system. For that, to calculate U , we need to
know the pair-wise potential between every pair of charges, then we sum up the terms
algebraically:


N 
N
qi q j
U= ke (3.29)
i=1 j=1>i
ri j

1   qi q j
N N
= ke
2 i=1 j=1=i ri j

where ri j is the distance between the charges i and j:


ri j = (xi − x j )2 + (yi − y j )2 + (z i − z j )2 (3.30)

For a physical interpretation of the expression given by Eq. (3.29), imagine that qi
is fixed at the position, but all other q j=i are at the infinity. In that case, an external
agent has to do work to move any of the charges q j from infinity to its position near qi
qi q j
given as ke . Furthermore, that is independent of the order in which the charges
ri j
are displaced.
Note that Eq. (3.29) can also be written as

1
N
U= qi φi (ri ) (3.31)
2 i=1
56 3 Electrostatic Potential

where φi (ri ) is electric potential created by all other charges at the position where
the ith charge is located:


N
qj
φi (ri ) = ke (3.32)
j=1=i
| ri − r j |

It is interesting to note the behavior of the electric potential φ given by Eq. (3.24)
for r → 0. From Eq. (3.24), we can see that φ(r → 0) → ∞. Now, let us consider
an elementary charged particle in its own electric potential, then the so-called self-
energy or the energy of the field created by that charge is given as

Usel f = qφ(r = 0) = ∞ (3.33)

Equation (3.33) is a contradiction because the self-energy of a static particle is finite

and given as mc2 , based on the theory of relativity. Therefore, that result (Eq. (3.33))
will give a rest mass of a charged particle, for example, consider an electron, infinite.
To solve that contradiction, we will assume that every charged particle is surrounded
by a screening sphere such that within that sphere the classical electrostatic theory
does not apply, and this theory is valid only outside that region. By simple con-
sideration, we can calculate the radius of that sphere, which is also called classical
radius, r0 , which identifies the boundary between the classical (r ≥ r0 ) and a higher
accuracy region called quantum electrodynamics region (r < r0 ). To calculate r0 , we
equalize the self-energy at r = r0 with the rest energy mc2 , where c is the speed of
light in vacuum and m is the rest mass:

q2
qφ(r0 ) = ke = mc2 (3.34)
r0

Solving it for r0 , we obtain the classical radius as

q2
r0 = ke (3.35)
mc2
For example, if the charged particle is an electron, then r0 is
 
N · m2 (−1.6 × 10−19 C)2
r0 = 8.9875 × 109  m
2
2  (3.36)
C
9.1 × 10−31 kg 3 × 108
s
= 0.28 × 10−14 m

Considering that the size of atoms is of the order 10−10 m, then the classical elec-
tromagnetic theory could be a valid method to describe the electron-electron and
electron-nucleus interactions within an atom; however, it may fail to describe the
interactions within nuclei, which have a size of the order 10−15 m.
3.7 Electric Potential of a Continuous Charge Distribution 57

3.7 Electric Potential of a Continuous Charge Distribution

We can partition the volume into macroscopically small charge elements dq, as
shown in Fig. 3.6. Then, we consider the potential due to macroscopically small
charge element dq by treating this element as a point charge:

dq
dφ = ke (3.37)
| r − r |

To obtain the total potential at point P, we integrate to include contributions from

all elements of the charge distribution:

dq
φ = ke (3.38)
| r − r |

If we assume that dq = ρd V , where ρ is the charge density inside the macroscopically

small volume d V = dr , then

dV
φ(r) = ke ρ(r ) (3.39)
V | r − r |

which is a function of r.
For a continuous charge distribution, the potential interaction energy can be writ-
ten as  
ke ρ(r)ρ(r )
U= dr dr (3.40)
2 | r − r |
V V

Using Eq. (3.39), we can write Eq. (3.40) in the following convenient form:

Fig. 3.6 A continuous

charge distribution in a
volume V with density ρ(r )
over a macroscopically small
volume centered at point r .
dq denotes an elementary
charge
58 3 Electrostatic Potential

1
U= ρ(r)φ(r) dr (3.41)
2
V

3.8 Differential form of Electric Potential

For an elementary displacement vector ds, the potential difference is given as the
following:
dφ = −E · ds (3.42)

or
dφ
E=− (3.43)
ds
It can also be written as
E = −∇φ (3.44)

where ∇ is the gradient. Furthermore, we can write in Cartesian coordinates as

 
∂φ(x, y, z) ∂φ(x, y, z) ∂φ(x, y, z)
E=− i+ j+ k (3.45)
∂x ∂y ∂z

where i, j, k are unit vectors along x, y, and z axes, respectively. Therefore,

∂φ(x, y, z)
Ex = − (3.46)
∂x
∂φ(x, y, z)
Ey = −
∂y
∂φ(x, y, z)
Ez = −
∂z

In spherical coordinates, Eq. (3.44) can be written as

 
∂φ(r, θ, ψ) 1 ∂φ(r, θ, ψ) 1 ∂φ(r, θ, ψ)
E = − r̂ + θ̂ + ψ̂ (3.47)
∂r r ∂θ r sin θ ∂ψ

where r̂, θ̂, ψ̂ are unit vectors along r , θ, and ψ directions, respectively. Therefore,

∂φ(r, θ, ψ)
Er = − (3.48)
∂r
1 ∂φ(r, θ, ψ)
Eθ = −
r ∂θ
1 ∂φ(r, θ, ψ)
Eψ = −
r sin θ ∂ψ
3.8 Differential form of Electric Potential 59

For a distribution of charges producing an electric field with a spherical symmetry

(i.e., the electrostatic potential depends only on r , φ = φ(r )), then E = E(r ):

∂φ(r )
Er = − (3.49)
∂r
Eθ = Eψ = 0

3.9 Multipole Expansion

In Eq. (3.39), the charge density ρ(r ) describes the distribution of charges localized
in a finite volume V , and it is zero outside that volume. Often, we are interested for the
potential φ outside the region of the volume V at the distances such that | r | | r |.
Therefore, an expansion of the term 1/ | r − r | around r = 0 in Eq. (3.39) can be
used to identify the value of the potential in different regions.
Using direct Taylor expansion of 1/ | r − r | around r = 0, we write

1 1
= (3.50)
| r − r | r
 
1
+ ∇r  ·r
| r − r | r =0
 
1  ∂2
3
1
+ xi x j + · · ·
2 i, j=1 ∂xi ∂x j | r − r | r =0
1
=
r
r · r
+ 3
r
 
1 
3
∂2 1
+ xi x j
2 i= j=1 ∂xi ∂x j | r − r | r =0
 
1  ∂2
3
1
+ xi2 + · · ·
2 i=1 ∂xi ∂xi | r − r | r =0
1
=
r
r · r
+ 3
r
1  3xi x j 1  3(xi )2 2
3   3
+ x x
i j − x + ···
2 i= j=1 r 5 2 i r5 i
1
=
r
60 3 Electrostatic Potential

r · r
+
r3
1  3xi x j − δi j (r  )2
3  
+ xi x j + · · ·
2 i, j=1 r5

where δi j is the Kronecker’s number: δi j = 1 if i = j, otherwise δi j = 0. Substituting

Eq. (3.50) into Eq. (3.39), we obtain

1
φ(r) = ke ρ(r )dr (3.51)
r V
 
1
+ ke 3 ρ(r )r dr · r
r V
⎛ ⎞

1 ⎝ 1  
3
+ ke 5 ρ(r ) 3xi x j − δi j (r  )2 dr ⎠ xi x j + · · ·
r i, j=1
2 V

We can introduce the following physical quantities:


Q= ρ(r )dr (monopole) (3.52)
 V

p= ρ(r )r dr (dipole)

 V
 
Qi j = ρ(r ) 3xi x j − δi j (r  )2 dr (quadrupole)
V

where the first expression gives the so-called electric “monopole”, or the total charge
in the volume V , the second term defines the electric dipole moment, and the third
term is the trace-less quadrupole moment tensor element. Using the quantities in
Eq. (3.52), we obtain the following convenient expression for the potential:
 
p · r ke 
3
Q xi x j 1
φ(r) = ke + ke 3 + Qi j 5 + Θ (3.53)
r r 2 i, j=1 r r4

where the first term is the zeroth-order approximation, which gives the electrostatic
potential that would have created a point charge equal to the total charge in the
confined free space volume V as it was placed at the origin of the reference frame
(i.e., r = 0); the second term gives the dipole electrostatic potential, which takes
into account polarity of the charge distribution; the third term gives the quadrupole
electrostatic potential. In typical calculations, higher-order terms in Eq. (3.53) (i.e.,
terms of order Θ(1/r 4 )) can be ignored, in particular, when the electrostatic potential
is required at some point far from the origin (or, r r  ).
3.9 Multipole Expansion 61

This formalism can straightforward apply to a system of static discrete charges qi

for i = 1, 2, . . . , N . In that case, the charge density at any point r in space is written
as
N
ρ(r) = qi δ (r − ri ) (3.54)
i=1

where δ is the delta function such that


1, r = ri
δ (r − ri ) = (3.55)
0, r = ri

Using Eq. (3.52), we can calculate the dipole moment as

  
N
  N
p= qi δ r − ri r dr = qi ri (3.56)
V i=1 i=1

Combining Eqs. (3.54) and (3.52), we can also calculate the total charge of system
(or monopole) as
  
N
  N
Q= qi δ r − ri dr = qi (3.57)
V i=1 i=1

which is an expected result. Furthermore, the quadrupole of the discrete system of

static charges is
  
N
      
Qi j = qk δ r − rk 3xi x j − δi j (r  )2 dr (3.58)
V k=1


N
 
= qk 3xki xk j − δi j rk2
k=1

Then, Eq. (3.53) can be used to calculate the electric potential at any point P as in
the case of the continuous charge distribution.
Note that dipole moment depends on the origin of the coordinate system, and
it is independent of that origin only if the system is neutral (see the Exercises).
Furthermore,
3 it is straightforward to show that the quadrupole has a trace equal to
zero: i=1 Q ii = 0 (see also exercises).
We can calculate the electric field using Eq. (3.44). In the following we are eval-
uating the gradient of each term in Eq. (3.53). The first term is rather simple to be
evaluated:  
1 r
∇φ(0) = ke Q∇ = −ke Q 3 (3.59)
r r
62 3 Electrostatic Potential

The gradient of the second term gives

p · r
p · r
p · r
p · r

∇ ke 3 = ke ∇x i + ∇ y j + ∇ z k (3.60)
r r3 r3 r3
 2
r px − 3(p · r)x
= ke i
r5
r 2 p y − 3(p · r)y
+ j
r5

r 2 pz − 3(p · r)z
+ k
r5
(r · r)p − 3(p · r)r
= ke
r5
where i, j, and k are unit vectors along the x-, y-, and z-axis, respectively.
Furthermore, the gradient of the third term is
⎛ ⎞ ⎛ ⎛ ⎞ ⎞
k 3
x x k 3
∂ 3
x x
∇⎝ Qi j 5 ⎠ = ⎝ ⎝ Q i j 5 ⎠ x̂k ⎠
e i j e i j
(3.61)
2 i, j=1 r 2 k=1 ∂xk i, j=1 r
⎛ ⎞
ke ⎝   ∂ xi x j
⎠
3 3
= Qi j x̂k
2 i, j=1 k=1
∂xk r 5
⎛ ⎞
ke ⎝ 
3 3
r (δik x j + xi δ jk ) − 5xi x j xk ⎠
2
= Qi j x̂k
2 i, j=1 k=1
r7

which are all terms of order Θ(1/r 4 ) that are ignored from our discussion. In
Eq. (3.60), x̂k denotes a unit vector along one of the axes (k = x, y, z).
Combining Eqs. (3.44), (3.59), and (3.60), we obtain multipole expansion of the
electric field as follows:
r
E(r) = ke Q (3.62)
r3
3(p · r)r − (r · r)p
+ ke 5
  r
1
+Θ
r4

For convenience, we use the expression of the unit vector r̂ along the direction from
the origin of reference frame toward the point P:
r
r̂ = (3.63)
r
3.9 Multipole Expansion 63

and Eq. (3.62) is written in the following convenient form:

 
Q 3(p · r̂)r̂ − p 1
E(r) = ke 2 r̂ + ke +Θ (3.64)
r r3 r4

In Eq. (3.64), the first term gives electric field created by a point charge placed at the
origin of the reference frame with a charge equal to the charge distributed in all free
space, and the second term gives the electric dipole moment field of a dipole at the
origin of coordinate system. Note that if the dipole has its center some position r0
related to the origin of coordinate system, then Eq. (3.64) can be modified as
 
Q 3(p · r̂)r̂ − p 1
E(r) = ke r̂ + ke +Θ (3.65)
r 2 | r − r0 |3 r4

3.10 Electric Potential of a Charged Conductor

We can show that the surface of a charged conductor in equilibrium is an equipotential

surface. For that, consider two points A and B on the surface of a charged conductor,
as shown in Fig. 3.7. Along a surface path connecting these points, E is always
perpendicular to the displacement ds; that is, E · ds = 0. Therefore,
 B
φB − φ A = − E · ds = 0 (3.66)
A

which implies that φ B = φ A .

Therefore, φ is constant everywhere on the surface of a charged conductor in
equilibrium. That is, the surface of any charged conductor in electrostatic equilibrium
is an equipotential surface. Besides, the electric field is zero inside the conductor,
that is
Er = −dφ/dr = 0 (3.67)

Fig. 3.7 Electric potential

on the surface of a charged
conductor in electrostatic
equilibrium
64 3 Electrostatic Potential

Fig. 3.8 A conducting

sphere of radius R and total
positive charge Q. Two
Gaussian surfaces are drawn
for r > R and r < R

which indicates that the electric potential φ is constant everywhere inside the con-
ductor, including the surface. Therefore, the potential φ inside the conductor in
equilibrium is equal to its value at the surface.
As an example, let us consider a solid metal conducting sphere of radius R and
total positive charge Q (see also Fig. 3.8). Using Gauss’s law with Gaussian surfaces
as indicated in Fig. 3.8, we can calculate the electric flux as

Q
ΦE = E · dA = E4πr 2 = (3.68)
S 0

Equation (3.68) implies that the magnitude of electric field is

 Q
ke ,r ≥R
E= r2 (3.69)
0, r<R

In addition, electric field vector is given as

 Q
ke r̂, r ≥ R
E= r2 (3.70)
0, r<R

Using Eq. (3.14), we can calculate the potential difference between a point B at
the infinity and a point A at the surface of the sphere with radius R as
 ∞
Q Q
φ∞ − φ R = − ke 2
dr = −ke (3.71)
R r R

Therefore, we get
Q
φ R = ke (3.72)
R
If B is an interior point, r < R, then
3.10 Electric Potential of a Charged Conductor 65
 r
φr − φ R = − E · dr (3.73)
 Rr
=− 0 · dr = 0
R

Then, we obtain
Q
φr = ke (3.74)
R
On the other hand, if B is an outside point, r > R, then
 ∞
φ∞ − φr = − E · dr (3.75)
r ∞
Q Q
=− ke 2
dr = −ke
r r r

Therefore,
Q
φr = ke (3.76)
r
For a nonspherical conductor, the surface charge density σ is not uniformly dis-
tributed along the surface. That is, higher surface charge density is obtained at the
regions with small radius of curvature or where the surface is convex. In contrast,
σ is low at the small radius of curvature regions or where the surface is concave.
Furthermore, we know that the electric field just outside the conductor is given as
E n = σ/ 0 . Therefore, the electric field is large near convex points having small radii
of curvature and reaches very high values at sharp points.

3.10.1 Cavity Within a Conductor

Let us consider a cavity of no charge within a conductor of any shape, as shown

in Fig. 3.9. Based on Gauss’s law (Q in = 0!), the electric field inside the cavity is
zero independent of the charge distribution on the outside surface of the conductor.
Furthermore, the field within the cavity is zero, even if the conductor is in an external
electric field because every point inside the conductor is at the same electric potential.
For instance, consider the points A and B inside the conductor and on the surface of
the cavity. Then, for those points:

φ A = φB (3.77)

On the other hand, using Eq. (3.14), we have


φB − φ A = − E · ds (3.78)
66 3 Electrostatic Potential

Fig. 3.9 A conductor of

arbitrary shape containing a
cavity with no charges inside
the cavity


where E is electric field inside cavity. Since φ A = φ B , E · ds = 0, or E = 0.

3.11 Exercises

Exercise 3.1 Assume a battery produces a specified potential difference of

12 V between conductors attached to the battery terminals. It is connected
between two parallel plates (see Fig. 3.10) at a separation of d = 0.30 cm.
Assuming that the electric field between the plates is uniform, what is the
electric field between the plates?

Solution 3.1 The field points from the positive plate (A) to the negative plate (B).
Furthermore, the positive plate is at a higher electric potential than the negative plate.
The potential difference between the plates must be equal to the potential difference
between the battery terminals:

φ B − φ A = −Ed (3.79)

The magnitude of E is

| φB − φ A | 12 V
E= = = 4.0 × 103 V/m (3.80)
d 0.30 × 10−2 m
3.11 Exercises 67

Fig. 3.10 A battery of 12 V

is connected between two
parallel plates at separation
d = 0.30 cm

Exercise 3.2 Consider a proton released at rest in a uniform electric field

of magnitude of 8.0 × 104 V/m and is directed along the positive x-axis, as
indicated in Fig. 3.11. The proton undergoes a displacement of 0.50 m in the
direction of E. What is (a) the electric potential difference of points A and B?
(b) the proton’s potential energy change for this displacement?

Solution 3.2 For proton q0 =| e |= 1.60 × 10−19 C.

(a) Potential difference is

Δφ = −Ed = −(8.0 × 104 V/m)(0.50 m) = −4.0 × 104 V (3.81)

(b) Change in potential energy is

ΔU = q0 Δφ (3.82)
−19
= (1.60 × 10 C)(−4.0 × 10 V)
4

−15
= −6.4 × 10 J

(−) sign indicates that the proton’s potential energy decreases during the displace-
ment along the electric field direction. That is, proton gains kinetic energy and at the
same time loses electric potential energy.
68 3 Electrostatic Potential

Fig. 3.11 A proton

undergoing a displacement
of d in the direction of E

Exercise 3.3 A charge q1 = 2.00 µC is located at the origin, and a charge

q2 = −6.00 µC is located at (0, 3.00) m (see also Fig. 3.12). (a) What is the
total electric potential created by these charges at the point P with coordinates
(4.00, 0) m? (b) What is the potential energy change of a charge of 3.00 µC
when it moves from infinity to point P?

Solution 3.3 (a) Total electric potential at P is

 
q1 q2
φ P = ke + (3.83)
r1 r2
   
N · m2 2.00 × 10−6 C −6.00 × 10−6 C
= 8.99 × 109 +
C2 4.00 m 5.00 m
= −6.29 × 103 V

(b) For the charge at infinity: Ui = 0, then charge is at point P,

U P = q3 φ P = −18.9 × 10−3 J (3.84)

Fig. 3.12 A system of

interacting charges
3.11 Exercises 69

Then the potential energy ΔU = U P − Ui = −18.9 × 10−3 J.

The work done by an external agent to move the charge from point P back to
infinity is
W = −ΔU = 18.9 × 10−3 J (3.85)

which is positive work.

Exercise 3.4 A charge q1 = 2.00 µC is located at the origin, and a charge

q2 = −6.00 µC is located at (0, 3.00) m. Find the total potential energy.

Solution 3.4 Total potential energy is


3 
3
qi q j
U = ke (3.86)
i=1 j>i
ri j
 
q1 q2 q1 q3 q2 q3
= ke + +
r12 r13 r23
= −5.48 × 10−2 J

Exercise 3.5 Consider an electric dipole consisting of two charges of equal

magnitude and opposite sign at a distance 2a from each other (see Fig. 3.13).
The dipole vector is taken parallel to x-axis and it is centered at the origin. (a)
Calculate the electric potential at point P. (b) Calculate φ and E x at a point
far from the dipole.

Fig. 3.13 An electric dipole

70 3 Electrostatic Potential

Solution 3.5 (a) Total electric potential

q1 q2
φ P = ke + ke (3.87)
r1 r2
 
q q
= ke −
x −a x +a
qa
= 2ke 2
x − a2

The electric field

dφ qax
Ex = − = 4ke 2 (3.88)
dx (x − a 2 )2

(b) For x a, we have x 2 − a 2 ≈ x 2 , thus

qa
φ P ≈ 2ke (3.89)
x2
and the electric field
dφ qa
Ex = − ≈ 4ke 3 (3.90)
dx x

Exercise 3.6 (a) Define the expression of the electric potential at point P
located at the central axis perpendicular to a uniformly charged ring of radius a
and total charge Q (see Fig. 3.14). (b) What is the expression for the magnitude
of the electric field at point P?

Solution 3.6 (a) Orient the ring with its plane perpendicular to an x-axis and its
center is at the origin. Then take point P to be√at a distance x from the center of
ring. The charge element dq is at a distance x 2 + a 2 from point P. Hence, we
can express φ as

Fig. 3.14 A uniformly

charged ring of radius a and
total charge Q
3.11 Exercises 71

dq
φ= ke (3.91)
r

dq
= ke √
x2 + a2

1
= ke √ dq
x2 + a2
Q
= ke √
x 2 + a2

(b) From symmetry, we see that along the x-axis E can have only an x component.
Then, from
E = −∇φ (3.92)

we get

dφ
Ex = − (3.93)
dx
d 2
= −ke Q (x + a 2 )−1/2
dx
Qx
= ke 2
(x + a 2 )3/2

Exercise 3.7 Find (a) the electric potential and (b) the magnitude of electric
field along the perpendicular central axis of a uniformly charged disk of radius
a and surface charge density σ (see Fig. 3.15).

Solution 3.7 (a) Orient the disk with its plane perpendicular to an x-axis and its
center is at the origin. Then take point P to be at a distance x from the center of
disk. Take a ring with width dr and radius r , then charge element dq = σ(2πr dr )

Fig. 3.15 A uniformly

charged disk of radius a and
surface charge density σ
72 3 Electrostatic Potential
√
is at a distance x 2 + r 2 from point P. Hence, we can express dφ as in the
previous exercise.
σr dr
dφ = 2πke √ (3.94)
x2 + r2

Then, the total φ is

 a
σr dr
φ= 2πke √ (3.95)
0 x2 + r2
 a
d(x 2 + r 2 )
= πke σ √
0 x2 + r2
  a
= πke σ 2 x 2 + r 2
 0

= 2πke σ x +a −x
2 2

(b) From symmetry, we see that along the x-axis E can have only an x component.
Then, from
E = −∇φ (3.96)

we get

dφ
Ex = − (3.97)
dx  
x
= 2πke σ 1 − √
x 2 + a2

Exercise 3.8 Show that the electric dipole moment of a system of discrete
charges depends on the origin of the coordinate system and it is independent
of that only if the system of charges is neutral.

Solution 3.8 Consider that the origin of the coordinate system is shifted by r0 , then
the position vector of a charge i, ri related to the origin O of the old system can
be related to the position vector ri of that charge in the new coordinate system as:
ri = ri − r0 , see also Fig. 3.16. Using Eq. (3.56), we can write the dipole moment in
the new reference frame as


N
p = qi ri (3.98)
i=1
3.11 Exercises 73

Fig. 3.16 A discrete set of

static charges relative to two
reference frames, which have
a shift on their origins given
by r0


N
= qi (ri − r0 )
i=1


N 
N
= qi ri − r0 qi
i=1 i=1
= p − r0 Q
N
where Q = i=1 qi . That result implies that dipole moment depends on the origin.
However, if the system of charges is neutral, that is, Q = 0, then p = p .

Exercise 3.9 Calculate the electric dipole moment and the quadrupole moment
of the system of four charges with a configuration as shown in Fig. 3.17.
q = 3 µC gives the magnitude of each charge.

Solution 3.9 Using Eq. (3.56), we can write the dipole moment in the given reference
frame as

Fig. 3.17 A configuration of

four charges with q = 3 µC
74 3 Electrostatic Potential


4
p= qi ri (3.99)
i=1

Then, each component of the dipole moment vector is


4
px = qi x i (3.100)
i=1
= (+3.0 µC)(0.0 m) + (−3.0 µC)(0.1 m)
+ (3.0 µC)(0.1 m) + (−3.0 µC)(0.0 m) = 0.0 (C · m)

4
py = qi yi (3.101)
i=1
= (+3.0 µC)(0.0 m) + (−3.0 µC)(0.0 m)
+ (3.0 µC)(0.1 m) + (−3.0 µC)(0.1 m) = 0.0 (C · m)
pz = 0.0 (C · m) (3.102)

Now, we can calculate the components of the quadrupole moment as


N
 
Qxx = qi 3xi xi − ri2 (3.103)
i=1
 
= q −3(0.1)2 + (0.1)2 + 3(0.1)2 − 2(0.1)2 − 3(0.0)2 + (0.1)2
= (3.0 µC) (−0.03 + 0.01 + 0.03 − 0.02 + 0.01) = 0.0 (C · m2 )

N
Qxy = qi 3xi yi (3.104)
i=1
= (3.0 µC)3(0.1)(0.1) = 0.09 × 10−6 (C · m2 )
Q x z = 0.0 (C · m2 ) (3.105)

N
Q yx = qi 3yi xi (3.106)
i=1
= (3.0 µC)3(0.1)(0.1) = 0.09 × 10−6 (C · m2 )

N
 
Q yy = qi 3yi yi − ri2 (3.107)
i=1
 
= q −3(0.0)2 + (0.1)2 + 3(0.1)2 − 2(0.1)2 − 3(0.1)2 + (0.1)2
= (3.0 µC) (0.01 + 0.03 − 0.02 − 0.03 + 0.01) = 0.0 (C · m2 )
Q yz = 0.0 (C · m2 ) (3.108)
Q zz = Q zy = Q zx = 0.0 (C · m ) 2
(3.109)
3.11 Exercises 75

Exercise 3.10 Show that the quadrupole moment tensor has a trace equal to
zero.

Solution 3.10 Using Eq. (3.56), we can write the diagonal of the quadrupole as


N
 
Q ii = qk 3xki xki − rk2 (3.110)
k=1

Then, we calculate the trace of the quadrupole moment as follows:


3
Trace(Q) = Q ii (3.111)
i=1


3 
N
 
= qk 3xki xki − rk2
i=1 k=1


N 
3
 
= qk 3xki xki − rk2
k=1 i=1


N
 
= qk 3rk2 − 3rk2 = 0
k=1

References

Holliday D, Resnick R, Walker J (2011) Fundamentals of physics. John Wiley and Sons, New York
Jackson JD (1999) Classical electrodynamics, 3rd edn. John Wiley and Sons, New York
Griffiths DJ (1999) Introduction to electrodynamics, 3rd edn. Prentice Hall, Hoboken
Chapter 4
Capacitance and Dielectrics

This chapter aims to introduce capacitance and molecular

theory of dielectrics.

In this chapter, we will introduce capacitance and dielectrics. Then, we discuss the
electrostatics of macroscopic media and introduce a molecular theory of dielectrics.
Also, we will introduce electric polarization, and then derive Maxwell’s equations
for an electrostatic field in both free space and dielectric media. As extra reading
material, the reader can also consider other literature (Holliday et al. 2011; Griffiths
1999).

4.1 Capacitance
Capac-
We first start with the definition of a capacitor. For that, consider two charged con- itor

ductors with charges of equal magnitude but of opposite sign, as shown in Fig. 4.1.
Suppose they are combined, forming a so-called capacitor. The conductors are also
called plates. Note that a potential difference Δφ exists between the conductors due
to the presence of two different types of charges in each conductor. Voltage
Moreover, because the unit of potential difference is the volt, a potential difference
is often called a voltage. Here, we denoted it ΔV , which is equal to absolute value
of Δφ, ΔV ≡| Δφ |. Capac-
ity
The capacity of an electric circuit to accumulate electric charge at a particular
value of ΔV is called the capacity. Based on the experimental results, the amount of
charge Q on a capacitor depends linearly on the potential difference ΔV between the
two conductors. Furthermore, the constant of proportionality depends on the shape
and distance between the conductors, as in the following demonstrated for a planar
capacitor.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 77

H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_4
 78 4 Capacitance and Dielectrics

Fig. 4.1 A capacitor formed

by two conductors carrying
charges Q and −Q

We can write this relationship, mathematically, as

Q = CΔV (4.1)

Here, Q is the amount of charge in each capacitor, that is, Q =| +Q |=| −Q |, and
Capac- hence, C > 0.
itance
Therefore, the capacitance C of a capacitor (see also Eq. (4.1)) represents the ratio
of the magnitude of charge on each of conductors to the magnitude of the potential
difference between them:
Q
C≡ (4.2)
ΔV

From the definition (Eq. (4.2)), capacitance is always a positive quantity.

Furthermore, the potential difference ΔV in Eq. (4.2) is a positive quantity
(ΔV > 0). Moreover, the potential difference ΔV increases linearly with the
stored charge Q, and hence the ratio Q/ΔV is constant for a given capacitor
C.
Therefore, capacitance measures the ability of a capacitor to store charge
and electric potential energy, and hence it depends on the shape and distance
of the capacitor’s plates. The capacitance has SI units of coulombs per volt.
The SI unit of capacitance is named farad (F), in honor of Michael Faraday:

C
1F=1 (4.3)
V
4.1 Capacitance 79

The capacitance of electronic device ranges from picofarads to microfarads,

and hence, the capacitance is labeled in “mF” for microfarads, “mmF” for
micromicrofarads, or “pF” for picofarads.

4.2 Calculating Capacitance

To demonstrate the calculation of the capacitance of a capacitor, we consider a

capacitor formed from a pair of parallel plates; each plate is connected to one terminal
of a battery acting as a source of potential difference, as shown in Fig. 4.2. Thus, the
voltage is ΔV .
We assume that the capacitor is initially uncharged, and the battery establishes an
electric field in the connecting wires when the connections are made. The directions
of the electric field lines are explained in Fig. 4.2 (on the left); that is,
 (+)
φ+ − φ A = − E · ds (4.4)
A

where ds is a small displacement vector along the left wire, and

 (−)
φ− − φ B = − E · ds (4.5)
B

where ds is a small displacement vector along the right wire. Initially, when Q = 0
on both plates φ A = φ B = 0, and hence E = 0. Therefore, on the plate connected
to the negative terminal of the battery, the electric field exerts a force on electrons,
which are in the wire just outside this plate; the electrons accelerate to move onto the
plate and hence starts charging the plate negatively. On the other hand, the electric
field exerts forces on electrons of the side (which is closer to the wire) of the plate

Fig. 4.2 A capacitor formed

by two parallel plates
connected to the terminals of
a battery at the potential
difference ΔV
80 4 Capacitance and Dielectrics

connected to the positive terminal of the battery. It accelerates the electrons to move
onto the wire. Hence, leaving on the plate more positive charges in comparison to
negative one (electrons); therefore, this plate is charged positively.
Those accelerations continue until the plates, the wires, and the terminals are all at
the same electric potential. That is also illustrated in Fig. 4.2 (on the right). Once the
equilibrium point is attained, no potential differences exist between the terminals and
the plates on both sides, and hence; as a result, no electric field is present in the wire,
E = 0, and the movement of electrons stops. The right plate carries a negative charge,
−Q, and left plate a positive charge, +Q. When the equilibrium establishes, the
potential difference between the capacitor plates becomes that between the terminals
of the battery.
To calculate the capacitance of a pair of oppositely charged conductors by an
amount of charge Q, we calculate the potential difference ΔV as

ΔV =| φ+ − φ− | (4.6)

Then, the capacitance is evaluated using the expression given by Eq. (4.2). Note that
for simple geometry of the capacitor, these calculations are relatively easy. In the
following, we will discuss spherical shape and planar shapes of conductors.

4.2.1 Spherical Conductors

Let us consider first the capacitance of an isolated spherical conductor of radius R

and charge Q concentric with a hollow sphere of infinite radius, which forms the
second conductor making up the capacitor. We derived the electric potential of the
sphere of radius R as
Q
φ R = ke (4.7)
R
Since at infinity φ∞ = 0 , we obtain

Q Q R
C= = = = 4π0 R (4.8)
ΔV ke Q/R ke

where ΔV =| φ R − φ∞ |. Equation (4.8) indicates that the capacitance of an isolated

charged sphere depends only on its radius R, and it is independent of both the charge
Q on the sphere and potential difference ΔV .
4.2 Calculating Capacitance 81

Fig. 4.3 The electric field

between the plates of a
parallel-plate capacitor is
uniform near the center but
nonuniform near the edges

4.2.2 Parallel-Plate Capacitors

Now, let us consider a capacitor composed of two parallel conductor plates of equal
area A, which are at a distance d, see also Fig. 4.3. One of the plates carries a
charge +Q, and the other −Q. Note that charges of like sign repel one another and
that charges of opposite signs attract one another (see also Chap. 1). As a battery is
charging a capacitor, electrons flow into the negative plate and out of the positive
plate (see Fig. 4.2).
Note that the electric field between the plates of a parallel-plate capacitor is uni-
form near the center but nonuniform near the edges. When the capacitor plates are
large, the accumulated charges can distribute themselves over a substantial area, and
hence the amount of charge stored on each plate Q, for a given potential difference
ΔV , increases as the plate area increases to ensure a constant surface charge density
σ. A simple argument can be used for that: because the electric field just outside
the conductor is perpendicular to the surface of the conductor and with magnitude
E = σ/0 , where E is proportional to constant ΔV , then σ is constant. Thus, we
expect the capacitance C to be proportional to the plate area A.
Above we derived a relationship between the electric field between the plates and
magnitude of potential difference, given as

ΔV
E= (4.9)
d
From Eq. (4.9), we see that when d decreases, E increases, for fixed ΔV . If we
move the plates closer together (that is, d decreases), We also consider the situation
before the charges have moved in response to that change, such that no charges have
moved. Hence, the electric field between the plates is the same but extends over a
shorter distance between plates. That situation corresponds to a new capacitor with
a potential difference between the plates that is different from the terminal voltage
of the battery. Now, across the wires connecting the battery to the capacitor exists a
potential difference (see also Fig. 4.2 for an illustration).
Based on the arguments that we discussed for a situation in Fig. 4.2, that potential
difference creates an electric field in the wires that drives more charges onto the
plates, which in turn increases the potential difference between the plates of the
capacitor. When it becomes equal to the potential difference between the terminals
82 4 Capacitance and Dielectrics

of the battery (Fig. 4.2), the potential difference across the wires falls back to zero.
Then, the flow of charge stops.
This simple experiment shows that decreasing the distance between the plates, d,
yields an increase in the amount of charge stored on the capacitor. The opposite is
also true; if d increases, the charge Q decreases. Therefore, this experiment indicates
that the capacitance is inversely proportional to d.
To verify these physical arguments, we start writing the surface charge density on
either plate as
Q
σ= (4.10)
A
If the plates are very close together (in comparison with their length and width), we
can assume that the electric field is uniform between the plates and is zero elsewhere.
The value of the electric field between the plates is

σ Q
E= = (4.11)
0 0 A

Because the field between the plates is uniform, the magnitude of the potential dif-
ference between the plates equals Ed; therefore,

Qd
ΔV = Ed = (4.12)
0 A

The capacitance is
Q Q
C= = (4.13)
ΔV Qd
0 A

Or
0 A
C= (4.14)
d
Equation (4.14) implies that the capacitance of a capacitor made up of two parallel
plates is proportional to the area A of each plate and inversely proportional to the
plate separation d. That is in agreement with our expectations from the conceptual
argument given above.

4.3 Combination of Capacitors

Often, in an electric circuit, two or more capacitors are combined. If that is the case,
for convenience, we can calculate the equivalent capacitance of certain combinations
using the following methods. Note that there exist circuit symbols for capacitors and
batteries, as well as the color codes used for them, as shown in Fig. 4.4. As we can see,
4.3 Combination of Capacitors 83

Fig. 4.4 The circuit symbols

and color code for capacitors,
batteries, and switches

the symbol for the capacitor manifests the geometry of the most common capacitor
model of a pair of parallel plates. Also, the longer vertical line in a circuit symbol
with appropriate color indicates that the positive polarity of the battery is at the higher
potential.

4.3.1 Parallel Combination

Figure 4.5 presents a combination of two capacitors connected in parallel. Also,

we show a circuit diagram for this combination of capacitors, as often seen in an
electric circuit. Note from Fig. 4.5 that the left plates of the capacitors connect to the
positive terminal of the battery using conducting wires; therefore, those plates, after
equilibrium of the electric potential establishes, are at the same electric potential as
the positive terminal of the battery. For the same reason, the right plate connecting
to the negative terminal of the battery has equal electric potential with the negative
terminal after the equilibrium of the electric potential establishes. As a result, the
potential differences across each capacitor connected in parallel are the same and
equal to the voltage applied to the battery; that is, ΔV1 = ΔV2 = ΔV .
Applying the model described above in Fig. 4.2, when two capacitors are initially
connected in a circuit, as shown in Fig. 4.5, electrons migrate between the wires
and the plates. As a result, the left plates charge positively, and the right plates

Fig. 4.5 Two capacitors

connected in parallel in an
electric circuit
84 4 Capacitance and Dielectrics

negatively. In other words, the internal chemical energy stored in the battery is the
source of that migration; that is, the internal chemical energy of the battery converts
into electric potential energy associated with the surface charges in the plates of
the capacitors at a separation d. During the process of the electrons migration, the
voltage across the capacitors becomes equal to that across the battery terminals and
then charge transfer stops. When that establishes in the circuit, the capacitors load
to their maximum charge capacity.
In the following, we show a few steps to calculate the equivalent capacitance, Ceq ,
of the combinations of C1 and C2 . For that, we denote by Q 1 and Q 2 the maximum
charges on each capacitor, respectively, and by Q the total charge stored by the two
capacitors:
Q = Q1 + Q2 (4.15)

Q is also the charge stored in the capacitor Ceq . The voltages applied across each
capacitor are the same, see also Fig. 4.5, and hence the charges in each capacitor are

Q 1 = C1 ΔV (4.16)
Q 2 = C2 ΔV

Replacing these two capacitors by one equivalent capacitor with a capacitance

Ceq is equivalent to a new electric circuit having one capacitor, as shown in Fig. 4.5;
that is, the effect of capacitor Ceq on the circuit is exactly the same as the effect of
the combination of two capacitors. In other words, the equivalent capacitor stores
the same charge Q when connected to the same battery with a voltage across the
battery terminals ΔV , and hence the voltage across the equivalent capacitor is ΔV .
Therefore, we have
Q = Ceq ΔV (4.17)

Then, we obtain

Ceq ΔV = C1 ΔV + C2 ΔV = (C1 + C2 )ΔV (4.18)

or
Ceq = C1 + C2 (4.19)

Note that the above description can be extended to three or more capacitors connected
in parallel. For example, if N capacitors, C1 , C2 , . . . , C N , connect in parallel in an
electric circuit, we find the equivalent capacitance to be

Ceq = C1 + C2 + · · · + C N (4.20)
4.3 Combination of Capacitors 85

4.3.2 Series Combination

Next, we consider an electric circuit in which two capacitors are combined in series,
as shown in Fig. 4.6. That is known as a series combination of capacitors. In that com-
bination, the left plate of capacitor 1 connects to one of the terminals of a battery (for
example, the positive terminal in Fig. 4.6) and the right plate of capacitor 2 connects
to the other terminal (for example, the negative terminal in Fig. 4.6). Furthermore,
the other two plates, from each capacitor, connect each other via a conducting wire
and to nothing else, as shown in Fig. 4.6. Two capacitors connected that way form an
isolated conductor that is initially uncharged and must continue to have a net charge
zero.
In the following, we will analyze the combination of two capacitors in series.
When the two capacitors are initially uncharged and just connect to a battery in the
circuit, then the electrons transfer from the left plate of C1 and into the right plate of
C2 . That is, during the process, a negative charge (electrons) stores on the right plate
of C2 and the same amount of negative charge leaves the left plate of C2 as electrons
migrating from that plate to the conducting wire leave behind the left plate having
an excess positive charge. Therefore, we can say that the negative charge leaving the
left plate of C2 transfers via the conducting wire and stores on the right plate of C1 .
As a result, the right plates, when the equilibrium establishes, accumulate a charge
−Q, and the left plates a charge +Q. That indicates that the charges on capacitors
connected as in Fig. 4.6 are the same.
It can be seen that the ΔV across the battery terminals is split between two
capacitors:
ΔV = ΔV1 + ΔV2 (4.21)

In Eq. (4.21), ΔV1 and ΔV2 are the potential across C1 and C2 , respectively. In
general, the total potential difference across any number of capacitors connected in
series is the sum of the potential differences across the individual capacitors. Now,
consider an equivalent capacitor, Ceq , with same effect on the circuit as the series
combination of the capacitors. After it is fully charged, the equivalent capacitor must
have a charge of −Q on its right plate and a charge of +Q on its left plate. Using
the definition of capacitance to the equivalent circuit in Fig. 4.6, we have

Fig. 4.6 Two capacitors

connected in series in an
electric circuit
86 4 Capacitance and Dielectrics

Q
ΔV = (4.22)
Ceq

For each capacitor we can write

Q Q
ΔV1 = ; ΔV2 = (4.23)
C1 C2

Then,
Q Q Q
= + (4.24)
Ceq C1 C2

or
1 1 1
= + (4.25)
Ceq C1 C2

When this analysis is applied to three or more capacitors connected in series, for
instance N capacitors C1 , C2 , . . . , C N , the equivalent capacitance is given as

1 1 1 1 1
= + + + ··· + (4.26)
Ceq C1 C2 C3 CN

4.4 Energy Storage in the Electric Field

To transfer an amount of charge from one plate of a capacitor to the other during the
process of charging the capacitor, an external work is done against the electric field.
That work stores in the capacitor in the form of the potential energy. For that, let q
be the charge on the capacitor at some instant during the charging process when the
potential difference across the capacitor is ΔV = q/C. At that instant, one of the
plates is carrying a charge +q and the other −q. To transfer an increment of charge
dq from the plate with charge −q (which is at a lower electric potential) to the plate
carrying charge +q (which is at a higher electric potential) an elementary work is
done against the electric field:
q
dW = ΔV dq = dq (4.27)
C
To calculate the total work required to charge the capacitor from q = 0 to final charge
Q, we integrate Eq. (4.27) as follows:

Q
q 1 Q2
W = dq = (4.28)
C 2 C
0
4.4 Energy Storage in the Electric Field 87

This work done to charge the capacitor stores in the capacitor as an electric potential
energy U . Therefore, U = W . Also, we can express the potential energy U in the
following forms:

1 Q2
U= (4.29)
2 C
1
= QΔV (4.30)
2
1
= C(ΔV )2 (4.31)
2
Note that all expressions given by Eqs. (4.29)–(4.31) are equivalent; that is, they can
all be used to calculate the potential energy stored in a capacitor depending on what
is known. We can consider the energy stored in a capacitor as being stored in the
electric field created between the plates as the capacitor is charged. This description
is reasonable from the viewpoint that the electric field is proportional to the charge Q
stored on a capacitor. For a capacitor of two parallel plates, the potential difference
is related to the electric field through a simple relationship ΔV = Ed. Furthermore,
A
its capacitance is C = 0 . Then, we obtain
d
 
1 A 1
U= 0 (Ed)2 = 0 (Ad)E 2 (4.32)
2 d 2

Since the volume is Ad, then the energy density is given

U 1
uE = = 0 E 2 (4.33)
Ad 2
This expression is generally valid. That is, the energy density in any electric field is
proportional to the square of the magnitude of the electric field at a given point.

4.5 Electrostatics of Macroscopic Media and Dielectrics

Until now, we have introduced the electric potential and electric field in the presence
of other charges or conductors. Therefore, there was no need for distinguishing
between microscopic and macroscopic fields. In fact, in the conductors, the surface
charge densities imply a macroscopic description. However, other media may exist,
where their effect on the electric charge movement is not negligible. In that case, the
electrical response of the medium to the external charges and fields must be taken
into account.
88 4 Capacitance and Dielectrics

4.5.1 Dielectrics

There exist many materials that do not allow electric charges to move freely within
them, or may allow such motion to occur only very slowly. Those materials are used
to block the flow of electrical current, and to form the insulators. For example, they
can create insulating layers between the plates of a capacitor. Those materials are
known as dielectric materials. As an application, the use of the dielectric material for
a capacitor reduces its size for a given capacitance or increases its working voltage.
Note that a dielectric material subject to a high enough electric field becomes a
conductor; that is, the dielectric material experiences a dielectric breakdown. Thus,
there exists a maximum voltage for dielectric capacitors to work. For example, there
is a maximum power that a coaxial cable can adequately function in high-power
applications such as radio transmitters; similarly, for microcircuits there are maxi-
mum voltages, which can be handled.

4.5.2 Comparison Between Dielectric Materials and

Conductors

To know about the differences between dielectric and conducting materials, we can
consider their behavior in electric fields. In particular, we have shown in Fig. 4.7
a conducting and dielectric sheet between the parallel plates in which a potential
difference exists. That is, there are an equal amount of opposite charges on the two
plates.
In the conducting sheet, the conducting electrons are free to move, and they estab-
lish a surface charge which exactly cancels the electric field within the conductor,
as shown in Fig. 4.7. That is, the surface charge density of the plates and conducting
sheet is the same but with opposite sign. On the other hand, the electrons in the
dielectric material are bound to atoms, and the external electric field causes only
a displacement of the electronic configuration of atoms (see Fig. 4.7). However, it
is sufficient to produce some surface charge with density σind (called an induced
charge). We say that the dielectric is polarized. Note that the surface charge is not

Fig. 4.7 Conductors and insulators in an external electric field. σ denotes the charge density of the
plates of the capacitor creating the external field, and σind denotes the induced charge density in
the surface of insulator
4.5 Electrostatics of Macroscopic Media and Dielectrics 89

Fig. 4.8 a Random

orientation of dipoles in
absence of external electric
field, and b orientation of the
dipoles along the external
electric field

able to cancel the external electric field within the sheet; however, it does reduce.
In the following, we will introduce a simplified molecular theory of dielectrics to
understand the behavior of dielectric materials in the presence of an external elec-
tric field.1 A more complicated, but more precise theory, will be introduced in the
following sections, accounting for electric polarization of the ponderable media.2

4.5.3 Molecular Theory of Dielectrics

Consider a dielectric material placed in the electric field between the plates of a
capacitor. The dielectric material is made up of polar molecules. Note that dielectric
material can also be made up of the nonpolar molecules, which can be polarized in the
presence of the external electric fields. In the dielectric made up of polar molecules,
the permanent dipoles of molecules are also called permanent dipoles. For dielectrics
made up of nonpolar molecules, the dipoles created due to polarization are called
induced dipoles.
The permanent dipoles in the dielectric arising from the polar molecules of the
dielectric material are randomly distributed and oriented in the absence of an electric
field, as shown in Fig. 4.8. When an external field E0 applies, for example, created
by charges on the capacitor plates, the forces exerted on the dipoles produce torques,
causing them to align almost in the direction of the field. We can say that the dielectric
is polarized; that is, macroscopic charge separation occurs. The degree of alignment
of molecules with an electric field depends on temperature and the magnitude of
the field. In general, the alignment increases with decreasing temperature and with
increasing electric field. If the molecules of the dielectric are nonpolar, then the
electric field due to the plates produces some charge separation and an induced
dipole moment. These induced dipole moments tend to align with the external field,
and the dielectric is polarized.

1 See Holliday et al. (2011).

2 See Jackson (1999).
90 4 Capacitance and Dielectrics

Fig. 4.9 a Orientation of the

dipoles of a dielectric in the
external electric field; and b
Induced charge density σind
in the surface of dielectric

It is essential to say that we can polarize with an external field, both a dielectric
made up of polar molecules or nonpolar molecules. Consider a dielectric material
placed between the plates of a capacitor so that it is in a uniform electric field E0 , as
shown in Fig. 4.9. Furthermore, the external electric field, E0 , directed to the right,
polarizes the dielectric material. The net effect of E0 is the formation of a positive
surface-induced charge density σind on the right face and an equal negative surface-
induced charge density −σind on the left side.
These surface-induced charges on the dielectric create an induced electric field
Eind in the direction opposite the external field E0 . Therefore, the net electric field
E in the dielectric is
E = E0 + Eind (4.34)

Projecting Eq. (4.34) along the direction of the external field, we obtain

E = E 0 − E ind (4.35)

Moreover, we suppose that the external field is created by two conducting plates
(electrodes), where each carries a surface charge σ per unit area. Also, the surface
charge on the dielectric is σind per unit area. Now, assuming that the electric field is
everywhere uniform and normal to the plates, the field outside the dielectric is equal
to the field created by plates and is given as
σ
E0 = (4.36)
0

Similarly the induced field within the dielectric is

σind
E ind = (4.37)
0

Equation (4.37) can further be written as

4.5 Electrostatics of Macroscopic Media and Dielectrics 91

σ − σind
E= (4.38)
0
σ− σind σ
=
σ 0
= r E 0 (4.39)

where σ− σind
r = (4.40)
σ
is called relative permittivity of the material, which is smaller than one: r < 1. Often,

1
ε= (4.41)
r

is called dielectric constant and it is greater than one: ε > 1, then

E0
E= (4.42)
ε
Equation (4.42) indicates that the magnitude of electric field decreases in presence
of dielectric material. Consider a dielectric completely filling the space between the
plates of a parallel-plate capacitor. From the definition the capacitance C is

Q
C= (4.43)
ΔV
Let E be the magnitude of electric field between the plates of the capacitor, then,
Q
the magnitude of potential difference is ΔV = Ed, and hence, C = . Using the
Ed
E0
relation E = , then
ε
Q
C= (4.44)
E0
d
ε
Q
=ε
E0 d
= εC0

where C0 is the capacitance in absence of dielectric media. Equation (4.44) implies

that the capacitance increases by the factor ε in the presence of the dielectric material
completely filling the region between the plates.
92 4 Capacitance and Dielectrics

4.5.4 Energy Stored in Capacitor

The energy stored in the absence of the dielectric is

Q 20
U0 = (4.45)
2C0

After the battery is removed and the dielectric inserted, the charge on the capacitor
remains the same. Hence, the energy stored in the presence of the dielectric is

Q 20
U= (4.46)
2C
Using the relation C = εC0 , then

Q 20
U= (4.47)
2εC0
or
U0
U= (4.48)
ε
Because ε > 1, the final energy is less than the initial energy (see also Eq. (4.48))
ΔU = U − U0 < 0. We can account for the “missing” energy by noting that the
dielectric, when inserted, gets pulled into the device. An external agent must do
negative work to keep the dielectric from accelerating.
This work is simply the difference

Wa = U − U 0 (4.49)

Alternatively, the positive work done on the external agent by the system is

W = −Wa = U0 − U (4.50)

In summary, a dielectric provides the following advantages:

1. Increase in capacitance;
2. Increase in maximum operating voltage;
3. Possible mechanical support between the plates, which allows the plates to
be close together without touching, thereby decreasing d and increasing C.
4.6 Electric Polarization 93

4.6 Electric Polarization

Consider an electric field applied to a medium made up of a large number of particles,

such as atoms or molecules. The charges bound in molecules will then respond to
the external electric field, and they will follow the perturbed motion to align with
the external field. Thus, the charge density within the molecules will be distorted.
The dipole moments3 of each molecule will be different in comparison to the dipole
moments in the absence of the applied electric field. That is, in the absence of the
external field, the average dipole moments over all molecules of the substance are
zero because the dipole vectors are oriented randomly. In contrast, in the presence
of the applied electric field, the net dipole moment of the substance is different from
zero. Therefore, in the medium, there is an average dipole moment per unit volume,
which is called electric polarization P, given as

P(r) = n i pi  (4.51)
i

In Eq. (4.51), pi is the dipole moment of the molecule type i in the medium, · · · 
denotes the average over a small volume around r, and n i is the average number per
unit volume of the molecule type i at the position r.
If the net charge of the molecule i is Q i , and there is a macroscopic excess or free
charge, the charge density at the macroscopic level is

ρ(r) = n i Q i  + ρ f r ee (4.52)
i

Note that, in general, average charge of a molecule i is zero, Q i  = 0, and hence,

the charge density ρ is equal to the macroscopic excess or free charge, ρ f r ee .
In the following, we will consider the case of a continuous charge distribution, as in
Fig. 3.6 (Chap. 3), and see the medium from a macroscopic viewpoint. The potential
at some point P at the position r from a macroscopic small volume element d V at the
position r is the sum of the potential created by the charge of d V , dq = ρ(r )d V and
the dipole moment of d V is P(r )d V , assuming that there are no higher macroscopic
multipole moment densities:
 
 ρ(r )d V P(r ) · (r − r )d V
dφ(r, r ) = ke + (4.53)
| r − r | | r − r  |3

In Eq. (4.53), P is outside the volume d V . To obtain the electric potential, we integrate
over all space by treating the element volume d V as macroscopically infinitesimal,
and hence d V = dr :

3The electric field of a molecule is characterized by the multipole moments of the molecule.
However, here, we assume that the dominant molecular multipole with the external field is a dipole.
 94 4 Capacitance and Dielectrics
   
ρ(r ) 1
φ(r) = ke dr + P(r ) · ∇r (4.54)
| r − r | | r − r |

In Eq. (4.54), the second term is the dipole contribution to the potential. This term
can be integrated by parts as the following:
     
  1  P(r )
ke dr P(r ) · ∇r = ke dr ∇r · (4.55)
| r − r | | r − r |
  
1
− ke dr ∇r · P(r )
| r − r |
  
 1
= −ke dr ∇r · P(r )
| r − r |

because the first integral is equal to zero. Therefore, Eq. (4.54) can be simplified as

ρ(r ) − ∇r · P(r )
φ(r) = ke dr (4.56)
| r − r |

Equation (4.56) indicates that potential φ(r) is created by the effective charge distri-
bution
ρe f f = ρ(r ) − ∇r · P(r ) (4.57)

where the first term is the macroscopic excess charge density in the dielectric and
the second term is the polarization-charge density in the dielectric medium, which
for a nonuniform polarization can either increase or decrease the net charge within
Elec- a small volume.
tric
dis- We can define the electric displacement vector D as
place-
ment
D = 0 E + P (4.58)

Note that a relationship between the vectors D and E is important for obtaining
a solution for the electric potential or field. If we exclude from the discussion the
ferroelectricity of the macroscopic medium and assume a linear response of the
medium to an external electric field, then the induced polarization P is parallel to E:

P = 0 χe E (4.59)

Here, the proportionality constant is independent of the direction; that is, the medium
is isotropic. Furthermore, we have assumed that the electric field does not become
extremely large. In Eq. (4.59), χe is the electric susceptibility of the medium. Com-
bining Eqs. (4.58) and (4.59), we obtain

D = 0 (1 + χe )E (4.60)
4.6 Electric Polarization 95

Denoting the relative electric permittivity of the medium as

1
r = (4.61)
1 + χe

then, the dielectric constant is defined as

1
= = 1 + χe (4.62)
r

Therefore,
D = 0 E (4.63)

If the dielectric is anisotropic and nonuniform, then  is dependent on the position,

(r). Furthermore, if the isotropic and uniform dielectric medium does not fill all of
the space where the electric field is applied, then we have to consider the boundary
conditions on both D and E at the interface between media. The boundary conditions
include the normal components of D and tangential components of E on each side
of the interface:

(D2 − D1 ) · n = σ (4.64)
(E2 − E1 ) × n = 0

In Eq. (4.64), n is an outward unit normal vector to the surface; that is, it is directed
from region 1 to region 2. Furthermore, σ is the macroscopic surface charge density
at the boundary surface, excluding the polarization charge.

4.7 Set of Maxwell Equations for Electrostatic Field

4.7.1 Maxwell Equations for Free Space Electrostatic Field

First, we introduce the set of Maxwell equations for the electrostatic field in free
space. Using Gauss’s Law (see Chap. 2), we can write the electric flux of electric
field created by continuous charge distribution in a volume V enclosed by the surface
A as 
Q in
E · dA = (4.65)
A 0

Note that in Eq. (4.65) E is the electrostatic field created by all charges in space,
and Q in is the electric charge inside the volume V enclosed by the surface A. The
left-hand side of Eq. (4.65) can be written in the following form using Gauss formula:
96 4 Capacitance and Dielectrics
 
E · dA = ∇ · E dV (4.66)
A V

where V is the volume enclosed by the surface A. In addition, the right-hand side of
Eq. (4.65) can be written as 
Q in ρ(r)
= dV (4.67)
0 V 0

Combining Eqs. (4.65), (4.66) and (4.67), we get

 
ρ(r)
∇ · E dV = dV (4.68)
V V 0

where ∇ · E is the divergence of the vector E, which produces a scalar.

Comparing both sides of Eq. (4.68), we obtain the first Maxwell equation in free
space:
ρ(r)
∇ · E(r) = (4.69)
0

where both E and ρ can be functions of the position r.

Using the expression of the electrostatic potential difference in free space,
Eq. (4.10) (Chap. 3), we have
 B
Δφ = − E · ds (4.70)
A

where A and B are two points in free space, and ds is an infinitesimal displacement
along the curve joining points A and B. If we consider a closed path, that is, A = B,
then Δφ = φ B − φ A = φ A − φ A = 0, and hence

E · ds = 0 (4.71)
L

where L is a closed path. Using Stokes’ formula, we can write

 
E · ds = (∇ × E) · dA (4.72)
L A

where A is the an arbitrary surface enclosed by the path L. Combining Eqs. (4.71)
and (4.72), we obtain 
(∇ × E) · dA = 0 (4.73)
A

Equation (4.73) implies that

∇ ×E=0 (4.74)

Equation (4.74) is the second Maxwell equation of the electrostatic field in free space.
4.7 Set of Maxwell Equations for Electrostatic Field 97

4.7.2 Maxwell Equations for Dielectric Media Electrostatic

Field

We mentioned that in the dielectric medium, an average over macroscopically small

volumes, which are microscopically large, is necessary to obtain the Maxwell equa-
tions of the macroscopic phenomena.
The first observation is that Eq. (4.74) holds microscopically, that is

∇ × Emicr o = 0 (4.75)

When averaging is made of the homogeneous Eq. (4.75), we obtain

∇ ×E=0 (4.76)

Equation (4.76) indicates that Eq. (4.74) holds for the averaged macroscopic electric
field E.
Using Eq. (4.57) for the effective charge density in the medium, Eq. (4.69)
becomes
ρ(r) − ∇ · P(r)
∇ · E(r) = (4.77)
0

Rearranging Eq. (4.77), we get

∇ · (0 E(r) + P(r)) = ρ(r) (4.78)

Using the definition of the electric displacement vector given by Eq. (4.58), we write
Eq. (4.78) as
∇ · D(r) = ρ(r) (4.79)

Note that Eqs. (4.76) and (4.79) are the macroscopic Maxwell equations in the
dielectric medium, which are the counterparts of Eqs. (4.69) and (4.74).

4.8 Potential Energy of Electrostatic Field

Often, it is practical to interpret the electrostatic potential energy that emphasizes

the interactions between the charges of a system as the energy stored in the electric
field surrounding the charges. In that way, we emphasize the electric field instead of
electric potential.
For that, we can use Eq. (3.41) (Chap. 3), and the first Maxwell’s equation in the
free space as ρ = 0 (∇ · E), then we write:
98 4 Capacitance and Dielectrics

1
U= 0 (∇ · E)φ(r) dr (4.80)
2
V

Furthermore, using Eq. (3.44) (Chap. 3), we obtain


0
U =− (∇ 2 φ(r))φ(r) dr (4.81)
2
V

If we integrate by parts in Eq. (4.81), we get

 
0 0
U= (∇φ(r))2 dr = | E |2 dr (4.82)
2 2
V V

The integrand in Eq. (4.82) can be identified as the energy density, u:

0
u= | E |2 (4.83)
2
It is worth noting that the form of the right-hand side of Eq. (4.83) implies that u ≥
0; therefore, the total electrostatic potential energy U ≥ 0. However, the electrostatic
potential of the system of two charges discussed in Chap. 3 (see Eq. (3.28)) implies
that when the two charges have opposite sign, then electrostatic potential, U , is
negative. The reason for that contradiction is that the expression of U given by
Eqs. (4.82) and (4.83) includes the self-energy term to the energy density; while
Eq. (3.28), or more general Eq. (3.29), given in Chap. 3, does not.
In the presence of the dielectric medium, ρ = ∇ · D, Eq. (4.80) can be written as

1
U= (∇ · D)φ(r) dr (4.84)
2
V

Integrating Eq. (4.84) by parts, we get


1
U =− D · (∇φ(r)) dr (4.85)
2
V

Using the differential form of electric potential (Eq. (3.44), Chap. 3), we finally get

1
U= D · E dr (4.86)
2
V
4.9 Exercises 99

4.9 Exercises

Exercise 4.1 Consider a two parallel plates capacitor, each with an area A =
2.00 × 10−4 m2 . The distance between the plates is d = 1.00 mm. Determine
the capacitance.

Solution 4.1 From the formula of capacitance:

A
C = 0 (4.87)
d
 2
2.00 × 10−4 m
= (8.85 × 10−12 C2 /N · m2 )
1.00 × 10−3 m
= 1.77 × 10−12 F = 1.77 pF

Exercise 4.2 Consider a solid cylindrical shape conductor with radius a and
charge Q, which is coaxial with a cylindrical shell of negligible thickness,
radius b > a, and opposite charge −Q, as shown in Fig. 4.10. Find the capac-
itance of this cylindrical capacitor if its length is .

Solution 4.2 First, we calculate the difference of potential between the two cylin-
ders:
b
Δφ = φb − φa = − E · ds (4.88)
a

where E is electric field inside a < r < b, which for a cylindrical charge distribution
is

Fig. 4.10 A solid cylindrical

conductor of a radius and
charge Q is coaxial with a
cylindrical shell of negligible
thickness, radius b > a, and
charge −Q
100 4 Capacitance and Dielectrics

λ
Er = 2ke (4.89)
r
where λ = Q/ is linear charge density.
Thus,

b b
dr
φb − φa = − Er dr = −2ke λ (4.90)
r
a a
 
b
= −2ke λ ln
a

From C = Q/ΔV with ΔV =| φb − φa |, we get

Q
C=   (4.91)
b
2ke λ ln
a
Q
=  
Q b
2ke ln
 a

=  
b
2ke ln
a

Exercise 4.3 A spherical capacitor consists of a spherical conducting shell of

radius b and charge −Q concentric with a smaller conducting sphere of radius
a and charge Q (see Fig. 4.11). Find the capacitance of this device.

Solution 4.3 The difference on the potential

b
φb − φa = − E · ds (4.92)
a

The electric field outside a spherically symmetric charge distribution has a radial
symmetry, and it is given by the expression

Q
E = ke r̂ (4.93)
r2
Thus,
4.9 Exercises 101

Fig. 4.11 A spherical

capacitor consists of a
spherical conducting shell of
radius b and charge −Q
concentric with a smaller
conducting sphere of radius
a and charge Q

b
φb − φa = − Er dr (4.94)
a
b  b
dr 1
= −ke Q = k e Q
r2 r a
a
 
1 1
= ke Q −
b a

From C = Q/ΔV with

ΔV =| φb − φa | (4.95)
(b − a)
= ke Q
ab
we get
ab
C= (4.96)
ke (b − a)

In the limit of b → ∞:
ab
C = lim a
(4.97)
b→∞
ke b 1 −
b
a
=
ke
= 4π0 a
102 4 Capacitance and Dielectrics

Exercise 4.4 A capacitor consists of two parallel plates in the form of disks
with radius r = 2 mm at a separation d = 2 mm. Find the capacitance of this
device.

Solution 4.4 From the formula of capacitance:

A
C = 0 (4.98)
d
 2
π(2 × 10−3 )2 m
= (8.85 × 10−12 C2 /N · m2 )
2.00 × 10−3 m
= 55.6 × 10−15 F = 0.0556 pF

Exercise 4.5 A capacitor consists of two parallel plates connected to the ter-
minals of a battery at the voltage ΔV = 10 V. If the capacitor stores a net
charge of Q = 60 µC, calculate its capacitance.

Solution 4.5 From the formula of capacitance:

Q
C= (4.99)
ΔV
we obtain
60 µC
C= = 6 × 10−6 F = 6 µF (4.100)
10 V

Exercise 4.6 Consider a two parallel plates capacitor, each with an area A =
4.00 × 10−4 m2 . The distance between the plates is d = 1.50 mm. Determine
the capacitance if the space between the plates is filled with a dielectric material
of type glass with  = 6.

Solution 4.6 From the formula of capacitance, we find the capacitance in absence
of dielectric material:
4.9 Exercises 103

Fig. 4.12 Two long

cylindrical conductors of
radius r1 and r2 , respectively,
are parallel and separated by
a distance d

Fig. 4.13 A cylindrical

Gaussian surface with radius
r and length  around a long
cylindrical conductors of
radius R and uniform charge
Q

A
C0 = 0 (4.101)
d
 2
4.00 × 10−4 m
= (8.85 × 10−12 C2 /N · m2 )
1.50 × 10−3 m
= 23.6 × 10−13 F = 2.36 pF

In presence of dielectric, we get

C = C0 = 14.16 pF (4.102)

Exercise 4.7 Two long cylindrical conductors of radius r1 and r2 , respectively,

are parallel and separated by a distance d, as shown in Fig. 4.12. Supposing
that r1 , r2  d and the charge on cylinder A is Q and the charge on cylinder
B is −Q, calculate the capacitance per unit length of that parallel-wire line.

Solution 4.7 Assume the length of cylinders is . First, we have to calculate the
electric field created by a cylindrical conductor of radius R and length  with a
uniform charge Q. For that, we can use Gauss’s law with a Gaussian surface chosen
as in Fig. 4.13. Then,
104 4 Capacitance and Dielectrics

Q
E ·ndA = (4.103)
0
A

Because of the symmetry, electric field E is radial, therefore, the basis of the cylinder
give a zero contribution to the net electric flux. We obtain

Q
E2πr  = (4.104)
0
or,
Q
E= (4.105)
2π0 r 

Therefore, the electric field vectors created by the two cylinders are

QA
EA = r̂ AB (4.106)
2π0 r 
QB
EB = − r̂ AB (4.107)
2π0 r 

where Q A = Q and Q B = −Q, and r̂ AB is a unit vector pointing from A to B. The

capacitance of the two conductors of this two-wire line is

Q
C= (4.108)
ΔV AB

where ΔV AB is the difference in potential between the two cylinders:

ΔV AB = φ A − φ B = ΔV AB
A
+ ΔV AB
B
(4.109)

where ΔV ABA
is the voltage drop due to the charge Q A in A and ΔV AB
B
is the voltage
drop due to the charge Q B on conductor B. Then, using the principle of superposition
the voltage drop from conductor B to conductor A due to the charges is sum of the
voltage drops by each charge individually. Therefore,

A r1
QA Q D
ΔV AB
A
=− E A · ds = dr = − ln (4.110)
2π0 r  2π0  r1
B D

Similarly,

A r2
QB Q r2
ΔV AB
B
=− E B · ds = dr = ln (4.111)
2π0 r  2π0  D
B D
4.9 Exercises 105

ds is a small element length vector pointing from B to A.

Then, the net voltage drop is given as
 
Q D r2
ΔV AB = ΔV AB
A
+ ΔV AB
B
=− ln − ln (4.112)
2π0  r1 D

The capacitance is calculated as

Q 2π0 
C= = (4.113)
| ΔV AB | D r2
ln − ln
r1 D

Then, the capacitance per unit length is calculated as

C 2π0
c= = (4.114)
 D r2
ln − ln
r1 D

Exercise 4.8 Consider a two parallel plates capacitor with capacitance C =

2.0 pF connected to a battery of voltage ΔV = 1.2 V. Determine the energy
stored in the capacitor.

Solution 4.8 Using the formula of energy stored in a capacitor

1 1
U= C(ΔV )2 = (2.0 × 10−12 F)(1.2 V)2 = 1.44 × 10−12 J (4.115)
2 2
Knowing that
1 eV = 1.602 × 10−19 J (4.116)

we get
U = 9 MeV (4.117)

Exercise 4.9 The voltage across a capacitor with capacitance C = 1.0 pF is

ΔV = 0.6 V. Determine the how much does energy stored in the capacitor
increase if the voltage increases to ΔV = 1.2 V.

Solution 4.9 Using the formula of energy stored in a capacitor

106 4 Capacitance and Dielectrics

Fig. 4.14 An electric circuit

of four capacitors
C1 = 1.0 pF, C2 = 0.5 pF,
C3 = 1.5 pF and
C4 = 2.0 pF connected to a
battery of ΔV = 2.0 V

1
U= C(ΔV )2 (4.118)
2
For ΔV = 0.6 V, we find

1
U1 = (1.0 × 10−12 F)(0.6 V)2 = 18 × 10−10 J (4.119)
2
For ΔV = 1.2 V , we find

1
U2 = (1.0 × 10−12 F)(1.2 V)2 = 72 × 10−10 J (4.120)
2
Therefore, the increase is
U2
= 4 times (4.121)
U1

Exercise 4.10 Consider the electric circuit of four capacitors C1 = 1.0 pF,
C2 = 0.5 pF, C3 = 1.5 pF and C4 = 2.0 pF connected to a battery of ΔV =
2.0 V, as shown in Fig. 4.14. Calculate the charge stored in each capacitor and
the voltage across capacitors.

Solution 4.10 Since C2 and C3 are connected in parallel, they are equivalent to one
capacitor, C23 :
C23 = C2 + C3 = 2.0 pF (4.122)

Then, C1 , C23 and C4 are connected in series, and hence, they are equivalent to C:

1 1 1 1
= + + (4.123)
C C1 C23 C4
4.9 Exercises 107

Replacing the numerical values, we get

1 1 1 1
= + + = (1.0 + 0.5 + 0.5) pF−1 (4.124)
C 1.0 pF 2.0 pF 2.0 pF

or C = 0.5 pF. Then, the charge stored in C is

Q = CΔV = (0.5 pF)(2.0 V) = 1.0 pC (4.125)

which is also the charge stored in C1 , C23 and C4 . Therefore, the voltage across each
of these capacitors is

Q 1.0 pC
ΔV1 = = = 1.0 V (4.126)
C1 1.0 pF
Q 1.0 pC
ΔV23 = = = 0.5 V (4.127)
C23 2.0 pF
Q 1.0 pC
ΔV4 = = = 0.5 V (4.128)
C4 2.0 pF

To find the charge stored in C2 and C3 , first note that the voltage across C2 and C3
is the same; that is,
ΔV2 = ΔV3 = ΔV23 = 0.5 V (4.129)

Thus, the charges stored in C2 and C3 are calculated as

Q 2 = C2 ΔV2 = (0.5 pF)(0.5 V) = 0.25 pC (4.130)

Q 3 = C3 ΔV3 = (1.5 pF)(0.5 V) = 0.75 pC (4.131)

References

Holliday D, Resnick R, Walker J (2011) Fundamentals of physics. John Wiley and Sons, New York
Griffiths DJ (1999) Introduction to electrodynamics, 3rd edn. Prentice Hall, Hoboken
Jackson JD (1999) Classical electrodynamics, 3rd edn. John Wiley and Sons, New York
Chapter 5
Electric Current

The chapter aims to introduce the electric current and its

theoretical microscopic view, including Ohm’s law.

In this chapter, we will introduce the electric current. We introduced charges at rest,
or electrostatics, so far. In this chapter, we will consider the phenomena associated
with electric charges in motion. We will introduce the electric current, or simply
current, which describes the rate of charge flow through some region of space. Also,
in this chapter, we will introduce resistance and Ohm’s law. As extra reading material,
the reader can also consider other literature (Holliday et al. 2011).

5.1 Electric Current

Consider the motion in a system of electric charges, as presented in Fig. 5.1. A

current will exist, if there is a net flow of charge through a region. To define current,
we consider the charges moving as in Fig. 5.1 and a surface of area A perpendicular
to the direction of motion of the charges. Aver-
age
By definition, the ratio of the amount of charge ΔQ that passes through the surface electric
area A in a time interval Δt is the average current Iav : current

ΔQ
Iav = (5.1)
Δt
which represents the charge that passes through A per unit time. If the charge flow
rate, ΔQ/Δt varies in time, then the current varies in time.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 109
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_5
 110 5 Electric Current

Fig. 5.1 A system of

electric charges in motion.
Charge passing through a
cross-sectional area A

Instan-
ta- Then, the instantaneous current I is define as
neous
electric
ΔQ dQ
current
I = lim = (5.2)
Δt→0 Δt dt

Note that the instantaneous current I is simply called electric current or current.
In the SI units, the current has a unit of the ampere (A):

C
1A=1 (5.3)
s
Equation (5.2) implies that a current of 1 A is equivalent to a charge of 1 C passing
through the surface area in 1 s.

5.1.1 Direction of Electric Current

Since the charges passing through the surface can be positive or negative, or both,
by convention, the direction of the current flow is assigned to the direction of the
positive charge displacement. In electric conductors, the current is due to the motion
of negatively charged electrons; therefore, the direction of the current flow is opposite
the direction of the movement of electrons. On the other hand, for a beam of positively
charged ions (such as protons in an accelerator), the current flows in the same direction
as the motion of the positive ions. Note that there are situations when the current
is due to the flow of both positive and negative charges, such as in the gases and
electrolytes.

5.1.2 Charge Carrier

In Fig. 5.2a, we show a conducting wire in which the two ends connect to form a loop,
all points on the loop are at the same electric potential, for example, for the points A
5.1 Electric Current 111

Fig. 5.2 a A conducting

loop, and b a conducting
wire connected to a battery
with voltage ΔV

and B, φ A = φ B . Hence,
 B the electric field within the conductor and at its surface is
zero (φ B − φ A = − A E · ds); that is, there is no net drift of charge along the wire,
and therefore, there is no current, I = 0. Moreover, the current in a conductor is zero
even when the conductor has some excess charge.
Now, if the ends of the conduction wire connect to the terminals of a battery
with a potential difference ΔV (see also Fig. 5.2b), not all points on the loop are
at the same potential. For example, the potential difference between the points A
and B of the conductor is φ B − φ A = ΔV . The battery applies a potential difference
between the ends A and B of the loop. Hence, an electric field is created within the
wire. Furthermore, that electric field applies forces on the conduction electrons of the
wire, and therefore, they accelerate around the loop, creating a current, I , with a flow
direction as shown in Fig. 5.2b. Usually, we will refer to the moving charge (positive
or negative) as a mobile charge carrier. Therefore, the mobile charge carriers in the
conductor of Fig. 5.2 are electrons, and they move opposite to the direction of the
current I flow. In general, for a metal, the mobile charge carrier is electron.

5.2 Microscopic Model of Current

In the following, we describe a microscopic model of conduction in a conductor to

relate the current to the motion of the charge carriers. In particular, we will consider
the current in a conductor with a cross-sectional area A, as shown in Fig. 5.3. Consider
a section of the conductor with a length Δx. The volume of that section is

ΔV = AΔx = Avd Δt (5.4)

Suppose that n is the volume number density of mobile charge carriers (or the charge
carrier density), then, the total number of carriers in the volume ΔV is

N = n AΔx = n Avd Δt (5.5)

 112 5 Electric Current

Fig. 5.3 Current in a

conductor of cross-sectional
area A

Fig. 5.4 A conductor in an

external electric field in
which the charge carriers are
free electrons

Therefore, the charge ΔQ in this volume is

ΔQ = N q = qn AΔx = qn Avd Δt (5.6)

From Eq. (5.1), the average current in the conductor is

ΔQ
Iav = = qn Avd (5.7)
Δt
Drift
speed By definition, the drift speed represents the average speed of the charge carriers,
denoted as vd . To understand the drift speed, we will consider a conductor, and
hence the charge carriers are free electrons. For an isolated conductor, the potential
difference across it is zero, as described above for Fig. 5.3, thus these electrons move
randomly as the motion of molecules of the gas in a container. If we apply a potential
difference across the conductor utilizing a battery, as also described above in Fig. 5.4,
an electric field sets up in the conductor. That field exerts an electric force on the
electrons, accelerating them in a given direction. That directed movement of electrons
produces a current, as shown in Fig. 5.4. It is important to note that the electrons do
not move in straight lines along the conductor. Indeed, they collide regularly with the
atoms of the conductor, and hence their resultant motion is a complicated movement,
considered here as a spiral motion. However, the collision just slows down the motion,
because the electrons move slowly along the conductor (in a direction opposite to E)
with a drift velocity vd , as shown in Fig. 5.4.
5.2 Microscopic Model of Current 113

The atom-electron collisions in a conductor is considered an effective internal

friction (or drag force) exerted by the atoms on the electrons. The amount of
energy transferred to the atoms of the conductor from the electrons during
collisions increases the vibration energy of the atoms, and hence an increase
in the temperature of the conductor.

5.3 Resistance and Ohm’s Law

Note that the electric field inside a conductor is zero, which is valid only if the
conductor is in static equilibrium. However, when charges move in a conductor, they
produce a current as a result of an electric field, which is maintained by connecting
the conductor to a battery. The charges move because of the electric field, and hence
a non-electrostatic situation exists in the conductor. Cur-
rent
Let I be the current flowing in a conductor of cross-sectional area A. The ratio of density
the current I with cross-sectional surface area A defines the current density J in the
conductor or the current per unit area:

I
J= (5.8)
A
Since the current I = nqvd A, the current density is

J = nqvd (5.9)

Note that Eq. (5.8) implies that J has SI units of A/m2 . Furthermore, Eq. (5.9) is valid
only if the current density is uniform, and only if the surface of the cross-sectional
surface is perpendicular to the direction of the current flow.
The current density vector is defined as

J = nqvd (5.10)

Equation (5.10) indicates that current density is in the direction of charge motion for
positive charge carriers (q > 0) and opposite the direction of motion for negative
charge carriers (q < 0). Therefore, it is similar to the current I ; however, current I
is not a vector quantity but the current density J is a vector.
 114 5 Electric Current

5.3.1 Ohm’s Law

A current density J and an electric field E are established in a conductor if a potential

difference ΔV is applied across the conductor. If a constant potential difference is
maintained by connecting the two ends of the conductor to the terminals of a battery,
Ohm’s then the current is constant.
Law
Ohm’s law states that the current density is linearly proportional to the electric
field:

J = σE (5.11)

In Eq. (5.11), the constant of proportionality σ is known as the conductivity of the

conductor. Note that not all materials follow Eq. (5.11). The conducting materials
that obey Ohm’s law are often called ohmic, and the ratio between the current density
and the electric field is a constant σ that is independent of the electric field producing
the current.
Consider a straight conducting wire with uniform cross-sectional area A and
length , as shown in Fig. 5.5. Suppose a potential difference ΔV = φb − φa > 0 is
maintained across the wire. Then, a uniform electric field and a current are created
in the wire. We can relate the potential difference to the field as the following:

ΔV = E (5.12)

The magnitude of the current density, based on Ohm’s law given by Eq. (5.11), is

ΔV
J = σE = σ (5.13)

Resis- Using Eq. (5.8), we obtain
tance
 

ΔV = I ≡ RI (5.14)
σA

where

R= (5.15)
σA

Fig. 5.5 A segment of

straight wire of uniform
cross-sectional area A and
length 
5.3 Resistance and Ohm’s Law 115

is called resistance.
Thus, the resistance R can be written as

ΔV
R= (5.16)
I
or in a different known form as:
ΔV = R I (5.17)

Equation (5.16) implies that resistance has SI units of volts per ampere, V/A. One
ohm () is defined as one volt per ampere:

1V
1= (5.18)
1A
By definition, the inverse of conductivity is called resistivity ρ: Resis-
tivity

1
ρ= (5.19)
σ
From Eq. (5.19), ρ has the units ohm-meters ( · m). Using this definition and
Eq. (5.15), we can express the resistance for a uniform block of a conducting material
as

R=ρ (5.20)
A
Every ohmic material has a characteristic resistivity ρ that depends on the properties
of the material and on temperature. Note that ρ is an important property in selection
of the conducting materials for applications used in electronic devices.

5.3.2 Classical Model for Electrical Conduction

In the following, we will describe a classical model of electrical conduction in con-

ductors. Note that Paul Drude first proposed that model in 1900. The model explains
Ohm’s law, and it shows that resistivity can be related to the motion of electrons in a
conductor. Drude model has its limitations because it is a classical model; however,
it introduces concepts that can be applied to more sophisticated models.
We will consider a conductor formed by atoms positioned in a regular array and
a set of electrons that can freely move, which are called conduction electrons. If the
atoms are not part of a solid, the conduction electrons are bound to atoms and are not
free to move; however, when atoms condense into a solid, then conduction electrons
become mobile. In there is no external electric field, the conduction electrons move in
random directions in all the conductors with average speeds of the order of 106 m/s,
similar to the motion of molecules in gas inside a container. Often, the conduction
116 5 Electric Current

Fig. 5.6 The random motion

of electrons in a conductor in
absence of electric field, and
drift velocity zero along the
horizontal

electrons in a conductor are assumed to form an electron gas. In the absence of

the electric field, the drift velocity of the free electrons is zero, and hence there is
no current, as presented in Fig. 5.6. In other words, on average, the same number
electrons move in one direction as in the opposite direction, and thus the net charge
flow is zero.
When an electric field is applied, as shown Fig. 5.7, the free electrons drift gently
in the opposite direction of the electric field. Besides, the free electrons still undergo
a random motion, as described in Fig. 5.6. Now, the average drift speed vd is much
smaller (typically 10−4 m/s) than average speed between collisions (typically 106
m/s). Therefore, the electric field E modified the random motion and made the elec-
trons to drift in the opposite direction to the field. It is important to emphasize that
there is slight curvature in the trajectories of the free electrons, as indicated in Fig. 5.7
because of the acceleration of the electrons between collisions. That is because the
electric field applies a force on the free electrons.
In the classical model, we assume that the motion of an electron after a collision is
independent of its motion before the collision. Besides, the excess energy gained by
the electrons in the electric field is transferred to the atoms of the conductor during
5.3 Resistance and Ohm’s Law 117

Fig. 5.7 Motion of electrons

in a conductor in presence of
electric field, E

their collision. The energy transferred to the atoms increases the vibration energy
of atoms, and therefore, the temperature of the conductor increases. Note that the
temperature increase of a conductor because of the resistance can be used efficiently,
such as in electric toasters and other familiar appliances. To derive a mathematical
model, we will consider a free electron of mass m e and charge q (−e) in an electric
field E. The electric field applies a force on the electron:

F = qE (5.21)

Therefore, the electron gains an acceleration:

F qE
a= = (5.22)
me me

The acceleration, as given by Eq. (5.22), occurs for only a short time between colli-
sions, which results only on a small drift velocity for the electron. We denote by t
the time between two collisions, and vi is the initial velocity of the electron, which is
the velocity just after the first collision. Then, the velocity of the electron after time
t (just before the second collision) is

qE
v f = vi + at = vi + t (5.23)
me

Taking the time average of v f over all possible times t and all possible values of
vi :
qE
v̄ f = v̄i + t¯ (5.24)
me

Assuming that the initial velocities are uniformly distributed over all possible direc-
tions, we obtain
v̄i = 0 (5.25)
118 5 Electric Current

In Eq. (5.24), t¯ is the average time interval (τ ) between successive collisions:

τ = t¯ (5.26)

The average value of v̄ f equals the drift velocity:

qE
vd = τ (5.27)
me

Then, the current density is

qE
J = nqvd = nq τ (5.28)
me

and its magnitude:

E
J = nq 2 τ (5.29)
me

In Eq. (5.29), n is the number of charge carriers per unit volume.

Using the expression of Ohm’s law, Eq. (5.12), we obtain the following relation-
ships for conductivity:
nq 2
σ = τ (5.30)
me

and resistivity from Eq. (5.21) becomes

1 me
ρ= = (5.31)
σ nq 2 τ

Equation (5.31) indicates that in this classical model, conductivity and resistivity
do not depend on the strength of the electric field, which is a characteristic of the
conductors obeying Ohm’s law. The average time between collisions τ is related to
the average distance between collisions  and the average speed v̄ as:


τ= (5.32)
v̄

5.3.3 Resistance and Temperature

It is found that the resistivity of a metal varies approximately linearly with tempera-
ture in a limited range of temperatures as follows:

ρ = ρ0 (1 + α(T − T0 )) (5.33)
5.3 Resistance and Ohm’s Law 119

where ρ is the resistivity at some temperature T (in degrees Celsius), ρ0 is the

resistivity at some reference temperature T0 (usually taken to be 20 ◦ C), and α is the
temperature coefficient of resistivity. It is easy to obtain the temperature coefficient
of resistivity as
1 ρ − ρ0 1 Δρ
α= = (5.34)
ρ0 T − T0 ρ0 ΔT

The unit for α is degrees Celsius−1 [(◦ C)−1 ]. Because resistance is proportional to
resistivity, we can write the variation of resistance as

R = R0 (1 + α(T − T0 )) (5.35)

5.4 Superconductors

There exists a class of materials whose resistance decreases to zero below a specific
temperature of Tc , known as the critical temperature. These materials are known as
superconductors. If we would plot the resistance as a function of temperature for a
superconductor, it follows that a superconductor behaves like a standard metal for
T > Tc , and for T ≤ Tc its resistivity suddenly becomes zero. That was discovered by
the Dutch physicist Heike Kamerlingh-Onnes (1853–1926), in 1911, working with
mercury (a superconductor material below 4.2 K). Recently, it has been shown that
the resistivity of superconductors for T < Tc is less than 4 × 10−25  · m; that is,
around 1017 times lower than the resistivity of copper metal, which can practically be
considered zero. There are thousands of superconductors with critical temperatures
that are substantially higher than initially thought possible. Because of low values of
resistivity of superconductors, once a current set up in a superconductor wire, it will
persist without any applied potential difference. Note that, already, steady currents
are observed to persist in superconducting loops for several years with no apparent
decay.

5.5 Electric Energy and Power

In Fig. 5.8, we show a simple circuit consisting of a battery and a resistor. Consider
a positive amount of charge ΔQ moves clockwise around the circuit. That is, the
charge passes from point a through the battery, then resistor and back to point a.
Points a and d are grounded; that is, we take the electric potential at these two points
to be zero. Electric potential energy U increases as the charge moves from point a Elec-
tric
to b through the battery as poten-
tial
energy
ΔU = ΔQΔV = ΔQ(φb − φa ) (5.36)
 120 5 Electric Current

Fig. 5.8 A simple circuit

consisting of a battery whose
terminals are connected to a
resistor

where φa and φb are electric potentials at the points a and b, respectively.

In contrast, as the charge moves through the resistor, from c to d, it loses the
same amount of electric potential energy due to collisions with atoms of the resistor.
That potential energy transfers into internal energy in the resistor. Note that we have
neglected the resistance of the connecting wires; that is, electric potential energy
faces no loss of energy in the paths bc and da. Therefore, when the charge arrives
back at point a, it has lost all the electric potential energy, and hence its potential
Elec- energy is zero.
tric
Power The rate at which the charge ΔQ loses potential energy in going through the
resistor is
ΔU ΔQ
= ΔV = I ΔV (5.37)
Δt Δt
where I is the current in circuit. The charge loses energy equals the electric power:

P = I ΔV (5.38)

Using Ohm’s Law: ΔV = I R, we get

(ΔV )2
P = I2R = (5.39)
R
When I expresses in amperes, ΔV in volts and R in ohms, the SI unit of power is
the watt.
5.6 Electromotive Force 121

5.6 Electromotive Force

A battery is a device that supplies electrical energy. Often, it is called either a source
of electromotive force or emf source. In general, the internal resistance of the battery
is neglected, and the potential difference between points a and b, see Fig. 5.8, is equal
to the emf
of the battery:
ΔV = φb − φa =
(5.40)

Therefore, the current in the circuit, based on Ohm’s law, is

ΔV

I = = (5.41)
R R
Because ΔV =
, the power supplied by the emf source can be expressed as

P = I
(5.42)

5.7 Exercises

Exercise 5.1 Consider 1 mol, or 63.5 g, copper wire in a residential build-

ing. The wire has a cross-sectional area of 3.31 × 10−6 m2 , and a current of
10.0 A, determine the drift speed of the electrons. Assume that each copper
atom contributes one free electron to the current. The density of copper is
8.95 g/cm3 .

Solution 5.1 The volume is

m 63.5 g
V = = = 7.09 cm3 (5.43)
ρ 8.95 g/cm3

Since in each mol of a substance there are N A = 6.02 × 1023 atoms, and since each
atom contributes with one electron, the total number of free electrons is

6.02 × 1023
n= = 8.49 × 1028 electrons/m3 (5.44)
7.09 cm3

Using the equation of current I = nq Avd , with q = 1.6 × 10−19 C being the absolute
value of charge of electron, we get

I
vd = = 2.22 × 10−4 m/s (5.45)
nq A
122 5 Electric Current

Exercise 5.2 Find the resistance of an 10.0 cm long aluminum cylinder

with a cross-sectional area of 2.00 × 10−4 m2 . Compare the calculations
for a cylinder of the same dimensions made of glass having a resistivity of
3.0 × 1010  · m.

Solution 5.2 The resistance for aluminum with ρ = 2.82 × 10−8  · m is calcu-
lated as

R Al = ρ (5.46)
A  
0.100 m
= (2.82 × 10−8  · m)
2.00 × 10−4 m2
= 1.41 × 10−5 

For the glass, we get


Rgl = ρ (5.47)
A  
0.100 m
= (3.0 × 1010  · m)
2.00 × 10−4 m2
= 1.5 × 1013 

Exercise 5.3 Cable television uses a coaxial cable, which consists of two
cylindrical conductors (see Fig. 5.9). The empty space between the conductors
is completely filled with silicon material. Assume that the current leakage
through the silicon is unwanted; that is, the cable is designed to conduct current
along its length. The radii of the inner and outer conductors are a = 0.500 cm
and b = 1.75 cm, respectively. The length of the cable is L = 15.0 cm. What
is the resistance of the silicon between the two conductors?

Solution 5.3 We divide the object whose resistance we are calculating into concen-
tric elements of infinitesimal thickness dr .
Then, equation R = ρ/A is written as

dr
dR = ρ (5.48)
A
d R is the resistance of a small element of silicon that has a thickness dr and surface
area A: A = 2πr L. A current that passes from the inner to outer conductor must
5.7 Exercises 123

Fig. 5.9 A coaxial cable

consists of two cylindrical
conductors with radius a and
b, respectively

pass radially through this concentric small element, and the area through which this
current passes is A. Therefore, we can write the resistance of that hollow silicon
cylinder as
dr
dR = ρ (5.49)
2πr L
Total resistance is   
b
dr ρ b
R= ρ = ln (5.50)
a 2πr L 2π L a

Given ρ = 640  · m for silicon, we obtain

 
640  · m 1.75 cm
R= ln (5.51)
2π 0.150 m 0.500 cm
= 851 

Exercise 5.4 Consider again an 63.5 g (1 mol) copper wire with a cross-
sectional area of 3.31 × 10−6 m2 , carrying a current of 10.0 A. Assume that
each copper atom contributes to the current with one free electron. The density
of copper is 8.95 g/cm3 . What is: (a) the average time between collisions for
electrons? (b) the mean free path for electrons in copper? Assume that the
average speed for free electrons in copper is 1.6 × 106 m/s.

Solution 5.4 (a) The average collision time is

me
τ= (5.52)
nq 2 ρ
124 5 Electric Current

where ρ = 1.7 × 10−8  · m for the cooper and charge carrier density is n = 8.49 ×
1028 electrons/m3 , thus

9.11 × 10−31 kg
τ= (5.53)
(8.49 × 1028 m−3 )(1.6 × 10−19 C)2 (1.7 × 10−8  · m)
= 2.5 × 10−14 s

(b) Mean free path is

 = τ v̄ = (2.5 × 10−14 s)(1.6 × 106 m/s) (5.54)

−8
= 4.0 × 10 m = 40 nm

Knowing that the atomic spacing is about 0.2 nm, that result indicates that the electron
in the wire travels about 200 atomic spacing until the next collision.

Exercise 5.5 A resistance thermometer measures the temperature by measur-

ing the change in resistance of a conductor. It is made from platinum and has
a resistance of 50.0  at 20.0 ◦ C. The resistance increases to 76.8  when
immersed in a vessel containing melting indium. What is the melting point of
the indium?

Solution 5.5 Using R = R0 (1 + α(T − T0 )), we get

R − R0
ΔT = (5.55)
α R0

where T0 = 20.0 ◦ C, R = 76.8  and R0 = 50.0 , then

ΔT = 137 ◦ C (5.56)

Then
T = T0 + ΔT = 157 ◦ C (5.57)

Exercise 5.6 Consider an electric heater where we apply a potential difference

of 120 V to a Nichrome wire with a total resistance of 8.00 . What are the
current in the wire and the power rating of the heater?

Solution 5.6 Using Ohm’s law:

5.7 Exercises 125

ΔV
I = (5.58)
R
120 V
= = 15.0 A
8.00 
The power is

P = I ΔV (5.59)
= (15.0 A)(120.0 V)
= 1800 W = 1.80 kW

Exercise 5.7 (a) How much energy will it take to cook a turkey in an oven
that works at 20.0 A and 240 V for 4 h? (b) If the energy is purchased at an
estimated price of 8.00 cent per kilowatt hour, what is the cost?

Solution 5.7 (a) The power is

P = I ΔV = (20.0 A)(240.0 V) = 4800 W = 4.80 kW (5.60)

Energy is
Energy = Pt = (4.80 kW)(4 h) = 19.2 kWh (5.61)

(b) The cost is

Cost = (19.2 kWh)($0.080/kWh) = $1.54 (5.62)

Exercise 5.8 Find the cost to operate a 100 W light-bulb for 24 h if the power
company charges $0.08/kWh.

Solution 5.8 Energy is

Energy = Pt = (0.10 kW)(24 h) = 2.4 kWh (5.63)

The cost is
Cost = (2.4 kWh)($0.08/kWh) = $0.19 (5.64)
126 5 Electric Current

Exercise 5.9 Electrons that emerge in a particle accelerator have energy of

40.0 MeV (1 MeV = 1.60 × 10−13 J). Assume that the electrons emerge in
pulses at the rate of 250 pulses/s, corresponding to a time between pulses of 4.00
ms. Duration of the pulse is 200 ns, and the electrons in the pulse constitute a
current of 250 mA. The current is zero between pulses. (a) How many electrons
are delivered by the accelerator per pulse? (b) What is the average current per
pulse delivered by the accelerator? (c) What is the maximum power delivered
by the electron beam?

Solution 5.9 (a) We know that

dQ
I = (5.65)
dt
or,
d Q = I dt (5.66)

then

Q pulse = I dt (5.67)

= I Δt = (250 × 10−3 A)(200 × 10−9 s)

= 5.00 × 10−8 C

Dividing this quantity of charge per pulse by the electronic charge gives the number
of electrons per pulse:

5.00 × 10−8 C/pulse

Electrons per pulse = (5.68)
1.60 × 10−19 C/electron
= 3.13 × 1011 electrons/pulse

(b) Average current is given as Iav = ΔQ/Δt. Because the time interval between
pulses is 4.00 ms, and because we know the charge per pulse from part (a), we obtain

Q pulse
Iav = (5.69)
t
5.00 × 10−8 C
=
4.00 × 10−3 s
= 12.5 µA

(c) By definition, power is energy delivered per unit time. Thus, the maximum power
is equal to the energy delivered by a pulse divided by the pulse duration (Fig. 5.10):
5.7 Exercises 127

Fig. 5.10 Train of pulses of

electrons in a particle
accelerator

E
P= (5.70)
Δt
(3.13 × 1011 electrons/pulse)(40.0 MeV/electron)
=
2.00 × 10−7 s/pulse
= (6.26 × 1019 MeV/s)(1.60 × 1013 J/MeV)
= 1.00 × 107 W = 10.0 MW

Exercise 5.10 In a particular cathode ray tube, the measured beam current is
30.0 µA. How many electrons strike the tube screen every 40.0 s?

Solution 5.10 From the definition

Q Ne e
I = = (5.71)
Δt Δt

where e is the magnitude of the electron charge 1.6 × 10−19 C, and Ne number of
electrons, then

I Δt
Ne = (5.72)
e
(30.0 × 10−6 A)(40.0 s)
=
1.6 × 10−19 C
= 75 × 1014
128 5 Electric Current

Exercise 5.11 If the drift velocity of free electrons in a copper wire is 7.84 ×
10−4 m/s, what is the electric field in the conductor?

Solution 5.11 For cooper ρ = 1.7 × 10−8  · m and charge carrier density is n =
8.49 × 1028 electrons/m3 , therefore, the average collision time is:
me
τ= (5.73)
nq 2 ρ
or

9.11 × 10−31 kg
τ= (5.74)
(8.49 × 1028 m−3 )(1.6 × 10−19 C)2 (1.7 × 10−8  · m)
= 2.5 × 10−14 s

Then, using the equation

E
vd = q τ (5.75)
me

we get
m e vd
E= (5.76)
qτ
(9.11 × 10−31 kg)(7.84 × 10−4 m/s)
=
(1.6 × 10−19 C)(2.5 × 10−14 s)
= 0.18 N/C

Exercise 5.12 Use data from the previous exercise to calculate the collision
mean free path of electrons in copper, assuming that the average thermal speed
of conduction electrons is 8.60 × 105 m/s.

Solution 5.12 The collision mean free path is

 = vd τ (5.77)

For cooper ρ = 1.7 × 10−8  · m and charge carrier density is n = 8.49 × 1028
electrons/m3 , therefore, the average collision time is
me
τ= (5.78)
nq 2 ρ
5.7 Exercises 129

or

9.11 × 10−31 kg
τ= (5.79)
(8.49 × 1028 m−3 )(1.6× 10−19 C)2 (1.7 × 10−8  · m)
= 2.5 × 10−14 s

then

 = (8.60 × 105 m/s)(2.5 × 10−14 s) (5.80)

−8
= 2.15 × 10 m
= 21.5 nm

Reference

Holliday D, Resnick R, Walker J (2011) Fundamentals of physics, John Wiley and Sons, New York
Chapter 6
Magnetic Field

This chapter aims to introduce the magnetic field and the motion
of charges in a magnetic field.

In this chapter, we will introduce the magnetic field. We described the interactions
between charged objects in terms of electric fields. An electric field associates with
any stationary or moving electric charge. In addition to an electric field, the region
of space surrounding any moving electric charge also contains a magnetic field. A
magnetic field also encompasses any magnetic substance. As extra reading material,
the reader can also consider other literature (Holliday et al. 2011).

6.1 Magnetic Field

The magnetic field is usually represented by symbol B. To determine the direction

of the magnetic field B at some location the compass needle is used, which points
along B at that location. The magnetic field lines outside a magnet go from north
poles to south poles, as shown in Fig. 6.1. It is common to use small iron filings to
display magnetic field line patterns of a bar magnet. To define the magnetic field B
at any location in space, the magnetic force F B that the field exerts on a test object
can be used. For that, we can use a charged particle moving with some velocity of
v. Furthermore, we ignore the presence of the gravitational field or an electric field
at the position of the test object.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 131
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_6
132 6 Magnetic Field

Fig. 6.1 A diagram of

magnetic field lines going
from north pole to south pole

Fig. 6.2 Direction of the

magnetic force acting on a
charge particle q (positive or
negative) moving with
velocity v

From the experimental results of the charged particles moving in a magnetic

field (see also Fig. 6.2), the following statements can be written
1. The magnitude of the magnetic force, FB , acting on a charged particle is
proportional to the charge q and the speed v of that particle.
2. The magnitude and direction of the magnetic force F B depend on the veloc-
ity of the particle and the magnitude and direction of the magnetic field B.
3. A charged particle moving parallel to the magnetic field vector B experi-
ences no magnetic force, F B = 0.
4. For a charged particle moving with velocity v making an angle θ = 0 with
the magnetic field B, the magnetic force F B exerted on that particle is
perpendicular to both v and B. Thus, F B is normal to the plane formed by
v and B.
5. Furthermore, the direction of the magnetic force F B acting on a positive
charge moving with velocity v is opposite to the direction of the magnetic
force acting on a negative charge moving with the same velocity v.
6.1 Magnetic Field 133

Fig. 6.3 The right-hand rule demonstration for finding the direction of the magnetic force acting
on a charge particle q (positive or negative) moving with velocity v

6. Moreover, the magnitude of the magnetic force F B acting on the moving

charged particle is proportional to sin θ , where θ is the angle between the
particle velocity v and the direction of the magnetic field B.

Mathematically, those observations can be formulated in the following form:

F B = qv × B (6.1)

Equation (6.1) indicates that the direction of F B applied on a positive charge particle
q is in the direction of v × B, and hence, by definition of the cross product, it is
perpendicular to both v and B (see Fig. 6.3). Furthermore, if q is a negative charge,
then F B is opposite to the direction of v × B. Moreover, Eq. (6.1) implies that the
magnitude of the magnetic force FB is

FB =| q | v B sin θ (6.2)

Here, θ is the smaller angle between v and B. Equation (6.2) implies that FB = 0 if v
is parallel or antiparallel to B (that is, θ = 0 or 180◦ ) and maximum (FB,max = qv B)
if v is perpendicular to B (that is, θ = 90◦ ). Equation (6.2) defines the operational
function of the magnetic field B at some point in space; that is, if a moving charged
particle is placed at that location in space, the magnetic field is defined in terms of
the force acting on that charge.

It is important to summarize the differences between electric and magnetic

forces as follows:
• The electric field applies a force on a charged particle that acts along the
direction of the field. In contrast, the magnetic force is perpendicular to the
direction of the magnetic field.
• The electric force acts on both moving and charged particle at rest, whereas
the magnetic force acts only on a moving charged particle.
134 6 Magnetic Field

• The electric force does work in moving a charged particle in space; in con-
trast, the magnetic force associated with a uniform magnetic field does not
do any work on a moving particle.
• Therefore, using the work-kinetic energy theorem, we can say that the kinetic
energy of a charged particle moving through only a magnetic field remains
constant.
• That is, when on a moving charged particle with a velocity v acts a magnetic
field, the direction of the velocity may change due to the field, but its speed
or kinetic energy remains constant.

From Eq. (6.1), the SI unit of the magnetic field is newtons per coulomb-meter per
second, which is called the tesla (T):

N
1T = 1 (6.3)
C · m/s

Since we defined a coulomb per second as an ampere, we write that

N
1T = 1 (6.4)
A·m
Note that gauss (G) is another unit used for the magnetic field, which relates to the
SI unit, tesla, as follows:
1T = 104 G (6.5)

6.2 Magnetic Force Acting on a Current-Carrying

Conductor

Consider the magnetic force exerted on a single charged particle that moves through
a magnetic field given by Eq. (6.1). If we suppose having a conducting wire in which
a current is maintained, for example, by utilizing a battery, then it should experience
a force when placed in a magnetic field. That is because the current is a stream of
mobile charged particles. Thus, the net force acting by the field on the wire is the
directorial sum of the individual forces exerted on all the charged particles making up
that current. When the mobile charged particles collide with atoms of the conducting
wire, then those magnetic forces used on the charged particles transmit to the wire.
We will consider a straight segment of wire with length L and cross-sectional area
A, carrying a steady current I , to quantify the magnetic force applied by the magnetic
force on a current-carrying wire. Furthermore, suppose that the wire is placed in a
uniform magnetic field B, as shown in Fig. 6.4. The magnetic force exerted on a
charge q moving with a drift velocity vd is given by Eq. (6.1) as
6.2 Magnetic Force Acting on a Current-Carrying Conductor 135

Fig. 6.4 Magnetic force

acting on a charge particle q
moving with a drift velocity
vd

f B = qvd × B (6.6)

We denote by Nq the total number of charges in that segment, given as: Nq = n AL,
where n is the number of charges per unit volume and AL gives the volume of the
segment. Then, to find the total force exerted on the wire, we multiply the force f B
exerted on one charge by Nq . Hence, the total magnetic force on the wire of length
L is
F B = Nq f B = n ALqvd × B (6.7)

Since the current is I = nqvd A, then

F B = I (L × B) (6.8)

In Eq. (6.8), L is a vector pointing in the same direction as the current I and has a
magnitude equal to the length L of the segment.
Note that the expression in Eq. (6.8) applies only to a straight segment of wire in
a uniform magnetic field. To generalize it for any arbitrary shaped wire, consider the
wire segment of uniform cross-section in a magnetic field, as shown in Fig. 6.5. We
partition the segment into small linear segments of length ds, where the vector ds is
in the direction of the current flow in the wire. Then, the magnetic force exerted on
ds in a field B is
dF B = I (ds × B) (6.9)

dF B , in Eq. (6.9), is directed out of the page, if we assume that the magnetic field is
pointing along y-axis. The total force F B acting on the wire is

Fig. 6.5 An arbitrarily

shaped wire segment of
uniform cross-section in
which is passing a current I
in a magnetic field
136 6 Magnetic Field

Fig. 6.6 An arbitrarily

shaped wire segment of
uniform cross-section in
which is passing a current I
in a uniform magnetic field

 b
FB = I ds × B (6.10)
a

a and b, in Eq. (6.10), represent the end points of the wire. Note that the magnitude
of the magnetic field and the direction the field makes with the vector ds may differ
at different points.
To evaluate the formula given by Eq. (6.10), we can consider the following situ-
ations. First, we consider that the magnetic field is uniform; that is, B is constant, as
shown in Fig. 6.6, then
 b
FB = I ds × B (6.11)
a
 b 
=I ds × B
a
= I (L × B)

Using the rule of sum of vectors, L in Eq. (6.11) denotes the vector pointing from a
to b.
In a second situation, we consider an arbitrarily shaped closed loop carrying the
steady current I , which is placed in a uniform magnetic field (see Fig. 6.7). We can
again express the force acting on the loop in the form:

FB = I ds × B (6.12)
 
=I ds × B

Fig. 6.7 A closed shaped

wire segment of uniform
cross-section in which is
passing a current I in a
uniform magnetic field
6.2 Magnetic Force Acting on a Current-Carrying Conductor 137

Because the set of length elements ds forms a closed polygon, the vector sum must
be zero: 
ds = 0 (6.13)

Hence,
FB = 0 (6.14)

Therefore, the net magnetic force acting on any closed current loop in a uniform
magnetic field is zero.

6.3 Torque on a Current Loop in a Uniform Magnetic Field

We showed that a force acts on a current-carrying conductor placed in a magnetic

field, given mathematically by Eq. (6.10). Now, consider a current loop placed in a
magnetic field, as shown in Fig. 6.8. Equation (6.10) implied that the net magnetic
force acting on the loop is zero (see also Eq. (6.14)). However, we will prove in
the following that the magnetic field applies a torque on the current loop. For that,
consider a rectangular loop carrying a current I in the presence of a uniform magnetic
field. First, we assume that the magnetic field is directed parallel to the plane of the
loop, as shown in Fig. 6.8.
The magnetic forces exerted on sides 12 and 34 are both zeroes because these
wire segments are parallel to the field B; hence, L × B = 0 for those two segments.
However, the magnetic field is normal to the portions 23 and 34; therefore, the
magnetic forces on Sects. 23 and 41 are different from zero.
We calculate the magnitude of the forces acting on segments 23 and 41 carrying
the current I and length L = a from Eq. (6.8), as

F23 = F41 =| I (L × B |= I a B (6.15)

Fig. 6.8 A current loop in a

uniform magnetic field
directed parallel to the plane
of loop
138 6 Magnetic Field

Fig. 6.9 A diagram of forces

acting on the current loop

Fig. 6.10 A current loop in

a uniform magnetic field
forming an angle θ to the
plane of loop

However, the direction of F41 , the force exerted on segment 41, is out of the page,
and that of F23 , the force exerted on segment 23, is into the page, as indicated in
Fig. 6.9. Therefore, those two forces point in opposite directions, and they are not
directed along the same line of action; that is, the distance between their lines of
action is b. Thus, those two forces exert a torque on the loop, such that the loop
rotates counterclockwise about the axis passing through the point O, as shown in
Fig. 6.9. The magnitude of that torque τ is

b b b b
τ = F23 + F41 = (I a B) + (I a B) = I abB (6.16)
2 2 2 2
where b/2 is the moment arm about O of each force. Denoting A = ab the area of
closed loop, then
τ = I AB (6.17)

We can generalize that result by considering the same current-carrying closed

loop, but in a uniform magnetic field making an angle θ < 90◦ with the normal
vector to the plane of the loop, as indicated in Fig. 6.10. We also assume that B is
perpendicular to the segment lines 12 and 34.
6.3 Torque on a Current Loop in a Uniform Magnetic Field 139

Fig. 6.11 Diagram of

magnetic forces acting on a
current loop in a uniform
magnetic field forming an
angle θ to the plane of loop

Magnetic forces F23 and F41 exerted on segments 23 and 41 are perpendicular
to those segment lines, and they have the same line of action, and they point in the
opposite direction to one another (see also Fig. 6.11); thus, they do not produce
any torque. On the other hand, F12 and F34 acting on segment lines 12 and 34 have
different lines of action, they point perpendicular to the plane of the loop, and they
have opposite direction. Hence, they produce a torque about the horizontal axis
passing through point O, as indicated in Fig. 6.11.
The moment arm of F12 about the point O is equal to (a/2) sin θ . Similarly, the
moment arm of F3 about O is also (a/2) sin θ . Because F12 = F34 = I bB, the net
torque about O has the magnitude
a a
τ = F1 sin θ + F3 sin θ (6.18)
2 2 a 
a
= I bB sin θ + I bB sin θ
2 2
= I abB sin θ = I AB sin θ

where A = ab is the area of the loop.

Equation (6.18) indicates that the torque is maximum when the magnetic field is
perpendicular to the normal vector of the plane of the loop (θ = 90◦ ):

τmax = I AB (6.19)

That is when the field B is parallel with the plane of the loop. Furthermore, it is zero
when the field B is parallel to the normal vector of the plane of the loop (θ = 0◦ ):

τ =0 (6.20)
 140 6 Magnetic Field

Note that due to the magnetic field, the loop rotates in the direction of decreasing
the angle θ . That is, the loop rotates in the direction such that the normal vector n to
the surface of the loop (or equivalently, surface area vector A = An) aligns with the
Torque magnetic field vector B.
vector
A convenient expression for the torque applied on a loop in a uniform magnetic
field B is
τ = I (A × B) (6.21)

where A, the vector of the surface area of the loop, is perpendicular to the plane of
the loop, that is, A = An (see also Fig. 6.11). Direction of A is determined using the
Mag- right-hand rule shown in Fig. 6.10.
netic
dipole The product I A is defined to be the magnetic dipole moment μ (or the “magnetic
moment") of the loop:
μ = IA (6.22)

The SI unit of magnetic dipole moment is ampere-meter2 (A · m 2 ). Using this def-

inition, we can express the torque exerted on a current-carrying loop in a magnetic
field B as
τ =μ×B (6.23)

The direction of the magnetic moment is the same as the direction of A (Fig. 6.12).

Fig. 6.12 Right-hand rule

demonstration in
determining the direction of
magnetic dipole
6.4 Motion of a Charged Particle in a Uniform Magnetic Field 141

6.4 Motion of a Charged Particle in a Uniform Magnetic

Field

From Eq. (6.1), the force exerted on a charged particle moving in a magnetic field is
normal to the velocity of the particle, and hence, perpendicular to its displacement.
Therefore, the work done on the particle by the magnetic force is zero because the
force is perpendicular to the displacement vector of the particle.
As a particular case, we consider a positively charged particle moving in a uniform
magnetic field with an initial velocity vector of the particle perpendicular to the field
(see Fig. 6.13). The direction of the magnetic field is pointing into the page. Using
the right-hand rule in Eq. (6.1), we find that the direction of the magnetic force is
pointing toward a single point at the center of a circle. Therefore, the particle is going
to move in a circle in a plane perpendicular to the magnetic field.
The particle moves in this way because the magnetic force F B is perpendicular to
both v and B and has a constant magnitude of qv B. As the force deflects the particle,
the directions of both v and F B change continuously, and F B points toward the center
of the circle at each position of the particle. Thus, force changes only the direction of
v, but it does change its magnitude. The rotation is counterclockwise for that positive
charge, and if q is negative, the rotation would be clockwise.
Using the second law of Newton for circular motion, we get

Fir = mar (6.24)
i

or
v2
FB = qv B = m (6.25)
r
The radius of the circle is
mv 2 mv
r= = (6.26)
qv B qB

Fig. 6.13 A positive charge

particle moving in a uniform
magnetic field with initial
velocity perpendicular to the
field
142 6 Magnetic Field

That is, the radius of the path is proportional to the linear momentum mv of the
particle and inversely proportional to the magnitude of the charge on the particle and
to the magnitude of the magnetic field. The angular speed of the particle is

v qB
ω= = (6.27)
r m
The period of the motion (the time that the particle takes to complete one revolution)
is equal to the circumference of the circle divided by the linear speed of the particle:

2πr 2π 2π m
T = = = (6.28)
v ω qB

6.5 Exercises

Exercise 6.1 Consider an electron in a television picture tube, which moves

toward the front of the tube (that is, toward us) with a speed of 8.0 × 106 m/s
along the x axis (see Fig. 6.14). The faces of the tube are coils of wire that
create a magnetic field of magnitude 0.025 T, at an angle of 60◦ to the x axis
and lying in the x y plane. Determine: (a) the magnetic force exerted on the
electron; (b) its acceleration.

Solution 6.1 From Eq. (6.2), we obtain

FB = (1.6 × 10−19 C)(8.0 × 106 m/s)(0.025 T)(sin 60◦ ) (6.29)

−14
= 2.8 × 10 N

Fig. 6.14 An electron in a

television picture tube moves
toward the front of the tube
with a constant speed
6.5 Exercises 143

In addition, the vector v × B is along the positive z direction (from the right-hand
rule) and the charge is negative, and hence, F B is along the negative z direction. Con-
sidering that the electron mass is m e = 9.11 × 10−31 kg, the electron acceleration
is
FB
a= (6.30)
me
2.8 × 10−14 N
=
9.11 × 10−31 kg
= 3.1 × 1016 m/s2

In addition, the vector a is along the direction negative z axis.

Exercise 6.2 Consider a conducting wire loop in the form of a semicircle

of radius R carrying a current I , as shown in Fig. 6.15. Suppose a uniform
magnetic field is applied in the direction along the positive vertical direction.
Determine the magnitude and direction of the magnetic force acting on the
straight and the curved portions of the wire.

Solution 6.2 The two forces acting on the wire are F1 applied on the straight portion
and F2 applied on the semicircle portion.
The magnitude of F1 is
F1 = I L B = 2I R B (6.31)

because the straight portion of the wire is perpendicular to B. The direction of F1 is

out of the page, since L × B is along the positive z axis.
We first write an expression for the force dF2 on a small element of length ds:

d F2 = I | ds × B |= I B sin θ ds (6.32)

Since s = Rθ , then ds = Rdθ , then d F2 = I R B sin θ dθ . Hence, the total force is

Fig. 6.15 A wire bent into a

semicircle of radius R forms
a closed circuit and carries a
current I
144 6 Magnetic Field
 π
F2 = I R B sin θ dθ (6.33)
0
= I R B [− cos θ ]π0
= 2I R B

Because F2 , with a magnitude of 2I R B, is directed into the page and because F1 ,

with a magnitude of 2I R B, is directed out of the page, the net force on the closed
loop is zero.

Exercise 6.3 A rectangular coil with dimensions of 5.40 cm × 8.50 cm and

25 turns of wire. It carries a steady current of 15.0 mA. In addition, the magnetic
force of 0.350 T is applied parallel to the plane of the loop. Determine: (a) the
magnitude of the magnetic dipole moment; (b) the magnitude of the torque
acting on the loop.

Solution 6.3 (a) For a coil with N = 25 turns, the magnetic dipole is

μcoil = N I A = 25(15.0 × 10−3 A)(0.0540 × 0.0850 m2 ) (6.34)

= 1.72 × 10−3 A · m2

(b) Since B is perpendicular to μcoil , then

τ = μcoil B = (1.72 × 10−3 A · m2 )(0.350 T) (6.35)

−4
= 6.02 × 10 N·m

Exercise 6.4 A satellite uses a device made up of coils to adjust the orientation
by interacting with the Earth’s magnetic field. A torque is created on the satellite
in the x, y, or z direction. This control system is used solar-generated electricity.
If we assume that the device has a magnetic dipole moment of 250 A · m2 , find
the maximum torque applied to the satellite when it is at a location where the
magnitude of the Earth’s magnetic field is 3.0 × 10−5 T.
6.5 Exercises 145

Solution 6.4 The torque is given as

τ =μ×B (6.36)

and the magnitude is

τ = μB sin θ (6.37)

where θ is angle between μ and B. The maximum torque is obtained when

τmax = μB = (250 A · m2 )(3.0 × 10−5 T) (6.38)

 
−5 N
= 750 × 10 A·m · 2
A·m
= 750 × 10−5 N · m

Exercise 6.5 A proton of mass 1.67 × 10−27 kg moves in a circular orbit

of radius 14 cm. Assume a uniform magnetic field of magnitude 0.35 T is
perpendicular to plane of the orbit. What is the linear speed of the proton?

Solution 6.5 The linear speed is of proton is

q Br
v= (6.39)
mp
(1.60 × 10−19 C)(0.35 T)(14 × 10−2 m)
=
1.67 × 10−27 kg
= 4.7 × 10 m/s
6

Exercise 6.6 An electron of charge q = −1.60 × 10−19 C and mass 9.11 ×

10−31 kg moves in a direction perpendicular to a uniform magnetic field of
magnitude 0.35 T with a linear speed 4.7 × 106 m/s. Find the radius of its
circular orbit.

Solution 6.6 The radius of the circle is given by

mev
r= (6.40)
qB
146 6 Magnetic Field

Then,

(9.11 × 10−31 kg)(4.7 × 106 m/s)

r= (6.41)
(1.60 × 10−19 C)(0.35 T)
= 7.6 × 10−5 m

Exercise 6.7 To measure the magnitude of a uniform magnetic field, electrons

are accelerated from rest in a potential difference of 350 V. They travel along
a curved path because of the magnetic force exerted on electrons. Assume that
the radius of the path is 7.5 cm and the magnetic field is perpendicular to the
beam of electrons. Determine: (a) the magnitude of the field; (b) the angular
speed of the electrons; (c) the period of revolution of the electrons.

Solution 6.7 (a) The change in the electric potential energy is

ΔU = (U f − Ui ) = − | e | ΔV (6.42)

ΔV is the potential difference. Using the conservation law of energy: ΔK = −ΔU .

Since
K i = 0; K f = m e v 2 /2 (6.43)

Thus, we get
v2
me =| e | ΔV (6.44)
2
Or,
2 | e | ΔV
v= (6.45)
me

Replacing the numerical values:

2(1.60 × 10−19 C)(350 V)
v= (6.46)
9.11 × 10−31 kg
= 1.11 × 107 m/s

and the magnetic field is

6.5 Exercises 147

mev
B= (6.47)
|e|r
(9.11 × 10−31 kg)(1.11 × 107 m/s)
=
(1.60 × 10−19 C)(0.075 m)
= 8.4 × 10−4 T

(b) From the equation

v
ω= (6.48)
r
we get
1.11 × 107 m/s
ω= = 1.5 × 108 rad/s (6.49)
0.075 m
(c) For the period
2π
T = = 4.1888 × 10−8 s ≈ 42 ns (6.50)
ω

Exercise 6.8 Suppose a proton of charge 1.6 × 10−19 C moves with a veloc-
ity of v = (2i − 4j + k) m/s in a region in which the magnetic field is
B = (i + 2j − 3k) T. What is the magnitude of the magnetic force this charge
experiences?

Solution 6.8 The magnetic force is

F B = qv × B (6.51)

where q is the charge of proton q = 1.6 × 10−19 C. The components are

FBx = q(v × B)x = q(v y Bz − vz B y ) (6.52)

FBy = q(v × B) y = −q(vx Bz − vz Bx )
FBz = q(v × B)z = q(vx B y − v y Bx )

or

FBx = (1.6 × 10−19 C)((−4.0)(−3.0) − (1.0)(2.0)) T · m/s (6.53)

−19
= 16 × 10 N
−19
FBy = −(1.6 × 10 C)((2.0)(−3.0) − (1.0)(1.0)) T · m/s (6.54)
−19
= 11.2 × 10 N
148 6 Magnetic Field

FBz = (1.6 × 10−19 C)((2.0)(2.0) − (−4.0)(1.0)) T · m/s (6.55)

−19
= 12.8 × 10 N

The magnitude of the force is

FB = 2
FBx + FBy
2
+ FBz
2
(6.56)
√
= 256 + 125.44 + 163.84 × 10−19 N
≈ 23.4 × 10−19 N

Exercise 6.9 An electron of charge −1.6 × 10−19 C is moving in a uniform

magnetic field of B = (1.40i + 2.10j) T. Find the magnetic force on the elec-
tron if its velocity is v = 3.70 × 105 j m/s.

Solution 6.9 The magnetic force given by Eq. (6.1), where q is the charge of electron
q = −1.6 × 10−19 C. The components are

FBx = q(v × B)x = q(v y Bz − vz B y ) (6.57)

FBy = q(v × B) y = −q(vx Bz − vz Bx )
FBz = q(v × B)z = q(vx B y − v y Bx )

or

FBx = (−1.6 × 10−19 C)((3.70)(0.0) − (0.0)(2.10)) × 105 T · m/s (6.58)

= 0.0 N
FBy = −(−1.6 × 10−19 C)((0.0)(0.0) − (0.0)(1.40)) T · m/s
= 0.0 N
FBz = (−1.6 × 10−19 C)((0.0)(2.10) − (3.70)(1.40)) × 105 T · m/s
= 8.3 × 10−14 N

The magnetic force is

F B = 8.3 × 10−14 k N (6.59)
6.5 Exercises 149

Exercise 6.10 Consider a wire with a mass per unit length of 0.500 g/cm. A
current of 2.00 A flows horizontally to the east. Determine the direction and
magnitude of the minimum magnetic field needed to lift this wire vertically
upward.

Solution 6.10 The magnetic force on the current-carrying wire is

FB =| I L × B | (6.60)
= I L B sin θ

where θ is the angle between the direction of the current and magnetic field. The
force per unit length is then

FB
fB = = I B sin θ (6.61)
L
In order to lift the wire vertically up, it should be larger than the force of gravity
f g = mg/L per unit length and directed upwards to lift the wire, so θ = 90◦ : f B =
I B. Hence the minimum magnetic field is given by

(m/L)g
f B = fg → B = (6.62)
I
where m/L = 0.05 kg/m. For the force to point upward, the field, by the right-hand
rule, should point east. Its magnitude, from the above formula

B = 0.05 × 9.8/2 = 0.25 T (6.63)

Exercise 6.11 Consider a wire loop in the form of a square of side length
10.0 cm, which carries a steady current of 10 A. The loop is placed in a uniform
magnetic field of magnitude 4.0 T. The magnetic field vector makes an angle of
30◦ with the normal vector to the plane of the loop. Determine the magnitude
of the torque acting on the loop.

Solution 6.11 The torque is given by

τ = IA × B (6.64)

and its magnitude is

τ = I AB sin θ (6.65)
150 6 Magnetic Field

where θ is the angle between cross-sectional area vector A of the plane and B:
θ = 30◦ , hence

τ = (10 A)(0.10 × 0.10 m2 )(4.0 T) sin(30◦ ) (6.66)

= 0.20 N · m

Exercise 6.12 Consider an electron that undergoes a circular motion with a

radius of 23 mm, and its kinetic energy is 6.4 × 10−17 J in a uniform magnetic
field. Find the magnetic field.

Solution 6.12 The radius is given as

mv mv
r= →B= (6.67)
qB qr

where mv is the momentum of the particle, i.e.,

√
mv = 2m K (6.68)

where K is the kinetic energy. Using this expression, we get

√
2m K
B= = 2.93 mT (6.69)
qr

Exercise 6.13 Consider a steady current of 2.40 A in a wire with a straight

section of 0.750 m long placed along the x-axis in a uniform magnetic field of
magnitude B = 1.60 T applied in the positive z direction. Suppose the current
is in the +x direction. Find the magnetic force on the section of wire.

Solution 6.13 The magnetic force is

FB = I L × B (6.70)

where L is pointing in the direction of I , so

L = 0.750i m (6.71)
6.5 Exercises 151

and
B = 1.60k T (6.72)

Therefore,

FBx = I (L y Bz − L z B y ) = 0 (6.73)
FBy = −I (L x Bz − L z Bz ) = −(2.40 A)(0.750 m)(1.60 T) = −2.88 N
FBz = I (L x B y − L y Bx ) = 0

Hence, F B is in the negative y direction, and its magnitude is 2.88 N.

Reference

Holliday D, Resnick R, Walker J (2011) Fundamentals of physics. John Wiley and Sons
Chapter 7
Sources of Magnetic Field

The chapter aims to introduce sources of magnetic field, Gauss’s

and Maxwell-Ampére’s laws.

In this chapter, we introduce the sources of the magnetic field. In particular, we

introduce Biot-Savart law and the magnetic forces between the current-carrying
parallel conductors. Also, we introduce Ampére’s law and the magnetic field flux.
Then, Gauss’s law and Maxwell-Ampére’s law will be presented. As extra reading
material, the reader can also consider other literature Holliday et al. (2011), Jackson
(1999), Griffiths (1999).

7.1 Biot-Savart Law

Biot-Savart law gives a mathematical formula about the magnetic field at a location
in space in terms of the current that produces the field. According to Biot-Savart law, Biot-
Savart
the magnetic field dB produced by a length element ds of a wire carrying a steady Law
current I at any point P (see also Fig. 7.1):

μ0 I ds × r̂
dB = (7.1)
4π r 2
where μ0 is a constant called the permeability of free space:

μ0 = 4π × 10−7 T · m/A (7.2)

In Eq. (7.1), r gives the distance from the position of element ds to the point P, and
r̂ is a unit vector along that direction, as indicated in Fig. 7.1.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 153
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_7
154 7 Sources of Magnetic Field

Fig. 7.1 The magnetic field

dB created by a small
segment of current-carrying
wire at the point P and P  :
dB is proportional to the
vector ds × r̂, and hence, at
point P it points out of the
viewing plane and at the
point P  it points in the
viewing plane

To find the total magnetic field B created at the point P by a current-carrying wire
of finite size and arbitrary shape, we sum up those contributions from all current
elements I ds that make up the current (see Eq. (7.1)) to obtain

μ0 I ds × r̂
B= (7.3)
4π r2
L

where L is the path along the wire.

In Eq. (7.3), the integral is about the entire current distribution over the wire.
Furthermore, notice that the unit vector r̂ is along the line that joins the position of
the length element ds along the direction of the current in the wire and the point P,
and hence, it is part of the integral.
Biot-Savart law also applies to a current consisting of charges flowing through
space, such as the electron beam in a television tube, or any other beam of some
charges (positive or negative). In that case, ds denotes the length of a small segment
vector of space along the direction of charge flow.

Similarities between Biot-Savart law for magnetism and Coulomb’s law for
electrostatics exist as follows.
First, the current element produces a magnetic field, whereas a fixed point
charge in space produces an electric field. Hence, the current is a source of the
magnetic field in space, and the charge at rest is a source of the electrostatic
field in that space.
Furthermore, both the magnitude of the magnetic field and electrostatic field
vary as the inverse square of the distance from the source (current element or
static charge) of that field to any location P in space. However, the directions
of the two fields are substantially different. The electric field created by a
point charge is radial, that is, it is either parallel (if the charge is positive) or
7.1 Biot-Savart Law 155

antiparallel (if the charge is negative) to the unit vector r̂. But, the magnetic
field created by a current element is perpendicular to both the small portion
element ds and the unit vector r̂.
Moreover, the electric field is due to an isolated electric charge at some
point, which exists either an isolated single point-like charge or as a portion of
an extended distribution of charges. Biot-Savart law, on the other hand, relates
the magnetic field of an isolated current element at some point that exists
only as a portion of an extended current distribution in a complete circuit
where the charges flow. Therefore, Biot-Savart law is only the first step in the
determination of a magnetic field; mathematically, it is the integrand of an
integral over the current distribution.

The magnetic field produced by a current-carrying long and straight wire often
occurs, as shown in Fig. 7.2 where the magnetic field surrounding a long, straight
current-carrying wire is presented. The magnetic field lines are circles concentric
with the wire, due to the symmetry, laying in planes perpendicular to the wire.
Furthermore, the magnitude of B is constant on any point of a circle of radius a
given by (see solved exercises)

Fig. 7.2 Right-hand rule

demonstration
 156 7 Sources of Magnetic Field

μ0 I
B= (7.4)
2πa
The
right- The right-hand rule is a convenient way for determining the direction of B; that is,
hand
rule we grasp the wire with the right hand, positioning the thumb along the direction of
the current, then the four fingers curl in the direction of the magnetic field, as shown
in Fig. 7.2.

7.2 Magnetic Force Between Two Parallel Conductors

Suppose we have two long, straight, parallel current-carrying wires (named 1 and 2,
respectively) at a distance a from one another. Let I1 and I2 be the currents in each
wire that flow in the same direction, as shown in Fig. 7.3. Based on Biot-Savart law,
wire 2, carrying the current I2 , creates a magnetic field B2 at the location of wire 1.
Using the right-hand rule, the direction of B2 is perpendicular to wire 1 at its location,
and the magnetic force on a length  of wire 1 is given by Eq. (6.8) (Chap. 6):

F1 = I1  × B2 (7.5)

Direction of the force F1 is shown in Fig. 7.3(A).

Because  is perpendicular to B2 , the magnitude of force in Eq. (7.5) is

F1 = I1 B2 (7.6)

Using Eq. (7.4) for the magnitude of B2 , we have

μ0 I2
B2 = (7.7)
2πa
Substituting Eq. (7.7) into Eq. (7.6), we obtain

Fig. 7.3 Demonstration of the interaction between two long and straight current-carrying wires I1
and I2
7.2 Magnetic Force Between Two Parallel Conductors 157

μ0 I1 I2
F1 = I1 B2 =  (7.8)
2πa
Similarly, we can now calculate the force F2 acting on wire 2 that is placed in the
field set up by wire 1, as shown in Fig. 7.3b. Based on third law of Newton, F2 is
equal in magnitude and opposite in direction to F1 :

F2 = I2  × B1 = −F1 (7.9)
μ0 I1 I2
F2 = I2 B1 =  = F1 (7.10)
2πa
For the currents flowing in opposite directions (that is when one of the currents
is reversed), the forces exerted on each current-carrying wire are reversed, and the
wires repel each other. Hence, we find that when parallel conductors carry currents
in the same direction, they attract each other, and when they take currents in opposite
directions repel each other. Because the magnitudes of the forces are the same on
both wires, we denote the magnitude of the magnetic force between the wires as
simply FB .
We can rewrite this magnitude in terms of the force per unit length:

FB μ0 I1 I2
= (7.11)
 2πa
Following the result of the force per unit length, given by Eq. (7.11), between any
two parallel wires, Ampére defined the unit of ampere as 1 ampere (denoted 1 A) Defini-
tion of
is the current carried in each of two long, parallel identical wires at distance of 1 m Ampere
from one another, which interact between them with a force of magnitude per unit
length 2 × 10−7 N/m:

(4π × 10−7 T · m/A)(1 A)(1 A)

2 × 10−7 N/m = (7.12)
2π(1 m)

7.3 Ampére’s Law

The following experiment by Oersted, in 1819, about deflection of the compass

needles demonstrated that a current-carrying conductor produces a magnetic field,
and it also lead to Ampére’s law.
158 7 Sources of Magnetic Field

Fig. 7.4 Compass needles

are placed in a horizontal
plane nearby a long vertical
wire

In this experiment, several needles of a compass are distributed in a horizontal

plane nearby a long vertical wire, as shown in Fig. 7.4. If no current flows in the
wire, all the needles point in the same direction along the Earth’s magnetic field
because the magnetic field of Earth exerts a torque on the magnetic moment of
each needle. If a steady current flows in each wire, all needles deflect and align
tangent to the circle along the direction of the magnetic field produced by the
long current-carrying wire. The observation demonstrated that the direction of
the magnetic field produced by the current in the wire is consistent with the
right-hand rule. Furthermore, if the current changes in the opposite direction,
the needles reverse.

The experiment demonstrated that the compass needles point in the direction of
B, and the magnetic field of B forms circles around the wire. The magnitude of B
is the same at every point on a circular path centered on the wire, by symmetry,
and it is lying in the circle plane perpendicular to the wire. This is illustrated in
Fig. 7.5. Moreover, varying the current and distance a from the wire, B varies, and
it is proportional to the current and inversely proportional to the distance a from the
wire as in Eq. (7.4).
We can evaluate the product B · ds for a small element ds on the circular path L,
as shown in Fig. 7.5. Note that the vector B is tangent to the circular path at the tail
of vector ds, and hence, they are approximately parallel. Summing up the products
for all small elements over the closed circular path, we obtain

B · ds = Bds (7.13)

Furthermore, the magnitude of B is constant along the path L at every point, and
hence, the sum of the products Bds over the closed line becomes the line integral of
B · ds:
7.3 Ampére’s Law 159

Fig. 7.5 The magnetic field

at every point on a circular
path L centered at a long
vertical current-carrying
wire, I . ds is a small potion
of the path L such that B is
approximately parallel to ds

 
B · ds = B ds (7.14)
L L
μ0 I
= (2πr )
2πr
= μ0 I

Note that Eq. (7.14) is derived for the special case of a circular path surrounding
a wire; it, however, holds for any arbitrary shaped closed path surrounding a current-
carrying wire that is part of a complete circuit. In the general case of an arbitrary
closed path, Ampére’s law states: The closed path line integral of B · ds is μ0 I ,
Ampére’s
where I is the total continuous current passing through any surface supported by the law
closed path L. Mathematically, we write

B · ds = μ0 I (7.15)
L

7.4 Magnetic Flux

The flux related to a magnetic field is defined similarly to electric flux. For that,
we consider a small surface element of area d A on an arbitrarily shaped surface, as
indicated in Fig. 7.6. Let B be the magnetic field vector at this small surface element
vector dA, the magnetic flux through the surface element is B · dA. Here, dA defines
a vector that is perpendicular to the surface and has a magnitude equal to the area
160 7 Sources of Magnetic Field

Fig. 7.6 Magnetic flux

through an arbitrary shaped
surface

d A: dA = d An, where n is a unit normal vector at the center of surface element.

Hence, the total magnetic flux Φ B through the surface is

ΦB = B · dA (7.16)
A

If we consider a planar surface of area A and a uniform field B that makes an

angle θ with dA at every point on the surface. Then, the magnetic flux through the
plane is
Φ B = B A cos θ (7.17)

From Eq. (7.17), when the magnetic field B is parallel to the plane, θ = 90◦ (that
is, cos θ = 0) and the flux is zero: Φ B = 0. If the magnetic field B is perpendicular
to the plane of the surface, then θ = 0 (that is, cos θ = 1) and the flux is maximum
Φ B = B A.
The SI unit of flux is the T · m2 , defined as the weber (Wb):

1 Wb = 1 T · m2 (7.18)

7.5 Gauss’s Law in Magnetism

In Chap. 2, we expressed the electric flux through a closed surface surrounding a

net charge is proportional to that charge inside the volume, as Gauss’s law. That
is, the net number of electric field lines leaving the surface depends only on the net
charge within it. That is related to the fact that electric field lines originate from some
positive electrical charges and terminate on some other electric negative charges; that
is, often, there is a non-zero net difference between the number of electric field lines
leaving and entering any closed surface surrounding the distribution of charges.
7.5 Gauss’s Law in Magnetism 161

However, that is substantially different for magnetic field lines, which are contin-
uous closed loops. Hence, magnetic field lines do not begin at some point and end
at any other location. Furthermore, the number of lines entering any closed surface
is equal to the number of lines leaving that surface, and thus, the net magnetic flux
must be zero. In contrast, the net electric flux through a closed surface enclosing only
one of the charges of an electric dipole moment is not zero.
Gauss’s
In magnetism, Gauss’s law defines the net magnetic flux through any closed law in
surface to be always zero: mag-
 netism
ΦB = B · dA = 0 (7.19)
A

The experimental facts support Gauss’s law in magnetism that isolated magnetic
poles (monopoles) have not been detected yet, and they perhaps do not exist. The
research for identifying magnetic monopoles continues because specific theories
suggest the possibility of finding monopoles experimentally.

7.6 Displacement Current

We showed that mobile charges produce magnetic fields, such as the free charge
carrier of a current-carrying conductor, as indicated in Ampére’s law. Furthermore,
if a current-carrying conductor has high symmetry, we can use Ampére’s law to
calculate the produced magnetic field, such as a long, straight current-carrying wire.
In Eq. (7.15), the line integral is over any closed path L through which the conduction
current passes, defined as
dq
I = (7.20)
dt
We now suppose that an electric field that varies with time is present, and show
that Ampére’s law in the form given by Eq. (7.15) is valid only if that electric field is
constant in time, which was proved by Maxwell. For that, let us consider a capacitor
that is being charged, as shown in Fig. 7.7. In the presence of a conduction current,
the charge on the positive plate changes, let say in the time interval dt it changes by
an amount dq = I dt. However, the conduction current between the plates is zero,
I = 0.
Let us take two surfaces S1 and S2 , bounded by the same closed line path P,
as indicated in Fig. 7.7. Note that the choice of the shape of these two surfaces is
completely arbitrary. In Fig. 7.7, S1 is one of the faces of a cube, and the other five
faces of the same cube construct the open surface S2 .
Based on Ampére’s law, Eq. (7.15), when the closed line path L is the path P
bounding the surface S1 , then 
B · ds = μ0 I (7.21)
P
162 7 Sources of Magnetic Field

Fig. 7.7 A capacitor formed

by two charged plates at the
charge Q

where I is the conduction current passing through S1 . In contrast, when the closed
line path is considered as bounding S2 , then the right-hand side of Eq. (7.15) is zero
because no conduction current passes through S2 :

B · ds = 0 (7.22)
P

Equations (7.21) and (7.22) show a disagreement by applying Ampére’s law as

in Eq. (7.15), which arises from the discontinuity of the current, as discovered by
Maxwell. To solve this contradiction, Maxwell postulated to add another term on the
right-hand side of Eq. (7.15) that included a factor called the displacement current
Id , as
dΦ E
Id = 0 (7.23)
dt

where 0 is the permittivity of free space and Φ E = A E · dA is the electric flux.
Physical basis of that term can be explained as follows. When the capacitor is
being charged (or discharged), the change on electric field between the plates may
be considered equivalent to adding a current (that is, the displacement current) flow-
ing only inside the capacitor that makes the conduction current in the wire to be
continuous. That current, Id , can be added to the right of the expression in Eq. (7.15)
as 
dΦ E
B · ds = μ0 (I + Id ) = μ0 I + μ0 0 (7.24)
L dt

Now, it does matter which surface bounded by the closed path (path P in Fig. 7.7) is
chosen, either conduction current (I ) for the surface S1 or displacement current (Id )
for the surface S2 passes through it. Equation (7.24) is a general form of Ampére’s
law, and it is called the Ampére-Maxwell law.
The electric flux through surface S2 is
7.6 Displacement Current 163

ΦE = E · dA = E A (7.25)
S2

where A is the surface area of the capacitor plates and E is the magnitude of the
uniform electric field between the plates. If Q is the charge on the plates at any
instant, then
Q
E= (7.26)
0 A

Therefore, the electric flux through S2 is simply

Q
ΦE = E A = (7.27)
0

Thus,

dΦ E
Id = 0 (7.28)
dt
dQ
=
dt
Equation (7.28) implies that the displacement current through open surface S2
is precisely equal to the conduction current I through S1 . Thus, the displacement
current is the source of the magnetic field on the surface boundary of S2 . Moreover,

Fig. 7.8 A capacitor formed

by two charged plates at the
charge Q.Electric field inside
the capacitor
164 7 Sources of Magnetic Field

the physical origin of the displacement current is the time-varying electric field
(see Fig. 7.8). The central point of this formalism, then, is that magnetic fields are
produced both by conduction currents and by time-varying electric fields.

7.7 Exercises

Exercise 7.1 Consider a thin, straight wire carrying a constant current I and
placed along the x axis (Fig. 7.9). Find (a) the magnitude and (b) direction of
the magnetic field at point P due to this current.

Solution 7.1 Using Biot-Savart law:

μ0 I ds × r̂
dB = (7.29)
4π r 2

where ds × r̂ has the direction along positive z axis, thus

μ0 I d x sin θ
dB = k (7.30)
4π r2
where | ds |= d x. We can get
 
1 dθ
d x = −a d =a (7.31)
tan θ sin2 θ

Then,

Fig. 7.9 A straight wire

carrying a steady current I
and placed along the x axis
7.7 Exercises 165

μ0 I adθ
dB = k (7.32)
4π sin θr 2
μ0 I sin θ dθ
= k
4π a
Then, for the total magnitude of magnetic field we get:

μ0 I θ2
B= sin θ dθ (7.33)
4aπ θ1
μ0 I
= (cos θ1 − cos θ2 )
4aπ

Consider the special case of an infinitely long, straight wire: θ1 = 0 and θ2 = π, thus

μ0 I
B= (7.34)
2aπ

Exercise 7.2 What is the magnetic field at point O for the current-carrying
wire segment in Fig. 7.10? The wire consists of two straight segments and a
circular arc of radius R, which subtends an angle θ. The arrowheads on each
portion of the wire indicate the direction of the current (see also Fig. 7.10).

Solution 7.2 The magnetic field at O due to the current in the straight segments
A A and CC  is zero because ds is parallel to r̂ along these paths; this means that

ds × r̂ = 0 (7.35)

Fig. 7.10 A wire consists of

two straight portions and a
circular arc of radius R,
which subtends an angle θ.
The arrowheads on the wire
indicate the direction of the
current
166 7 Sources of Magnetic Field

Each length element ds along path AC is at the same distance R from O, and the cur-
rent in each contributes a field element dB directed into the page at O. Furthermore,
at every point on AC, ds is perpendicular to r̂; hence,

| ds × r̂ |= ds (7.36)

The magnitude of the field at O due to the current in an element of length ds is:

μ0 I ds
dB = (7.37)
4π R 2
Therefore, the total magnetic field magnitude is

μ0 I ds
B= (7.38)
4π R2
μ0 I s
=
4π R2
μ0 I θ
=
4π R

Exercise 7.3 Consider a circular wire loop of radius R located in the yz plane
and carrying a steady current I (see Fig. 7.11). Calculate the magnetic field at
an axial point P a distance x from the center of the loop.

Solution 7.3 Every length element ds is perpendicular to vector r̂ at the location of

the element. Thus,
| ds × r̂ |= ds (7.39)

Furthermore, all small segment elements around the loop are at the same distance r
from P, where r 2 = x 2 + R 2 . Hence, the magnitude of dB due to the current in any
length element ds is

Fig. 7.11 A circular wire

loop of radius R located in
the yz-plane and carrying a
steady current I
7.7 Exercises 167

μ0 I | ds × r̂ | μ0 I ds
dB = = (7.40)
4π r2 4π x 2 + R 2

We can write
dB = d Bx i + d B y j (7.41)

and   
B= d Bx i + d By j = d Bx i = Bx i (7.42)

where d Bx = d B cos θ, hence


μ0 I ds cos θ
Bx = (7.43)
4π x 2 + R 2

where
R
cos θ = √ (7.44)
x2 + R2

Thus, we obtain

μ0 I Rds
Bx = (7.45)
4π (x + R 2 )3/2
2

μ0 I R2π R
=
4π (x 2 + R 2 )3/2
μ0 I R 2
=
2(x 2 + R 2 )3/2

Thus,
μ0 I R 2
B= i (7.46)
2(x 2 + R 2 )3/2

Exercise 7.4 A long, straight cylindrical wire with a radius R carries a steady
current I0 , which is uniformly distributed through its cross-section, as shown
in Fig. 7.12. Calculate the magnetic field at the region 1 (r ≥ R) and 2 (r < R).

Solution 7.4 (a) Let us choose for our path of integration a circle, labeled 1 in
Fig. 7.12. From symmetry, B must be constant in magnitude and approximately
parallel to ds at every point of the circle 1. Because the total current passing through
the plane of the circle is I0 , Ampére’s law gives
 
B · ds = B ds = B(2πr ) = μ0 I0 (7.47)
168 7 Sources of Magnetic Field

Fig. 7.12 A long, straight

cylindrical wire of radius R
carrying a steady current I0
uniformly distributed
through its cross section

or (for r ≥ R) we get
μ0 I0
B= (7.48)
2πr
(b) Let us choose, for our path of integration, the circle 2. The same symmetry applies
in this case. Using Ampére’s law:
 
B · ds = B ds = B(2πr ) = μ0 I (7.49)

Then, the magnetic field magnitude is

μ0 I
B= (7.50)
2πr
where the current I is determined according to

I0 I
= 2 (7.51)
π R2 πr
From here, the current is
r2
I = I0 (7.52)
R2
Therefore, we obtain (for r < R)

μ0 I0
B= r (7.53)
2π R 2
7.7 Exercises 169

Fig. 7.13 A toroid with

inner radius a and outer
radius b consists of a
conducting wire wrapped
around a ring (a torus) made
of a nonconducting material

Exercise 7.5 A device called a toroid is often used to create an almost uniform
magnetic field in some enclosed area (see also Fig. 7.13). The device consists
of a conducting wire wrapped around a ring (a torus) made of a nonconducting
material. For a toroid having N closely spaced turns of wire, calculate the
magnetic field in the region occupied by the torus at a distance r from the
center.

Solution 7.5 Using Ampére’s law:


B · ds = μ0 N I (7.54)

where N is the number of wires. By symmetry, we see that the magnitude of the field
is constant on this circle and tangent to it, so B · ds = Bds. Hence,

B ds = B(2πr ) = μ0 N I (7.55)

Therefore, B is
μ0 N I
B= (7.56)
2πr

Exercise 7.6 Let us now consider an extended object carrying a current. A

thin, infinitely large sheet lying in the yz plane carries a current of linear
current density Js , as shown in Fig. 7.14. The current is in the y direction, and
Js represents the current per unit length measured along the z axis. Find the
magnetic field near the sheet.
170 7 Sources of Magnetic Field

Fig. 7.14 A thin, infinitely

large sheet lying in the
yz-plane carries a current of
linear current density Js

Solution 7.6 Using Ampére’s law:


B · ds = μ0 I = μ0 Js  (7.57)

where the rectangle has dimensions  and w, with the sides of length  parallel to
the sheet surface. Or,
2B = μ0 Js  (7.58)

Thus,
μ0 Js
B= (7.59)
2

Exercise 7.7 A current-carrying wire (labeled 1) is oriented along the y axis

in which the current is I1 . A rectangular loop is located to the right of the wire
1 and in the x y plane. Wire 2 carries a current I2 (see also Fig. 7.15). What is
(a) the magnetic force applied on the top wire of length b in the loop, labeled
Wire 2 in Fig. 7.15 and (b) the force exerted on the bottom wire of length b?

Solution 7.7 Let consider the force exerted by wire 1 on a small segment ds of wire
2 by using:
dF B = I2 ds × B (7.60)

From Ampére’s law, the field at a distance x from wire 1 is

μ0 I1
B= (−k) (7.61)
2πx
where k is a unit vector along z axis. Here, ds = d xi, then we obtain
7.7 Exercises 171

Fig. 7.15 A wire 1 oriented

along the y axis and carrying
a steady current I1 , and
another rectangular loop
located to the right of the
wire and in the x y-plane
carries a current I2

μ0 I1
dF B = I2 d xi × (−k) (7.62)
2πx
μ0 I1 I2 d x
= j
2π x
Total force is
 a+b
μ0 I1 I2 d x
FB = j (7.63)
a 2π x
μ0 I1 I2
= j [ln x]a+b
a
2π  
μ0 I1 I2 b
= j ln 1 +
2π a

(b) Force exerted on the bottom wire of length b is

 
μ0 I1 I2 b
FB = − j ln 1 + (7.64)
2π a

Exercise 7.8 Consider a rectangular loop of length b and width a located

close to a long wire carrying a current I , as shown in Fig. 7.16. Assume that a
distance c between the wire and the nearest side of the rectangle, and that the
wire is parallel to the long side of the rectangle. Find the total magnetic flux
through the loop due to the current in the wire.

Solution 7.8 The magnitude of the magnetic field created by the wire at a distance
r from the wire is
μ0 I
B= (7.65)
2πr
172 7 Sources of Magnetic Field

Fig. 7.16 A rectangular

loop of width a and length b
is located near a long wire
carrying a current I

The factor 1/r indicates that the field varies over the loop, and the field is directed into
the page. Because B and dA are parallel at any point within the loop, the magnetic
flux through a small surface element d A is

ΦB = B dA (7.66)

μ0 I
= dA
2πr

μ0 I
= bdr
2πr

Performing the integration

 a+c
μ0 I μ0 I b  a
ΦB = bdr = ln 1 + (7.67)
c 2πr 2π c

Apply the series expansion formula for ln(1 + x) ≈ x (for x  1) to this equation
to show that it gives a reasonable result when the loop is far from the wire relative
to the loop dimensions, c  a:
ΦB ≈ 0 (7.68)

Exercise 7.9 A sinusoidally varying voltage is applied across an 8.00 µF

capacitor. The angular frequency of the voltage is ω = 2.00 × 104 rad/s, and
the voltage amplitude is 30.0 V. Find the displacement current between the
plates of the capacitor.

Solution 7.9 The angular frequency of the source is

ω = 2π f = 2π(3.00 × 103 Hz) = 1.88 × 104 rad/s (7.69)

7.7 Exercises 173

Hence the voltage across the capacitor in terms of t is

ΔV = ΔVmax sin ωt = (30.0 V) sin(1.88 × 104 t) (7.70)

More, we know Q = CΔV . Thus,

dQ d d
Id = = (CΔV ) = C (ΔV ) (7.71)
dt dt dt
−6 d

= (8.00 × 10 F) (30.0 V) sin(1.88 × 104 t)

dt
= (4.52 A) cos(1.88 × 104 t)

The displacement current varies sinusoidally with time and has a maximum value of
4.52 A.

Exercise 7.10 In Niels Bohr’s 1913 model of the hydrogen atom, an electron
circles the proton at a distance of 5.29 × 10−11 m with a speed of 2.19 × 106
m/s. Compute the magnitude of the magnetic field that this motion produces
at the location of the proton.

Solution 7.10 The current is

| e | vd
I = (7.72)
2πr

where r = 5.29 × 10−11 m, vd = 2.19 × 106 m/s, and | e |= 1.6 × 10−19 C. Using
the Biot-Savart law:
μ0 I (ds × r̂)
dB = (7.73)
4π r2
Therefore, we obtain

μ0 I ds sin θ
dB = (7.74)
4π r2
μ0 I r sin θdθ
=
4π r2
Total magnetic field is
 π
μ0 I sin θdθ
B= (7.75)
0 4π r
μ0 I μ0 | e | vd
= =
2πr (2πr )2
174 7 Sources of Magnetic Field

(4π × 10−7 T · m/A)(1.6 × 10−19 C)(2.19 × 106 m/s)

=
(2π5.29 × 10−11 m)2
= 3.99 T

References

Holliday D, Resnick R, Walker J (2011) Fundamentals of physics. John Wiley and Sons
Jackson JD (1999) Classical electrodynamics, 3rd edn. John Wiley and Sons
Griffiths DJ (1999) Introduction to electrodynamics, 3rd edn. Prentice Hall
Chapter 8
Magnetism in Matter

The chapter aims to introduce Maxwell equations of

electromagnetism in free space and the medium.

In this chapter, we introduce magnetism in the matter. We first discuss the magnetic
moments of an atom of matter. Then, we will discuss the magnetic substances, such as
ferromagnetic, paramagnetic, and diamagnetic substances. Next, we will introduce
Faraday’s law of induction. Finally, we will introduce the famous Maxwell equations
of electromagnetism in free space and the medium. As extra reading material, the
reader can also consider other literature Jackson (1999), Landau and Lifshitz (1971),
Sykja (2006), Griffiths (1999).

8.1 Magnetic Moments of Atoms

The magnetic field created by a current-carrying coil of wire can explain what makes
some materials to have strong magnetic properties. In general, based on Biot–Savart
law (Eq. (7.1), Chap. 7), any current loop produces a magnetic field, as shown in
Fig. 8.1. Thus, it has a magnetic dipole moment, including the atomic-level current
loops described in some models of the atom. Those atomic-level current loops may
define the magnetic moments in a magnetized substance. In the Bohr model of the
atom, the current loops are associated with the circular motion of electrons around the
nucleus. Besides, another magnetic moment, which is intrinsic to electrons, protons,
neutrons, and other particles, arises from a property called spin.
Now, let us consider the classical model of the atom in which electrons move
in circular orbits around the nucleus. Each orbiting electron creates a current loop
because it is a moving charge e. Therefore, there exists a magnetic moment of the

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 175
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_8
176 8 Magnetism in Matter

Fig. 8.1 Magnetic field and

magnetic dipole moment
created by a current loop.
A = πa 2 is the surface area
of the loop and n is unit
normal vector to the surface
of the current loop

Fig. 8.2 Angular

momentum L = r × p
created by an orbiting
electron and the magnetic
dipole moment µ of the
current loop I

electron associated with its orbital motion (see also Fig. 8.2). Suppose electron is
moving with constant speed v in a circular orbit of radius r about the nucleus coun-
terclockwise, as shown in Fig. 8.2. During a full period T , the electron travels the
length 2πr , which is the circumference of the circle, and hence its speed is

2πr
v= (8.1)
T
The current I associated with this orbiting electron with a charge e is given as
e
I = (8.2)
T
Using the following relations:

2π
T = (8.3)
ω
v
ω=
r
we get
eω ev
I = = (8.4)
2π 2πr
8.1 Magnetic Moments of Atoms 177

Fig. 8.3 The orbital

magnetic moments µ
produced by two current
orbits of two electrons
moving in opposite direction

The magnitude of magnetic moment associated with this current loop is μ = I A

with a direction of the vector shown in Fig. 8.2, where A = πr 2 is the surface area
enclosed by the orbit. Therefore, we obtain the magnetic moment

ev 1
μ= IA= πr 2 = evr (8.5)
2πr 2
Knowing that the orbital angular momentum of electron is L = r × p and its mag-
nitude
L = m e vr sin 90◦ = m e vr (8.6)

the magnetic dipole moment can be written as

eL
μ= (8.7)
2m e

Equation (8.7) implies that the magnetic moment of the electron is proportional to
its orbital angular momentum. Note that because the electron is negatively charged,
the vectors μ and L point in opposite directions, as shown in Fig. 8.2. Both vectors
are perpendicular to the plane of the orbit. Quanti-
zation
Based on the quantum physics, the orbital angular momentum is quantized and it
is equal to multiples of , which relates to Planck’s constant h:

h
= = 1.05 × 10−34 J · s (8.8)
2π
The smallest non-zero value of the electron’s magnetic moment resulting from its
orbital motion is √ e
μ= 2  (8.9)
2m e

Although the orbital motion of electrons in atoms of every substance produces

magnetic dipole moments, not all materials are magnetic. That is because the orbital
magnetic moment created by an electron in an atom may be canceled by the orbital
magnetic moment of an electron in another atom that is moving in the opposite direc-
tion in orbit, as illustrated in Fig. 8.3. Therefore, the net magnetic effect produced by
the orbital motion of the electrons in atoms is approximately zero in some materials.
178 8 Magnetism in Matter

Fig. 8.4 Classical

description of the electron
viewed as spinning about its
axis while it orbits the
nucleus

In addition to the orbital magnetic moment, an electron has a spin that may also
contribute to its net magnetic moment. In classical description, the electron spins
about its axis while it moves around the nucleus in an orbit, as shown in Fig. 8.4. The
magnitudes of the angular momentum S of the spinning motion and of the angular
momentum L due to the orbital motion are of the orders given in the following. Based
on the quantum physics, the magnitude of the spin angular momentum is
√
3
S=  (8.10)
2
Therefore, the spin magnetic moment of an electron is

e
μspin = (8.11)
2m e

That is also called the Bohr magneton:

e
μB = = 9.27 × 10−24 J/T = 9.27 × 10−24 A · m2 (8.12)
2m e

Thus, atomic magnetic moments can be expressed as multiples of the Bohr magneton.
Usually, in an atom with many electrons, the electrons pair up with their spins
opposite to one another, giving a net spin magnetic moment of zero. However, in an
atom with an odd number of electrons, there exists at least one unpaired electron,
and hence there is a spin magnetic moment different from zero.
The total magnetic moment of an atom is the sum of the orbital and spin mag-
netic moment vectors. The nucleus of an atom also contributes with a net magnetic
moment that originates from protons and neutrons. However, in practice, the mag-
netic moments of the protons and neutrons are much smaller than that of the electrons
and can usually be omitted. For instance, the magnetic moment associated with a
proton or neutron is given as in Eq. (8.12), by replacing the mass of the electron with
the mass of a proton or neutron:
8.1 Magnetic Moments of Atoms 179

p,n e
μB = (8.13)
2m p,n

Since the masses of the proton (m p ) and neutron (m n ) are much greater than the mass
of an electron (m e ), we get
μeB m p,n
p,n = ∼ 103 (8.14)
μB me

Equation (8.14) indicates that the magnetic moments of the proton and neutron are
about 1000 times smaller than that of the electron.

8.2 Magnetization Vector and Magnetic Field Strength

To describe the magnetic properties of a substance, the magnetization vector M is

introduced. In a substance of volume V with a large number of molecules or atoms
per unit of volume, where each has a magnetic moment µi , the average macroscopic
magnetization vector is 
M(r) = n i µi  (8.15)
i

where n i is the average number density of molecules or atoms of type i and µi  is
the average magnetic moment in a macroscopically small volume element centered
at position r of type i. The magnetization contributes an effective current density:

JM = ∇ × M (8.16)

In the microscopic description, Ampére’s law for static electric fields can be rewritten
as  
Bmicr o · ds = μ0 Jmicr o · dA (8.17)
L A

where A is any surface enclosed by the path L. Using Stokes’ formula, we obtain
 
(∇ × Bmicr o ) · dA = μ0 Jmicr o · dA (8.18)
A A

The macroscopic equivalent of Eq. (8.18) is given by

 
(∇ × B) · dA = μ0 J · dA (8.19)
A A

where the macroscopic surface current density, J , is

 180 8 Magnetism in Matter

J = J + J M (8.20)

where J represents the surface charge density from the flow of free charge in the
medium. Substituting Eq. (8.20) into Eq. (8.19), we obtain
 
(∇ × B) · dA = μ0 (J + J M ) · dA (8.21)
A A

Comparing both sides in Eq. (8.21), we have

∇ × B = μ0 (J + ∇ × M) (8.22)

where Eq. (8.16) is used. Equation (8.22) can be re-arranged in a convenient form as

∇ × (B − μ0 M) = μ0 J (8.23)

The The contribution of the magnetization of the medium is expressed in terms of the
magne-
tization magnetization vector as
vector Bm = μ0 M (8.24)

The total magnetic field B at a point within a substance will be sum of both the
applied (external) field B0 and the magnetization of the substance, Bm :

B = B0 + Bm (8.25)

Therefore, the total magnetic field in the region where both the external magnetic
field and magnetization of the substance are present becomes

B = B0 + μ0 M (8.26)

To analyze the magnetic fields produced from the magnetization, we introduce

another field quantity, namely, the magnetic field strength H within the substance.
The magnetic field strength characterizes the effect of the conduction currents in
wires on a material. To indicate the differences between the magnetic field strength
H and the magnetic field B, often, B is called the magnetic flux density or the magnetic
The induction.
mag-
netic The magnetic field strength H is a vector quantity defined as
field
strength
B0 B
H= = −M (8.27)
μ0 μ0

Or,
B = μ0 (H + M) (8.28)
8.2 Magnetization Vector and Magnetic Field Strength 181

Both H and M have the same units. In SI units, M, as a magnetic moment per unit
volume, has the units of (ampere)(meter)2 /(meter)3 or equivalently amperes per
meter. Then, Eq. (8.23) can be written in terms of H as

∇ ×H=J (8.29)

which represents the differential form of Ampére’s law for magnetism in a medium
with a static electric field.

8.3 Classification of Magnetic Substances

Materials can be classified based on their magnetic properties as

1. Paramagnetic;
2. Diamagnetic;
3. Ferromagnetic.
Diamagnetic materials are made up of atoms that do not have permanent magnetic
moments. In contrast, paramagnetic and ferromagnetic materials are made up of
atoms that have permanent magnetic moments. For a paramagnetic and diamagnetic
material, the magnetization vector M depends linearly on the external magnetic field
strength H as
M = χH (8.30)

In Eq. (8.30), χ (Greek letter chi) is a dimensionless proportionality constant, called

the magnetic susceptibility. For a paramagnetic material, χ is positive, and hence M
is parallel with H. On the other hand, for a diamagnetic material, χ is negative, and
M points the opposite direction to H. The linearity of Eq. (8.30) between M and H
does not obey for ferromagnetic materials. Mag-
netic
Using the above equation: perme-
B = μ0 (H + M) (8.31) ability

we obtain
B = μ0 (H + χ H) = μ0 (1 + χ )H (8.32)

Or,
B = μm H (8.33)

where
μm = μ0 (1 + χ ) (8.34)

is called magnetic permeability of the material.

By comparing the magnetic permeability, μm , with the permeability of free space,
μ0 , materials classify as follows:
182 8 Magnetism in Matter

1. Paramagnetic: μm > μ0 ;
2. Diamagnetic: μm < μ0 .
Since χ is very small for paramagnetic and diamagnetic materials, μm ≈ μ0 for those
materials. However, for a ferromagnetic substance, μm is up to several thousand times
greater than μ0 , and hence χ is very high for ferromagnetic materials.

8.3.1 Ferromagnetism

In Fig. 8.5, we show a substance in which the atoms have permanent magnetic
moments. That material exhibits strong magnetic effects called ferromagnetism. Such
ferromagnetic substances include iron, cobalt, and nickel. In a ferromagnetic mate-
rial, there exist atomic magnetic moments (see Fig. 8.5) that when we put them in an
external magnetic field, the atomic magnetic moments align parallel to one another
and external magnetic field. Interestingly, the magnetic moments remain aligned
after the external field is removed due to strong interactions between neighboring
moments. Therefore, we say that the material remains magnetized.
Ferromagnetic materials are made up of microscopic domains with aligned mag-
netic moments of about 1017 to 1021 atoms. The volume of each of those domains is
about 10−12 to 10−8 m3 . There exist walls between the domains. When the material
is not magnetized, the net microscopic magnetic moments of domains are randomly
oriented, and hence the net macroscopic magnetic moment vanishes. When an exter-
nal magnetic field is applied across the sample, the microscopic magnetic moments
of the domains tend to align with the field, yielding a non-zero net macroscopic mag-
netization of the sample. When the external magnetic field reduces slowly to zero,
the sample may have a net magnetization in the direction of the applied external field.
Usually, at standard temperatures, thermal fluctuations are not significantly large to
cause any changes to that preferred orientation of domain magnetic moments.

Fig. 8.5 a Orientation of

permanent magnetic
moments of a ferromagnetic
material in absence of
external magnetic field and b
in presence of an external
magnetic field
8.3 Classification of Magnetic Substances 183

8.3.2 Paramagnetism

Paramagnetic materials are characterized by a small positive magnetic susceptibility

0 < χ
1 that is due to the existence of the permanent magnetic moments of atoms
or ions. The dipole moments interact very weakly with one another, and hence they
are randomly oriented when there is no external magnetic field. On the other hand,
atomic magnetic moments of the paramagnetic line up with the field when an external
magnetic field is present. The alignment process, however, is opposed by the thermal
motion, which tends to randomize the magnetic moment orientations.
It was found experimentally by Pierre Curie and others that, under a wide range
of conditions, the magnetization of a paramagnetic is proportional to the applied
magnetic field B0 and inversely proportional to the temperature T (in Kelvin):

B0
M =C (8.35)
T
Equation (8.51) is known as Curie’s law, and the constant C is called Curie’s constant.
The law implies that when B0 = 0, the magnetization is zero, corresponding to a ran-
dom orientation of magnetic moments. With increasing the ratio of the magnetic field
to temperature, the magnetization approaches a maximum value, which corresponds
to the complete alignment of its moments, and Eq. (8.51) becomes invalid.
When the temperature of a ferromagnetic substance is greater or equal to the
critical Curie temperature, TC , the substance loses its residual magnetization and
becomes paramagnetic. Below TC , the magnetic moments are aligned and the sub-
stance is ferromagnetic. For T > TC , the thermal fluctuations are high such that they
cause a random orientation of the magnetic moments, and the substance becomes
paramagnetic, as shown in Fig. 8.6.

Fig. 8.6 Magnetization

versus temperature for a
ferromagnetic and
paramagnetic
184 8 Magnetism in Matter

8.3.3 Diamagnetism

When an external magnetic field B0 is applied, a weak magnetic moment is induced

opposite to the applied field direction, and a magnet only weakly repels the dia-
magnetic substances. Usually, the ferromagnetic and paramagnetic effects are much
more significant than diamagnetic effects in all materials. Therefore, those effects
are dominant only when the ferromagnetic and paramagnetic effects do not exist.
To understand the diamagnetism, we may consider a classical model of two atomic
electrons orbiting around the nucleus with the same speed in opposite directions. The
electrons continue their circular orbits around the nucleus because of the attractive
electrostatic force exerted by the positive charge of the nucleus. Since the magnetic
moments of the two electrons are equal in magnitude but have opposite directions,
their net magnetic moment in the atom is zero. However, when an external magnetic
field is applied, it exerts an additional force on the electrons, qv × B. That magnetic
force combines with the electrostatic force increasing the orbital motion linear speed
of the electron with a magnetic moment opposite to the field and in a decrease of that
speed when the magnetic moment of the electron is parallel to the field. Therefore,
the two magnetic moments of the electrons no longer cancel, and the substance gains
a net magnetic moment opposite of the applied field.
As we mentioned, the superconductor is a material with zero electrical resistance
below some critical temperature. There exist some types of superconductors that
exhibit perfect diamagnetism in the superconducting state. Hence, the applied mag-
netic field is excluded by the superconductor such that the field is zero in its interior,
known as the Meissner effect. When a permanent magnet is nearby a superconductor,
they repel one another.

8.4 The Magnetic Field of the Earth

Usually, we say about a compass magnet having a “north-seeking” pole and a “south-
seeking” pole. That is, one pole of the magnet seeks or points to the north geographic
pole of the Earth.
Since the north pole of a magnet is attracted by the north geographic pole of the
Earth, we have that the Earth’s south magnetic pole is located near the north geo-
graphic pole, and the Earth’s north magnetic pole is located near the south geographic
pole. The configuration of the Earth’s magnetic field is very much like the one that
would be achieved by burying a gigantic bar magnet deep in the interior of the Earth.
8.5 Faraday’s Law of Induction 185

8.5 Faraday’s Law of Induction

In 1831, Faraday was the first who observed quantitatively the phenomena related to
time-dependent electric and magnetic fields. In particular, the behavior of currents in
circuits was observed when placed in a time-varying magnetic field. Faraday found
that a transient current is induced in a loop if the steady current flow in an adjacent
circuit is turned on or off. Also, he observed that when the circuit moves relative to
the circuit in which a constant current is flowing, then a current is induced in the
moving circuit. Moreover, Faraday observed that when a magnet is approaching or
moving away from a circuit, then a current is produced in the circuit. Similarly, he
found that no current is induced when the current flow on the second circuit was not
changing or when either the second circuit or magnet was not moving relative to the
first circuit. Fara-
day’s
Faraday explained the observation of the induced current with the change of the Law
magnetic flux linked by the circuit. That is, the change in the magnetic flux induces
an electric field around the circuit; the line integral of the induced electric field yields
a potential difference, called electromotive force, ε. Then, the electromotive force
produces a current flow based on Ohm’s law.
To obtain a mathematical formulation of Faraday’s law, we consider the circuit L
bounded by an open surface A with unit normal vector n, as shown in Fig. 8.7. Fur-
thermore, the magnetic field B near the circuit is shown. The magnetic flux through
the surface of circuit is given by

ΦB = B · dA (8.36)
A

where dA = nd A with d A being a small surface element (see Fig. 8.7). The elec-
tromotive force around the circuit or induced voltage is given as
   
ε = − − Eind · ds = Eind · ds (8.37)
L L
Lenz’s
The first minus sign in Eq. (8.37) indicates that the polarity of induced electro- Law
motive force, ε, is opposing the change on the magnetic flux, Φ B . That is, induced
electromotive force produces a current in the circuit, which creates a magnetic field,
based on Biot–Savart’s law, to oppose the change in the magnetic flux Φ B through
the area enclosed by the current circuit. That is known as Lenz’s law.
In Eq. (8.37), Eind is the induced electric field at ds. Faraday’s law, mathematically,
is formulated as1 :
dΦ B
ε=− (8.38)
dt

dΦ B
1 In general, Faraday’s law is written as ε = −k , where the constant k is used to adjust the
dt
units. For SI units, k = 1, and for Gaussian units, k = 1/c, where c is the speed of light. Here, we
use SI units, and thus k = 1.
186 8 Magnetism in Matter

Fig. 8.7 Magnetic field lines

passing through an arbitrary
surface A supported by the
contour L

where the magnetic flux is given by Eq. (8.36).

In particular, in the following, we consider a uniform magnetic field B. Then,
from Eq. (8.36), the magnetic flux is

Φ B = B · A = B A cos θ (8.39)

θ is the angle formed by the magnetic field vector and unit normal vector n. Equa-
tion (8.39) implies that Φ B = 0, when θ = 90◦ , that is, when the magnetic field B
and the open surface are parallel. Furthermore, Φ B,max = B A, maximum value of the
magnetic flux is obtained for θ = 0◦ , that is, when magnetic field B is perpendicular
to the open surface. In addition, Φ B is between zero and its maximum value, Φ B,max ,
for any other angle θ .

Moreover, Eq. (8.39) indicates that to change the magnetic flux Φ B value, and
hence to cause induction of an electromotive force, ε, we can perform it in
different ways:
1. By changing the magnitude of the magnetic field B and keeping constant
A and θ .
2. By fixing the magnitude of magnetic field B and surface area of the circuit
A and changing its orientation with respect to B, that is, the angle θ .
3. By changing the surface area A of the circuit and fixing the magnitude B
and the orientation of the surface relative to B, that is, the angle θ .
4. In addition, since the magnetic flux is equal to the net magnetic field lines
passing through the circuit surface, then we can cause induced electromotive
force by keeping B, A, and θ constant, but moving the circuit (either moving
it closer to the magnet or moving it away) to change the number of magnetic
field lines passing through the surface (either increasing or decreasing it).
8.5 Faraday’s Law of Induction 187

To explain the origin of the electromotive force ε, consider a conductor moving to

the right with constant velocity v in a uniform magnetic field B, which has a direction
in the viewing plane, as shown in Fig. 8.8. Knowing that in a conductor there are two
kind of charges, namely, the free electrons (negative charge) and positively charged
atoms, which move along with the conductor, then there is a Lorentz force acting
on both electrons and positive ions: F B = q(v × B), where q = −e for electrons
and q = +e for positive ions. Here, e is the magnitude of the charge of an electron.
The direction of the force acting on the electron and positive charge is indicated
in Fig. 8.8; they cause the electron to move down and positively charge atoms up.
Therefore, after some time, there will be a collection of negative charges (electrons) at
the lower end and positive charges at the upper end, as shown in Fig. 8.8. That causes
a potential difference between the upper and lower sides of the conductor, φ+ − φ− ,
which by definition is the electromotive force (or induced potential difference), ε:

ε = φ+ − φ − (8.40)

That is equivalent to the appearance of an induced current Iind , which is opposite to

the direction of the movement of electrons, that is, it flows upward (see Fig. 8.8). Thus,
in addition to magnetic force, on the charges is acting the electric force: Fe = qEind ,
which has the opposite direction to the magnetic force. The charges will continue
accumulating at each end of the conductor until the two forces balance each other.
Therefore, based on Newton’s third law:

F B = −Fe (8.41)

or
qv B = q E ind (8.42)

From Eq. (8.42), we obtain induced electric field:

E ind = v B (8.43)

Note that after the electrons stop moving, the induced electric field becomes constant
and produces a potential difference between the upper end and the lower end, which
is the induced electromotive force, given as

ε = L E ind = Lv B (8.44)

which is positive, and hence φ+ > φ− . Note that this polarity will change if the
direction of the movement of the conductor reverses, that is, if the conductor moves
to the left.
Note that this is in agreement with Faraday’s viewpoint. That is, the effect of the
electromotive force is to induce a current in the conductor to oppose the changes,
that is, to oppose the movement of the conductor to the right. Indeed, since in the
conductor the current Iind is induced, then the magnetic field exerts a force on the
188 8 Magnetism in Matter

Fig. 8.8 A conductor in a

uniform magnetic field B
point in the viewing plane.
Conductor is moving to the
right with constant velocity v

Fig. 8.9 The magnetic force

F B on the conductor moving
to the right with constant
velocity v in which is
induced a current Iind

current conductor, based on previous sections, which is given as (see also Fig. 8.9)

F B = Iind (L × B) (8.45)

As seen in Fig. 8.9, the direction of this force opposes the movement of the conductor
to the right.
Combining Eq. (8.37) and Eq. (8.38), we obtain Faraday’s law in a general form:
 
dΦ B d
Eind · ds = − =− B · dA (8.46)
L dt dt A

8.6 Rowland Ring Apparatus

Consider a ferromagnetic material, which consists of a torus made up of some material

(for example, iron) within N turns of wire, as shown in Fig. 8.10. Another coil is
connected to a galvanometer (G), which is used to measure the total magnetic flux
through the torus.
To measure the magnetic field B in the torus, the current in the toroid increases
from zero to some maximum value I :
8.6 Rowland Ring Apparatus 189

Fig. 8.10 Rowland ring

apparatus

B = μ0 N I /(2πr ) (8.47)

B · ds = μ0 N I

As the current increases, the magnetic flux through the second coil connected to G
changes by
ΦB = B A (8.48)

In Eq. (8.48), A is the cross-sectional surface area of the toroid. Based on Faraday’s
law, the change on flux induces an electromotive force in the secondary coil:

dΦ B
| ε |= (8.49)
dt
If the galvanometer is calibrated in advance, the value of the current in the primary
coil corresponds to a value of B. First, the magnetic field B is measured in the absence
of the torus, and then in the presence of the torus. In that way, the magnetic properties
of the torus material can be obtained by comparing the magnetic field in the two
measurements.
Consider a torus made of iron with no magnetization. When the current in the
primary coil increases from zero to its maximum value I , the magnitude of the
magnetic field strength H increases according to

H = nI (8.50)

Fig. 8.11 shows that the magnitude of the total field B also increases with I , along
the curve from point O to point a. At O, the domains in the iron are randomly
oriented, and hence Bm = 0. With increasing further the current in the primary coil,
190 8 Magnetism in Matter

Fig. 8.11 Magnetization

curve

the external field B0 increases, and the domains become nearly aligned at point a. At
point a, the iron core is approaching saturation in which all domains in the iron are
aligned. If the current decreases slowly to zero, the external field B0 also decreases
to zero, following the path ab, as shown in Fig. 8.11. At the point, b, B = 0, only
the external field is B0 = 0 because the iron is magnetized due to the alignment of
a large number of its domains (that is, B = Bm ). Therefore, the iron has a remanent
magnetization. The curve giving B as a function of H is called the magnetization
curve.
Now, if the current in the primary coil is reversed, then the external magnetic field
B0 direction is also changed. Therefore, the domains experience an opposite direction
torque to reorient; when the point c is reached, the sample is again unmagnetized
B = 0. A further increase of the current in the reverse direction produces an opposite
direction magnetization of the iron compared with that at point b. The saturation point
is reached at d. The same sequence of events occurs when the current decreases to
zero and then increases back in the original direction, shown by the path de f in
Fig. 8.11.
Further increase in the current returns the magnetization curve to point a, with
a maximum magnetization of the sample. That effect is called magnetic hysteresis,
and the closed loop is known as the hysteresis loop. That is, the magnetization of
a ferromagnetic substance depends on both the history of magnetization and on the
magnitude of the applied external field. Moreover, a ferromagnetic substance has
a memory that makes the material to remain magnetized after the external field is
removed.
The shape and size of the hysteresis loop depend on the ferromagnetic properties
and in addition to the strength of the applied magnetic field. For hard ferromagnetic
materials it is wide, that is, a large remanent magnetization (see Fig. 8.12). Those
materials are more difficult to be demagnetized by an external field. For soft fer-
romagnetic materials, such as iron, the hysteresis loop is very narrow, and hence
it has a small remanent magnetization. Those materials are easily magnetized and
demagnetized.
Ideally, a soft ferromagnet may not exhibit a hysteresis, and hence it would have no
remanent magnetization. Furthermore, a ferromagnetic sample can be demagnetized
8.6 Rowland Ring Apparatus 191

Fig. 8.12 Magnetic hysteresis loop

Fig. 8.13 Illustration of the

demagnetization of a
ferromagnetic material using
successive hysteresis loops

using successive hysteresis loops, by decreasing applied magnetic field, as shown in

Fig. 8.13. Moreover, the magnitude of the surface area enclosed by the magnetization
curve corresponds to the work required to perform one hysteresis loop. That energy
transferred to the material during a magnetization process comes from the source
of the external field, such as the em f in the circuit of the toroidal coil, that is, the
electrical energy transfers into magnetic energy.
When a magnetization cycle repeats, magnetic energy transfers into internal
energy due to the dissipation processes within the material as the domains try to
re-align. That will increase the temperature of the substance. Therefore, electric
devices, such as alternating current adapters for cell phones, or power tools, subject
to alternating fields, use soft ferromagnetic substance cores, having narrow hysteresis
loops, and hence little energy loss per cycle.
Magnetic computer disks store information by alternating the direction of B in a
thin layer of ferromagnetic material. Old floppy disks are constructed of a layer on a
circular plastic sheet. Hard drives are made up of several rigid platters, each with a
magnetic coating on each side. Also, audiotapes and videotapes are similar to floppy
disks with a ferromagnetic material on a very long strip of plastic.
192 8 Magnetism in Matter

8.7 Maxwell’s Equations of Magnetism

Now, we present Maxwell’s equations that characterize the magnetic phenomena.

These laws include all the laws of the magnetism discussed in this chapter. Thus, we
assume here that E = 0.
First, we start with Maxwell’s equations in free space. In the following, we write
these equations in the integral form:

B · ds = μ0 I (8.51)
L

B · dA = 0
A

In Eq. (8.51), the first Maxwell’s equation is Ampére’s law, where the first term
on the right-hand side gives the net current through the open surface enclosed by
the contour L. The second Maxwell’s equation in Eq. (8.51) implies that magnetic
field flux through a closed surface is equal to zero. Alternatively, the net number
of magnetic field lines passing through a closed surface is zero, that is, so many
magnetic lines are leaving the closed surface as entering it. That indicates that there
do not exist free magnetic poles.
It is important to note that in Eq. (8.51) we have assumed that B is only a function
of the position r. In the next chapter (Chap. 9), we will discuss the magnetic and
electrostatic fields that depend on both position r and time t.
Equation (8.51) can also be written in a differential form. For instance, the dif-
ferential form of the first Maxwell’s equation is defined using Stokes’ formula and
current density vector J as follows:
 
B · ds = (∇ × B) · dA (8.52)
L
 A

= μ0 J · dA
A

Comparing both sides in Eq. (8.52), we obtain the first Maxwell’s equation of mag-
netism in the following differential form:

∇ × B = μ0 J (8.53)

To derive the differential form of the second Maxwell’s equation of magnetism,

we use Gauss’s formula in the last expression of Eq. (8.51), as
 
B · dA = ∇ · B dV = 0 (8.54)
A V

Here, V is the volume enclosed by the surface A. Therefore, we obtain

8.7 Maxwell’s Equations of Magnetism 193

∇ ·B=0 (8.55)

Finally, we can summarize Maxwell’s equations of magnetism in the following

differential form:

∇ × B = μ0 J (8.56)
∇ ·B=0 (8.57)

Next, we will consider Maxwell’s equations of magnetism in a medium, which

are written in terms of the magnetic field strength H:

B = μ0 (H + M) (8.58)

where M is magnetization vector.

To obtain the first Maxwell’s equation of magnetism in the media, we start with
differential form of Ampére’s law :

∇ × B = μ0 ∇ × (H + M) (8.59)
= μ0 (∇ × H) + μ0 (∇ × M)
= μ0 J

where J is the equivalent macroscopic current, which is sum of the microscopic

current J and an effective current density due to magnetization:

JM = ∇ × M (8.60)

Combining Eqs. (8.59) and (8.60), we obtain

μ0 (∇ × H) = μ0 J (8.61)

or
∇ ×H=J (8.62)

The second Maxwell’s equation of magnetism in a medium remains the same

expression as in Eq. (8.57).

8.8 Vector Potential

We return to Biot–Savart law, and rewrite it as follows (refer also to Fig. 8.14):
194 8 Magnetism in Matter

Fig. 8.14 Biot–Savart law


μ0 J(r )
B= ∇× dr (8.63)
4π | r − r |
V

where we have used that

 
1 r − r
∇ =− (8.64)
| r − r | | r − r  |3

and   

I ds × r̂  (r − r ) × J(r )
= −dr (8.65)
| r − r  |2 | r − r  |3

where d V = dr is a small volume element, as indicated in Fig. 8.14. We can now
introduce a vector potential of the magnetic field as

μ0 J(r )
A(r) = dr (8.66)
4π | r − r |
V

and the magnetic field can be written as

B = ∇ × A(r) (8.67)

In general, the vector potential is defined up to the gradient of an arbitrary scalar

function Ψ (r), that is,

μ0 J(r )
A(r) = dr + ∇Ψ (r) (8.68)
4π | r − r |
V
8.8 Vector Potential 195

Equation (8.68) indicates that vector potential can be transformed as

A(r) → A(r) + ∇Ψ (r) (8.69)

which is known as gauge transformation. Transformation of the form given by

Eq. (8.69) is allowed, since they do not change the magnetic field. For instance,
using transformation in Eq. (8.69) and definition of the magnetic field in Eq. (8.67),
we get
B = ∇ × (A(r) + ∇Ψ (r)) = ∇ × A(r) (8.70)

since ∇ × ∇Ψ (r) = 0.
Consider the differential form of Maxwell’s equation of magnetism given by
Eq. (8.53):
∇ × B = μ0 J (8.71)

Substituting Eq. (8.67) into Eq. (8.71), we get

∇ × (∇ × A) = μ0 J (8.72)

Simplifying the left-hand side of Eq. (8.72), we obtain

∇(∇ · A) − ∇ 2 A = μ0 J (8.73)

The freedom of gauge transformations allows to make another convenient choice of

gauge as ∇ · A = 0, then the vector potential satisfies Poisson equation:

∇ 2 A = −μ0 J (8.74)

Note that the choice ∇ · A = 0, using Eq. (8.69), implies that ∇ 2 Ψ = 0. Therefore,
if ∇ 2 Ψ = 0 in all space, then Ψ is a constant assuming there are no currents at the
infinity.

8.9 Multipole Expansion

Equation (8.66) gives the vector potential of the magnetic field in terms of the current
density J(r ) in a localized finite volume V . Furthermore, J(r ) is zero outside the
volume. Suppose that we are interested on finding A outside that volume. For that,
similar to scalar potential in electrostatics, we expand the term 1/ | r − r | around
r = 0 using Taylor expansion, as given by Eq. (3.50) (Chap. 3).
Assuming that | r | | r |, we can rewrite Eq. (8.66) as follows:
196 8 Magnetism in Matter
⎛ ⎞
     2
μ0 1 r 
· r 1
3
3x x − δi j (r )
J(r ) ⎝ + 3 + xi x j + · · · ⎠ dr
i j
A(r) =
4π r r 2 i, j=1 r5
V
(8.75)
Equation (8.75) can be considered sum of three contributions, namely, A0 , A1 and
A3 , if we neglect higher order term in the expansion.
The first term, which corresponds to the monopole term in the electrostatic expan-
sion, is  
μ0   μ0  
A0 (r) = J(r )dr = ∇ · J(r ) d A (8.76)
4πr 4πr
V A

where the integration is over the surface enclosing the volume V and Stokes’ formula
is used. Using the continuity equation of the current density (ρ = 0):

∇ ·J=0 (8.77)

we obtain
A0 (r) = 0 (8.78)

which indicates that there are no isolated magnetic monopoles (or magnetic charges).
To calculate the second term of the expansion, we use the following mathematical
relation:
c × (a × b) = (b · c)a − b(c · a) + a(c · b) − (a · c)b (8.79)

with r ≡ c, r ≡ a and J ≡ b, we obtain

r × (r × J) = (J · r)r − J(r · r ) + r (r · J) − (r · r)J (8.80)

= 2(J · r)r − 2J(r · r)

Or,
1
J(r · r) = (J · r)r − r × (r × J) (8.81)
2
Substituting Eq. (8.81) into the second term of Eq. (8.75), we obtain

μ0 r · r 
A1 (r) = J dr (8.82)
4π r3
V
⎛ ⎞
  
μ0 r  μ0 1 r × J
= (J · r) 3 dr − r×⎝ dr ⎠

4π r 4π 2 r3
V V

The first integral in Eq. (8.82) vanishes since the integrand is an odd function of r .
Furthermore, we define the magnetic moment density or magnetization as
8.9 Multipole Expansion 197

1
M(r) = (r × J(r)) (8.83)
2
and its integral over the volume gives the magnetic dipole moment:

1   
m= r × J dr (8.84)
2
V

Therefore, Eq. (8.82) can be written as

μ0 m × r
A1 (r) = (8.85)
4π r 3
which is the lowest non-zero term in the expansion of A for a steady current distribu-
tion in a localized volume V . To evaluate the third term, we introduce the magnetic
quadrupole moment components as follows:

 
Q i jk = Jk (r ) 3xi x j − δi j (r  )2 dr (8.86)
V

where the third index k expresses the anisotropic distribution of the current along x-,
y-, and z-axes. Then, the third term of the expansion becomes

μ0 
3
A2 (r) = Q i jk xi x j x̂k (8.87)
8π i, j,k=1

where x̂k (for k = 1, 2, 3) are unit vectors along x-, y-, and z-axes.
To calculate the expansion terms of the magnetic field, we use Eq. (8.67) with
vector potential expanded according to Eq. (8.75). We consider only the first non-
vanishing term of the expansion, and write the first term of expansion of B as
μ0  r 
B1 (r) = ∇ × A1 = ∇ ×m× 3 (8.88)
4π r
r
Using Eq. (8.79), where ∇ ≡ c, m ≡ a and
≡ b, we can write
r3
 r r  r  r r
∇ × m × 3 = 3 · ∇ m − 3 (∇ · m) + m ∇ · 3 − (m · ∇) 3 (8.89)
r r r r r
where
198 8 Magnetism in Matter


3
 
(r · ∇) m = xi ∇i m j x̂ j = 0 (8.90)
i, j=1


3
r (∇ · m) = x j (∇i m i ) x̂ j = 0 (8.91)
i, j=1

because m is independent on r. Furthermore,

  
r ∂ x ∂ y ∂ z
m ∇· 3 =m + + (8.92)
r ∂x r3 ∂y r3 ∂z r 3
 2 
r − 3x 2 r 2 − 3y 2 r 2 − 3z 2
=m + +
r5 r5 r5
=0

and
 
r ∂ ∂ ∂ xi + yj + zk
(m · ∇) 3 = m x + my + mz (8.93)
r ∂x ∂y ∂z r3
 2 
r − 3x 2
3x y 3x z
= mx 5
i− 5 j− 5 k
r r r
 
3x y r 2 − 3y 2 3yz
+ my − 5 i + j − k
r r5 r5
 
3x z 3yz r 2 − 3z 2
+ mz − 5 i − 5 j + k
r r r5
m m·r m·r m·r
= 3 − 3 5 xi − 3 5 yj − 3 5 zk
r r r r
m (m · r̂)r̂
= 3 −3
r r3

where r̂ is a unit vector along r, that is, r = r r̂. Combining Eqs. (8.88), (8.89), (8.90),
(8.92), and (8.93), we obtain

μ0 3(m · r̂)r̂ − m
B1 (r) = ∇ × A1 = (8.94)
4π r3
Equation (8.94) gives the magnetic field created by a magnetic dipole moment,
which has the same form as the electric field of an electric dipole (see also the second
term in Eq. (3.64), Chap. 3).
8.10 Energy of the Magnetic Field 199

8.10 Energy of the Magnetic Field

Consider a steady current-carrying single circuit. When the magnetic flux through
the circuit changes, an electromotive force ε is induced around it, based on Faraday’s
law. To maintain a constant current in the circuit, the external sources, such as the
battery, must do work. The rate change of the work is

dW dΦ B
= −I ε = I (8.95)
dt dt
where Φ B is the magnetic flux through the circuit and the negative sign is due to
Lenz’s law. Form Eq. (8.95), the work done by the sources to keep current constant
for a change of the magnetic flux with dΦ B is

δW = I δΦ B (8.96)

Consider an element of the circuit with cross-sectional surface area ΔS perpendicular

to the direction of the current flow, then I = J ΔS. Then, Eq. (8.96) can be written
as 
Δ(δW ) = J ΔS δB · dA (8.97)
A

where A is the surface area of circuit, and dA = nd A is an small surface element

with n a unit normal vector to the surface of the circuit. Using Eq. (8.67), we obtain

Δ(δW ) = J ΔS (∇ × δA) · nd A (8.98)
A

Applying Stokes’ theorem, we write Eq. (8.98) in the following form:


Δ(δW ) = J ΔS δA · ds (8.99)
L

where L is a closed contour line of the portion of circuit and ds is a small element
in this contour parallel to the current density vector J. Therefore, J ΔSds = Jd V ,
where d V = dr is a volume element. Summing up all those closed path portion of
the circuit, we obtain the total increment of work done by the external sources due
to a change of magnetic field δA:

δW = δA(r) · J dr (8.100)
V

Using Ampére’s law, Eq. (8.62) for the static electric field, we write Eq. (8.100) as
200 8 Magnetism in Matter

δW = δA(r) · (∇ × H) dr (8.101)
V

Using the following vector identity:

∇ · (a × b) = b · (∇ × a) − a · (∇ × b) (8.102)

Equation (8.101) reduces to


δW = (H · (∇ × δA) + ∇ · (H × δA)) dr (8.103)
V

If we assume that the magnetic field distribution is localized, the second term is zero.
Therefore, using δB = ∇ × δA, we obtain

δW = H · δB dr (8.104)
V

For paramagnetic or diamagnetic medium, H and B are linearly dependent, and hence

1
H · δB = δ(H · B) (8.105)
2
Substituting Eq. (8.105) into Eq. (8.104), we get

1
δW = δ(H · B) dr (8.106)
2
V

If we consider a change of the fields from zero to their final values, we obtain the
total magnetic energy: 
1
W = H · B dr (8.107)
2
V

Equation (8.107) gives the magnetic analogue of the electrostatic energy.

We can obtain another expression for the magnetic energy analogue to the elec-
trostatic energy given in terms of the charge and electric potential. For that, we can
use Eq. (8.100), and assuming a linear relation between J and A:

1
δA · J = δ(J · A) (8.108)
2
Substituting Eq. (8.108) into Eq. (8.100), we get
8.10 Energy of the Magnetic Field 201

1
δW = δ(J · A) dr (8.109)
2
V

Summing up all small changes of the field, we obtain the total magnetic energy in
another analogue form: 
1
W = J · A dr (8.110)
2
V

8.11 Exercises

Exercise 8.1 A toroid wound with 60.0 turns/m of wire carries a current of
5.00 A. The torus is iron, which has a magnetic permeability of μm = 5000μ0
under the given conditions. Find H and B inside the iron.

Solution 8.1 Consider the torus region of a toroid that carries a current I . If this
region is a vacuum, M = 0 (because no magnetic material is present), the total
magnetic field is that arising from the current alone, and B = B0 = μ0 H. Because
B0 = μ0 n I in the torus region, where n is the number of turns per unit length of the
toroid,
H = B0 /μ0 = μ0 n I /μ0 (8.111)

or
A · tur ns
H = n I = (60.0 tur ns/m)(5.00 A) = 300 (8.112)
m
and

B = μm H = 5000μ0 H (8.113)
T·m A · tur ns
= (5000)(4π × 10−7 )(300 ) = 1.88 T
A m
This value of B is 5000 times the value in the absence of iron.

Exercise 8.2 Estimate the saturation magnetization in a long cylinder of iron,

assuming one unpaired electron spin per atom.
202 8 Magnetism in Matter

Solution 8.2 The saturation magnetization is obtained when all the magnetic
moments in the sample are aligned. If the sample contains n atoms per unit vol-
ume, then the saturation magnetization is

M S = nμ (8.114)

where μ is the magnetic moment of atom. Because the molar mass of iron is 55 g/mol
and its density is 7.9 g/cm3 , the value of n for iron is 8.6 × 1028 atoms/m3 . Assuming
that each atom contributes one Bohr magneton (due to one unpaired spin) to the
magnetic moment, we obtain

−24 A · m
2
atoms
M S = (8.6 × 1028 )(9.27 × 10 ) (8.115)
m3 atom
= 8.0 × 105 A/m

Exercise 8.3 The current in the long, straight wire is I1 = 5.00 A, and the
wire lies in the plane of the rectangular loop, which carries 10.0 A, as shown
in Fig. 8.15. The dimensions are c = 0.100 m, a = 0.150 m, and  = 0.450
m. Find the magnitude and direction of the net force exerted on the loop by
the magnetic field created by the wire.

Solution 8.3 The magnitude of magnetic field created by the current I1 at any dis-
tance r is
μ0 I1
B1 = (8.116)
2πr
and the direction is tangent to the circle in the place perpendicular to the direction of
the current I1 . The force exerted by B1 on the I2 is given as F = I2 L × B1 where L
is along the direction of I2 . Thus, the forces on the horizontal directions of the loop
where I2 passes (i.e., 12 and 34) are zero, and it is different from zero only on the
vertical directions (41 and 23). The directions of these two forces are shown in the
figure. Their magnitudes are

μ0 I1 I2
F41 = I2 B1 sin 90◦ =  (8.117)
2π c
μ0 I1 I2
F23 = I2 B1 sin 90◦ = 
2π(a + c)

Replacing the numerical values, we get

8.11 Exercises 203

Fig. 8.15 The current in the

long, straight wire is
I1 = 5.00 A, and the wire
lies in the plane of the
rectangular loop, which
carries 10.0 A

(4π × 10−7 T · m/A)(5.00 A)(10.0 A)

F41 = (0.450 m) (8.118)
2π(0.100 m)
= 0.450 × 10−4 N
(4π × 10−7 T · m/A)(5.00 A)(10.0 A)
F23 = (0.450 m)
2π(0.250 m)
= 1.80 × 10−5 N

The resultant force is then given as

F = −4.50 × 10−5 i N + 1.80 × 10−5 i N (8.119)

= −2.70 × 10−5 i N

Sign (−) indicates that force has opposite direction to positive x-axis.

Exercise 8.4 Two long, parallel conductors separated by 10.0 cm carry cur-
rents in the same direction (see Fig. 8.16). The first wire carries current
I1 = 5.00 A, and the second carries I2 = 8.00 A. (a) What is the magnitude
of the magnetic field created by I1 and acting on I2 ? (b) What is the force per
unit length exerted on I2 by I1 ? (c) What is the magnitude of the magnetic
field created by I2 at the location of I1 ? (d) What is the force per unit length
exerted by I2 on I1 ?

Solution 8.4 In Fig. 8.16, we have shown direction of the magnetic fields created
by the two currents and the forces they exert on each other.
(a) The magnitude of B1 is
204 8 Magnetism in Matter

Fig. 8.16 Two long, parallel

conductors separated by
10.0 cm carry currents in the
same direction

μ0 I1
B1 = (8.120)
2πr
(4π × 10−7 T · m/A)(5.00 A)
=
2π(0.100 m)
= 1.00 × 10−5 T

(b) The magnetic force is

F12 = I2 L × B1 (8.121)

and the magnitude:

F12 = I2 L B1 (8.122)

and its magnitude per unit length is

F12
f 12 = = I2 B1 (8.123)
L
μ0 I1 I2
=
2πr
(4π × 10−7 T · m/A)(5.00 A)(8.00 A)
=
2π(0.100 m)
= 8.00 × 10−5 N

(c) The magnitude of B2 is

μ0 I2
B2 = (8.124)
2πr
(4π × 10−7 T · m/A)(8.00 A)
=
2π(0.100 m)
= 1.60 × 10−5 T
8.11 Exercises 205

(d) The magnetic force is

F21 = I1 L × B2 (8.125)

and the magnitude:

F21 = I1 L B2 (8.126)

and its magnitude per unit length is

F21
f 21 = = I1 B2 (8.127)
L
μ0 I1 I2
=
2πr
(4π × 10−7 T · m/A)(5.00 A)(8.00 A)
=
2π(0.100 m)
−5
= 8.00 × 10 N

Note that the forces have the same magnitude, but have opposite direction, in agree-
ment with the third law of Newton.

Exercise 8.5 A circular wire loop of radius r carries a current I , as shown in

Fig. 8.17. What is the magnitude of the magnetic field at its center?

Solution 8.5 Using Biot–Savart law:

μ0 I ds × r̂
dB = (8.128)
4π r 2
For the magnitude, we obtain
μ0 I ds
dB = (8.129)
4π r 2
Total magnetic field is

μ0 I ds
B= (8.130)
4π r2
μ0 I
= (2πr )
4πr 2
μ0 I
=
2r
206 8 Magnetism in Matter

Fig. 8.17 A circular wire

loop of radius r carries a
current I

References

Fermi E (1930) Über die magnetischen Momente der Atomkerne. Z Phys 60:5–6
Griffiths DJ (1999) Introduction to electrodynamics, 3rd edn. Prentice Hall
Jackson JD (1999) Classical electrodynamics, 3rd edn. John Wiley and Sons
Landau LD, Lifshitz EM (1971) The classical theory of fields. Pergamon Press
Sykja H (2006) Bazat e Elektrodinamikës. SHBUT
Chapter 9
Maxwell’s Equations
of Electromagnetism

This chapter aims to introduce Maxwell’s equations of

electromagnetism for the free space and the medium.

In this chapter, we discuss Maxwell’s equations of electromagnetism. Magnetic and

electrostatic fields exist together, and hence it makes sense to always discuss the two
fields simultaneously under electromagnetism. That is, often, the electromagnetism
is represented by electromagnetic field, which is the pair of vectors (E, B) for the free
space or (D, H) for the medium. Also, in this chapter, we will present the energy of the
electromagnetic field and the motion of the charged particles in an electromagnetic
field. For further reading, one can also consider other available literature Jackson
(1999), Landau and Lifshitz (1971), Sykja (2006), Griffiths (1999).

9.1 Maxwell’s Equations of Electromagnetism

First, we start with Maxwell’s equations in free space. In the following, we write
these equations in the integral form:

Q
E · dA = (9.1)
0
 A

∂
E · ds = − B · dA
∂t
L  A
 
∂
B · ds = μ0 I + 0 E · dA
∂t A
L
B · dA = 0
A

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 207
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_9
208 9 Maxwell’s Equations of Electromagnetism

The first Maxwell’s equation in Eq. (9.1) is simply Gauss’s law of the electric field
flux through a surface A enclosing a volume V with a net charge inside that volume Q.
Note that here electric field E is created by all the charges in free space (including the
charge Q inside the volume). This equation indicates the existence of the free electric
poles of the electric field (charges). The second Maxwell’s equation in Eq. (9.1)
derives from Faraday’s law; the left-hand side is the induced electromotive force,
and the right-hand side gives the rate change with time of the magnetic field flux
through the surface enclosed by contour L.
In Eq. (9.1), the third Maxwell’s equation is Ampére’s law, where the first term
on the right-hand side gives the net current through the open surface enclosed by the
contour L, and the second term determines the displacement current.
The fourth Maxwell’s equation in Eq. (9.1) implies that magnetic field flux through
a closed surface is equal to zero. Alternatively, the net number of magnetic field lines
passing through a closed surface is zero; that is, so many magnetic lines are leaving
the closed surface as entering it. That indicates that there do not exist free magnetic
poles.
It is important to note that in Eq. (9.1) we have assumed that B and E are functions
of both position r and time t. Therefore, the formulae consider partial derivatives.
Equation (9.1) can also be written in a differential form. For instance, the dif-
ferential form of the first Maxwell’s equation is given by expression in Eq. (4.69)
(Chap. 4). To derive the second Maxwell’s equation in a differential form, we use
Stokes’ formula in Eq. (9.1):
 
∂
∇ × E dA = − B · dA (9.2)
A ∂t A

Comparing both sides of Eq. (9.2), we get the second Maxwell’s equation in a dif-
ferential form:
∂B
∇×E=− (9.3)
∂t
Similarly, using Stokes’ formula and current density vector J, we can rewrite the
third Maxwell’s equation as follows:
 
B · ds = (∇ × B) · dA (9.4)
L
 A
 
∂
= μ0 J · dA + 0 E · dA
A ∂t A

Comparing both sides in Eq. (9.4), we obtain the third Maxwell’s equation in the
following differential form:
 
∂E
∇ × B = μ0 J + 0 (9.5)
∂t
9.1 Maxwell’s Equations of Electromagnetism 209

To derive the differential form of the fourth Maxwell’s equation, we use Gauss’s
formula in the last expression of Eq. (9.1), as
 
B · dA = ∇ · B dV = 0 (9.6)
A V

In Eq. (9.6), V denotes the volume enclosed by the surface A. Therefore, we obtain

∇·B=0 (9.7)

Finally, we can summarize the four Maxwell’s equations in the following differ-
ential form:
ρ
∇·E= (9.8)
0
∂B
∇×E=−
∂t
 
∂E
∇ × B = μ0 J + 0
∂t
∇·B=0

Next, we will consider Maxwell’s equations in a medium, which are written in

terms of the displacement vector D and magnetic field strength H, such as

1
E= (D − P) (9.9)
0
B = μ0 (H + M)

where P is electric polarization and M is magnetization vector. The first Maxwell’s

equation in medium is derived in Chapter 4 (see Eq. (4.79)):

∇·D=ρ (9.10)

In Eq. (9.10), ρ is the excess charge density in the medium. The second and fourth
Maxwell’s equations are given by the second and fourth expressions, respectively,
in Eq. (9.8).
To obtain the third Maxwell’s equation, we start with differential form of Ampére’s
law:

∇ × B = μ0 ∇ × (H + M) (9.11)
= μ0 (∇ × H) + μ0 (∇ × M)
= μ0 J
210 9 Maxwell’s Equations of Electromagnetism

where J is the equivalent macroscopic current, which is sum of the microscopic

current J and an effective current density due to magnetization:

JM = ∇ × M (9.12)

Combining Eqs. (9.11) and (9.12), we obtain

μ0 (∇ × H) = μ0 J (9.13)

Using the continuity equation for the charge and current, we write

∂ρ
∇·J+ =0 (9.14)
∂t
Replacing Eq. (9.10) into Eq. (9.14), we get
 
∂D
∇· J+ =0 (9.15)
∂t

Replacing the current J in Ampére’s law (Eq. (9.13)) by its generalization

∂D
J→J+ (9.16)
∂t
we obtain the third Maxwell’s equation:
 
∂D
μ0 (∇ × H) = μ0 J + (9.17)
∂t
or
∂D
∇×H=J+ (9.18)
∂t

9.2 Vector and Scalar Potentials of Electromagnetic Field

Maxwell’s equations include a set of first-order partial differential equations related

to different electric and magnetic field components. In general, they can be solved for
simple problems. However, for convenience potentials that include a smaller number
of second-order equations are often introduced, which satisfy some of Maxwell’s
equations identically. We already introduced those potentials when we discussed the
electrostatic and magnetostatic fields, separately, as scalar electric potential φ and
the vector potential A, which relate to the fields as E = −∇φ and B = ∇ × A, as
discussed in the previous chapters.
9.2 Vector and Scalar Potentials of Electromagnetic Field 211

In the following, for convenience, we restrict ourselves to the free space form
of Maxwell’s equations given by Eq. (9.8). Since ∇ · B = 0, we can, again for the
electromagnetic field, write that

B=∇×A (9.19)

Using the second expression in Eqs. (9.8) and (9.19), we can write

∂(∇ × A)
∇×E=− (9.20)
∂t
which can be arranged in this form:
 
∂A
∇× E+ =0 (9.21)
∂t

Equation (9.21) indicates that the quantity of a vanishing curl can be written as a
gradient of a scalar function, namely, the potential scalar φ:

∂A
E+ = −∇φ (9.22)
∂t
or
∂A
E = −∇φ − (9.23)
∂t
Definitions of the electromagnetic fields E and B related to the potentials φ and A
given by Eq. (9.19) and Eq. (9.23) satisfy identically the two homogeneous Maxwell’s
equations in Eq. (9.8). The dynamics of φ and A can be determined using the other
two non-homogeneous Maxwell’s equation in Eq. (9.8).
For instance, using the first Maxwell’s equation in Eq. (9.8), we obtain

∂(∇ · A) ρ
∇2φ + =− (9.24)
∂t 0

In addition, using the third Maxwell’s equation in Eq. (9.8), we get

 
∂(∇φ) ∂ 2 A
∇ × (∇ × A) = μ0 J + μ0 0 − − 2 (9.25)
∂t ∂t

Using the following relation

∇ × (∇ × A) = ∇(∇ · A) − ∇ 2 A (9.26)

and that
√ 1
μ0 0 = (9.27)
c
212 9 Maxwell’s Equations of Electromagnetism

with c denoting the speed of light in vacuum, Eq. (9.25) reduces as follows:
 
1 ∂ 2A 1 ∂φ
∇ A− 2 2 −∇ ∇·A+ 2
2
= −μ0 J (9.28)
c ∂t c ∂t

Those derivations indicate that four first-order differential Maxwell’s equations

given as in Eq. (9.8) for (E, B) are equivalent to two second-order differential equa-
tions given by Eqs. (9.24) and (9.28) for (φ, A). Note, however, they are still coupled,
that is, solutions, (φ, A), of Eqs. (9.24) and (9.28) are coupled to each equation. To
decouple φ and A, we note that potentials φ and A are arbitrarily defined. That is, B,
defined by Eq. (9.20), remains unchanged by the following transformation:

A = A + ∇Ψ (9.29)

where Ψ is an arbitrary scalar function.

Furthermore, electric field E given by Eq. (9.24) remains unchanged under the
following transformation:
∂Ψ
φ = φ − (9.30)
∂t

Equations (9.29) and (9.30) imply that in choosing the pair of potentials (φ, A),
we determine the so-called Lorentz condition to be satisfied:
1 ∂φ
∇·A+ =0 (9.31)
c2 ∂t
Using the transformation given by Eqs. (9.30) and (9.31), it is straightforward to
show that equations for φ and A transform as follows:

1 ∂ 2φ ρ
∇2φ − =− (9.32)
c ∂t
2 2 0

and
1 ∂ 2A
∇2A − = −μ0 J (9.33)
c2 ∂t 2
The forms of the expressions given in Eqs. (9.32) and (9.33) indicate that equations
for φ and A are decoupled, and hence they can be solved separately.

9.3 Electromagnetic Field Energy and Conservation Law

As derived separately in the previous chapters, we can write now the total electro-
magnetic field energy as the sum of the electric field potential energy U E and the
magnetic field energy U B :
9.3 Electromagnetic Field Energy and Conservation Law 213
 
1 1
U = UE + UB = E · D dr + B · H dr (9.34)
2 2
V V

1
= (E · D + B · H) dr
2
V

In writing Eq. (9.34), we assume that electromagnetic field (E, B) exists in a finite
volume V of the space. Furthermore, we can introduce the electromagnetic field
energy density u as
1
u = (E · D + B · H) (9.35)
2
Thus, 
U= u dr (9.36)
V

and hence u gives the energy per unit of volume.

Consider there is a continuous distribution of charge and current in some finite
volume V , where the electromagnetic field is different from zero. We know that
only the electric field does work on the moving charges, but not the magnetic field
(since the magnetic force is perpendicular with direction of the velocity). Therefore,
the total element of work done by the electromagnetic field on the charge during a
displacement Δs is given as

ΔW = ρ(r)Δs · E dr (9.37)
V

where dq = ρ dr is the charge in a small element volume d V = dr and ρ is the

charge density at some location r at the center of that small element volume. Using
Eq. (9.37), we can calculate the average power of the fields as follows:

ΔW Δs
= ρ(r) · E dr (9.38)
Δt Δt
V

Taking the limit when Δt → 0 of both sides of Eq. (9.38) and knowing that
limΔ→0 (Δs/Δt) = v is the velocity of the charge dq, we obtain the rate of the
work done by the electromagnetic field on the charge:

dW
= J · E dr (9.39)
dt
V

where J is the usual current density.

214 9 Maxwell’s Equations of Electromagnetism

Substituting the current density from Eq. (9.18) into Eq. (9.39), we get
  
dW ∂D
= ∇×H− · E dr (9.40)
dt ∂t
V
  
∂D ∂B
=− ∇ · (E × H) + E · +H· dr
∂t ∂t
V

where the following vector relationship was used:

∇ · (E × H) = H · (∇ × E) − E · (∇ × H) (9.41)

and Maxwell’s equation for ∇ × E. Moreover, we assume that macroscopic medium

is linear in its magnetic and electric field properties and there are no dispersion or
energy losses. Using Eq. (9.35), we get the following form for Eq. (9.40):
   
∂u
− J · E dr = ∇ · (E × H) + dr (9.42)
∂t
V V

We introduce the so-called Poynting vector as

S=E×H (9.43)

Then, substituting Eq. (9.43) into Eq. (9.42), we get

   
∂u
− J · E dr = ∇·S+ dr (9.44)
∂t
V V

Comparing both sides of Eq. (9.44) and assuming that the volume V is chosen
arbitrary, we get the continuity equation in its differential form:

∂u
+ ∇ · S = −J · E (9.45)
∂t
Equation (9.45) is also known as the energy conservation law of the electromagnetic
field; that is, the rate change with time of the sum of the electromagnetic energy in
a volume V and the energy leaving the volume through the boundary surface of that
volume per unit of time is equal to the total work done by the fields on the sources
(current and charges) within the volume. Moreover, the work done per unit of time
and per unit of volume by the electromagnetic field (that is, J · E) is transformation
of electromagnetic energy into internal energy (mechanical or heat energy).
9.4 Conservation Law of Momentum 215

9.4 Conservation Law of Momentum

At a microscopic level, the medium comprises charged particles, such as electrons

and atomic nuclei. Therefore, we can think of the rate of energy conversion as a rate
of energy increase of the charged particles of the medium per unit of volume. Fur-
thermore, Eq. (9.44) can be treated as a conservation law of energy at the microscopic
level of the combined system of particles and electromagnetic fields (E, B).
For that, we denote by E mech the total mechanical energy of the charged particles
inside the volume V and assume that the particles do not leave the volume. The rate
change of the mechanical energy is

d E mech
= J · E dr (9.46)
dt
V

Combining Eqs. (9.44) and (9.46), we express the conservation law of energy of the
combined system as 
d  
E mech + E f ield = ∇ · S dr (9.47)
dt
V

Using Gauss’s theorem, we get


d  
E mech + E f ield = S·ndA (9.48)
dt
A

where A is the surface enclosing the volume V and n is an outward normal vector to
every point in the surface. In Eq. (9.48), E f ield is the total energy of electromagnetic
field given by
   
1 1 2
E f ield = 0 E +
2
B dr = u dr (9.49)
2 μ0
V V

The net electromagnetic force on a small volume charged particle dq moving

with velocity v is sum vector of the electric field force, dqE, and magnetic field
force (Lorentz force), dq(v × B):

dF = dq (E + v × B) (9.50)

Integrating Eq. (9.50) over a volume V of the charge distribution, we obtain the total
force exerted by the electromagnetic field on the volume charge distribution:

F= (E + v × B) ρ(r) dr (9.51)
V
216 9 Maxwell’s Equations of Electromagnetism

Using the second law of Newton, and denoting by Pmech the total momentum of the
particles inside the volume V , we get

dPmech
= (ρE + J × B) dr (9.52)
dt
V

To eliminate ρ and J from the equation, we use Maxwell’s equations given in

Eq. (9.8), and write
 
∂E
ρE + J × B = 0 E(∇ · E) + B × − c2 B × (∇ × B) (9.53)
∂t

Using the following relation:

∂E ∂B ∂
B× =E× − (E × B) (9.54)
∂t ∂t ∂t
then, Eq. (9.53) can be written as

ρE + J × B = 0 E(∇ · E) + c2 B(∇ · B) (9.55)

∂
− E × (∇ × E) − (E × B) − c2 B × (∇ × B)
∂t

where c2 B(∇ · B) = 0 is added, and the second Maxwell’s equation relation is used
as in Eq. (9.8). Substituting Eq. (9.55) into Eq. (9.52), we get

dPmech d
+ 0 (E × B) dr (9.56)
dt dt
V

 
= 0 E(∇ · E) − E × (∇ × E) + c2 B(∇ · B) − c2 B × (∇ × B) dr
V

The second term on the left-hand side of Eq. (9.56) defines the electromagnetic
momentum P f ield in the volume V as
 
1
P f ield = 0 (E × B) dr = (E × H) dr (9.57)
c2
V V

Therefore, the integrand of Eq. (9.57) represents the density of electromagnetic

momentum:
1
p f ield = 2 (E × H) (9.58)
c
9.4 Conservation Law of Momentum 217

The terms on the integrand of the integral on the right-hand side can be simplified
as follows:

(B(∇ · B) − B × (∇ × B))1 = (B(∇ · B))1 − (B × (∇ × B))1 (9.59)


3
∂ Bi
= B1 − B2 (∇ × B)3 + B3 (∇ × B)2
i=1
∂ xi
 3    
∂ Bi ∂ B2 ∂ B1 ∂ B3 ∂ B1
= B1 − B2 − + B3 − +
i=1
∂ xi ∂ x1 ∂ x2 ∂ x1 ∂ x3
   
∂ B1 ∂ B2 ∂ B3 ∂ B2 ∂ B1 ∂ B3 ∂ B1
= B1 + B1 + B1 − B2 − + B3 − +
∂ x1 ∂ x2 ∂ x3 ∂ x1 ∂ x2 ∂ x1 ∂ x3
∂ B12
∂ B1 B2 ∂ B1 B3 1 ∂(B1 + B2 + B3 )
2 2 2
= + + −
∂ x1 ∂ x2 ∂ x3 2 ∂ x1
 ∂
3  
1
= B1 Bi − B · Bδ1i
i=1
∂ x i 2

In general, the k-th component is

3  
∂ 1
(B(∇ · B) − B × (∇ × B))k = Bk Bi − B · Bδki (9.60)
i=1
∂ xi 2

Similarly,

3  
∂ 1
(E(∇ · E) − E × (∇ × E))k = E k E i − E · Eδki (9.61)
i=1
∂ xi 2

We can now determine the so-called Maxwell stress tensor as

 
1 
σki = 0 E k E i + c2 Bk Bi − E · E + c2 B · B δki (9.62)
2

Therefore, we can rewrite Eq. (9.56) in the following form:

 
d   3
∂
Pmech + P f ield k = σki dr (9.63)
dt i=1
∂ xi
V

Using Gauss’s formula, Eq. (9.63) can be written as

 
d   3
Pmech + P f ield k = (σki n i ) d A (9.64)
dt i=1
S
218 9 Maxwell’s Equations of Electromagnetism

where A is the surface enclosing the volume V and n i is the i-th component of the
outward unit vector to the surface at every point. 3
Equation (9.64) represents the conservation law of momentum, where i=1 σki n i
is the i-th component of the flow per unit
3 area of the momentum across the surface
into the volume V of sources. Thus, i=1 σki n i is also the force per unit surface
area acting along the surface A and applied to the combined system of particles plus
fields inside volume V .

9.5 Dynamics of Charged Particles in Electromagnetic

Fields

Maxwell’s equations form the basis of all classical electromagnetic phenomena in

matter. Those equations combined with the Lorentz force equation and Newton’s
second law of motion provide a complete description of the classical dynamics of
interacting charged particles and electromagnetic fields.
Consider a single particle of mass m and charge q moving with velocity v in an
electromagnetic field (E, B), as shown in Fig. 9.1. We can write the Lagrangian of
the particle as
mv 2
L = T −U = − qφ + qv · A (9.65)
2
where (φ, A) are the potentials of the electromagnetic field, and

v= ẋ12 + ẋ22 + ẋ32 (9.66)

is the speed. In Eq. (9.65), the first term gives the kinetic energy, and the last two
terms on the right-hand side give, respectively, potential energy of a charged particle
in the electromagnetic field. The term qφ is simply electrostatic potential energy,
as discussed in Chap. 3, while the term −qv · A gives the potential energy of the
moving charge in magnetic field. A physical explanation for these two terms follows
from that J · E gives the work done per unit of time and per unit of volume by the
electromagnetic field, and hence the decrease in potential energy of a moving particle
in electromagnetic field is

−dU = J · E dt d V (9.67)
= J · Edt (ds · dA)
= I dt ds · E
= dqds · E
 
 ∂A
= qδ(r − r ) ds · −∇φ −
∂t
9.5 Dynamics of Charged Particles in Electromagnetic Fields 219

Fig. 9.1 A charged particle

moving with velocity v at the
position r. The charge
density at any position r of
the volume V extended by
the electromagnetic field is
ρ(r ) = qδ(r − r )/d V ,
where δ is the function:
δ(r − r ) = 1 if r = r ,
otherwise it is zero. d V is a
small cylindrical volume
element along the path of
particle d V = ds · dA,
where ds is displacement of
particle during a small time
interval dt

Integrating over the all volume V extended by the electromagnetic field, we get
  
 ∂A
U =− qδ(r − r ) ds · ∇φ + (9.68)
∂t
V
= −q (φ(∞) − φ(r)) − qv · A

Taking that the electromagnetic field is zero at the infinity (φ(∞) = 0), we obtain
the potential energy of a charged particle in the electromagnetic field as

U (r) = qφ(r) − qv · A(r) (9.69)

Lagrangian equations are given as

 
d ∂L ∂L
− = 0, i = 1, 2, 3 (9.70)
dt ∂ ẋi ∂ xi

Substituting Eq. (9.65) into Eq. (9.70), we obtain

 
∂ Ai ∂φ ∂
m ẍi + q − −q −q (v · A) = 0, i = 1, 2, 3 (9.71)
∂t ∂ xi ∂ xi
or  
∂φ ∂ Ai
m ẍi = q − − + q(v × (∇ × A))i , i = 1, 2, 3 (9.72)
∂ xi ∂t

Using Eqs. (9.19) and (9.23), we can write Eq. (9.71) in the well-known form:

m ẍi = q E i + q(v × B)i , i = 1, 2, 3 (9.73)

220 9 Maxwell’s Equations of Electromagnetism

The first term on the right-hand side is the electric field force on q, and the second
is the Lorentz force applied on moving charge by a magnetic field. If we consider
now a system of N charges, q1 , q2 , . . . , q N inside some finite volume V . That could
correspond to the system of electrons in the field of the nucleus in an atom. Then,
Lagrangian of the system in the presence of the electromagnetic field is given as


N
m i v2 
N N  N
qi q j
L = T −U = i
− (qi φ + qi vi · A) − ke (9.74)
i=1
2 i=1
r
i=1 j=i+1 i j

where (φ, A) are again the potentials of the electromagnetic field, and the last term
is the potential interaction energy between charges in the system. In Eq. (9.74), ri j
is the distance between two charges i and j.
Using Lagrangian equations, Eq. (9.71), we obtain the following second-order
differential equations:

N
qi q j
m ẍi = qi E i + qi (vi × B)i + ke (xi − x j ), i = 1, 2, · · · , 3N (9.75)
j=1=i
ri3j

where 3N also gives the number of degrees of freedom of the system.

9.6 Macroscopic Maxwell Equations

When discussing the electromagnetic fields in a medium made up of the constituents

of matter, such as electrons, protons and neutrons, we would have to consider indi-
vidually those sources of the electric and magnetic moments. However, the number
of these sources may go up to 1023 , which is a big number. If we denote by small let-
ters e, b, j microscopic fields created by those sources, then microscopic Maxwell’s
equations will be written as
ρ
∇ ·e = (9.76)
0
∂b
∇ ×e =− (9.77)
∂t
∇ ·b=0 (9.78)
 
∂e
∇ × b = μ0 j + 0 (9.79)
∂t

Equations (9.76)–(9.79) imply that there will exist up to 1023 such equations to
be solved numerically for each source. Furthermore, we are not always interested in
considering the microscopic fields, but we would like to consider the electromagnetic
field behavior on classical length scales, which exceeds the atomic scales.
9.6 Macroscopic Maxwell Equations 221

For that, macroscopic fields are introduced by averaging the microscopic fields
as

E = Ē (9.80)
B = B̄ (9.81)

where the averaging is defined as


A(r) = a(r) = a(r − r )w(r ) dr (9.82)
V

where w is the weight function at the position of the source, which is normalized to
unity. Note that w extends over a sufficiently large region in space.
Therefore,

∂A ∂a
= (9.83)
∂t ∂t
∂A ∂a
= (9.84)
∂r ∂r
The averaged Maxwell’s equations can be written as follows:

ρ
∇ ·E= (9.85)
0

∂B
∇ ×E=− (9.86)
∂t
∇ ·B=0 (9.87)
 
∂E
∇ × B = μ0 j + 0 (9.88)
∂t

To understand the averages of the sources of the fields, such as the charge density and
current density, we will have to consider the atomic structure of a typical solid. There
are two types of sources of the charge, namely, the nuclei and valence electrons. In the
following discussion, we denote by rm the position of the center of atom or molecule
to a laboratory frame and am,i the position of the source i of molecule m to the
center of the molecule. Furthermore, in conductors, there exist the free conduction
electrons at some position ri (t) in the system at time t. We denote by ρb charge
density contribution from the bound sources of charge (nuclei and valence electrons)
and ρ f the charge density contribution from the free electrons in conduction region.
They are written as
222 9 Maxwell’s Equations of Electromagnetism

ρb (r, t) = qm,i δ(r − rm (t) − am,i (t)) (9.89)
m,i

ρ f (r, t) = e δ(r − ri (t)) (9.90)
m,i

where qm,i is the charge of the source i (nucleus or electron) in molecule m and e is
the charge of electron.
The average density of the bound charges over a macroscopic region is given by
 
ρb (r) = w(r ) qm,i δ(r − rm (t) − am,i (t)) dr (9.91)
V m,i

= w(r − rm − am,i )qm,i
m,i
  
≈ qm w(r − rm ) − ∇ w(r − rm ) · am,i qm,i
m m i

= qm δ(r − rm ) − ∇ · P(r)
m

The third line gives the Taylor expansion of the weights knowing that the weight
function changes exceed the atomic extensions and they decay to zero at distances
further than atomic dimensions | a |. At the zeroth order, Taylor expansion gives the
total charge of molecule: 
qm = qm,i (9.92)
i

The first-order term includes the molecular dipole moments:


dm = am,i qm,i (9.93)
i

In Eq. (9.91), P gives the average polarization of the medium:


P(r, t) =  δ(r − rm (t))dm (t) (9.94)
m

Equation (9.94) indicates that P is the density of the dipole moments carried by the
molecule in a system.
The average density of the mobile charge carriers in the medium (that is, conduc-
tion electrons) is given by

ρ f (r, t) = e δ(r − ri (t)) (9.95)
i
9.6 Macroscopic Maxwell Equations 223

Therefore, the average charge density is

 
ρ(r, t) = qm δ(r − rm ) − ∇ · P(r) + e δ(r − ri (t)) (9.96)
m i
= ρ(r, t) − ∇ · P(r, t)

where
 
ρ(r, t) = qm δ(r − rm ) + e δ(r − ri (t)) (9.97)
m i

denotes the macroscopic charge density in the system.

Substituting Eq. (9.96) into Eq. (9.85), as we have shown in the previous section,
the first Maxwell’s equation reduces to

∇ · D(r, t) = ρ(r, t) (9.98)

where D = 0 E + P.
To compute the average of the microscopic current density, we sum up the bound
and free charge current densities: jb + j f . The bound part of the charge current
density is given as
    
jb (r) = qm,i ṙ + ȧm,i δ r − rm − am,i (9.99)
m,i

Averaging expression given by Eq. (9.99), we get

     
jb (r) = w(r ) qm,i ṙm + ȧm,i δ r − rm − am,i dr (9.100)
m,i

= qm,i w(r − rm − am,i )(ṙm + ȧm,i )
m,i
 
   
≈ qm,i w(r − rm ) − ∇w(r − rm ) · am,i ṙm + ȧm,i
m,i m
(0) (1)
= jb (r) + jb (r) + jb (r) (2)

where
 
jb (r) (0) = qm w(r − rm )ṙm =  qm ṙm δ(r − rm ) (9.101)
m m

gives the current carried by a molecule in the zeroth-order approximation, where the
molecule is considered a point-like charge.
224 9 Maxwell’s Equations of Electromagnetism

The first-order approximation term is given as

 
jb (r) (1) = w(r − rm )ḋm − (∇w(r − rm ) · dm )ṙm (9.102)
m

Let us compare it with the time derivative of a polarization vector:


d 
Ṗ = w(r − rm )dm (9.103)
dt m
 
= w(r − rm )ḋm − (∇w(r − rm ) · dm )ṙm
m

Substituting Eq. (9.103) into Eq. (9.102)


jb (r) (1) = Ṗ + ∇w(r − rm ) × (dm × ṙm ) ≈ Ṗ (9.104)
m

The second-order contribution term is given by

  
jb (r) (2) = − qm,i am,i ȧm,i · ∇w(r − rm ) (9.105)
m,i

which relates to the magnetic dipole moments carried by the molecules:

1
M= qm,i am,i × ȧm,i w(r − rm ) (9.106)
2 m,i

Calculating the curl of the expression given by Eq. (9.106), we get


1 d 
∇ × M = jb (2) + qm,i am,i (am,i · ∇w(r − rm ) (9.107)
2 dt m,i

In the right-hand side of Eq. (9.107), the second expression is the electric quadrupole
contribution, which can be ignored since it is a small term. Therefore,

jb (2) ≈ ∇ × M (9.108)

The contribution from the free charge carrier is


j f (r) = e ṙi δ(r − ri ) (9.109)
i

Combining terms in Eqs. (9.101), (9.104), (9.108), and (9.109), we obtain the
expression for the average current density as
9.6 Macroscopic Maxwell Equations 225

j(r) = J(r) + Ṗ(r) + ∇ × M(r) (9.110)

where  
J(r) = e ṙi δ(r − ri ) +  qm ṙm δ(r − rm ) (9.111)
i m

gives the current carried by the free charge carrier and the point-like approximated
molecule. M is the density of the magnetic dipole moments in the system:
 1
M= δ(r − rm ) qm,i am,i × ȧm,i (9.112)
m
2 i

Substituting Eq. (9.110) into Eq. (9.88), as we have shown in the previous section,
the last Maxwell’s equation reduces to
 
1 ∂D
∇× B−M =J+ (9.113)
μ0 ∂t

1
Since H = B − M, we finally get the inhomogeneous Maxwell’s equation:
μ0

∂D
∇×H=J+ (9.114)
∂t

9.7 Exercises

Exercise 9.1 Show that −∇(v · A) = v × (∇ × A), where v is the velocity

of a charged particle and A is potential vector of the electromagnetic field.

Solution 9.1 We start with the following vector identity:

a × (∇ × b) = a(b · ∇) − ∇(a · b) (9.115)

Then,
v × (∇ × A) = (A · ∇)v − ∇(v · A) = −∇(v · A) (9.116)

because the velocity components do not depend on the coordinates, and thus

∂vi
= 0, i, j = 1, 2, 3 (9.117)
∂x j
226 9 Maxwell’s Equations of Electromagnetism

Exercise 9.2 An electric dipole with moment p and fixed direction is located
at the position r0 (t) with respect to the origin. Its velocity is v = dr0 /dt. Find
the dipole’s charge and current densities.

Solution 9.2 We start with definition of the charge density

2
qi
ρ(r) = δ (r − ri ) (9.118)
i=1
dr

We denote by Δr the vector from positive to negative charge, then

   
q Δr q Δr
ρ(r) = δ r − r0 + − δ r − r0 − (9.119)
dr 2 dr 2
    
q 1 Δr Δr
= Δr δ r − r0 + − δ r − r0 −
dr Δr 2 2
1 
= −p · δ (r − r0 )
dr
= −(p · ∇)δ(r − r0 )

The current density is

J = vρ(r) = −v(p · ∇)δ(r − r0 ) (9.120)

Exercise 9.3 Express A and φ in terms of the magnetic field vector B and
electric field vector E for static potentials and homogeneous fields.

Solution 9.3 We start with Eqs. (9.19) and (9.23), and since the potentials are static,
then
∂A
=0 (9.121)
∂t
and hence
∂φ(r)
E = −∇φ = − (9.122)
∂r
Since the field is homogeneous, then

φ(r) = −E · r (9.123)
9.7 Exercises 227

Using Eq. (9.19) and multiplying both sides by r, we get

B × r = r × (A × ∇) = 2A (9.124)

Or
1
A= (B × r) (9.125)
2

Exercise 9.4 Show that (A · ∇)r = A, where A is the vector potential.

Solution 9.4 We write

(A · ∇)x = A x ∇x x + A y ∇ y x + A z ∇z x = A x (9.126)
(A · ∇)y = A x ∇x y + A y ∇ y y + A z ∇z y = A y (9.127)
(A · ∇)z = A x ∇x z + A y ∇ y z + A z ∇z z = A z (9.128)

Therefore, combining those results, we write that

(A · ∇)r = A (9.129)

Exercise 9.5 Inside the volume V the potential vector A satisfies the condition
∇ · A = 0, while at the boundary surface S of that volume An = 0. Show that

A dV = 0 (9.130)
V

Solution 9.5 Using Gauss’s formula, we write

 
A dV = ∇ · A dS (9.131)
V S

where S is the surface enclosing the volume V . From the conditions, we obtain

∇ · A dS = 0 (9.132)
S
228 9 Maxwell’s Equations of Electromagnetism

Exercise 9.6 Find the charge density ρ if the electric field is E = (b · r)b,
where b is a constant vector.

Solution 9.6 Using Maxwell’s equation

ρ
∇·E= (9.133)
0

Using the definition of the divergence, we get

 
∇ · (b · r)b = (∇(b · r)) · b + (b · r)(∇ · b) = bx i + b y j + bz k · b = b2
(9.134)
because ∇ · b = 0. Therefore, we get

ρ = 0 b 2 (9.135)

Exercise 9.7 Find the current charge density J if the magnetic field is B =
f (r )(a × r), where a is a constant vector and f (r ) is a scalar function and a
is a constant vector.

Solution 9.7 Using Maxwell’s equation

∇ × B = μ0 J (9.136)

Consider the left-hand side:

∇ × B = ∇ × ( f (r )(a × r)) (9.137)

= ∇ f (r ) × (a × r) + f (r )∇ × (a × r)
f  (r )
= r × (a × r) + f (r )∇ × (a × r)
r
= f (r )∇ × (a × r)
= f (r ) [a(∇ · r) − r(∇ · a) + (r · ∇)a − (a · ∇)r]
= f (r ) [3a − a] = 2 f (r )a

where r
∇ f (r ) = f  (r ) (9.138)
r
9.7 Exercises 229

where f  (r ) = d f (r )/dr . Therefore, we get

2
J= f (r )a (9.139)
μ0

Exercise 9.8 Prove that if the current density is zero outside some finite vol-
ume and continuously differential everywhere, then

J dV = 0 (9.140)
V

Solution 9.8 Since the current density is differential function inside the volume,
then using Gauss’s formula
 
J dV = (∇ · J) d A (9.141)
V A

where A is the surface enclosing the volume V . Using Maxwell’s equation, we write
 
1
(∇ · J) d A = ∇ · (∇ × B) d A = 0 (9.142)
μ0
A A

because ∇ · (∇ × a) = 0 for any arbitrary vector a.

Exercise 9.9 Suppose that J = (a × ∇)δ(r − r0 ). Determine the magnetic

moment.

Solution 9.9 Magnetic moment is


1
m= (r × J) d V (9.143)
2
V

Substituting the expression for J, we get

230 9 Maxwell’s Equations of Electromagnetism

1
m= (r × (a × ∇)δ(r − r0 )) d V (9.144)
2
V
1
= r0 × (a × ∇)
2
1
= ((r0 · ∇)a − (r0 · a)∇)
2
1
= − (r0 · a)∇
2
where a is assumed to be a constant vector.

Exercise 9.10 Show that magnetic moment does not depend on the choice of
the coordinate system origin.

Solution 9.10 Let r0 be a displacement of the origin of the coordinate system, then
magnetic dipole to new origin is written as

1
m = ((r − r0 ) × J) d V (9.145)
2
V

Simplifying that expression, we get

 
1 1
m= (r × J) d V − (r0 × J) d V (9.146)
2 2
V V

1
= m − r0 × J d V
2
V

1
= m − r0 × ∇ · J d A
2
A
=m

because ∇ · J = 0 the current density continuity equation.

References 231

References

Griffiths DJ (1999) Introduction to electrodynamics, 3rd edn. Prentice Hall

Jackson JD (1999) Classical electrodynamics, 3rd edn. John Wiley and Sons
Landau LD, Lifshitz EM (1971) The classical theory of fields. Pergamon Press
Lorentz LV (1867) Phil Mag Ser 3(34):287
Sykja H (2006) Bazat e Elektrodinamikës. SHBUT
Chapter 10
More About Faraday’s Law of Induction

This chapter aims to extend the law of induction and introduce

alternating current circuits and the phenomena of resonance,
useful in wireless applications of electromagnetic theory.

In this chapter, we introduce some more information about Faraday’s law of induc-
tion. For further reading, one can also consider other available literature (Holliday
et al., 2011).

10.1 Moving Conductor in a Closed Circuit

Consider a conductor moving in a magnetic field B, which points toward the page,
along two fixed conductors. The conductor is part of a closed circuit as shown in
Fig. 10.1. The length of the stationary resistance (R) part is l. In Fig. 10.1, Fapp
denotes the applied force, which is perpendicular to the moving conductor part with
a constant velocity v. On the moving electrons and ions of the moving conductor is
acting Lorenz force: Fq = q (v × B), where q is the charge. The directions of these
two forces are depicted in Fig. 10.1. As a result, positive charges collect on the upper
side of the moving conductor and electrons (negative charges) on the lower side,
creating a difference of the electric potential ΔΦ ≡ ind = Φ+ − Φ− .
Furthermore, as the conductor moves to the right, the magnetic field flux Φ B
through the closed circuit path changes. When the moving conductor is at the position
x, then the magnetic flux through the closed surface of the circuit is

Φ B = B · (l · x) = Blx (10.1)

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 233
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_10
234 10 More About Faraday’s Law of Induction

Fig. 10.1 A moving conductor in a magnetic field B pointing toward the page part of a closed
conducting path. Fapp is the applied force to the moving conductor gaining the velocity v and FB is
the magnetic force acting on the current I flowing in the moving conductor

Using Faraday’s law, the induced electric potential is

dΦ B
ind = − = −Blv (10.2)
dt
where the speed is v = d x/dt, which is the speed of the moving conductor. The
induced electric potential ind creates a current flowing in the circuit, as shown in
Fig. 10.1. Therefore, upon the moving conductor of length l is exerted a magnetic
force FB pointing to the left (see Fig. 10.1):

FB = I (L × B) (10.3)

where L is a vector of length l pointing along the direction of the current flow. It has
a magnitude of
FB = l I B (10.4)

and direction is opposite to the direction of applied force, Fapp . Besides, since the
conductor is moving with constant velocity, the resultant force must be equal to zero,
based on Newton’s first law; therefore, ignoring the mass of the conductor, we have

Fapp = l I B (10.5)

Also, since the resistance of the circuit is R, the voltage drop across the resistance
equals the magnitude of the induced voltage, and hence
10.1 Moving Conductor in a Closed Circuit 235

Fig. 10.2 Equivalent circuit

| ind | Blv
I = = (10.6)
R R
The equivalent electric circuit is shown in Fig. 10.2. Furthermore, the work done by
applied force for moving the conductor by a displacement d to the right is

Wapp = Fapp · d (10.7)

which equals the work done on the charges moving through the magnetic field. Note
that the magnetic field force is moving the electrons downward with a certain average
drift velocity that establishes the current I in the circuit (see also Fig. 10.2).
Applying the conservation law of the energy, the work done by the applied force
on the moving conductor during the interval of time Δt is equal to the electric energy
supplied by the induced electric potential ind during the same time interval Δt. The
power delivered by the applied field is

dWapp
P= = Fapp · v = (I l B)v = I (Blv) = I | ind | (10.8)
dt
where Eqs. (10.6) and (10.7) are used. Therefore, we also write the conservation law
of energy as

| ind |2
P= = I | | = I2R
 (10.9)
R  ind 
   Electrical energy supplied by  Internal energy in R
ind
Mechanical energy

In Eq. (10.9), the first term represents the mechanical energy supplied by the applied
force, the second term represents the electrical energy supplied by induced emf, and
the last term is the internal energy in resistance.

10.1.1 Induced Electric Potential and Electric Field

As discussed above, the change in the magnetic flux induces electric potential in a
conducting loop. Therefore, an electric field is created in the conductor as a result
236 10 More About Faraday’s Law of Induction

Fig. 10.3 Induced electric

field in a closed loop placed
in a uniform magnetic field
pointing toward the page

of the changing magnetic flux. The induced electric field is varying in time and it is
non-conservative.
Consider a loop in a magnetic field perpendicular to the surface of the loop pointing
toward the page. If the magnetic field varies with time, based on Faraday’s law, an
induced electric voltage creates in the loop (see also Fig. 10.3):

dΦ B
ind = − (10.10)
dt

where Φ B = πr 2 B is the magnetic flux through the surface of the loop (r is the loop’s
radius) and A = πr 2 is the area of surface enclosed by the loop. Thus,

dB
ind = −πr 2 (10.11)
dt
Polarity of the induced potential difference is indicated in Fig. 10.3; therefore, the
induced electric field is tangent to the loop, and it has the same magnitude since all
the points on the loop are equivalent, as shown in Fig. 10.3.
The work done to move a test charge (q) around the loop equals to

Wq = q · ind (10.12)

The force acting on the charge q is

F=q ·E (10.13)
10.1 Moving Conductor in a Closed Circuit 237

The work done by electrical force to move the test charge around the loop is
 
Wq = Fq · dl = q E dl = 2πrq E (10.14)
L L

where the integration is over the closed-loop line L.

Combing Eqs. (10.12) and (10.14), we obtain

qind = 2πrq E (10.15)

From Eq. (10.15), we find the magnitude of the electric field

ind
E= (10.16)
2πr
Substituting Eq. (10.11) into Eq. (10.16), we obtain
 
1 dB r dB
E= −πr 2 =− (10.17)
2πr dt 2 dt

Equation 10.17 indicates that the induced electric field due to time-varying mag-
netic field is a function of time. The minus sign indicates that the electric field opposes
the change in magnetic field.
Furthermore, the induced electric voltage in any closed loop can be expressed as

dB
ind = E · dl = −πr 2 (10.18)
L dt

In general, the path may not be a closed loop and E may not be constant. However,
Eqs. (10.10) and (10.18) indicate that induced electric force is
non-conservative
because its work along a closed path is not zero, or equivalently L E · dl = 0.

10.1.2 Generators and Motors

Figure 10.4 shows an alternative current (AC) generator, which can be used to convert
mechanical energy to electrical energy. It comprises a loop of wire rotated using some
external mechanical rotor in a magnetic field. When the loop rotates in the magnetic
field, the magnetic flux passing through the area enclosed by the loop changes with
time, which induces emf and a current in the loop, based on Faraday’s law. If A is
the area enclosed by the loop and θ the angle between the magnetic field vector B
and surface area vector A = An (where n is a unit vector perpendicular to the loop’s
238 10 More About Faraday’s Law of Induction

Fig. 10.4 A generator design

surface), and ω the angular speed of rotation of the loop (which is assumed to be
constant). Then, the magnetic flux is

Φ B1 = B A cos θ (10.19)

where θ = ωt; therefore,

Φ B1 = B A cos (ωt) (10.20)

Assuming N turns, then

Φ B = N Φ B1 = N B A cos (ωt) (10.21)

Thus, the induced electric voltage is

dΦ B
ind = − (10.22)
dt
= N B Aω sin (ωt)

Equation 10.22 implies that maximum induced emf is

ind,max = N B Aω (10.23)

Motors are devices that convert electrical energy into mechanical energy, that is,
the reverse mode of a generator. In this case, a current supplied to the loop by a
battery and the torque acting on the current-carrying loop causes it to rotate.
10.2 Inductance 239

Fig. 10.5 A simple electric

circuit containing a switch
(S), source emf (), induced
emf ( L ), and a resistance
(R)

10.2 Inductance

10.2.1 Self-inductance

There is a difference between the emfs and the current produced by a battery or other
sources and those induced by changing the magnetic field flux.
In general, the source emf and source current describe parameters associated
with a physical source. In contrast, the induced emf and induced current describe
parameters associated with changing magnetic field flux.
Consider the simple electric circuit shown in Fig. 10.5. It consists of a source emf,
, a resistance, R, and a switch, S. When the switch is closed the source current does
not instantly increase to its maximum value Imax :

Imax = (10.24)
R
At some instance of time t the current in the circuit is I (t). Besides, the current
passing through straight wire line produces a magnetic field B = μ0 I /2πr (where r
is the distance from the wire), and thus B ∼ I . Since the current increases to reach
its maximum value Imax , so does the magnetic field. Furthermore, the magnetic flux
passing through the surface area enclosed by the circuit is

ΦB = B · dS (10.25)
S

where S is the surface area enclosed by the circuit. Therefore, since B increases, Φ B
increases with time, that is, dΦ B /dt = 0, which in turn creates an induced emf in
the circuit:
dΦ B
L = − (10.26)
dt
Based on Lenz’s law, the polarization of induced emf  L is such that it would create an
induced current in the circuit, such that this induced current would create a magnetic
field that would oppose the change of the magnetic field created by the source current.
240 10 More About Faraday’s Law of Induction

Fig. 10.6 An iron core and wire coil in which is passing the current I

Fig. 10.7 a The induced emf polarity when the current I increases; b The induced emf polarity
when the current I decreases

Therefore, the direction of the induced emf,  L , is opposite to the direction of

source emf, . As a result, a gradual increase of the current source occurs to its
maximum value Imax (equilibrium value) rather than an instantaneous. This effect is
called self-induction, and  L is called self-induced emf or back emf.
Consider the solenoid in Fig. 10.6, which consists of the cylindrical iron coil
and the wire coil of N turns. A current I flows along the wire coil. Magnetic field
direction B to the left is created due to the current in the wire coil.
Increasing the current yields an increase of the magnetic flux passing through the
coil, and hence dΦ B /dt = 0 and −dΦ B /dt < 0. Then, the induced emf has a polarity
as shown in Fig. 10.7a. In contrast, with decreasing the current I , the magnetic flux
through the coil decreases, and dΦ B /dt = 0, but −dΦ B /dt > 0. Therefore, the
induced emf has a polarity as shown in Fig. 10.7b.
Polarization of  L is defined based on Lenz’s law; the polarity of the induced emf
 L opposes the change of the magnetic field from the source current. Using Faraday’s
law, the induced emf is given by Eq. (10.26), with Φ B given by Eq. (10.25). Therefore,
Φ B ∼| B | and | B |∼ I . Here, I is the source current in the circuit. Therefore, the self-
induced emf,  L , is proportional to the rate change of the source current,  L ∼ d I /dt.
Suppose there is a closed spaced coil of N turns (for example, either a solenoid
or a toroid) carrying a source current I . Then, we write
10.2 Inductance 241

dΦ B dI
 L = −N = −L (10.27)
dt dt
where L is a proportionality constant, called the inductance of the coil. In general, L
depends on the geometry of the coil and other physical parameters. Equation 10.27
can also be arranged as follows:

d(N Φ B ) d(L I )
L = − =− (10.28)
dt dt
Thus, we can find that
N ΦB = L I (10.29)

Or,
N ΦB
L= (10.30)
I
Furthermore, using (10.27), we can write that
L
L = −  (10.31)
dI
dt

Equation 10.31 is in analogy with Ohm’s law: R = /I . Note that L is a measure
of the opposition to the change in the circuit source current. The SI units of L are
Henry (H):
V V·s
1H = 1 =1 (10.32)
A/s A

10.2.2 Mutual Inductance

The magnetic flux passing through the surface area of a loop is (see also Fig. 10.8):

ΦB = B · dS (10.33)
S

where | B |∼ I . Therefore, Φ B varies with time because I varies with time. Thus, an
induced emf occurs through the process of mutual inductance. This is related with
the fact that it depends on the interaction between two circuits.
Consider two parallel coils of N1 and N2 turns, respectively, as shown in Fig. 10.9.
Through the coil I is passing the current I1 and coil II the current I2 . Suppose the
current I1 is creating a magnetic field with magnetic field lines as depicted in Fig. 10.9.
Some of these lines pass through the coil II. We denote by Φ12 the magnetic flux
of the magnetic field created by coil I through the coil II. The mutual inductance,
namely, M12 , of coil II with respect to coil I is
242 10 More About Faraday’s Law of Induction

Fig. 10.8 The magnetic

field created by a current
flowing in a loop

Fig. 10.9 The magnetic

field created by two parallel
coils of N1 and N2 turns,
respectively, in which is
passing the current I1 and I2

Φ12
M12 = N2 (10.34)
I1

Assuming that the current I1 is varying with time, then an induced emf is created at
coil II, given as
dΦ12
ind,2 = −N2 (10.35)
dt
where Φ12 is calculated from Eq. (10.34) as

M12
Φ12 = I1 (10.36)
N2

Therefore, substituting Eq. (10.36) into Eq. (10.35), we get

 
d M12
ind,2 = −N2 I1 (10.37)
dt N2
d I1
= −M12
dt
10.2 Inductance 243

Similarly, if I2 is the current passing through the coil II, which varies on time,
then there exists a mutual inductance M21 such that

Φ21
M21 = N1 (10.38)
I2

In Eq. (10.38), Φ21 is the magnetic flux through coil I due to magnetic field created
by coil II. The induced emf at coil I is

dΦ21
ind,1 = −N1 (10.39)
dt
Substituting Φ21 from Eq. (10.38) into Eq. (10.39), we obtain

d I2
ind,1 = −M21 (10.40)
dt
Assuming that coil I and coil II are identical, then M12 = M21 ≡ M because they
depend only on the coil’s geometry. Therefore, we obtain the following:

d I2
ind,1 = −M , (10.41)
dt
d I1
ind,2 = −M
dt
A circuit containing a large self-inductance is called an inductor. Based on Lenz’s
law, the self-inductance prevents the current in the circuit from increasing or decreas-
ing instantaneously.

10.3 Oscillations in an LC Circuit

Consider a capacitor connected to an inductor, as shown in Fig. 10.10. That is also

called an LC circuit. Assuming initially that the capacitor is fully charged (that it, its
maximum charge is Q max ), then the switch is closed. The current and the charge will
oscillate between its maximum positive and negative values. There is no resistance
in the circuit (that is, R = 0); therefore, there is no internal energy losses. Besides,
we assume that there is no energy radiated away. Q max denotes the initial charge
in capacitor. Furthermore, the switch is closed at t = 0. Therefore, the energy of
capacitor at t = 0 is
Q2
UC = max (10.42)
2C
Moreover, at t = 0, the current in the circuit is zero, I = 0, and hence there is
no energy stored in the inductor. When the switch closes, the rate of charge leaving
244 10 More About Faraday’s Law of Induction

Fig. 10.10 The LC circuit

capacitor equals the current in the circuit. As the capacitor gets discharged, its electric
field gets decreased. During this process the current in the circuit increases, and thus
the magnetic field of the inductor increases (note that B ∼ I ). As a result of that, the
energy transfers from electric energy of the capacitor into magnetic field energy of
the inductor.
When capacitor is fully discharged, the energy in the capacitor becomes zero, the
current in the circuit reaches its maximum value, and hence all the energy is stored in
the inductor. Then, the current continues going to capacitor, and thus increasing its
charge and the current decreases. At the end, the capacitor gets back fully charged;
however, its polarity is opposite. This process repeats by transferring electrical energy
of the capacitor into magnetic field energy of inductor, and vice versa.
As analogy, one can consider the mechanical oscillations of a mass m around
its equilibrium position x = 0. The oscillations of charge in the capacitor of an LC
circuit are analogue of the oscillations in the displacement x of the mass m from
its equilibrium position, and oscillations in the current value of an LC circuit are
analogue of the oscillations in the velocity of the mass m.
Now, we consider an intermediate step at time t, when the charge in the capacitor
is Q (Q < Q max ), and thus the electrical energy of capacitor is

Q2
UC = (10.43)
2C
At the same time t, the current in the circuit is I , and the magnetic field energy stored
in the inductor is
LI2
UL = (10.44)
2
Then, the total energy in the LC circuit at the time t is the sum of the two terms given
by Eqs. (10.43) and (10.44), given by the expression:

Q2 LI2
U = UC + U L = + (10.45)
2C 2
Since the resistance is zero in the circuit, there is no energy transformation into
internal energy (which means no losses of the initial energy given by Eq. (10.42)).
Therefore, U is constant at any time t, and equal to its value at t = 0:
10.3 Oscillations in an LC Circuit 245

Q 2max
U= (10.46)
2C
Furthermore, dU/dt = 0:

Q dQ LI dI
0=2 +2 (10.47)
2C dt 2 dt
Alternatively, using the definition of the current I = d Q/dt, we obtain

d2 Q 1
+ Q=0 (10.48)
dt 2 LC

where the relation d I /dt = d 2 Q/dt 2 is used. Denoting by

1
ω2 = (10.49)
LC
we obtain the following second-order differential equation:

d2 Q
+ ω2 Q = 0 (10.50)
dt 2
The solution of Eq. (10.50) is given as

Q(t) = Q max cos (ωt + φ) (10.51)

where ω represents the angular frequency that depends on L and C (see Eq. (10.49)):

1
ω= (10.52)
LC

which is also called natural frequency. In Eq. (10.51), φ is the phase angle, determined
by the initial conditions, at t = 0. For instance, at t = 0, we have Q(t = 0) = Q max ,
and thus from Eq. (10.51):
Q max = Q max cos (φ) (10.53)

Thus, using Eq. (10.53), we find that φ = 0, and Eq. (10.51) can be written as

Q(t) = Q max cos (ωt) (10.54)

The current I flowing in the LC circuit is determined as

d Q(t)
I = = −ω Q max sin (ωt) ≡ −Imax sin (ωt) (10.55)
dt
where Imax = ω Q max .
246 10 More About Faraday’s Law of Induction

Fig. 10.11 The oscillations

of charge and current in the
LC circuit. In the numerical
calculations, ω = π and
Q max = 1

The period of oscillations is

2π
T = (10.56)
ω
Thus, we can also write
2π
ωt = t (10.57)
T
In Fig. 10.11, we show the oscillations of charge and current in the LC circuit. In
the numerical calculations, ω = π and Q max = 1, used as illustration.
Using Eq. (10.45), the total energy of the system is

1 2 L 2
U= Q max cos2 (ωt) + Imax sin2 (ωt) (10.58)
2C
   2  
electric energy magnetic energy

As it can be seen, Eq. (10.58) indicates that the energy in the LC circuit continuously
oscillates between the electric energy stored in capacitor and magnetic field energy
stored in inductor, as shown in Fig. 10.12. Because there is no loss of energy in
resistance, then
1 2 L 2
Q max = Imax (10.59)
2C 2
Therefore, Eq. (10.58) can also be written as

Q 2max  2 Q2
U= cos (ωt) + sin2 (ωt) = max (10.60)
2C 2C
which is constant.
10.4 The RL Circuit 247

Fig. 10.12 The oscillations

of the electrical energy
stored in capacitor and
magnetic field energy stored
in inductor for the LC
circuit. In the numerical
calculations: ω = π, C = 1,
and Q max = 1

10.4 The RL Circuit

Consider a circuit consisting of the resistance R and the inductance L, as shown in

Fig. 10.13. This is also called an RL circuit. The switch S is closed as t = 0. Because
of the inductance of the inductor a back emf is produced. Furthermore, an inductor
in a circuit opposes the change in the current passing through the circuit. The source
current I starts increasing, and back emf that opposes the increase of the current is
induced in the inductor:
dI
 L = −L (10.61)
dt
Since the current I is increasing, d I /dt > 0, and thus  L < 0. Therefore, the polarity
of  L is opposite of source emf  (see also Fig. 10.13).
Using Kirchhoff’s law, we write

 + VR +  L = 0 (10.62)

where VR = −I R (| VR |= I R) is the voltage drop in the resistance. Combining

Eqs. (10.61) and (10.62), we get

Fig. 10.13 The RL circuit

248 10 More About Faraday’s Law of Induction

dI
− IR− L =0 (10.63)
dt
Rearranging Eq. (10.63), we can write that

dI
L + IR =  (10.64)
dt
For finding the general solution of Eq. (10.64) concerning I , we first solve the
following homogeneous equation:

dI
L + IR = 0 (10.65)
dt
which can be arranged as
dI R
= − dt (10.66)
I L
Integrating both sides of Eq. (10.66), we obtain

I t
dI R
=− dt (10.67)
I L
I0 0

Or,  
I R
ln =− t (10.68)
I0 L

which can be solved for I to give

I = I0 exp (−t/τ ) (10.69)

where
L
τ= (10.70)
R
is a time constant. Assuming that the integration constant I0 is a function of time t,
then substituting Eq. (10.69) into Eq. (10.64), we find
 
d I0 I0
L exp (−t/τ ) − exp (−t/τ ) + R I0 exp (−t/τ ) =  (10.71)
dt τ

which can be simplified as

d I0 
= exp (t/τ ) (10.72)
dt L
Solving Eq. (10.72) for I0 , we find
10.4 The RL Circuit 249

 
I0 = exp (t/τ ) dt = exp (t/τ ) + C (10.73)
L R

where C is an integration constant determined by the initial conditions at t = 0 and

Eq. (10.70) is used. First, we substitute Eq. (10.73) into Eq. (10.69):

I (t) = + C exp (−t/τ ) (10.74)
R
Using the initial conditions that I (t = 0) = 0, from Eq. (10.74), we get

0 = I (0) = + C exp (−0/τ ) (10.75)
R
Or,

C =− (10.76)
R
Substituting the expression for C (Eq. (10.76)) into the expression for I (Eq. (10.74)),
we obtain the current as

I (t) = (1 − exp (−t/τ )) (10.77)
R
In Eq. (10.77), τ is the time that takes for the current in the circuit to become
 
I (τ ) = (1 − exp(−1)) ≈ 0.63 (10.78)
R R
In Fig. 10.14, we show I (t) as a function of time for the RL circuit. The induced
emf is given by Eq. (10.61), where the rate change of the current is
 
dI  1 
= exp (−t/τ ) = exp (−t/τ ) (10.79)
dt R τ L

Therefore, we find that

 L = − exp (−t/τ ) (10.80)

Graphically  L versus time t is shown in Fig. 10.14. It can be seen that as t increases,
the induced  L approaches zero, which corresponds with the time when the current
in the RL circuit reaches the equilibrium maximum value, Imax = /R.
Based on Lenz’s law, the induced emf in an inductor prevents the source emf (that
is, battery) from establishing an instantaneous current. In that case, the source emf
has to do work against the inductor to create the current. Therefore, part of the energy
supplied by the source emf appears as internal energy in the resistor:

Uint = I 2 Rt (10.81)
250 10 More About Faraday’s Law of Induction

Fig. 10.14 The current and

the induced emf as a function
of time for the RL circuit

where t is the time during which the current flows in the resistance. The other part
of energy is stored as the magnetic field energy of the inductor:

UL = I | L | t (10.82)

The rate of the total energy is then written as

dUtot dUint dU L
= + (10.83)
dt dt dt
Therefore, we find
dI
I = I2R + I L (10.84)
dt
Here, U L , which represents the energy stored in the inductor at any instance of time
t, is such that its rate is given as

dU L dI
= IL (10.85)
dt dt
Or, we can write that
dU L = I Ld I (10.86)

Integrating both sides of Eq. (10.86), we obtain

U L I
LI2
UL = dU L = I Ld I = (10.87)
2
0 0

which represents the energy stored in the inductor in the form of the magnetic field
when the current is I .
10.4 The RL Circuit 251

Consider a solenoid of N turns, with self-inductance given as L = N Φ B /I ,

which can be written as

L = μ0 n 2 V = μ0 n 2 A (10.88)

where A is the cross-sectional area of the solenoid and  its length, and V
the volume of the solenoid. Here, Φ B is the magnetic field flux through the
solenoid given as

ΦB = B · A (10.89)
μ0 N I
= A

= μ0 n I A

The magnetic field of the solenoid is

μ0 N I
B= = μ0 n I (10.90)

and thus
B
I = (10.91)
μ0 n

Then, the energy stored in magnetic field is

LI2
UB = (10.92)
2
 I2
= μ0 n 2 A
2
 B2
= μ0 n 2 A
2μ20 n 2
B2 B2
= A = V
2μ0 2μ0

Furthermore, the magnetic field energy density is

UB B2
uB = = (10.93)
V 2μ0
252 10 More About Faraday’s Law of Induction

10.5 The RLC Circuit

Now, we consider the RLC circuit, which contains a resistance (R), inductor (L),
and a capacitor (C), as shown in Fig. 10.15. Also, we assume that the capacitor is
initially fully charged, that is, its charge at t = 0 is Q max , which corresponds to the
state before the switch is closed. When the switch is closed, a current is established
in the circuit. The total energy stored in capacitor and inductor is given as

LI2 Q2
U LC = U L + UC = + (10.94)
2 2C
However, U LC , given by Eq. (10.94), is not constant because of the losses of the
energy as internal energy in the resistance R. The rate of the energy transformation
into internal energy within the resistor is given as

dU R
= −I 2 R (10.95)
dt
where minus sign indicates that is an energy lost, and hence the rate decreases with
time t (explaining the sign minus in front of I 2 R in Eq. (10.95)). Therefore, we can
write
dU LC dU R
= (10.96)
dt dt
Using Eqs. (10.94) and (10.95), we obtain

dI Q dQ
LI + = −I 2 R (10.97)
dt C dt
where Q is the charge in the capacitor at the time t and I is the current in the circuit
I = d Q/dt. The expression in Eq. (10.97) can further be written as

d2 Q dQ Q
L 2
+R + =0 (10.98)
dt dt C
which is a non-homogeneous second-order differential equation for the variable Q.

Fig. 10.15 The RLC circuit

10.5 The RLC Circuit 253

For solving Eq. (10.98), we assume the general form of the solution as follows:

Q(t) = exp (ρt) (10.99)

where ρ is a parameter to be defined. First, we substitute the expression for Q (given

by Eq. (10.99)) into Eq. (10.98), and obtain

1 ρt
Lρ2 eρt + Rρeρt + e =0 (10.100)
C
or
R 1
ρ2 + ρ+ =0 (10.101)
L LC
which is a quadratic equation for ρ. The roots of Eq. (10.101) are given as

R R2 1
ρ1,2 =− ± − (10.102)
2L 4L 2 LC

R 1 R2C
=− ±√ −1
2L LC 4L

In Eq. (10.102), we do the following substitutions:

1
ω0 = √ (10.103)
LC

and 
R2C
ξ= (10.104)
4L

The first term in Eq. (10.102) is rewritten as


R R2C 1
= ·√ ≡ ξω0 (10.105)
2L 4L LC

Then, Eq. (10.102) takes the form


ρ1,2 = −ξω0 ± ω0 ξ 2 − 1 (10.106)

where ω0 is the angular frequency of the free oscillations in the LC circuit. In analogy
with classical mechanics, the RLC circuit corresponds to the dumping motion of the
oscillator in presence of the friction force.
254 10 More About Faraday’s Law of Induction

10.5.1 Case 1

First, we consider the case for R = 0, and so ξ = 0. That is, the LC circuit, and hence
free oscillations establish where the energy is transformed between the electrical
energy stored in the capacitor and magnetic field energy stored in the inductor. From
Eq. (10.106), we have
ρ1,2 = ±iω0 (10.107)
√
where i = −1. Therefore, the charge at time t is

Q(t) = A exp (iω0 t) + B exp (−iω0 t) (10.108)

Using the initial conditions that at t = 0, Q(0) = Q max , and so A + B = Q max . Then,
Eq. (10.108) simplifies as

Q(t) = A exp (iω0 t) + Q max exp (−iω0 t) − A exp (−iω0 t) (10.109)

= (2 A − Q max )i sin (ω0 t) + Q max cos (ω0 t)

where the relation exp (±iφ) = cos φ ± i sin φ is used.

Moreover, at some finite time t = 3T /4 (where T = 2π/ω0 is the period), the
charge Q(3T /4) = 0 (see also Fig. 10.11); therefore, the first term must have zero
amplitude, 2 A − Q max = 0 (because cos (ω0 3T /4) = 0). Then, we obtain

Q(t) = Q max cos (ω0 t) (10.110)

which is the solution obtained for LC circuit.

10.5.2 Case 2

Consider the case when | ξ |< 1, hence ξ 2 − 1 < 0. Therefore,


ρ1,2 = −ξω0 ± ω0 i 1 − ξ 2 (10.111)

Substituting expression for ρ from Eq. (10.111) into Eq. (10.99), we obtain
  
Q(t) = A exp (−ξω0 t) exp iω0 1 − ξ 2 t (10.112)
  
+ B exp (−ξω0 t) exp −iω0 1 − ξ 2 t
      
= exp (−ξω0 t) A exp iω0 1 − ξ 2 t + B exp −iω0 1 − ξ 2 t
      
= exp (−ξω0 t) a cos ω0 1 − ξ 2 t + b sin ω0 1 − ξ 2 t
10.5 The RLC Circuit 255

Fig. 10.16 The dumping

oscillations in a RLC circuit.
Here, ω0 = π, ξ =  0.05,
A = 1, and α ≡ 1 − ξ 2

It can be seen that the amplitude is not constant, but it varies with time according to
exp (−ξω0 t), which decays to zero for t → ∞. That factor is also called the dump-
ing exponential factor. The dumping oscillations in the RLC circuit are illustrated
in Fig. 10.16, where we can distinguish the exponential decay of the amplitude,
oscillations of the cosine factor, and the dumping oscillations of Q(t).

10.5.3 Case 3

The third case is for | ξ |> 1. Then, the roots are


ρ1,2 = −ξω0 ± ω0 ξ 2 − 1 (10.113)

and the charge in the capacitor at any time t is

Q(t) = A exp (−ρ1 t) + B exp (+ρ2 t) (10.114)

10.6 Alternating Current Circuits

10.6.1 AC Sources and Phases

In general, the AC circuit consists of circuit elements and a generator (see also
Fig. 10.4). It provides an alternative current (AC). As discussed previously, the prin-
ciple of a generator is based on Faraday’s law of induction. From Fig. 10.4, a con-
ducting loop rotates at constant speed ω, then a sinusoidal voltage (or emf) is induced
in the loop with an instantaneous value of

ΔV = V0 sin (ωt) (10.115)

256 10 More About Faraday’s Law of Induction

In Eq. (10.115), V0 is the maximum voltage output of the AC generator (or alternating
voltage amplitude). Furthermore, the relationship between angular frequency ω and
linear frequency f (or period T = 1/ f ) is

2π
ω = 2π f = (10.116)
T
The analysis of the circuit is performed using the so-called phasor diagram.

10.6.2 Resistors in an AC Circuit

An AC generator has a symbol in the circuit as shown in Fig. 10.17a.

Now, we consider a circuit consisting of a resistor R connected to a generator
with a sinusoidal voltage as given in Eq. (10.115) (see also Fig. 10.17b). The voltage
drop in the resistance ΔVR equals the difference of the electric potential between the
points a and b:
ΔVR = Vab (10.117)

Using Kirchhoff’s law, the algebraic sum of the voltages around a closed loop in
a circuit is equal to zero, and hence we write

ΔV − ΔVR = 0 (10.118)

Combining Eqs. (10.115) and (10.118), we have

ΔV = ΔVR = V0 sin (ωt) (10.119)

Fig. 10.17 a AC generator

symbol and b the circuit
consisting of a resistor and
an AC generator of voltage
ΔV = V0 sin (ωt)
10.6 Alternating Current Circuits 257

Fig. 10.18 The plots of the

instantaneous current i R and
voltage ΔV R for R = 2 ,
V0 = 1 V, and ω = π rad/s

Therefore, instantaneous current in the resistor is

ΔVR V0
iR = = sin (ωt) (10.120)
R R
= I0 sin (ωt)

where I0 is the maximum value of current (or the amplitude):

V0
I0 = (10.121)
R
Then, we can write the instantaneous voltage across the resistor as

ΔVR = R I0 sin (ωt) (10.122)

The plots of the instantaneous current i R and voltage ΔVR are shown in Fig. 10.18,
for R = 2 , V0 = 1 V, and ω = π rad/s.
The phasor diagram can be used to represent current-voltage phase relationship.
For the circuit in Fig. 10.17, the phasor diagram is shown in Fig. 10.19. It can be
seen that

T
(R) 1
Iav = i R (t)dt (10.123)
T
0
T
1
= I0 sin (ωt) dt
T
0
1 I0
= (cos(0) − cos(ωT ))
T ω
I0
= ·0=0
ωT
258 10 More About Faraday’s Law of Induction

Fig. 10.19 The phasor

diagram

Thus, the average value of the current during one period T is equal to zero. That is,
the current for a period of time T /2 maintains one sign (e.g., negative) and for the
same period of time T /2 the opposite sign (e.g., positive). Hence, the sum is zero.
The direction of the current has no effect on the behavior of the resistance. That is,
as the current flows through the resistor, electrons will collide with the fixed atoms
of the material of the resistor, and if the current changes the direction, the effect
remains the same. That is, the temperature of the resistor will increase due to these
collisions.
Quantitatively, the rate at which the electrical energy is converted into internal
energy is
P = i R2 R (10.124)

where P is the power, i R is the instantaneous current in the resistor, and R is the
resistance. Since P is proportional to i R2 , then P does not depend on the sign of i R .
However, the temperature increase of the AC current having a maximum value of I0
is not the same as that produced by a direct current equal to I0 . Let us consider an
instantaneous current as in Eq. (10.112). The average power Pav (AC)
over a period T
is

T
1
Pav
(AC)
= P(t)dt (10.125)
T
0
T
1
= i R2 (t)Rdt
T
0
T
R
= I02 sin2 (ωt)dt
T
0
10.6 Alternating Current Circuits 259

T
R 1 − cos(2ωt)
= I02 dt
T 2
0
R T
= I02
T 2
I2R
= 0
2
= Irms
2
R

where Irms is the so-called the root mean square current:



 T
1
Irms = i R2 (t)dt (10.126)
T
0

which can be calculated as

T
1
2
Irms = i R2 (t)dt (10.127)
T
0
T
1
= I02 sin2 (ωt) dt
T
0
T
I2 1 − cos (2ωt)
= 0 dt
T 2
0
I02
=
2
Or,
I0
Irms = √ ≈ 0.707I0 (10.128)
2

On the other hand, for a direct current (DC) with I = I0 :

Pav
(DC)
= P = I02 R = 2Pav
(AC)
(10.129)

which implies that the average power of a DC is two times larger of an AC working
at the same maximum current, I0 .
260 10 More About Faraday’s Law of Induction

Note that the root mean square voltage, Vrms , is



 T
1
Vrms = ΔVR2 (t)dt (10.130)
T
0

which can be evaluated as follows:

T
1
2
Vrms = ΔVR2 (t)dt (10.131)
T
0
T
1
= V02 sin2 (ωt) dt
T
0
T
V2 1 − cos (2ωt)
= 0 dt
T 2
0
V02
=
2
Alternatively,
V0
Vrms = √ ≈ 0.707V0 (10.132)
2

10.6.3 Inductors in an AC Circuit

In Fig. 10.20, an inductor in the AC circuit is shown. The voltage drop across the
inductor equals the electric voltage between a and b endpoints: Vab = ΔVL . On the
other hand, the self-induced voltage across the inductor is given as

di L
ΔVL ≡  L = −L (10.133)
dt
Using Kirchhoff’s law:
ΔV + ΔVL = 0 (10.134)

and thus
di L
V0 sin (ωt) = ΔV = −ΔVL = L (10.135)
dt
From Eq. (10.135), we find that
10.6 Alternating Current Circuits 261

Fig. 10.20 The inductor in

an AC circuit


V0
i L (t) = sin (ωt) dt (10.136)
L
V0
=− cos (ωt)
Lω
V0  π
= sin ωt −
Lω 2
where the trigonometric relation cos φ = − sin (φ − π/2) is used. Comparing the
expression given by Eq. (10.136) with the general form of i L :

i L (t) = I0 sin (ωt ± φ) (10.137)

we write that
V0 V0
I0 = ≡ (10.138)
Lω XL

where
X L = Lω (10.139)

is the so-called inductive reactance.

Using Eq. (10.133), we can determine self-induced voltage drop across the induc-
tor as
di L
 L (t) = −L (10.140)
dt
V0
= −L ω sin (ωt)
Lω
= −V0 sin (ωt)
= −I0 X L sin (ωt)

where
V0 = X L I0 (10.141)
262 10 More About Faraday’s Law of Induction

Fig. 10.21 The phasor

diagram of an inductor in an
AC circuit

In Fig. 10.21, we show the phasor diagram of an inductor in an AC circuit.

Note that
Vrms
Irms = (10.142)
XL

10.6.4 Capacitors in an AC Circuit

Consider a capacitor in an AC circuit, as shown in Fig. 10.22. The output voltage of

the AC generator is
ΔV = V0 sin (ωt) (10.143)

The instantaneous voltage across the capacitor is

ΔVC = Vab (10.144)

Applying Kirchhoff’s law, we write

ΔVC = ΔV = V0 sin (ωt) (10.145)

Using the following relationship:

q
C= (10.146)
ΔVC

Alternatively, the instantaneous charge in the capacitor is

q = CΔVC (10.147)

and furthermore we have the instantaneous current as

10.6 Alternating Current Circuits 263

Fig. 10.22 The capacitor in

an AC circuit

dq d d
iC = = C (ΔVC ) = C (ΔV ) (10.148)
dt dt dt
= C V0 ω cos (ωt)
 π
= C V0 ω sin ωt +
2
Hence, we obtain  π
i C = I0 sin ωt + , (10.149)
2
where
V0
I0 = V0 (Cω) ≡ (10.150)
XC

where
1
XC = (10.151)
Cω
is the so-called capacitive reactance.
In Fig. 10.23, we present the phasor diagram of the capacitor in an AC circuit.
Besides, we can write that
Vrms
Irms = (10.152)
XC

10.6.5 The RLC Series in an AC Circuit

Next, consider the RLC series in an AC circuit, as presented in Fig. 10.24. The series
consists of resistor (R), inductor (L), and the capacitor (C). The generator produces
an alternating voltage given as
264 10 More About Faraday’s Law of Induction

Fig. 10.23 The phasor

diagram of an capacitor in an
AC circuit

Fig. 10.24 The RLC series

in an AC circuit

ΔV = V0 sin (ωt) (10.153)

The current varies with time as

i(t) = I0 sin (ωt − φ) (10.154)

where φ is a phase angle shift between the voltage and the current, determined in the
following.
Note that the current passing across each element of the series has the same ampli-
tude. Therefore, the voltage drops in resistor, inductor, and capacitor are, respectively,
given as

ΔVR = (I0 R) sin (ωt) = V0,R sin (ωt) , (10.155)

 π
ΔVL = (I0 X L ) sin ωt − = −V0,L cos (ωt) ,
 2
π
ΔVC = (I0 X C ) sin ωt + = V0,C cos (ωt)
2
where

V0,R = I0 R , (10.156)
10.6 Alternating Current Circuits 265

Fig. 10.25 The phasor

diagram of the RLC series in
an AC circuit

V0,L = I0 X L ,
V0,C = I0 X C

Using Kirchhoff’s law, we write

ΔV = ΔVR + ΔVL + ΔVC ≡ V0 sin (ωt + φ) (10.157)

where φ is a phase angle shift between the voltage and the current, which has to be
determined. The phasor diagram is used to determine V0 and φ, which is shown in
Fig. 10.25. From the phasor diagram, the amplitude of the resultant voltage drop in
the circuit is
  2
V0 = V0,R 2
+ V0,L − V0,C (10.158)

= (I0 R)2 + (I0 X L − I0 X C )2

= I0 R 2 + (X L − X C )2

Denoting by Z the so-called impedance:


Z= R 2 + (X L − X C )2 (10.159)

we obtain Ohm’s law for the RLC series in an AC circuit as

V0 = I0 Z (10.160)

Alternatively,
V0
I0 = (10.161)
Z
Furthermore, the phase angle shift is found from the tangent as
266 10 More About Faraday’s Law of Induction

Fig. 10.26 The variations of

the voltage drops in each
element of the RLC series in
an AC circuit

X L − XC
tan φ = (10.162)
R
as  
−1 X L − XC
φ = tan (10.163)
R

In general, the sign of the angle φ is defined by the following rules. If X L > X C , φ < 0
because it is measured clockwise; if X L < X C , φ > 0 because it is an angle measured
counterclockwise. For X L = X C , tan φ = 0, and thus φ = 0; that is, Z = R, and so
here is no phase angle shift between the voltage and the current:

V0
I0 = (10.164)
R
The frequency for which X L = X C is called the resonance frequency.
In Fig. 10.26, we show the plots of ΔVR , ΔVL , and ΔVC to indicate the phase
shifts between the alternating voltage drops in every element of the circuit.

10.7 Power in the AC Circuit

In the case of the AC circuit, the instantaneous power delivered by the generator is

P = i(t) · ΔV (t) (10.165)

= (I0 sin (ωt − φ)) · (V0 sin (ωt))

which corresponds to the general case of the RLC series in an AC circuit. Using the
trigonometric relationship:

sin(α ± β) = sin(α) cos(β) ± cos(α) sin(β) (10.166)

10.7 Power in the AC Circuit 267

we can write Eq. (10.165) as follows:

P = I0 V0 (sin(ωt) cos(φ) − cos(ωt) sin(φ)) sin (ωt) (10.167)


= I0 V0 sin2 (ωt) cos(φ) − sin(ωt) cos(ωt) sin(φ)
 
1
= I0 V0 sin (ωt) cos(φ) − sin(2ωt) sin(φ)
2
2

The average delivered power for a period T is then calculated as

T
1
Pav = P(t)dt (10.168)
T
0
T  
I0 V0 1
= sin (ωt) cos(φ) − sin(2ωt) sin(φ) dt
2
T 2
0
⎛ ⎞
T T
I0 V0 ⎝ 1
= cos(φ) sin2 (ωt)dt − sin(φ) sin(2ωt)dt ⎠
T 2
0 0
⎛
T
I0 V0 ⎝ 1 − cos(2ωt)
= cos(φ) dt
T 2
0

1
+ sin(φ) (cos(2ωT ) − cos(0))
4ω
 
I0 V0 T
= cos(φ) + 0
T 2
I0 V0
= cos(φ)
2
Therefore, the average power is

I0 V0
Pav = cos(φ) (10.169)
√2  √ 
2Irms 2Vrms
= cos(φ)
2
= Irms Vrms cos(φ)
≡ F Irms Vrms

where
F = cos(φ) (10.170)

is the so-called power factor.

268 10 More About Faraday’s Law of Induction

Moreover, from Fig. 10.25, V0,R = V0 cos(φ) = I0 R; therefore,

I0 R
cos(φ) = (10.171)
V0

and the average power is

I0 V0 I0 R I2R
Pav = = 0 (10.172)
2 V0 2

Alternatively,
Pav = Irms
2
R (10.173)

As a result, the average power delivered by a generator of AC circuit is converted into

the internal energy in the resistor (which is similar to the DC circuit). Furthermore,
if there is no resistor, then there is no power loss.
Moreover, if the load is just a resistor (that is, X C = X L = 0), then

Pav = Irms Vrms (10.174)

which indicates that the power factor is one, F = 1.

10.8 Resonance in the RLC Series Circuit

The current rms is given by

Vrms
Irms = (10.175)
Z
where Z is given by Eq. (10.159). Therefore, we can also write that

Vrms
Irms =  (10.176)
R 2 + (X L − X C )2

From Eq. (10.176), if X L = X C , we say that there is a resonance; that is,

Vrms
Irms = . (10.177)
R
That indicates that Irms has maximum value. The frequency for which that occurs
can be found using the following relation:

1
ω0 L = (10.178)
ω0 C

where ω0 denotes the resonance frequency, which can be obtained as

10.8 Resonance in the RLC Series Circuit 269

Fig. 10.27 The current rms

versus frequency ω for
different resistance values
R 4 < R 3 < R 2 < R 1 . ω0
represents the resonance
frequency

1
ω0 = √ (10.179)
LC

Besides, using Eq. (10.176), we obtain

Vrms
Irms =    (10.180)
1 2
R + ωL −
2
ωC
Vrms
=  2 2
ω LC − 1
R +
2
ωC

In Fig. 10.27, we plot the current rms versus frequency ω for different resistance
values R4 < R3 < R2 < R1 . ω0 represents the resonance frequency. That indicates
that Irms , given by Eq. (10.180), is maximum at ω = ω0 , and hence at this frequency,
the generator must deliver maximum power.
We can express the average power as a function of the frequency ω as
 2
Vrms
Pav = 2
Irms R = R (10.181)
Z
2
Vrms
= R
R 2 + (X L − X C )2
2
Vrms
=   R
1 2
R + ωL −
2
ωC
2
Vrms
= R
L2  2
2 2
R + 2 ω − ω0
2
ω
2
Vrms Rω 2
=  2
R 2 ω 2 + L 2 ω 2 − ω02
270 10 More About Faraday’s Law of Induction

Fig. 10.28 The average power versus frequency ω for two different resistance values. ω0 represents
the resonance frequency. Δω = ω2 − ω1 is the width measured at half of the maximum power

From Eq. (10.181), if ω = ω0 (that is, at resonance condition), then power delivered
by the generator is maximum, and given as
2
Vrms
Pav
(max)
= (10.182)
R
In Fig. 10.28, we plot the average power versus frequency ω. The sharpness of
the average power curve is characterized by the dimensionless parameter η:
ω0
η= (10.183)
Δω
which is also called quality factor. In Eq. (10.183), Δω is the width of the curve
calculated as the following difference:

Δω = ω2 − ω1 (10.184)

where ω1 and ω2 are the frequencies for which Pav = Pav

(max)
/2 (see also Fig. 10.28).
Here,
R
Δω = (10.185)
L
and thus
ω0 L
η= (10.186)
R
10.9 Exercises 271

10.9 Exercises

Exercise 10.1 Consider a motor in which the coils have a total resistance of
10  and the applied electric voltage is 120 V. When the motor is running at
its maximum speed, the back emf is 70 V. Find the current in the coil when a)
the motor is turned on and b) it has reached maximum speed.

Solution 10.1 (a) When the motor is turned on, the back emf is zero, and thus

ind 120 V
I = = = 12 A . (10.187)
R 10 
(b) At maximum speed, the back emf has its maximum value, and hence

ind − back 120 V − 70 V

I = = = 5.0 A . (10.188)
R 10 

Exercise 10.2 Consider an AC generator consisting of 8 turns, each with area

A = 0.09 m2 , and the total resistance of the wire is R = 12 . The loop rotates
in a B = 0.5 T magnetic field at constant frequency of f = 60 Hz. a) Find the
maximum induced emf and b) what is the maximum induced current?

Solution 10.2 (a) The angular speed is

ω = 2π f = 2π(60 Hz) ≈ 377 rad/s . (10.189)

Then, the induced emf is

ind = N ABω sin (ωt) (10.190)

and

ind,max = N ABω = 8(0.09 m2 )(0.5 T)(377 rad/s) ≈ 136 V . (10.191)

(b) The maximum induced current is

ind,max 136 V
Iind,max = = ≈ 11.3 A . (10.192)
R 12 
272 10 More About Faraday’s Law of Induction

Exercise 10.3 A long solenoid of radius R has n turns of wires per unit of
length and carries a time-varying current that as a function of time is

I = Imax cos (ωt) , (10.193)

where Imax is the maximum current and ω is the angular frequency. (a) Find
the magnitude of the induced electric field outside solenoid, at distance r > R
from its long central axis. (b) What is magnitude of the induced electric field
inside solenoid, at distance r from central axis.

Solution 10.3 (a) Using the following relationship:

dΦ B
E · dl = − (10.194)
L dt

where Φ B = B A = Bπ R 2 and E is constant. Therefore,

E · dl = Edl = E dl = E2πr (10.195)
L L L

Combining Eqs. (10.194) and (10.195), we get

d(Bπ R 2 ) dB
E2πr = − = −π R 2 (10.196)
dt dt
The magnitude of the electric field is

R2 d B
E =− (10.197)
2r dt
Using Ampére’s law, we write

B · dl = μ0 I (10.198)
L

Or,
B1 · l = μ0 I (10.199)

Then, the magnetic field strength is

μ0 I
B1 = (10.200)
l
10.9 Exercises 273

For N turns, we have

μ0 I
B=N = nμ0 I (10.201)
l
Writing the maximum magnetic field as Bmax = nμ0 Imax , then

B = nμ0 Imax cos (ωt) (10.202)

The derivative of B for time t is

dB
= −nμ0 Imax ω sin (ωt) (10.203)
dt
Substituting expression given by Eq. (10.203) into Eq. (10.197), we get

R2
E= nμ0 Imax ω sin (ωt) (10.204)
2r
(b) For r < R,
Φ B = Bπr 2 (10.205)

Then
dB
E2πr = −πr 2 (10.206)
dt
Therefore,
r
E= nμ0 Imax ω sin (ωt) (10.207)
2

Exercise 10.4 The conducting bar shown in Fig. 10.1 of mass m and length
l moves on two parallel frictionless rails in presence of a uniform magnetic
field directed into the page. The bar is given an initial velocity vi to the right
and it is released at t = 0. Find the speed of the bar as a function of time.

Solution 10.4 The induced current is counterclockwise and the magnetic force is

F B = I (L × B) = −I l Bi (10.208)

where i is the unit vector along the directions of motion (to the right). Thus,

vi = vi i (10.209)
274 10 More About Faraday’s Law of Induction

Using the second law of Newton:

dv
F B = ma = m (10.210)
dt
Combining Eqs. (10.208) and (10.210), we get

dv dv
− I l Bi = m =m i (10.211)
dt dt
Or,
dv
− Il B = m (10.212)
dt
Using Faraday’s law, we write

dΦ B d(Blx) dx
ind = − =− = −Bl = −Blv (10.213)
dt dt dt
Then, the induced current is

| ind | Blv
I = = (10.214)
R R
Or,
RI
Bl = (10.215)
v
Substituting Eq. (10.215) into Eq. (10.212), we obtain
 
Blv dv
− Bl =m (10.216)
R dt

Or,  
B 2l 2 v dv
− =m (10.217)
R dt

After rearranging it, we write

 
dv B 2l 2
=− dt (10.218)
v mR

Integrating both sides of Eq. (10.218), we obtain

v   t
dv B 2l 2
=− dt (10.219)
v mR
vi 0
10.9 Exercises 275

Or,    
v B 2l 2
ln =− t (10.220)
vi mR

Denoting by τ a time characteristic as

mR
τ= (10.221)
B 2l 2
Then, we get  
t
v(t) = vi exp − (10.222)
τ

The induced current is (from Eq. (10.214)

 
Bl t
I = vi exp − (10.223)
R τ

and using Eq. (10.213), the induced emf is

 
t
ind = −Blvi exp − (10.224)
τ

Exercise 10.5 Consider a conducting bar of length l rotates with constant

angular speed ω about a pivot at one end. A uniform magnetic field B is
directed perpendicular to the plane of rotation (see also Fig. 10.29). Find the
motion emf induced between the ends of the bar.

Solution 10.5 We consider a segment of length dr having a linear velocity v at a

distance r from the pivot, as shown in Fig. 10.29. The linear speed is given as v = ωr .
The magnitude of the induced emf in this segment is

dind = Bvdr (10.225)

The total induced emf is then calculated as

l l
ind = Bvdr = B(ωr )dr (10.226)
0 0

After integration, we find that

Bωl 2
ind = (10.227)
2
276 10 More About Faraday’s Law of Induction

Fig. 10.29 A conducting bar

of length l rotates with
constant angular speed about
a pivot at one end. A uniform
magnetic field B is directed
perpendicular to the plane of
rotation

Exercise 10.6 Find the inductance of a uniformly wound solenoid having N

turns and length l. Assume that l r , with r being the radius of the winding.
The core of the solenoid is air.

Solution 10.6 We can use Ampére’s law to find the magnetic field strength as

B · dl = N μ0 I (10.228)
L

where I is the current passing through the coil. Alternatively,

Bl = N μ0 I (10.229)

Therefore, the magnetic field strength is

N μ0 I
B= = nμ0 I (10.230)
l
where n is the number of turns per unit length of solenoid. The magnetic flux through
each turn is then calculated as follows:

ΦB = B · A (10.231)

where A is the cross-sectional area of the solenoid, given as A = πr 2 . Therefore,

Φ B = π Br 2 = πnμ0 I r 2 (10.232)

Then, the inductance is

10.9 Exercises 277

ΦB N N N 2r 2
L=N = π μ0 I r 2 = μ0 π (10.233)
I I l l
Thus, inductance L depends on the N and on the geometry of the solenoid, namely,
l and r . Furthermore, we can write Eq. (10.46) as follows:

L = μ0 (lπr 2 )n 2 = μ0 n 2 V (10.234)

where V = lπr 2 is the volume of the solenoid.

Exercise 10.7 Consider an air-core solenoid consisting of N = 300 turns of

length l = 25.0 cm, and its cross-sectional area is A = 4.00 cm2 . (a) Calculate
the inductance of solenoid. (b) Calculate self-induced emf in the solenoid if
the current through it is decreasing at a rate of 50.0 A/s.

Solution 10.7 (a) Using the derived formula for the inductance of a solenoid given
by Eq. (10.233), we have
N2 A
L = μ0 (10.235)
l
Replacing the numerical values, we get

 (300)2 · (4.00 × 10−4 m2 )

L = 4π × 10−7 T m/A (10.236)
25.0 × 10−2 m
−4
= 1.81 × 10 Tm /A = 0.181 mH
2

(b) The self-induced emf is

dI
 L = −L (10.237)
 dt
= − 1.81 × 10−4 Tm2 /A · (50.0 A/s)
= 9.05 × 10−3 V = 9.05 mV

Exercise 10.8 Consider the DC circuit consisting of RL series, shown in

Fig. 10.30. The switch is closed at t = 0. (a) Calculate the time constant τ
of the circuit. (b) Calculate the current in the circuit at t = 2.00 ms. c) Com-
pare the potential difference across the resistor with that across the inductor.
278 10 More About Faraday’s Law of Induction

Fig. 10.30 A DC circuit of

series of RL

Solution 10.8 (a) The time-varying current is

 
I (t) = 1 − e−t/τ . (10.238)
R
The time constant τ is given as

L 30.0 × 10−3 H
τ= = = 0.005 s = 5.00 ms . (10.239)
R 6.00 
(b) The current I at t = 2.00 ms is

12.0 V 
I (2.00 ms) = 1 − e−(2.00 ms)/(5.00 ms) = 0.659 A . (10.240)
6.00 
(c) For the potential difference across the inductor, we use the expression given by
Eq. (10.80):
 L (t) = −e−t/τ . (10.241)

The voltage drop across the resistor is


ΔVR = I (t)R =  1 − e−t/τ . (10.242)

Then, the difference between them is

ΔVR −  L =  , (10.243)

which is a constant equal to the source emf.

10.9 Exercises 279

Fig. 10.31 A DC circuit of

the RL series

Exercise 10.9 Consider a DC circuit of RL series, as shown in Fig. 10.31.

When the switch S2 is closed at the instance S1 is open (t = 0). The current in
the upper loop varies as
I (t) = I0 e−t/τ (10.244)

where I0 = /R is the initial current and τ = L/R is the time constant. Show
that the energy initially stored in the magnetic field of the inductor appears as
internal energy in the resistor as the current decays to zero.

Solution 10.9 The rate of energy change dU/dt at which the energy is delivered to
the resistor is E R = I 2 R, where I is the instantaneous current:

dU
= I 2 R = I02 Re−2t/τ (10.245)
dt
Thus, we can write
dU = I02 Re−2t/τ dt (10.246)

Integrating both sides of Eq. (10.246), we obtain

 ∞
dU = I02 R e−2t/τ dt (10.247)
0

Therefore, we find that

∞  
τ −2t/τ 2t
U= −I02 R e d − (10.248)
2 τ
0
280 10 More About Faraday’s Law of Induction

I02 Rτ  −2·∞/τ
=− e − e0
2
I2L
= 0
2

Exercise 10.10 Consider the CD circuit in Fig. 10.32. The capacitor is initially
charged when the switch S1 is open and S2 is closed. The switch S1 is then
closed and at the same instant S2 is opened, such that the capacitor is connected
to the inductor. (a) Calculate the frequency of oscillation in the circuit. (b) What
is the maximum Q max and Imax ? (c) Determine the current and charge as the
function of time t.

Solution 10.10 (a) The linear frequency is given as

ω 1
f = = √ (10.249)
2π 2π LC
1
= 
2π (2.80 × 10 H)(9.00 × 10−12 F)
−3

= 1.00 × 106 Hz .

(b) The initial charge in capacitor equals the maximum charge on the capacitor:

Q max
C= . (10.250)

Alternatively,

Q max = C = (12.0 V)(9.00 × 10−12 F) = 1.08 × 10−10 C (10.251)

The maximum current is then

Imax = Q max ω (10.252)

−10
= 2π(1.08 × 10 C)(1.00 × 10 Hz)
6

−4
= 6.79 × 10 A.

(c) The variation of charge Q is

Q = Q max cos (ωt) (10.253)

−10

= (1.08 × 10 C) cos 2π × 106 (rad/s) · t .
10.9 Exercises 281

Fig. 10.32 A DC circuit of

the LC

The current is varying according to

I = −Imax sin (ωt) (10.254)

−4

= −(6.79 × 10 A) sin 2π × 106 (rad/s) · t .

Reference

Holliday D, Resnick R, Walker J (2011) Fundamentals of physics. John Wiley and Sons
Chapter 11
Some Applications of Electromagnetic
Theory

This chapter aims to introduce two applications of

electromagnetic theory to electrostatic properties of
macromolecular solutions and wireless charging, important in
bio-nanotechnology and wireless technology development.

In this chapter, we discuss the applications of electromagnetic theory to electrostatic

properties of macromolecular solutions and wireless charging. For further reading,
one can also consider other available literature (Kamberaj 2020; Deuflhard et al.
1997) [Philips et al. (2013)].

11.1 Electrostatic Properties of Macromolecular Solutions

11.1.1 The pH and Equilibrium Constant

The dissociation constant of water molecules manifests the competition between the
energy of binding and the entropy of charge liberation. It requires considering the
interchange of energetic (E) and entropic (S) effects in controlling the separation to
examine pH from a quantitative viewpoint. The competition between energy mini-
mization (bound water molecule state) and entropy maximum (ionic dissociation of
water) sets the stage for many biological reactions.
The reaction for dissociation of a water molecule is

H2 O  H+ + OH− . (11.1)

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 283
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_11
284 11 Some Applications of Electromagnetic Theory

The problem is to find the fraction of water molecules that are in dissociated state in
a sample of water. Equilibrium constant is gives as
N   
 νi

N
K eq = ci0 exp −β μi0 νi . (11.2)
i i

In Eq. (11.2), β = 1/k B T (where k B is Boltzmann’s constant and T is the tempera-

ture), μi0 is the standard chemical potential, and
 
ci
μi = μi0 + k B T ln , (11.3)
ci0

where ci0 is the standard state concentration of the i component. νi is the stoichiomet-
ric coefficient that equals the change in the number of particles of the i-th component
during reaction:
A+B  AB . (11.4)

In Eq. (11.4), νA = νB = −1 and νAB = 1. On the other hand, for reaction given by
Eq. (11.1), νH+ = νOH− = +1 and νH2 O = −1.
The dissociation constant is defined as
1
Kd = . (11.5)
K eq

The law of mass action implies that


N
K eq = ciνi , (11.6)
i

which is known as the law of mass action.

Therefore, for the reaction given by Eq. (11.1), we write
cH+ · cOH−
Kd = (11.7)
cH2 O
cH+ ,0 · cOH− ,0  
= exp −β μH+ ,0 + μOH− ,0 − μH2 O,0 .
cH2 O,0

In Eq. (11.7), the left-hand side is the ratio of the concentrations of the products to
reactions at a given temperature and the right-hand side is related with the dissociation
constant. Furthermore, μ X,0 denotes the standard state chemical potential. Moreover,
note that pH of solution is given as

pH = − log10 (cH+ ) . (11.8)

11.1 Electrostatic Properties of Macromolecular Solutions 285

ci,0 represents the concentration of the species of the type i in some standard state.
Assuming that H+ is due to only dissociation, then cH+ = cOH− . Furthermore, cH2 O =
cH2 O,0 and cH+ ,0 = cOH− ,0 = 1 M. Therefore, we obtain

cH2 + 1M · 1M  
= exp −β μH+ ,0 + μOH− ,0 − μH2 O,0 . (11.9)
cH2 O,0 cH2 O,0

Using the fact, the change in energy of reaction is

kJ
μH+ ,0 + μOH− ,0 − μH2 O,0 = 79.9 . (11.10)
mol
We obtain that
 
kJ
cH2 + = exp −β79.9 = 1.0 × 10−14 M2 . (11.11)
mol

Or,
cH+ = 1.0 × 10−7 M . (11.12)

That is, using Eq. (11.8), we obtain

pH = 7 . (11.13)

This is a result for water under standard conditions (that is, T = 300 K).

11.1.2 Charge on DNA and Proteins

The charge state of a macromolecule depends on pH of the solution. On the other

hand, the charge state of the macromolecule is important in determining both their
structure and function in solution. Here, we discuss how the charge state is tuned, by
considering the following reaction:

HM  H+ + M− . (11.14)

Here, M is the macromolecule of interest. From the equation of the law of mass
action, the dissociation constant becomes:
cH+ · cM−
Kd = . (11.15)
cHM

We introduce pK to measure the tendency of the macromolecule to undergo the

dissociation reaction as
pK = − log10 K d . (11.16)
286 11 Some Applications of Electromagnetic Theory

Or,

+ log10 K d = log10 cH+ + log10 cM− − log10 cHM (11.17)

= −pH + log10 cM− − log10 cHM .

Combining Eqs. (11.16) and (11.17), we obtain

 
cM−
pH = pK + log10 , (11.18)
cHM

which is known as Hendersen-Hasselbalch equation.

If cM− = cHM , then pH = pK. That corresponds to the pH at which half of HM
molecules are dissociated. Therefore, pK equals pH at which half of macromolecules
have been dissociated.
Consider a DNA, which has a pK such that pK ≈ 1. Then,
 
cDNA−
pH = 1 + log10 . (11.19)
cHDNA

Using Eq. (11.19), at normal pH (that is, pH = 7), phosphates on the DNA backbones
are fully dissociated. That is, two electronic charges for every base pair, or a linear
charge density:
2e
λ= . (11.20)
0.34 nm

11.1.3 Charge States of Amino Acids

An important aspect of charge state of proteins is that different side chains of amino
acids have different dissociation tendency. That results in that at different pH values,
different side chains will be in different states of dissociation. Usually, the titration
curve is used to characterize the charge of a macromolecule at different pH values.
In the following, we introduce a theoretical framework for prediction of the titration
curve.1 If we assume that the concentrations are proportional to probabilities, then
Eq. (11.18) takes the following form:
  
ΔG ΔG
pH = pK + log10 exp − = pK + γ , (11.21)
k BT 2.303k B T

where ΔG equals the free energy of the ionized state of a protein relative to neutral
state and ln 10 ≈ 2.303. In Eq. (11.21), γ = +1 for bases and γ = −1 for acids. For
a protein with n ionizable groups:

1 Deuflhard et al. (1997).

11.1 Electrostatic Properties of Macromolecular Solutions 287


n
ΔG = ΔG i , (11.22)
i=1

where ΔG i is expressed from Eq. (11.21), and it gives the contribution of each
ionizable group: 
ΔG i = 2.303γi k B T pH − pKi . (11.23)

Here, an ionizable group can be in an isolated amino acid in solvent, and we have a
model compound environment with pKi,model , or in one of the protein’s amino acids
in solvent, and we have a protein environment with pKi,protein . We write

ΔG i,protein
0
pKi,protein = , (11.24)
2.303k B T
ΔG i,model
0
pKi,model = ,
2.303k B T

where ΔG i,X
0
is the standard free energy change for a state with dissociated molecule
and proton in solution relative to the state with proton bound to a molecule (such as
the protein).
For a protein with n ionizable groups, Eq. (11.22), we have


n

ΔG (x1 , x2 , . . . , xn ) = 2.303k B T xi γi pH − pKi,protein , (11.25)
i=1

where xi = 0, if the group i is neutral; otherwise, xi = 1, if the group is ionized.

Introducing the standard free energy difference as

ΔG i,m→p
0
= ΔG i,protein
0
− ΔG i,model
0
(11.26)

= 2.303k B T pKi,protein − pKi,model .

Alternatively,
ΔG i,m→p
0
pKi,protein = + pKi,model . (11.27)
2.303k B T

Then, substituting the expression given by Eq. (11.27) into (11.25), we have:


n

ΔG (x1 , x2 , . . . , xn ) = 2.303k B T xi γi pH − pKi,model (11.28)
i=1

ΔG i,m→p
0
− .
2.303k B T
288 11 Some Applications of Electromagnetic Theory

In general, ΔG i,m→p
0
has an electrostatic nature, and hence only the electrostatic
contribution to ΔG i,m→p is practically calculated. Therefore, Eq. (11.28) can also be
0

written as


n

ΔG (x1 , x2 , . . . , xn ) = 2.303k B T xi γi pH − pKi,intrinsic (11.29)
i=1


n−1 
n
+ xi x j Ψi j ,
i=1 j=i+1

where

ΔG i,m→p
0
pKi,intrinsic = pKi,model + (11.30)
2.303k B T
ΔΔG i(elec)
= pKi,model −
2.303k B T
= pKi,model + pKi,shift ,

and Ψi j is the interaction energy between the ionized groups. Furthermore, according
to this approach,
ΔΔG i(elec)
pKi,shift = − , (11.31)
2.303k B T

and
ΔΔG i(elec) = ΔG (elec) (elec)
i,protein − ΔG i,model , (11.32)

In Eq. (11.32), the first and second terms in the right-hand side are given as

ΔG (elec)
i,protein = Ψi,protein − Ψi,protein ,
ionized neutral
(11.33)
ΔG (elec)
i,model = Ψi,model
ionized
− Ψi,model
neutral
,

where Ψi,X
ionized
and Ψi,X
neutral
give the electrostatic solvation free energies of the ionized
and neutral states, respectively, for protein and model compound environments of
the group i. Those terms are calculated using Poisson-Boltzmann approach (which
is introduced in the following):

1
N
Ψ = qi φi , (11.34)
2 i=1

where φi is the electrostatic potential of the protein assembly of partial charges at

the position of charge qi .
11.1 Electrostatic Properties of Macromolecular Solutions 289

Fig. 11.1 (Middle) Titration Curve for the Entire Spike Protein COVID-19 bound to Antibodies.
(Left) Three-dimensional structure colored according to segment name; (Right) Three-dimensional
structure colored according to partial charge associate to each atom. Structure was prepared using
CHARMM36 force field and titration curve was produced using H+ web server

Figure 11.1 shows the Titration Curve for the Entire Spike Protein COVID-19-
Antibodies complex (middle). Three-dimensional structure is colored according to
segment name (left) and colored according to partial charge associate to each atom
(right). Structure was prepared using CHARMM36 force field2 and titration curve
was produced using H+ web server.3

11.1.4 Salt Binding

In cells, many aspects of biological molecules involve recognition of one molecule by

another. For instance, binding of RNA polymerase to DNA promotor sites, receptors
binding to their ligands, protein-membrane binding, and protein-protein binding.
The surface charge distribution at these macromolecules plays an important role
in determining specificity of these interactions and their strength. The importance of
electrostatics in macromolecular binding is observed in vitro experiments because
such interactions are almost strongly dependent on the concentration of ions in the
solution in which they are measured. For example, protein-DNA binding interactions
is affected by salt. In particular, regulatory proteins transcription factors binding to
DNA is modulated by salt concentration.
One can explain that the ions in salty solutions assemble in a screening cloud
around macromolecules that have the potential to bind. When the reaction binding
occurs, then ions of the screening clouds release, and hence that results in an increase
of entropy. Secondly, increasing ion concentration, the ions can interact with receptor
and so they compete with ligand.

2 MacKerell et al. (1998).

3 Anandakrishman et al. (2012).
290 11 Some Applications of Electromagnetic Theory

Fig. 11.2 Interacting

charges q1 and q2 in a
medium with dielectric
constant ε. The size of
charges d1 and d2 are such
that d1 , d2  r , where r is
their separation; therefore,
the charges are point-like
charges

The equilibrium or dissociation constants can lead toward understanding of charge

state of macromolecules in solutions. We need to discuss the electrostatics to under-
stand the macroscopic level. The charges in solution move around. The extent
to which charges leave their macromolecular hosts is dictated by the competition
between energy and entropy. The energy of a positive charge and negative charge
favors the reduction of separation, that is, the binding of the charges with opposite
sign of the charges. On the other hand, the entropic contribution favors the charges
being separated and moving freely in a solution, and hence it drives the system to
the charge liberation.
In Fig. 11.2 are shown two interacting charges q1 and q2 in a medium with dielec-
tric constant ε. The size of charges d1 and d2 are such that d1 , d2  r , where r is
their separation; therefore, the charges are point-like charges. The magnitude of force
between the charges is given as

1 q1 q2
F= . (11.35)
4πε0 ε r 2

In Eq. (11.35), the dielectric constant of the medium for air and nonpolar medium
is ε ≈ 1, and for water (a polar medium) is ε ≈ 80. Therefore, the dielectric constant
of the medium is larger for polar medium. From the microscopic point of view, it
represents the screening of the electrostatic interactions between the charges due
to permanent dipoles of polar molecules. Figure 11.3 illustrates the formation of
permanent dipole in the water molecule due to the distribution of the electronic
charges in the water molecule (which is a polar molecule).

Fig. 11.3 Illustration of the

permanent dipole of a water
molecule
11.1 Electrostatic Properties of Macromolecular Solutions 291

Fig. 11.4 Charge-charge

interaction screening by
permanent dipoles of the
polar molecules of the
medium

Figure 11.4 represents the charge-charge screening because of the reorientation

of permanent dipoles of polar molecules of the medium. This is equivalent to saying
that the electric field E of the charge q at a point P at the distance r from the charge is
reduced due to the screening by the permanent dipoles of the medium polar molecules
by ε:
1 q r̂
E(r ) = , (11.36)
4πε0 ε r 2

where r̂ is a unit vector pointing to the point P.

Using the first Maxwell’s equation given by Eq. (4.79), where D is given by
Eq. (4.63), we can write
ε0 ∇ (ε(r)E(r)) = ρ(r) , (11.37)

where, in general, ε can also be a function of r, which is the general case of the non-
homogeneous medium, and the polarized charges created at the dielectric boundaries
must be taken into account as well. In Eq. (11.37), ρ is the external charge density; for
example, a macromolecule is immersed in a solvent medium. Using the relationship
E = −∇φ(r), where φ(r) is the scalar potential function, then Eq. (11.37) becomes

ρ(r)
∇ · (ε(r)∇φ(r)) = − . (11.38)
ε0

In Eq. (11.38), ρ is the sum of the distribution of the macromolecule fixed charge
density ρm (r) and ionic charge density ρ I (r):

ρ(r) = ρm (r) + ρ I (r) . (11.39)

Equation (11.38) is known as Poisson equation. A rigorous solution of that equation

is provided using the Poisson-Boltzmann’s approach as discussed in the following.
292 11 Some Applications of Electromagnetic Theory

11.1.5 Energy Cost of Assembling a Collection of Charges

Cells spend a great amount of energy in moving charges around. Understanding the
energetics associated with charge management, we discuss energy associated with
a charge distribution. For that, we calculate the work done to bring isolated charges
from infinity, where they do not interact, to a form, such as their distribution in a
macromolecular configuration, where they interact with each other. If the electrical
energy is positive, then an external agent is doing work to bring charges together;
otherwise, it is the system of charges that is doing work.
Consider the potential electrical energy of the spherical ball of radius R and charge
Q uniformly distributed, as shown in Fig. 11.5, immersed in a medium with dielectric
constant ε. For that, we calculate the work done to assemble the ball. First, we divide
the sphere into spherical shells of thickness dr and charge dq, and then calculate the
work done to bring together these spherical shells.
The potential electrical energy to bring a shell of thickness dr and charge dq from
infinity to a sphere of radius r and charge q uniformly distributed (see also Fig. 11.5)
is given as
dUelec = φ(r )dq , (11.40)

where the scalar electric potential is given as

∞
q
φ(r ) = dr . (11.41)
4πε0 εr 2
r

After integration in Eq. (11.41), we obtain

q
φ(r ) = . (11.42)
4πε0 εr

Fig. 11.5 A sphere of radius

R containing a charge Q.
The inner part of the sphere
of radius r (r < R)
containing the charge q. The
spherical ring of thickness
dr with charge dq
11.1 Electrostatic Properties of Macromolecular Solutions 293

Assuming a uniform distribution of the charge in the sphere of radius R, we write

for the charge density of the spherical volume as

3Q
ρ= . (11.43)
4π R 3
Then, the total charge q in a sphere of radius r is

3Q 4π 3  r 3
q= r = Q . (11.44)
4π R 3 3 R
Substituting Eq. (11.44) into (11.42), we obtain an expression for the scalar electric
potential at the surface of a sphere of radius r and charge q:

Qr 2
φ(r ) = . (11.45)
4πε0 εR 3

Differentiating Eq. (11.44), we get

r2
dq = 3Q dr . (11.46)
R3
Substituting Eqs. (11.46) and (11.45) into (11.40), we find the work done to bring a
spherical shell of charge dq from infinity at the surface of a sphere of radius r and
charge q:
Qr 2 r2 3Q 2 r 4
dUelec = 3Q 3 dr = dr . (11.47)
4πε0 εR 3 R 4πε0 ε R 6

Integrating Eq. (11.47) for r from zero to R, we calculate the work done to assemble
a charge Q uniformly distributed in a sphere of radius R as

3Q 2 1
Uelec = . (11.48)
20πε0 ε R

Equation (11.48) indicates that Uelec is positive, which is expected since the assem-
bling charges with the same sign requires external work.

Example 11.1 As an example, we consider the DNA condensation in Bac-

teriophage φ29. DNA inside a virus is contained in the volume provided by
the viral capsid. Since DNA is a strong acid, it is highly charged in solution.
Therefore, we need to add energy to bring all charges carried by the DNA into
close proximity of the DNA configuration inside capsid.
294 11 Some Applications of Electromagnetic Theory

Fig. 11.6 Force (in pN)

versus DNA’s length (in nm)

For bacteriophage φ29, this energy is provided in the form of mechanical

work done by portal motor (which is a protein machine that translates DNA),
which in turn is fueled by ATP hydrolysis.
The measurements indicate that there is an internal force build-up as DNA
is packing. The work against this internal is schematically shown in Fig. 11.6.
From here, the internal work done is approximated as

1
Wint ≈ (700 nm) · (60 pN) = 210000 (pN · nm) . (11.49)
2
The total charge carried by DNA of φ29 genome of about 20000 base pairs
is calculated using Eq. (11.20) as

2e
Q= · 20000 bp = 40000e , (11.50)
bp

where e = 1.6 × 10−19 C is the magnitude of the charge of one electron.

Assuming that this charge is distributed uniformly inside a capsid, consid-
ered as spherical with a radius R ≈ 20 nm. The work done to condense this
DNA charge inside the capsid sphere is

3Q 2 1
Uelec = ≈ 108 (pN · nm) . (11.51)
20πε0 ε R

Comparing Eqs. (11.49) and (11.51), we find that Uelec  Wint . This result
indicates that the potential energy value calculated in Eq. (11.51) is not cor-
rect when compared with experimental results because it does not include the
screening of the forces between charges of DNA by the presence of the counter
ions in the medium. Therefore, for a better treatment of the forces between the
charges of the macromolecule, we introduce the Poisson-Boltzmann’s model
in the following.
11.1 Electrostatic Properties of Macromolecular Solutions 295

11.1.6 The Poisson-Boltzmann Equation

Here, we introduce the Poisson-Boltzmann equation as described elsewhere.4 In the

Poisson-Boltzmann approach, all the macromolecular atoms are considered explic-
itly as particles with partial point charges at the atomic positions, and the dielectric
constant of the macromolecule is ε p (it is often considered to be low, typically, ε p is in
the range of 2–4). The solvent environment surrounding the macromolecule is taken
implicitly into account as a dielectric medium with the dielectric constant of εw (typ-
ically, about 80). The macromolecular dielectric value does not take into account the
rearrangement of polar and charged amino acids with external electric fields, which
could result into a larger dielectric constants. For example, it is suggested that the
increase of the dielectric can compensate for the need for group re-orientations.
In non-homogeneous interacting particles system, density of a particle at any point
r can be written as
σ I,i (r) = gi (r)σ 0I,i (r), (11.52)

where σ 0I,i (r) is the particle density of the same system considered as ideal gas (i.e.,
non-interacting particles system), and gi (r) is the i-th particle distribution, which is
taken to follow the Boltzmann distribution

gi (r) = exp (−βWi (r)) . (11.53)

In Eq. (11.53), Wi (r) is the potential of mean force for the particle i, which is equal
to electric potential energy; that is, the average electrostatic potential at the charge’s
position, φ(r), multiplied by the charge of particle, qi :

Wi (r) = qi φ(r)

where qi = z i e with z i being its valency and e being the charge of proton.
Thus, Eq. (11.52) can be written as

σ I,i (r) = σ 0I,i (r) exp (−βqi φ(r)) . (11.54)

Then, the charge density is given as

 
ρ I (r) = qi σ I,i (r) = qi σ 0I,i (r) exp (−βqi φ(r)) , (11.55)
i i

where
σ 0I,i (r) = ci∞ λ(r),

4 Kamberaj (2020).
296 11 Some Applications of Electromagnetic Theory

where ci∞ is the bulk constant concentration of the i-th ionic species, satisfying the
condition of the electrostatic neutrality:

qi ci∞ = 0
i

λ(r) is the accessibility of ions at point r (i.e., λ(r) = 0 in the region inside the
macromolecule and λ(r) = 1 in the solvent region). Therefore, we can write

ρ I (r) = λ(r) qi ci∞ exp (−βqi φ(r)) . (11.56)
i

Using Eq. (11.56), the Poisson equation (see Eq. (11.38)) takes the form of the
so-called nonlinear Poisson-Boltzmann equation

λ(r)  ∞ ρm (r)
∇ · (ε(r)∇φ(r)) + qi ci exp (−βqi φ(r)) = − . (11.57)
ε0 i ε0

For an electrostatic neutral solvent, we can write


N+

N−
qi(+) ci+,∞ = qi(−) ci−,∞
i=1 i=1

where two kind of ionic species are assumed to exist in the solution, positive and
negative with N+ and N− being the number of positive and negative ions, respectively.
Assuming that N+ = N− = N I , and since qi(+) = −qi(−) ≡ qi and ci+,∞ = ci−,∞ =
ci∞ /2, we get from Eq. (11.57) that

λ(r) 
N
I
ρm (r)
∇ · (ε(r)∇φ(r)) − qi ci∞ sinh (βqi φ(r)) = − , (11.58)
ε0 i=1 ε0

which is a form often found in the literature and it represents a nonlinear partial differ-

ential equation. In Eq. (11.58), sinh represents the function: sinh(x) = e x − e−x /2.
Assuming that the potential is small, the linear form of the equation can be obtained
as 
ρm (r) β  2 ∞
∇ · (ε(r)∇φ(r)) = − +ε q c λ(r)φ(r). (11.59)
ε0 εε0 i i i

We can determine the so-called Debye screening constant κ as

β  2 ∞ β 1
κ2 = qi ci = I ≡ 2, (11.60)
εε0 i εε0 lD
11.1 Electrostatic Properties of Macromolecular Solutions 297

which also describes the exponential decay of the potential in the solvent, with l D
being the Debye length, and I 
I = qi2 ci∞
i

being the ionic strength. Note that κ = 0 in the macromolecule region because the
mobile ions are present only in the solvent region.
Equation (11.59) can then be written as

ρm (r)
∇ · (ε(r)∇φ(r)) − ε(r)κ2 λ(r)φ(r) = − . (11.61)
ε0

Although for biological systems φ is not small, and therefore the linearization
condition does not hold, comparisons between the linear and nonlinear forms of the
Poisson-Boltzmann equation show that both forms are in good agreement with each
other. Moreover, these comparisons have shown that small differences are related to
the charge density, and hence to the electric field magnitude, at the interface solvent-
solute.
From solving either the linear Poisson-Boltzmann equation (see Eq. (11.61)) or the
nonlinear Poisson-Boltzmann equation (see Eq. (11.57)), the electrostatic potential,
φ(r), will be obtained at any point r in space. It can be seen, that knowing φ, we may
calculate the local concentration of ions through the formula

ci (r) = ci∞ exp (−βqi φ(r)) , (11.62)

which involves the Boltzmann distribution. Moreover, the gradient of the electrostatic
potential can give the electric field, E(r) = −∇φ(r).
Another quantity of interest calculated using the electrostatic potential is the elec-
trostatic component of the solvation free energy. The electrostatic term of solvation
free energy gives the work done for a possible process of charging the macromolecule
and ions in an ionic discharged atmosphere. Using these processes in thermodynamic
cycles, we can compute the electrostatic component of free energies for real processes
such as solvation. The free energy for charging the solute (e.g., a macromolecule)
in an ionic environment can be calculated using different approaches, for example,
by direct integration of the charge, by considering a variation principle, or using
thermodynamic arguments.
Based on Marcus theory, the electrostatic energy G solv elec contains three different
terms. The first term is the classical electrostatic energy, G cl
elec

1
elec =
G cl d 3 rρm (r)φ(r). (11.63)
2

The second term is arising from mixing the mobile species G mob
elec :
298 11 Some Applications of Electromagnetic Theory

 ci (r)
elec = k B T
G mob d 3r ci (r) ln . (11.64)
i
ci∞

Combining Eqs. (11.62) and (11.64), we obtain

 

G mob
elec =− d r 3
ci (r)qi φ(r). (11.65)
i

The third term is the so-called osmotic term, which due to nonuniform ionic concen-
tration, and it is calculated as a volume integral:


G solvent
elec = kB T d 3r ci∞ − ci (r) (11.66)
i

  
= kB T d 3r ci (r) exp (βqi φ(r)) − 1
i

 

= 3
d r ci (r)qi φ(r),
i

where the linearity of the exponential term is applied for φ small.

Therefore, the total electrostatic energy is

G elec = G cl
elec + G elec + G elec
mob solvent

or

1
G elec = G cl
elec = d 3 rρm (r)φ(r). (11.67)
2

where ρm is the charge density of fixed charges (i.e., nonionic charges such as partial
atomic charges of macromolecule). It is possible to also calculate the free energy by
knowing the electrostatic potential, for example, for the possible charging process
of a macromolecule in an ionic environment. That is done by combining different
possible processes in a thermodynamic cycle that can lead to the computation of the
theoretical free energy of some real process.
For instance, the thermodynamic cycle shown in Fig. 11.7 can be used to calculate
the electrostatic component of the solvation free energy. This thermodynamic cycle
indicates that the electrostatic component of the solvation free energy is the difference
in the free energies related to two purely hypothetical charging processes; one in
some reference surrounding environment phase (e.g., with dielectric constant equal
to that of the macromolecule) and the other the solvent surrounding environment
with dielectric constant εw . Thus, using Eq. (11.66), we can write
11.1 Electrostatic Properties of Macromolecular Solutions 299

Fig. 11.7 Thermodynamic cycle for calculation of electrostatic solvation free energy. The differ-
ence in the charging energy, ΔG elec in reference surrounding phase and in solvent is the electrostatic
solvation free energy. Colors in bottom plots indicate the partial atom charges. In the top plots, the
blue color indicates no partial charges on atoms

ΔG solv
elec = G elec (solvent) − G elec (ref) (11.68)

1
= d 3 rρm (r) [φw (r) − φr (r)] .
2

Often, φw (r) − φr (r) is called reaction potential, φreac (r), and thus Eq. (11.68) can
be written as

1
ΔG solv
elec = d 3 rρm (r)φreac (r). (11.69)
2

In Eq. (11.69), ΔG solv

elec represents the work done by electrostatic forces for trans-
ferring a set of partial atomic charges of macromolecule from a fixed point in some
reference surrounding environment with dielectric constant εr to a fixed point in
solvent surrounding environment with dielectric constant εw .
Depending on the shape and charge distribution of macromolecule, numerical
solutions of the Poisson-Boltzmann equation could be difficult. Solvated macro-
molecular systems are in general modeled by regions with different dielectric con-
stants. Figure 11.8 illustrates a solvated macromolecule occupying the region Ω,
where the macromolecule region is represented by Ωm as a solid surface; the solvent
region is represented by Ωw . The dielectric interface σ is defined by the molecular
surface and represents the region not penetrated by mobile ions, and n will represent
an unit vector normal to σ pointing from Ωm to Ωw . The transition from solute (with
300 11 Some Applications of Electromagnetic Theory

Fig. 11.8 Different

computational regions of
interest: The solute
(macromolecule) region,
Ωm , with dielectric constant
εm ; solvent region, Ωw , with
dielectric constant εw where
different mobile ions are
denoted; dielectric interface,
which characterize the
solvent accessible surface
are constructed with a probe
radius of r = 1.4 Å (surface
colored in blue)

low-dielectric constant) to solvent (with high-dielectric constant) is modeled to be

abrupt, giving rise to the dielectric interface σ. There are two conditions on σ that
are usually satisfied:

(φ(r))Ωm = (φ(r))Ωw (11.70)

   
∂φ(r) ∂φ(r)
ε = ε .
∂n Ωm ∂n Ωw

These conditions are used in the methods based on boundary integral equations, but
may not apply to finite difference methods. Usually, the boundary of the entire com-
putational domain is also defined, Γ . In addition, approximated Dirichlet boundary
condition is imposed in the boundary Γ .
The widely used numerical methods include finite difference method (FDM), the
boundary element method (BEM), and finite element method (FEM).
Figure 11.9 shows solvation free energies, ΔG solvelec , decomposed for each atom
(in kcal/mol) for Bovine Pancreatic Trypsin Inhibitor (BPTI) protein of 58 amino
acids.5 The protein was prepared using CHARMM36 force field. Inset is the three-
dimensional structure colored according to solvation free energy values. In the numer-
ical calculations, the dielectric constant of protein was ε p = 2 and for the water
εw = 80. The ionic concentration in the solvent was 0.154 M/L.

5 Czapinska et al. (1999).

11.1 Electrostatic Properties of Macromolecular Solutions 301

Fig. 11.9 Solvation free energy (in kcal/mol), ΔG solv

elec , for Bovine Pancreatic Trypsin Inhibitor
(BPTI) protein shown for every atom. Inset is the three-dimensional structure colored according
to solvation free energy values. The dielectric constant of protein was ε p = 2 and for the water
εw = 80. The ionic concentration in the solvent was 0.154 M/L

11.1.7 Calculation of pKa of Amino Acids in Macromolecules

Using Eq. (11.18), and again assuming that the probabilities are proportional to
concentrations, we write  
cM− −ΔG
= exp . (11.71)
cHM kB T

In Eq. (11.71), ΔG is the free energy of the ionized state relative to neutral state of
a macromolecule. Therefore, Eq. (11.18) can also be written as
  
−ΔG
pKa = pH − log10 exp . (11.72)
kB T

Or,
ΔG = 2.303k B T γ (pH − pKa) . (11.73)

In general, the group can be part of an isolated amino acid in solution (called model
compound environment, pKamodel ), or part of an amino acid in a macromolecule that
is solvated (called protein environment model, pKaprotein ). Solving Eq. (11.73) for
pKa, we obtain
ΔG
pKa = pH − γ , (11.74)
2.303k B T

because 1/γ = γ.
We can consider an amino acid of a protein without N-terminal and C-terminal
patches in an isolated state in solution to calculate pKamodel in a neutral state, then
302 11 Some Applications of Electromagnetic Theory

ΔG 0model
pKamodel = , (11.75)
2.303k B T

where ΔG 0model is the free energy change for the state with dissociated amino acid
and proton in solution relative to the state with proton bound to amino acid. The
experimental pKa for the compound model can be derived from molecules that are
close to the compound model (such as organic molecules).
For the amino acid in a protein environment, we have

ΔG 0protein
pKaprotein = , (11.76)
2.303k B T

where ΔG 0protein is the free energy change for the state with dissociated protein and
proton in solution relative to the state with proton bound to protein.
Then, ΔG of a protein assuming only one ionizable group is

ΔG = 2.303γk B T pH − pKaprotein . (11.77)

Furthermore, the standard free energy difference is

ΔΔG 0m→p = ΔG 0protein − ΔG 0model . (11.78)

Therefore, Eq. (11.77) can also be written as

 
ΔG 0 m→p
ΔG = 2.303γk B T pH − pKamodel − . (11.79)
2.303k B T

Therefore,
ΔG 0m→p
pKacalc = pKamodel + . (11.80)
2.303k B T

It can be seen (see also Eq. (11.80)) that by knowing pKamodel and ΔG 0m→p , we can
calculate pKacalc . The second term in Eq. (11.80) represents the so-called pKashift :

ΔG 0m→p
pKashift = . (11.81)
2.303k B T

Therefore,
pKacalc = pKamodel + pKashift . (11.82)

Note that ΔG 0m→p has an electrostatic origin, and hence only the electrostatic con-
tribution is considered, using Poisson-Boltzmann approach. Based on this approach,

ΔΔG (elec)
pKashift = − , (11.83)
2.303k B T
11.1 Electrostatic Properties of Macromolecular Solutions 303

Table 11.1 The values of calculated pKamodel for some amino acids using CHARMM36 force
field. In addition, the experimental pKa are shown, along with their standard deviations
Amino acid pKamodel pKashift pKacalc pKaexp
GLU 4.07 −2.63 1.44 2.1 ± 0.1
ASP 3.86 −0.45 3.41 3.1 ± 0.1
HIS 6.10 2.10 8.20 7.75 ± 0.02
CYS 8.23 −1.85 6.38 10.2 ± 0.2
LYS 10.53 −0.011 10.52 10.6 ± 0.1

where
ΔΔG (elec) = ΔG (elec) (elec)
protein − ΔG model . (11.84)

In Eq. (11.84), the first and second terms in the right-hand side are given as

ΔG (elec)
protein = Ψprotein − Ψprotein ,
ionized neutral
(11.85)
ΔG (elec)
model = Ψmodel
ionized
− Ψmodel
neutral
.

In Eq. (11.85), Ψ Xionized and Ψ Xneutral give the electrostatic solvation free energies
of the ionized and neutral states, respectively, for protein and model compound
environments. Those terms are calculated using Poisson-Boltzmann approach, given
by Eq. (11.34).
In Table 11.1, we show the values of calculated pKamodel for some amino acids
using CHARMM36 force field. Besides, we present the predicted values pKacalc for
different amino acids in Bernase protein6 environment, calculated using CHARMM
program.7 The experimental values of pKa are also shown, including the statisti-
cal error. In the numerical calculations using Poisson-Boltzmann approach, ε p = 2,
εw = 80, and the solvent probe radius is r p = 1.4 Å. The ionic concentration of
solvent is 0.154 M/L, and the temperature is 300 K.

11.2 Wireless Charging

There exists an increasing interest on inductive technologies. Furthermore, there are

products provided in the markets from companies, such as Powermal and Wireless
Power Consortium. These technologies are also called tightly coupled technologies.
Some of those technologies offer more specific capabilities in the network, such as
supervising and measuring the power to each device. In fact, these features may pro-
vide some attractive services to providers of charging services (such as airports, Star-

6 Martin et al. (1999).

7 Brooks et al. (2009).
304 11 Some Applications of Electromagnetic Theory

bucks, and others). Some other efforts include development of dual mode approaches;
for instance, wireless charging combining the resonance and inductive technologies
together under a single technology. Moreover, some companies are also focusing on
newer technologies referred as highly resonant loosely coupled technology.

11.2.1 Tightly Coupled Wireless Power Systems

First, we can introduce the transformer model system, as shown in Fig. 11.10, to
describe a wireless charging system. The transformer model system comprises the
primary coil (PC), a secondary coil (SC), and the iron core that couples the magnetic
field lines of the PC into SC (see also Fig. 11.10).
The current flowing through the primary coil creates a magnetic field given by
Eq. (7.1), based on Biot-Savart’s law. The change in the current passing through the
PC produces a magnetic field that changes with time, and hence based on Faraday’s
law, a change of magnetic flux through the SC, which results in an induced current
in the secondary coil. Here, the iron core enables collection of magnetic field lines
around the primary coil and pass them to the secondary coil. The induced current in
the secondary coil creates a wireless power transfer system. Based on Faraday’s law,
the voltage V p across the primary circuit is

dΦ B
V p = −N1 , (11.86)
dt
where N1 is the number of turns at the primary coil. Φ B is the magnetic flux through
every turn. If all magnetic flux remains within the iron, the flux through each turn in

Fig. 11.10 A transformer model system

11.2 Wireless Charging 305

the primary equals the flux through each turn of the secondary coil, then the voltage
across the secondary coil is
dΦ B
Vs = −N2 , (11.87)
dt
where N2 is the number of turns in the secondary coil. Then, we obtain
 
Vp N2
Vs = −N2 − = Vp . (11.88)
N1 N1

If N2 > N1 , then Vs > V p , corresponding to the so-called step up transformer. If

N2 < N1 , then Vs < V p , and we have the so-called step down transformer.
If the secondary circuit is open (for example, using some switch), then the primary
coil acts as a simple circuit having an alternative source emf and an inductor (assum-
ing no resistance). That is, power factor F = cos φ = 0 (φ = π/2), and the average
power delivered from AC generator to primary circuit is P av = Irms Vrms cos φ = 0.
If the secondary circuit is closed, I2 current is induced in the secondary circuit. If
the load resistance R L in the secondary circuit is purely resistance, the I2 is in phase
with Vs . If there are no power losses, then

I1 V p = I2 Vs , (11.89)

where the first term gives the input power and the second term is the output power.
Therefore,

Vp
I1 = , (11.90)
Req
Vs
I2 = .
RL

In Eq. (11.90), Req is the equivalent resistance of R L when viewed from primary
circuit side. Thus,

Vp Vp
Req = = (11.91)
I1 I2 Vs
Vp
V p2 V p2
= =
I2 Vs Vs2
RL
 2  2
Vp N1
= RL = RL .
Vs N2

Now, the tightly coupled wireless charger works in the same way, except that
the iron core is removed. Furthermore, the planar coil is used instead of windings,
306 11 Some Applications of Electromagnetic Theory

Fig. 11.11 Tightly coupled

wireless inductive system

as shown in Fig. 11.11. Since there is no iron core, the magnetic field lines passes
through the air, which has a lower permeability compared with iron, and hence there
is lower ability to allow magnetic field to pass through. In fact, the factor is around
7000 lower.8 As a result, the amount of magnetic flux and hence power coupled
in the second coil through air is much lower than iron core. Therefore, to obtain a
high-power transfer (as efficient as possible, typically 70 % DC into DC in a wireless
power system), the primary coil and secondary coil are kept very close proximity to
each other, and in a concentric alignment. That is, the secondary coil couples to the
strongest and largest part of the primary coil magnetic field. Figure 11.11 illustrates
a tightly coupled wireless inductive system. Specifically, any magnetic flux from
primary coil that does not couple to secondary coil is a leakage inductance, which
causes energy to be wasted because it is equivalent to adding an impedance into the
primary coil source that does not induce voltage in the secondary coil. Increasing
impedance in the primary coil source causes the energy looses of the amount I 2 R,
which implies that as the current increases to maintain charging at the target rates in
the receiver the wasted energy increases.
For normal use, the leakage inductance should be minimal; that is, the primary
coil and secondary coil should be kept far from each other. In contrast, the coil to
coil efficiency requires that the primary coil and secondary coil to be close to each
other. At lower efficiency charger has to send more power, and hence higher charger
losses for the same power delivered. Thus, primary and secondary coils must be
arranged tightly coupled. In general, the tightly coupled wireless power systems are
concentrically aligned, approximately the same size, and kept in very close proximity
(as illustrated in Fig. 11.11).
The laws of physics indicate some limitations on the use of tightly coupled sys-
tems. Specifically, they can not be used in many types of consumer devices under
the same standards, and they can charge only one device at a time. These limitations
on the use of the tightly coupled systems can be avoided by using the phenomena of
resonance leading to the development of resonant systems.

8 McGraw-Hill Science and Technology Encyclopedia: Ampére Law.

11.2 Wireless Charging 307

11.2.2 Loosely Coupled Highly Resonant Systems

For these systems, the secondary coil may be coupled to a fewer magnetic field lines.
Besides, they have a larger distance from the primary coil in comparison with tightly
coupled inductive systems. As a result, there exists a higher degree of spatial freedom,
which is sometimes desired, between the primary and secondary coils. However,
we could expect much lower efficiency compared with tightly coupled inductive
systems. Therefore, the primary and secondary coils are at high-resonance coupling
of the magnetic fields to avoid that. This helps to achieve the desired efficiency power
transfer between the coils.
Nikolla Tesla was the first to demonstrate that the resonance principle could be
used to transfer power through the wireless systems. That principle is used to develop
loosely coupled systems (such as phone devices).
In electrical systems, resonance is achieved by an RLC (resistance R-inductance
L-capacitance C) circuit system, which occurs at a specific frequency, determined
by the R, L, and C values. The quality factor, η, is defined by the resistance in the
circuit (see also Eq. (10.186), Chap. 10). High η indicates a high efficiency of energy
transfer. The inductive reactance (ωL) and capacitance reactance (1/ωC) are equal
in magnitude at resonance for which η is determined by the resistance in the circuit.
Here, the resonance frequency is given by

1
ω0 = √ = 2π f , (11.92)
LC

where f is the frequency at the resonance. High value of η is obtained for high-
frequency values. Many resonant systems work at f = 6.78 MHz. Some resonant
wireless power systems work at much lower frequency of f ≈ 100 kHz with simpler
circuit design.
Resonant wireless systems are similar to inductive systems, as illustrated in
Fig. 11.12. Both use a primary and secondary coil. However, the difference is that
in the loosely coupled resonant systems, the secondary coil is not necessary at close
proximity to a large percentage of magnetic field coming from primary coil. For
that high η value coils, however, are necessary. Therefore, efficient power transfer is
not strictly dependent on the alignment, size, shape, or positioning of primary and
secondary coils relative to each other.
The most import thing, multiple secondary coils can also be used to capture
power since each coil can share the overall coupling with the primary coil and still
have efficient power transfer. The advantages of using the loosely resonant coupled
systems include their applicability to different portable electronic devices charged
and their use at all times.
308 11 Some Applications of Electromagnetic Theory

Fig. 11.12 Loosely coupled

resonant systems

11.3 Exercises

Exercise 11.1 The voltage output of a generator is given by

ΔV = (200 V) sin (ωt) . (11.93)

Calculate the rms current in the circuit when this generator is connected to a
100 Ω resistor.

Solution 11.1 First, the maximum voltage is

V0 = 200 V . (11.94)

Then, the rms is given as

V0 200 V
Vrms = √ = √ = 141 V . (11.95)
2 2

The rms current is then

Vrms 141 V
Irms = = = 1.41 A . (11.96)
R 100 Ω
The current passing through the resistor is

ΔVR 200 V
iR = = sin (ωt) = (2.00 A) sin (ωt) . (11.97)
R 100 Ω
From Eq. (11.97), maximum current is I0 = 2.00 A.
11.3 Exercises 309

Exercise 11.2 In a purely inductive AC circuit, L = 25.0 mH and the rms

voltage is 150 V. Calculate the inductive resistance and rms current in the
circuit if the frequency is 60.0 Hz.

Solution 11.2 The inductive reactance is

X L = Lω = L(2π f ) (11.98)
−3
= 2π(25.0 × 10 H)(60.0 Hz)
= 9.42 Ω .

The rms current is

V0
I0 XL V0 Vrms
Irms =√ = √ =√ = (11.99)
2 2 2X L XL
150 V
= = 15.9 A .
9.42 Ω

Exercise 11.3 An 8.00 µF capacitor is connected to the terminal of a 60.0 Hz

AC generator whose rms voltage is 150 V. Calculate the capacitive reactance
and the rms current in the circuit.

Solution 11.3 The angular frequency is

ω = 2π f = 2π(60.0 Hz) = 377 rad/s . (11.100)

The capacitive reactance is

1
XC = (11.101)
Cω
1
=
(8.00 × 10−6 F)(377 rad/s)
= 332 Ω .

The rms current is

310 11 Some Applications of Electromagnetic Theory

Vrms
Irms = (11.102)
XC
150 V
=
332 Ω
= 0.452 A = 452 mA .

Exercise 11.4 In a series RLC circuit, the applied voltage has a maximum
value of 120 V and oscillates with frequency of 60.0 Hz. The circuit consists
of an inductor with a varying inductance, a 200 Ω resistor, and a 4.00 µF
capacitor. Calculate the value of L such that the voltage across the capacitor
lags the applied voltage by 30◦ .

Solution 11.4 In Fig. 11.13, we present the phasor diagram. From Fig. 11.13,
φ = −60◦ , where the minus sign indicates that the angle is measured clockwise.
Furthermore,
X L − XC
tan φ = . (11.103)
R
Or,
2π f L = X L = X C + R tan φ . (11.104)

Thus,

Fig. 11.13 The phasor

diagram of the problem
11.3 Exercises 311

1 R
L= + tan φ (11.105)
(2π f ) C
2 2π f
1 200 Ω
= −6
+ tan(−60.0◦ )
(2π60.0 Hz) (4.00 × 10 F) 2π60.0 Hz
2

= 0.840 H = 840 mH .

Exercise 11.5 A series of RLC AC circuit has

R = 425 Ω; L = 1.25 H; C = 3.50 µF; (11.106)

ω = 377 rad/s; V0 = 150 V .

(a) Calculate the inductive reactance, capacitive reactance, and the impedance.
(b) Calculate maximum current.
(c) Calculate the phase angle shift between the current and voltage.
(d) Determine the maximum voltage and instantaneous voltage across any
element.

Solution 11.5 (a) The inductive reactance is

X L = Lω (11.107)
= (1.25 H)(377 rad/s)
= 471 Ω .

The capacitive reactance is

1
XC = (11.108)
Cω
1
=
(3.50 × 10−4 , F)(377 rad/s)
= 758 Ω .

The impedance is given by


Z= R 2 + (X L − X C )2 (11.109)

= (425 Ω)2 + (471 Ω − 758 Ω)2
= 513 Ω .

(b) The maximum current is

312 11 Some Applications of Electromagnetic Theory

V0
I0 = (11.110)
Z
150 V
=
513 Ω
= 0.292 A = 292 mA .

(c) The phase angle shift is

 
−1 X L − XC
φ = tan (11.111)
R
 
−1 471 Ω − 758 Ω
= tan
425 Ω
 
287
= tan−1 −
425
◦
= −34.0 .

(d) The voltage drop across the resistor is

ΔVR = V0,R sin (ωt) . (11.112)

Thus,

V0,R = I0 R (11.113)
= (0.292 A)(425 Ω) = 124 V .

Substituting Eq. (11.113) into (11.112), we have

ΔVR = (120 V) sin (377t) . (11.114)

The voltage drop across the inductor is

 π
ΔVL = V0,L sin ωt − = −V0,L cos (377t) , (11.115)
2
where the amplitude is

V0,L = I0 X L (11.116)
= (0.292 A)(471 Ω) = 138 V .

Therefore,
ΔVL = −(138 V) cos (377t) , (11.117)

Next, we determine the voltage drop across the capacitor as

11.3 Exercises 313
 π
ΔVC = V0,C sin ωt + = V0,C cos (377t) , (11.118)
2
with amplitude determined as

V0,C = I0 X C (11.119)
= (0.292 A)(758 Ω) = 221 V .

The voltage across the capacitor is then

ΔVC = (221 V) cos (377t) , (11.120)

Exercise 11.6 Calculate the average power delivered to the series RLC circuit
with R = 425 Ω, L = 1.25 H, C = 3.50 µF, ω = 377 rad/s, and V0 = 150 V.

Solution 11.6 The rms current is

I0 V0
Irms = √ = √ , (11.121)
2 2Z

where

Z= R 2 + (X L − X C )2 (11.122)
  2
1
= R + 2π f L −
2
2π f C
= 513 Ω .

Therefore, the rms current is

150 V
Irms = √ = 0.206 A = 206 mA . (11.123)
2(513 Ω)

The rms voltage is

V0 150 V
Vrms = √ = √ = 106 V . (11.124)
2 2

The phase angle shift is

314 11 Some Applications of Electromagnetic Theory
 
X L − XC
φ = tan−1 (11.125)
R
 
471 Ω − 758 Ω
= tan−1
425 Ω
◦
= −34.0 .

Therefore, the power factor is

F = cos φ = cos(−34.0◦ ) = 0.829 . (11.126)

The average power delivered by the AC generator is

Pav = Irms Vrms cos φ (11.127)

= (0.206 A)(106 V)(0.829)
= 18.1 W .

Exercise 11.7 Let us assume a series RLC circuit in which R = 150 Ω, L =

20.0 mH, Vrms = 20.0 V, and ω = 5000 rad/s. Calculate the capacitance C for
which the current is maximum.

Solution 11.7 The current is maximum at resonance. The resonance frequency is

1
ω0 = √ . (11.128)
LC

Solving it for C, we get

1
C= (11.129)
Lω02
1
=
(20.0 × 10−3
H)(5.00 × 103 rad/s)2
= 2.00 × 10−6 F = 2.00 µF .

Exercise 11.8 An electricity generator needs to deliver 20 MW of power to a

city 1.0 km away.
(a) If the resistance of the wire is 2.0 Ω and the electricity costs about
0.10 $/kWh calculate the cost of the utilities to the company to send the
11.3 Exercises 315

power to the city for one day. A common voltage for commercial power
generators is 22 kV, but a step up transformer is used to boost the voltage
to 230 kV before transmission.
(b) Do the same calculations when the power plant delivers the electricity at
its original voltage of 22 kV.

Solution 11.8 (a) The power losses in the transmission line are due to the line
resistance:
P = I2R = IV . (11.130)

The current is
P 20 × 106 W
I = = = 87 A . (11.131)
V 230 × 103 V

Therefore,
P = (87 A)2 (2.0 Ω) = 15 kW . (11.132)

The energy loss in 1 day (that is 24 h) is

E loss = P(24 h) = (15 kW)(24 h) = 360 kWh . (11.133)

The cost in dollars is

Cost = E loss (0.10 $/kWh) = (360 kWh)(0.10 $/kWh) = 36 $ . (11.134)

(b) The current is

P 20 × 106 W
I = = = 910 A . (11.135)
V 22 × 103 V

The power loss is

P = (910 A)2 (2.0 Ω) = 1700 kW . (11.136)

The energy loss in 24 h is

E loss = P(24 h) = (1700 kW)(24 h) = 40800 kWh . (11.137)

The cost in dollars is

Cost = E loss (0.10 $/kWh) (11.138)

= (40800 kWh)(0.10 $/kWh)
= 4080 $ .
316 11 Some Applications of Electromagnetic Theory

Exercise 11.9 The rms voltage in output of an AC generator is 200 V and

operating frequency is 100 Hz. Determine the expression of the output voltage
as a function of time t.

Solution 11.9 First, the angular frequency is

ω = 2π f = 2π(100 Hz) = 628 rad/s . (11.139)

The amplitude of the output voltage is

√ √
V0 = Vrms 2 = (200 V) 2 = 288 V . (11.140)

Therefore, the instantaneous output voltage is

ΔV (t) = V0 sin (ωt) = (288 V) sin (628t) . (11.141)

Exercise 11.10 (a) Calculate the resistance of a light bulb that has an average
power of Pav = 75 W, connected to a 60 Hz power source having a maximum
voltage of 170 V. (b) Calculate the resistance of a 100 W bulb.

Solution 11.10 (a) For an AC generator, the output voltage is

ΔV (t) = V0 sin (ωt) , (11.142)

where V0 = 170 V and ω is

ω = 2π f = 2π(60 Hz) = 377 rad/s . (11.143)

Thus,
ΔV (t) = (170 V) sin (377t) . (11.144)

The average power is

V02
Pav = . (11.145)
2R
Therefore, the resistance is

V02 (170 V)2

R= = = 192.7 Ω . (11.146)
2Pav 2(75 W)
11.3 Exercises 317

(b) If Pav = 100 W, then

V02 (170 V)2

R= = = 144.5 Ω . (11.147)
2Pav 2(100 W)

References

Anandakrishman R, Aguilar B, Onufriev AV (2012) Nucleic acids research 40:W537–W541

Brooks BR, Brooks CL, MacKerell AD, Nilsson L, Petrella RJ, Roux B, Won Y, Archontis G, Bartels
C, Boresch S, Caflisch A, Caves L, Cui Q, Dinner AR, Feig M, Fischer S, Gao J, Hodoscek M, Im
W, Kuczera K, Lazaridis T, Ma J, Ovchinnikov V, Paci E, Pastor RW, Post CB, Pu JZ, Schaefer M,
Tidor B, Venable RM, Woodcock HL, Wu X, Yang W, York DM, Karplus M (2009) CHARMM:
The biomolecular simulation program. J Comput Chem 30:1545
Czapinska H, Krzywda S, Sheldrick GM, Otlewski J, Jaskolski M (1999) High-resolution structure
of bovine pancreatic trypsin inhibitor with altered binding loop sequence. J Mol Biol 295:1237
Deuflhard P, Hermans J, Leimkuhler B, Mark AE, Reich S, Skeel RD (eds) (1997) Computational
molecular dynamics: challenges. Methods, ideas. Springer
Kamberaj H (2020) Molecular dynamics simulations in statistical physics: theory and applications.
Springer Nature
MacKerell AD Jr, Brooks B, Brooks CL III, Nilsson L, Roux B, Won Y, Karplus M (1998)
CHARMM: the energy function and its parametrization with an overview of the program. Wiley,
Chichester
Martin C, Richard V, Salem M, Hartley RW, Mauguen Y (1999) Acta Crystallogr Sect D 55:386
R. Phillips and J. Kondev and J. Theriot and H. G. Garcia (2013) Physical Biology of the Cell, 2nd
edn. Garland Science, Taylor and Francis Group
Chapter 12
Electromagnetic Waves in Vacuum
and Linear Medium

This chapter aims to derive electromagnetic wave equations in

vacuum and a linear medium. Also, this chapter aims to
introduce reflection’s law and Snell’s law of optics, and the
Fresnel’s equations.

In this chapter, we discuss the electromagnetic wave equations in vacuum and linear
media. Furthermore, the reflection’s law and the Snell’s law are derived, along with
the polarization of electromagnetic planar waves and Fresnel’s equations. For further
reading, one can also consider other available literature (Holliday et al. 2011; Jackson
1999; Landau and Lifshitz 1971; Sykja 2006; Griffiths 1999; Altland and Simons
2010; Protheroe 2013).

12.1 Electromagnetic Wave Equations in Vacuum

Considering the Maxwell’s equations (see Eq. (9.8) in Chap. 9), the electromagnetic
wave equations in vacuum are derived. Furthermore, we consider charge-free and
current-free medium. These are waves traveling with speed c in vacuum.
Taking the curl of the second equation in Eq. (9.8) (Chap. 9), known as Faraday
law, we obtain  
∂B
∇ × (∇ × E) = −∇ × . (12.1)
∂t

Using the expression

∇ × (∇ × a) = ∇(∇ · a) − ∇ 2 a, (12.2)

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 319
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_12
320 12 Electromagnetic Waves in Vacuum and Linear Medium

we can write that

∂
∇(∇ · E) − ∇ 2 E = − (∇ × B) . (12.3)
∂t
Substituting the third expression of Eq. (9.8) (Chap. 9) into Eq. (12.3), we find

∂J ∂ 2E
∇(∇ · E) − ∇ 2 E = −μ0 − μ0 0 2 . (12.4)
∂t ∂t
Since there is vacuum, the current density is zero, J = 0, and the charge density
ρ = 0; therefore, ∇ · E = ρ/0 = 0 (see also Eq. (9.8), Chap. 9). Thus, we finally
obtain
∂ 2E
∇ 2 E − μ0 0 2 = 0. (12.5)
∂t
√
Using the relation c = 1/ μ0 0 for the speed of light in vacuum, we write

1 ∂ 2E
∇2 E − = 0. (12.6)
c2 ∂t 2
Now, we can take the curl of the third expression in Eq. (9.8) (Chap. 9), known
as Ampére’s law:

∂
∇ × (∇ × B) = μ0 ∇ × J + μ0 0 (∇ × E) . (12.7)
∂t
Using Eq. (12.2), the fourth expression in Eq. (9.8) (Chap. 9), and J = 0, we find

1 ∂ 2B
∇2 B − = 0. (12.8)
c2 ∂t 2
Equations (12.6) and (12.8) give the wave equations for the electric field vector E
and magnetic field vector B, respectively. Note that these equations represent planar
wave equations in three dimensions and they have the same form. Furthermore, the
speed of the wave propagation in vacuum is the speed of light in vacuum c. The
monochromatic plane wave solutions of Eqs. (12.6) and (12.8) are given as

E(r, t) = E0 exp (i (k · r − ωt)) , (12.9)

B(r, t) = B0 exp (i (k · r − ωt)) ,
√
where i = −1. In Eq. (12.9), k is the wave vector, with magnitude given as

2π
k= , (12.10)
λ
12.1 Electromagnetic Wave Equations in Vacuum 321

Fig. 12.1 Illustration of the electromagnetic wave propagation along y-axis

where λ is the wavelength, which denotes the distance between adjacency maximum
points:
λ = cT, (12.11)

where T is the period. Direction of the vector k gives the direction of the wave’s
propagation. The magnetic and electric field vectors are identical at all points at with
the phase angle is
k · r − ωt = 2mπ + constant, (12.12)

where m = 0, 1, 2, . . .. That is, the vectors (E, B) are identical in planes perpendic-
ular to vector k that are separated one wavelength λ.
Figure 12.1 illustrates propagation of electromagnetic wave along positive y-axis,
and relative orientation of vectors (E, B) and k.
ω is the angular frequency, which is related
ω
c= , (12.13)
k
where c defines the phase velocity v p (which is equal to the propagation of the
electromagnetic waves velocity in vacuum c). In the dispersion medium, k = k(ω).
Therefore, combining Eqs. (12.10), (12.11), and (12.13), we write

2π
λ= . (12.14)
ω
Note that the expressions in Eq. (12.9) represent the solutions for the wave equa-
tions; however, they will also represent electromagnetic waves only they satisfy the
Maxwell’s equations (see Eq. (9.8), Chap. 9).
322 12 Electromagnetic Waves in Vacuum and Linear Medium

12.2 Relationships Between k, E, B

The solutions (E, B) monochromatic plane waves have the constants (E0 , B0 ), ω and
k such that (E, B) satisfy Maxwell’s equations. For that, the time derivative, curl and
divergence of the vectors E and B are calculated using Eq. (12.9). For instance, the
time derivatives of the electric field vector E and magnetic field B give

∂E
= −iωE, (12.15)
∂t
∂B
= −iωB.
∂t
Taking the curl of E and B, we write

∇ × E = i (k × E) , (12.16)
∇ × B = i (k × B) .

Using the second Maxwell’s equation (∇ × E = −∂B/∂t), and combining Eqs.

(12.15) and (12.16), we find that

k × E = ωB. (12.17)

Using the fourth Maxwell’s equation (∇ × B = μ0 0 ∂E/∂t for J = 0), we obtain

k × B = −ωμ0 0 E. (12.18)

Taking the divergence of both sides in expressions given by Eq. (12.9), it can be
written that

∇ · E = i (k · E) , (12.19)
∇ · B = i (k · B) .

Using the first Maxwell’s equation (∇ · E = 0 for ρ = 0), we get

k · E = 0, (12.20)

and from the third Maxwell’s equation (∇ · B = 0), we obtain

k · B = 0. (12.21)

From Eqs. (12.20) and (12.21), we conclude that both electric field vector E and
magnetic field vector k are perpendicular to vector k. Note that k represents the
direction of propagation of the electromagnetic wave; therefore, E and B are per-
pendicular to the direction of propagation of electromagnetic wave, and hence, the
electromagnetic waves are transverse waves.
12.2 Relationships Between k, E, B 323

Furthermore,
E · B = E · (k × E)/ω = 0, (12.22)

which indicates that E and B are perpendicular. Therefore, we can write that

E(r, t) = E0 exp (i (k · r − ωt)) , (12.23)

k × E(r, t)
B(r, t) = ,
ω
where Eq. (12.17) is used, and k · E0 = 0. Using Eq. (12.13), we write

k E(r, t) 1
B(r, t) = = E(r, t). (12.24)
ω c

12.3 Electromagnetic Waves Equations in Linear Medium

Next, we consider the electromagnetic waves in a linear medium of charge-free

and current-free medium. For a linear medium, the electric polarization vector and
magnetization are considered to depend linearly on the fields:

P =  0 χe E (12.25)
M = χm H

and hence

D = 0 E (12.26)
B = μμ0 H

where  = 1 + χe and μ = 1 + χm . For the linear medium, μ0 and 0 of the vacuum

are considered the limit of the magnetic permeability μm ≡ μμ0 and the electric
permittivity of medium m ≡ 0  of the linear medium, respectively, when χm → 0
and χe → 0. That is, the propagation speed of the electromagnetic wave in a linear
medium is
1 c
vp = √ =√ . (12.27)
μm m μ

The curl of the second Maxwell’s equation for a linear medium becomes

∂
∇ × (∇ × E) = − (μ0 μ∇ × H) . (12.28)
∂t
Using Eq. (12.2) and the fourth Maxwell’s equation for a linear medium with no
currents (∇ × H = ∂D/∂t), we obtain
324 12 Electromagnetic Waves in Vacuum and Linear Medium

∂ 2E
∇ 2 E = m μm , (12.29)
∂t 2
where D = m E. Thus, we obtain the wave equation for electric field in a linear
medium as
1 ∂ 2E
∇ 2 E − 2 2 = 0. (12.30)
v p ∂t

Similarly, taking the curl of the fourth Maxwell’s equation in the medium (∇ ×
H = ∂D/∂t for J = 0), we write

∂
∇ × (∇ × H) = − (∇ × D) . (12.31)
∂t
Or,
1 ∂
∇ × (∇ × B) = −m (∇ × E) . (12.32)
μm ∂t

Finally, Eq. (12.32) can be written in the form of the wave equation for B as

1 ∂ 2B
∇2 B − = 0. (12.33)
v 2p ∂t 2

Here, Eqs. (12.30) and (12.33) represent the electromagnetic wave equations in a
linear medium propagating with speed v p .
Note that all the relations derived in the previous section for vacuum between k,
E and B also hold for the linear medium by replacing c with v p . In addition, from
Eq. (12.24), can write

B(r, t)
H (r, t) = (12.34)
μm
1 E(r, t)
=
μm v p

m 1
= E(r, t) ≡ E(r, t),
μm Z

where Z is the wave impedance, a characteristic of the medium. For the vacuum,
 
μm μ0
Z= = ≈ 377 Ω. (12.35)
m 0
12.4 Energy and Momentum of Electromagnetic Waves 325

12.4 Energy and Momentum of Electromagnetic Waves

Using the relation in Eq. (12.24), the contributions of electric and magnetic field to
the total energy density are equal; that is,

E2 B2
m = . (12.36)
2 2μm

Therefore, the total energy density in the electromagnetic wave is

 
E2 B2 B2
u = m + = m E 2 = . (12.37)
2 2μm μm

Furthermore, we can write that

u(r, t) = m E 02 cos2 (k · r − ωt + δ) , (12.38)

where δ denotes a phase angle shift. The flux of energy is given by Poynting vector
S = E × H. For the electromagnetic waves in a linear non-dispersion medium, we
have
 
k×E
S=E× (12.39)
μm ω
1
= (k(E · E) − (E · k)E)
μm ω
1
= E 2k
μm ω
1 ω 2
= E k̂
μm ω v p
√
μm m 2
= E k̂
μm

m 2
= E k̂ = v p m E 2 k̂,
μm

where E · k = 0 is used, and k̂ is a unit vector along k. Therefore, using Eqs. (12.38)
and (12.39), we have
S(r, t) = v p u(r, t)k̂. (12.40)

The intensity of the electromagnetic wave is defined as the time average of the
magnitude of the Poynting vector S as follows:

m | E 0 | 2
I = v p m | E 0 |2 cos2 (k · r − ωt + δ) = v p . (12.41)
2
326 12 Electromagnetic Waves in Vacuum and Linear Medium

The units of the intensity are W/m2 .

The momentum density is defined as

S
g= . (12.42)
c2
Then, the momentum flux of an electromagnetic wave in a linear medium is given as

S
p = vpg = vp . (12.43)
c2
The magnitude of the momentum flux of the electromagnetic wave gives the amount
of momentum crossing unit area of a surface perpendicular to the vector k̂ per unit of
time. Thus, it can be defined as the radiation pressure. If a parallel beam of radiation
is perfectly absorbed by a surface perpendicular to the direction of propagation of
the electromagnetic wave, then the radiation pressure is

vp v 2p
prad,abs = S = u(r, t), (12.44)
c2 c2
where · · ·  denotes a time average. Note that if the radiation is perfectly reflected,
then the radiation pressure is twice as large:

v 2p
prad,ref = 2 u(r, t). (12.45)
c2
In the case of the isotropic electromagnetic wave, the radiation pressure is obtained
by integrating the momentum flux incident on the area A of one side of a plane surface,
as shown in Fig. 12.2. Considering only the component of the momentum density
perpendicular to the plane (g cos θ ) and the projected area A cos θ (as depicted in
Fig. 12.2), we obtain

1
1 v 2p 1 v 2p
prad,iso = 2
u(r, t) cos θ A cos θ d(cos θ ) = u(r, t). (12.46)
A c 3 c2
0

Fig. 12.2 Illustration of the

radiation pressure of an
isotropic electromagnetic
wave
12.4 Energy and Momentum of Electromagnetic Waves 327

In vacuum, v p = c, and hence from Eq. (12.46), we find that

1
prad,iso = u(r, t), (12.47)
3
which represents the state’s equation for the gas of photons.

12.5 Coherence of Electromagnetic Waves

The monochromatic plane electromagnetic waves are an idealization picture because

they have infinite extent in all directions. Furthermore, they are just sinusoidal waves,
and hence their angular frequency spread is zero, Δω = 0. The same would have been
true for the monochromatic spherical waves. These waves (that is, having Δω = 0)
are called completely coherent. If two identical of such waves superimpose, then
the interference can be seen. For instance, consider two electromagnetic waves with
magnitudes of electric field vectors as

E 1 = E 0 cos(k · r1 − ωt), (12.48)

E 2 = E 0 cos(k · r2 − ωt + δ),

where δ is a phase angle shift; moreover, we have assumed they have the same
amplitude E 0 . The resultant wave is given as follows:

E R = E1 + E2 (12.49)
= E 0 cos(k · r1 − ωt) + E 0 cos(k · r2 − ωt + δ)
 
k · r2 − ωt + δ − k · r1 + ωt
= 2E 0 cos
2
 
k · r2 − ωt + δ + k · r1 − ωt
× sin
2
   
k · (r2 − r1 ) + δ k · (r2 + r1 ) δ
= 2E 0 cos sin − ωt + ,
2 2 2

which is the equation of a traveling wave, and its amplitude is

 
k · (r2 − r1 ) + δ
E R0 = 2E 0 cos . (12.50)
2

From Eq. (12.50), when

k · (r2 − r1 ) + δ = 2nπ, n = 0, 1, 2, . . . , (12.51)

328 12 Electromagnetic Waves in Vacuum and Linear Medium

the amplitude of the resultant wave is maximum E R0 = ±2E 0 , and we have con-
structive interference, whereas when

k · (r2 − r1 ) + δ = (2n + 1)π, n = 0, 1, 2, . . . , (12.52)

the amplitude of the resultant wave is minimum E R0 = 0, and we have destruc-

tive interference. Therefore, the interference pattern will be constructed of alternat-
ing bright and dark fringes between 100 % constructive interference with Imax =
v p (m /2)E 2R0 = 4I0 , and 100 % destructive interference with Imin = 0. Thus, the
so-called visibility of the fringes is

Imax − Imin
V = = 1. (12.53)
Imax + Imin

In reality, the electromagnetic waves are only partially coherent; that is, they
may exhibit coherence at some particular location over a limited time. Let tc be the
coherence time and the coherence length L c = v p tc . Then, electromagnetic waves
may display coherence due to the partially coherent wave train of length L c passing
through that location over the time tc . The width of the wave front across which the
coherence is maintained is the so-called coherence width wc . The time or distance
over which the phase is significantly different from the pure sinusoidal wave is used
to determine tc , L c , and wc , which can be used to determine the conditions under
which a monochromatic electromagnetic planar wave may be a good approximation
to the electromagnetic field present.
The real sources of electromagnetic waves will produce waves having a finite
extent and have a spread of frequencies. For instance, the atomic transition have a
natural line width of about Δω ∼ , where is the transition rate. Therefore, the
spectral line emission can be considered quasi-monochromatic.
It can be shown that
ΔxΔk x ∼ 1. (12.54)

Since v p = ω/k and that the wave travels in the direction x so that k x = k, and k̂ = x̂,
then
Δk x = Δω/v p . (12.55)

Then, the wave trains associated with emitted photons have a coherence length of
vp
L c = Δx ∼ . (12.56)

In general, the coherence time will be shorter, and so the observed width of a
spectral emission line broader. For example, if during the transition, an atom collides
with another atom, then a phase angle shift occurs. Therefore, the coherence length
will be reduced. That effect becomes more significant when the density and temper-
ature increase because the probability of a collision to occur increases. This effect
12.5 Coherence of Electromagnetic Waves 329

Fig. 12.3 Illustration of the

coherence width

is called pressure broadening. Besides, the frequency of the emitted photon by an

individual atom will experience a Doppler shift because of the thermal motion. The
Doppler shift amount depend on the velocity component of the atom in the direction
toward the observer. The Doppler broadening produces a Gaussian line shape with
a line width proportional to T 1/2 , where T is the temperature.
In Fig. 12.3, it is illustrated the experiment on determining the coherence width.
For that, consider a quasi-monochromatic plane electromagnetic wave reaching an
aperture with diameter a, which is at distance L from a screen. If we consider the
electromagnetic field at the point O1 on the screen, the contributions from the entire
source add up (that is, all points between S1 and S2 ). All these contributions will
have a finite phase at O1 given that the difference between the paths S1 O1 and S2 O1
is much smaller than λ/2. That is equivalent to the conditions that ka 2 L. If we
consider the point P on the screen, which is at distance y from the center of the screen
O1 , the coherence will be lost as the difference in the paths r1 and r2 is comparable
to λ/2 (or equivalently, a phase difference of Δφ = π ). The distance yc for which
this occurs denotes the radius of the patch of the wave’s front called coherence area:

Ac = π yc2 . (12.57)

This area defines space over which the front of the wave may be considered spatially
coherent, with a coherence width dc = 2yc . If we consider that there is no phase
angle difference between the incident electromagnetic waves reaching the point P,
then the phase difference is due to the path difference given as (for a < y L)

kya
Δφ = k(r2 − r1 ) ≈ . (12.58)
L
Therefore,
2π L λL
dc = = . (12.59)
ka a
There are artificial and natural sources of the coherent electromagnetic waves. For
example, a natural source of the coherent electromagnetic wave is the astrophysical
MASER (Microwave Amplification by Stimulated Emission of Radiation) in which
a molecular line is observed at the microwave region of the spectrum. Astrophysical
MASER occur in the shocked regions in molecular clouds in the Galaxy. Long-lived
330 12 Electromagnetic Waves in Vacuum and Linear Medium

upper level of a molecular species is created by excitation through the collisions or

infrared photons giving rise to a population inversion; that is, more molecules exist in
the upper level than in lower level. That population inversion give rise to stimulated
emission to dominate over the absorption, and hence giving a negative absorption
coefficient.
A single photon can initiate MASER action from the spontaneous decay. The
photon propagates across the region of the molecular cloud having population inver-
sion and the photon repeated stimulated emission of essentially identical photons in
phase with itself, similarly to the photons produced by stimulated emission. There-
fore, there exists a cascade where the number of photons build up exponentially with
distance. Because the path lengths across the cloud are huge the resulting bright-
ness of coherent microwave radiation is sufficiently high to be observed with radio
telescopes.
The artificial sources of coherent electromagnetic waves include MASER and
LASER. For instance, in a laboratory setup artificial MASERs using hyperfine split-
ting of the 21 cm line spin-flip transition of the ground state of neutral atom hydrogen
in a weak magnetic field, or Zeeman effect, are used as a frequency standard. The
LASER (Light Amplifier by Stimulated Emission of Radiation) has many applica-
tions. The working principle of a LASER is the same as the astrophysical MASER,
introduced above. However, the main difference is that transitions in the optical or
infrared are used in LASERs, and depending on the medium various methods are
used to create the upper metastable level. Furthermore, in the LASER, the enormous
path lengths needed are achieved by creating a LASER cavity containing LASER
medium between two parallel mirrors in which the light is reflected back and forth,
where only a small fraction of radiation is escaping away from one of the mirrors
(which is allowed to partially transmit light) to form a LASER beam.
It is worthy to note that coherence, including the quantum optics, is essential
in physical optics. Furthermore, it is a field with a broad area of research, and the
electromagnetic field for studying the LASER beam is a good approximation to the
monochromatic electromagnetic plane wave.

12.6 Polarization of Electromagnetic Waves

Consider a monochromatic plane electromagnetic wave propagating along the posi-

tive z-axis; that is, k = k ẑ, where ẑ is a unit vector along +z-direction. Thus, we can
write the electric field vector of the electromagnetic field as
 
E(r, t) = E 0x eiδx x̂ + E 0y eiδ y ŷ ei(kz−ωt) . (12.60)

The real component of the electric field is

 
Re (E(r, t)) = E 0x cos (kz − ωt + δx ) x̂ + E 0y cos kz − ωt + δ y ŷ. (12.61)
12.6 Polarization of Electromagnetic Waves 331

Fig. 12.4 Linear

polarization

There exist linearly, circularly, and elliptically polarization of electromagnetic plane

waves depending on the ration E 0y /E 0x and the relative phase shift (δ y − δx ).

12.6.1 Linear Polarization

When δx = δ y ≡ δ, the linear polarization is obtained. From Eq. (12.61), we have

 
Re (E(r, t)) = E 0x x̂ + E 0y ŷ cos (kz − ωt + δ) . (12.62)

The locus of the tip of the vector E undergoes a simple periodic motion along the line
that forms the angle θ (such that tan θ = E 0y /E 0x ) related to the E x -axis in a plane
of E y and E x axes, as depicted in Fig. 12.4. In other words, the electromagnetic wave
is linearly polarized where the electric field direction is an angle θ to the E x -axis, or
equivalently a plane polarized electromagnetic wave.
The quasi-monochromatic electromagnetic plane waves can also be linearly
polarized, which can be produced by passing the unpolarized light waves across
a monochromatic filter and then across a polarized (for example, a Polaroid Film),
or by passing a laser beam across a polarizer.

12.6.2 Circular and Elliptical Polarization

In the circular and elliptical polarization, the electric and magnetic field vectors
(E, B) rotate with angular frequency ω. In the case of a circular polarization, E 0y =
E 0x ≡ E 0 and δ y − δx = ±π/2. Therefore, the locus of the tip of vector E undergoes
a circular motion with a radius E 0 in the x − y-plane with an angular frequency ω,
as shown in Fig. 12.5. Depending on the sign of δ y − δx there exist two possibilities,
namely the left polarization (or positive helicity) and right polarization (or negative
helicity). Practically, one can define δx = 0, and discuss the possible polarization in
terms of the sign of δ y . For circularly polarized wave traveling along the positive
z-axis (that is, k = k ẑ), we define the basis unit vectors as
332 12 Electromagnetic Waves in Vacuum and Linear Medium

Fig. 12.5 Circular

polarization

1  
e R = √ x̂ − i ŷ , (12.63)
2
1  
e L = √ x̂ + i ŷ .
2

The right circular polarized electromagnetic wave has the following electric field
vectors:
1  
E(r, t) = E 0 ei(kz−ωt) e R = √ E 0 ei(kz−ωt) x̂ + e−iπ/2 ŷ , (12.64)
2

where ±i = ∓e−iπ/2 is used. Therefore,

1  
Re (E(r, t)) = √ E 0 cos(kz − ωt + δx )x̂ + sin(kz − ωt + δx )ŷ . (12.65)
2

For the left circular polarized electromagnetic wave, we write

1  
E(r, t) = E 0 ei(kz−ωt) e L = √ E 0 ei(kz−ωt) x̂ + e+iπ/2 ŷ . (12.66)
2

Therefore,

1  
Re (E(r, t)) = √ E 0 cos(kz − ωt + δx )x̂ − sin(kz − ωt + δx )ŷ . (12.67)
2

Figure 12.5 presents a situation where the observer is at the z-axis and is looking the
wave approaching towards the observer. The right-hand rule can be used to determine
the direction of rotation of electric field vector E; that is, with the right hand, we let
the thumb, which gives direction of the source of the electromagnetic waves, to point
in the direction of the negative z-axis (along −ẑ), then if the fingers of the right-hand
curl in the direction of rotation of the vector E, the polarization is right-handed.
For the elliptical polarization the tip of the electric field vector E rotates with an
angular speed ω along an ellipse in the x − y-plane. For simplicity, we are assuming
12.6 Polarization of Electromagnetic Waves 333

Fig. 12.6 Elliptical

polarization

that the major axis of the ellipse is along the y-axis; however, in general, it can point
in any arbitrary direction. Besides, we assume that δ y − δx = ±π/2 as in the case
of circular polarization. Then, we have

E 0x E 0y
Re (E(r, t)) = √ cos(kz − ωt + δx )x̂ + √ sin(kz − ωt + δ y )ŷ (right),
2 2
(12.68)
E 0x E 0y
Re (E(r, t)) = √ cos(kz − ωt + δx )x̂ − √ sin(kz − ωt + δ y )ŷ (left).
2 2

Figure 12.6 shows the elliptical polarization for E 0x < E 0y .

12.7 Reflection and Refraction of Electromagnetic Waves

Consider the reflection and transmission of electromagnetic waves at the boundaries

between linear media. That will derive the so-called laws of reflection and refraction.
For a linear medium, the so-called refraction index is defined by
c
n= . (12.69)
vp

Using Eq. (12.27), we obtain that

√
n= μ. (12.70)

Furthermore, we write

λ0 = nλ, (12.71)
k
k0 = ,
n
334 12 Electromagnetic Waves in Vacuum and Linear Medium

where λ0 and k0 are the wavelength and magnitude of wave number in vacuum,
and λ and k the corresponding quantities in the linear medium. Using the relation
k = ω/v p , the electromagnetic planar wave vectors are also related as

1
B(r, t) = (k × E(r, t)) (12.72)
ω
1 
= k̂ × E(r, t)
vp
n
= k̂ × E(r, t) .
c

12.7.1 Laws of Reflection and Refraction

Figure 12.7 presents a monochromatic electromagnetic plane wave incident at the

boundary separating two linear media (dielectrics) with (n 1 , μ1 , 1 ) and (n 2 , μ2 , 2 ),
respectively. The incident wave is partly reflected in the first medium and partly
transmitted in the second medium (called refracted wave). The reflected and refracted
rays are in the same plane with the incident ray and the normal to the interface
boundary, which is also called plane of incidence.
The electric and magnetic fields of the incident, reflected, and refracted waves are

E1 (r, t) = E10 exp (i (k1 · r − ω1 t)) (incident wave), (12.73)

  
E1 (r, t) = E10

exp i k1 · r − ω1 t (reflected wave),
E2 (r, t) = E20 exp (i (k2 · r − ω2 t)) (refracted wave),
n
B(r, t) = (k × E(r, t)) (for every wave).
c

Fig. 12.7 Reflection and

refraction of the incident
electromagnetic planar
waves at the boundary
between two linear media
(dielectrics) with (n 1 , μ1 , 1 )
and (n 2 , μ2 , 2 ),
respectively. θ1 is the
incident angle, which is the
angle between the incident
planar wave direction and
normal line to the boundary,
θ1 is the reflection angle, and
θ2 is the refraction angle
12.7 Reflection and Refraction of Electromagnetic Waves 335

The electromagnetic field vectors (E, B) satisfy the boundary conditions; that is,
at all the times ω is the same for all three waves (namely, incident, reflected, and
refracted), and the vectors E, D, B, and H have the same phase angle at any point at
the boundary. Therefore, for any two points r1 and r2 at the boundary plane between
the two media, we obtain

k1 · Δr = k1 · Δr = k2 · Δr, (12.74)
k1 sin θ1 = k1 sin θ1 = k2 sin θ2 ,

where Δr = r2 − r1 is the difference between any two arbitrary points along the
boundary of the two dielectrics. Since the incident and reflected wave are traveling
in the same medium, the magnitudes k1 = k1 ; therefore,

θ1 = θ1 . (12.75)

This is known as the reflection’s law.

Furthermore, from Eq. (12.74), we have

k1 sin θ1 = k2 sin θ2 . (12.76)

Using relations in Eq. (12.71), we obtain

k0 n 1 sin θ1 = k0 n 2 sin θ2 , (12.77)

or,
n 1 sin θ1 = n 2 sin θ2 . (12.78)

Equation (12.78) is also known as refraction’s law or Snell’s law.

The reflection’s law and Snell’s law determine the directions of the reflected
and refracted (transmitted) waves; that is, using Eqs. (12.75) and (12.78), we can
determine the reflected angle θ1 and refracted angle θ2 , if the incident angle θ1 is
known, and if the index of refraction n is known for both dielectrics. Next, the
fraction of the reflected or transmitted intensity is also of interest. The fraction of
the incident intensity that is reflected is called reflectance (denoted by R), and the
fraction of the incident intensity that is refracted is called transmittance, denoted by
T.
Note that the reflectance and transmittance are different if the polarization plane
is oriented parallel or perpendicular to the incident plane. The incident electromag-
netic plane wave can be considered as superposition of two linearly polarized waves
with polarization’s planes being orthogonal, where one of the polarization plane is
parallel and the other perpendicular to the plane of the incidence. The polarization
perpendicular to the plane of the incidence is denoted as TE mode, and the polar-
ization parallel to that plane is called TM mode. In the TE mode (transverse electric
field), the vector E is perpendicular to the plane of incidence, and in the TM mode
336 12 Electromagnetic Waves in Vacuum and Linear Medium

(transverse magnetic field), the vector B is perpendicular to the plane of incidence.

Note that in the electromagnetic plane waves, the vectors E and B are both perpen-
dicular to each other and perpendicular to the vector k (which indicates the direction
of the wave propagation).

12.8 Fresnel Equations

Fresnel equations give the relationships between the amplitude reflection coefficients
and the reflectance and transmittance with the incidence angle θ for the two polar-
ization.

12.8.1 Boundary Conditions

Consider an electromagnetic plane wave intersects the interface between two linear
medium, as depicted in Fig. 12.8. The boundary conditions on the electromagnetic
field determine the characteristics of the reflected and transmitted waves. In the first
linear medium where the incident wave is passing through, the electric and magnetic
field vectors (E, B) are the sum of the incident and reflected waves. The boundary
conditions to be satisfied are derived from Maxwell’s equations.
The Gauss’s law can be applied for the closed surface S. For electric field,

D · dS = q f , (12.79)
S

and for magnetic field (no-magnetic charges exist),

B · dS = 0. (12.80)
S

Fig. 12.8 Boundary

conditions on the
electromagnetic plane wave
12.8 Fresnel Equations 337

If the closed surface is taken infinitesimally small (such that volume charge density
converges to surface charge density σ f ), and since there is no charges (that is, σ f = 0),
we obtain the following boundary conditions:
 
1 E1 + E1 · ẑ = 2 E2 · ẑ for D, (12.81)
 
B1 + B1 · ẑ = B2 · ẑ for B.

Using Ampére’s law for the loop C2 (see also Fig. 12.8), we can write

  
∂D
H · dr = Jf + · dS, (12.82)
C2 S ∂t

and using the Faraday’s law for the loop C1 , we obtain


∂B
E · dr = − · dS. (12.83)
C1 S ∂t

If we assume that the loops becomes infinitesimally thin such that contributions from
∂D/∂t ≈ 0 and ∂B/∂t ≈ 0, then only the free surface charge density contributes.
Therefore,

1   1
B1 + B1 · x = B2 x for H, (12.84)
μ1 μ2
1   1
B1 + B1 · y = B2 y for H,
μ1 μ2
 
E1 + E1 · x = E2 x for E,
 
E1 + E1 · y = E2 y for E.

Also, we can use the relationship between the magnitude of the electric and magnetic
field: B = n E/c.

12.8.2 Perpendicular Polarization

Consider the perpendicular polarization of the electromagnetic plane wave with a

geometry as depicted in Fig. 12.9. For this geometry, we have

E1 · ẑ = 0, E1 · ŷ = 0, B1 · x̂ = 0, (12.85)
E1 · ẑ = 0, E1 · ŷ = 0, B1 · x̂ = 0,
E2 · ẑ = 0, E2 · ŷ = 0, B2 · x̂ = 0.

Therefore, there are three equations remaining only as follows:

338 12 Electromagnetic Waves in Vacuum and Linear Medium

Fig. 12.9 Perpendicular

polarization of the
electromagnetic plane wave

 n π n2 π
E 1⊥ + E 1⊥, − θ1 = E 2⊥ cos
1
cos − θ2 for B, (12.86)
 c n 2
n2
c 2
−E 1⊥ + E 1⊥, cos θ1 = −E 2⊥
1
cos θ2 for H,
cμm cμm

E 1⊥ + E 1⊥, = E 2⊥ for E.

Solving Eq. (12.86) for E 1⊥, and E 2⊥ , we obtain

n2
cos θ1 − cos θ2
n1
E 1⊥, = n2 E 1⊥ (12.87)
cos θ1 + cos θ2
n1
2 cos θ1
E 2⊥ = n2 E 1⊥ .
cos θ1 + cos θ2
n1

Using the Snell’s law, Eq. (12.79), we obtain that

  2
n1
cos θ2 = 1− sin θ1 . (12.88)
n2

Substituting Eq. (12.88) into (12.87), we get

12.8 Fresnel Equations 339
 2
n2
cos θ1 − − sin2 θ1
n1
E 1⊥, =  2 E 1⊥ (12.89)
n2
cos θ1 + − sin2 θ1
n1
2 cos θ1
E 2⊥ =   E 1⊥ .
2
n2
cos θ1 + − sin2 θ1
n1

Then, the amplitude of the reflection and transmission coefficient for perpendicular
polarization are as follows:
 2
n2
cos θ1 − − sin2 θ1
E ⊥, n1
R⊥ = 1⊥ =  2 , (12.90)
E1 n2
cos θ1 + − sin θ1
2
n1
E 2⊥ 2 cos θ1
T⊥ = =   .
E 1⊥ n2 2
cos θ1 + − sin2 θ1
n1

In Fig. 12.10 are plotted the amplitude coefficients, R⊥ and T⊥ versus the incident
angle for crown glass (n 2 = 1.52) and water (n 2 = 1.33) in the case of perpendicular
polarization of the electromagnetic plane wave. For the medium 1, the index of
refraction was n 1 = 1.00.

Fig. 12.10 The amplitude coefficients versus the incident angle for crown glass (n 2 = 1.52) and
water (n 2 = 1.33) in the case of perpendicular polarization of the electromagnetic plane wave. For
the medium 1, the index of refraction was n 1 = 1.00
340 12 Electromagnetic Waves in Vacuum and Linear Medium

12.8.3 Parallel Polarization

Figure 12.11 shows the geometry of a monochromatic electromagnetic plane wave

with parallel polarization. In this case, the following relations hold:

B1 · ẑ = 0, B1 · ŷ = 0, E1 · x̂ = 0, (12.91)
B1 · ẑ = 0, B1 · ŷ = 0, E1 · x̂ = 0,
B2 · ẑ = 0, B2 · ŷ = 0, E2 · x̂ = 0.

Then, we obtain the following three equations to be solved:


,
1 E 1 + 1 E 1 sin θ1 = 2 E 2 sin θ2 (for D), (12.92)

,
E 1 + E 1 n 1 = E 2 n 2 (for H),

,
−E 1 + E 1 cos θ1 = −E 2 cos θ2 (for E).

,
Solving Eq. (12.92) for E 1 and E 2 , we obtain

n2
cos θ1 − cos θ2
, n
E1 = n1 E1 , (12.93)
2
cos θ1 + cos θ2
n1
2 cos θ1
E2 = n E1 .
2
cos θ1 + cos θ2
n1

Using the Snell’s law, Eq. (12.89) is substituted in Eq. (12.93) to obtain the
following:

 2  2
n2 n2
cos θ1 − − sin2 θ1
, n1 n1
E1 = 2  2 E1 , (12.94)
n2 n2
cos θ1 + − sin θ1
2
n1 n1
 
n2
2 cos θ1
n1
E2 =  2   2 E1 .
n2 n2
cos θ1 + − sin2 θ1
n1 n1

From Eq. (12.94), we obtain the amplitude reflection and transmission coefficients
for parallel polarization as follows:
12.8 Fresnel Equations 341

Fig. 12.11 Parallel

polarization of the
electromagnetic plane wave

 2  2
n2 n2
, cos θ1 − − sin2 θ1
E1 n1 n1
R =
= 2  2 , (12.95)
E1 n2 n2
cos θ1 + − sin2 θ1
n1 n1
 
n2
2 cos θ1
E2 n1
T =
=    2 .
E1 n2 2 n2
cos θ1 + − sin2 θ1
n1 n1

In Fig. 12.12 are plotted the amplitude coefficients, R and T versus the incident
angle for crown glass (n 2 = 1.52) and water (n 2 = 1.33) in the case of parallel
polarization of the electromagnetic plane wave. For the incident wave medium, the
index of refraction was n 1 = 1.00.

12.8.4 External and Internal Reflection

 reflection is defined for n 2 > n 1 . Therefore, in Eqs. (12.91) and (12.95),

The external
the term (n 2 /n 1 )2 − sin2 θ1 is real for every 0◦ ≤ θ1 ≤ 90◦ . This was also illustrated
in Figs. 12.10 and 12.12. For the perpendicular polarization, the amplitude transmis-
sion coefficient, T⊥ , reduces smoothly from about +0.86 to 0.00 (see also Fig. 12.10)
for n 1 = 1.00 and n 2 = 1.52, 1.33 (respectively, for crown glass and water). Further-
more, the amplitude reflection coefficient, R⊥ , reduces gradually from about -0.21
to -1.00. On the other hand, for the parallel polarization, the amplitude reflection
coefficient, R , goes smoothly from about +0.21 to -1.00, as shown in Fig. 12.12.
By definition, the incident angle θ1 for the parallel polarization for which R is zero
342 12 Electromagnetic Waves in Vacuum and Linear Medium

Fig. 12.12 The amplitude coefficients versus the incident angle for crown glass (n 2 = 1.52) and
water (n 2 = 1.33) in the case of parallel polarization of the electromagnetic plane wave. For the
medium 1, the index of refraction was n 1 = 1.00

is called Brewster’s angle, denoted by θ B . That is, at the angle of incidence waves
θ B only perpendicular polarized waves are reflected, and hence, creating linearly
polarized light. Replacing R = 0 in Eq. (12.95), we obtain the Brewster’s angle θ B
as follows:
1
cos θ B =   2 , (12.96)
n2
1+
n1
n2
tan θ B = .
n1

For R⊥, ≥ 0, the phase shift is zero, and for R⊥, < 0, the phase shift on the reflection
is π (or 180◦ ) because e±π = −1. In the case of perpendicular polarization R⊥ ≥ 0,
and hence the phase shift is zero. In the case of the parallel polarization, R ≥ 0 for
θ1 ≥ θ B , and hence the phase shift is zero; however, for θ1 > θ B , R > 0, and hence
there is a phase shift of π (or 180◦ ).
In the case of the internal reflection, n 1 > n 2 , and thus n 2 /n 1 < 1. In Fig. 12.13
are shown the amplitude coefficients as a function of the incident angle θ1 for the
perpendicular polarization. The incident wave medium was crown glass (n 1 = 1.52)
and water (n 1 = 1.33), and the refracted wave medium has an index of refraction
n 2 = 1.00. Fig. 12.14 presents similar plot for parallel polarization. There is a critical
angle θc such that for θ1 > θc the Snell’s law does not apply because sin θ2 becomes
greater than one. For these incident angles (that is, θ1 > θc ), there will be no refracted
waves (i.e., T⊥, = 0); however, there is a total internal reflection such that | R⊥,
|= 1.
Using the Snell’s law, Eq. (12.79), for θ2 = 90◦ , we can determine the critical
angle as follows:
12.8 Fresnel Equations 343

Fig. 12.13 The amplitude coefficients versus the incident angle for crown glass (n 1 = 1.52) and
water (n 1 = 1.33) in the case of perpendicular polarization of the electromagnetic plane wave. For
the refracted wave medium, the index of refraction was n 2 = 1.00. For θ1 > θc , | R⊥ | is plotted

Fig. 12.14 The amplitude coefficients versus the incident angle for crown glass (n 1 = 1.52) and
water (n 1 = 1.33) in the case of parallel polarization of the electromagnetic plane wave. For the
refracted wave medium, the index of refraction was n 2 = 1.00. For θ1 > θc , | R | is plotted

n2
sin θc = , (12.97)
n1
or,  
n2
θc = sin−1 . (12.98)
n1

Since for the internal reflection to occur, n 2 /n 1 < 1, and thus

 2
n2 > 0, θ1 < θc
− sin2 θ1 = (12.99)
n1 < 0, θ1 > θc
344 12 Electromagnetic Waves in Vacuum and Linear Medium

Therefore, from Eqs. (12.90) and (12.95), the amplitude reflection coefficients are
real numbers for θ1 < θc ; that is, for θ1 < θc the phase shift is φ = 0◦ , if R⊥, > 0,
and φ = 180◦ if R⊥, < 0, which is similar to the external reflection. On the other
hand, for θ1 > θc , the amplitude reflection coefficients are complex numbers. In that
case, the amplitude reflection coefficient is written as

R⊥, =| R⊥, | e−iφ , (12.100)

where φ is the phase shift. From Eqs. (12.90) and (12.95), since θ1 > θc , the amplitude
reflection coefficients can be written as follows:

x − iy x 2 + y 2 e−iψ
R⊥, = = = e−2iψ ≡ e−iφ , (12.101)
x + iy x +y e
2 2 +iψ

where φ is the phase shift defined as

y
φ = 2ψ = 2 tan−1 . (12.102)
x
Here, for the perpendicular polarization of the electromagnetic plane wave, x and y
are given as

x = cos θ1 , (12.103)
  2
n2
y= sin2 θ1 − .
n1

For the parallel polarization of the electromagnetic plane wave, x and y are as follows:
 2
n2
x= cos θ1 , (12.104)
n1
  2
n2
y= sin2 θ1 − .
n1

The phase shift as a function of the incident angle for crown glass (n 1 = 1.52) and
water (n 1 = 1.33) in the case of perpendicular polarization of the electromagnetic
plane wave is plotted in Fig. 12.15, and for parallel polarization is shown in Fig. 12.16.
In both cases, for the refracted wave medium, the index of refraction was n 2 = 1.00.
12.8 Fresnel Equations 345

Fig. 12.15 The phase shift versus the incident angle for crown glass (n 1 = 1.52) and water (n 1 =
1.33) in the case of perpendicular polarization of the electromagnetic plane wave. For the refracted
wave medium, the index of refraction was n 2 = 1.00

Fig. 12.16 The phase shift versus the incident angle for crown glass (n 1 = 1.52) and water (n 1 =
1.33) in the case of parallel polarization of the electromagnetic plane wave. For the refracted wave
medium, the index of refraction was n 2 = 1.00

12.8.5 Normal Incidence of Electromagnetic Waves

Consider normal incidence of the electromagnetic plane waves to the boundary of

the two linear media. Since the plane of incidence is not defined, the results for the
perpendicular and parallel polarization are obtained as limiting case for the θ1 going
to zero (θ1 → 0).
Thus, in the case of the perpendicular polarization, Eq. (12.90), for θ1 → 0,
becomes
346 12 Electromagnetic Waves in Vacuum and Linear Medium

n2
1−
E 1⊥, n1
R⊥ = ⊥ = n , (12.105)
E1 2
1+
n1
⊥
E 2
T⊥ = 2⊥ = n2 .
E1 1+
n1

For the parallel polarization (see Eq. (12.95)), the amplitude coefficients in the case
of normal incidence are as follows:
n2
, −1
E1 n
R =
= 1 n , (12.106)
2
E1 1+
n1

E2 2
T =
= n2 .
E1 1+
n1

Comparing Eqs. (12.105) and (12.106), we find that R⊥ = −R for θ1 = 0◦ ; on

the other hand, we expect that for normal incidence the perpendicular and parallel
polarization to be identical. Note that the amplitude coefficients and the phase shifts
are defined relative to the electric field directions of the incident, reflected, and
refracted waves, as depicted in Figs. 12.9 and 12.11.
Figure 12.17 illustrates the direction of electric field for close to normal incidence
in the case of internal reflection (n 2 < n 1 ), as in Fig. 12.17a, external reflection
(n 2 > n 1 ), as in Fig. 12.17b for perpendicular polarization, and internal reflection
(see Fig. 12.17c) and external reflection (see Fig. 12.17d) for parallel polarization.
In the case of external reflection at the normal incidence, the electric field direction
of the reflected wave is opposite to that of the incident wave (see also Fig. 12.17b, d),
whereas in the case of the internal reflection (Fig. 12.17a, c), the electric field direction
is the same as the incident wave. Therefore, for the reflection at normal incidence,
there is a phase shift of 180◦ (or π ) on the reflection for the external reflection;
however, no phase shift occurs for the internal reflection for both polarization waves.

12.8.6 Reflectance and Transmittance

The reflectance and transmittance determine the rate of energy flow incident on the
unit area of the interface: S1 · ẑ = S1 cos θ1 . Then, the reflectance is given as

S1 cos θ1 S  cos θ1 S
r= = 1 = 1, (12.107)
S1 cos θ1 S1 cos θ1 S1

where the reflection’s law is used.

The transmittance is as follows:
12.8 Fresnel Equations 347

Fig. 12.17 The direction of electric field for close to normal incidence: a internal reflection (n 2 <
n 1 ) and b external reflection (n 2 > n 1 ) for perpendicular polarization; c internal reflection and d
external reflection for parallel polarization

S2 cos θ2
t= . (12.108)
S1 cos θ1

In Eqs. (12.107) and (12.108), S is the magnitude of the Poynting vector (for μm ≈ 1;
that is, a dielectric medium):

1 1 n 2
S = EH = EB = E2 = E . (12.109)
m m v p m c

Here, the factor cos θ considers the projected area of the interface as seen by the
incoming or outgoing wave. Therefore, we obtain the reflectance and transmittance
for the perpendicular and parallel polarization as follows:

r⊥, =| R⊥, |2 , (12.110)

 
n2
cos θ2
n1  2
t⊥, = T⊥, = 1 − r⊥, .
cos θ1
348 12 Electromagnetic Waves in Vacuum and Linear Medium

In the case of the internal reflection, as θ1 → θc , R⊥, → 1, while the amplitude

reflection coefficients T⊥ and T remain finite. Because cos θ2 → 0 as θ1 → θc (or
equivalently, cos θ1 → cos θc ), which is finite, then we obtain t⊥, → 0 as θ1 → θc .
Also, r⊥, + t⊥, = 1 in agreement with energy conservation law.
Using the Snell’s law (see Eq. (12.79)), we have
  2
n1
cos θ2 = 1− sin2 θ1 . (12.111)
n2

Substituting Eq. (12.111) into the second expression in Eq. (12.110), we obtain
 2
n2
− sin2 θ1
n1  2
t⊥, = T⊥, . (12.112)
cos θ1

Combining Eq. (12.90) with (12.112), we obtain transmittance for perpendicular

polarization as follows:
 2
n2
4 cos θ1 − sin2 θ1
n1
t⊥ = ⎛  ⎞2 . (12.113)
2
⎝cos θ1 + n2
− sin2 θ1 ⎠
n1

For the parallel polarization, the transmittance is obtained by substituting Eq. (12.95)
into (12.112) as follows:

 2  2
n2 n2
4 cos θ1 − sin2 θ1
n1 n1
t = ⎛   ⎞2 . (12.114)
 2 2
n
⎝ 2 cos θ1 + n
− sin2 θ1 ⎠
2
n1 n1

The reflectance is calculated as follows:

r⊥, = 1 − t⊥, . (12.115)

The reflectance for the external reflection is shown in Fig. 12.18 for perpendicular
and parallel polarization with n 1 = 1.00 and n 2 = 1.52 (crown glass). In Fig. 12.19,
it is shown the reflectance for the internal reflection in the case of perpendicular and
parallel polarization with n 1 = 1.52 (crown glass) and n 1 = 1.00. In the case of the
12.8 Fresnel Equations 349

Fig. 12.18 The reflectance for the external reflection (n 2 > n 1 ) for perpendicular and parallel
polarization with n 1 = 1.00 and n 2 = 1.52 (crown glass)

Fig. 12.19 The reflectance for internal reflection (n 2 < n 1 ) for perpendicular and parallel polar-
ization with n 1 = 1.52 (crown glass) and n 2 = 1.00

external reflection (see also Fig. 12.18) when θ1 → 90◦ , the reflectance approaches
to one. This is a property used in the build of imaging X-ray space telescopes.
If θ1 > θc , then total internal reflection occurs. However, because of the boundary
conditions there is still an electromagnetic field present on the other side of the
interface. The transmitted electromagnetic wave on the other side has a wave vector
as follows: ωn 2  
k2 = sin θ2 ŷ + cos θ2 ẑ . (12.116)
c
For θ1 > θc , using the Snell’s law, we obtain
350 12 Electromagnetic Waves in Vacuum and Linear Medium

cos θ2 = 1 − sin2 θ2 (12.117)
  2
n1
= 1− sin2 θ1
n2

 2
n1
=i sin2 θ1 − 1.
n2

Substituting Eq. (12.117) into (12.116) and using the Snell’s law, we find that
⎛  ⎞
 2
ωn 2 ⎝ n 1 n1
k2 = sin θ1 ŷ + i sin2 θ1 − 1 ẑ⎠ . (12.118)
c n2 n2

The refracted electromagnetic wave electric field vector is

E2 (r, t) = E 02 exp (i(k2 · r − ωt)) (12.119)

⎛ ⎛ ⎛   ⎞
ωn 2 ⎝ n 1 n1 2 2
= E 02 exp ⎝i ⎝ sin θ1 ŷ + i sin θ1 − 1 ẑ⎠ · r
c n2 n2

− ωt))
⎛   ⎞
2
ωn n
= E 02 exp ⎝− sin2 θ1 − 1 ẑ · r⎠
2 1
c n2
  ωn
1
× exp i sin θ1 r · ŷ − ωt ,
c

which indicates that E2 (r, t) is a wave propagating along the y-axis and its amplitude
is ⎛   ⎞
2
ωn n
E 0 = E 02 exp ⎝− sin2 θ1 − 1 ẑ · r⎠ .
2 1
(12.120)
c n2

Therefore, the wave is exponentially decaying along the z-axis direction, which
means in the direction perpendicular to the surface of dielectric; thus, this is a surface
wave. This wave is called evanescent wave. Besides, Eq. (12.120) indicates that the
decay distance of the electric field is of order c/ω, which of order of the wavelength;
therefore, the wave can not propagate into the region outside the dielectric as a
monochromatic plane wave. The evanescent wave may travel a tiny distance (such
as in tunneling phenomena in quantum mechanics), if a slab of similar dielectric
is brought within a few wavelength of the dielectric in which the wave is incident.
If that occurs, a part of the incident wave tunnels through the gap and propagates
in the slab as a monochromatic plane wave in the same direction as the original
incident wave; however, its intensity reduces depending on the width of the gap.
12.8 Fresnel Equations 351

This phenomena is called frustrated total internal reflection, and it can be applied to
optics with beam-splitters and optical fiber junctions.

12.9 Exercises

Exercise 12.1 Assume that Fizeau’s wheel has 360 teeth and is rotating at
27.5 rev/s when a pulse of light passing through opening A is blocked by tooth
B on its return. If the distance to the mirror is 7 500 m, what is the speed of
light?

Solution 12.1 The wheel has 360 teeth, and so it must have 360 openings. Therefore,
because the light passes through opening A but is blocked by the tooth immediately
adjacent to A, the wheel must rotate through an angular displacement of (1/720)
rev in the time interval during which the light pulse makes its round trip. From the
definition of angular speed, that time interval is

Δθ
Δt = (12.121)
ω
(1/720) rev
= = 5.05 × 10−5 s
27.5 rev/s

Thus the speed of light calculated from this data is

2d
c= (12.122)
Δt
2(7500 m)
= = 2.97 × 108 m/s
5.05 × 10−5 s

Exercise 12.2 A beam of light of wavelength 550 nm traveling in air is inci-

dent on a slab of transparent material. The incident beam makes an angle of
40.0◦ with the normal, and the refracted beam makes an angle of 26.0◦ with
the normal. Find the index of refraction of the material.

Solution 12.2 Using the Snell’s law of refraction

n 1 sin θ1 = n 2 sin θ2
352 12 Electromagnetic Waves in Vacuum and Linear Medium

with n 1 = 1.00 for air, we get

sin θ1
n2 = n1 (12.123)
sin θ2
sin 40◦
= (1.00) = 1.47
sin 26◦

Exercise 12.3 A light ray of wavelength 589 nm traveling through air is inci-
dent on a smooth, flat slab of crown glass at an angle of 30.0◦ to the normal.
Find the angle of refraction.

Solution 12.3 Using the Snell’s law

n 1 sin θ1 = n 2 sin θ2

and thus,
n1
sin θ2 = sin θ1
n2

for the air n 1 = 1.00 and for the crown glass n 2 = 1.52. Therefore,
 
1.00
sin θ2 = sin 30◦ = 0.329
1.52
or,
θ2 = sin−1 (0.329) = 19.2◦

Because this is less than the incident angle of 30◦ , the refracted ray is bent toward
the normal, as expected. Its change in direction is called the angle of deviation and
is given by
δ =| θ1 − θ2 |= 10.8◦

Exercise 12.4 A laser in a compact disc player generates light that has a
wavelength of 780 nm in air. (A) Find the speed of this light once it enters the
plastic of a compact disc (n = 1.55). (B) What is the wavelength of this light
in the plastic?

Solution 12.4 (A) We expect to find a value less than 3.00 × 108 m/s because n > 1.
We can obtain the speed of light in the plastic by using Equation
12.9 Exercises 353

c 3.00 × 108 m/s

v= = = 1.94 × 108 m/s
n 1.55
(B) We use Equation
λ
λn =
n
to calculate the wavelength in plastic, noting that we are given the wavelength in air
to be λ = 780 nm:
λ 780 nm
λn = = = 503 nm
n 1.55

Exercise 12.5 A light beam passes from medium 1 to medium 2, with the
latter medium being a thick slab of material whose index of refraction is n 2 .
Show that the emerging beam is parallel to the incident beam.

Solution 12.5 Using the Snell’s law of refraction to the upper surface, we get
n1
n 1 sin θ1 = n 2 sin θ2 → sin θ2 = sin θ1
n2
n2
for the lower surface we get sin θ3 = sin θ2 . Combining these two equations, we
n1
get  
n2 n1
sin θ3 = sin θ1 = sin θ1 → θ3 = θ1
n1 n2

Exercise 12.6 Find the critical angle for an air-water boundary. (The index of
refraction of water is 1.33.)

Solution 12.6 The air above the water having index of refraction n 2 and the water
having index of refraction n 1 . Applying the equation

n2 1.00
sin θc = = = 0.752
n1 1.33
or
θc = sin−1 (0.752) = 48.8◦
354 12 Electromagnetic Waves in Vacuum and Linear Medium

Exercise 12.7 A narrow beam of sodium yellow light, with wavelength

589 nm in vacuum, is incident from air onto a smooth water surface at an
angle θ1 = 35.0◦ . Determine the angle of refraction θ2 and the wavelength of
the light in water. Find the critical angle for an air-water boundary. (The index
of refraction of water is 1.33.)

Solution 12.7 Using the Snell’s law:

n 1 sin θ1 = n 2 sin θ2

where n 1 = 1.00, n 2 = 1.33, and θ1 = 35.0◦ . Thus,

n 1 sin θ1 (1.00)(sin 35.0◦ )

sin θ2 = = ≈ 0.43126
n2 1.33

Or,
θ2 = sin−1 0.43126 ≈ 25.5◦

The wavelength of light in water is:

λ 589 nm
λn = = ≈ 443 nm
n2 1.33

Exercise 12.8 The wavelength of red helium - neon laser light in air is λ =
632.8 nm. (a) What is its frequency? (b) What is its wavelength in glass that
has an index of refraction of 1.50? (c) What is its speed in the glass?

Solution 12.8 We consider that the speed of light in air (n 1 = 1.00) is c = 3.00 ×
108 m/s.
(a) The frequency of light wave is:

1 c 3.00 × 108 m/s

f = = = ≈ 4.74 × 1014 Hz
T λ 632.8 × 10−9 m

(b) The wavelength of light in glass is

λ 632.8 nm
λn = = ≈ 421.9 nm
n 1.50
12.9 Exercises 355

(c) The speed of light in glass is:

c 3.00 × 108 m/s

v= = = 2.00 × 108 m/s
n 1.50

Exercise 12.9 An underwater scuba diver sees the Sun at an apparent angle
of 45.0◦ from the vertical. What is the actual direction of the Sun?

Solution 12.9 The indices of refraction for air and water are respectively: n air =
1.00 and n wat = 1.33. Using the Snell’s law, we write

n air sin θ1 = n wat sin θ2

Here, θ2 = 45.0◦ . Therefore,

   ◦
−1 n wat sin θ2 −1 (1.33) sin 45.0
θ1 = sin = sin
n air 1.00

Or,
θ1 ≈ sin−1 (0.94) ≈ 70.1◦

which is the angle with vertical direction.

Exercise 12.10 A monochromatic plane wave with electric field vector

E(r, t) = E0 exp (i (k · r − ωt))

is traveling along the positive z axis through a linear medium with permittivity
 = 4 and permeability μ = 1, which is polarized along the x axis. The fre-
quency is f = 1 GHz and E has a maximum value of +10−3 V/m at t = 5 ns
and z = 1 m. (a) Determine the angular frequency, phase velocity, wave num-
ber, wave vector, and wavelength. (b) Determine the expression for E(r, t). (c)
Determine the expression for H(r, t). (d) Determine the Poynting vector and
its time-averaged value. (e) Determine the position where E x is maximum at
t = 0 s.

Solution 12.10 (a) The angular frequency is

ω = 2π f ≈ 6.28 × 106 rad/s.

356 12 Electromagnetic Waves in Vacuum and Linear Medium

The phase velocity is

c 3.00 × 108 m/s

vp = √ = √ = 1.5 × 108 m/s.
μ 4

The wave number magnitude is

ω 6.28 × 106 rad/s

k= = = 4.19 × 10−2 m−1 .
vp 1.5 × 108 m/s

The wave number vector is

k = k ẑ = 0.0419ẑ m−1 .

The wavelength is given as

2π 2π
λ= = = 149.88 m.
k 0.0419 m−1
(b) The expression for the electric field vector is
  
E(z, t) = E 0 exp i 0.0419z − 6.28 × 106 t x̂.

At t = 5 ns = 5 × 10−9 s and z = 1 m, we have

E(1 m, 5 × 10−9 s) = E 0 exp (i (0.0419 − 0.0314)) x̂ = E 0 exp (i0.0105) x̂.

The real part is

Re{E(1 m, 5 × 10−9 s)} = E 0 cos(0.0105)x̂,

and the maximum is given as

10−3 = E 0 cos(0.0105)

or,
10−3
E0 = ≈ 10−3 V/m.
0.99989
Thus,     
E(z, t) = 10−3 V/m exp i 0.0419z − 6.28 × 106 t x̂.

(c) The magnetic field vector is

H(r, t) = H (r, t)ŷ,

12.9 Exercises 357

and its magnitude is

1 E(r, t)
H (r, t) = E=  (12.124)
Z μm
m
E(r, t) E(r, t)
=  = 
μ0 1
Z0
0  
E(r, t)
≈ ,
188.5

where  
μm 1
Z= = Z0 = 188.5 Ω,
m 

is the impedance of dielectric medium and Z 0 ≈ 377 Ω is the impedance of the

vacuum. Therefore,
  
H(r, t) = (5.31 × 10−6 T) exp i 0.0419z − 6.28 × 106 t ŷ.

(d) Using Eq. (12.39), the Poynting vector is

E 2 (r, t)
S= ẑ (12.125)
Z
 
V2   
= 5.31 × 10−9 exp i 0.0838z − 12.56 × 106 t ẑ.
(m · Ω)
2

The period is
2π
T = = 1.00 × 10−6 s
ω
The time average of the Poynting vector is given by Eq. (12.41) as follows:

m | E 0 | 2
S = v p (12.126)
2

m | E 0 | 2
=
μm 2
| E 0 |2 10−6 V2
= = ≈ 2.65 × 10−9 2
2Z 377 (m · Ω)
358 12 Electromagnetic Waves in Vacuum and Linear Medium

(e) First, we write E x = E · x̂ as

    
E x (z, t) = 10−3 V/m exp i 0.0419z − 6.28 × 106 t .

At t = 0, we obtain
 
E x (z, 0) = 10−3 V/m exp (i0.0419z) ,

which has a maximum for

0.0419z = 2π.

This
2π
z= ≈ 149.88 m.
0.0419

References

Altland A, Simons B (2010) Condensed matter field theory, 2nd edn. Cambridge University Press
Griffiths DJ (1999) Introduction to electrodynamics, 3rd edn. Prentice Hall
Holliday D, Resnick R, Walker J (2011) Fundamentals of physics. John Wiley and Sons
Jackson JD (1999) Classical electrodynamics, 3rd edn. John Wiley and Sons
Landau LD, Lifshitz EM (1971) The classical theory of fields. Pergamon Press
Protheroe RJ (2013) Essential electrodynamics, 1st edn. Bookboon
Sykja H (2006) Bazat e Elektrodinamikës. SHBUT
Chapter 13
Electromagnetic Waves in Dispersive
Media

This chapter aims to derive electromagnetic wave equations in

dispersive media. In particular, the aims of this chapter include
the understanding the absorption, Lorentz’s oscillator model of
a dielectric, deriving the wave equation of a conductor, deriving
the wave equation of a dilute plasma and understanding the
magnetized plasma or dielectric.

In this chapter, we derive electromagnetic wave equations in dispersive media. This

chapter describes the absorption, Lorentz’s oscillator model of a dielectric, the wave
equation of a conductor, the wave equation of a dilute plasma, and the magnetized
plasma or dielectric. For further reading, one can also consider other available liter-
ature (Jackson 1999; Landau and Lifshitz 1971; Sykja 2006; Griffiths 1999; Altland
and Simons 2010; Protheroe 2013).

13.1 Dispersion and Absorption

In Eq. (12.13) (Chap. 12), it was assumed that the wave number k is a constant;
however, in general, the index of refraction n(ω) and the wave number k(ω) are
functions of the angular frequency:
ω ω
k(ω) = = n(ω) . (13.1)
vp c

For a monochromatic plane wave in matter, we have

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 359
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2_13
360 13 Electromagnetic Waves in Dispersive Media

E(r, t) = E0 exp (i (k · r − ωt)) (13.2)

    
= E0 exp i kr k̂ · r − ωt exp −ki k̂ · r ,

where kr = Re (k) (real part) and ki = Im (k) (imaginary part) such that k =
(kr + iki ) k̂. The intensity of the electromagnetic wave is given as the time aver-
age of the Poynting vector S(r, t)

I (r) = S(r, t) ∝ E 2 (r, t) (13.3)

   
= E 02 exp −α(ω)k̂ · r exp 2ikr (ω)k̂ · r ,

which indicates that intensity decays exponentially with distance, and α =

2Im (k(ω)) = 2ki (ω) is the so-called absorption coefficient.

13.1.1 Lorentz’s Model of Oscillations in Dielectrics

The polarizability of atoms in molecules can be approximately modeled using the

semi-classical approaches, which consider the atoms as nucleus with effective charge
+e surrounded by a spherical cloud of electrons with radius a0 ∼ 10−10 m corre-
sponding to a single valence electron of charge −e (where e magnitude of the elec-
tron’s charge e = 1.6 × 10−19 C). In this approximation, the nucleus is screened by
all the electrons in the atom except one electron (the valence electron), and hence it
has an effective charge of +e.
In the presence of an external electric field, the electron cloud displaces by an
amount re relative to the nucleus, and hence it experiences a restoring force from the
nucleus equal to −κre , where
e2
κ≈ . (13.4)
4π0 a03

Here, κ is a spring constant, which is a measure of the strength of restoring force of

the nucleus. Applying the second law of Newton for the valence electron, we have

m e r̈e = −eE − κre , (13.5)

where r̈e is the second time derivative of the valence electron displacement re and E is
the external electric field vector. In equilibrium, the right-hand side (which is the net
force acting on the valence electron) is zero, and thus the equilibrium displacement
for the valence electron re is
eE
re = − . (13.6)
κ
The dipole momentum is
13.1 Dispersion and Absorption 361

Fig. 13.1 Displacement re of the cloud electron charge (−e) surrounding the nucleus with a charge
+e under the external electric field E due to the electrical force −eE

e2 E
p = ere = − , (13.7)
κ
and the molecular polarizability is

e2
αpol = = 4π0 a03 . (13.8)
κ
If the external electric field E oscillates in time, then the electrons will oscillate in
response. Note there are various dragging forces (such as friction forces and radiation
from stopping1 ) that apply to the valence electron, which are added as a dumping
term proportional to the velocity that are added to the equation of motion as follows:

m e r̈e = −eE(t) − κre − m e γ ṙe , (13.9)

where ve = ṙe = dre /dt is the valence electron velocity and γ is a proportionality
constant.
Figure 13.1 illustrates the displacement of the cloud electron charge under the
external electric field E.

1 Sykja (2006).
362 13 Electromagnetic Waves in Dispersive Media

The frequency of simple harmonic motion for the valence electrons of atoms is
defined as
 
κ e2
ω0 = = ≈ 2 × 1016 rad/s. (13.10)
me 4π0 a03 m e

This is also the resonance frequency and it corresponds to the frequencies in the
ultraviolet range. When a monochromatic plane wave (or photons) with frequency ω0
(or close to that value) strikes an atom, then the work done against the friction forces
to allow the atom oscillating causes an absorption of photons. Note that the atom
is not a classical system with just one resonance, but a quantum system with many
resonant frequencies. Therefore, this semi-classical system is a good approximation
to describe some important properties of dielectrics.
Thus, Eq. (13.9) takes the form

m e r̈e = −eE(t) − m e ω02 re − m e γ ṙe . (13.11)

Consider a sinusoidal electric field of the electromagnetic wave; that is, E(t) ≡
E0 e−iωt . Then, the solution of Eq. (13.11) can be required in the following form:

re (t) = r0 eρt , (13.12)

where ρ and r0 are constants to be determined. Replacing re from Eq. (13.12) into
(13.11), we obtain

m e ρ2 eρt r0 = −eE0 e−iωt − m e ω02 r0 eρt − m e γρr0 eρt , (13.13)

or  
m e ρ2 + ω02 + γρ eρt r0 = −eE0 e−iωt . (13.14)

Equation (13.14) implies that ρ = −iω, and

 
m e ρ2 + ω02 + γρ r0 = −eE0 . (13.15)

Therefore, we obtain
e E0
r0 = − . (13.16)
m e ω02 − ω 2 − iγω

The displacement of the valence electrons is given as follows:

13.1 Dispersion and Absorption 363

e E0
re = − e−iωt (13.17)
m e ω0 − ω 2 − iγω
2

e ω02 − ω 2 + iγω
=−  2  E0 e−iωt
m e ω − ω 2 2 + (γω)2
0

2 +iφ
e (ω0 − ω ) + (γω) e
2 2 2
=−  2 2 E0 e−iωt
me ω − ω + (γω)
0
2 2

e/m e −i(ωt−φ)
= −  2 3/2 E0 e ,
ω0 − ω + (γω)
2 2 2

where

γω
φ = tan−1 . (13.18)
ω02 − ω 2

The minus sign in Eq. (13.17) indicates that the cloud electron displacement is
opposite of the electric field. The dipole moment is as follows:
 
e2 ω02 − ω 2 + iγω
p(t) = −ere (t) =   E(t). (13.19)
m e ω 2 − ω 2 2 + (γω)2
0

Consider a molecule of N atoms, n oscillators with frequency ω0,k , dumping

coefficients γk , and f k cloud electrons (for k = 1, 2, . . . , n). The polarization of
molecule is
 n  2 
N e2  ω0,k − ω 2 + iγk ω
P(r, t) = fk  2 E(t). (13.20)
m e k=1 ω 2 − ω 2 + (γk ω)2
0,k

Using Eq. (12.25) (see Chap. 12), the refraction index is (for μ ≈ 1)

√ √  P
n(ω) = μ ≈  = 1 + χe = 1+ (13.21)
0 E
  n  2  1/2
N e2  ω0,k − ω 2 + iγk ω
= 1+ fk  2 .
0 m e k=1 ω0,k
2
− ω 2 + (γk ω)2

Equation (13.21) indicates that refractive index is a complex number, and thus there
is an attenuation of the wave occurring in a dispersive media, which is more important
close to the resonance frequencies (ωk ≈ ω0 ), as described in the following.
First, considering that the second term in Eq. (13.21) is small, we can retain only
the second-order terms, and thus:
364 13 Electromagnetic Waves in Dispersive Media
 n  2 
N e2  ω0,k − ω 2 + iγk ω
n(ω) ≈ 1 + fk  2 (13.22)
20 m e k=1 ω0,k
2
− ω 2 + (γk ω)2
  2 
N e2 
n
f k ω0,k − ω2
= 1+  
20 m e k=1 ω 2 − ω 2 2 + (γk ω)2
0,k
  
nr (ω)≡Re(n(ω))

N e2 
n
f k γk ω
+i  2 .
20 m e k=1 ω0,k
2
− ω 2 + (γk ω)2
  
n i (ω)≡Im(n(ω))

Then, the wave number is complex number, and the real part kr (ω) is
  2 
N e2 
n
ωn r (ω) ω f k ω0,k − ω2
kr (ω) = = 1+   . (13.23)
c c 20 m e k=1 ω 2 − ω 2 2 + (γk ω)2
0,k

Here, kr (ω) is the so-called dispersion relation for the medium. Using Eq. (13.23), the
phase velocity v p (ω) can be determined for a given frequency of the electromagnetic
wave ω: ω
v p (ω) = . (13.24)
kr (ω)

Figure 13.2 shows the real part n r (ω) and imaginary part n i (ω) of the index of
refraction as a function of frequency ω for illustration.
The absorption coefficient α(ω) is determined from the imaginary part of k(ω) as
follows:

Fig. 13.2 The illustration of the real part nr (ω) and imaginary part n i (ω) of the index of refraction
as a function of frequency ω for n = 2
13.1 Dispersion and Absorption 365

N e2 
n
ωn i (ω) f k γk ω 2
α(ω) = 2ki (ω) = 2 =  2  . (13.25)
c 0 m e c k=1 ω − ω 2 2 + (γk ω)2
0,k

Close to the resonance frequencies, ω0,k ≈ ω, and so ω0,k + ω ≈ 2ω; therefore, the
absorption coefficient can be approximated as

N e2 
n
f k γk ω 2
α(ω) =  2  (13.26)
0 m e c k=1 ω − ω 2 2 + (γk ω)2
0,k

N e2 
n
f k γk ω 2
=  
0 m e c k=1 (ω0,k − ω)(ω0,k + ω) 2 + (γk ω)2
N e2 
n
f k γk
≈   .
0 m e c k=1 4 ω0,k − ω 2 + γ 2
k

Replacing the Lorentzian function, given as

1 (γk /2)
L(ω, ωk , γk ) =   , (13.27)
π ω0,k − ω 2 + (γk /2)2

Equation (13.26) can also be written as

π N e2 
n
α(ω) = f k L(ω, ωk , γk ). (13.28)
0 m e c k=1

The Lorentz’s function has its full width at half maximum (FWHM) of γk . It is
normalized to one, and it gives the shape of absorption and emission lines, applied
to driven resonant systems. The heavy damped oscillators have broad line widths
and respond to a wider range of driving frequencies around the resonant frequency
ωk . The width of the spectral line depends on the quality factor η, which (see also
Chap. 10) is a measure of the sharpness of the resonance and depends on γk :
ωk ωk
η= = . (13.29)
Δω γk

13.2 Dispersion

The phase velocity is determined by Eq. (13.24), which can also be written as
c
v p (ω) = . (13.30)
Re (n(ω))
366 13 Electromagnetic Waves in Dispersive Media

Equation (13.29) indicates that the refraction index is a function of frequency n(ω),
and thus there is a dispersion. Because of the dispersion, different colors present
in a white light separate when passed through a glass prism. Besides, due to the
dispersion, the wave packets travel with a velocity called the group velocity.

13.2.1 Wave Packets and Group Velocity

A wave packet is a superposition of a group of waves traveling together and localized

in space at some initial time. Consider the general expression of a planar wave solution

ψ(r, t) = ψ0 exp (i (k(ω) · r − ωt)) . (13.31)

Let us consider a superposition of such plane waves as follows:


ψ(r, t) = dk ψ0 (k) exp (i (k · r − ω(k)t)) , (13.32)

where ψ0 (k) is some arbitrary function of k, which is chosen such that ψ(r, t =
0) = 0 (i.e., it is localized in a finite region in space). This kind of configurations in
space and time are also called wave packets. There are various choices for the spatially
localized envelope functions; however, a convenient choice for the weighting function
ψ0 is the Gaussian shape function as follows:
2

2ξ
ψ0 (k) = ψ0 exp −(k − k0 ) , (13.33)
4

where ψ0 is an amplitude coefficient, ξ is a coefficient that has the units of lengths

determining the spatial extent of the wave package at t = 0. For simplicity, we
consider a wave packet traveling in the positive z-axis direction; that is, k = k ẑ, and
k · r = kz. Therefore, the spatial profile ψ(z, 0) of the wave package at the initial
time is

2 ξ2
ψ(z, 0) = ψ0 dk e−(k−k0 ) 4 +ikz (13.34)

2 ξ2
= ψ0 ds e−s 4 ei(s+k0 )z (s ≡ k − k0 )

2 ξ2
= ψ0 e ik0 z
dk e−k 4 eikz (k ≡ s)

2 ξ2
= ψ0 eik0 z dk e−k 4 e−(z/ξ)
2

√
π
ψ0 ξeik0 z e−(z/ξ) ,
2
=
2
13.2 Dispersion 367

where the Gaussian integral is used

 +∞

−αk 2 π
dk e = . (13.35)
0 α

Thus, ψ(z, 0) is concentrated in a volume side ξ, and it describes the profile of a small
wave packet at initial time t = 0. Next, we discuss the time evolution of the plane
wave ψ(z, t). For that, we assume that ω(k) varies slowly with k compared with the
extension of the wave package in the wave number space k, which is considered the
Fourier space of the real space z. Thus, we can expand ω(k) around k0 as follows:

dω 1 d 2ω
ω(k) = ω(k0 ) + (k − k0 ) + (k − k0 )2 + ··· . (13.36)
dk k0 2 dk 2 k0

Denoting ω0 ≡ ω(k0 ), vg ≡ (dω/dk)k0 , and β ≡ (d 2 ω/dk 2 )k0 , we obtain

β
ω(k) ≈ ω0 + (k − k0 )vg + (k − k0 )2 . (13.37)
2
Therefore,

dk e−(k−k0 ) ξ /4 ei(kz−ω(k)t)
2 2
ψ(z, t) = ψ0 (13.38)

≈ ψ0 eik0 (z−v p t) dk e−(ξ +2iβt)(k−k0 ) /4 ei(k−k0 )(z−vg t)
2 2

√
π
ψ0 ξ(t)eik0 (z−v p t) e−((z−vg t)/ξ(t)) ,
2
=
2

where ξ(t) = ξ 2 + 2iβt, and v p = ω0 /k0 (phase velocity at ω0 ). These finding
indicate that the center of the wave package moves with group velocity vg to the
right along the z-axis:

dω
vg = , (13.39)
dk k0

which is the group velocity at k0 (or, at ω0 ). vg determines the effective center velocity
of a superposition of a continuum or group of plane waves. In vacuum where c(ω) = c
is constant and hence independent of frequency ω, we have ω = ck, and thus vg = c,
which equals the speed of light in vacuum.
Note that the group velocity is physical velocity; that is, it describes the speed of
propagation of the physical information. On the other hand, the phase velocity v p is
characteristic of an individual plane wave:

ω(k)   
exp (i (kz − ω(k)t)) = exp ik z − t = exp ik z − v p t . (13.40)
k
368 13 Electromagnetic Waves in Dispersive Media

Thus, the phase velocity v p is the velocity of the traveling of the wave. However, the
wave extends over the entire system, and hence the phase velocity does not represent
a physical information. Furthermore, from the definition v p = c/n(ω), and since
n > 1, it implies that v p < c; however, the phase velocity can exceed the speed of
light, and this is not in conflict with the Einstein’s special theory of relativity postulate
about the speed of light because the dynamics of a uniform wave train is not related
to the transport of energy of any other form of physical information.
The group velocity, on the other hand, is such that

dω(k) dk(ω) −1
vg = = (13.41)
dk dω

−1

dωn(ω) dn(ω) −1
= c = n(ω) + ω c
dω dω
vp
= .
ω dn
1+
n(ω) dω

Close to the normal dispersion region, dn/dω > 0, which implies that vg < v p ; for
the anomalous cases, however, dn/dω < 0, and thus vg > v p or even the speed of
light in vacuum, c.2

13.2.2 Normal and Anomalous Dispersion

Note that the energy is transported with the wave packet at the group velocity vg .
Besides, since vg < c, then the Einstein’s special theory of relativity is satisfied. That
is also called normal dispersion.
From Eq. (13.41), we have
c
vg = . (13.42)
dRe (n(ω))
Re (n(ω)) + ω
dω

Equation (13.42) indicates that normal dispersion occurs for dRe(n(ω))/dω > 0.
However, for dRe(n(ω))/dω < 0, the group speed vg can exceed the speed of light
in vacuum c, and this is the so-called anomalous dispersion. Note that the anoma-
lous dispersion is related to the absorption, and hence the group velocity does not
characterize any longer physical quantity; in that case, we can say that it does not
represent the speed of transferring energy of information. Therefore, the Einstein’s
special theory of relativity is not violated.
The dispersive spreading of the electromagnetic wave packets is an important
phenomenon, such as the dispersive deformation is one of the main factors that

2 Jackson (1999).
13.2 Dispersion 369

limits the information load transported through the fiber optical cables; if the load
is too high, the wave packets constituting the bit stream through the fibers begin to
overload and thus loosing the identity. Construction of more sophisticated counter
measures optimizing the data capacity of optical fibers represents an important area
of research.

13.3 Refractive Index of a Conductor

If an electromagnetic wave propagates in a conductor, the current density is J = σE

(where σ is the conductivity). Therefore, when deriving the wave equations from
Maxwell’s equations in a medium, the current density appears in these equations.
Thus, taking the curl of the second Maxwell’s equations in a conductor,

∂
∇ × (∇ × E) = − (μ0 μ∇ × H) . (13.43)
∂t

Using the fourth Maxwell’s equation for a conductor (∇ × H = σE + ∂D/∂t), we

have

∂E ∂2E
− ∇ 2 E = − μm σ + m μm 2 , (13.44)
∂t ∂t

where D = m E. Equation (13.44) can be re-arranged in the following form:

∂E ∂2E
∇ 2 E − μm σ − m μm 2 = 0. (13.45)
∂t ∂t

Similarly, using the curl of the fourth Maxwell’s equation for Δ × B, we can obtain
the wave equation for the magnetic field as follows:

∂B ∂2B
∇ 2 B − μm σ − m μm 2 = 0. (13.46)
∂t ∂t
Consider the monochromatic plane wave solutions, representing waves propagat-
ing to the right, as

E(r, t) = E0 exp (i (k · r − ωt)) , (13.47)

B(r, t) = B0 exp (i (k · r − ωt)) .

Substituting expression for E from Eq. (13.47) into (13.45), we obtain

 
(ik)2 − μm σ(−iω) − μm m (iω)2 E = 0. (13.48)
370 13 Electromagnetic Waves in Dispersive Media

From Eq. (13.48), we obtain the dispersion equation for the electromagnetic waves
in a conductor as follows:

k(ω) = iμm σω + μm m ω 2 . (13.49)

If we consider a good conductor such that σ

m ω, then from Eq. (13.49):
√√ √ √
k(ω) ≈ i μm σω = eiπ/2 μm σω (13.50)
√
= eiπ/4 μm σω
1+i√
= √ μm σω.
2

Equation (13.50) indicates that wave number k is complex number with real and
imaginary part as follows:

1 √
kr (ω) ≡ Re(k(ω)) = √ μm σω, (13.51)
2
1 √
ki (ω) ≡ Im(k(ω)) = √ μm σω.
2

The time average of the magnitude of the electric field vector is

   
E(r, t) ∝ exp −ki k̂ · r ≡ exp −k̂ · r/δ(ω) , (13.52)

where δ(ω) is the so-called skin depth defined as


2
δ(ω) = . (13.53)
μm σω

Therefore, when the electromagnetic waves propagate through a conductor, the inten-
sity decreases as follows:
 
I = Imax exp −2k̂ · r/δ(ω) , (13.54)

where the absorption coefficient is given as

2
α(ω) = . (13.55)
δ(ω)

Figure 13.3 presents a log plot of the skin depth versus linear frequency for
soft iron (σ = 11.2 × 106 Ω −1 m −1 ), silver (σ = 62.1 × 106 Ω −1 m −1 ), lead (σ =
4.7 × 106 Ω −1 m −1 ), and seawater at 20 ◦ C and a salt concentration of 3 % (σ =
4.1 Ω −1 m −1 ), and μm = 2 × 10−5 . At high electromagnetic wave frequencies, the
13.3 Refractive Index of a Conductor 371

Fig. 13.3 Log plot of the skin depth versus linear frequency for soft iron (σ = 11.2 ×
106 Ω −1 m−1 ), silver (σ = 62.1 × 106 Ω −1 m−1 ), lead (σ = 4.7 × 106 Ω −1 m−1 ), and seawater at
20 ◦ C and a salt concentration of 3 % (σ = 4.1 Ω −1 m−1 ). μm = 2 × 10−5

waves can only penetrate small distances inside the conductor. For example, in sea-
water the skin depth is about δ = 30 cm at f = 1 MHz and much longer of δ = 10
m at f = 1 kHz for a temperature of 20 ◦ C and salt concentration of 3 %. That is
why, there is a communication problem of radio waves at frequency range 3–30 kHz.
In the case of a conductor at high-frequency oscillating currents, magnetic fields
are created that induce electric fields opposing the change in magnetic field; these
induced electric fields produce the so-called eddy currents, and hence a resistive
power loss. Furthermore, the induced electric fields are stronger at the center of the
conductor, producing a high resistance at the center of the conductor, and thus the
conduction current is confined to a thin outer layer with thickness the skin depth.
That is equivalent to using a hollow wire, which is lighter than a conducting wire.
The reflectance of a good conductor is high; therefore, the parabolic or spherical
glass surfaces of a reflecting telescope are coated with a thin silver or aluminum,
typically 0.1 µm. Also, the index of refraction of a conductor is complex number. The
amplitude reflection coefficient for the external reflection (for example, air/metal) at
the normal incidence at θ1 = 0◦ (for both perpendicular and parallel polarization),
and obtain a reflectance as follows:

(1 − n r (ω))2 + (n i (ω))2
r⊥, (θ1 = 0◦ ) = . (13.56)
(1 + n r (ω))2 + (n i (ω))2

Using expressions given by Eq. (13.51), we derive the index of the refraction (both
the real and imaginary parts) as follows:
372 13 Electromagnetic Waves in Dispersive Media

kr (ω)c μm σ
n r (ω) = =c , (13.57)
ω 2ω

ki (ω)c μm σ
n i (ω) = =c .
ω 2ω

Knowing the conductivity σ (or equivalently, the resistivity ρ = 1/σ) and μm , one
can calculate the reflectance of a conductor depending on the frequency ω for normal
incidence waves from air to metal.

13.4 Wave Propagation in a Dilute Plasma

Consider the wave propagation in an ionized plasma consisting of n e electrons and n i

ions per unit of volume. Besides, we will ignore the distant ions as they will provide
just a little damping to the electrons motion, and because of their large mass, their
contribution to the current density can be omitted.
Applying the second law of Newton for the motion of an individual electron, we
have
dv
me = −eE(t), (13.58)
dt

where E(t) is the time-varying external electric field of an electromagnetic wave.

The current density arriving from all bulk electrons in the plasma is

J(t) = n e (−e)v(t). (13.59)

The time derivative of the current density is

dJ(t) n e e2
= E(t). (13.60)
dt me

Consider an oscillating electric field such that E(t) = E0 e−iωt , then from Eq. (13.60),
we obtain
dJ(t) n e e2
= E0 e−iωt . (13.61)
dt me

Or,
13.4 Wave Propagation in a Dilute Plasma 373

n e e2
J(t) = E0 dt e−iωt (13.62)
me

n e e2
= E0 d(−iωt) e−iωt
−iωm e
n e e2
=i E.
ωm e

Equation (13.62) implies that conductivity is

n e e2
σ=i , (13.63)
ωm e

which is purely imaginary quantity. Using Eq. (13.62) and i = eiπ/2 , we write

n e e2
J(t) = E0 e−i(ωt−π/2) . (13.64)
ωm e

Therefore, the current density and the electric field have a phase shift of π/2, or
90◦ . If we calculate the average power, then it equals to zero. This is analogue to the
LC circuit discussed in Chap. 10 where the energy was continuously exchanging
between the electric field energy of the capacitor C and magnetic field energy of the
inductor L.

13.4.1 Electromagnetic Waves in a Dilute Plasma

Let us consider a dilute (i.e., low density) plasma and neural (i.e., the charge density
is ρ = 0). We may also assume that μm = μ0 and m = 0 (i.e., the vacuum). Then,
the wave equation for the electric field vector E can be obtained by taking the curl
of the second Maxwell’s equation (Faraday’s law) as usually

∂
∇ × (∇ × E) = − (∇ × B) , (13.65)
∂t

∂ ∂E
−∇ E=−2
μ0 J + μ0 0 , (13.66)
∂t ∂t
2

∂J ∂ E
∇ 2 E = μ0 + μ0 0 , (13.67)
∂t ∂t 2
2

n e e2 ∂ E
∇ 2 E = μ0 E + μ0 0 , (13.68)
me ∂t 2

where Eq. (13.61) is used. Therefore, the wave equation for E becomes
374 13 Electromagnetic Waves in Dispersive Media

n e e2 1 ∂2E
∇ E − μ0
2
E− 2 = 0. (13.69)
me c ∂t 2

Substituting the monochromatic electromagnetic plane wave solution, E(r, t) =

E0 exp (i (k · r − ωt)) into Eq. (13.69), we get

n e e2 1
(ik)2 E − μ0 E− (−iω)2 E = 0 (13.70)
me c2
or,

n e e2 ω2
−k − 2
2
+ 2 E = 0. (13.71)
c 0 m e c

Therefore, the following dispersion relation is obtained:

k 2 c2 = ω 2 − ω 2p , (13.72)

where ω p is the so-called plasma frequency:


n e e2
ωp = . (13.73)
0 m e

Equation (13.72) implies that for ω > ω p , k is purely real number, and hence the
electromagnetic wave propagates in the plasma, whereas for ω < ω p , wave number
vector k is purely imaginary, the electromagnetic waves can not propagate in plasma.

13.4.2 Phase and Group Velocity in a Dilute Plasma

Using Eq. (13.72), the index of refraction n, group velocity vg and phase velocity v p
are given as follows:

kc ω 2p
n(ω) = = 1− 2, (13.74)
ω ω


−1
dk ω 2p
vg (ω) = = c 1− 2, (13.75)
dω ω
ω c
v p (ω) = =  . (13.76)
k ω 2p
1− 2
ω

From Eqs. (13.75) and (13.76),

13.4 Wave Propagation in a Dilute Plasma 375

vg v p = c2 . (13.77)

Thus,
c vg
n(ω) ≡ = . (13.78)
vp c

Therefore, for ω < ω p , there is no electromagnetic wave propagation in plasma,

which is also called cutoff frequency of a dilute plasma. For ω > ω p the phase velocity
magnitude exceeds the speed of light in vacuum c; however, vg < c.

13.4.3 Plasma and Dielectric at High Frequency

Consider ω
ω p , then the index of refraction n(ω) of a dilute plasma from
Eq. (13.74) becomes 
ω 2p ω 2p
n(ω) = 1 − 2 ≈ 1 − . (13.79)
ω 2ω 2

On the other hand, using Eq. (13.22), the real part of the index of refraction for a
dielectric using the Lorentz’s oscillator model is

N e2 
n
ω0,k
2
− ω2
n r (ω) ≈ 1 + fk 2 (13.80)
20 m e k=1 (ω0,k − ω 2 )2 + (γk ω)2
N e2  f k
n
≈1− .
20 m e k=1 ω 2

Therefore,
ω 2p
n r (ω) ≈ 1 − , (13.81)
2ω 2
where 
N  e2
ωp = , (13.82)
0 m e

and N  is the total number density of electron oscillators:


n
N = N fk . (13.83)
k=1

Comparing Eq. (13.79) with (13.83), we find that the index of refraction for the
dielectric and plasma at high frequency are identical.
376 13 Electromagnetic Waves in Dispersive Media

13.5 Exercises

Exercise 13.1 Show that the time-averaged power density E · J in a dilute

plasma is zero.

Solution 13.1 Consider E(t) = E0 e−iωt , then using Eq. (13.62), we have

n e e2 2 n e e2 2 −i(2ωt−π/2)
E·J=i E = E e , (13.84)
ωm e ωm e 0

where i = eπ/2 is used. Thus,

n e e2 2
E·J= E (cos(2ωt − π/2) − i sin(2ωt − π/2)) (13.85)
ωm e 0
n e e2 2
= E (sin(2ωt) + i cos(2ωt)) .
ωm e 0

The time average of the following functions are

  T
1 T 1
sin(2ωt) = sin(2ωt)dt = sin(2ωt)d(2ωt) = 0, (13.86)
T 0 2T ω 0
  T
1 T 1
cos(2ωt) = cos(2ωt)dt = cos(2ωt)d(2ωt) = 0. (13.87)
T 0 2T ω 0

Therefore,
E · J = 0. (13.88)

Exercise 13.2 The resistivity of silver is ρ = 1.6 × 10−8 Ωm, and its perme-
ability is μm = 0.9998μ0 , find the reflectance for light with wavelength 500 nm
at normal incidence from air.

Solution 13.2 Using Eq. (13.56) for the reflectance:

(1 − n r (ω))2 + (n i (ω))2
r⊥, (θ1 = 0◦ ) = . (13.89)
(1 + n r (ω))2 + (n i (ω))2

For both n r and n i , we can use Eq. (13.57) as follows:

13.5 Exercises 377

 
kr c c μm σω c2 μm
nr = n i = = = . (13.90)
ω ω 2 2ωρ

μm = 4π × 10−7 (T· m/A)0.9998 ≈ 12.56 × 10−7 (T· m/A)

Furthermore,
2π 2πc
ω= = ,
T λ
and thus
 
cλμm (3 × 108 )(500 × 10−9 )(12.56 × 10−7 )
nr = n i = = ≈ 30.6.
4πρ 4π(1.6 × 10−8 )

Therefore,
r⊥, = 0.937.

Exercise 13.3 The density of electrons in a dilute plasma is n e = 1.0 × 1020

electrons/m3 . Determine the plasma frequency ω p .

Solution 13.3 Using Eq. (13.78), m e = 9.11 × 10−31 kg, e = 1.6 × 10−19 C, then
the plasma frequency is

1/2
n e e2
ωp = (13.91)
0 m e

1/2
(1020 )(1.6 × 10−19 )2
=
(8.8542 × 10−12 )(9.11 × 10−31 )
≈ 0.563 × 1012 Hz.

Exercise 13.4 Determine the absorption coefficient of seawater knowing that

at low frequency, seawater has an electrical conductivity σ = 4.0 Ω −1 m−1 .
The frequency is f = 108 Hz.

Solution 13.4 Using Eqs. (13.53) and (13.55) for μ ≈ 1, the absorption coefficient
of seawater is
378 13 Electromagnetic Waves in Dispersive Media

α= 4πμ0 σ f (13.92)

= 4π(4π × 10−7 )(4.0)(108 )
≈ 79.5 m−1 .

Exercise 13.5 Determine the skin length of seawater knowing that at low
frequency, seawater has an electrical conductivity σ = 4.0 Ω −1 m−1 . The fre-
quency is f = 100 Hz.

Solution 13.5 Using Eqs. (13.53) and (13.55) for μ ≈ 1, the absorption coefficient
of seawater is

α= 4πμ0 σ f (13.93)

= 4π(4π × 10−7 )(4.0)(100)
≈ 0.079 m−1 .

Therefore, the skin length is

1
δ= ≈ 12.6 m.
α

References

Altland A, Simons B (2010) Condensed matter field theory, 2nd edn. Cambridge University Press
Griffiths DJ (1999) Introduction to electrodynamics, 3rd edn. Prentice Hall
Jackson JD (1999) Classical electrodynamics, 3rd edn. John Wiley and Sons
Landau LD, Lifshitz EM (1971) The classical theory of fields. Pergamon Press
Protheroe RJ (2013) Essential electrodynamics, 1st edn. Bookboon
Sykja H (2006) Bazat e Elektrodinamikës. SHBUT
Appendix
Vectorial Analysis

Some mathematical background on vector analysis, useful to

electromagnetism.

In this appendix, we discuss some applications of vectorial calculus and vector differ-
ential operators to electromagnetism. In addition, we present some useful expressions
of the gradient operator acting on a scalar and vector field in Cartesian, spherical,
and cylindrical coordinates. Moreover, Laplacian expressions in Cartesian, spherical,
and cylindrical coordinates are provided.

A.1 Vector Calculus

Consider A a vector, which is a function of a scalar quantity s. We can express A in

terms of its components as

A(s) = A x (s)i + A y (s)j + A z (s)k. (A.1)

Definition A.1 (Derivative) The derivative of A for the variable s is defined as

dA(s) A(s + Δs) − A(s)

= lim . (A.2)
ds Δs→0 Δs

In terms of the components of A, we can write

dA(s) d Ax d Ay d Az
= i+ j+ k. (A.3)
ds ds ds ds

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer 379
Nature Switzerland AG 2022
H. Kamberaj, Electromagnetism, Undergraduate Texts in Physics,
https://doi.org/10.1007/978-3-030-96780-2
380 Appendix: Vectorial Analysis

Fig. A.1 Graphical

interpretation of the
derivative

A graphical interpretation of derivative of vector A is shown in Fig. A.1. As it can

be seen that as Δs → 0, the vector difference ∇A = A(s + Δs) − A(s) approaches
along the line tangent to the curve A(s) at point s. Therefore, dA/ds is a vector, and
it is tangent to the curve.
The magnitude of the vector dA/ds is given as
 2  2  2
dA(s) dA d Ax d Ay d Az
| |≡ = + + . (A.4)
ds ds ds ds ds

The following rules hold for the derivative:

d dA dB
(A(s) ± B(s)) = ± , (A.5)
ds ds ds
d df dA
( f (s)A(s)) = A(s) + f (s) ,
ds ds ds
d dA dB
(A(s) · B(s)) = · B(s) + A · ,
ds ds ds
d dA dB
(A(s) × B(s)) = × B(s) + A × .
ds ds ds
Note that in the last expression of Eq. (A.5) the order of the vectors A and B matters.
In Eq. (A.5), f (s) is a scalar function of s.
Usually, in three-dimensional space, we have to distinguish two types of functions
or fields. Namely, the scalar point functions of the form:

f (r) ≡ f (x, y, z) (A.6)

and the vector point functions

 
A(r) ≡ A(x, y, z) = A x (x, y, z), A y (x, y, z), A z (x, y, z) . (A.7)
Appendix: Vectorial Analysis 381

Fig. A.2 A line L in

three-dimensional space, and
a vector A defined at every
point along the curve L

For example, a scalar function is the electric potential, φ(r) = φ(x, y, z), and a
vector function is electric field vector E(r) = E(x, y, z). These functions can also
be explicit function of some other scalar, such as s, which is often time t.
Consider a line L in three-dimensional space, and a vector A defined at every
point along the curve L, as shown in Fig. A.2.

Definition A.2 (Line integral) The line integral of A along the curve L is defined as

A(r) · ds . (A.8)
L

If we consider the displacement vector dr such that | dr |= ds, then the line integral
can also be written as
  
A(r) · ds = A(r) · dr = A(r) cos θ ds . (A.9)
L L L

Furthermore, if r = xi + yj + zk, then dr = d xi + dyj + dzk, and hence, from

Eq. (A.9), we can write
 
 
A(r) · ds = A x d x + A y dy + A z dz (A.10)
L L
  
= Ax d x + A y dy + A z dz .
L L L

If we denote ds a distance measured along the line, which parametrizes the line, then
Eq. (A.9) can also be written as
382 Appendix: Vectorial Analysis
   
dr
A(r) · dr = A· ds (A.11)
ds
L L
  
dx dy dz
= Ax + Ay + Az ds .
ds ds ds
L

A.2 Vector Differential Operators

Definition A.3 (Gradient operator) The vector gradient operator is denoted by grad
or ∇, and it is not a vector but it is a vector operator, which in Cartesian coordinates
is defined as
∂ ∂ ∂
∇=i +j +k (A.12)
∂x ∂y ∂z
 
∂ ∂ ∂
= , , .
∂ x ∂ y ∂z

When this operator acts on a scalar function φ, the result is a vector; for example, let
φ(x, y, z) be a scalar function of (x, y, z), then ∇φ is a vector given as

∂φ ∂φ ∂φ
∇φ = i +j +k (A.13)
∂x ∂y ∂z
 
∂φ ∂φ ∂φ
= , , .
∂ x ∂ y ∂z

A practical aspect of the gradient is as the following. Suppose we want to calculate a

small change on the scalar function φ due to a change on the variables dr. For that,
using Eq. (A.13), we write

dφ = ∇φ · dr =| ∇φ || dr | cos θ , (A.14)

where θ is the angle between ∇φ and dr.

When this operator acts on a vector function A, it performs a scalar product and
hence the result is a scalar; for example, let A(x, y, z) be a vector function of (x, y, z),
then ∇ · A is a scalar given as dot product:

∂ ∂ ∂
∇·A= (i · A) + (j · A) + (k · A) (A.15)
∂x ∂y ∂z
∂ Ax ∂ Ay ∂ Az
= + + ,
∂x ∂y ∂z

which is also called divergence of A, and it is denoted by divA.

Appendix: Vectorial Analysis 383

Moreover, when the gradient operator performs a cross-product with another vec-
tor function, it gives as a result the curl of A or the rot (rotation) of A, denoted by
∇ × A or curlA, or rotA. This is a vector, and it is defined as
⎡ ⎤
i j k
⎢ ∂ ∂ ∂ ⎥
∇×A=⎢ ⎣ ∂ x ∂ y ∂z ⎦
⎥ (A.16)
Ax A y Az
⎛⎡ ⎤⎞ ⎛⎡ ⎤⎞ ⎛⎡ ⎤⎞
∂ ∂ ∂ ∂ ∂ ∂
= det ⎝⎣ ∂ y ∂z ⎦⎠ i − det ⎝⎣ ∂ x ∂z ⎦⎠ j + det ⎝⎣ ∂ x ∂ y ⎦⎠ k
A y Az Ax Az Ax A y
     
∂ Az ∂ Ay ∂ Ax ∂ Az ∂ Ay ∂ Ax
= − i+ − j+ − k.
∂y ∂z ∂z ∂x ∂x ∂y

A.3 Stokes’ Formula

Consider the vector field lines J passing through the cross-sectional surface area A
enclosed by the closed line L, as shown in Fig. A.3. Here, the surface area vector
A = An, where n is a unit vector perpendicular to surface area. Then, according to
Stokes’ formula, the line integral of F along the curve L is
 
F · ds = (∇ × F) · dA . (A.17)
L A

In Eq. (A.17), the left-hand side gives the line integral of the vector field F, and the
right-hand side gives the flux of the vector J ≡ ∇ × F (the rot of F) through the
surface area enclosed by the line L.

Fig. A.3 Vector field lines J

passing through the
cross-sectional surface area
A enclosed by the closed line
L
384 Appendix: Vectorial Analysis

Fig. A.4 Vector field lines F

passing through the closed
surface area A enclosing the
volume V . d A is a small
surface element and n is an
outward unit vector normal
to the surface element

A.4 Gauss’s Formula

Consider the vector field lines F passing through the closed surface area A enclosing
the volume V , as shown in Fig. A.4. d A is a small surface element and n is an outward
unit vector normal to the surface element. The vector F is tangent to the line at every
point of the line. Then, according to Gauss’s formula, the closed surface integral of
F along the surface A is
 
F · dA = (∇ · F) d V , (A.18)
A V

where dA = d An. In Eq. (A.18), the left-hand side gives the net number of lines
leaving the surface A and the right-hand side gives the total amount of source creating
the field F (e.g., charge) inside the volume V (Fig. A.4).

A.5 Some Useful Formula

Let φ and A be a scalar field and vector field, respectively. Then,

∂(φ A x ) ∂(φ A y ) ∂(φ A z )

∇ · (φA) = + + (A.19)
∂x ∂y ∂z
∂φ ∂ Ax ∂φ ∂ Ay ∂φ ∂ Az
= Ax + φ + Ay + φ + Az + φ
∂x ∂x ∂y ∂y ∂z ∂z
= (∇φ) · A + φ(∇ · A) .
Appendix: Vectorial Analysis 385

Furthermore, the following formula can be proved:

 
∂(φ A z ) ∂(φ A y )
∇ × (φA) = − i (A.20)
∂y ∂z
 
∂(φ A x ) ∂(φ A z )
+ − j
∂z ∂x
 
∂(φ A y ) ∂(φ A x )
+ − k
∂x ∂y
 
∂φ ∂ Az ∂φ ∂ Ay
= Az + φ − Ay − φ i
∂y ∂y ∂z ∂z
 
∂φ ∂ Ax ∂φ ∂ Az
+ Ax + φ − Az − φ j
∂z ∂z ∂x ∂x
 
∂φ ∂ Ay ∂φ ∂ Ax
+ Ay + φ − Ax − φ k
∂x ∂x ∂y ∂y
= (∇φ) × A + φ(∇ × A) .

Moreover, since i, j, and k are unit and constant vectors, then

∇ ·i = ∇ ·j = ∇ ·k = 0, (A.21)
∇ ×i = ∇ ×j = ∇ ×k = 0.

It is easy to show that

∂x ∂y ∂z
∇·r = + + =3 (A.22)
∂x ∂y ∂z

and
⎡ ⎤
i j k
⎢∂ ∂ ∂ ⎥
∇×r =⎢⎣ ∂ x ∂ y ∂z ⎦
⎥ (A.23)
x y z
⎛⎡ ⎤⎞ ⎛⎡ ⎤⎞ ⎛⎡ ⎤⎞
∂ ∂ ∂ ∂ ∂ ∂
= det ⎝⎣ ∂ y ∂z ⎦⎠ i − det ⎝⎣ ∂ x ∂z ⎦⎠ j + det ⎝⎣ ∂ x ∂ y ⎦⎠ k
y z x z x y
     
∂z ∂y ∂x ∂z ∂y ∂x
= − i+ − j+ − k = 0.
∂y ∂z ∂z ∂x ∂x ∂y

If r̂ is a unit vector along the direction r such that r̂ = r/r , then using Eq. (A.19)
with φ = 1/r and A = r, we get
386 Appendix: Vectorial Analysis
r
∇ · r̂ = ∇ · (A.24)
 r
1 1
=∇ · r + (∇ · r)
r r
      
∂ 1 ∂ 1 ∂ 1 3
= i +j +k ·r+
∂x r ∂y r ∂z r r
r·r 3
=− 3 +
r r
2
= ,
r
and
∇ × r̂ = 0 . (A.25)

A.6 Laplacian

Laplacian formulas can be obtained by applying the vector gradient operator twice.
Definition A.4 (Laplacian) Let φ be a scalar field function and A a vector field. The
Laplacian are defined as follows:

Δφ = ∇ · (∇φ) (A.26)
ΔA = ∇(∇ · A) − ∇ × ∇ × A .

In Eq. (A.26), Δφ is called scalar Laplacian and ΔA is called vector Laplacian. It

can be seen that the scalar Laplacian is given in Cartesian coordinates as

∂ 2φ ∂ 2φ ∂ 2φ
Δφ = + 2 + 2 . (A.27)
∂x 2 ∂y ∂z

Furthermore,
    
∂ ∂ Ay ∂ Ax ∂ ∂ Az ∂ Ax
∇×∇×A=i − + − (A.28)
∂y ∂x ∂y ∂z ∂ x ∂z
    
∂ ∂ Ax ∂ Ay ∂ ∂ Az ∂ Ay
+j − + −
∂x ∂y ∂x ∂z ∂ y ∂z
    
∂ ∂ Ax ∂ Az ∂ ∂ Ay ∂ Az
+k − + − ,
∂x ∂z ∂x ∂ y ∂z ∂y

and
Appendix: Vectorial Analysis 387
 
∂ ∂ Ax ∂ Ay ∂ Az
∇(∇ · A) = i + + (A.29)
∂x ∂x ∂y ∂z
 
∂ ∂ Ax ∂ Ay ∂ Az
+j + +
∂y ∂x ∂y ∂z
 
∂ ∂ Ax ∂ Ay ∂ Az
+k + + .
∂z ∂ x ∂y ∂z

Therefore, substituting Eqs. (A.28) and (A.29) into (A.26), we write that
 
∂2 ∂2 ∂2
ΔA = i + + Ax (A.30)
∂x2 ∂ y2 ∂z 2
 2 
∂ ∂2 ∂2
+j + + Ay
∂x2 ∂ y2 ∂z 2
 2 
∂ ∂2 ∂2
+k + + Az .
∂x2 ∂ y2 ∂z 2

A.7 Curvilinear Coordinates

In the following, we introduce two other coordinate systems, namely spherical coor-
dinate system (r, θ, φ) and cylindrical coordinate system (ρ, φ, z), as presented in
Fig. A.5. Furthermore, we introduce a metric for each geometry, defined as follows:

g ≡ (1, 1, 1) , Cartesian coordinate system (A.31)

g ≡ (1, r, r sin θ ) , Spherical coordinate system
g ≡ (1, ρ, 1) , Cylindrical coordinate system .

An interval length ds relates to the change on variables in these coordinate systems

in the following general form:


3
(ds)2 = gi2 (d xi )2 . (A.32)
i=1

Therefore,

(ds)2 = (d x)2 + (dy)2 + (dz)2 , Cartesian coordinate system, (A.33)

(ds) = (dr ) + r (dθ ) + r sin θ (dφ) , Spherical coordinate system,
2 2 2 2 2 2 2

(ds)2 = (dρ)2 + ρ 2 (dφ)2 + (dz)2 , Cylindrical coordinate system .

Moreover, the generals form of the vector gradient operator acting on a scalar field
ψ or vector field A are as follows:
388 Appendix: Vectorial Analysis

Fig. A.5 a A Cartesian coordinate system (x, y, z); b A spherical coordinate system (r, θ, φ); c
A cylindrical coordinate system(ρ, φ, z)


3
1 ∂ψ
∇ψ = x̂i , (A.34)
i=1
gi ∂ xi
 
1  ∂
3
g1 g2 g3
∇·A= Ai ,
g1 g2 g3 i=1 ∂ xi gi
 
1  ∂  
3
1 ∂
∇×A= x̂i i jk (gk Ak ) − gj Aj
2 i= j=k=1 g j gk ∂ x j ∂ xk

  
∂  
3
1 ∂
= x̂i i jk (gk Ak ) − gj Aj ,
i= j>k=1
g j gk ∂x j ∂ xk

where x̂i is a unit vector along the coordinate xi , and i jk is Levi-Civita number:
⎧
⎨ +1, if (i → j → k) rotates clockwise,
i jk = 0, if i = j, or i = k, or j = k, (A.35)
⎩
−1, if (i → j → k) rotates counterclockwise.

For example, for any scalar field and vector field functions, Eqs. (A.13) and (A.15)
give the expressions of the gradient and divergence in Cartesian coordinates. In
spherical coordinates, then for a scalar field ψ and vector field A, we find

∂ψ 1 ∂ψ 1 ∂ψ
∇ψ = r̂ + θ̂ + φ̂ , (A.36)
∂r r ∂θ r sin θ ∂φ

and
Appendix: Vectorial Analysis 389
    
1 ∂ g1 g2 g3 ∂ g1 g2 g3
∇·A= A1 + A2 (A.37)
g1 g2 g3 ∂ x1 g1 ∂ x2 g2
 
∂ g1 g2 g3
+ A3
∂ x3 g3
 
1 ∂ ∂ ∂
= (g2 g3 A1 ) + (g1 g3 A2 ) + (g1 g2 A3 )
g1 g2 g3 ∂ x1 ∂ x2 ∂ x3
 
1 ∂  2  ∂ ∂  
= sin θ r A r + r (sin θ A θ ) + r A φ
r 2 sin θ ∂r ∂θ ∂φ
1 ∂  2  1 ∂ 1 ∂ Aφ
= r Ar + (sin θ Aθ ) + .
r ∂r
2 r sin θ ∂θ r sin θ ∂φ

Furthermore, ∇ × A in spherical coordinates is given as

 
1 ∂ ∂
∇ × A = x̂1 123 (g3 A3 ) − (g2 A2 ) (A.38)
g2 g3 ∂ x2 ∂ x3
 
1 ∂ ∂
+ x̂2 213 (g3 A3 ) − (g1 A1 )
g1 g3 ∂ x1 ∂ x3
 
1 ∂ ∂
+ x̂3 312 (g2 A2 ) − (g1 A1 )
g1 g2 ∂ x1 ∂ x2
 
1 ∂   ∂ Aθ
= r̂ sin θ Aφ −
r sin θ ∂θ ∂φ
 
1 ∂ Ar ∂  
+ θ̂ − sin θ r Aφ
r sin θ ∂φ ∂r
 
1 ∂ ∂ Ar
+ φ̂ (r Aθ ) − .
r ∂r ∂θ

Laplacian of a scalar field ψ in spherical coordinates is calculated as

Δψ = ∇ · (∇ψ) (A.39)
 
∂ψ 1 ∂ψ 1 ∂ψ
= ∇ · r̂ + θ̂ + φ̂ (A.40)
∂r r ∂θ r sin θ ∂φ
     
∂ψ 1 ∂ψ 1 ∂ψ
= ∇ · r̂ + ∇ · θ̂ + ∇ · φ̂
∂r r ∂θ r sin θ ∂φ
        
T1 T2 T3

= T1 + T2 + T3 ,

where the terms T1 , T2 , and T3 are calculated using Eq. (A.34) as

390 Appendix: Vectorial Analysis
 
1 ∂ ∂ ∂
T1 = (g2 g3 A1 ) + (g1 g3 A2 ) + (g1 g2 A3 ) (A.41)
∂ x1
g1 g2 g3 ∂ x2 ∂ x3
⎡ ⎤
1 ⎢∂  2  ∂ ∂  ⎥
= 2 ⎣ r sin θ Ar + (r sin θ Aθ ) + r Aφ ⎦
r sin θ ∂r    ∂θ    ∂φ   
Ar =∂ψ/∂r Aθ =0 Aφ =0
 
1 ∂ ∂ψ
= 2 r 2 sin θ
r sin θ ∂r ∂r
2 ∂ψ ∂ψ 2
= + ,
r ∂r ∂r 2

 
1 ∂ ∂ ∂
T2 = (g2 g3 A1 ) + (g1 g3 A2 ) + (g1 g2 A3 ) (A.42)
g1 g2 g3 ∂ x1 ∂ x2 ∂ x3
⎡ ⎤
⎢ ⎥
⎢∂   ∂ ∂  ⎥
1 ⎢ ⎥
= 2 ⎢ r sin θ Ar + (r sin θ Aθ ) +
2
r Aφ ⎥
r sin θ ⎢ ∂r    ∂θ    ∂φ   ⎥
⎣ A =0 1 ∂ψ A =0
⎦
r φ
Aθ =
r ∂θ
 
1 ∂ ∂ψ
= 2 sin θ
r sin θ ∂θ ∂θ
1 ∂ψ 1 ∂ 2ψ
= 2 + 2 2 ,
r tan θ ∂θ r ∂θ
and
 
1 ∂ ∂ ∂
T3 = (g2 g3 A1 ) + (g1 g3 A2 ) + (g1 g2 A3 ) (A.43)
g1 g2 g3 ∂ x1 ∂ x2 ∂ x3
⎡ ⎤
⎢ ⎥
⎢ ⎥
⎢∂   ∂ ∂   ⎥
1 ⎢ ⎥
= 2 ⎢ r 2 sin θ Ar + (r sin θ Aθ ) + r Aφ ⎥
r sin θ ⎢ ∂r    ∂θ    ∂φ    ⎥
⎢ Aθ =0 ⎥
⎣ Ar =0 1 ∂ψ ⎦
Aφ =
r sin θ ∂φ
 
1 ∂ 1 ∂ψ
= 2
r sin θ ∂φ sin θ ∂φ
1 ∂ 2ψ
= 2 2 .
r sin θ ∂φ 2

Therefore,
Appendix: Vectorial Analysis 391

∂ψ 2 1 ∂ 2ψ 1 ∂ 2ψ 2 ∂ψ 1 ∂ψ
Δψ = + + + + 2 . (A.44)
∂r 2 r 2 ∂θ 2 r 2 sin θ ∂φ 2
2 r ∂r r tan θ ∂θ

For Laplacian of a vector field (the second expression in Eq. (A.26)), we first evaluate

∂
∇(∇ · A) = r̂ (∇ · A) (A.45)
 ∂r  
I1
1 ∂
+ θ̂ (∇ · A)
 r ∂θ 
I2
1 ∂
+ φ̂ (∇ · A)
r sin θ ∂φ
  
I3

= I1 + I2 + I3 ,

where ∇ · A is given by Eq. (A.37), and thus we have

  
r̂ ∂ 1 ∂  2  ∂ ∂ Aφ
I1 = sin θ r Ar + r (sin θ Aθ ) + r . (A.46)
sin θ ∂r r2 ∂r ∂θ ∂φ

Then, I2 is
  
θ̂ ∂ 1 ∂  2  ∂ ∂ Aφ
I2 = 3 sin θ r Ar + r (sin θ Aθ ) + r , (A.47)
r ∂θ sin θ ∂r ∂θ ∂φ

and I3 is
 
φ̂ ∂ ∂  2  ∂ ∂ Aφ
I3 = 3 2 sin θ r Ar + r (sin θ Aθ ) + r . (A.48)
r sin θ ∂φ ∂r ∂θ ∂φ

Furthermore, we calculate
 
1 ∂   ∂
∇ × ∇ × A = r̂ r sin θ (∇ × A) φ − (r (∇ × A) θ ) (A.49)
r 2 sin θ ∂θ ∂φ
 
1 ∂ ∂  
+ θ̂ ((∇ × A)r ) − r sin θ (∇ × A)φ
r sin θ ∂φ ∂r
 
1 ∂ ∂
+ φ̂ (r (∇ × A)θ ) − ((∇ × A)r ) ,
r ∂r ∂θ

where
392 Appendix: Vectorial Analysis
 
1 ∂   ∂ Aθ
(∇ × A)r = sin θ Aφ − , (A.50)
r sin θ ∂θ ∂φ
 
1 ∂ Ar ∂  
(∇ × A)θ = − sin θ r Aφ ,
r sin θ ∂φ ∂r
 
1 ∂ ∂ Ar
(∇ × A)φ = (r Aθ ) − .
r ∂r ∂θ

Substituting Eq. (A.49) and (A.50) into second expression of Eq. (A.26), we calculate
Laplacian of a vector field A.
In cylindrical coordinates, for the scalar field ψ, we can calculate the gradient as

∂ψ 1 ∂ψ ∂ψ
∇ψ = ρ̂ + φ̂ + ẑ , (A.51)
∂ρ ρ ∂φ ∂z

where the first expression of Eq. (A.34) is used. Then, for a vector field function A,
we calculate the divergence in cylindrical coordinates as follows:
    
1 ∂ g1 g2 g3 ∂ g1 g2 g3
∇·A= A1 + A2 (A.52)
g1 g2 g3 ∂ x1 g1 ∂ x2 g2
 
∂ g1 g2 g3
+ A3
∂ x3 g3
 
1 ∂ ∂ ∂
= (g2 g3 A1 ) + (g1 g3 A2 ) + (g1 g2 A3 )
g1 g2 g3 ∂ x1 ∂ x2 ∂ x3
 
1 ∂   ∂ Aφ ∂
= ρ Aρ + + (ρ A z )
ρ ∂ρ ∂φ ∂z
 
1 ∂   ∂ Aφ ∂ Az
= ρ Aρ + +ρ .
ρ ∂ρ ∂φ ∂z

The rotor ∇ × A in cylindrical coordinates is given as

 
1 ∂ ∂
∇ × A = x̂1 123 (g3 3 )
A − (g2 2 )
A (A.53)
g2 g3 ∂ x2 ∂ x3
 
1 ∂ ∂
+ x̂2 213 (g3 A3 ) − (g1 A1 )
g1 g3 ∂ x1 ∂ x3
 
1 ∂ ∂
+ x̂3 312 (g2 A2 ) − (g1 A1 )
g1 g2 ∂ x1 ∂ x2
 
1 ∂ Az ∂  
= ρ̂ − ρ Aφ
ρ ∂φ ∂z
 
∂ Aρ ∂ Az
+ φ̂ −
∂z ∂ρ
 
1 ∂   ∂ Aρ
+ ẑ ρ Aφ − .
ρ ∂ρ ∂φ
Appendix: Vectorial Analysis 393

Similarly, Laplacian of the scalar field ψ in cylindrical coordinates is calculated

as

Δψ = ∇ · (∇ψ) (A.54)
 
∂ψ 1 ∂ψ ∂ψ
= ∇ · ρ̂ + φ̂ + ẑ (A.55)
∂ρ ρ ∂φ ∂z
     
∂ψ 1 ∂ψ ∂ψ
= ∇ · ρ̂ + ∇ · φ̂ + ∇ · ẑ
∂ρ ρ ∂φ ∂z
        
T1 T2 T3

= T1 + T2 + T3 .

Each of the terms T1 , T2 , and T3 is calculated using Eq. (A.34) as follows:

 
1 ∂ ∂ ∂
T1 = (g2 g3 A1 ) + (g1 g3 A2 ) + (g1 g2 A3 ) (A.56)
g1 g2 g3 ∂ x1 ∂ x2 ∂ x3
⎡ ⎤
⎢ ⎥
⎢ ⎥
⎢
1⎢∂   ∂   ∂ ⎥
⎥
= ⎢ ρ Aρ + Aφ + (ρ A z )⎥
ρ ⎢ ∂ρ    ∂φ    ∂z   ⎥
⎢ A z =0 ⎥
⎣ ∂ψ Aφ =0 ⎦
Aρ =
∂ρ
 
1 ∂ ∂ψ
= ρ
ρ ∂ρ ∂ρ
1 ∂ψ ∂ψ 2
= + ,
ρ ∂ρ ∂ρ 2

 
1 ∂ ∂ ∂
T2 = (g2 g3 A1 ) + (g1 g3 A2 ) + (g1 g2 A3 ) (A.57)
g1 g2 g3 ∂ x1 ∂ x2 ∂ x3
⎡ ⎤
⎢ ⎥
⎢ ⎥
1⎢⎢∂   ∂   ∂ ⎥
⎥
= ⎢ ρ Aρ + Aφ + (ρ A z )⎥
ρ ⎢ ∂ρ    ∂φ    ∂z   ⎥
⎢ A z =0 ⎥
⎣ Aρ =0 1 ∂ψ ⎦
Aφ =
ρ ∂φ
 
1 ∂ 1 ∂ψ
=
ρ ∂φ ρ ∂φ
1 ∂ 2ψ
= 2 .
ρ ∂φ 2
394 Appendix: Vectorial Analysis

The last term is

 
1 ∂ ∂ ∂
T3 = (g2 g3 A1 ) + (g1 g3 A2 ) + (g1 g2 A3 ) (A.58)
g1 g2 g3 ∂ x1 ∂ x2 ∂ x3
⎡ ⎤
⎢ ⎥
⎢ ⎥
1 ⎢∂   ∂   ∂ ⎥
= ⎢ (ρ A z ) ⎥
ρ ⎢ ∂ρ ρAρ + Aφ +
 ∂φ    ∂z    ⎥
⎢ ⎥
⎣ Aρ =0 Aφ =0 ∂ψ ⎦
Az =
∂z
 
1 ∂ ∂ψ
= ρ
ρ ∂z ∂z
∂ 2ψ
= .
∂z 2

Thus, combining T1 , T2 , and T3 , we obtain

∂ψ 2 1 ∂ 2ψ ∂ 2ψ 1 ∂ψ
Δψ = + + + . (A.59)
∂ρ 2 ρ 2 ∂φ 2 ∂z 2 ρ ∂ρ

For Laplacian of a vector field in cylindrical coordinates (the second expression in

Eq. (A.26)), we first calculate

∂
∇(∇ · A) = ρ̂ (∇ · A) (A.60)
∂ρ
  
I1
1 ∂
+ φ̂ (∇ · A)
ρ ∂φ
  
I2
∂
+ ẑ (∇ · A)
 ∂z  
I3

= I1 + I2 + I3 ,

with ∇ · A given by Eq. (A.52). Hence, we have

 
∂ 1 ∂   1 ∂ Aφ ∂ Az
I1 = ρ̂ ρ Aρ + + . (A.61)
∂ρ ρ ∂ρ ρ ∂φ ∂z

The term, I2 is
 
φ̂ ∂ 1 ∂   1 ∂ Aφ ∂ Az
I2 = ρ Aρ + + , (A.62)
ρ ∂φ ρ ∂ρ ρ ∂φ ∂z
Appendix: Vectorial Analysis 395

and the term I3 is

 
∂ 1 ∂   1 ∂ Aφ ∂ Az
I3 = ẑ ρ Aρ + + . (A.63)
∂z ρ ∂ρ ρ ∂φ ∂z

Furthermore, we calculate
 
1 ∂(∇ × A)z ∂  
∇ × ∇ × A = ρ̂ − ρ(∇ × A)φ (A.64)
ρ ∂φ ∂z
 
∂(∇ × A)ρ ∂(∇ × A)z
+ φ̂ −
∂z ∂ρ
 
1 ∂   ∂(∇ × A)ρ
+ ẑ ρ(∇ × A)φ − ,
ρ ∂ρ ∂φ

where
 
1 ∂ Az ∂  
(∇ × A)ρ = − ρ Aφ (A.65)
ρ ∂φ ∂z
∂ Aρ ∂ Az
(∇ × A)φ = −
∂z ∂ρ
 
1 ∂   ∂ Aρ
(∇ × A)z = ρ Aφ − .
ρ ∂ρ ∂φ

Some useful applications of the spherical and cylindrical coordinates include the
cases when there is some symmetry of the problem such that either the scalar field
function ψ or vector field function A depends on fewer coordinates. For example,
in problems with cylindrical symmetry scalar electric potential function depends
only on ρ, and it is independent on φ and z, then the formulas simplify significantly
compare with those using the Cartesian coordinate system.