Classic Electrodynamics

Lecture Notes 2023/2024, Técnico Lisboa

Prof. Rogério Jorge, Zakary Buras-Stubbs & Tiago Jorge

Contents
Section 1: Propagation of electromagnetic waves - September 19 2023, Part
1 1
Subsection 1.1: The wave equation and its solution . . . . . . . . . . . . . . . . . 1
Subsection 1.2: Plane waves in a homogeneous medium . . . . . . . . . . . . . . . 2
Subsection 1.3: Plane Monochromatic Waves . . . . . . . . . . . . . . . . . . . . . 2
Subsection 1.4: Energy and Momentum in Electromagnetic Waves . . . . . . . . . 3

Section 2: Electromagnetic Waves in Matter - September 19 2023, Part 2 4

Subsection 2.1: Propagation in linear media . . . . . . . . . . . . . . . . . . . . . 4
Subsection 2.2: Reflection and Refraction of Waves . . . . . . . . . . . . . . . . . 5
Subsection 2.3: Propagation in conductive media . . . . . . . . . . . . . . . . . . 6

Electromagnetic Waves

Section 1: Propagation of electromagnetic waves - September 19 2023, Part 1

................................................................................................................................
Subsection 1.1: The wave equation and its solution
A wave can be seen as information that travels at a given velocity in a medium along a
given direction. For now, we’ll treat them as perturbations of the fields that travel with a
fixed shape at constant velocity. A typical wave equation in one dimension x looks like

@ 2y 1 @ 2y
= , (1.1)
@x2 v 2 @t2
where y is the displacement of the perturbation, t is time and v the velocity of the wave. A
typical solution of the wave looks like a sinusoidal wave

y = A cos[k(z vt) + ], (1.2)

where A is the amplitude of the wave, the argument of the cosine is the phase, is the phase
constant and k is the wave number, which is related to the wavelength of the wave by
= 2⇡/k as when x advances by 2⇡/k the cosine completes one cycle. The period of the
wave is given by T = 1/f with f = v/ = kv/2⇡ the frequency of the wave. Its angular

1
frequency is ! = 2⇡f = kv. The solution can also be written as y = A cos(kz !t + ).
Using Euler’s formula ei✓ = cos ✓ + i sin ✓, we can write the complex function

ỹ = Ãei(kz !t)
, (1.3)

with à = ei and the solution written as

y = Re[ỹ]. (1.4)

Sometimes, the complex notation is easier to manipulate due to the exponential function
instead of sines and cosines.
In more than one dimension, we can define the perturbation vector as

f̃ = Ãei(kz !t)
n, (1.5)

where n is the polarization vector. When the waves are transverse, n is perpendicular to the
direction of propagation z, n · z = 0. The polarization of the wave can be characterized by
the direction of n with respect to x, y and z. It can be written in terms of the polarization
angle ✓,
n = cos ✓x + sin ✓y. (1.6)

................................................................................................................................
Subsection 1.2: Plane waves in a homogeneous medium
In vacuum, in the absence of charges and currents, the electromagnetic field vectors in a
dielectric satisfy the equations
@B @D
r⇥E= , r⇥H= , r · D = 0, r · B = 0. (1.7)
@t @t
In a nondispersive medium, the field vectors are related by D = ✏E, B = µH where ✏ is the
permittivity and µ the magnetic permeability. In the case of homogeneous media, where ✏
and µ are constant, Eq. (1.7) can be combined in the following second order equations

1 @ 2E 1 @ 2B
2 2
r E= 2 2, r B= 2 2 , (1.8)
v @t v @t
p
where v = 1/ ✏µ = c = 3 ⇥ 108 m/s is the phase velocity of the wave. Therefore, each
component of the electric and magnetic field satisfies, independently, the 3D wave equation
1 @ 2f
r2 f = , (1.9)
v 2 @t2
where the wave travels at the speed of light v = c.
................................................................................................................................
Subsection 1.3: Plane Monochromatic Waves
A plane monochromatic wave propagating in the direction of the wavevector k where
k = 2⇡/ with the wavelength, is described by

E = E0 ei(k·r !t)
. (1.10)

2
The wave amplitude E0 is, in general, a complex vector. Due to the fact that r · E = 0 in
vacuum, the waves are transverse and E0 is perpendicular to the direction of propagation k.
Depending on the magnitude and direction of the real and imaginary parts of E0 , the waves
will be plane, circularly or elliptically polarised.
However, such waves are a mathematical idealisation. In reality, waves are in fact only
close to monochromatic, that is, they are a superposition of monochromatic waves with
frequencies in some range !. The superposition of plane monochromatic waves with different
wave vectors and frequencies is often referred to as a wave group or wavepacket. This is
described by ˆ
(r, t) = (k)ei(k·r !t) dk, (1.11)

where is a Cartesian component of E or H. The function (k) is called the amplitude

function. The maximum of the amplitude of a wavepacket moves through space with the
group velocity
d!
vg = . (1.12)
dk
While each wave equation is independent from the other, the components of the electric
and magnetic fields are related between them via the Maxwell’s equations. As r · E =
r · B = 0 (in vacuum), it shows that k · E = k · B = 0, so that electric and magnetic fields
are perpendicular to the direction of propagation. For waves travelling along z, we have that
Ez = Bz = 0. With Faraday’s law r ⇥ E = @B/@t, for waves travelling along z, k = kz,
we find that
k
B0 = (z ⇥ E0 ). (1.13)
!
Then E and B are in phase and mutually perpendicular, with the real amplitudes related
by E0 = !B0 /k = cB0 .
................................................................................................................................
Subsection 1.4: Energy and Momentum in Electromagnetic Waves
The energy per unit volume in the electromagnetic fields is given by
✓ ◆
1 B2
u= 2
✏0 E + . (1.14)
2 µ0

In case of a monochromatic plane wave, B 2 = E 2 /c = µ0 ✏0 E 2 so that the electric and

magnetic contributions to the energy are equal

u = ✏0 E 2 = ✏0 E02 cos2 (kz !t + ). (1.15)

To find the energy that it carries with it, we wil use the Poynting vector
1
S= E ⇥ B. (1.16)
µ0
For a monochromatic plane wave propagating in the z direction, we find that

S = c✏0 E02 cos2 (kz !t + )z = cuz. (1.17)

3
 As well, the momentum stored in the fields is given by g = S/c2 . For monochromatic
plane waves, we find that
1 1
g = ✏0 E02 cos2 (kz !t + )z = uz. (1.18)
c c
We also write the average value of each of these quantities by noting that the average of
cos is 1/2, leading to
2

✏0 E02 c✏0 E02 ✏0 E02

hui = , hSi = , hgi = . (1.19)
2 2 2c
The intensity of the wave is the average power per unit area I = hSi = c✏0 E02 /2. The
radiation pressure is the average force per unit area. For the momentum transfer p =
hgi Ac t, the radiation pressure is given by
1 p 1 I
P = = ✏0 E02 = . (1.20)
A t 2 c

Waves in Matter

Section 2: Electromagnetic Waves in Matter - September 19 2023, Part 2

................................................................................................................................
Subsection 2.1: Propagation in linear media
In linear media, D = ✏E and H = B/µ. These are regions where there is no free charge
or free current. Furthermore, if the medium is homogeneous, so that ✏ and µ are constant,
Maxwell’s equations reduce to
@B
r · E = 0, r ⇥ E = , (2.1)
@t
@E
r · B = 0, r ⇥ B = µ✏ . (2.2)
@t
Waves propagating through such media then travel at the speed
1 c
v=p = , (2.3)
✏µ n

where r
✏ µ p
n= = ✏ r µr , (2.4)
✏ 0 µ0
with n the index of refraction. As, for most materials, µr ' 1, we typically use the approxi-
mation
p
n ' ✏r . (2.5)
For more info about the role of this relation in the transparency of materials, you can read
about the Ewald-Oseen extinction theorem. All the results for the energy, power, intensity

4
and pressure of the electromagnetic wave that we found in vacuum carries through to waves
in linear media by the replacement ✏0 ! ✏, µ0 ! µ and c ! v.
However, the boundary conditions for the fields passing from one medium to another are
altered. Here, we find that
k k
✏1 E1? = ✏2 E2? , E1 = E2 , (2.6)
k k
B1 B2
B1? = B2? , = , (2.7)
µ1 µ2
(2.8)

................................................................................................................................
Subsection 2.2: Reflection and Refraction of Waves
We now look at the propagation of waves in linear media and the resulting phenomena
when it passes from one medium to another. Let’s use simple plane waves for this problem
travelling in the z direction

EI = E0I ei(k1 z !t)

x, BI = B0I ei(k1 z !t)
y, (2.9)
p
where B0I = E0I /v1 and v1 = c/n1 with n1 ' ✏r1 the refraction index in the medium
1. We call this the incident wave. We then put another medium at z = 0 with index of
p
refration n2 ' ✏r2 so that waves propagating there have different velocities v2 = c/n2 and
different wave numbers k2 . This interface between two media leads to a reflected wave back
to medium 1
ER = E0R ei( k1 z !t) x, BR = B0R ei( k1 z !t) y, (2.10)
where B0R = E0R /v1 as k changed direction and B = k ⇥ E/! with v = !/k. It also leads
to a tramsitted wave in medium 2

ET = E0T ei(k2 z !t)

x, BT = B0T ei(k2 z !t)
y, (2.11)

where B0T = E0T /v2 .

The total wave from the left EI + ER and BI + BR and the total wave from the right ET
and BT should obey the boundary conditions derived for linear media. We first note that
there are no z components of the wave, so that the perpendicular boundary conditions are
k k
not needed. The parallel boundary conditions for the electric field E1 = E2 lead to

E0I + E0R = E0T , (2.12)

k k
while the parallel boundary conditions for the magnetic field B1 /µ1 = B2 /µ2 lead to
✓ ◆
1 E0I E0R 1 E0T
= . (2.13)
µ1 v 1 v1 µ2 v 2

or E0I E0R = E0T with

µ1 v 1 µ1 n 2
= = . (2.14)
µ2 v 2 µ2 n 1

5
 Solving for the amplitudes, we find that
1 v2 v1 2E0I 2v2
E0R = E0I ' E0I , E0T = ' E0I , (2.15)
1+ v2 + v1 1+ v2 + v1

where we approximated µ ' µ0 . In this particular case, the ratios of the intensity I = ✏vE02 /2
for each one of the waves gives rise to the following reflection R and transmission T coefficients
✓ ◆2
IR n1 n2 IT 4n1 n2
R= = , T = = . (2.16)
II n1 + n2 II (n1 + n2 )2

These satisfy R + T = 1 as required by conservation of energy. Usually, most of the light is

transmitted, that is, T R in general.
The case of oblique incidence, with ER = E0R ei(kR ·r !t) and BR = kR ⇥ER /v1 kR , requires
substantially more algebra, but the approach is exactly the same as we did before. Applying
the boundary conditions at the interface leads to a relation between the wave-vectors of each
wave kI , kR and kT as each cartesian component should be equal, so that we can define an
angle of incidence ✓I , reflection ✓R and refraction or transmission ✓T as

kI sin ✓I = kR sin ✓R = kT sin ✓T . (2.17)

The fact that the frequency stays constant leads to kI v1 = kR v1 = kT v2 = !, which together
with Eq. (2.17) leads to the law of reflection ✓I = ✓R and to Snell’s law sin ✓T / sin ✓I = n1 /n2 .
Applying then the boundary conditions at the interface to the cartesion components of the
fields leads to Fresnel’s equations
↵ 2
E0R = E0I , E0T = E0I , (2.18)
↵+ ↵+

with ↵ = cos ✓T / cos ✓I . Interestingly, for ↵ = , the reflected wave vanishes, leading to the
Brewster’s angle
2
1 n2
sin2 ✓B = ! tan ✓ B ' . (2.19)
(n1 /n2 )2 2 n1
Regarding the power striking the interface S · z, we have II = ✏1 v1 E0I 2
cos ✓I /2, IR =
2
✏1 v1 E0R cos ✓R /2 and IT = ✏T vT E0T
2
cos ✓T /2, leading to the reflection and transmission co-
efficients ✓ ◆2 ✓ ◆2
IR ↵ IT 2
R= = , T = =↵ , (2.20)
II ↵+ II ↵+
with R + T = 1.
................................................................................................................................
Subsection 2.3: Propagation in conductive media
In conductive media, charges respond to the electric and magnetic fields. Such fields and
create a current flow by pushing the charges in a particular direction. For most substances,
the current density J is proportional to the force per unit charge f = E + v ⇥ B. The
proportionality factor, , is called the conductivity of the material and is often obtained
experimentally for each material. Its inverse is called the resistivity ⇢ = 1/ . Insulators

6
have = 0 while perfect conductors have = 1. In most materials (but not for plasmas),
the current is mostly due to the electric field applied, so that
J = E. (2.21)
Equation (2.21) is called Ohm’s law. In this case, Maxwell’s equation for H should be
replaced by
@D
r⇥H= E+ . (2.22)
@t
What about charge conservation? Using the continuity equation for free charge
@⇢
r·J= , (2.23)
@t
together with Ohm’s law r · E = ⇢/✏, we find that
@⇢
= (r · E) = ⇢, (2.24)
@t ✏
with a solution in a homogeneous linear medium
⇢ e ( /✏)t
⇢0 . (2.25)
This means that any initial free charge ⇢0 dissipates in a characteristic time ⌧ = ✏/ .
How are the wave equations modified in conductors? For time periods greater than ⌧ , so
that ⇢ ' 0, we can write Maxwell’s equations as
@B
r · E = 0, r⇥E= , (2.26)
@t
@E
r · B = 0, r ⇥ B = µ✏ + µ E. (2.27)
@t
This system leads to the modified wave equations for E and B
@ 2E @E @ 2B @B
r2 E = µ✏ 2
+ µ , r 2
B = µ✏ 2
+µ . (2.28)
@t @t @t @t
The plane wave solutions for waves travelling in the z direction look similar to the previous
solutions
E = E0 ei(kz !t) , B = B0 ei(kz !t) , (2.29)
with the difference that the wave number k is now complex
k 2 = µ✏! 2 + iµ !, (2.30)
with solutions k = kR + ikI ,
sr sr
r ⇣ ⌘ r ⇣ ⌘2
✏µ 2 ✏µ
kR = ! 1+ + 1, kI = ! 1+ 1. (2.31)
2 ✏! 2 ✏!
The imaginary part of k leads to an attenuation of the wave, that is, a decreasing ampli-
tude with increasing z,
E = E0 e kI z i(kR z !t)
e , B = B0 e kI z i(kR z !t)
e . (2.32)

7
The distance d = 1/kI is called the skin depth and is a measure of how fat the wave penetrates
into the conductor. The real part determines, as before the propagation speed and the index
of refraction, = 2⇡/kR , v = !/kR and n = ckR /!.
In terms of its modulus and phase, we can write
q q p
i 2 2
k = Ke , K = kR + kI = ! ✏µ 1 + 2 /✏2 ! 2 , (2.33)

and = tan 1 (kI /kR ).

Without loss of generality, we assume that the waves are travelling along z and that E
is polarized along x which, using Faraday’s law, leads to
k
E = E0 e kI z i(kR z !t)
e x, B = E0 e kI z i(kR z !t)
e y. (2.34)
!
This shows that E and B are no longer in phase but have a phase difference of and that
their real amplitudes are related by |B|/|E| = K/!.