ELECTROMAGNETISM

Prof. D P Hampshire – Part II: Summary 2nd Year Physics Lecture Course 20240103

Part I was :
Lectures and learning classical physics.
Maxwell I and Maxwell II: the divergence equations
Maxwell III and Maxwell IV: the curl equations
Electromagnetic waves and communication
Maxwell and Einstein
Figs: 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33 and 34 .

6 Conductors and superconductors

6.1 Maxwell’s contribution – shirts to meet other life forms
6.2 Properties of conductors

6.2.1 Current density J in conductors

When first considering resistors in circuits at school, we all found

V = IR (Ohm’s Law)
6-1

V
𝐸 = and 𝐽 = I/A 6-2
L

Gives

𝑬 = 𝑱ρn (Ohm’s law + Definition of resistivity, ρn ).

6-3

𝑱 = σn 𝑬 (Ohm’s law + Definition of conductivity, σn ),

6-4

where σn = 1/ρn .

6.2.2 Charge Density in conductors

𝜌 ∂ρ
Using Ohm’s law (𝑬 = 𝑱ρn ), Maxwell I ( 𝛁. 𝑬 = ε ) and the continuity equation (𝛁 ⋅ 𝑱 = − ∂t ), we can
0
use a dimensionality argument to find that the characteristic life-time, 𝜏 , is:

𝜏 = ρn ε0
6-5

𝛁. 𝑬 = 0 (Maxwell I for conductors) .

6-6

pg. 1 Confidential © Damian Peter Hampshire 2023. All rights reserved.

6.2.3 Drude semi-classical model (microscopic properties)

We use the general Coulomb-Lorentz expression for the force, F , on the charge carriers

𝑭 = q(𝑬 + 𝒗 × 𝑩)
6-7

6-8
𝑭 ≈ q𝑬.,

Figure 35 : Top: The Drude model - The electric field accelerates the charges which then collide with the
scattering sites. Bottom: The semi-classical Drude model – only the conduction charges near the Fermi
surface can change their state – change velocity and scatter.

pg. 2 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Hence the momentum of the charge carriers at time t + ∂t is given by
∂t ∂t ∂t
p(t + ∂t) = (1 − ) p(t) + (1 − ) F(t) ∂t + {F(t) ∂t } 6-9
τ τ τ

∂p = p(t + ∂t) − p(t)

6-10

Gives
∂p p(t)
=− + F(t) + O(∂t) 6-11
∂t τ

Using F = qE and p = mv gives:

dv mv
m = qE − (Equation of motion for charge carriers in a conductor) 6-12
dt τ
mv
where the term can be interpreted as a frictional damping term.
τ

𝑱 = Nq𝒗
6-13

d𝑱 m
m = N𝑞 2 𝑬 − 𝑱 6-14
dt τ

d𝑱̃ m
m ̃−
= N𝑞 2 𝑬 𝑱̃ 6-15
dt τ

̃ = 𝑬0 exp −iωt 6-16

𝑬

𝑱̃ = 𝑱0 exp −i(ωt − θ)
6-17

N𝑞 2
𝑱̃ = ̃
⋅𝑬
m(τ−1 − iω) 6-18

Nq2
̃n =
σ (The Drude conductivity)
m(τ−1 − iω) 6-19

𝝉Drude classical = 𝝉Drude semi−classical (EF )

6-20
∗
𝑚Drude classical = 𝑚Drude semi−classical (EF ) .
6-21

6.3 Electromagnetic waves propagating through conductors

pg. 3 Confidential © Damian Peter Hampshire 2023. All rights reserved.

6.3.1 Dispersion relation and Drude model for conductors

Nq2
σn =
m (τ−1 − iω) 6-22

To calculate the dispersion relation, we again use Maxwell’s equations. Taking the curl of Maxwell III
gives:

∂
𝛁×𝛁×𝑬= − 𝛁×𝑩 6-23
∂t

∂𝑬 ∂𝑬
𝛁 × 𝑩 = μ0 𝑱 + μ0 ε0 = μ0 𝜎𝑁 𝑬 + μ0 ε0 6-24
∂t ∂t

∂𝑬 ∂2 𝑬
𝛁 × 𝛁 × 𝑬 == −μ0 𝜎𝑁 − μ0 ε0 2 6-25
∂t ∂𝑡

Using the vector identity

𝛁 × (𝛁 × 𝑬) = 𝛁(𝛁 ⋅ 𝑬) − ∇2 𝑬
6-26

∂𝑬 ∂2 𝑬
∇2 𝑬 = μ0 𝜎𝑁 + μ0 ε0 2 6-27
∂t ∂𝑡

This equation is of the form of a travelling wave travelling along the x-direction. We assume wave-
equation given by equation 9-43a trial solution of the complex, where the electric field is complex and
given by:
̃ (r, t) = 𝑬𝑜 expi(𝑘𝑥 − ωt)
𝑬
6-28

k 2 = μ0 ε0 ω2 + iωμ0 𝜎𝑁
6-29

The general solution for the E-field is simply found by substituting k back into the plane wave equation

̃ = 𝑬o exp[i( (k real + 𝑖k imaginary )𝑥 − 𝜔𝑡)]

𝑬
6-30

𝑬 = 𝑬o cos(k real x − ωt)exp(−k imaginary 𝑥)

6-31

The wave propagates in the conductor but is exponentially attenuated within the conductor. The
cos(k real x − ωt) is the oscillating wave part where the wavelength is

2𝜋
𝜆= 6-32
k 𝑟𝑒𝑎𝑙
pg. 4 Confidential © Damian Peter Hampshire 2023. All rights reserved.
The exp(−k imaginary 𝑥) term is the decaying part. Rewriting exp(−k imaginary 𝑥) = exp(−𝑥/δ), where

1
δ= 6-33
k 𝑖𝑚𝑎𝑔𝑖𝑛𝑎𝑟𝑦

6.3.2 The high frequency limit (i.e. ω ≫ τ−1).

In the high frequency limit ω ≫ τ−1

Nq2 iNq2
σn = lim−1 = +
ω≫τ me (τ−1 − iω) me ω 6-34

Substituting into the dispersion relation for conductors gives

iNq2 Nq2
k 2 = μ0 ε0 ω2 + iωμ0 = μ0 ε0 ω2 {1 − }
me ω me εo ω2 6-35

using c 2 = 1/μ0 ε0 . This can be rewritten

ω2 ω𝑝 2
k2 = {1 − ( ) } 6-36
c2 ω

where the (angular) plasma frequency is given by,

1
Nq2 2
ωp = ( ) 6-37
me εo

The dispersion relation has two distinct regimes:

i) ω < ωp ; k is imaginary and there is an attenuated propagation of the
electromagnetic waves.
ii) ω > ωp ; k is real and the electromagnetic waves propagate without attenuation
through the plasma.
c
vphase = (> 𝑐 for 𝜔 > ωp )
ω 2
√1 − ( p ) 6-38
ω

ωp 2
vgroup = c√1 − ( ) (< 𝑐 for 𝜔 > ωp ) 6-39
ω

pg. 5 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 36 : The velocity of an electromagnetic wave in a plasma as a function of angular frequency.
Electromagnetic waves propagating are attenuated at angular frequencies below the angular plasma
frequency (ωp ).

Elemental metal: Li Na K Rb
λ𝑝 (Calculated from λ𝑝 = 2𝜋𝑐/𝜔𝑝 ) : 1550 2090 2870 3220
λ𝑝 (Experimental data) : 1550 2100 3150 3400

6.3.3 The low frequency limit (i.e. ω < τ−1 )

Nq2 τ
σn = 6-40
m

In the low frequency limit,

k 2 = iωμ0 σn
6-41

Note this approximation for the dispersion relation is also know as the “good conductor” limit since 𝜎𝑁 ≫
ωε0 .

pg. 6 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 37 : An Argand diagram showing the two roots of the imaginary number 𝑖 = √−1

1+i
√i =
√2 6-42

ωμ0 σn
k = (1 + i)√ = k real + 𝑖k imaginary 6-43
2

ωμ0 σn ωμ0 σn
where k real = √ and k imaginary = √ .
2 2

̃ = 𝑬o exp[i( (k real + 𝑖k imaginary )𝑥 − 𝜔𝑡)]

𝑬
6-44

𝑬 = 𝑬o cos(k real x − ωt)exp(−k imaginary 𝑥)

6-45

2𝜋 2
𝜆= = 2𝜋√ 6-46
k 𝑟𝑒𝑎𝑙 ωμ0 σn

1 2
δ= =√ 6-47
k 𝑖𝑚𝑎𝑔𝑖𝑛𝑎𝑟𝑦 ωμ0 σn

pg. 7 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 38 : The reflectivity of a 1 mm slab of copper as a function of frequency.

6.4 Superconducting Materials

i) The materials properties and definitions for superconductors

The two London equations describe the fundamental properties of superconductors that are found in low
magnetic fields:
a) Zero resistivity
If we assume there is no scattering of the charge carriers we have
∂𝒗
𝑭 = m𝒂 = m 6-48
∂t

Using 𝑭 = q𝑬 and J = Nqv, we find the first London equation,

∂𝑱
𝑬 = μ0 λ2𝐿 (1st London Equation), 6-49
∂t

pg. 8 Confidential © Damian Peter Hampshire 2023. All rights reserved.

 1
me
where the London penetration depth , λ𝐿 , is given by λ𝐿 = (μ )2 and Ns is the density of
0 Ns e2
superelectrons.

b) The Meissner state – exclusion of magnetic flux

The 2nd London equation is given by,

𝑩 = −μ0 λ2𝐿 ∇ × 𝑱. (2nd London Equation).

6-50

Substituting Maxwell’s 4th equation for J , where (∂𝑬⁄∂t = 0), into the second London equation gives

𝑩 = −λ2𝐿 ∇ × ∇ × 𝑩
6-51

Using the vector relation:

∇ × (∇ × 𝑩) = ∇(∇ ⋅ 𝑩) − ∇2 𝑩
6-52

Gives

1 6-53
∇2 𝑩 = 𝑩
λ2𝐿
For a semi-infinite slab, the equation has solutions of the form:
𝑩(x) = 𝐁0 exp(−𝑥/λ𝐿 ) for 𝑥 > 0 ,
6-54

which describes the Meissner effect.

pg. 9 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 39 : At the critical temperature, TC, the material becomes superconducting and the resistivity
drops to zero.

Figure 40 : The magnetic field profile for a (LHS) semi-infinite slab and (RHS) a cylinder of
superconductor in an applied field (Bapplied). The field decays exponentially at the surface to zero over a
characteristic distance λ𝐿 .

pg. 10 Confidential © Damian Peter Hampshire 2023. All rights reserved.

7 Dielectric and Magnetic Materials
7.1 The materials playground

7.2 The permittivity of dielectrics

7.2.1 Polar and non-polar dielectrics

Figure 41 : A polarized cube of dielectric between two capacitor plates.

pg. 11 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 42 : The polar dielectric H2O (water) which has a permanent electric dipole moment.

Figure 43 : The non-polar dielectric argon (Ar), which has no dipole moment until an E-field is applied.

pg. 12 Confidential © Damian Peter Hampshire 2023. All rights reserved.

7.2.2 Current density and charge density in dielectrics

Figure 44 : An electric dipole moment p.

The fundamental response of a dielectric:

𝒑 = q𝒅 − Definition of the electric dipole moment (𝒑)

7-1

𝑷 = N𝒑 − Definition of the polarisation, 𝑷 (Cm−2 )

7-2

𝑷 = εo (εr − 1)𝑬 − Definition of the relative dielectric constant or permittivityy (εr )

7-3

pg. 13 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 45 : A cube of polarized dielectric. The +ve charge is displaced in the E-field.

In an applied E field, we consider the +ve charges to be displaced. The positive charge at (x+Δx,y,z) is:

Q+ve = Nq
⏟ × Sd
⏟
charge/volume Volumeof+vecharge 7-4

which can be written in terms of surface charge density (σ):

Q+ve
σ= = Nqd = P 7-5
S

More generally, the surface charge density σ (C.m-2) is:

̂
σ=𝑷⋅𝐧
7-6

𝑷
𝑬 = εr 𝑬 −
εo 7-7

𝑷
where 𝑬 (as ever) is the net local field, εr 𝑬 is the field produced by the plates and − ε is the field
o
produced by the dielectric.

pg. 14 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Now consider a cube of non-uniformly polarized dielectric (you may consider the negative to be fixed in
space) inside the bulk of a dielectric:

QIn (𝑥) = P(𝑥)Δ𝑦Δ𝑧

7-8

QOut (𝑥 + Δ𝑥) = P(𝑥 + Δ𝑥)Δ𝑦Δ𝑧

7-9

∂P
∂Q = QIn (𝑥) − QOut (𝑥 + Δ𝑥) = −[P(𝑥 + Δ𝑥) − P(𝑥)]Δ𝑦Δ𝑧 = − Δ𝑥Δ𝑦Δ𝑧 7-10
∂𝑥

∂P𝑥 ∂P𝑦 ∂P𝑧

ρb = − − − = −𝛁 ⋅ 𝑷
∂𝑥 ∂𝑦 ∂𝑧 7-11

Figure 46 : An arbitrary volume of polarized material

∂Q ∂ρb ∂ ∂
= −∫ dV = ∫ 𝛁. 𝑷dV = ∫ 𝑷 ⋅ d𝑺
∂t ∂t ∂t ∂t 7-12

pg. 15 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Equally
∂Q
= ∫ 𝑱b ⋅ d𝑺
∂t 7-13

Given this is true for an arbitrary surface, the current density at any point is

∂𝑷
𝑱b = 7-14
∂t

7.3 The permeability (μr ) of magnetic materials

7.3.1 Ampère’s model

A magnetic material produces its own magnetic field.

Figure 47 : The local and macroscopic fields produced by a magnetic material

pg. 16 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 48 : A magnetic dipole moment m.

The fundamental response of a magnetic material:

̂ − Definition of the magnetic dipole moment (𝒎)

𝒎 = IS𝒏
7-15

where I is the current flowing around a loop of area S.

𝑴 = N𝒎 − Definition of the magnetization (𝑴)

7-16

where N is the number of magnetic dipoles per unit volume.

The definition for H , is

𝑩 = μo [𝑯 + 𝑴]
7-17

Where μo 𝑯 is roughly the field produced by the solenoid/scientist.

The permeability is

𝑩 = μo μr 𝑯 − Definition of relative permeability (μr ):

7-18
μr −1
or equivalently μo 𝑴 = 𝑩
μr

7.3.2 Currents densities in magnetic materials

Bulk current density

pg. 17 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 49 : Two magnetized cubes next to each other. The magnetization in both cubes point in the z-
direction in the top panel, but in the x-direction in the bottom panel. The net current density points in the
y-direction in both panels.

pg. 18 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Consider the currents flowing and their equivalent magnetization. For the cube on the left:

𝑚z = 𝐼1 (𝑥)𝑆 = 𝐼1 (𝑥)Δx. Δy = 𝑀z (𝑥). Δx. Δy. Δz

7-19

𝐼1 = 𝑀z Δz
7-20

The current along the interface between the cubes (Inet)#

𝐼net = 𝑀𝑧 . Δ𝑧 − (𝑀𝑧 (𝑥 + Δ𝑥). Δ𝑧 = −Δ𝑀𝑧 . Δ𝑧

7-21

𝐼𝑛𝑒𝑡 Δ𝑀𝑧 ∂𝑀𝑧

𝐽𝑦 = =− =− 7-22
Δ𝑥. Δ𝑧 Δ𝑥 ∂x

There is also a contribution to 𝐽𝑦 if Mx varies, given by

∂𝑀𝑥
𝐽𝑦 = 7-23
∂𝑧

Hence in 3D,
𝑱=𝛁×𝑴
7-24

Figure 50 : For a uniformly magnetised material, the sum of the small path integrals on the LHS is
equivalent to the large path integral on the RHS.

a) Surface Current Density

In a magnetized cylinder, the circulating currents in the bulk of the material cancel. The field from the
material comes entirely from the circulating surface current.

F2F Lecture 14 -2020

pg. 19 Confidential © Damian Peter Hampshire 2023. All rights reserved.
For a uniformly magnetized material, there is a discontinuous change in M at the surface. The current
through the rectangle shown in the figure (normal to the board) can equally well be considered as either a
surface current per unit length (𝐼𝐿 ) or a bulk current (𝑱) where

∫ 𝑱 ⋅ d𝑺 = 𝐼𝐿 𝐿
7-25

Figure 51 : Top: a uniformly magnetized cylinder. Bottom: a uniformly magnetised arbitrary shape.

pg. 20 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Using 𝑱 = 𝛁 × 𝑴 we have

∫ 𝑱 ⋅ d𝑺 = ∫(𝛁 × 𝑴) ⋅ d𝑺 = ∫ 𝑴 ⋅ d𝒍 = 𝑀𝐿 = 𝐼𝐿 𝐿
7-26

⇒ 𝐼𝐿 = 𝑀
7-27

More generally
̂
⇒ 𝑰L = 𝑴 × 𝐧
7-28

There is no charge density ρ, for magnetic materials

7.4 The general dispersion relation

Derived for a very general material that is simultaneously conducting, a dielectric, magnetic and Infinite
Linear-Isotropic-Homogeneous(ILIH).

∂𝑬 ∂𝑷 ∂𝑬
𝛁 × 𝑩 = μ0 𝑱 + μ0 ε0 = μ0 (σn 𝑬 + + 𝛁 × 𝑴 ) + μ0 ε0 − General equation 7-29
∂t ∂t ∂t

𝑷 = εo (εr − 1)𝑬 and μo 𝐌 = (μr − 1)𝑩 /μr ,

7-30

∂
𝛁 × 𝑩 = μ⏟
r μo σn 𝑬 + μ
⏟ r μo (ε⏟
ε 𝑬) .
∂t r o 7-31
μ μ ε

μ = μr μo 𝑎𝑛𝑑 ε = εr εo .
7-32

∂ ∂ ∂
𝛁 × (𝛁 × 𝑬) = − 𝛁 × 𝑩 = − [μ (σn 𝑬 + (ε𝑬))]
∂t ∂t ∂t 7-33

We now use the vector identity and Maxwell I for a conducting material

𝛁 × (𝛁 × 𝑬) = 𝛁(𝛁 ⋅ 𝑬) − ∇2 𝑬 and 𝛁 ⋅ 𝑬 = 0
7-34

∂𝑬 ∂2 𝑬
∇2 𝑬 − μσn − με 2 = 0 − General wave equation 7-35
∂t ∂t

2
∂𝑩 ∂2 𝑩
∇ 𝑩 − 𝜇σn − μϵ 2 = 0 − General Wave equation 7-36
∂t ∂t

We can substitute a plane wave solution into the wave equation.

𝑬(r, t) = 𝐄𝑜 expi(𝑘𝑥 − ωt) − complex plane wave solution

7-37

k 2 = μεω2 + iωμσn – General dispersion relation

7-38

pg. 21 Confidential © Damian Peter Hampshire 2023. All rights reserved.

7.5 Why does light change speed?

Maxwell II gives:
𝒌 × 𝑬 = ω𝑩
7-39

The phase velocity is given by

ω 𝐸
𝑣phase = =
k 𝑟𝑒𝑎𝑙 𝐵 7-40

The general dispersion relation can be written

iμσn 2
k 2 = (με + )ω 7-41
ω

So waves slow down because the material either decreases E, increases B or both.

Conductors Dielectrics Magnetic materials

Visualisation

Basic building Scattering time, τ 𝒑 = q𝒅 ̂

𝒎 = 𝐼𝑆𝒏
block
Bulk materials 𝑱 = Nq𝒗 𝑷 = N𝒑 𝑴 = 𝑁𝒎
response
Bulk materials 𝑱 = σn 𝑬 𝑷 = εo (εr − 1)𝑬 𝑩 = μo μr 𝑯 ,
characteristic (μo 𝑴 = (μr − 1)𝑩 )
properties 𝑬 = 𝑱ρn 𝑷 = εo 𝜒𝑒 𝑬 𝑴 = χ𝑯
(μo 𝑴 = χ𝑩/(1 + χ) )
Charge density ~ Zero. −𝛁. 𝑷 Zero

Current density 𝑱𝑓𝑟𝑒𝑒 ∂𝑷 𝑱=𝛁×𝑴

∂t ̂
𝑰L = 𝑴 × 𝐧

Auxiliary fields 𝑫 = εo 𝑬 + 𝑷 𝐁 = μo [𝑯 + 𝑴]
Equation of motion d𝑣 m𝑣 d𝑣 m𝑣
m = q𝐸 − m = q𝐸 −
d𝑡 τ d𝑡 τ
− mω20 𝑥

Table 1 : Useful definitions for conductors, dielectrics and magnetic materials.

pg. 22 Confidential © Damian Peter Hampshire 2023. All rights reserved.

8 Electromagnetic waves propagating through materials

8.1 The auxiliary fields D and H

8.1.1 The auxiliary field D.

We can write Maxwell I :

𝜌 ρfree + ρbound ρfree − 𝛁 ⋅ 𝑷

𝛁⋅𝑬 = = =
εo εo εo 8-1

Maxwell I can be written in terms of free current density alone where

𝛁 ⋅ 𝑫 = ρfree
8-2

This leads to a definition for the electric displacement field D where:

𝑫 = εo 𝑬 + 𝑷 − definition of 𝑫
8-3

or equivalently 𝑫 = εr εo 𝑬. This equation says that for electrostatics, 𝑫/εo is the E-field produced by the
free charges alone.

Figure 52 : The displacement field (D) between two capacitor plates.

pg. 23 Confidential © Damian Peter Hampshire 2023. All rights reserved.
8.1.2 The auxiliary field H and displacement current density, 𝑱Disp .

∂𝑬
𝛁 × 𝑩 = μ0 𝑱 + μ0 ε0 8-4
∂t

∂𝑷
𝑱 = 𝑱free + 𝑱bound = 𝑱free + +𝛁×𝑴 8-5
∂t

where: 𝑱free = 𝑱Ohms law + 𝑱Experimentalist .

∂
𝛁 × (𝑩 − μ0 𝑴) = μ0 𝑱free + μ0 (𝑷 + ε0 𝑬) 8-6
∂t

∂𝑫
𝛁 × 𝑯 = 𝑱free + 8-7
∂t

Where the definition for the magnetic field strength, H is

𝐁 = μo [𝑯 + 𝑴]
8-8

∂𝑫 ∂(εo 𝑬 + 𝑷)
𝑱Displacement Current Density = = 8-9
∂t ∂t

Hence, we can re-write Maxwell’s equations in terms of auxiliary fields and free charge density and free
current density where,

𝛁 ⋅ 𝑫 = ρfree
8-10

8-11
𝛁⋅𝑯=0

∂𝑯 μr εr ∂𝑯 8-12
𝛁 × 𝑫 = −με =− 2
∂t c ∂t

∂𝑫
𝛁 × 𝑯 = 𝐉free + 8-13
∂t

pg. 24 Confidential © Damian Peter Hampshire 2023. All rights reserved.

8.2 An optical interface

Figure 53 : Top: Static E-fields and B-fields. Bottom: A snapshot in time for an electromagnetic wave
incident on an interface.

pg. 25 Confidential © Damian Peter Hampshire 2023. All rights reserved.

8.2.1 A static E-field that crosses two dielectrics

Figure 54 : Different static E-fields in two different dielectrics.

A) For an insulating material Maxwell I

𝛁 ⋅ 𝐃 = ρfree = 0
8-14

∫ 𝛁 ⋅ 𝐃 𝑑𝑉 = 0 8-15
𝐶𝑦𝑙𝑖𝑛𝑑𝑒𝑟

∫ 𝑫1 ⋅ d𝑺 + ∫ 𝑫2 ⋅ d𝑺 = 0 8-16
𝐶𝑖𝑟𝑐𝑢𝑙𝑎𝑟 𝑡𝑜𝑝 𝐶𝑖𝑟𝑐𝑢𝑙𝑎𝑟 𝑏𝑜𝑡𝑡𝑜𝑚

or equivalently
8-17
⇒ D1 ⊥ = D2⊥ or equivalently ε1 E1 ⊥ = ε2 E2⊥ . (𝑫 = εo εr 𝑬)

B) Using Maxwell III (consider path B) integrating over the surface inside the dotted path

∂𝑩
∫ 𝛁 × 𝑬 ⋅ d𝑺 = − ∫ ⋅ d𝑺
Rectangle ∂t
8-18
Rectangle

∂ 8-19
∮ 𝑬 ⋅ d𝒍 = − ∫ 𝑩 ⋅ d𝑺
Rectangle ∂t Rectangle

pg. 26 Confidential © Damian Peter Hampshire 2023. All rights reserved.

 ∂𝜙
E1// l − E2// l = − =0 8-20
∂t

⇒ E1// = E2//
8-21

D⊥ (or equivalently εr E⊥ ) and E// are continuous across the interfacial plane between two dielectrics.

8.2.2 A static B-field that crosses two magnetic materials

Figure 55 Different static B-fields in two different magnetic media.

∫ 𝛁 ⋅ 𝑩 𝑑𝑉 = 0
8-22
Cylinder

8-23
∫ 𝑩 ⋅ d𝑺 = 0 ⇒ B1⊥ = B2⊥
Cylinder

B1⊥ = B2⊥
8-24

Maxwell IV

pg. 27 Confidential © Damian Peter Hampshire 2023. All rights reserved.

 ∂𝑫
𝛁 × 𝑯 = 𝑱𝐟𝐫𝐞𝐞 + 8-25
∂t

∂𝛁. 𝑫
∫(𝛁 × 𝑯) ⋅ d𝑺 = ∮ 𝑯 ⋅ d𝒍 = ∫ 𝑱free ⋅ d𝑺 + ∫ ⋅ dV 8-26
⏟ ∂t
−no free current density
−no free charge density

8-27
⇒ H1// l = H2// l ⇒ H1// = H2//

∴ B⊥ , H// are continuous across the boundary.

8.3 The Laws of geometrical optics

Figure 56 : An electromagnetic wave incident on an interface – it is partially reflected and partially

transmitted.

͠Let’s make no assumptions about the polarisation or phase of the reflected and transmitted wave:

̃ 𝑖 (𝒓, 𝑡) = 𝑬
𝑬 ̃ 0𝑖 exp 𝑖(𝒌𝑖 ⋅ 𝒓 − 𝜔𝑖 𝑡) 8-28

pg. 28 Confidential © Damian Peter Hampshire 2023. All rights reserved.

 ̃ r (𝒓, 𝑡) = 𝐄̃0r exp i(𝐤 r ⋅ 𝒓 − ωr 𝑡)
𝑬
8-29
̃ t (𝒓, 𝑡) = 𝐄̃0t exp i(𝐤 t ⋅ 𝒓 − ωt 𝑡)
𝑬
8-30
For a very thin interface, we can assume there is no decay of the waves across the interface so we need only consider
the real 𝐤 terms.

𝐸̃0i// exp [i((𝒌𝑖 ⋅ 𝑹1 − ωi t) + 𝐸̃0r// exp [i((𝒌𝑟 ⋅ 𝑹1 − ωr t) = 𝐸̃0t// exp [i((𝒌𝑡 ⋅ 𝑹2 − ωt t) 8-31

𝑹1 = 𝑹2 = 𝑹 8-32

Now for simplicity, we assume the origin of the coordinate system is in the interfacial plane and consider special
cases:

i) 𝑹=0

̃0i// e−iωit + E
E ̃0r// e−iωr t = E
̃0t// e−iωtt 8-33

E0i// cos(ωi t + 𝜙𝑖 ) + +E0r// cos( ωr t + 𝜙𝑟 ) = E0t// cos( ωt t + 𝜙𝑡 )

8-34

ωi = ωr = ωt ,
8-35

𝜙𝑖 = 𝜙𝑟 = 𝜙𝑡
8-36

ii) t=0

E0i// ei𝐤i⋅𝑹 + E0r// ei𝐤r ⋅𝑹 = E0t// ei𝐤 t⋅𝑹

8-37

This condition is true for all 𝑹 in the plane. The E0// terms are all real constants so:

⇒ 𝐤i ⋅ 𝑹 = 𝐤r ⋅ 𝑹 = 𝐤t ⋅ 𝑹 8-38

Consider reflection (i.e. 𝐤 i ⋅ 𝑹 = 𝐤 r ⋅ 𝑹):

(𝐤 i − 𝐤 r ) ⋅ 𝑹 = 0
8-39

⇒ (𝐤 i − 𝐤 r ) is normal to the interface

8-40

pg. 29 Confidential © Damian Peter Hampshire 2023. All rights reserved.

 k //i = k //r .
8-41

Since the phase velocities are also the same,

|𝐤 i | = |𝐤 r | 8-42

⇒ θi = θr − 1st law of geometrical optics: Angle of incidence = Angle of reflection.

8-43

F2F Lecture 19 -2020

Consider transmission (i.e. 𝐤 i ⋅ 𝑹 = 𝐤 t ⋅ 𝑹):

(𝐤 i − 𝐤 t ) ⋅ 𝑹 = 0 8-44

⇒ k //i = k //t
8-45

k i sinθi = k t sinθt
8-46

𝑐 8-47
𝑛=
𝑣𝑝ℎ𝑎𝑠𝑒
cki ckt
gives ni = and nt = and therefore
ω ω

ni sinθt
= − 2nd law of geometrical optics (Snell’s law).
nt sinθi 8-48

Note that Snell’s law is valid for all polarisations.

𝐤 i , 𝐤 r , and 𝐤 t are all in the plane of incidence (3rd law of geometrical optics).
8-49

8.4 Fresnel’s equations for insulating magnetic and dielectric materials..

8.4.1 E-field parallel to the plane of incidence

pg. 30 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 57 : An electromagnetic wave incident on an interfacial plane where E is parallel to the plane of
incidence

8-50
(E0i + E0r )cosθi = E0t cosθt

H0i − H0r = H0t

8-51

𝒌 × 𝑬 = ω𝑩
8-52

E
B=
vphase 8-53

E μ
=√ 8-54
𝐻 𝜀

𝜀∗
⇒ (E0i − E0r ) = √ E 8-55
𝜇 ∗ 0t

pg. 31 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Then combining (1) and (2) to eliminate Eot and Eor in turn, we obtain the important ratios:

𝜀∗
Eor cosθt − √𝜇 ∗ cosθi
{ } =
Eoi E−parallel 𝜀∗ 8-56
cosθt + √𝜇 ∗ cosθi

Eot 2cosθi
{ } =
Eoi E−parallel 𝜀∗
cosθt + √𝜇 ∗ cosθi 8-57

Fresnel’s equations for E// to the plane of incidence

Figure 58 : The Fresnel coefficients for a wave with the E - field parallel and normal to the plane of
incidence for an air-to-water interface.

pg. 32 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Figure 59: Top: The reflection and transmission of light into the air that has originated in the water.
Bottom: The Fresnel coefficients for a wave with the E - field parallel and normal to the plane of
incidence for a water-to-air interface .

pg. 33 Confidential © Damian Peter Hampshire 2023. All rights reserved.

8.4.2 E-field normal to the plane of incidence.

Figure 60: An electromagnetic wave incident on an interfacial plane where E is normal to the plane of
incidence.

The continuity equations for H// and E// are now,

8-58
E0i + E0r = E0t (1)

H0i cosθi − H0r cosθr = H0t cosθt

8-59

𝜀∗
(E0i − E0r )cosθi = √ E cosθt
𝜇 ∗ 0t 8-60

Then combining (1) and (2) to eliminate Eot and Eor in turn, we obtain the important ratios:
𝜀∗
cosθ − √
Eor i 𝜇 ∗ cosθt
{ } =
Eoi E−orthogonal 𝜀∗ 8-61
cosθi + √𝜇 ∗ cosθt

Eot 2cosθi
{ } =
Eoi E−orthogonal 𝜀∗ 8-62
cosθi + √𝜇 ∗ cosθt

pg. 34 Confidential © Damian Peter Hampshire 2023. All rights reserved.

Fresnel’s second pair of equations – E orthogonal to plane of incidence.

A microscopic understanding of the Brewster angle.

9 Surviving University

Figure 61 : Commenting on Orwell’s maxim near his statue outside the BBC in London: ‘If you value liberty, you
have a responsibility to understand those who say things you do not want to hear’.

pg. 35 Confidential © Damian Peter Hampshire 2023. All rights reserved.