Waves and Electromagnetism Wavefunctions Lecture Notes: Alessandro de Angelis University of Udine, December 2012

Waves and Electromagnetism
Wavefunctions
Lecture notes
Alessandro De Angelis
University of Udine, December 2012

Contents
1 Differential calculus applied to vector fields 5

1.1 Scalar and vector fields . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 The gradient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 The divergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.4 The curl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.5 Second order derivatives, and the Laplacian . . . . . . . . . . . . 9
1.6 The differential equation of heat flow . . . . . . . . . . . . . . . . 10
1.7 Differential operators in polar coordinates . . . . . . . . . . . . . 11
1.8 The gradient theorem . . . . . . . . . . . . . . . . . . . . . . . . 12
1.9 The divergence theorem (Gauss’ theorem) . . . . . . . . . . . . . 14
1.9.1 Flux of a vector field . . . . . . . . . . . . . . . . . . . . . 14
1.9.2 Flux through a closed surface . . . . . . . . . . . . . . . . 15
1.9.3 The flux from a cube . . . . . . . . . . . . . . . . . . . . . 17
1.9.4 Heat conduction; the diffusion equation . . . . . . . . . . 19
1.10 The curl theorem (Stokes’ theorem) . . . . . . . . . . . . . . . . 19
1.10.1 Circulation of a vector field . . . . . . . . . . . . . . . . . 19
1.10.2 The circulation around a square . . . . . . . . . . . . . . 20
1.11 Curl-free fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.12 Dirac’s delta function . . . . . . . . . . . . . . . . . . . . . . . . 23
I Waves and electromagnetism 25

2 Waves 26
2.1 Translations and the wave equation . . . . . . . . . . . . . . . . . 27
2.1.1 An example: mechanical waves in one dimension . . . . . 28
2.2 Energy transported by a wave . . . . . . . . . . . . . . . . . . . . 29
2.3 Sinusoidal waves and the Fourier theorem . . . . . . . . . . . . . 30
2.4 Amplitude, wavelength, period . . . . . . . . . . . . . . . . . . . 31
2.5 Transverse and longitudinal waves . . . . . . . . . . . . . . . . . 32
2.6 An example: sound . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.7 The classical Doppler effect . . . . . . . . . . . . . . . . . . . . . 33
2.7.1 Redshift of galaxies and the expansion of the Universe . . 34
2.7.2 The Vavilov-Cherenkov effect . . . . . . . . . . . . . . . . 37
1
2.8 Composition of waves . . . . . . . . . . . . . . . . . . . . . . . . 38
2.8.1 Boundary conditions and steady waves . . . . . . . . . . . 38
2.8.2 Beats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.8.3 Group velocity and phase velocity . . . . . . . . . . . . . 42
2.9 Waves in three dimensions . . . . . . . . . . . . . . . . . . . . . . 44
2.9.1 Spherical waves . . . . . . . . . . . . . . . . . . . . . . . . 44
3 Electromagnetism and the Maxwell’s equations 46

3.1 Charge density and current density . . . . . . . . . . . . . . . . . 47
3.2 Maxwell’s equations in differential form . . . . . . . . . . . . . . 47
3.3 Maxwell’s equations and continuity equation for charge . . . . . 48
3.4 The potentials, vector and scalar . . . . . . . . . . . . . . . . . . 49
3.5 Maxwell’s equations and electrostatics . . . . . . . . . . . . . . . 50
3.6 Maxwell’s equations and magnetostatics . . . . . . . . . . . . . . 50
4 Solutions of the Maxwell’s equations in vacuo 51

4.1 Maxwell’s waves and light . . . . . . . . . . . . . . . . . . . . . . 52
4.2 Properties of the electromagnetic waves . . . . . . . . . . . . . . 54
4.2.1 Energy transported an the electromagnetic wave . . . . . 56
4.2.2 The Poynting vector . . . . . . . . . . . . . . . . . . . . . 57
4.3 Photoelectric effect; the photon hypothesis . . . . . . . . . . . . . 57
4.4 The perception of electromagnetic waves: visible light . . . . . . 60
4.5 Natural units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5 Geometrical optics 64
5.1 Light propagation through different materials; transmission of
electromagnetic waves . . . . . . . . . . . . . . . . . . . . . . . . 64
5.2 Huygens’ principle . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.3 The laws of reflection and refraction . . . . . . . . . . . . . . . . 65
5.3.1 The Fermat principle . . . . . . . . . . . . . . . . . . . . . 68
6 Interference and diffraction 70

6.1 Young’s interference experiment . . . . . . . . . . . . . . . . . . . 70
6.1.1 The two-slit case . . . . . . . . . . . . . . . . . . . . . . . 74
6.2 Diffraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.2.1 Diffraction through a single slit . . . . . . . . . . . . . . . 76
6.2.2 Diffraction grating . . . . . . . . . . . . . . . . . . . . . . 77
6.2.3 The diffraction limit . . . . . . . . . . . . . . . . . . . . . 78
7 Maxwell’s equations and Einstein’s special relativity 79

7.1 Classical electromagnetism is not a consistent theory . . . . . . . 79
7.2 Galilean transformations, relativity, and the ether . . . . . . . . . 80
7.2.1 Maxwell’s equations and the ether . . . . . . . . . . . . . 81
7.3 The Michelson-Morley experiment . . . . . . . . . . . . . . . . . 81
7.4 Einsteins’s postulates and relativity . . . . . . . . . . . . . . . . 83
7.4.1 Relativity of simultaneity . . . . . . . . . . . . . . . . . . 84
2
7.4.2 Time dilation . . . . . . . . . . . . . . . . . . . . . . . . . 86
7.4.3 Length contraction . . . . . . . . . . . . . . . . . . . . . . 87
7.5 Invariance of the interval . . . . . . . . . . . . . . . . . . . . . . . 87
8 Lorentz transformations and the formalism of special relativity 88

8.1 The Lorentz transformations . . . . . . . . . . . . . . . . . . . . 88
8.1.1 A theorem by von Ignatowski . . . . . . . . . . . . . . . . 88
8.1.2 Transformation of velocities . . . . . . . . . . . . . . . . . 88
8.1.3 The relativistic Doppler effect . . . . . . . . . . . . . . . . 88
8.2 4-vectors; covariant and controvariant representation . . . . . . . 90
8.2.1 Covariant derivatives . . . . . . . . . . . . . . . . . . . . . 92
8.2.2 Four-dimensional velocity . . . . . . . . . . . . . . . . . . 93
8.3 Spacelike and timelike events; future and past . . . . . . . . . . . 94
8.4 E=mc2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
8.5 4-momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
8.6 The photon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
9 Covariant formulation of electromagnetism 95

9.1 Relativistic force . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
9.2 Electromagnetism and relativity . . . . . . . . . . . . . . . . . . 95
9.2.1 The equations for the potentials . . . . . . . . . . . . . . 95
9.2.2 The electromagnetic tensor . . . . . . . . . . . . . . . . . 97
9.2.3 Covariant expression of Maxwell’s equations . . . . . . . . 98
9.3 The transformation of the fields . . . . . . . . . . . . . . . . . . . 98
9.3.1 Invariants . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
9.4 Covariant expression of the Lorentz force . . . . . . . . . . . . . 98
9.4.1 Motion of a particle under a constant force . . . . . . . . 98
9.5 Radiation from an accelerated particle . . . . . . . . . . . . . . . 98
II Wave functions; introduction to quantum physics 99

10 The crisis of classical physics 100
10.1 The quantum properties of radiation . . . . . . . . . . . . . . . . 100
10.1.1 The photoelectric effect . . . . . . . . . . . . . . . . . . . 100
10.1.2 The Compton effect . . . . . . . . . . . . . . . . . . . . . 100
10.1.3 Blackbody radiation . . . . . . . . . . . . . . . . . . . . . 100
10.1.4 Pair production . . . . . . . . . . . . . . . . . . . . . . . . 100
10.2 The wave properties of matter . . . . . . . . . . . . . . . . . . . . 100
10.2.1 Diffraction of electrons . . . . . . . . . . . . . . . . . . . . 100
10.2.2 de Broglie’s wavelength . . . . . . . . . . . . . . . . . . . 100
10.3 Discrete versus continuum phenomena: atomic spectra . . . . . . 100
3
11 The Schrödinger equation 101
11.1 Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
11.2 Interpretation of the wavefunction . . . . . . . . . . . . . . . . . 101
11.3 The operation of measurement; collapse of the wavefunction . . . 101
11.4 Reality, ortodoxy, agnosticism . . . . . . . . . . . . . . . . . . . . 101
11.5 Expectation values . . . . . . . . . . . . . . . . . . . . . . . . . . 101
11.6 Momentum, and operators . . . . . . . . . . . . . . . . . . . . . . 101
11.6.1 Angular momentum . . . . . . . . . . . . . . . . . . . . . 101
11.7 The Hamiltonian operator . . . . . . . . . . . . . . . . . . . . . . 102
11.7.1 * A quantum view of Nöther’s theorem . . . . . . . . . . 102
12 Solving the Schrödinger equation in one dimension 103

12.1 Decoupling the space and time parts of the equation . . . . . . . 103
12.2 Free particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
12.3 The infinite square well . . . . . . . . . . . . . . . . . . . . . . . 103
12.3.1 Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . 103
12.4 Potential step . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
12.5 The finite square well . . . . . . . . . . . . . . . . . . . . . . . . 103
12.6 Potential barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
12.6.1 Reflection and Transmission . . . . . . . . . . . . . . . . . 103
12.6.2 Tunneling . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
12.7 * The harmonic oscillator . . . . . . . . . . . . . . . . . . . . . . 103
13 Central potentials and the Hydrogen atom 104

13.1 The Schrödinger equation in 3 dimensions . . . . . . . . . . . . . 104
13.2 Spherically symmetric potentials . . . . . . . . . . . . . . . . . . 104
13.3 Separation of variables . . . . . . . . . . . . . . . . . . . . . . . . 107
13.4 The angular part . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
13.4.1 Angular momentum . . . . . . . . . . . . . . . . . . . . . 107
13.5 Radial probability density . . . . . . . . . . . . . . . . . . . . . . 108
13.6 The H atom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
4
Chapter 1
Differential calculus applied

to vector fields
1.1 Scalar and vector fields

The concept of field formalizes physical properties of space. To each point of
space (let us start with the usual 3-dimensional space) a physical quantity is
associated. Mathematically, a physical field is a function whose domain is space.
The simplest possible physical field is a scalar field. By a scalar field we
mean a field which is characterized at each point by a single number – a scalar.
Of course the number may change in time; we shall talk for now about what
the field looks like at a given instant.
As an example of a scalar field, let us consider a solid block of material which
has been heated at some places, so that the temperature of the body varies
from point to point. Then the temperature will be a function of x, y, and z, the
position in space measured in a rectangular coordinate system. Temperature
T (x, y, z) is a scalar field.
One way of picturing scalar fields is to draw contours, which are imaginary
surfaces through points for which the field has the same value, just as contour
lines on a map connect points with the same height. For a temperature field
the contours are called “isothermal surfaces” or isotherms. Figure 8.1 illustrates
a temperature field and shows the dependence of T on x and y when z = 0.
Several isotherms are drawn.
There are also vector fields. In this case, a vector is given for each point in
(a region of) space. As an example, consider a rotating body: the velocity of
the material of the body at any point is a vector which is a function of position.
As a second example, consider the flow of heat in a block of material. If the
temperature in the block is high at one piece and low at another, there will be a
flow of heat from the hotter places to the colder ones. The heat will be flowing
in different directions in different parts of the block.
The heat flow is a directional quantity which we call ~h. Its magnitude is
5
Figure 1.1: The temperature field; isotherms.
a measure of how much heat is flowing, and can be defined as the amount
of thermal energy that passes, per unit time and per unit area, through an
infinitesimal surface element, perpendicular to the direction of flow. The vector
points in the direction of flow. In symbols: if ∆J is the thermal energy that
passes per unit time through the surface element ∆a, then
~h = lim ∆J ~ef (1.1)

∆a→0 ∆a
where ~ef is the versor of the heat flow. Examples of the heat flow vector are
also shown in Figure 8.1.
1.2 The gradient

~ (x, y, z))
For a real-valued function T (x, y, z) on <3 , the gradient ∇T (x, y, z) (or ∇T
3
is a function on < , that is, its value at a point (x, y, z) is the triple

∂T ∂T ∂T ∂T ∂T ∂T
∇T (x, y, z) = , , = i+ j+ k
∂x ∂y ∂z ∂x ∂y ∂z
in <3 , where each of the partial derivatives is evaluated at the point (x, y, z).
One can think of the symbol ∇ as being “applied” to a real-valued function T
to produce a 3-dimensional function ∇T .
Is ∇T (x, y, z) a vector? Of course it is not generally true that any three
numbers form a vector. It is true only if, when we rotate the coordinate system,
6
the components of the vector transform among themselves in the correct way
for a vector. So it is necessary to analyze how these derivatives are changed by
a rotation of the coordinate system. We shall show that ∇T (x, y, z) is indeed
a vector: the derivatives do transform in the correct way when the coordinate
system is rotated. We can observe that, by the total differential theorem,
∂T ∂T ∂T
dT = dx + dy + dz (1.2)
∂x ∂y ∂z
for any d~r = (dx, dy, dz). And since df is a scalar and d~r is a vector, ∇T (we
can now call it ∇T~ ) must be a vector (of course a demonstration based on the
transformations of coordinates is also possible, but ennoying).
Is ∇~ a vector? Strictly speaking, no, since ∂ , ∂ and ∂ are not numbers.
∂x ∂y ∂z
But it helps to think of it as a vector, as we shall see. The process of “applying”
∂ ∂ ∂
∂x , ∂y , ∂z to a real-valued function T (x, y, z) can be thought of as multiplying
the quantities:

∂ ∂T ∂ ∂T ∂ ∂T
(T ) = , (T ) = , (T ) =
∂x ∂x ∂y ∂y ∂z ∂z
~ is often referred to as the “del operator”, since it “operates”

For this reason, ∇
on functions.
We can write the expression (1.2) as
~ ) · (d~r) .
dT = (∇f (1.3)
~ represents the spatial rate of change of T . The x−component of ∇T

∇T ~ shows
how fast T changes in the x−direction, and in the same way for the other
coordinates. What is the direction of the vector ∇T ~ ? Equation (1.3) shows
that the rate of change of T in any direction is the component of ∇T ~ in that
~ is that in which it has the largest
direction. It follows that the direction of ∇T
possible component- in other words, the direction in which T changes the fastest.
~ is
The gradient of T has the direction of the steepest uphill slope (in T ). ∇T
perpendicular to the isotherms.
What would it mean for the gradient to vanish? If ∇T ~ = 0 at (x, y, z),
then dT = 0 for small displacements about the point (x, y, z). This is, then, a
stationary point of the function T (x, y, z). It could be a maximum (a summit),
a minimum (a valley), a saddle (a pass).
1.3 The divergence

What if we make the ∇ ~ operator to operate on a vector field (a function of <3
3
into < ) instead than on a scalar field? In this case we when two possible kinds
of products, the scalar product and the vector product; we shall see that both
correspond to interesting functions.
7
Figure 1.2: Divergence.
Giving a vector function F~ defined in a domain in <3 , we define as the

divergence of F~ the formal scalar product ∇
~ and F~ :

~ ~ ∂Fx ∂Fy ∂Fz
∇·F = + + , (1.4)
∂x ∂y ∂z
and it is clearly a scalar.
What is its geometrical interpretation? ∇ ~ · F~ is a measure of how much the
~
vector F spreads out (diverges) from a point. For example, the vector function
in Figure 1.2(a) has a positive divergence (if the arrows pointed in, it would be
a large negative divergence), the function in Figure 1.2(b) has zero divergence,
and the function in Figure 1.2(c) again has a positive divergence.
Exercise: if the function pictured in Figure 1.2(a) is F~(a) = x~i + y~j, and the
function pictured in Figure 1.2(b) is F~(b) = ~j, compute their divergence.
1.4 The curl

~ times a vector field F~ .
In a similar way, we can define the vector product of ∇
This is called the curl of F~ :
~ux ~uy ~uz
~ × F~ = ∂/∂x
∇ ∂/∂y ∂/∂z . (1.5)
Fx Fy Fz
This can be demonstrated to be a vector function.
8
Figure 1.3: Curl.
Geometrical interpretation: ∇~ × F~ is a measure of how much the vector F~

“curls around” the point in question. Thus the three functions in Figure 1.2 all
when zero curl (as you can easily check for yourself), whereas the functions in
Figure 1.3 when a nonzero curl, pointing in the z−direction, as the right-hand
rule would suggest. A whirlpool would be a region of large curl.
Exercise: if the function pictured in Figure 1.3(a) is F~(a) = −y~i + x~j, and
the function pictured in Figure 1.3(b) is F~(b) = x~j, compute their curl.
1.5 Second order derivatives, and the Laplacian

So far we when had only first derivatives. Why not second derivatives? We can
write several combinations:
~ × (∇T
1. ∇ ~ )
~ · (∇
2. ∇ ~ × F~ )
~ · (∇T
3. ∇ ~ )
~ ∇
4. ∇( ~ · F~ )
~ × (∇
5. ∇ ~ × F~ )
You can check that these are all the legal combinations.
1. and 2. ∇ ~ × (∇T
~ ) and ∇ ~ · (∇
~ × F~ ). The first two are identically zero. Let
us check it for the first one: one has, by the Schwartz’s lemma:
~ux ~uy ~uz

~
~ × ∇T
~ =
∇ ∂/∂x ∂/∂y ∂/∂z = 0.
∂T /∂x ∂T /∂y ∂T /∂z
We when thus demonstrated that the curl of a gradient is zero, which is easy
to remember because of the way the vectors work. It can be demonstrated that
9
if the curl of a field F~ is zero, then this field is always the gradient of a scalar
field: there is some scalar field φ such that F~ = ∇φ.~
Also ∇~ · (∇~ × F~ ) = 0, and there is a similar theorem stating that if the
divergence of F~ is zero, F~ is the curl of some vector field A.
~
3. ∇~ · (∇T
~ ). Let us examine the third expression now. For a real-valued
function f (x, y, z), the Laplacian1 of T , denoted by ∇2 T , is defined as
2 2 2
∇2 T (x, y, z) = ∇ ~ )= ∂ T +∂ T +∂ T .
~ · (∇T (1.6)
∂x2 ∂y 2 ∂z 2
Sometimes the notation ∆T is used instead.
The Laplacian can be thought as a scalar operator
2
∂2 ∂2

2 ∂
∇ = + 2+ 2
∂x2 ∂y ∂z
and as such it can be applied to vector functions as well, giving as output a

vector:
~ y, z) = ∇2 Ax (x, y, z), ∇2 Ay (x, y, z), ∇2 Az (x, y, z) .
∇2 A(x,

The explicit calculation confirms that this approach is justified.
~ ∇
4. ∇( ~ · F~ ). It is a possible vector field, which may occasionally come up (for
example, see next point).
~ × (∇
5. ∇ ~ × F~ ). Let us compare this expression with the vector identity
~ × (B
A ~ × C)
~ = B(
~ A~ · C)
~ − (A
~ · B)
~ C~.
~ and B
In order to use this formula, we should replace A ~ by the operator ∇
~ and
~ ~
put C = F . If we do that, we get
~ × (∇
∇ ~ × F~ ) = ∇(
~ ∇~ · F~ ) − ∇2 F~ . (1.7)
1.6 The differential equation of heat flow

Let us give an example of a law of physics written in vector notation. For heat
conductors, the energy flows through the material from a surface at temperature
T2 to a surface at temperature T2 < T1 . The total energy flow is proportional
to the area A of the faces, and to the temperature difference; it is also inversely
1 Pierre-Simon de Laplace (1749 - 1827) was a French mathematician and astronomer whose
work was pivotal to the development of mathematical astronomy and statistics. He was one
of the first scientists to postulate the existence of black holes and the notion of gravitational
collapse.
10
proportional to d, the distance between the plates (for a given temperature
difference, the thinner the slab the greater the heat flow). Letting J be the
thermal energy that passes per unit time through the slab, we write
J = k(T2 − T1 ) A/d .
The constant of proportionality k is called the thermal conductivity of the ma-

terial.
For a small slab of area ∆a the heat flow per unit time is
∆a
∆J = k∆T (1.8)
∆d
where ∆d is the thickness of the slab. ∆J/∆a is what we when defined earlier
as the magnitude of the vector ~h, whose direction is the heat flow. The heat
flow will be perpendicular to the isotherms. Also, ∆T /∆d is just the rate of
change of T with position. And since the position change is perpendicular to the
isotherms, ∆T /∆d is the maximum rate of change. It is, therefore, in the limit
~ Since the direction of ∇T
for ∆d → 0, just the magnitude of ∇T. ~ is apposite to
~
that of h, we can write the previous equation as a vector equation:
~h = −k ∇T
~
(the minus sign is necessary because heat flows “downhill” in temperature.)

This is the differential equation of heat conduction.
1.7 Differential operators in polar coordinates

Often – for example, in the case of central symmetries – it is convenient to
use radial coordinate systems when dealing with quantities such as the gradi-
ent, divergence, curl and Laplacian. We will present the expressions for these
operators in spherical coordinates.
A point (x, y, z) can be represented in spherical (polar) coordinates (r, θ, φ),
where x = r sin θ cos φ, y = r sin θ sin φ, z = r cos θ. θ (the angle down from the
z axis) is called the polar angle, and φ (the angle around from the x axis) is the
azimuthal angle.
At each point (r, θ, φ), ~er , ~eθ , ~eφ are unit vectors in the direction of increasing
r, θ, φ, respectively (see Figure 1.4). Then the vectors ~er , ~eθ , ~eφ are orthonormal.
By the right-hand rule, we see that ~eθ × ~eφ = ~er .
We can summarize the expressions for the gradient and the Laplacian applied
to a scalar field T and to a vector field F~ in spherical coordinates in the following
equations:
~ = ∂T ~er + 1 ∂T ~eθ + 1 ∂T ~eφ
∇T (1.9)
∂r r ∂θ r sin θ ∂φ
∂2T

2 1 ∂ 2 ∂T 1 ∂ ∂T 1
∇ T = 2 r + 2 sin θ + 2 2 . (1.10)
r ∂r ∂r r sin θ ∂θ ∂θ r sin θ ∂φ2
11
Figure 1.4: Spherical coordinates.
Figure 1.5: Line integral.
The derivation of the above formulas is conceptually straightforward but

boring. The basic idea is to take the Cartesian equivalent of the quantity in
question and to substitute into that formula using the appropriate coordinate
transformation.
1.8 The gradient theorem

We found previously that there were various ways of taking derivatives of fields.
Some gave vector fields; some gave scalar fields. Although we developed many
different formulas, everything could be summarized in one rule: the operators
∂/∂x, ∂/∂y, ∂/∂z are the three components of a vector operator ∇. ~ We would
now like to get some understanding of the significance of the derivatives of fields.
We shall then when a better feeling for what a vector field equation means.
We when already discussed the meaning of the gradient operation (appli-
cation of ∇~ on a scalar). We take up now an integral formula involving the
12
Figure 1.6: Line integral and gradient theorem.
gradient. The relation contains a very simple idea: since the gradient repre-
sents the rate of change of a field quantity, if we integrate that rate of change,
we should get the total change (like in the fundamental theorem of calculus).
Let us first introduce the concept of line integral for a vector field. The line
integral from a point a to a point b of the curve is nothing but the integral of
the dot product of the value of the function times the line element (i.e., at each
point the value of the function is weighted by the cosine of the angle formed by
the vector itself and by the tangent to the line):
Z b
~
E(x, y, z) · d~l
a
(see Figure 1.5). For example, the work done by the force F~ from a to b along
a given path is the line integral of F~ along that path.
Suppose we have a scalar function of three variables T (x, y, z). Starting at
point a, we move by a small distance d~l1 (Figure 1.6). The function T will
change by an amount
dT = ∇T~ · d~l1 .
Now we move a little further, by an additional small displacement d~l2 ; the

~ ·d~l2 . In this way, proceeding by infinitesimal
incremental change in T will be ∇T
steps, we make a journey to point b. At each step we compute the gradient of T
(at that point) and dot it into the displacement d~l: this gives us the change in
T. Evidently the total change in T in going from a to b along the path selected
is Z b
~ · d~l = T (~b) − T (~a) .
∇T
a
This is called the fundamental theorem for gradients; like the “ordinary” fun-
damental theorem of calculus, it says that the integral (here a line integral) of
a derivative (here the gradient) is given by the difference between the values of
the function at the boundaries (a and b).
13
A geometrical interpretation will make use of an example. Suppose you
wanted to determine the height of the Eiffel Tower. You could climb the stairs,
using a ruler to measure the rise at each step, and adding them all up, or you
could place altimeters at the top and the bottom, and subtract the two readings;
you should get the same answer either way.
Line integrals ordinarily depend on the path taken from a to b. But the right
side of in the gradient theorem makes no reference to the path - only to the end
points. Evidently, gradients when the special property that their line integrals
are path independent:
Rb
Corollary 1: a (∇T ~ ) · d~l is independent of the path taken from a to b.
Corollary 2: (∇T )·d~l = 0, since the beginning and end points are identical,
~
H
and hence T (b) − T (a) = 0.
1.9 The divergence theorem (Gauss’ theorem)

1.9.1 Flux of a vector field
A surface integral is an expression of the form
Z
~h · ~n da
S
where ~h is a vector function, and da is an infinitesimal patch of area; ~n is a

unit vector with direction perpendicular to the surface. There are, of course,
two directions perpendicular to any surface, so the sign of a surface integral
is intrinsically ambiguous. If the surface is closed (forming a “balloon”), one
frequently puts a circle on the integral sign
I
~h · ~n da ;
S
in this case tradition dictates that “outward” is positive, but for open surfaces
it is, again, arbitrary.
We will identify sometimes a flat surface with a vector perpendicular to the
surface itself, and with intensity equal to the area of the surface. The expression
of the surface integral becomes then
Z
~h · d~a .
S
If ~h describes the flow of a fluid (mass per unit area per unit time), then
~h · d~a represents the total mass per unit time passing through the surface -
R
S
hence the alternative name, flux Φ. In the case of heat flow, we may think: ~h is
the “current density” of heat flow and the surface integral of it is the total heat
current directed out of the surface: that is, the thermal energy per unit time
(joule per second).
14
Figure 1.7: The closed surface S defines the volume V. The unit vector ~n is the
outward normal to the surface element da, and ~h is the heat-flow vector at the
surface element.
We can generalize this idea to the case to any vector field; for instance, it
might be the electric field. We can certainly still integrate the normal component
of the electric field over an area if we wish. Although it does not appear to be
the flow of anything, we still call it the “flux”. We say flux of E~ through the
surface S the quantity Z
Φ= E~ · d~a .
S
We thus generalize the word “flux” to mean the “surface integral of the normal
component” of a vector.
1.9.2 Flux through a closed surface

We defined the vector ~h, which represents the heat through a unit area in a
unit time. Suppose that inside a block of material we have some closed surface
S which encloses the volume V. We would like to find out how much heat is
flowing out of this volume. We can, of course, find it by calculating the total
heat flow out of the surface S (Figure 1.7).
We write da for the area of an element of the surface. The symbol stands for
a two-dimensional differential. If, for instance, the area happened to be in the
xy−plane, we would have da = dxdy (later we shall have integrals over volume
and for these it is convenient to consider a differential volume that is a little
cube; so when we write dV we mean dV = dxdydz)2 .
The heat flow out through the surface element da is the area times the
component of ~h perpendicular to da. Returning to the special case of heat flow,
let us take a situation in which heat is conserved. For example, imagine some
2 Some texts write d2 a instead of da to remind that it is kind of a second-order quantity.
They also write d3 V instead of dV. We will use the simpler notation, and assume that you
remember that an area has two dimensions and a volume has three.
15
Figure 1.8: A volume V contained inside the surface S is divided into two
pieces by a “cut” at the surface Sab . We now when the volume V1 enclosed in
the surface S1 = Sa + Sab and the volume V2 enclosed in the surface
S2 = Sb + Sab .
material in which after an initial heating no further heat energy is generated

or absorbed. Then, if there is a net heat flow out of a closed surface, the heat
content of the volume inside must decrease. So, in circumstances in which heat
would be conserved, we say that
I
~h · d~a = − dQ , (1.11)
dt
where Q is the heat inside the surface.
We shall point out an interesting fact about the flux of any vector. Imagine
that we have a closed surface S that encloses the volume V. We now separate
the volume into two parts by some kind of a “cut”, as in Figure 1.8. Now we
have two closed surfaces and volumes: the volume V1 is enclosed in the surface
S1 , which is made up of part of the original surface Sa and of the surface of the
cut, Sab ; the volume V2 is enclosed by S2 , which is made up of the rest of the
original surface S, and closed off by the cut Sab .
The sum of the fluxes through S1 and S2 equals the flux through the whole
surface that we started with. The flux through the part of the surfaces Sab
common to both S1 and S2 just exactly cancels out. For the flux of a generic
vector C~ out of V1 , we can write
Z Z
ΦS1 = ~
C · ~n da + C~ · ~n1 da
Sa Sab
and for the flux out of V2

Z Z
ΦS2 = ~ · ~n da +
C ~ · ~n2 da .
C
Sb Sab
Since ~n1 = −~n2 , the sum of the fluxes through S1 and S2 is just the sum of two
integrals which, taken together, give the flux through the original surface S.
16
~ out of a small cube.
Figure 1.9: Computation of the flux of C
We can similarly subdivide again the volume - say by cutting V1 into two
pieces. You see that the same arguments apply. So for any way of dividing
the original volume, it is generally true that the flux through the outer surface,
which is the original integral, is equal to a sum of the fluxes out of all the little
interior pieces. Then we can divide it by a large number of small cubes.
1.9.3 The flux from a cube

We now take the special case of a small cube (or parallelepiped) and find an
interesting formula for the flux out of it.
Consider a cube whose edges are lined up with the axes (Figure 1.9). Let
us suppose that the coordinates of the corner nearest the origin are x, y, z.
Let ∆x be the length of the cube in the x−direction, ∆y be the length in the
y−direction, and ∆z be the length in the z−direction (they are small quantities).
We wish to find the flux of a vector field C~ through the surface of the cube. We
shall do this by making a sum of the fluxes through each of the six faces.
First, consider the face marked 1 in the figure. The flux outward on this
face is the negative of the x−component of C, ~ integrated over the area of the
face. This flux is approximately
Φ1 = −Cx (1) ∆y∆z .
(since we are considering a small cube, we approximate the integral by the value
of Cx at the center of the face - which we call the point (1) - multiplied by the
17
area of the face). Similarly, the flux out of face 2 is approximately
Φ2 = Cx (2) ∆y∆z .
Cx (1) and Cx (2) are, in general, slightly different. If ∆x is small enough, we
can write
∂Cx
Cx (2) = Cx (1) + ∆x
∂x
(there are, of course, higher order terms, but we are considering the limit for
∆x → 0). So the flux through faces 1 and 2 is
∂Cx
Φ1 + Φ 2 = ∆x∆y∆z .
∂x
The derivative should really be evaluated at the center of face 1; that is, at
(x, y + ∆y/2, z + ∆z/2). But in the limit of an infinitesimal cube, we make a
negligible error if we evaluate it at the corner (x, y, z).
Applying the same reasoning to each of the other pairs of faces, we find
∂Cy
Φ3 + Φ 4 = ∆x∆y∆z ,
∂y
and
∂Cz
Φ5 + Φ 6 = ∆x∆y∆z .
∂z
The total flux through all the faces is the sum of these terms. We find that
Z
~ · d~a = ∂Cx + ∂Cy + ∂Cz ∆x∆y∆z
C
cube ∂x ∂y ∂z
and so we can say that for an infinitesimal cube
dΦ = (∇ ~ · C)dV
~ .
We have shown that the outward flux from the surface of an infinitesimal
cube is equal to the divergence of the vector multiplied by the volume of the
cube. We now see the meaning of the divergence of a vector. The divergence of
~ per unit volume,
a vector at the point P is the flux - the outgoing “flow” of C
in the neighborhood of a point P .
We have connected the divergence to the flux out of an infinitesimal volume.
For any finite volume we can use the fact we proved above - that the total
flux from a volume is the sum of the fluxes out of each part. We can, that is,
integrate the divergence over the entire volume. This gives us the theorem that
the integral of the normal component of any vector over any closed surface can
also be written as the integral of the divergence of the vector over the volume
enclosed by the surface. This theorem is named after Gauss3 .
I Z
~
C · d~a = ~ · C)
(∇ ~ dV
S V
3 Carl Friedrich Gauss (Brunswick, 1777 - Göttingen, 1855) was a German mathematician,
generally regarded as one of the greatest mathematicians of all time for his contributions
to number theory, geometry, probability theory, geodesy, planetary astronomy, the theory of
functions, and potential theory (including electromagnetism). Gauss was the only child of
poor parents. He was rare among mathematicians in that he was a calculating prodigy.
18
~ around the curve Γ. Right: The
Figure 1.10: Left: the circulation of C
circulation around the whole loop is the sum of the circulations around the
two loops Γ1 and Γ2 .
where S is any closed surface and V is the volume inside it.
1.9.4 Heat conduction; the diffusion equation

~h · d~a = − dQ , becomes, if we call q
H
By the Gauss’ theorem, equation (1.11), dt
the amount of heat per unit volume,
d ~ · ~h)∆V
− (q∆V ) = (∇
dt
and thus we can transform the integral equation (1.11) into a differential equa-
tion defined locally, i.e., point by point:
dq ~ · ~h .
− =∇
dt
This kind of law appears frequently in physics: it is a conservation law, or a
continuity equation.
1.10 The curl theorem (Stokes’ theorem)

1.10.1 Circulation of a vector field
The circulation of a vector field is the line integral around a closed loop.
Playing the same kind of game we did with the flux, we can show that the
circulation around a loop is the sum of the circulations around two partial loops.
Suppose we break up our curve of Figure 1.10(left) into two loops, by joining
two points (1) and (2) on the original curve by some line that cuts across as
shown in Figure 1.10(right).
There are now two loops, Γ1 and Γ2 ; Γ1 is made up of Γa , which is that part
of the original curve to the left of (1) and (2), plus Γab , the “short cut”; Γ2 is
made up of the rest of the original curve plus the short cut.
19
Figure 1.11: Some surface bounded by the loop Γ is chosen. The surface is
divided into a number of small areas, each approximately a square. The
circulation around Γ is the sum of the circulations around the little loops.
The integral along Γab will have, for the curve Γ2 , the opposite sign compared
to Γ1 because the direction of travel is opposite - we must take both our line
integrals with the same “sense” of rotation. Following the same kind of argument
we used before, you can see that the sum of the two circulations will give just
the line integral around the original curve Γ: the parts due to Γab cancel. The
circulation around the one part plus the circulation around the second part
equals the circulation about the outer line.
We can continue the process of cutting the original loop into any number
of smaller loops. When we add the circulations of the smaller loops, there is
always a cancellation of the parts on their adjacent portions, so that the sum is
equivalent to the circulation around the original single loop.
Now let us suppose that the original loop is the boundary of some surface.
There are, of course, an infinite number of surfaces which all have the original
loop as the boundary. Our results will not, however, depend on which surface
we choose. First, we break our original loop into a number of small loops that
all lie on the surface we have chosen. No matter what the shape of the surface,
if we choose our small loops small enough, we can assume that each of the small
loops will enclose an area which is essentially flat. Also, we can choose our small
loops so that each is very nearly a square (Figure 1.11). Now we can calculate
the circulation around the big loop by finding the circulations around all of the
little squares and then taking their sum.
1.10.2 The circulation around a square

How shall we find the circulation for each little square? One question is, how is
the square oriented in space? We could easily make the calculation if it had a
special orientation. For example, if it were in one of the coordinate planes. Since
we have not assumed anything as yet about the orientation of the coordinate
axes, we can just as well choose the axes so that the one little square we are
concentrating on at the moment lies in the xy−plane, as in Figure 1.12. If our
result is expressed in vector notation, we can say that it will be the same no
20
Figure 1.12: Computing the circulation of C around a small square.
matter what the particular orientation of the plane.

We want now to find the circulation of the field C ~ around our little square. It
will be easy to do the line integral if we make the square small enough that the
vector C~ does not change much along any one side of the square (the assumption
is better the smaller the square, so we are really talking about infinitesimal
squares).
Starting at the point (x, y)−the lower left corner of the figure - we go around
in the direction indicated by the arrows. Along the first side - marked (1) - the
tangential component is Cx (1) and the distance is ∆x. The first part of the
integral is Cx (1)∆x. Along the second leg, we get Cy (2)∆y. Along the third, we
get −Cx (3)∆x, and along the fourth, −Cy (4)∆y. The minus signs are required
because we want the tangential component in the direction of travel. The whole
line integral is then
I
~ · d~l = (Cx (1) − Cx (3))∆x + (Cy (2) − Cy (4))∆y .
C
Since
∂Cx
Cx (3) ' Cx (1) + ∆y
∂y
and
∂Cy
Cy (4) ' Cy (2) + ∆x
∂x
one can write, at first order,
I
~ ~ ∂Cy ∂Cx
C · dl = − ∆x∆y .
∂x ∂y
21
Neglecting second order terms, the derivative can be evaluated at (x, y). The
above expression can be written in vector form:
I
~ · d~l = ∇
C ~ ×C~ · ∆~a .
The circulation around any loop Γ can now be easily related to the curl of the
vector field. We fill in the loop with any convenient surface S, as in Figure 1.11,
and add the circulations around a set of infinitesimal squares in this surface;
the sum can be written as an integral.
Our result is a very useful theorem called Stokes’ theorem4
I Z
~ · d~l = (∇
C ~ × C)
~ · d~a (1.12)
Γ S
where S is any surface bounded by Γ.

We must now speak about a convention of signs. The z−axis in Figure 1.12
would point toward the reader in a “usual”-that is, “right-handed”-system of
axes. When we took our line integral with a “positive” sense of rotation, we
found that the circulation was equal to the z−component of ∇~ × C.
~ If we had
gone around the other way, we would have gotten the apposite sign. Now how
shall we know, in general, what direction to choose for the positive direction
of the “normal” component of ∇ ~ × C?
~ The “positive” normal must always be
related to the sense of rotation by the “right-hand rule”.
1.11 Curl-free fields

We would like, now, to consider some consequences of our new theorems. Take
first the case of a vector whose curl is everywhere zero. Then Stokes’ theorem
says that the circulation around any loop is zero. Now if we choose two points
(1) and (2) on a closed curve, it follows that the line integral of the tangential
component from (1) to (2) is independent of which of the two possible paths is
taken. We can conclude that the integral from (1) to (2) can depend only on
the location of these points - that is to say, it is some function of position only.
The same logic was used where we proved that if the integral around a closed
loop of some quantity is always zero, then that integral can be represented as
the difference of a function of the position of the two ends. This fact allowed
us to invent the idea of a potential. We proved, furthermore, that the vector
field was the gradient of this potential function. It follows that any vector field
whose curl is zero is equal to the gradient of some scalar function. That is, if
~ ×C
∇ ~ = 0 everywhere, there is some scalar field ψ for which C ~ = ∇ψ.
~ We can,
if we wish, describe this special kind of vector field by means of a scalar field
(this shows how restrictive is the class of the conservative fields).
4 Sir George Gabriel Stokes (1819 - 1903) was a mathematician and physicist. Born in
Ireland, he spent all of his career at University of Cambridge, where he served as the Lucasian
Professor of Mathematics from 1849 until his death. Stokes made seminal contributions to
fluid dynamics (including the Navier-Stokes equations), optics, and mathematical physics.
22
Let’s show something else. Suppose we have any scalar field φ. If we take its
~ the integral of this vector around any closed loop must be zero:
gradient, ∇φ,
I
~ · d~l = 0 .
∇φ
loop
Using Stokes’ theorem, we conclude that

Z
~ × (∇φ)
∇ ~ · d~a = 0
surf ace
over any surface; but this means that the integrand is zero:
~ × (∇φ)
∇ ~ = 0.
We obtained in a different way a result we had obtained with vector algebra.

We stated, without demonstration, that a field C ~ with zero curl can be
expressed as the gradient of a scaler field. Now, thanks to the Stokes’ theorem,
we can demonstrate it. If the curl is zero, then the circulation along whatever
loop is zero; thus the line integral from a point A to a point B is independent
of the path. I can thus define, choosing arbitrarily a point A,
Z B
ϕ(B) = ~ · d~l
C
A
and thus
~ = ∇ϕ
C ~ .
Note that I can choose arbitrarily the point A; this reflects on the fact that,
adding an arbitrary constant to ϕ, the above relation is still valid.
1.12 Dirac’s delta function

The Dirac5 delta function is used to define fields which are nonzero in a region
of measure zero, but for which the integral over space is different from zero.
In one dimension, it can be defined as a real function δ(x) being zero in all
points apart from the origin, with the constraint
Z ∞
δ(x) dx = 1
−∞
(i.e., its value must be infinite at the origin, see Figure 1.13) for a sketch.
5 Paul Dirac (1902 -1984) was an English theoretical physicist who made fundamental con-
tributions to the early development of both quantum mechanics and quantum electrodynamics.
Among other discoveries, he formulated the Dirac equation, which describes the behaviour of
fermions, and predicted the existence of antimatter. Dirac shared the Nobel Prize in Physics
for 1933 with Erwin Schrödinger, for his contributions to Quantum Mechanics.
23
Figure 1.13: A sketch of the function δ(x).
In the mathematical literature it is known as a generalized function, or

distribution. It can be obtained as the limit of a class of legitimate functions,
for example, of the Gaussian functions:
1 x2
δ(x) = lim √ e− 2σ2 .
σ→0 σ 2π
Integrals including δ(x) are perfectly legitimate, and in particular:

Z ∞
δ(x − a)f (x) dx = f (a) .
−∞
It is easy to generalize the delta function to three dimensions:
δ 3 (~r) = δ(x)δ(y)δ(z) .
This three-dimensional delta function is zero everywhere except at (0, 0, 0),

where it blows up. Its volume integral is 1.
As in the one-dimensional case, integration with δ 3 (~r) picks out the value of
a function f at the location of the spike (if the spike is included in the domain
of integration): Z
δ(~r − ~a)f (~r) dV = f (~a) .
all space
24
Part I
Waves and
electromagnetism
25
Chapter 2
Waves
Until 1600, scientists were dreaming to give a complete description of nature

through the reduction to a set of elementary particles. An elementary particle is
ideally a point-like structure; the density of its charges (mass, electrical charge,
...) can be represented by a Dirac’s delta function. Isaac Newton1 in his studies
related to optics used only the concept of particle to treat light’s propagation.
In the XVII century, however, some particular phenomena were observed,
which could not be described by such model: their description needed the dis-
placement of a delocalized perturbation. Phenomena like destructive interfer-
ence (the sum of two perturbations could cancel) needed an extension of the
particle model. In order to explain such phenomena, the concept of wave was
formally introduced in 1600 in the Netherlands, where people had a long tradi-
tion in the field of optics. Christiaan Huygens2 was one of the main actors of
this intellectual revolution.
The wave is a model that describes the general propagation of a perturbation.
Typical examples of phenomena usually described as waves are:
• sound waves;
• elastic waves (local deformations);
1 Sir Isaac Newton (1642 - 1727) was an English physicist, mathematician, astronomer,
natural philosopher, alchemist and theologian, who has been considered by many to be the
greatest and most influential scientist who ever lived. His monograph Philosophiae Naturalis
Principia Mathematica, published in 1687, laid the foundations for most of classical mechanics.
Newton built the first practical reflecting telescope and developed a theory of colour based on
the observation that a prism decomposes white light into the many colours that form the visible
spectrum. He also formulated an empirical law of cooling and studied the speed of sound. In
mathematics, Newton shares the credit with Leibniz for the development of differential and
integral calculus. Newton was also deeply involved in occult studies and interpretations of
religion.
2 Christiaan Huygens (1629 - 1695) was a prominent Dutch mathematician, astronomer,
physicist and horologist. His work included early telescopic studies elucidating the nature of
the rings of Saturn and the discovery of its moon Titan, the invention of the pendulum clock
and other investigations in timekeeping, and studies of both optics and the centrifugal force.
Huygens achieved note for his argument that light consists of waves.
26
Figure 2.1: A perturbation moving right with speed c.
• waves on the water surface (gravity waves);

• electromagnetic waves.
The representation of a generic wave is a function ξ(x, y, z, t).
2.1 Translations and the wave equation

If
∂ξ/∂y = ∂ξ/∂z = 0 ,
then ξ(x, t) is called a plane wave; a function satisfying these conditions might
propagate along the x axis only.
Let us consider the function of the space coordinate ξ(x, 0) at a fixed time
t = 0 (we set it arbitrarily as the origin of time). Suppose that we want that after
a time t the same space function is displaced by ∆x = vt (i.e., the perturbation
represented by that function is moving in the positive direction of the x axis at
a speed v). The new function will be ξ(x − vt).
In the same way, the functional form ξ(x + vt) will represent a perturbation
moving left with speed v.
If we set u± = (x ± vt), then:
∂ξ dξ ∂u± dξ


 = =
 ∂x du± ∂x du±
∂ξ dξ ∂u ± dξ
= ±v


 =
∂t du± ∂t du±
27
Figure 2.2: Perturbation along a string.
 2
d2 ξ

∂ ξ d dξ ∂u±
= =


 ∂x2 du2±

du± du± ∂x
2
d2 ξ

∂ ξ d dξ ∂u±
(±v) = v 2 2

 2 =


∂t du± du± ∂t du±
∂2ξ 1 ∂2ξ
⇒ 2
= 2 2 (2.1)
∂x v ∂t
This is called the d’Alembert’s3 equation, or wave equation, in 1 dimension. A
plane wave is defined as a function satisfying the equation (2.1).
Since the wave equation is linear, the sum of two or more waves is still a
wave. The general solution ξ(x, t), however, is not in general a function that
moves left or right, but a linear combination of tho functions ξ+ and ξ− moving
right and left respectively:
ξ(x, t) = ξ+ (x − vt) + ξ− (x + vt) . (2.2)
2.1.1 An example: mechanical waves in one dimension

Let us consider a string of length L with a tension T . In equilibrium the string
is straight. Suppose that the string is perturbed from its rest state; a section is
moved by a small distance compared to its length, perpendicularly to the string
itself.
Let AB be a piece of the string of length dx and mass dm, witch is moved
to a distance ξ from the equilibrium state. We have a force on each extreme of
the element AB. Because of the bending of the string these two forces have the
same intensity but not the same direction. Analyzing the forces acting on the
element AB, we have:
Fx = T (cos α0 − cos α) = dm ax
(2.3)
Fξ = T (sin α0 − sin α) = dm aξ .
3 Jean-Baptiste d’Alembert (1717 - 1783) was a French mathematician, physicist, philoso-
pher, and music theorist. He was also co-editor with Denis Diderot of the Encyclopd́ie.
28
Because of the little bending, the two angles α and α0 are small, so we can write:
sin α0 − sin α ' α0 − α ' tan α0 − tan α

(2.4)
cos α0 − cos α ' 0 ,
and then:
Fx = dm ax ' 0
(2.5)
Fξ = dm aξ ' T (tan α0 − tan α).
We see that the horizontal net force is very small compared to the vertical
one. We can associate tan α to the slope of the string, i.e., to the derivative
with respect to x: ∂ξ/∂x, so we have:
∂ ∂ξ ∂2ξ
Fξ = T dx = T 2 dx . (2.6)
∂x ∂x ∂x
Let µ = M/L be the linear density of the string. We have that the mass
of element AB is µdx; moreover the vertical acceleration is ∂ 2 ξ/∂t2 . By the
second law of motion we can write:
∂2ξ ∂2ξ
µdx = T 2 dx. (2.7)
∂t2 ∂x
Thus,
∂2ξ T ∂2ξ
2
= . (2.8)
∂t µ ∂x2
As we can see this is the d’Alembert wave equation so weqnow know that
the perturbabion moves through the string with velocity v = Tµ .
2.2 Energy transported by a wave

A wave transfers a perturbation, and thus, ultimately, energy. If u is the density
of energy per unit volume associated to the perturbation, the the intensity of
transmitted energy U per units of area and time is:
dU dU dx
I= = = uv, (2.9)
dSdt dSdx dt
where I is called intensity and is measured in Wm−2 . In a sinusoidal mechan-
ical wave ξ0 sin(kx − ωt) (v = ω/k) a fixed point describes a simple harmonic
oscillation, thus:
1 1 1
u = kξ02 ∝ ω 2 ξ02 ⇒ I ∝ ω 2 ξ02 v . (2.10)
2 2 2
29
2.3 Sinusoidal waves and the Fourier theorem
A periodical phenomenon is something that repeats itself equally on regular
periods of time. More precisely in mathematics a function f (t) is said to be
periodical of period T if:
f (t + nT ) = f (t), ∀n ∈ N and ∀t. (2.11)
Well known examples of periodical functions are the trigonometric functions sin
and cos.
Sinusoidal (or, which is equivalent, cosinusoidal) waves are very important
because of the Fourier theorem.
Figure 2.3: Example of Fourier decomposition.
The Fourier theorem states that every periodical function f (t) of period
T = 2π/ω which is finite, continuos and differentiable can be expressed by the
Fourier series:
∞
X ∞
X
f (t) = a0 + (an cos nωt + bn sin nωt = a0 + cn cos(nωt + ϕn )) (2.12)
n=1 n=1
30
where:
p
cn = a2n + b2n (2.13)
an
tan ϕn = (2.14)
bn
Z T
1
a0 = f (t)dt (2.15)
T 0
T
2 T
Z Z
2
an = f (t) cos(nωt)dtbn = f (t) sin(nωt)dt (2.16)
T 0 T 0
Thus a generic wave can be written as a linear combination of sine waves.
Let us analyze the proprieties of (co)sinusoidal waves:
ξ(x, t) = ξ0 cos[k(x − vt) + δ] (2.17)
To represent a (co)sinusoidal wave we can use the complex representation:
ξ(x, t) = Re[ξ0 ei[k(x−vt)+δ] ] (2.18)
and so we can refer to the complex wave:

˜ t) = ξ˜0 eik(x−vt)
ξ(x, (2.19)
where ξ˜0 = ξ0 eiδ (the amplitude absorbs the phase). As we stated before, the
real part of the complex wave is the physical wave in this context. Since the
wave equation is linear, we can carry on our calculations using exponentials (the
advantage of the complex notation is that exponentials are much easier to ma-
nipulate than sines and cosines), and then go back to the cosine representation
when we want.
2.4 Amplitude, wavelength, period

We introduce some important definitions related to of a harmonic (sinusoidal)
wave ξ(x, t) = ξ0 cos[k(x − vt) + δ]:
• ξ0 = max{|ξ(x, t)|} is defined as the amplitude.

• T = 2π/ω is the period, i.e., the minimum time after which the function
makes a replica of itself (the time that a point takes to do a complete
oscillation).
• ν = 1/T is the frequency, i.e., the number of periods per second; the
unit of the frequency in the SI is the hertz Hz: [Hz]= [s−1 ].
• λ = 2π/k is the wavelength, the spatial period of the wave - i.e., the
minimum distance over which the wave’s shape repeats.
• v = λ/T = λν is the velocity of the wave.
31
• k = 2π/λ is defined as the wave number.
• ω = kv = 2π/T is the angular frequency.
• δ is the phase, and it depends on the choice of the starting time.
2.5 Transverse and longitudinal waves

Let us suppose that the quantity associated to a wave ξ is a vector, for example
representing an oscillation, or the electric field. We define:
• Transverse waves] the waves for which the direction of the perturbation is
the same as their direction of propagation.
• Longitudinal waves the waves for which the direction of the perturbation
is perpendicular to the direction of propagation.
Figure 2.4: Transverse and longitudinal waves.
Transverse waves are said to be polarized if there is a simple law describing

the direction of oscillation ξ in the plane of oscillation, say, yz. Particular cases
of polarized waves are:
• Linearly polarized waves: waves in which the direction of oscillation ξ is
fixed.
• Elliptically polarized waves. These are waves that satisfy the condition:
ξy = ξ0y sin(kx − ωt) (2.20)

ξz = ξ0z cos(kx − ωt) (2.21)
ξy2 ξz2
2 + 2 =1 (2.22)
ξ0y ξ0z
2 2
In particular, if ξ0y = ξ0x the wave is said to be circularly polarized.
2.6 An example: sound

Sound is a sequence of (approximately longitudinal) waves of pressure that prop-
agates through compressible media such as air, water, or solids. Sound that is
32
Figure 2.5: A stationary source of waves and a moving source.
perceptible by humans has frequencies from about 20 Hz to 20 000 Hz; in air

at standard temperature and pressure, the corresponding wavelengths of sound
waves range from 17 m to 17 mm.
Pitch is an auditory sensation in which a listener assigns musical tones to rel-
ative positions on a musical scale based primarily on the frequency of vibration.
Pitch is closely related to frequency, but the two are not equivalent, apart from
purely sinusoidal waves. Frequency is an objective, scientific concept, whereas
pitch is subjective, and related to the sector of psychoacoustics.
The speed of sound depends on the medium the waves pass through, and is a
fundamental property of the material. In general, the speed of sound is propor-
tional to the square root of the ratio of the elastic modulus K = −V dP/dV of
the medium (where V is the volume and P is the pressure) to its density. Those
physical properties and the speed of sound change with ambient conditions. For
example, the speed of sound in gases depends on temperature. In air at NTP,
the speed of sound is approximately 340 m/s. In fresh water at 20 ◦ C, the speed
of sound is approximately 1400 m/s. In steel, the speed of sound is about 6000
m/s.
2.7 The classical Doppler effect

A peculiar effect of wave propagation is the so-called Doppler effect: the char-
acteristics of a wave depend on the motion of the emitter and of the receiver. It
is a phenomenon that affects all type of waves, and it is of primary importance
in astrophysics, as we shall see next.
The Doppler effect is the apparent change of frequency of the wave, when
a source or an observer are in movement. We notice this effect, for example,
when we are in relative movement with respect to a source of sound waves: if
we approach the source, the frequency of the sound is higher; if we get far away,
the frequency appears lower.
Let us consider a source, that emits a wave with frequency f , and an ob-
33
server. Suppose that the source and the observer move with constant relative
velocity, and that the direction of the velocity lies on the straight line joining
the two objects. We want to analize how the Doppler effect manifests itself.
We make also the semplification that one of the two objects is stationary with
respect to the medium in which the wave propagates (for example, in the case
of sound, the air of the atmosphere). We can immediately note that the phe-
nomenon depends on who is moving and who is stationary with respect to the
medium in which the wave propagates.
Source moving, observer at rest. Let us consider first the case in which
the source moves with velocity vs < v in the direction of the observer. The
length of the wave received λ0 changes. We have that
vs
λ0 = λ − .
f
Thus:
v v vs
= − ;
f0 f f
1
=⇒ f 0 = f · ;
1 − vvs
These conditions hold only when the velocity of propagation v is larger than vs .
In particular if vs v, we can make the approximation
vs
f0 ' f 1 + .
v
Source at rest, observer moving. Let us consider the case in which the
source lies on a fixed point and the observer moves with velocity vo in the
direction of the source. The velocity of the wave relative to the observer is
v + vo , so we get
v + vo v + vo vo
f0 = = =f 1+ .
λ v/f v
We can notice that only in the first approximation the variation of frequency
is the same in both cases. The case in which both the source and the observer
move is more complicated, as the case in which the direction of the movement is
not directed along the line joining them; however, there is nothing conceptually
new.
2.7.1 Redshift of galaxies and the expansion of the Uni-

verse
The Doppler effect has important applications in astrophysics. Observing the
Doppler effect on the spectrum of emission of galaxies and stars in the Universe,
34
Figure 2.6: Experimental plot of the relative velocity (in km/s) of known
astrophysical objects as a function of the distance from the Earth (in Mpc; 1
pc ' 3.3 ly). The line is a fit to Hubble’s Law.
35
Figure 2.7: Redshift of emission spectrum of stars and galaxies at different
distances. Using this shift we can calculate the relative velocity between these
objects and the Earth.
we can compute the relative velocity of these objects with respect to us, and
then the distance, thanks to the so-called Hubble’s law.
In 1929 Edwin Hubble4 , studying the emission of galaxies, observed from
their Doppler redshift that objects in Universe move away with us with velocity
v = H0 d , (2.23)
where d is the distance between the objects, and H0 is a parameter called the
Hubble constant (whose value is known today to be about 24km/s/M ly). The
above relation is called Hubble’s law.
To give an isea of what H − 0 means, the speed of revolution of the Earth
around the Sun is about 30 km/s; Andromeda, the large galaxy closest to the
Milky Way, has a distance d of about 2.5 Mly. However, the Hubble’s law is
just statistical and working for large distances, where gravitational attraction
becomes negligible: Andromeda is indeed approaching us.
Dimensionally we note that H0 is a frequency: H0 ' (14 × 109 years)−1 .
A simple interpretation of this law is that, if the Universe has always been
expanding at a constant rate, about 14 · 109 years ago its volume was zero. This
result is consistent with present estimates of the age of the Universe within the
so-called big bang theory.
The redshift
λ0
z= −1
λ
is also used as a metric of distance of objects.
4 Edwin Hubble (1889 - 1953) was an American astronomer who played a crucial role in
establishing the field of extragalactic astronomy and is generally regarded as one of the most
important observational cosmologists of the 20th century.
36
Figure 2.8: The Cherenkov cone.
2.7.2 The Vavilov-Cherenkov effect

The Vavilov-Cherenkov5 (commonly just called Cherenkov) effect occurs when
a wave emitter moves through a medium faster than the speed of the wave in
that medium.
In the case vs > v, as we can see in the Figure 2.8, the wave front is a cone,
that takes the name of Cherenkov cone. The following relation connects the
angle θ with the velocities v and vs :
cos θ = v/vs . (2.24)
To find the value of cos θ let us consider two positions of the source S1 and
S2 , and the corresponding points P and Q on the wave front: the wave emitted
in S1 has, in P , the same phase of the one emitted in S2 in Q. For the same
reason also the points S10 and S2 have the same phase. The time that the source
spends to go from S1 to S2 is equal to the time that the wave spends to go from
S1 to S10 . If we call a the distance S1 S2 we have S1 S10 = a cos θ. Thus we get:
a a cos θ
=
vs v
v
=⇒ cos θ = .
vs
As an example, when cosmic rays interact with the atmosphere they generate
showers of particles. The charged particles radiate light, and some of them
are faster than light in the atmosphere, thus generating cones of collimated
Cherenkov light. This light is detected by special-purpose telescopes.
5 Pavel Alekseyevich Cherenkov (1904-1990) was a Soviet physicist who shared the Nobel
Prize in physics in 1958 with Ilya Frank and Igor Tamm for the discovery of Cherenkov
radiation, made in 1934. The discovery was made during Cherenkov’s thesis, directed by the
academician Nikolai Vavilov; when the Nobel prize was assigned, however, Vavilov was dead
since 15 years.
37
2.8 Composition of waves
Since the wave equation is linear, when we want to sum two waves we can
just sum the functions representing them. We should remember that energy is
proportional to the square of the amplitude: this can create effects of positive
and negative interference that we shall discuss later.
Figure 2.9: Examples of algebraic sum of two waves.
2.8.1 Boundary conditions and steady waves

Let ξ(x, t) be the equation of a plane wave along the x axis direction. Generally
we can write
ξ(x, t) = ξ+ (x − vt) + ξ− (x + vt) (2.25)
where ξ+ (x − vt) propagates in the positive direction of the x coordinate, while
ξ− (x + vt) propagates in the negative direction (Figure 2.10).
Consider now the case in which the wave is confined in a finite region, e.g., a
wave on a rope between two walls. When a wave ξ+ meets the wall, it changes
verse of propagation, and “creates” a second wave ξ− with the same character-
Figure 2.10: Waves with opposite velocity.
38
istics but opposite velocity. If we sum the two waves we obtain a wave with
constant speed zero, also called stationary wave.
Let us analyze mathematically what happens. Since v = ω/k, and ω remains
the same, the change of velocity from +v to −v corresponds to a change of wave
number from +k to −k. So the two wave equations are:
ξ+ (x, t) = ξ0 sin (kx − ωt) ;
ξ− (x, t) = ξ0 sin (−kx − ωt) .
Using the appropriate trigonometric formulae, the sum of the waves is:
ξ(x, t) = ξ+ (x, t) + ξ− (x, t) =
= ξ0 sin (kx − ωt) + ξ0 sin (−kx − ωt) =
= 2ξ0 sin (−ωt) cos (kx) =
= −2ξ0 sin (ωt) cos (kx) .
The oscillation is maximum in the points x such that cos(kx) = 1, while is
minimum in the points that satisfy the condition cos(kx) = 0. These points are
called respectively antinodes and nodes.
Observe that only waves that have nodes on the extremities of the rope will
survive: otherwise the amplitude of the sum of the waves in an extreme is not
null, and so also the energy (dispersed) is not null, but this contradicts the
conservation of energy, because we are supposing the reflected wave to have the
same amplitude ξ0 .
Imagine you are perturbing a violin string: the extremities P and Q are
fixed and you generate different waves displacing the string in different points.
The wave you have generated can be expressed as the sum of sinusoidal waves,
from the Fourier’s theorem. All these waves have a same property: the points
P and Q are nodes for them. After a transient, only the waves for which the
points P and Q are nodes survive, otherwise the wave, when reflected, generates
a destructive interference. So only discrete values of λ are permitted, those for
which the boundaries are nodal points. These are called the harmonics:
• the fundamental frequency with frequency f1 is the wave with maximum
wavelenght: it is such that λ2 = L where λ is the wavelenght and L is the
string lenght. Observe that λ = 2L;
2λ
• the second harmonic f2 is a wave with wave lenght such that 2 = L, that
is λ = L;
3λ
• the third harmonic f3 is such that 2 = L, that is λ = 32 L;
4λ
• the fourth harmonic f4 has a wave lenght such that 2 = L, i.e. λ = 21 L;
• and so on . . .
The wave generated perturbing the violin string is, after a transient, the sum
of these waves, as we can see in Figure 2.12. The fundamental frequency deter-
mines the pitch of the note, and together with the higher harmonics determine
39
Figure 2.11: The first four harmonics.
40
Figure 2.12: A string is plucked in a certain point. This creates a wave that is
sum of three waves: the foundamental frequency, the second and the third
harmonics. Their sum produce a determined sound.
the timbre of the sound. We can obtain sounds with same pitch, but different
timbre, just plucking the string in different places.
2.8.2 Beats
We will now analyze another interesting example of wave composition. Let us
consider two sinusoidal waves that propagate in the same direction, with the
same amplitude ξ0 and velocity v, but slightly different frequencies ωi .
The waves are defined by the following equations:
ξ1 (x, t) = ξ0 cos (k1 x − ω1 t) ;

ξ2 (x, t) = ξ0 cos (k2 x − ω2 t) .
What happens if we sum them? Let us put:

k1 + k2
∆k = k1 − k2 ; hki = ;
2
ω1 + ω2
∆ω = ω1 − ω2 ; hωi = .
2
By trigonometric identities:

α+β α−β
cos(α) + cos(β) = 2 cos cos ,
2 2
41
Figure 2.13: Graphic representation of beats.
we obtain
f1 (x, t) + f2 (x, t) = ξ0 cos(k1 x − ω1 t) + ξ0 cos(k2 x − ω2 t) =

∆k ∆ω
= 2ξ0 cos x− t cos(hkix − hωit). (2.26)
2 2
The resulting function is a product of two cosinusoidal functions. Figure 2.13

explains the behaviour of the new wave. In this case the listener perceives an
oscillating volume level; the frequency of the volume oscillation is much lower
than the sound frequency.
This effect happens for example when two singers are not able to take the
same tonality.
2.8.3 Group velocity and phase velocity

In this section we want to analize the behaviour of a wave packet, starting
with the prerequisite that the propagation velocity of a wave with equation
ξ(x, t) = ξ0 cos(kx − ωt), is given by v = ω/k.
A wave packet is a short envelope of waves that travels as a unit. If the
wave packet propagates in one direction (e.g. the x axis), using the Fourier’s
theorem, its general form can be written as:
Z
ξ(x, t) = Ξ(k)ei(kx−ω(k)t) dk,
where Ξ(k) is a function that takes a large value in a region of area ∆k̄ around a
certain point k̄, and goes to zero elsewhere. For example f could be a Gaussian
function with very low variance so that the funcion has a great peak near k̄. An
example of wave packet is represented in Figure 2.14.
We want now to analize the speed of propagation of a wave packet. To do
this we observe that beats are a simple wave packet, made by two waves. So we
begin studying this case, that is simpler than the general one. We know that
42
Figure 2.14: A wave packet.
Figure 2.15: In the example of beats the envelope wave (green) is given by the
cosinus with longer wavelenght.
the speed of a wave whose equation is f (x, t) = ξ0 cos(kx − ωt) is given by:
ω
v= . (2.27)
k
In the previous section we obtained that beats have an equation that is the
product of two cosinusoidal functions. The speed of the envelope is given by the
speed of the factor with longer wavelenght, i.e., lower wavenumber. If we return
to (2.26), we have to compare the numbers hki and ∆k 2 , to understand which
term has the lower wavenumber. It is easy to prove that ∆k 2 ≤ hki, so we have
∆k ∆ω

to consider the factor cos 2 x − 2 t . Using the (2.27) we find the velocity
∆ω
venvelope = .
∆k
If we want to extend the result to general wave packets, we have to take the
limit for ∆k that goes to zero. So the velocity of the envelope, also called group
velocity, is given by:
d(ω(k))
vg = . (2.28)
dk
The phase velocity is the speed of propagation of a phase, for example of the
point P represented in Figure 2.15. This velocity, in the case of beats, is equal
to the propagation velocity of the factor with shorter wavelenght and higher
43
wavenumber, i.e., we have to look at the term cos(hkix − hωit). Remembering
that v = ω1 /k1 = ω2 /k2 the phase velocity of beats is:
hωi ω1 + ω2 vk1 + vk2 ω1 ω2
v= = = = = = v.
hki k1 + k2 k1 + k2 k1 k2
Note that the phase velocity is equal to the speed of the two waves that generate
beats. In general, in a dispersive medium (where v is a function of ω) the phase
and group velocities can be different.
2.9 Waves in three dimensions

Every wave we have considered so far was a planar wave moving along the x axis,
whose equation is ξ(x, t) = ξ(x ± vt). A wave of this form can be decomposed
in its harmonics which can be written as ξ(x, t) = ξ0 sin((kx ± ωt) + φ).
To describe a wave moving in a general direction we define a new vector
~k = 2π ûv . So we have that
λ
ξ = ξ0 sin((~k · ~r ± ωt) + φ) = ξ0 sin((kx x + ky y + kz z ± ωt) + φ)
ω
knowing that |~k| = we have
v
∂2ξ ∂2ξ ∂2ξ 1 ∂2ξ
2
+ 2 + 2 = ∇2 ξ = 2 2 .
∂x ∂y ∂z v ∂t
This last equation also has nonplanar waves as solutions.
We define wavefront a surface whose phase is constant in a given moment
of time, and ray the line orthogonal to the wavefront which represents in that
point the direction of the wave and of the energy associated to it. If |v| is the
same in every direction we have a spherical wavefront; if |v| is the same in every
direction perpendicular to a given axis we have a cylindrical wavefront.
2.9.1 Spherical waves

A small spherical object which pulsates periodically produces a sound wave,
whose wavefronts are spheres concentric to the object. These waves are an
example of spherical waves.
The equation for a single harmonic component is ξ(r, t) = A(r) sin (kr − ωt).
If the medium in which the wave is travelling in is motionless and has uniform
density then the waves propagate outwards with constant velocity, therefore the
wavelength will not depend on the distance whereas, due to energy conservation,
the amplitude will, decreasing as we get further from the source.
The energy flow per unit surface is I(r) = CA2 (r), where C is a constant.
If we have no dispersion, the power carried through a surface of radius r is
constant, and equal to
cost ξ0
I(r)S(r) = CA2 (r) · 4πr2 = cost ⇒ = A(r) =
r r
44
where S(r) = 4πr2 is the surface of the wavefront.
Therefore the equation of a spherical wave in a nondispersive medium is
ξ0
ξ(r, t) = sin(kr − ωt).
r
45
Chapter 3
Electromagnetism and the

Maxwell’s equations
The Maxwell’s equations1

I
q
E~ · dS ~ = (3.1)
S 0
I Z
d
E~ · d~l = − ~ · dS
B ~ (3.2)
C dt S(C)
I
~ · dS
B ~ = 0 (3.3)
S
I Z
~ · d~l d
B = µ0 I + 0 µ0 E~ · dS
~, (3.4)
C dt S(C)
together with the equation describing the motion of a particle of electrical charge
q in an electromagnetic field
F~ = q(E~ + ~v × B)
~ (3.5)
(Lorentz2 force), provide a complete description of electromagnetic field and of
its dynamical effects.
We want to write Maxwell’s equations in a local form, and to transform the
integro-differential equations above into purely differential equations.
1 James Clerk Maxwell (1831 - 1879) was a Scottish physicist. His most prominent achieve-
ment was formulating classical electromagnetic theory. Maxwell’s equations, published in

1865, demonstrate that electricity, magnetism and light are all manifestations of the same
phenomenon, namely the electromagnetic field. Maxwell also contributed to the Maxwell-
Boltzmann distribution, which gives the statistical distribution of velocities in a classical per-
fect gas in equilibrium. Einstein kept a photograph of Maxwell on his study wall, alongside
pictures of Faraday and Newton.
2 Hendrik Lorentz (1853 - 1928) was a Dutch physicist who gave important contributions
to electromagnetism. He also derived the equations subsequently used by Albert Einstein

to describe the transformation of space and time coordinates in different inertial reference
frames. He was awarded the 1902 Nobel Prize in Physics.
46
3.1 Charge density and current density
We first write in terms of local variables the electric charge and the electric
current.
The charge density ρ(t, ~r) is defined as the charge per unit volume in a point
~r at a time t: Z
q= ρ(t, ~r) dV . (3.6)
V
The current density ~(t, ~r) is defined as the intensity of electrical current per
unit surface: Z
~.
I = ~(t, ~r) · dS (3.7)
S
3.2 Maxwell’s equations in differential form

Let us examine Equation (3.1). The charge q is contained in the volume volume
V , thus we can write Z
q= (S)ρ dV .
V
By Gauss’ theorem, I Z
E~ · dS
~= ~ · E)
(∇ ~ dV
S V (S)
and thus
~ · E~ = ρ .
∇ (3.8)
0
In the same way we get from Equation (3.3), by applying Gauss’ theorem,
I
B~ · dS
~ = 0 =⇒ ∇
~ ·B
~ = 0. (3.9)
Equations
I Z
d
E~ · d~l = − B~ · dS
~
C(S) dt S
I Z
~ · d~l = d
B µ0 I + 0 µ0 E~ · dS
~
C(S) dt S
become respectively, by the application of Stokes’ theorem, and of Equation

(3.7),
Z Z
~ ~ ~ d ~ · dS
~
(∇ × E) · dS = − B
S dt S
Z Z Z
~ ~ ~ ~ d
(∇ × B) · dS = µ0 ~ · dS + 0 µ0 E~ · dS
~
S S dt S
47
and thus
~
∂B
~ × E~
∇ = −
∂t
~ ×B
~ ∂ E~
∇ = 0 µ0 + µ0~ .
∂t
These are called respectively the law of Faraday3 -Lenz4 law and the Ampère5 -
Maxwell law.
Thus Maxwell’s equations can be written in differential form as:
~ · E~ = ρ
∇
0
~
~ × E~ = − ∂ B
∇
∂t
∇~ ·B ~ =0
~
~ ×B
∇ ~ = 0 µ0 ∂ E + µ0~
∂t
These equations allow calculating the electric and magnetic fields from the
charge distribution ρ(t, ~r) and the current density ~(t, ~r).
3.3 Maxwell’s equations and continuity equation

for charge
If we take the divergence of both sides of the Ampère-Maxwell law
~
~ = 0 µ0 ∂ E + µ0~
~ ×B
∇
∂t
we obtain
~
~ · (∇
∇ ~ × B) ~ · ∂ E + µ0 ∇
~ = 0 µ0 ∇ ~ ·~ .
∂t
~ · (∇
But ∇ ~ × B)
~ = 0, and we can exchange the derivatives with respect to space
and time:
∂ ~ ~ ~ ·~ = 0 .
0 µ0 (∇ · E) + µ0 ∇
∂t
Gauss’ law tells us that ∇~ · E~ = ρ/0 ; thus
~ ·~ + ∂ρ = 0 .
∇ (3.10)
∂t
3 Michael Faraday (1791 - 1867) was an English scientist who contributed to the fields of
electromagnetism and electrochemistry. His main discoveries include those of electromagnetic

induction, diamagnetism and electrolysis. Although Faraday received little formal education
he was one of the most influential scientists in history.
4 Heinrich Lenz (1804 -1865) was a Russian physicist of Baltic ethnicity. He is most noted
for formulating Lenz’s law in electrodynamics in 1833.

5 André-Marie Ampère (1775 - 1836) was a French physicist and mathematician who is
generally regarded as one of the main founders of the science of electrodynamics.
48
The equation above is a continuity equation for charge. If there is a net
electric current is flowing out of a region, then the charge in that region must
be decreasing by the same amount. Charge is conserved.
3.4 The potentials, vector and scalar

The physics in the Maxwell equations depends only on the magnetic and electric
fields. We shall tray to express these fields in terms of some generating fields,
called potentials, which have no immediate relation to physics.
The divergence of B~ is zero, and this means that we can always represent B ~
as the curi of another vector field. Conversely, the divergence of a curl is always
zero. Thus we can always relate the magnetic field to a field we will call A ~ by
~ =∇
B ~ ×A
~.
~ is called the vector potential.

The field A
We remind that the scalar potential φ, such that
E~ = −∇φ
~
in electrostatics, was not completely specified by its definition. If we have found

φ, we can always find another potential φ0 that is equivalent from the point of
view of physics by adding a constant:
φ0 = φ + C .
The new potential φ0 gives the same electric fields, since the gradient of a con-
stant is zero; φ0 and φ represent the same physics.
Similarly, we can have different vector potentials A ~ which give the same
~ ~
magnetic fields. Again, because B is obtained from A by differentiation, adding
a constant to A ~ does not change anything physical. But we can add to A ~ any
field which is the gradient of some scalar field, without changing the physics:
~ =∇
B ~ ×A
~ =⇒ ∇
~ × (A
~ + ∇ψ)
~ ~ ×A
=∇ ~+∇
~ × ∇ψ
~ =∇
~ ×A
~=B
~.
~ by arbitrarily
It is usually convenient to take some of the freedom out of A
placing some other condition on it (in much the same way that we found it
convenient - often - to choose to make the potential zero at large distances).
We can, for instance, restrict A ~ by choosing arbitrarily what the divergence
~
of A must be. We can always do that without affecting B. ~ This is because
~ ~ 0 ~
although A and A have the same curi, and give the same B, they do not need
to have the same divergence. By a suitable choice of ψ we can make ∇~ ×A ~ 0 any
well-behaved function we wish.
What should we choose for ∇ ~ · A?
~ The choice should be made to get the
greatest mathematical convenience and will depend on the problem we are doing.
For magnetostatics, we make the simple choice
~ ·A
∇ ~=0 (3.11)
49
(later, when we take up electrodynamics, we shall make a different choice). Our
complete definition of A is then, for the moment,
~ ×A
∇ ~=B
~ and ∇
˜ · Ã = 0 .
3.5 Maxwell’s equations and electrostatics

Two of the Maxwell’s laws are the traditional Ampère-Maxwell and Faraday-
Lenz laws.
Can we obtain the two other laws of electrostatics and magnetostatics : the
Coulomb’s law, stating that the electric field by a point charge q is

1 ~ur
E~ = (3.12)
4π0 r2
and the Biot-Savart6 law, stating that the elementary magnetic field generated
by a current I over an element of conductor d~l at a radius ~r is
µ d~l × ~u
~ = 0 r
dB I . (3.13)
4π r2
from the Maxwell equations? It turns out that this is indeed possible.
We do the demonstration for the Coulomb’s law. The differential form of the
Maxwell’s law is equivalent to its integral form (3.1). Due to symmetry reasons,
the field must be directed radially. Then if we choose as a surface a sphere of
radius r, equation (3.1) becomes
~ 4πr2 = q
|E| ,
0
which demonstrates the statement.
3.6 Maxwell’s equations and magnetostatics

For the demonstration of the Biot-Savart law from Maxwell’s equations, see
the Feynman’s lectures on Physics, Volume 2, Chapter 14 (it is not part of the
program of the exam).
6 Jean-Baptiste Biot (1774 - 1862) was a French physicist, astronomer, and mathemati-
cian who established the reality of meteorites, made an early balloon flight, and studied the
polarization of light. Félix Savart (1791 - 1841), professor at Collège de France, was the co-
originator of the Biot-Savart Law, along with Biot. Together, they worked on the theory of
magnetism and electrical currents. Their law was developed about 1820.
50
Chapter 4
Solutions of the Maxwell’s

equations in vacuo
Maxwell’s equations are four partial-derivatives first order equations; two of

them are coupled in the unknown fields E~ and B. ~ We shall see now that in
regions of space and time where no charges or currents are present, these can be
decoupled in two equations, one for E~ and one for B.
~ The resulting equations
are wave equations; we shall discuss them, together with their implications.
Let us go back to the Maxwell’s equations in differential form:
• Coulomb’s law:
~ · E~ = ρ
∇
ε0
• Faraday-Lenz’s law:
~
~ × E~ = − ∂ B
∇
∂t
• the equation stating that there are no magnetic charges (monopoles):
~ ·B
∇ ~ =0
• Ampère-Maxwell’s law:
~
~ = ε0 µ0 ∂ E + µ0~ .
~ ×B
∇
∂t
Now we consider a region without charges and we are going to analyze the
situation. We can write the Maxwell’s equations in vacuo as:
~ · E~ = 0
∇ (4.1)
~
~ × E~ = − ∂ B
∇ (4.2)
∂t
51
~ ·B
∇ ~ =0 (4.3)
~ ×B
∇ ~ = ε0 µ0 ∂E (4.4)
∂t
A trivial solution would be that the electromagnetic field is everywhere zero.
Let us examine the characteristics of nontrivial solutions.
We suppose that there would be field generators outside the no-charges re-
gion. Consequently the region inherits information about the charges in the
Universe. Equation 4.2 can be written as:
~ × E~ = − ∂ ∇

~ × ∇
∇ ~ ×B ~
∂t
∂ 2 E~
= −ε0 µ0 2 thanks to equation 4.4
∂t
On the other hand
∂ 2 E~
~ × ∇

~ × E~ = ∇
~

~ · E~ −∇2 E~ = −∇2 E~
−µ0 ε0 = ∇ ∇
∂t2 | {z }
=0, eq. 4.1
and thus we obtain:

∂ 2 E~ 1
= ∇2 E~ . (4.5)
∂t2 ε0 µ0
We observe that the electric field in the empty space satisfies the d’Alembert’s
equation: it is thus a wave.
Similarly to what was done for the electric field we can write
∂2B~ ∂ ~
−µ0 ε0 2
= ε0 µ0 ∇ × E~ = ∇
~ × ∇~ ×B
~ =∇~ ∇ ~ ·B
~ −∇2 B
~ = −∇2 B
~
∂t ∂t | {z } | {z }
eq.4.2 =0,eq. 4.3
and thus we obtain

∂2B~ 1
= ~.
∇2 B (4.6)
∂t2 ε0 µ0
This is the wave equation for the magnetic field in empty space; we note that
it has the same form as that of the electric field.
4.1 Maxwell’s waves and light

Since
1 N · m2
' 9 · 109
4πε0 C
µ0 T·m
= 10−7
4π A
52
we can write the velocity of this wave:
1
1 4πε0 9 · 109 m2
v2 = = µ0 ' ' 9 · 1016 .
ε0 µ0 4π 10−7 s2
Then: r
1
v= ' 3 · 108 m/s (4.7)
ε0 µ0
that is consistent with the value of the speed of light. It becomes then natural
to assume that light is an electromagnetic wave.
Hertz1 presented the decisive experimental confirmation that light is a Max-
well’s wave in 1888, by generating light through an oscillating electromagnetic
field. He wrote: “The connection between light and electricity is now estab-
lished ... In every flame, in every luminous particle, we see an electrical process
... Thus, the domain of electricity extends over the whole of nature. It even
affects ourselves intimately: we perceive that we possess ... an electrical organ -
the eye.” By 1900, then, three great branches of physics, electricity, magnetism,
and optics, had merged into a single unified theory (and it was soon apparent
that visible light represents only a tiny window in the vast spectrum of electro-
magnetic radiation, from radio though microwaves, infrared and ultraviolet, to
X-rays and gamma rays.)
The Ampère-Maxwell equation (4.4) becomes, if we do this assumption,
~ ~
~ = ε0 µ0 ∂ E = 1 ∂ E .
~ ×B
∇
∂t 2
c ∂t
Thus the speed of light enters in the fondamental laws of nature.
This fact has an important consequence: in classical mechanics the Maxwell’s
equations are not the same for different reference frames in relative motion with
constant velocity. Indeed, in classical physics, all the laws depend on acceler-
ation, which is invariant between the inertial reference frames (the coordinates
are transformed from one reference frame to another by means of the Galilei2
transformations); velocities cannot enter in a fundamental law valid in all iner-
tial frames, since they are not an invariant quantity.
Since the Maxwell’s equations violate the Galilean relativity, if we assume
that they are correct, there could be two possibilities (not mutually exclusive):
• Relativity is not a legitimate assumption: there is a preferred reference
frame, in which electromagnetism can be described (this has been called
1 Heinrich Hertz (1857 - 1894) was a German physicist who clarified and expanded James
Clerk Maxwell’s electromagnetic theory of light.

2 Galileo Galilei (1564 - 1642) was an Italian physicist, mathematician, astronomer, and
philosopher who played a major role in the scientific revolution. His achievements include
improvements to the telescope and consequent astronomical observations and support for
Copernicanism. His contributions to observational astronomy include the discovery of the
phases of Venus, of the four largest satellites of Jupiter (named the Galilean moons in his
honour), and the observation and analysis of sunspots. Galilei also worked in military science
and technology.
53
the aether). For example, this would be the case if electromagnetic fields
would be the perturbation of a static medium.
• The relativistic transformations of Galilei do not work.
We shall see that the answer to this question brought to us one of the most
fascinating theories of the XX century - Einstein’s3 special relativity.
4.2 Properties of the electromagnetic waves

Now we study the caracteristics of elettromagnetic waves in vacuo. We suppose
to be very far away from the electromagnetic charges, so the equation of the
~ y, z, t) ≡ E(z,
wave depends on a single variable. Then E(x, ~ t) and
∂ E~ ∂ E~ ~
∂B ~
∂B
= = = =0
∂x ∂y ∂x ∂y
(the waves must be plane waves). Thus the Maxwell’s equations become:
→
−
∂ Ez
=0 (4.8)
∂z
∂ E~ ~
∂B
ûz × =− . (4.9)
∂z ∂t
Indeed
→
− −
→! →
− −
→! −
→ −
→!
~ × E~ = ûx ∂ Ez ∂ Ey ∂ Ez ∂ Ex ∂ Ey ∂ Ex
∇ − − ûy − + ûz − =
∂y ∂z ∂x ∂z ∂x ∂y
∂ E~ ~
∂B
= ûz × =− .
∂z ∂t
Analogously for the magnetic field:
−
→
∂ Bz
=0 (4.10)
∂z
3 Albert Einstein (1879 - 1955) was a German-born physicist who deeply changed the rep-
resentation of the Universe by the human species. While best known for his mass-energy
equivalence formula E = mc2 (published in 1905), he received the 1921 Nobel Prize in Physics
for his discovery of the law of the photoelectric effect (also in 1905), which was pivotal in es-
tablishing quantum theory within physics. Near the beginning of his career, Einstein thought
that Newtonian mechanics could not reconcile the laws dynamics with the laws of the electro-
magnetic field. This led to the development of his special theory of relativity (again in 1905).
He realized, however, that the principle of relativity could also be extended to gravitational
fields, and with his subsequent theory of gravitation in 1916, he published a paper on the gen-
eral theory of relativity. He was visiting the United States when Adolf Hitler came to power
in 1933, and did not go back to Germany, where he was a professor in Berlin. He settled in
the U.S., becoming a citizen in 1940. On the eve of World War II, he helped the set up of
the Manhattan Project, and ultimately the construction of the atomic bomb. Later, however,
he highlighted the danger of nuclear weapons. Einstein was affiliated with the Institute for
Advanced Study in Princeton until his death.
54
~
∂B 1 ∂ E~
ûz × = 2 . (4.11)
∂z c ∂t
We observe that
~
! →
− →
−
∂B 1 ∂ Ez ∂ Ez
0 = ûz · ûz × = 2 =⇒ =0
∂z c ∂t ∂t
! −
→ −
→
∂ E~ ∂ Bz ∂ Bz
0 = ûz · ûz × =− =⇒ =0
∂z ∂t ∂t
and thus we can conclude that Ez and Bz are constants, and in particular they
do not depend on the variables x, y, z, t. If we postulate that the Universe has
finite energy, we conclude that Ez = Bz = 0. Recall that the energy density
associated with the presence of electric and magnetic field are respectively
ε0 E 2 B2
uE~ = uB~ = . (4.12)
2 2µ0
Unless the Universe has infinite energy, Ez = Bz = 0: the electromagnetic wave
must be transverse.
We rewrite all the equations obtained to understand the relations between
the electric field and the magnetic field in an electromagnetic wave.
Ez = 0 (4.13)
Bz = 0 (4.14)
∂Ex ∂By
=− (4.15)
∂z ∂t
∂Ey ∂Bx
= (4.16)
∂z ∂t
∂Ex ∂By
= c2 (4.17)
∂t ∂z
∂Ey ∂Bx
= −c2 (4.18)
∂t ∂z
If we set υ = z − ct we have that ∂υ ∂υ
∂z = 1 and ∂t = −c. We can obtain from
equation 4.15
∂By ∂Ex ∂Ex ∂υ ∂Ex
=− =− =−
∂t ∂z ∂υ ∂z ∂υ
Z Z Z
∂By ∂Ex 1 ∂Ex
⇒ By = dt = − dt = dυ
∂t ∂υ c ∂υ
Ex
⇒ By = + constant
c
The constant must be equal to zero, not to have infinite energy; in the end we
have then
Ex
By = . (4.19)
c
55
Analogously from Equation (4.16)
Ey
Bx = − . (4.20)
c
In conclusion we have obtained that
~
|E|
~ =
|B|
c
and
~ · E~ = Ey Ex
B − Ex + Ey + 0 = 0.
c c
In other words, the electric wave is perpendicular and proportional to the mag-
netic wave.
We also observe from (4.19) and (4.20) that
E~ × B
~ = EB u~z : (4.21)
i.e., the vector product E~ × B

~ gives the direction of propagation of the wave.
4.2.1 Energy transported an the electromagnetic wave

We found the following equations:
~ 2~
∂ 2 ~E ~
2∂ E 1
2
= c c2 =
∂t ∂x2 ε0 µ0
~
~ 2~
~ ~
∂2B 2∂ B ~~ |~E|
2
= c |B| =
∂t ∂x2 c
The energy densities associated with the electric and magnetic fields are
respectively
ε0 E 2 B2
uE~ = uB~ =
2 2µ0
so that the total energy density is
ε0 E 2 B2
u= + .
2 2µ0
Considering that B = E/c and ε0 µ0 = 1/c2 ,we have that in an electromag-
netic wave the magnetic and electric components of the total energy are equal:
B2 E2
uB~ = = = uE~
2µ0 2µ0 c2
so we can write the total energy density as
u = 2uE~ = ε0 E 2 .
56
4.2.2 The Poynting vector
Let u be the energy density per volume unit, the energy carried by a wave,
perpendicularly to its direction, per time and surface unit is
dU dU dx
I= = = uv.
dSdt dSdx dt
We call I the intensity and it is measured in W/m2 .
So far we assumed that the origin of the energy flux is coincident to the
wave source, ignoring possible dissipative effects. The energy carried by an
electromagnetic wave in vacuum per time and surface unit is
1
S = uc = ε0 E 2 c = ε0 EBc2 = EB
µ0
By the properties of E~ and B

~ in an electromagnetic wave we have that
1 ~~
S~ = ~
E ×B
µ0
has the same direction of the wave and its magnitude is equal to the power per
surface unit. This vector is called the Poynting’s vector 4 .
A sine wave of the form E = E0 cos(kz − ωt), B = B0 cos(kz − ωt) has an
average power of EB/2µ0 .
The electromagnetic wave also carries momentum. Based on considerations
supported by Einstein’s relativity theory (see later), the momentum carried per
~
surface unit is related to energy as S/c.
4.3 Photoelectric effect; the photon hypothesis

When light interacts with material objects, it often reveals a particle-like nature,
with behaving like a group of point-like objects interacting with individual atoms
or molecules.
In the phenomenon called photoelectric effect, light is able to knock electrons
out of the surface of a metal, creating an electric current. The phenomenon was
first observed by Hertz in 1887. The characteristics of such an emission cannot
be explained unless one assumes (Einstein suggested this explanation in one
of his famous articles published in 1905, and he was awarded the Nobel prize
for this) that light interacts like a group of wave packets (we shall call them
photons) interacting as particles.
Let us analyse the photoelectric effect using the apparatus in figure 4.1, in
which light of frequency ν is directed onto a target and ejects electrons from it.
A potential difference V is maintained between the target and the collector
to sweep up these electrons, called photoelectrons. This collection produces a
4 John Poynting (1852 - 1914) was an English physicist, college professor in Birmingham.
57
Figure 4.1: Sketch of the apparatus used to study the photoelectric effect.
Figure 4.2: The photoelectric effect.
58
photoelectric current I that is measured with meter A. We adjust the poten-
tial difference V by moving the sliding contact so that the collector is slightly
negative with respect to the target. This potential difference acts to slow down
the ejected electrons. We then vary V until it reaches a certain value, called the
stopping potential Vstop , at which point the reading of meter A has just dropped
to zero. When V = Vstop , the most energetic ejected electrons are turned back
just before reaching the collector.
Then Kmax , the kinetic energy of
these most energetic electrons, is
Kmax
Kmax = eVstop
where e is the elementary charge. Mea-

surements show that for light of a given
dKmax
frequency, Kmax does not depend on the
dν = h intensity of the light source.
Now let us vary the frequency ν of the
ν0 ν incident light and measure the associated
stopping potential Vstop . Figure 4.3 is
a plot of Vstop versus f . Note that the
Figure 4.3: The kinetic energy photoelectric effect does not occur if the
Kmax , as a function of the frequency is smaller than a certain cutoff
frequency. frequency ν0 or, equivalently, if the wave-
length is greater than the corresponding
cutoff wavelength λ0 = νc0 . This is true no matter how intense the incident light
is.
This observation is in sharp contrast with a classical view of reality, which
would make us expect to see electrons ejected for every frequency, given a bright
enough light. This is not what happens. For light below the cutoff frequency ν0 ,
the photoelectric effect does not occur, no matter how bright the light source
is.
The existence of a cutoff frequency is, however, just what we should expect
if the energy is transferred via photons. The electrons within the target are held
there by electric forces. To just escape from the target, an electron must pick
up a certain minimum energy Φ, where Φ is a property of the target material
called its work function. If the energy hν transferred to an electron by a photon
exceeds the work function of the material (if hν > Φ), the electron can escape
the target. If the energy transferred does not exceed the work function (that is,
if hν < Φ), the electron cannot escape. This is what figure 4.3 shows.
Einstein summed up the results of such photoelectric experiments in the
equation
hν = Kmax + Φ (photoelectric equation) . (4.22)
This is a statement of the conservation of energy for a single photon absorption
by a target with work function Φ. Energy equal to the photon’s energy hν is
transferred to a single electron in the material of the target. If the electron
is to escape from the target, it must pick up energy at least equal to Φ. Any
59
additional energy (hν − Φ) that the electron acquires from the photon appears
as kinetic energy K of the electron. In the most favourable circumstance, the
electron can escape through the surface without losing any of this kinetic energy
in the process; it then appears outside the target with the maximum possible
kinetic energy Kmax .
Therefore we can conclude that the a light wave of frequency ν is made of
quanta of energy
h
hν = ω = ~ω
2π
where h is the Planck’s constant, which has value
h ' 6.63 × 10−34 J s = 4.14 × 10−15 eV s .
4.4 The perception of electromagnetic waves: vis-

ible light
A typical human eye will respond to wavelengths from about 750 nm (the wave-
length of red) to 390 nm (the wavelength of purple). This range is similar to the
Sun’s radiation spectrum.
If we take a green laser (of wavelength 0.5 µm, that is 5.00 × 10−7 m), we
have
3 × 108 m/s
ν= = 0.6 × 1015 Hz.
5.00 × 10−7 m
Therefore the energy associated to the green light is
E = hν = 6 × 10−34 · 0.6 × 1015 ' 3 × 10−19 J ' 2 eV.
4.5 Natural units

The International system of units (SI) can be constructed on the basis of four
fundamental units: of length (the meter m), of time (the second s), of mass (the
kilogram kg), of charge (the coulomb C)5 .
These units are inappropriate for the world of fundamental physics: the
radius of a nucleus is of the order of 10−15 m, also called one femtometer (fm)
or one fermi; the mass me of an electron is of the order of 10−30 kg; the charge
of an electron is (in absolute value) of the order of 10−19 C. By using such units
we would carry along a lot of exponents! Thus in particle physics we better use
units like the electron charge for the electrostatic charge, and the electronvolt
eV and its multiples (keV, MeV, GeV, TeV) for the energy.
Length 1 fm 10−15 m
Mass 1 MeV/c2 1.78 × 10−30 kg
Charge |e| 1.60 × 10−19 C
5 For reasons related only to metrology, i.e., of reproducibility and accuracy of the definition,
in the standard SI the unit of electrical current, the ampere A, is used instead of the coulomb;
the two definitions are however conceptually equivalent.
60
Note the unit of mass, in which the relation E = mc2 establishing equivalence
of mass and energy (we shall discuss this relationship later) is used implicitly:
what one is doing here is to use 1 eV ' 1.60 × 10−19 J as the new fundamental
unit of energy.
With these new units, the mass of a proton is about 0.938 GeV/c2 , and the
mass of the electron is about 0.511 MeV/c2 . The fundamental energy level of
Hydrogen is about −13.6 eV.
In addition to this, nature is providing us with two constants which are
particularly appropriate for the world of fundamental physics: the speed of light
c ' 3.00×108 m/s = 3.00×1023 fm/s, and the Planck’s constant ~ ' 1.05×10−34
J s ' 6.58 × 10−16 eV s. It seems then natural to express speeds in terms of
c, and angular moments in terms of ~. When we do this we switch to the
so-called Natural Units (NU). The minimal set of natural units (not including
electromagnetism) could then be
Speed 1c 3.00 × 108 m/s
Angular momentum 1~ 1.05 × 1034 J s
Energy 1 eV 1.60 × 10−19 J
In such a system, ~ = c = 1.
After these conventions, just one unit can be used to describe the mechanical
Universe: we choose energy, and thus we can express al mechanical quantities
in terms of eV and of its multiples. It is immediate to express momenta and
masses directly in terms of energy; related to lengths and energies, we can use
the fact that
~c ' 1.97 × 10−13 MeV m = 3.15 × 10−26 J m

~' 6.58 × 10−22 MeV s = 1.05 × 10−34 J s
(the first relation can also be written as ~c ' 0.197 GeV fm. By choosing natural
units, all factors of ~ and c may be omitted from equations, which leads to
considerable simplifications. For example, the relativistic energy relation
E 2 = p 2 c2 + m 2 c4
becomes
E 2 = p2 + m2 .
To express 1 m and 1 s in NU, we can just write
1m 12 −1
1m = ~c ' 5.10 × 10 MeV
1s 21 −1
1s = ~ ' 1.52 × 10 .MeV
Both length and time are thus, in natural units, expressed as inverse of energy.
The first relation can also be written as 1fm ' 5.10GeV−1 : note that when you
when a quantity expressed in MeV−1 , in order to express it in GeV−1 , you must
multiply (and not divide) by a factor of 1000.
61
mks NU
Quantity p q r n
Action (~) 1 2 -1 0
Velocity (c) 0 1 -1 0
Mass 1 0 0 0
Length 0 1 0 -1
Time 0 0 1 -1
Momentum 1 1 -1 1
Energy 1 2 -2 1
Table 4.1: Dimensions of different physical quantities in the SI and NU.
Let us now find a general rule to transform quantities expressed in natural

units into SI units, and vice-versa.
To express a quantity in NU back to SI we first restore the ~ and c factors
by dimensional arguments and then use the conversion factors ~ and c (or ~c)
to evaluate the result. The dimension of c is [c] = [m/s]; the dimension of ~ is
[kgm2 s−1 ].
The vice-versa (from SI to NU) also easy. A quantity with metre-kilogram-
second (mks) dimensions M p Lq T r (where M represents the mass, L the length
and T the time) has the NU dimensions [E p−q−r ], where E represents energy.
Since ~ and c do not appear in NU, this is the only relevant dimension, and
dimensional checks and estimates are very simple. The quantity Q in the SI can
be expressed in NU as
p q
MeV−1

MeV
QNU = QSI 5.62 × 1029 5.10 × 1012
kg m
−1 r

MeV
× 1.52 × 1021 MeVp−q−r
s
The NU and SI dimensions are listed for some important quantities in Table
4.1.
Finally, let us discuss how to treat electromagnetism. To do so, we must
introduce a new unit, charge for example. We can redefine the unit charge by
observing that
e2
4π0
has the dimension of [J m], and thus is a pure number in NU. By dividing by
~c one has:
e2 1
' .
4π0 ~c 137
Imposing that the electric permeability of vacuum 0 = 1 (thus automatically
µ0 = 1 for the magnetic permeability of vacuum, since from Maxwell’s equations
62
0 µ0 = 1/c2 ) we obtain the new definition of charge, and with such a definition:
e2 1
α= ' .
4π 137
This is called the Lorentz-Heaviside convention. Elementary charge in NU be-
comes then a pure number:
e ' 0.303 .
63
Chapter 5
Geometrical optics
In cases where the wavelength is small compared to other length scales in a phys-
ical system, light waves can be modeled by light rays, moving on straight-line
trajectories representing the direction of the propagation. This is the domain
of the so-called geometrical optics.
5.1 Light propagation through different materi-

als; transmission of electromagnetic waves
If light travels through a transparent material, rather than in vacuum, the
electromagnetic equations still hold, substituting effective constants ε and µ
for ε0 and µ0 , where ε = εr ε0 and µ = µr µ0 (µr and εr are respectively the
relative magnetic constant and relative dielectric constant). µr ' 1 for non-
ferromagnetic materials, and εr ≥ 1; for example εr ' 1 + 6 × 10−4 in air at
NTP, εr ' 80 in water (which has a highly polar molecule), and εr ' 5 to
εr ' 10 in common glass.
So, in transparent (non-polarizable) materials, light travels with speed:
1 c c c
v = c0 = √ = √ ' √ =:
εµ εr µr εr n
√
where n ' εr ≥ 1 is the ratio between speed of light in vacuum and speed of
light through the material, and it is called refractive index.
5.2 Huygens’ principle

In the XVII century Huygens formulated a conjecture about the propagation of
waves that only later, around 1882, Kirchhoff1 demonstrated for electromagnetic
waves on the basis of Maxwell’s equations, introducing some improvements.
1 Gustav Kirchhoff (1824 - 1887) was a German physicist who contributed to the funda-
mental understanding of electrical circuits, spectroscopy, and the emission of black-body (he
coined the term “black body”) radiation by heated objects.
64
Figure 5.1: Huygens’ principle.
The Huygens’ principle states that all points on a given wave front can be
taken as point sources for the production of spherical secondary waves, called
wavelets, which propagate in the forward direction with the speed characteristic
of waves in that medium.
The corrections introduced by Kirchhoff were that the amplitude varies ac-
cording to a decreasing function of the angle θ with respect to the direction of
propagation (f (θ) ' (1 + cos θ)/2), and that the phase of the emitter is antic-
ipated by π/2 with respect to the phase of the wavefront. In many problems,
however, these two points can be neglected.
5.3 The laws of reflection and refraction

A Huygens’ construction can be used to derive the laws of reflection and refrac-
tion of light between two optical media with different indices of refraction; these
are called Snelll’s laws2 .
Assume a wave with fronts separated by a wavelength λ1 traveling with
speed v1 in an optically clear medium incident on the boundary with a second
optically clear medium.
Theorem 1. When a wave meets an obstacle on his path, the angle of incidence
θ1 is equal to the angle of reflection θ2 .
2 Willebrord Snell, or Snellius (1580 - 1626), was a Dutch astronomer and mathematician.
His name has been attached to the laws of reflection and of refraction of light.
65
To demonstrate it, we can look at the Figure 5.3: when the wave front
touches the surface of reflection, every point behaves like a source of a new
wave, that propagates form the point from which the principal wave came. The
result of these propagations is a wave that appears like the original wave, just
reflected by an angle equal to the angle θ1 .
A second application of the rule that Huygens postulated is the demonstra-
tion of the phenomenon of refraction.
Theorem 2. If a wave passes from a medium with rifraction index n1 to one
with index n2 , then the direction of the light obeys the following equation
n1 sin θ1 = n2 sin θ2 , (5.1)
where θ1 and θ2 are the angles formed by the direction of the wave with the
normal of the surface, before and after the change of medium of propagation
respectively (vi = c/ni is the speed of light in the two media).
Let us look at Figure 5.3. If we call T the time at which the wavefront
begins to cross the second medium, and T + ∆t the time at which a second ray
hits the second medium, then in a time of ∆t the wave formed by the wavelets
has moved inside the second medium of a distance of v2 ∆t. We obtain that
AC sin θ1 = v1 ∆t, and AC sin θ2 = v2 ∆t. From that we obtain
sin θ1 v1 n2
= = . (5.2)
sin θ2 v2 n1
Note that if n1 > n2 , an angle of incidency θc exists, called the critical angle,
for which the wave does not propagate in the second medium. This is because
the equation (5.2) becomes:

n1
θ2 = arcsin sin θ1 .
n2
Thus refraction (transmission) is possible only if

n1
−1 < sin θ1 < 1;
n2

n2 n2
arcsin − < θ1 < arcsin .
n1 n1
This principle is used in optical fibres: the waves in a tiny tube cannot exit from
the fibres because of the elevate refraction index.
We also observe that in the transmission between two media, frequency of
waves does not change, consistent with Huygens’ principle: in fact the wavelets
have all the same frequency as the frequency of the principal wave.
66
Figure 5.2: The phenomenon of reflection.
67
Figure 5.3: The phenomenon of refraction.
5.3.1 The Fermat principle

The Fermat3 principle is equivalent to the Huygens’ principle, and states that
a wave that propagates in more than one medium, covers the path that request
the minimum time.
To better understand this principle we can look at the following example.
When we are in a point S on a beach and we want to reach a point P in the
sea, the fastest path for us is not along the straight line that connects S and
P : it is better for us cover a longer distance on the sands, where our speed is
faster.
We try now to demonstrate the law of refraction using this principle. As we
can see in Figure 5.4 we have to move from S to P . We call C the point along
3 Pierre de Fermat (1607 - 1665) was a French lawyer in Toulouse, and an amateur math-
ematician who is given credit for early developments that led to infinitesimal calculus. He
made notable contributions to analytic geometry, probability, and optics, and in particular to
number theory.
68
Figure 5.4: The Fermat’s Principle.
our path in which the medium changes. The total time is given by
√ p
a 2 + x2 b2 + (c − x)2
t= + ;
v1 v2
dt x c−x
= √ − p =
dx 2
v1 a + x 2 v2 b + (c − x)2
2
sin θ1 sin θ2
= − .
v1 v2
Imposing the derivative of the function to equal 0, we get
sin θ1 v1 n2
= = .
sin θ2 v2 n1
that is the law of refraction.
69
Chapter 6
Interference and diffraction
Light waves interfere with each other much like mechanical waves do: inter-
ference associated with light waves arises when the electromagnetic fields that
constitute the individual waves combine.
Since light frequency is very high compared to the sensitivity of the human
eye (and of most instruments: the typical frequency of 1015 Hz is very difficult
to reach), for interference between two sources of light to be observed, there are
two conditions which must be met: the sources must be coherent (they must
maintain a constant phase with respect to each other), and the waves must have
identical wavelengths.
6.1 Young’s interference experiment

In 1801, Thomas Young1 experimentally proved that light is a wave, contrary to
what most other scientists then thought. He did so by demonstrating that light
undergoes interference, as do water waves, sound waves, and waves of all other
types. In addition, he was able to measure the average wavelength of sunlight;
his value, 570 nm, is impressively close to the modern accepted value of 555 nm.
We shall here examine Young’s experiment as an example of the interference of
light waves.
Figure (6.1) shows the basic arrangement of Young’s experiment. Light from
a distant monochromatic source of wavelength λ illuminates slit S0 in screen A.
The emerging light then spreads via diffraction to illuminate two slits S1 and
S2 in screen B (this is done to guarantee coherence of the two sources).
The snapshot of Figure (6.1) depicts the interference of the overlapping
waves. We observe the interference on a viewing screen C intercepting the light.
Where it does so, points of interference maxima form visible bright rows-called
bright bands, bright fringes, or (loosely speaking) maxima - that extend across
1 Thomas Young (1773 - 1829) was an English medical doctor and polymath. He made
notable scientific contributions to the fields of vision, light, solid mechanics, energy, physiology,
language, musical harmony, and in the decipherment of Egyptian hieroglyphs.
70
Figure 6.1: Arrangement of Young’s interference experiment.
the screen. Dark regions - called dark bands, dark fringes, or (loosely speaking)
minima - result from fully destructive interference and are visible between ad-
jacent pairs of bright fringes (maxima and minima more properly refer to the
center of a band.) The pattern of bright and dark fringes on the screen is called
an interference pattern. Figure (6.2) is a photograph of part of the interference
pattern that would be seen by an observer standing to the left of screen C at a
distance L d, where d is the distance between S1 and S2 .
Where are the dark fringes and the bright fringes located? Let us examine
the difference in optical path between the two rays in Figure (6.3).
When one wave travels an integer number of wavelengths farther than the
other, the waves arrive in phase, and a bright fringe occurs. This condition is
verified for
δ = r2 − r1 ' d sin θ = mλ
(this assumes the paths are parallel; they are not exactly parallel, but the above
is a very good approximation since L d). The absolute value of the integer
m is called the order of the maximum.
When destructive interference occurs, a dark fringe is observed. This needs
a path difference of an odd half wavelength

1
δ = r2 − r1 ' d sin θ = m + λ.
2
Let us take a coordinate z on the screen, starting from z = 0 at θ = 0. Then,
71
Figure 6.2: Interference pattern.
Figure 6.3: Difference of path length.
72
Figure 6.4: N-slits interference.
z = L tan θ, and, for small θ, z ' L sin θ. One has thus

L
zbright ' λ m (6.1)
d
L 1
zdark ' λ m+ (6.2)
d 2
(with m = 0, ±1, ±2, ...). Young’s setup is thus effective to measure the wave-
length of light, which is amplified by a factor L/d.
What happens in between the maxima and the minima? Let us solve the
general problem of the interference of N equally spaced sources in phase, each
of amplitude A (Figure (6.4). Let A(θ) be the common amplitude of each wave
(which can be a function of θ due to the attenuation).
The total field Etot as a function of θ is
N
X N
X N
X
Etot = En = A(θ)ei(krn −ωt) = A(θ)ei(kr1 −ωt) eik(n−1)d sin θ . (6.3)
n=1 n=1 n=1
With ζ = eikd sin θ , the sum is a sum of the geometrical series 1 + ζ + ζ 2 + ... +
73
ζ N −1 = (ζ N − 1)/(ζ − 1). Thus
eikN d sin θ − 1
Etot = A(θ)ei(kr1 −ωt)
eikd sin θ − 1
eik(N/2)d sin θ eik(N/2)d sin θ − e−ik(N/2)d sin θ
= A(θ)ei(kr1 −ωt) ik(1/2)d sin θ ik(1/2)d sin θ
e e − e−ik(1/2)d sin θ
sin (N/2)kd sin θ
= A(θ)ei(kr1 −ωt) eik((N −1)/2)d sin θ .
sin (1/2)kd sin θ
The amplitude is thus
sin (N/2)kd sin θ sin (N α/2)

Atot (θ) = A(0) ≡ A(0)
sin (1/2)kd sin θ sin (α/2)
with
2πd sin θ
α ≡ kd sin θ = .
λ
The amplitude at θ = 0 is
Atot (0) = lim Atot (θ) = N A(0) .

θ→0
and thus
Atot (θ) A(θ) sin (N α/2)
= .
Atot (0) A(0) N sin (α/2)
When we go to intensities,
2
Itot (θ) A(θ) sin (N α/2)
= .
Itot (0) A(0) N sin (α/2)
Even for large angles, the effect of A(θ) is to simply act as an envelope
function of the oscillating sine functions. We can always bring A(θ) back in if
we want to, but the more interesting behavior of Atot (θ)) is the oscillatory part.
We are generally concerned with the locations of the maxima and minima of
the oscillations and not with the actual value of the amplitude. We can pose,
for the moment, Atot (θ) ' Atot (0).
What does the Itot (θ)/Itot (0) ratio look like as a function of θ? The plot for
N = 4 is shown in Figure (6.5). If we are actually talking about small angles,
then we have α = kd sin θ ' kdz/D.
6.1.1 The two-slit case

In particular, for the two-slit case,
2
Itot (θ) sin (2α/2)
=' = cos2 (α/2) . (6.4)
Itot (0) N sin (α/2)
74
Figure 6.5: 4-slits interference.
6.2 Diffraction
Any wave passing through an opening experiences diffraction. Diffraction means
that the wave spreads out on the other side of the opening rather than the
opening casting a sharp shadow.
Diffraction is most noticeable when the opening is about the same size as
the wavelength of the wave. If light passes through a narrow slit, it produces a
characteristic pattern of light and dark areas called a diffraction pattern, which
can be explained by the interference of light traveling different optical paths.
Light passing a sharp edge also exhibits a diffraction pattern. Huygens’
Principle describes this spreading out, and a Huygens’ construction can be used
to quantify the diffraction phenomenon. For example, Figure (6.6) shows co-
herent light incident on an opening, which has dimensions comparable to the
wavelength of the light. Rather than casting a sharp shadow, light spreads out
on the other side of the opening. We can describe this spreading out by using a
Huygens’ construction and assuming that spherical wavelets are emitted at sev-
eral points inside the opening. The resulting light waves on the right side of the
opening undergo interference and produce a characteristic diffraction pattern.
Light waves can also go around the edges of barriers. In this case, the light
far from the edge of the barrier continues to travel like the light waves shown in
Figure (6.6). The light near the edge of the barrier seems to bend around the
barrier and is described by the sources of wavelets near the edge.
75
Figure 6.6: Diffraction through a single slit.
Figure 6.7: Diffraction through a single slit.
6.2.1 Diffraction through a single slit

Huygens’ principle requires that the waves spread out after they pass through
slits. This spreading out of light from its initial line of travel is called diffraction;
in general, diffraction occurs when waves pass through small openings, around
obstacles or by sharp edges.
Fraunhofer diffraction occurs when the rays leave the diffracting object in
parallel directions (screen very far from the slit).
We shall now discuss what happens when a plane wave impinges on just
one wide slit with width a instead of a number of infinitesimally thin ones (see
Figure 6.7).
By Huygens’ principle, we can consider the wide slit to consist of an infinite
number of line sources (or point sources, if we ignore the direction perpendicular
to the page) next to each other, each creating a cylindrical wave. In other words,
the diffraction pattern from one continuous wide slit is equivalent to the large-N
76
limit of the N −slit result.
We shall perform this calculation by an integral over all the phases from
the possible paths from different parts of the wide slit. Let the slit run from
y = −a/2 to y = a/2, and let B(θ)dy be the amplitude that would be present
at a location θ on the screen if only an infinitesimal slit of width dy was open.
So B(θ) is the amplitude (on the screen) per unit length (in the slit): B(θ)dy is
the analog of the A(θ) in the case of interference. If we measure the pathlengths
relative to the midpoint of the slit, then the path that starts at position y is
shorter by y sin θ. It therefore has a relative phase of e−ky sin θ .
Integrating over all the paths that emerge from the different values of y
(through slits of width dy) gives the total wave at position θ on the screen as
(up to an overall phase from the y = 0 point, and ignoring the temporal part of
the phase)
Z +a/2
Etot (θ) = dy B(θ) e−iky sin θ .
−a/2
This is the continuous version of√the discrete sum in the case of the multi-slit
interference. B(θ) falls off like 1 = r; however, we shall assume that θ is small,
which means that we can assume cosθ ' 1.
Z +a/2
B(0) −ik(a/2) sin θ
Etot (θ) ' B(0) dy e−iky sin θ = e − eik(a/2) sin θ
−a/2 −ik sin θ
−2i sin ka sin θ

2
= B(0)
−ik sin θ
sin 1 ka sin θ

= B(0)a 1 2 .
2 ka sin θ
There is no phase here, so this itself is the amplitude Atot (θ). Taking the
usual limit at θ = 0, we obtain Atot (0) = B(0)a. Therefore, Atot (θ)/Atot (0) =
sin(β/2)/(β/2), where β = ka sin θ. Since the intensity is proportional to the
square of the amplitude, we arrive at the equation
2
Itot (θ) sin(β/2)
=
Itot (0) β/2
with β = ka sin θ = 2πa sin θ/λ.
From Figure (6.8), we see that most of the intensity of the diffraction pattern
is contained within the main bump where β < 2π. Numerically, you can shown
that the fraction of the total area that lies under the main bump is about 90%.
So it makes sense to say that the angular width of the pattern is given by
λ
sin θ ' .
a
6.2.2 Diffraction grating

The diffracting grating consists of many equally spaced parallel slits. A typical
grating contains several thousand lines per centimeter.
77
Figure 6.8: The function sin2 (β/2)/(β/2)2 .
The intensity of the pattern on the screen is the result of the combined effects
of interference and diffraction.
6.2.3 The diffraction limit

Making a perfect lens that produces flawless images has been a dream of lens
makers for centuries. This is however impossible: whenever an object is imaged
by an optical system, such as the lens of a camera, or our eyes, fine features are
permanently lost in the image.
Roughly, the diffraction limit is given by the fact that a lens of aperture a
has an intrinsic diffraction of a light of wavelength λ such that the minimum
uncertainty δθ on the measurement of angles is, as seen in the previous Section,
λ
δθ ' k
a
where k depends on the geometry of the lens and on the definition of uncertainty,
and it is of the order of unity.
For example, the human eye has an opening of a diameter of about 1mm.
Thus the minimum angle we can appreciate (let us assume a wavelength of 500
nm) is about 500 nm/1mm = 0.5 mrad. The Moon is about d ' 4 × 108 km
away from us: then, when we look to the Moon, one pixel of our view has – at
best: the atmosphere introduces additional distortions – a typical size of 200
km. A perfect 10-m telescope could register pixels of 20 m: size matters in
astronomy and optics in general. In any case, we could not take from the Earth
a picture of the astronauts on the Moon.
78

Consequences of the invariance of the speed of light:

a reminder
• From the principle of relativity it follows that the velocity

of propagation of electromagnetic interactions is the same
in all inertial systems of reference
• We assume that the velocity of propagation of interactions
is a universal constant, and equal to c
– Relativity of simultaneity
– Time dilation
– Length contraction
(time and space intervals are not absolute)

More on length contraction

• The observers on the car, whose frame of reference we
have chosen to call S’, measure the length of their car
to be L’. They verify this by watching the front end of
the car and then the rear end of the car pass the center
of the platform, and measuring the amount of time t’
that passes between those two events. According to the
observers on the car, the car is at rest and the subway
platform is moving at a velocity -V relative to the car.
L’ = V t’
• To an observer standing at the center of the platform when the subway car goes by, the car
has a length L that can be calculated by measuring the time that passes between the time
when the front end of the car reaches the center of the platform and the time when the rear
end of the car reaches the center of the platform: L=Vt
• The interval t is a proper time interval for the observer standing on the platform, because the
two events whose times are being measured both happen at the center of the platform.
Therefore the time intervals are related by
L'
t'= t L =

Examples of time dilation

and length contraction:
The muon experiment
Think of the muon’s point

of view…

The propagation of signal
• Let’s look at two inertial reference systems K and K' with coordinate axes XYZ and
X’ Y’ Z’, where the system K' moves relative to K along the X(X') axis.
• Suppose signals start out from some point A on the X' axis in two opposite
directions. Since the velocity of propagation of a signal in the K' system, as in all
inertial systems, is equal (for both directions) to c, signals will reach points B and C,
equidistant from A, at the same time (in the K' system)
• The same two events (arrival of the signal at B and C) can by no means be
simultaneous for an observer in the K system. In fact, the velocity of a signal
relative to the K system has, according to the principle of relativity, the same value
c, and since the point B moves (relative to the K system) toward the source of its
signal, while the point C moves in the direction away from the signal (sent from A
to C), in the K system the signal will reach point B earlier than point C.
• We shall frequently use the concept of event. An event is described by the place
where it occurred and the time when it occurred. Thus an event occurring in a
certain material particle is defined by the three coordinates of that particle
• We designate the time in the systems K and K' by t and t'
Intervals

Transformation of intervals
As already shown, if ds = 0 in one inertial system, then ds' = 0 in any other system.
On the other hand, ds and ds' are infinitesimals of the same order. From these two
conditions it follows that ds2 and ds'2 must be proportional to each other:

Timelike events
Timelike events (2)

Spacelike events
• Let us take some event O as our origin of time and

space coordinates. Let us now consider what relation
Future and past
other events bear to the given event O. For
visualization, we shall consider only one space
dimension and the time, marking them on two axes.
• Uniform rectilinear motion of a particle, passing
through x = 0 at t = 0, is represented by a straight line
going through O and inclined to the t axis at an angle
whose tangent is the velocity of the particle. Since the
maximum possible velocity is c, there is a maximum
angle which this line can subtend with the t axis.
• The two lines represent the propagation of signals (with
the velocity of light) in opposite directions passing
through O (i.e. going through x=0 at t=0). All lines
representing the motion of particles can lie only in the
regions aOc and dOb. On the lines ab and cd, x = ± ct.
• First consider events whose world points lie within the
region aOc. For all the points of this region c2t2-x2 >0.
The interval between any event in this region and the
event O is timelike. In this region t>0, i.e. all the events
in this region occur "after" the event O. But two events
which are separated by a timelike interval cannot occur
simultaneously in any reference system. Consequently
it is impossible to find a reference system in which any
of the events in region aOe occurred "before" the event
O, i.e. at time t<0. Thus all the events in region aOc are
future events relative to O in all reference systems. This
region can be called the absolute future relative to O.

• In the same way, all events in the region bOd are in the
Future and past
absolute past relative to O; i.e. events in this region
occur before the event O in all systems of reference. (2)
• Next consider regions dOa and eOb. The interval
between any event in this region and the event O is
spacelike. These events occur at different points in
space in every reference system. Therefore these
regions can be said to be absolutely remote relative to
O.
• Note that if we consider all three space coordinates
instead of just one, then instead of the two intersecting
lines of Fig. 2 we would have a “cone" x2 + y2 + z2 –
c2t2 = 0 in the 4-dimensional coordinate system x, y, z.
t, the axis of the cone coinciding with the t axis. (This
cone is called the light cone.) The regions of absolute
future and absolute past are then represented by the two
interior portions of this cone.
• Two events can be related causally to each other only if
the interval between them is timelike; this follows
immediately from the fact that no interaction can
propagate with a velocity greater than the velocity of
light. As we have just seen, it is precisely for these
events that the concepts "earlier" and "later" have an
absolute significance, which is a necessary condition for
the concepts of cause and effect to have meaning.
The Lorentz transformations

The Lorentz transformations (2)

Length contraction, again

Transformation of velocities
Exercises at home
• Are Lorentz transformations commutative?

• Derive time dilation from the Lorentz
transformations

Doppler effect in classical physics:

General solution
• Numerator – Receiver (observer)

– Toward +
– Away –
• Denominator – Source
– Toward –
– Away +
Doppler effect from Lorentz transformations
• Now, the signal transmitted by one observer, and received by another, is a light
wave. It makes a difference compared to the situation in Doppler effect with sound
waves: there is no medium, and the velocity of the wave (i.e., of light) is the same
for both observers. And the the “transmitter” and “receiver” move relative to each
other with such a speed that relativistic effects have to be taken into account.
• We will consider the following situation: the “transmitter” is in the frame O that
moves away with speed -u (meaning: to the left) from the frame O’ in which the
“receiver” located. At a moment the “transmitter” starts to broadcast a light wave.

--(/(-.

•"'*
)
&
• $

"'0
"'1+0
,"'
•
$’
$%
•
)
t0$

+ ,& $
1#t0
t'(c + u)
'=
t 0

c t'(c + u) t 0 1
'=
= '=
' t 0 t' 1+ u /c
t'

=
t 0
1 u 2 /c 2
'=
1+ u /c
(1 u /c)(1+ u /c)
=
(1+ u /c)(1+ u /c)
1 u /c
=
1+ u /c
-.

#

"

"

! "

4-vectors

4-tensors

Unit tensor and metric tensor
The completely antisymmetric tensor of rank 4

Covariant derivatives

μ
μ ; μ
x x μ
μ 1 2
μ = 2 2 2 =
c t
4-dimensional velocity

Relativistic dynamics: mass and energy
E = h

( (

E=mc2
Relativistic momentum

Relativistic momentum
In addition, it is proportional to the space part of the 4-velocity
Energy-momentum vector (4-momentum)

pμ = m uμ = mc (1 , v/c) is instead a 4-vector
pμpμ = m2c2 = E2/c2 – p2

The photon
Relativistic force

Vector potential and scalar potential

By defining Aμ = (/c , A) and jμ = (c , j) the above equations become

Aμ = 0 jμ

The continuity equation tells that • j + = 0 μ j μ = 0 ; thus jμ is a 4-vector.
t
Also Aμ is then a 4-vector.
Electromagnetism and relativity

The electromagnetic tensor
E i = 0 A i + i A 0 F 0i = E i
1ˆ 2ˆ 3ˆ
B i = 1 2 3
1
A A2 A3
Covariant expression of Maxwell’s equations
μ F μ = μ ( μ A A μ ) = (μ μ ) A ( μ A μ )

Ma μ A μ = + A = 0
t
μ F μ = j
• Non ce ne sarebbe bisogno, ma…

Invariants constructed from the electromagnetic tensor

Lorentz force: covariant expression

Lorentz force: covariant expression
Motion of a particle under constant force
dp 0 dp 0 m
qeEu = ds dt = dt 1 v 2 = Fv
d

Decay of an unstable particle

Elastic collision (1 + 2 1’+2’)

Elastic collision: target at rest

Compton scattering
• If the incident particle 1 has 0 mass (photon)
E'1 E1 + m(E'1 E1 ) 1 1

cos = = 1+ m

E'1 E1 E1 E'1
1 1 mc
= 1 m
= 1 (')
h ' h h

The fall of Classical Physics
Classical physics: Fundamental Models
Particle Model (particles, bodies)

Motion in 3 dimension; for each time t, position and speed
are known (they are well-defined numbers, regardless we
know them). Mass is known.
Systems and rigid objects
Extension of particle model
Wave Model (light, sound, …)

Generalization of the particle model: energy is transported,
which can be spread (de-localized)
Interference
2
1
Classical physics at the end of XIX Century
Scientists are convinced that the particle and wave model can describe the
evolution of the Universe, when folded with
Newton’s laws (dynamics)
Description of forces
Maxwell’s equations
Law of gravity.
…
We live in a 3-d world, and motion happens in an absolute time. Time and
space (distances) intervals are absolute.
The Universe is homogeneous and isotropical; time is homogeneous.

Relativity
The physics entities can be described either in the particle or in the wave
model.
Natura non facit saltus (the variables involved in the description are
continuous).
Something is wrong
Relativity, continuity, wave/particle (I)
Maxwell equations are

not relativistically
covariant!
Moreover, a series of
experiments seems to
indicate that the
speed of light is
constant (Michelson-
A speed!
Morley, …)
4
2
Something is wrong
Relativity, continuity, wave/particle (IIa)
In the beginning of the XX
century, it was known that
atoms were made of a
heavy nucleus, with
positive charge, and by
light negative electrons
Electrostatics like gravity:
planetary model
All orbits allowed
But: electrons, being
accelerated, should radiate
and eventually fall into the
nucleus
Something is wrong
Relativity, continuity, wave/particle (IIb)
If atoms emit energy in the form of photons due to
level transitions, and if color is a measure of energy,
they should emit at all wavelengths – but they don’t
3
Something is wrong
Relativity, continuity, wave/particle (III)
Radiation has a particle-like behaviour, sometimes
Particles display a wave-like behaviour, sometimes
=> In summary, something wrong involving the

foundations:
Relativity
Continuity
Wave/Particle duality
Need for a new physics

A reformulation of physics was needed
This is fascinating!!! Involved philosophy, logics, contacts
with civilizations far away from us…
A charming story in the evolution of mankind
But… just a moment… I leaved up to now with classical

physics, and nothing bad happened to me!
Because classical physics fails at very small scales, comparable with
the atom’s dimensions, 10-10 m, or at speeds comparable with the
speed of light, c ~ 3 108 m/s
Under usual conditions, classical physics makes a good job.
Warning: What follows is logically correct, although

sometimes historically inappropriate.
8
4
I
Light behaves like a particle,
sometimes
Photoelectric Effect Features

and Photon Model explanation
The experimental results contradict all four classical
predictions
Einstein interpretation: All electromagnetic radiation can be
considered a stream of quanta, called photons
A photon of incident light gives all its energy hƒ to a single
electron in the metal
h is called the Planck constant, and plays a

fundamental role in Quantum Physics
10
5
The Compton Effect
Compton dealt with Einstein’s idea of

photon momentum
Einstein: a photon with energy E carries a
momentum of E/c = hƒ / c
According to the classical theory,
electromagnetic waves of frequency ƒo
incident on electrons should scatter,
keeping the same frequency – they
scatter the electron as well…
11
Compton’s experiment showed that, at

any given angle, a different frequency of
radiation is observed
The graphs show the scattered x-ray for
various angles
Again, treating the photon as a particle of
energy hf explains the phenomenon. The
shifted peak, ‘> 0, is caused by the
scattering of free electrons
This is called the Compton shift equation

12
6
Compton Effect, Explanation
The results could be explained, again, by treating the
photons as point-like particles having
energy hƒ
momentum hƒ / c
Assume the energy and momentum of the isolated
system of the colliding photon-electron are conserved
Adopted a particle model for a well-known wave
The unshifted wavelength, o, is caused by x-rays
scattered from the electrons that are tightly bound to
the target atoms
The shifted peak, ', is caused by x-rays scattered
from free electrons in the target
13
Blackbody radiation
Every object at T > 0 radiates electromagnetically, and
absorbes radiation as well
Stefan-Boltzmann law:
Blackbody: the
perfect absorber/emitter
“Black” body
Classical interpretation: atoms in the object vibrate; since <E> ~

kT, the hotter the object, the more energetic the vibration, the
higher the frequency
The nature of the radiation leaving the cavity through the
hole depends only on the temperature of the cavity walls 14
7
Experimental findings
& classical calculation
Wien’s law: the emission
peaks at
Example: for Sun T ~ 6000K
But the classical calculation

(Rayleigh-Jeans) gives a
completely different result…
Ultraviolet catastrophe
15
Experimental findings
& classical calculation
Classical calculation (Raileigh-
Jeans): the blackbody is a set of
oscillators which can absorb any
frequency, and in level transition
emit/absorb quanta of energy:
No maximum; a ultraviolet
catastrophe should absorb all
energy Experiment
16
8
Planck’s hypothesis
Only the oscillation modes for which
E = hf
are allowed…
17
Interpretation
Elementary oscillators can have only E n
quantized energies, which satisfy 4hf 4
E=nhf (h is an universal constant, n is 3
3hf
an integer –quantum- number)
2hf 2
Transitions are accompanied by the
hf 1
emission of quanta of energy (photons)
The classical calculation is accurate for large

wavelengths, and is the limit for h -> 0
18
9
Which lamp emits e.m. radiation ?
1) A
2) B
3) A & B
4) None
19
Particle-like behavior of light:

now smoking guns…
The reaction
has been
recorded
millions of
times…
20
10
Bremsstrahlung
"Bremsstrahlung" means in German
"braking radiation“; it is the radiation
emitted when electrons are decelerated or
"braked" when they are fired at a metal
target. Accelerated charges give off
electromagnetic radiation, and when the
energy of the bombarding electrons is
high enough, that radiation is in the x-ray
region of the electromagnetic spectrum. It
is characterized by a continuous
distribution of radiation which becomes
more intense and shifts toward higher
frequencies when the energy of the
bombarding electrons is increased.
21
Summary
The wave model cannot explain the behavior of light
in certain conditions
Photoelectric effect
Compton effect
Blackbody radiation
Gamma conversion/Bremsstrahlung
Light behaves like a particle, and has to be
considered in some conditions as made by single
particles (photons) each with energy
h ~ 6.6 10-34 Js is called the Planck’s constant

22
11
II
Particles behave like waves,
sometimes
23
Should, symmetrically, particles display

radiation-like properties?
The key is a diffraction experiment: do particles show
interference?
A small cloud of Ne atoms was cooled down to T~0. It
was then released and fell with zero initial velocity onto
a plate pierced with two parallel slits of width 2 μm,
separated by a distance of d=6 μm. The plate was
located H=3.5 cm below the center of the laser trap.
The atoms were detected when they reached a screen
located D=85 cm below the plane of the two slits. This
screen registered the impacts of the atoms: each dot
represents a single impact. The distance between two
maxima, y, is 1mm.
The diffraction pattern is consistent with the diffraction
of waves with
24
12
Diffraction of electrons
Davisson & Germer 1925:
Electrons display diffraction patterns !!!
25
de Broglie’s wavelength
What is the wavelength associated to a particle?
de Broglie’s wavelength:
Explains quantitatively the diffraction by Davisson and Germer……
Note the symmetry
What is the wavelength of an electron moving at 107 m/s ?
(smaller than an atomic length; note the dependence on m)

26
13
Atomic spectra
Why atoms emit according to a discrete energy spectrum?
Balmer
Something must
be there... 27
Electrons in atoms: a semiclassical model

Similar to waves on a cord, let’s imagine that
the only possible stable waves are stationary…
2r=n n=1,2,3,…
=> Angular momentum is quantized (Bohr

postulated it…)
28
14
Hydrogen (Z=1)
v
m
r
F
NB:
• In SI, ke = (1/40) ~ 9 x 109 SI units
• Total energy < 0 (bound state)
• <Ek> = -<Ep/2> (true in general for bound states, virial theorem)
Only special values are possible for the radius !
29
Energy levels
The radius can only assume
values
The smallest radius (Bohr’s radius) is
Radius and energy are related:
And thus energy is quantized:
30
15
Transitions
An electron, passing from an orbit of energy Ei
to an orbit with Ef < Ei, emits energy [a photon
such that f = (Ei-Ef)/h]
31
Level transitions and energy quanta
We obtain Balmer’s relation!

32
16
Limitations
Semiclassical models wave-particle duality can explain
phenomena, but the thing is still insatisfactory,
When do particles behave as particles, when do they behave
as waves?
Why is the atom stable, contrary to Maxwell’s equations?
We need to rewrite the fundamental models, rebuilding

the foundations of physics…
33
Wavefunction
Change the basic model!

We can describe the position of a particle
through a wavefunction (r,t). This can account
for the concepts of wave and particle (extension
and simplification).
Can we simply use the D’Alembert waves, real
waves? No…
34
17
Wavefunction - II
We want a new kind of “waves” which can account for particles, old waves,
and obey to F=ma.
And they should reproduce the characteristics of “real” particles: a
particle can display interference corresponding to a size of 10-7 m, but

have a radius smaller than 10-10 m
Waves of what, then? No more of energy,
but of probability
The square of the wavefunction is the intensity, and it gives the

probability to find the particle in a given time in a given place.
Waves such that F=ma? We’ll see that they cannot be a function in R,
but that C is the minimum space needed for the model.
35
SUMMARY
Close to the beginning of the XX century, people thought that
physics was understood. Two models (waves, particles). But:
Quantization at atomic level became experimentally evident
Particle-like behavior of radiation: radiation can be considered in some
conditions as a set of particles (photons) each with energy
Wave-like property of particles: particles behave in certain condistions as

waves with wavenumber
Role of Planck’s constant, h ~ 6.6 10-34 Js

Concepts of wave and particle need to be unified: wavefunction
(r,t).
36
18
L’equazione di Schroedinger
37
Proprieta’ della funzione d’onda
38
19
L’equazione di S.
39
20
L1 – The Schroedinger
equation
• A particle of mass m on the x-axis is subject to
a force F(x, t)
• The program of classical mechanics is to
determine the position of the particle at any
given time: x(t). Once we know that, we can
figure out the velocity v = dx/dt, the momentum
p = mv, the kinetic energy T = (1/2)mv2, or any
other dynamical variable of interest.
• How to determine x(t) ? Newton's second law:
F = ma.
– For conservative systems - the only kind that occur
at microscopic level - the force can be expressed as
the derivative of a potential energy function, F = -dV/
dx, and Newton's law reads m d2x/dt2 = -dV/ dx
– This, together with appropriate initial conditions
(typically the position and velocity at t = 0),
determines x(t).
1
• A particle of mass m, moving along the
x axis, is subject to a force
F(x, t) = -dV/ dx
• Quantum mechanics approaches this
same problem quite differently. In this
case what we're looking for is the wave
function, (x, t), of the particle, and we
get it by solving the Schroedinger
equation:
~ 10-34 Js
• In 3 dimensions,
2
The statistical interpretation
• What is this "wave function", and what can it tell you?
After all, a particle, by its nature, is localized at a point,
whereas the wave function is spread out in space (it's
a function of x, for any given time t). How can such an
object be said to describe the state of a particle?
• Born's statistical interpretation:
Quite likely to find the particle near A, and relatively

unlikely near B.
• The statistical interpretation introduces a kind of
indeterminacy into quantum mechanics, for even if you
know everything the theory has to tell you about the
particle (its wave function), you cannot predict with
certainty the outcome of a simple experiment to
measure its position
– all quantum mechanics gives is statistical information
about the possible results
• This indeterminacy has been profoundly disturbing
3
Realism, ortodoxy,
agnosticism - 1
• Suppose I measure the position of the particle,

and I find C. Question: Where was the particle
just before I made the measurement?
There seem to be three plausible answers to
this question…
1. The realist position: The particle was at C. This
seems a sensible response, and it is the one
Einstein advocated. However, if this is true QM
is an incomplete theory, since the particle
really was at C, and yet QM was unable to tell
us so. The position of the particle was never
indeterminate, but was merely unknown to the
experimenter. Evidently is not the whole
story: some additional information (a hidden
variable) is needed to provide a complete
description of the particle
4
Realism, ortodoxy,
agnosticism - 2

2. The orthodox position: The particle wasn't
really anywhere. It was the act of
measurement that forced the particle to "take a
stand“. Observations not only disturb what is to
be measured, they produce it .... We compel
the particle to assume a definite position. This
view (the so-called Copenhagen interpretation)
is associated with Bohr and his followers.
Among physicists it has always been the most
widely accepted position. Note, however, that
if it is correct there is something very peculiar
about the act of measurement - something that
over half a century of debate has done
precious little to illuminate. 5
Realism, ortodoxy,
agnosticism - 3

3. The agnostic position: Refuse to answer. This
is not as silly as it sounds - after all, what
sense can there be in making assertions about
the status of a particle before a measurement,
when the only way of knowing whether you
were right is precisely to conduct a
measurement, in which case what you get is
no longer "before the measurement"? It is
metaphysics to worry about something that
cannot, by its nature, be tested. One should
not think about the problem of whether
something one cannot know anything about
exists
6
Realism, ortodoxy,
or agnosticism?

• Until recently, all three positions had their
partisans. But in 1964 John Bell demonstrated
that it makes an observable difference if the
particle had a precise (though unknown)
position prior to the measurement. Bell's
theorem made it an experimental question
whether 1 or 2 is correct. The experiments
have confirmed the orthodox interpretation: a
particle does not have a precise position prior
to measurement; it is the measurement that
insists on one particular number, and in a
sense creates the specific result, statistically
guided by the wave function.
• Still some agnosticism is tolerated…
7
• Suppose I measure the
Collapse of position of the particle, and I
find C. Question: Where will
the be the particle immediately
wavefunction after?
• Of course in C. How does the
orthodox interpretation explain
that the second measurement
is bound to give the value C?
Evidently the first
measurement radically alters
the wave function, so that it is
now sharply peaked about C.
The wave function collapses
upon measurement (but soon
spreads out again, following
the Schroedinger equation, so
the second measurement
must be made quickly). There
are, then, two entirely distinct
kinds of physical processes:
"ordinary", in which evolves
under the Schroedinger
equation, and
"measurements", in which
suddenly collapses.
8
Normalization
• | (x, t)|2 is the probability density for finding

the particle at point x at time t.
The integral of | (x, t)|2 over space must be 1
(the particle has to be somewhere).
• The wave function is supposed to be
determined by the Schroedinger equation--we
can't impose an extraneous condition on
without checking that the two are consistent.
• Fortunately, the Schroedinger equation is
linear: if is a solution, so too is A , where
A is any (complex) constant. What we must
do, then, is pick this undetermined
multiplicative factor so that The integral of |
(x, t)|2 over space must be 1 This process is
called normalizing the wave function.
• Physically realizable states correspond to the
"square-integrable" solutions to Schroedinger's
equation.
9
Will a normalized function stay as such?
10
Expectation values
• For a particle in state , the expectation value

of x is
• It does not mean that if you measure the

position of one particle over and over again,
this is the average of the results
– On the contrary, the first measurement (whose
outcome is indeterminate) will collapse the wave
function to a spike at the value obtained, and the
subsequent measurements (if they're performed
quickly) will simply repeat that same result.
• Rather, <x> is the average of measurements
performed on particles all in the state , which
means that either you must find some way of
returning the particle to its original state after
each measurement, or you prepare an
ensemble of particles, each in the same state
, and measure the positions of all of them:
<x> is the average of them.
11
Momentum
12
More on operators
• One could also simply observe that

Schroedinger’s equations works as if
(exercise: apply on the plane wave). In

3 dimensions,
Compound operators
• Kinetic energy is
13
Angular momentum
14
Exercise
• A particle is represented at t=0 by the wavefunction
(x, 0) = A(a2-x2) |x| < a (a>0).
=0 elsewhere
a Determine the normalization constant A
b, c What is the expectation value for x and for p at t=0?
15
Exercise (cont.)
• A particle is represented at t=0 by the wavefunction
(x, 0) = A(a2-x2) |x| < a (a>0).
=0 elsewhere
d, e Compute <x2>, <p2>
f, g Compute the uncertainty on x, p
h Verify the uncertainty principle in this case
16
L2 – The time-independent
Schroedinger equation
• Supponiamo che il potenziale U sia

indipendente dal tempo
17
18
, soluzione della prima equazione (eq.agli autovalori detta
equazione di S. stazionaria), e’ detta autofunzione
19
3 comments on the stationary solutions: 1
1. They are stationary states. Although

the wave function itself
does (obviously) depend on t, the

probability density does not - the time
dependence cancels out. The same
thing happens in calculating the
expectation value of any dynamical
variable
20
2. They are states of definite energy. In
mechanics, the total energy is called the
Hamiltonian:
H(x, p) = p2/2m + V(x).
The corresponding Hamiltonian operator,
obtained by the substitution p -> p operator
Note: it is true in general that, if is

eigenfunction of an operator, the measurement
gives as a result certainly the eigenvalue 21
3. They are a basis. The general solution

is a linear combination of separable
solutions. The time-independent
Schroedinger equation might yield an
infinite collection of solutions, each with
its associated value of the separation
constant; thus there is a different wave
function for each allowed energy.
The S. equation is linear: a linear
combination of solutions is itself a
solution.
It so happens that every solution to the
(time-dependent) S. equation can be
written as a linear combination of
stationary solutions.
To really play the game, mow we must

input some values for V 22
Free particle (V=0, everywhere)
23
The infinite square well
24
Infinite square well, 2
25
But…
26
27
11/6/12
Intermezzo: Heisenberg principle (theorem)
10
11
1
11/6/12
The uncertainty principle (qualitative)
x(m)
Imagine that you're holding one end of a long rope, and you generate a
wave by shaking it up and down rhythmically.
Where is that wave? Nowhere, precisely - spread out over 50 m or so.
What is its wavelength? It looks like ~6 m
x(m)
By contrast, if you gave the rope a sudden jerk you'd get a relatively narrow
bump traveling down the line. This time the first question (Where precisely is
the wave?) is a sensible one, and the second (What is its wavelength?)
seems difficult - it isn't even vaguely periodic, so how can you assign a
wavelength to it?
12
=> Uncertainty is a characteristic of the wave representation
The uncertainty principle (qualitative, II)
x(m)
x(m)
The more precise a wave's x is, the less precise is , and vice versa. A
theorem in Fourier analysis makes this rigorous…
This applies to any wave, and in particular to the QM wave function. is
related to p by the de Broglie formula
Thus a spread in corresponds to a spread in p, and our observation says

that the more precisely determined a particle's position is, the less precisely
is p
This is Heisenberg's famous uncertainty principle. (we'll prove it later, but I

want to anticipate it now)
13
2
11/6/12
General rules for the linear piece-wise potential
Divide the interval in N regions with constant

potential Vi
For each region, solve the Schroedinger equation
Real exponentials for E <Vi; imaginary otherwise
Impose boundary/initial conditions
Impose continuity conditions
14
15
3
11/6/12
16
17
4
11/6/12
Finite Square-Well Potential
The finite square-well potential is
The Schrödinger equation outside the finite well in regions I and III is
or using
yields . Considering that the wave function must be zero at

infinity, the solutions for this equation are
18
Finite Square-Well Solution

Inside the square well, where the potential V is zero, the wave equation
becomes where
Instead of a sinusoidal solution we have
The boundary conditions require that
and the wave function must be smooth where the regions meet.
Note that the

wave function is
nonzero outside
of the box.
19
5
11/6/12
Penetration Depth
The penetration depth is the distance outside the potential well where
the probability significantly decreases. It is given by
mx m
Et (V0 E) = (V0 E)
p 2m(V0 E) 2m(V0 E) 2
It should not be surprising to find that the penetration distance that

violates classical physics is proportional to Planck’s constant.
20
Barriers and Tunneling

Consider a particle of energy E approaching a potential barrier of height V0 and
the potential everywhere else is zero.
We will first consider the case when the energy is greater than the potential
barrier.
In regions I and III the wave numbers are:
In the barrier region we have
21
6
11/6/12
Reflection and Transmission
The wave function will consist of an incident wave, a reflected wave, and a
transmitted wave.
The potentials and the Schrödinger wave equation for the three regions are
as follows:
The corresponding solutions are:
As the wave moves from left to right, we can simplify the wave functions to:
22
Probability of Reflection and Transmission

The probability of the particles being reflected R or transmitted T is:
Because the particles must be either reflected or transmitted we have:

R + T = 1.
By applying the boundary conditions x ±, x = 0, and x = L, we arrive
at the transmission probability:
Notice that there is a situation in which the transmission probability is 1.
23
7
11/6/12
Tunneling
Now we consider the situation where classically the particle does not have enough
energy to surmount the potential barrier, E < V0.
The quantum mechanical result is one of the most remarkable features of modern
physics, and there is ample experimental proof of its existence. There is a small,
but finite, probability that the particle can penetrate the barrier and even emerge
on the other side.
The wave function in region II becomes
The transmission probability that

describes the phenomenon of tunneling is
24
Uncertainty Explanation
Consider when L >> 1 then the transmission probability becomes:
This violation allowed by the uncertainty principle is equal to the

negative kinetic energy required! The particle is allowed by quantum
mechanics and the uncertainty principle to penetrate into a classically
forbidden region. The minimum such kinetic energy is:
25
8
QM in 3 dimensions
1
V(r): separation of variables
2
ˆ = 2 2 ,
,
ˆ
ˆ

Y = ll 1Y 2 = ll 1
2

ˆ

2 . 3
The angular equation
4
The angular equation -

= i ,
ˆ

ˆ = m

ˆ = m

ˆ

. 5
The angular equation -
Pl are the Legendre polynomials, defined by the Rodrigues formula:
6
7
ˆ = l(l + 1)

8
Spherical harmonics
= (-1)m for m>=0 and =1 for m<0. The Y are orthogonal, so
9
The radial equation
10
The H atom
u(r) = rR(r)
11
Asymptotic behavior
u(r) = rR(r)
12
The radial solution
u(r) = rR(r)
n>l
is the q-th Laguerre polynomial.

Remember: only the solutions for n>l are valid functions 13
The radial
solution: energy
14
Ground state
15
n>1
16
n > 1 (continued)
Since they are eigenvectors for different eigenvalues
17
Example
18
Graphs
19
Angular momentum
20
21
22

Waves and Electromagnetism Wavefunctions Lecture Notes: Alessandro de Angelis University of Udine, December 2012

Uploaded by

Copyright:

Available Formats

Waves and Electromagnetism Wavefunctions Lecture Notes: Alessandro de Angelis University of Udine, December 2012

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Waves and Electromagnetism Wavefunctions Lecture Notes: Alessandro de Angelis University of Udine, December 2012

Uploaded by

Copyright:

Available Formats

Waves and Electromagnetism

University of Udine, December 2012

1 Differential calculus applied to vector fields 5

I Waves and electromagnetism 25

3 Electromagnetism and the Maxwell’s equations 46

4 Solutions of the Maxwell’s equations in vacuo 51

6 Interference and diffraction 70

7 Maxwell’s equations and Einstein’s special relativity 79

8 Lorentz transformations and the formalism of special relativity 88

9 Covariant formulation of electromagnetism 95

II Wave functions; introduction to quantum physics 99

12 Solving the Schrödinger equation in one dimension 103

13 Central potentials and the Hydrogen atom 104

Differential calculus applied

1.1 Scalar and vector fields

~h = lim ∆J ~ef (1.1)

1.2 The gradient

~ is often referred to as the “del operator”, since it “operates”

~ represents the spatial rate of change of T . The x−component of ∇T

1.3 The divergence

Giving a vector function F~ defined in a domain in <3 , we define as the

1.4 The curl

Geometrical interpretation: ∇~ × F~ is a measure of how much the vector F~

1.5 Second order derivatives, and the Laplacian

~ux ~uy ~uz

and as such it can be applied to vector functions as well, giving as output a

The explicit calculation confirms that this approach is justified.

1.6 The differential equation of heat flow

The constant of proportionality k is called the thermal conductivity of the ma-

(the minus sign is necessary because heat flows “downhill” in temperature.)

1.7 Differential operators in polar coordinates

Figure 1.5: Line integral.

The derivation of the above formulas is conceptually straightforward but

1.8 The gradient theorem

Now we move a little further, by an additional small displacement d~l2 ; the

and hence T (b) − T (a) = 0.

1.9 The divergence theorem (Gauss’ theorem)

where ~h is a vector function, and da is an infinitesimal patch of area; ~n is a

1.9.2 Flux through a closed surface

material in which after an initial heating no further heat energy is generated

and for the flux out of V2

1.9.3 The flux from a cube

Φ1 = −Cx (1) ∆y∆z .

where S is any closed surface and V is the volume inside it.

1.9.4 Heat conduction; the diffusion equation

1.10 The curl theorem (Stokes’ theorem)

1.10.2 The circulation around a square

matter what the particular orientation of the plane.

where S is any surface bounded by Γ.

1.11 Curl-free fields

Using Stokes’ theorem, we conclude that

We obtained in a different way a result we had obtained with vector algebra.

1.12 Dirac’s delta function

In the mathematical literature it is known as a generalized function, or

Integrals including δ(x) are perfectly legitimate, and in particular:

It is easy to generalize the delta function to three dimensions:

This three-dimensional delta function is zero everywhere except at (0, 0, 0),

Until 1600, scientists were dreaming to give a complete description of nature