Classical Electrodynamics: Selected Topics

C.U., Physics, PG 2nd semester
Anirban Kundu
Contents
1 Maxwell’s Equations 3
2 Four-vectors 5
2.1 The Relativistic Action and Distribution Functions . . . . . . . . . . . . . . . . . . . 7
3 Relativistic Formulation of Maxwell’s Equations 9
3.1 Field Tensor and Its Dual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2 Transformation of the Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.3 Fields due to an Uniformly Moving Particle . . . . . . . . . . . . . . . . . . . . . . . 11
4 Lorentz Force Equation and Its Generalisation 13
4.1 Motion in Combined Uniform and Static Electric and Magnetic Fields . . . . . . . . 14
5 Lagrangian and Equation of Motion 15
5.1 Charged Particle in an Electromagnetic Field: the Generalised Momentum . . . . . . 16
5.2 Lagrangian for the Electromagnetic Field . . . . . . . . . . . . . . . . . . . . . . . . 17
5.3 Energy and Momentum of the Electromagnetic Field: Poynting’s Theorem . . . . . . 19
6 Potential Formulation 21
6.1 Retarded Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6.2 Digression: Lorentz Invariance of Electric Charge . . . . . . . . . . . . . . . . . . . . 22
6.3 The Li´enard-Wiechert Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
6.4 Fields due to a Moving Charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
6.5 The Fate of the Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
7 Accelerated Charge 24
7.1 Radiation from a Slow-moving Charge . . . . . . . . . . . . . . . . . . . . . . . . . . 25
7.2 Relativistic Generalisation of Larmor’s Formula . . . . . . . . . . . . . . . . . . . . . 25
7.3 Relativistic Motion: v | a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
7.4 Frequency Distribution: Bremsstrahlung for Slow Electrons . . . . . . . . . . . . . . 28
7.5 Relativistic Motion: v ⊥ a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
7.6 Thomson Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
7.7 Modifications to the Thomson formula: Compton, Klein-Nishina . . . . . . . . . . . 33
7.8 Scattering from Bound Electrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
7.9 Cherenkov Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
8 Radiation Reaction 35
8.1 When is the Radiation Reaction Important? . . . . . . . . . . . . . . . . . . . . . . . 36
8.2 The Abraham-Lorentz Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
8.3 Problems with the Abraham-Lorentz Formula . . . . . . . . . . . . . . . . . . . . . . 38
8.4 Relativistic Motion: Dirac Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
8.5 Radiation Reaction on a Charged Oscillator . . . . . . . . . . . . . . . . . . . . . . . 40
8.6 Scattering and Absorption of Radiation by a Charged Oscillator . . . . . . . . . . . 41
This note is based upon the following excellent textbooks:
Jackson: Classical Electrodynamics
Panofsky and Phillips: Classical Electricity and Magnetism
Griffiths: Introduction to Electrodynamics
Raychaudhuri: The Theory of Electricity and Magnetism
Feynman Lectures, vol. 2
You are always advised to read the original textbooks. Remember that the supplementary problems
form an integral part of the course.
I will use the rationalised Lorentz-Heaviside system throughout. That system is explained in
Section 1. I will also assume, as prerequisite, that you know (i) the ordinary noncovariant form of
Maxwell’s equations; (ii) the basic postulates of the Special Theory of Relativity, and (iii) how to
get the Lagrangian and the Hamiltonian of a particle. In fact, for a certain section, I have to use
the Lagrangian formulation for a field, but probably you have already encountered that in your
Classical Mechanics course. Even if you have not, the formulation is analogous.
I will always consider electromagnetic fields in vacuum, ı.e., E = D and B = H. In fact, I will
never use D and H and associated quantities like the electric polarisation P or magnetisation M.
A four-dimensional vector will be labeled by a Greek index; all Greek indices will run
from 0 to 3, 0 being the time component. I will use the flat space-time Minkowski metric
η
µν
= diag(1, −1, −1, −1). All repeated indices are implicitly summed over.
This note is brief, and please remember that the problems are an integral part of the course; you
must do them before proceeding to the next section. I could not discuss a number of interesting
topics just for want of time. For example, I would like to discuss the physics of accelerated charges
in more detail, including the frequency dependence of radiated power; also the theory of half-
advanced and half-retarded potentials. There may be numerous typos. Please feel free to inform
me about them. You can contact me either at my office, or at akphy@caluniv.ac.in, where I am
always available.
2
1 Maxwell’s Equations
Why should we take this course? Is something exciting happening in classical electrodynamics?
The answer, unfortunately, is no; it is an age-old and established subject, like classical mechanics,
and almost all the fundamental discoveries were made in the 19th century.
Like classical mechanics, it is a tool (well, it is classical mechanics, only the force is provided
by electric or magnetic fields, and you study the motion of charged particles in these fields, or the
properties of the fields themselves), which is applied in many branches of physics, most notably in
astrophysics, plasma physics (either man-made or astrophysical), and particle physics, and in a lot of
applied branches as well. Let me give you just an example. You know that Special Relativity comes
into play when the velocities are very large, comparable to that of light, and General Relativity is
relevant for dealing with gravity. In astrophysics, you encounter both these situations. There are
all sorts of radiation in the sky, and a proper knowledge of them can help us unravel the mystery of
the origin of the universe, or at least the galaxies. You can map the sky not only with visible light
but also with invisible electromagnetic waves, from microwaves to gamma-rays. So you must know
how and where these radiations are generated. Relativistic classical electrodynamics helps you to
do that.
But electrodynamics has been quantised, and quantum electrodynamics is known to be one of the
greatest intellectual achievements of mankind. No other theory, not even gravitation (Newtonian
or Einsteinian), has been tested to such precision. Why, then, we still use the classical theory?
Indeed, there are places (e.g., if you wish to calculate the scattering cross-section of an electron in
the field of another electron, or the well-known Compton scattering) where a quantum calculation
is easier and more precise. But these systems are microscopic; when you apply your theory to a
macroscopic system (like a star), you have to use some kind of averaging over the ensemble, and
come back to the classical regime, thanks to the correspondence principle. Also, the classical theory
deals with concepts like electric and magnetic fields, which are easier to measure.
A good point to start this course is the set of equations that tells you almost everything about
classical electrodynamics (the force law of Lorentz completes the set). By the word classical we
mean that there are no photons, only the electromagnetic wave. Electrons are the objects with
which this classical field interacts. Sometimes the electrons are treated in a quantum-mechanical
way; that is called a semiclassical approximation. Of course, we have no time here to go into a full
quantum-mechanical theory of the interaction of electrons with photons.
In most textbooks (e.g., Panofsky and Phillips) you will see the use of the SI, sometimes called
the MKS, system. In that system Maxwell’s equations (in vacuum) read
∇.E =
ρ
ε
0
,
∇B = µ
0
j +µ
0
ε
0
∂E
∂t
,
∇E = −
∂B
∂t
,
∇.B = 0. (1)
The electric charge is expressed in coulomb; the charge of an electron is −1.6 10
−19
coulomb.
In Raychaudhuri, as well as in the second edition of Jackson, you will encounter the Gaussian
system, which, I must say, is gradually going out of use
1
:
∇.E = 4πρ,
1
Even though the system has become almost archaic, I will still recommend Raychaudhuri’s book to those who
are not lucky to have him as a teacher — he died in 2005 — but would like to have a glimpse of the physics insight
of a teacher sans pareil.
3
∇B =

c
j +
1
c
∂E
∂t
,
∇E = −
1
c
∂B
∂t
,
∇.B = 0. (2)
The Lorentz force law in these two systems is
F = q(E+v B) (SI), F = q

E+
1
c
v B

(Gaussian). (3)
In this note we will use the so-called rationalised Lorentz-Heaviside (RLH) system. This system
is characterised by a scaling of the fields and charge and current densities of the SI system:
E
SI
=
1

ε
0
E
RLH
, B
SI
=

µ
0
B
RLH
, q
SI
=

ε
0
q
RLH
, j
SI
=
1

µ
0
j
RLH
. (4)
This seems to be a complicated scaling, and to top it all, we set c = 1. That is not a blasphemy.
It just tells you that the units of length and time are related, and when you say the length of 1
second you actually mean the length light travels in 1 second (so the length of one year is actually
one light-year). It also tells you that any velocity, apart from being a dimensionless quantity, must
be a number whose magnitude is less than or equal to one. Some quantities get related too: energy
and mass now have same dimension, so we can safely talk about a proton having a mass of 938
MeV
2
.
But what do we gain? You may not believe it, but we have abolished all factors of µ
0
, ε
0
, and
c from our subsequent discussions. E and B now stand on the same footing, so do ρ and j. Of
course, physics does not change an iota, and if you like, at the end of your calculation, you can go
back to the familiar SI result with the help of eq. (4). And after rescaling, it is better to make a
dimensional analysis; you may have to introduce suitable powers of c in the expression
3
To verify this claim of elegance, let us look at the Maxwell’s equations, with eq. (4) and the
relation µ
0
ε
0
= 1/c
2
:
∇.E = ρ,
∇B = j +
∂E
∂t
,
∇E = −
∂B
∂t
,
∇.B = 0. (5)
An important point to note is that if ρ = j = 0, the equations are symmetric under the interchange
E →B, B →−E. We will come back to this later.
Under the parity transformation x →−x, it is clear that E →−E, B →B, j →−j. Under time
reversal t →−t, E does not change sign — it cannot if it is created by a static charge configuration
— but B →−B, which is intuitively easy to understand: a magnetic field is created by a current
and under time reversal, the current reverses its direction, so does B.
Of course, there is no free lunch, but the price to pay is small. We cannot express the electric
charge in the conventional unit of coulomb. Rather, the charge is something like 0.3 unit (this
comes from the fine structure constant, something you will learn later, whose experimental value is
2
This is the system that is used by astrophysicists and particle physicists. They also use ¯h = 1 which relates mass
with length and time: [M] = [L]
−1
= [T]
−1
. We won’t need it for a discussion of classical electrodynamics. You may
go further and use kB = 1; that will relate temperature with mass.
3
Our system only abolishes µ0, ε0 and c, but not numerical factors like 4π, so there should not be any problem
with dimensional analysis. An example will be given later.
4
approximately 1/137, and the theoretical expression, in the RLH system, and with ¯h = 1, is e
2
/4π
— in the SI system it is e
2
/4πε
0
¯hc). There are people who believe that this number is more handy
than something like 10
−19
!
Q. B is derivable from a vector potential A: B = ∇A. Show that for a uniform magnetic field
B, one can write A =
1
2
Br.
Q. Check eq. (5). How does the Lorentz force law look like?
Q. With c = 1, show that the Lorentz transformation equations in 1 + 1 (i.e., one space and one
time) dimension can be written as t

= t cosh θ −xsinh θ, x

= xcosh θ −t sinh θ. How is θ related
with v?
Q. With c = ¯ h = 1, find how MeV is related to s
−1
and fm
−1
(1 fm= 10
−13
m).
Q. Show that the energy density of the electromagnetic field and the Poynting vector are respec-
tively given by
1
2

E
2
+B
2

and EB in the RLH system.
2 Four-vectors
There are two postulates of the Special Theory of Relativity (STR): physical laws are invariant
in all frames which are mutually inertial, and the velocity of light in vacuum is a constant in all
inertial frames. A consequence of the second postulate is that nothing can travel faster than light
(in vacuum). This consequence rules out action-at-a-distance; every information must take some
time to proceed from one point to another
4
. So if I shake an electron here, another electron at
a distance of one light-year will feel it not before one year, and if the sun vanishes right now, the
earth will fly off in a tangent only after eight minutes (assuming that the gravitational information
travels at the speed of light). The objects which carry these informations are called fields. We may
even quantise these fields and get the corresponding field excitations, the particles. It is just a small
matter that electrodynamics was quantised long ago (and some other forces which are analogous to
electrodynamics were born quantised), and gravitation resisted all attempts (and by the smartest
minds on this planet!) of quantisation.
Now to the more mundane subject of four-vectors. A three-vector has three spatial components,
they satisfy certain transformation laws, and their products are defined in a certain way. Four-
vectors have one temporal and three spatial components. They are written as
A
µ
≡ (A
0
, A). (6)
Conventionally, the Greek indices run from 0 to 3 (and the Latin indices from 1 to 3, i.e., over the
spatial components only). The zero-th component is the time component, and components 1 to 3
are the usual spatial components.
A
µ
will be a four-vector if and only if its components transform like the transformation of x
µ
,
the position four-vector, defined as x
µ
≡ (t, x) (remember c = 1):
x

µ
= Λ
µ
ν
x
ν
(7)
where Λ
µ
ν
is the Lorentz transformation matrix. For a boost along the x-direction, it looks like
Λ
µ
ν
=

¸
¸
¸
γ −γv 0 0
−γv γ 0 0
0 0 1 0
0 0 0 1
¸

(8)
4
That’s the problem with Coulomb’s law. It just tells you the force between two charges, and is of an action-at-
a-distance form. Lorentz force law, with v = 0 (electrostatics), contains more: the concept of a field E, the mediator
of the force, so that action is not instantaneous. But there is one nice feature of Coulomb’s law: it tells you that our
space (not space-time) must be 3-dimensional. How? Try to write down in d-dimensions Gauss’ law of constancy of
electric flux over concentric spheres enclosing a point charge q, and you will get a Coulomb-like law: the field goes as
1/r
d−1
. That it goes as inverse square tells you that d = 3.
5
with γ =

1 −v
2

−1/2
. Note that Λ
0
0
> 1 and detΛ = 1; the transformations that satisfy these two
conditions are known as proper Lorentz transformations.
A note of caution here. One should write either Λ
µ
ν
or Λ
ν
µ
to indicate clearly which one is the
first index and which one is the second. Otherwise, one may write, for an antisymmetric mixed
tensor, A
µ
ν
= −A
µ
ν
, which is confusing to say the least. However, we will hardly fall in such a trap
during this course, so can afford to be a little cavalier in the positioning of the indices.
Eq. (7) defines a contravariant four-vector; the transformation law is
A

µ
= Λ
µ
ν
A
ν
(9)
which can also be written as (∂x

µ
/∂x
ν
)A
ν
. Note that repeated indices are summed over.
The invariant interval between two nearby space-time points in the Minkowski space is given
by
ds
2
= dt
2
−dx
2
−dy
2
−dz
2
= η
µν
dx
µ
dx
ν
(10)
where η
µν
= diag(1, −1, −1, −1) is called the metric tensor or simply the metric. People who deal
with gravity and curved space-time call this flat space-time metric tensor and denote the full metric
tensor by g
µν
, but let us not go into that.
Let’s see what we get.
• There are quantities with more than one Greek index (also called Lorentz index). Quantities
with two such indices are called rank-2 tensors, η
µν
is the simplest example of that. One can
construct higher rank tensors too.
• The metric tensor can be used to lower the Lorentz index. We can define another four-vector
B
µ
= η
µν
A
ν
. This also has four components, but they are (B
0
, −B). So, the time component
is unchanged (that’s why whether you call it B
0
or B
0
is absolutely immaterial) but the
spatial components reverse sign
5
. This is called a covariant four-vector (also known as a
1-form) and transform as
B

µ
= Λ
ν
µ
B
ν
=
∂x
ν
∂x

µ
B
ν
. (11)
Note that Λ
ν
µ
is just the inverse of Λ
µ
ν
.
• Eq. (10) also tells you something about forming a Lorentz scalar from two or more Lorentz
vectors or tensors (vectors are nothing but tensors of rank 1). If in a product a Lorentz index
occurs in pair, once as a superscript (contravariant) and once as a subscript (covariant) that
particular index is to be summed over from 0 to 3. This index becomes a dummy index; the
process is called contraction. Thus, A.B = A
µ
B
µ
= A
0
B
0
− A.B is a Lorentz scalar. Note
that A
µ
B
µ
= A
µ
B
µ
. This is nothing but a dot product of two four-vectors.
• Eq. (10) can be written as dx

µ
dx

µ
= Λ
µ
ν
Λ
α
µ
dx
ν
dx
α
. If ds
2
is invariant, we must have Λ
µ
ν
Λ
α
µ
=
δ
α
ν
. (Is it correct to write Λ
µ
ν
Λ
ν
µ
= 1?)
One can also form a contravariant metric tensor η
µν
to raise the index:
A
µ
= η
µν
A
ν
. (12)
5
This statement depends on the choice of the metric tensor. Griffiths, for example, uses ηµν = diag(−1, 1, 1, 1), and
in his case, only the zero-th component reverses sign. This is generally the choice that people who work with gravity
make. The reason is that the curvature of a sphere in this metric comes out to be positive. On the other hand, the
energy-momentum relation reads p
2
= −m
2
. Since I am a particle physicist, I will use the metric diag(1, −1, −1, −1)
that keeps p
2
= m
2
, and will not bother about the sign of the curvature. For my case, timelike separations have
ds
2
> 0 and spacelike separations have ds
2
< 0, which is just the opposite of Griffiths. Anyway, before looking at a
textbook, first check the metric it uses.
6
Since we can write ds
2
= η
µν
dx
µ
dx
ν
, and as x
µ
≡ (t, −x), it follows that η
µν
is also
diag(1, −1, −1, −1). Obviously, η
µν
η
µα
= δ
ν
α
and η
µν
η
µν
= 4.
One can form tensors of rank 2 (or higher) by taking products (not contractions) of Lorentz
vectors; e.g., I can write A
µν
= C
µ
D
ν
. The transformation law is obvious:
A

µν
= Λ
µ
α
Λ
ν
β
A
αβ
. (13)
The transformation law of a covariant tensor of rank 2 is B
µν
= Λ
α
µ
Λ
β
ν
B
αβ
. There can be tensors
of mixed type, with one (or more) contravariant and one (or more) covariant index; an example is
Λ
µ
ν
. You can easily formulate the transformation laws of higher rank tensors.
A rank-2 tensor has 16 components, and is generally written as a 4 4 matrix. It can be
symmetric (A
µν
= A
νµ
), antisymmetric (A
µν
= −A
νµ
), or with no such obvious property. Metric
tensor is symmetric; we will encounter the most important antisymmetric tensor in physics shortly.
Examples of commonly used four-vector are (i) x
µ
≡ (t, x), (ii) p
µ
≡ (E, p), (iii) j
µ
≡ (ρ, j)
(charge and current densities), (iv) A
µ
≡ (φ, A) (scalar and vector potentials), etc. Use of four-
vectors can make life more elegant; for example, the Einstein mass-energy relationship is just
p
µ
p
µ
= m
2
. (14)
Another important four-vector is ∂
µ
= ∂/∂x
µ
≡ (∂/∂t, ∇) [and ∂
µ
≡ (∂/∂t, −∇)]. Note that here
the negative sign comes for the spatial part of the contravariant vector. (That’s only natural:
you expect ∂
µ
x
µ
= 4, right?) The contraction of ∂
µ
with any four-vector A
µ
is known as the
four-divergence
6
.
Q. Show that a symmetric rank-2 tensor has 10 independent components, and an antisymmetric
rank-2 tensor has 6.
Q. If A
αβ
is a symmetric tensor and B
αβ
is an antisymmetric one, show that A
αβ
B
αβ
= 0.
Q. How many components does the tensor A
αβ
µ
have? How does it transform?
Q. Show that the Lorentz transformation equations for the coordinates when the velocity v of the
moving frame is in an arbitrary direction is given by
t

= γ(t −v.x), x

= x +
γ −1
v
2
(v.x)v −γvt. (15)
Q. From eq. (15), find the general form of Λ
µ
ν
.
Q. Show that A.B is a Lorentz scalar: A

.B

= A.B.
Q. Suppose that, in a coordinate system, the metric tensor g
µν
is diag(1, −a
2
/(1 −
kr
2
), −a
2
r
2
, −a
2
r
2
sin
2
θ) where a is a function of time. If g
µν
g
µα
= δ
α
ν
, what should be the
form of g
µν
? (This metric is called the Friedmann-Robertson-Walker metric and is used to describe
the smooth expanding universe, but some texts use an overall minus sign in the metric.)
Q. Consider the equation (
αβ
= 8πGT
αβ
, where G is a constant and T
αβ
is a symmetric tensor
satisfying ∂
α
T
αβ
= 0. How many independent equations does this tensor equation represent?
Q. All components of a four-vector must have the same dimension. How should you insert factors
of c in p
µ
and j
µ
in the SI system?
Q. How should A
0
and A, the components of the four-potential scale when you go from RLH to
SI system? Show that A
µ
= (A
0
, cA) is the correct four-vector in the SI system.
2.1 The Relativistic Action and Distribution Functions
This is a digression and has almost nothing to do with the material that follows, but since we are
going to discuss relativistic electrodynamics, the discussion will be incomplete without this.
6
A contravariant vector in the denominator of a fraction is equivalent to a covariant vector in the numerator for
contraction.
7
In Newtonian mechanics, the equations of motion are obtained by minimising the action S =

Ldt. In relativistic theories the volume element d
4
x is invariant. This can be seen from the fact
that the Lorentz transformation is a ‘rotation’ in the 4-dimensional space. Another way to see this
is to check that the Jacobian is unity (do this). Anyway, I can write S =

Ld
4
x where L =

Ld
3
x,
L being the Lagrangian density. If S is invariant, so is L. This formulation is more relevant for field
theories. We will see later how to get the equations of motion (i.e., Maxwell’s equations) from the
Lagrangian density of the electromagnetic field. In quantised versions, all particles are assumed to
be excitations of their corresponding fields, so one only needs to know L there.
For classical and relativistic single-particle dynamics, the action must be only a function of s,
where ds
2
= dt
2
−dx
2
−dy
2
−dz
2
= dt
2
(1 −v
2
) and hence ds = dt

1 −v
2
. Let
S = −α

a
b
ds = −α

t
2
t
1

1 −v
2
dt, (16)
where α is some constant. In the nonrelativistic limit (v ≪1) we have
L = −α

1 −v
2
≈ −α +
1
2
αv
2
+ , (17)
so by comparing with the nonrelativistic free Lagrangian (
1
2
mv
2
) we see that α = m and hence
S = −m

ds
7
.
Now, ds
2
= dsds = dx
µ
dx
µ
, so dsδ(ds) = dx
µ
δ(dx
µ
). To determine the dynamics, we vary the
action and get
δS = −m

b
a
δ(ds) = −m

b
a
dx
µ
δ(dx
µ
)
ds
= −m

b
a
u
µ
dδx
µ
= −mu
µ
δx
µ
[
b
a
+m

b
a
δx
µ
du
µ
ds
ds, (18)
where u
µ
= dx
µ
/ds is the proper 4-velocity. If δx
µ
vanishes at the end points, we obtain
du
µ
ds
= 0, (19)
which is a generalisation of force-free motion in relativistic mechanics
8
.
If we wish to deal not with a single particle but with a large collection of them, we introduce a
distribution function in nonrelativistic mechanics (remember the Maxwell-Boltzmann distribution).
We do the same for relativistic motion. Let us consider a set of N particles, each of mass m,
described by a distribution function f(p
µ
) at any given location in space. The total number of
particles can be written as
N =

d
4
pθ(p
0
)δ(p
2
−m
2
)f(p
µ
). (20)
The delta function ensures that all particles are real and the theta function tells us that the energies
are positive. Since N, d
4
p, θ and δ(p
2
− m
2
) are all Lorentz invariant, so is f. Now we can write
the delta function as
δ(p
2
−m
2
) = δ(p
2
0
−E
2
) =
1
2E
[δ(p
0
−E) +δ(p
0
+E)] , (21)
7
There is an extra term, −m, in the Lagrangian. But the Lagrangian is never unique and the equation of motion
does not change by adding a constant term to L. However, it shifts H, the Hamiltonian, by +m: we now have the
rest energy also apart from the standard kinetic energy.
8
The trajectory is called a geodesic. In flat space-time, it is a straight line, analogous to the ordinary force-free
motion. If the space-time is curved, the metric depends on the coordinates, and the geodesic equation looks more
complicated. It can be shown (and this is the central point of the General Theory of Relativity) that a locally non-
inertial frame is equivalent to the presence of a gravitational field. Thus, gravitational effects can be taken care of
by studying the force-free trajectory in a non-inertial frame. The reason for such an equivalence, or the equivalence
between gravitational and inertial mass, is that the gravitational field equally affects all particles. One cannot make
electromagnetism purely a property of the space-time because the charged particle trajectories depend on q/m and
are different for different charged particles.
8
where E
2
= p
2
+m
2
. Because of the theta function, only the first delta function will be effective,
and the p
0
integration will just replace p
0
by E, so
N =
1
2

d
3
p
E
f(p
0
= E, p). (22)
As N and f are Lorentz invariant, so is d
3
p/E.
Finally, note that E = γm and ds/dt = γ
−1
, so Eds/dt is Lorentz invariant. Multiplying the
numerator and denominator by d
3
x, we get Ed
3
x(ds/d
4
x). The quantity in parenthesis is Lorentz
invariant, so Ed
3
x is also invariant. Taking the product, we see that the phase space element
d
3
pd
3
x is relativistically invariant, though none of the individual terms are.
Q. Prove eq. (22).
Q. Show that the Jacobian of Lorentz transformation is unity.
3 Relativistic Formulation of Maxwell’s Equations
3.1 Field Tensor and Its Dual
The electric and magnetic fields in eq. (5) can be derived from scalar and vector potentials:
B = ∇A, E = −∇φ −
∂A
∂t
. (23)
They automatically satisfy ∇.B = 0 and ∇ E = −∂B/∂t (Faraday’s law), but the other two
equations give interesting results. Ampere’s law (with Maxwell’s correction) gives

∇.A+
∂φ
∂t

+


2
∂t
2
−∇
2

A = j, (24)
while Gauss’s law gives, after the addition and subtraction of ∂
2
φ/∂t
2
,


∂t

∇.A +
∂φ
∂t

+


2
∂t
2
−∇
2

φ = ρ. (25)
If you remember that ∂
µ


∂t
, −∇

, you can easily combine eqs. (24) and (25) into

µ

µ
A
ν
−∂
ν

µ
A
µ
= j
ν
, (26)
or in an even more elegant way:

µ
F
µν
= j
ν
(27)
where
F
µν
= ∂
µ
A
ν
−∂
ν
A
µ
(28)
is the electromagnetic field tensor, a rank-2 tensor which is, by construction, antisymmetric. This
is by far the most important antisymmetric tensor in physics. If you have done the exercises in the
last section, you know that this should have six independent components. Let’s find them out and
explicitly construct F
µν
.
The six independent components of F
µν
are F
01
, F
02
, F
03
, F
12
, F
13
, and F
23
. For the first
term
F
01
= ∂
0
A
1
−∂
1
A
0
= −E
x
. (29)
Similarly, F
02
= −E
y
and F
03
= −E
z
. From B = ∇A, we have
F
12
= ∂
1
A
2
−∂
2
A
1
= −B
z
. (30)
9
Thus,
F
µν
=

¸
¸
¸
0 −E
x
−E
y
−E
z
E
x
0 −B
z
B
y
E
y
B
z
0 −B
x
E
z
−B
y
B
x
0
¸

. (31)
The covariant tensor F
µν
is obtained by
F
µν
= η
µα
η
νβ
F
αβ
=

¸
¸
¸
0 E
x
E
y
E
z
−E
x
0 −B
z
B
y
−E
y
B
z
0 −B
x
−E
z
−B
y
B
x
0
¸

, (32)
i.e., by the substitution E →−E. The reason for this sign change is that we need one η
00
and one
η
ii
for the electric field components, whose product is −1. For the magnetic field components there
are two −1s from the metric tensors, so there is no sign change.
One can construct another rank-2 antisymmetric tensor from F
µν
. Define a rank-4 completely
antisymmetric tensor ǫ
µναβ
as
ǫ
µναβ
= +1 for 0123 or even permutations,
= −1 for odd permutations of 0123,
= 0 otherwise (33)
and ǫ
µναβ
= −ǫ
µναβ
(this convention is opposite to that of Jackson, but the final results will be
identical). Then
G
µν
= −
1
2
ǫ
µναβ
F
αβ
=

¸
¸
¸
0 −B
x
−B
y
−B
z
B
x
0 E
z
−E
y
B
y
−E
z
0 E
x
B
z
E
y
−E
x
0
¸

, (34)
is called the dual tensor to F
µν
and is obtained by the substitution E →B, B →−E in eq. (31).
Do you remember that we have talked about such a symmetry of the Maxwell’s equations earlier?
So what does G
µν
have to do with electrodynamics?
It is quite straightforward to show that

µ
G
µν
= 0 (35)
leads to the second pair of Maxwell’s equations.
Q. Given eq. (31), show that eq. (27) gives two of the Maxwell’s equations.
Q. Show that eq. (35) indeed gives the other two equations of Maxwell.
Q. As F
µν
and G
µν
are both antisymmetric rank-2 tensors, one can construct three Lorentz scalars
out of them: F
µν
F
µν
, G
µν
G
µν
, and F
µν
G
µν
. Show that the first two are proportional to E
2
− B
2
while the third one is proportional to E.B. Hence these two quantities are invariant under Lorentz
transformations.
Q. Show that under the parity operation x →−x, F
µν
F
µν
is invariant but F
µν
G
µν
is not.
Q. Why one does not construct a scalar like p
µ
p
ν
F
µν
where p is some four-momentum (say that
of the electromagnetic field)?
Q. Convince yourself that ǫ
3012
= −1. What are the values of ǫ
2103
, ǫ
0321
and ǫ
2032
?
Q. Show that η
µν
ǫ
µναβ
= 0.
Q. Show that the transformation A
µ
→A
µ
+∂
µ
λ, where λ is some scalar function of x, keeps F
µν
invariant. Such transformations are generically called gauge transformations; this particular one is
called the Lorentz gauge.
10
3.2 Transformation of the Fields
How does the electric and magnetic fields transform under Lorentz transformations? It has been
beautifully demonstrated in Griffiths how, from a physical perspective, one can understand the
field transormations; in particular, if E (or B) is zero in a frame, how it can have a nonzero
value in another frame. Thus, what appears as an electric phenomena in one frame may appear
to be a magnetic phenomena in another frame (this led Einstein to his celebrated 1905 paper On
the Electrodynamics of Moving Bodies that introduced STR). Thus, unless one always deals with
nonrelativistic motions, it is better not to talk about E or B separately, but to talk about F
µν
.
We will derive the field transformation laws simply, from the transformation law of a rank-2
tensor, eq. (13):
F
01

= Λ
0
α
Λ
1
β
F
αβ
. (36)
Consider the motion to be along the common x-axis; thus v = (v, 0, 0). Only Λ
0
0
, Λ
0
1
, Λ
1
0
and Λ
1
1
are nonzero, but F
00
= F
11
= 0, and F
10
= −F
01
. This gives
−E

x
= Λ
0
0
Λ
1
1
F
01
+ Λ
0
1
Λ
1
0
F
10
=

γ
2
−γ
2
v
2

F
01
= −E
x
. (37)
Let us also work out a couple more:
F
03

= Λ
0
α
Λ
3
β
F
αβ
⇒−E

z
= Λ
0
0
Λ
3
3
F
03
+ Λ
0
1
Λ
3
3
F
13
= γ (−E
z
−vB
y
) . (38)
F
12

= Λ
1
α
Λ
2
β
F
αβ
⇒−B

z
= Λ
1
0
Λ
2
2
F
02
+ Λ
1
1
Λ
2
2
F
12
= γ (vE
y
−B
z
) . (39)
The complete set looks like
E

x
= E
x
, E

y
= γ(E
y
−vB
z
), E

z
= γ(E
z
+vB
y
),
B

x
= B
x
, B

y
= γ(B
y
+vE
z
), B

z
= γ(B
z
−vE
y
). (40)
Eq. (40) is symmetric under the exchange of E → B, B → −E. So you will get the same
transformation laws if you start from the dual tensor G
µν
.
Q. Show that for a general Lorentz transformation
E

= γ (E+v B) −
γ
2
γ + 1
v (v.E) , B

= γ (B−v E) −
γ
2
γ + 1
v (v.B) . (41)
3.3 Fields due to an Uniformly Moving Particle
Eq. (40) tells us how to get the fields in a comoving frame. Let us now look at the other side of the
coin: suppose a particle is moving uniformly in the lab frame (say along the positive x direction);
what should the fields be due to that particle?
The approach is very easy; go to the frame where the particle is at rest, calculate the fields
there, and boost back to the lab frame. What we need is the reverse transformations of eq. (40).
Suppose we are in the frame S, and the particle with charge q moves with a velocity v along the
x-axis. The detector is located at the point (0, b, 0) in the S frame (this can always be arranged;
call the nearest distance of approach b and let the y-axis pass through the detector). Let t = t

= 0
11
S
S’
x’ x
y
y’
b
vt
Figure 1: Fields in S due to a uniformly moving particle which is at rest in S

.
when the particle was at the origin of S, i.e., at the nearest distance of approach. Let S

be the
frame where the particle is at rest, so the two frames coincide at t = t

= 0. Also, let n be the unit
vector along the line joining the instantaneous position of the charge (at the origin of S

) and the
detector, and n.ˆ v = cos ψ. Thus, b = r sin ψ and vt = −r cos ψ.
At a time t in S and t

in S

, the coordinate of the detector in S

is x

1
= −vt

, x

2
= b, and
x

3
= 0. The distance is r

=

(vt

)
2
+b
2
. The electric field components in S

are
E

1
= −
qvt

4πr

3
, E

2
=
qb
4πr

3
, E

3
= 0. (42)
The magnetic field components B

1
, B

2
, B

3
are all zero.
In unprimed coordinates, the fields in S

(I have not yet implemented the field transformations)
look like
E

1
= −
qγvt
4π (b
2

2
v
2
t
2
)
3/2
, E

2
=
qb
4π (b
2

2
v
2
t
2
)
3/2
, E

3
= 0. (43)
Now we boost the fields back to the lab frame with the inverse of eq. (40), i.e., replacing v by −v:
E
1
= E

1
= −
qγvt
4π (b
2

2
v
2
t
2
)
3/2
,
E
2
= γE

2
=
γqb
4π (b
2

2
v
2
t
2
)
3/2
,
B
3
= γvE

2
=
γvqb
4π (b
2

2
v
2
t
2
)
3/2
. (44)
This we can write in a more compact way. Note that E
1
/E
2
= −vt/b, so E is always directed
along n, the unit radial vector joining the present position of the charge to the detector, just as a
static Coulomb field. Also, the denominator (b
2

2
v
2
t
2
)
3/2
can be written as r
3
γ
3
(1−v
2
sin
2
ψ)
3/2
(replace b and vt by r and ψ and use γ
2
−1 = γ
2
v
2
), so
E =
qr
4πr
3
γ
2

1 −v
2
sin
2
ψ

3/2
(45)
and
B = v E. (46)
Thus the passing charge induces a magnetic field. That’s nothing new; in fact, for low velocities
(γ ≈ 1) the expressions for E and B are identical to the ones that you get from the Coulomb
and the Biot-Savart law. More interesting is the high-velocity limit (v → 1, γ ≫ 1). At t = 0,
transverse electric and magnetic fields E
2
and B
3
show a sharp peak, of approximate height γq/b
2
,
and the peak falls off sharply as even for a small t the b
2
term in the denominator will be utterly
12
negligible. This is indistinguishable from the effect of a pulse of plane polarised radiation moving
in the x-direction. The longitudinal field E
1
varies rapidly from negative to positive, has a zero
time integral, and is actually zero at t = 0; the detector won’t feel its presence.
Let us note an important point right here. If you calculate the Poynting vector E B from
eqs. (45) and (46), you will get a nonzero result, so the field carries some energy. However, it does
not radiate. We will see later that radiation means the presence of energy at an infinite distance
(as if the energy decouples itself from the charge configuration and moves to infinity). Here if you
integrate the Poynting vector over a surface at infinity, the answer will be zero, since both the fields
fall off as 1/r
2
(and hence the Poynting vector as 1/r
4
) while the surface area of the sphere only
grows as r
2
. Thus, to get a radiation, the fields cannot fall off faster than 1/r, and you will soon
see that only if the charge is accelerated, there is a 1/r component in the field. Thus, a charge
moving with constant velocity never radiates; to get radiation one must have an accelerated charge.
There is another very elegant way to see that a charge moving with uniform velocity won’t
radiate. A static charge does not radiate; E ∝ 1/r
2
and B = 0. But a charge moving with uniform
velocity can be made static in another inertial frame moving with the same velocity. Physical laws
must be invariant in all inertial frames, so such charges do not radiate. There is no cheating here;
this is as watertight an argument (if not better) as the one given earlier, or the one that we will
see later using retarded potentials. If you know the full power of the Special Theory of Relativity,
a lot of relations can be deduced very quickly.
Q. A charge e moves along the x-axis with a constant velocity v and passes the origin at t = 0.
How do you write the form of j? [Hint: you need some Dirac δ-functions.] Show that for v ≪ 1
you indeed get the Biot-Savart law.
Q. Show that the full width at half maximum of the transverse pulse is of the order of b/γv.
Q. What should be the form of eq. (44) in the SI system?
4 Lorentz Force Equation and Its Generalisation
We know that x is the spatial component of a four-vector x
µ
. Is dx/dt the spatial component of a
four-vector? The answer is no, since both x and t change under Lorentz transformation. Rather,
we define a proper time τ which is the time that passes in a clock of the system where the particle
is at rest. Clearly, this is a unique quantity and does not undergo any Lorentz transformation.
So the hybrid quantity dx/dτ behaves like the spatial component of a four-vector
9
. This is the
four-velocity, which we will call u
µ
, and whose zero-th component is dt/dτ = γ. On the other hand,
v = dx/dt is the ordinary three-velocity. Remember that it is wrong to talk about an ordinary
four-velocity v
µ
.
Similarly, ordinary force F = dp/dt is not the component of a four-vector. Rather, K = dp/dτ
is. Obviously, K = γF and K
0
= dE/dτ is the power (call it proper power if you like).
Lorentz force is an ordinary force. Let us find the four-force K
µ
(sometimes called the Minkowski
force) which behaves as a Lorentz vector. The force should be constructed from the products of
fields and velocity components, and the only option is u
ν
F
µν
. There should be the electric charge,
so we throw q also, and try with the product qu
ν
F
µν
.
Wait. First, why don’t we start with the dual tensor G
µν
? Well, we wish to get the Lorentz force
in the nonrelativistic limit, but if we started from G
µν
, we would have ended up with q(B−vE),
something which we don’t want. Second, is the electric charge Lorentz invariant? The answer is yes;
experimentally we never saw a fast-moving electron to have any charge but −1.6 10
−19
coulomb,
and theoretically there are good reasons why it should be a Lorentz scalar (there are effects that
change the electric charge, but that has nothing to do with Lorentz invariance). More about this
9
I call it hybrid since x is measured in S and τ in S

.
13
later.
The proper velocity is given by u
ν
= (γ, −γv), and the spatial components of K
µ
can be easily
computed:
K
1
= q

u
0
F
10
+u
2
F
12
+u
3
F
13

= qγ (E+v B)
x
. (47)
Thus the Lorentz force law is restored at the nonrelativistic limit γ ≈ 1. K
0
= qγv.E gives the
rate of change of energy of the particle.
Another way to get the equation of motion in an electromagetic field, analogous to that of eq.
(19), is this. The action can always be written as S = −m

L(x
µ
, u
µ
)ds. We can expand L (which,
to maintain translational invariance, should not better be an explicit function of x
µ
) as
L(x
µ
, u
µ
) = 1 +c
0
φ(x) +c
1
A
µ
(x)u
µ
+c
2
g
µν
(x)u
µ
u
ν
+ , (48)
where the argument x of the fields is a shorthand for the four-vector x
ν
.
In practical applications, terms upto c
2
are enough. The curved space-time metric tensor g
µν
is important for gravitational interaction. The scalar term can also be included in this by adding
η
µν
φ(x) to g
µν
, as u
µ
u
µ
= 1. We are left with the c
1
term, which is relevant for electrodynamics.
Let us write the action as
S =

b
a
(−mds −qA
µ
dx
µ
) , (49)
where we have substituted q, the charge of the particle, for c
1
since ultimately we must have the
Lorentz force law. This tells us the total Lagrangian is L = −m

1 −v
2
−qA
0
+qA.v as dx/dt = v.
Varying the action with respect to the trajectory x
µ
(s), we get
δS = −

b
a

m
dx
µ
δdx
µ
ds
+qA
µ
d(δx
µ
) +qδA
µ
dx
µ

= 0. (50)
Using δA
µ
= (∂A
µ
/∂x
ν
)δx
ν
= ∂
ν
A
µ
δx
ν
, and dA
µ
= ∂
ν
A
µ
dx
ν
, and A
µ
d(δx
µ
) = d(A
µ
δx
µ
) −

ν
A
µ
dx
ν
δx
µ
, we get

b
a
¸
m
du
µ
ds
δx
µ
+q(∂
ν
A
µ
)u
ν
δx
µ
−q(∂
ν
A
µ
)u
µ
δx
ν

ds −(mu
µ
+qA
µ
) δx
µ
[
b
a
= 0. (51)
In the third term, we interchange the dummy indices µ and ν, which changes nothing. The variations
for which δx
µ
vanishes at the end points lead to the equation of motion:
m
du
µ
ds
= qF
µν
u
ν
. (52)
4.1 Motion in Combined Uniform and Static Electric and Magnetic Fields
Consider a particle moving in constant uniform electric and magnetic fields. They need not be
parallel; in fact, here we will only discuss the case where they are perpendicular. In such a combi-
nation, K
0
is not zero, so the particle receives energy from the field and its velocity always changes.
Still, we will try to make life simpler by going to a frame where one of E and B is zero. Since
E.B is a Lorentz invariant quantity, if the fields are not perpendicular, the dot product will always
retain the same value in different frames and it is impossible to make one of them zero. Moreover,
since E
2
− B
2
is also Lorentz invariant, only that field which is smaller in the starting frame can
be made zero.
Suppose [E[ < [B[. Also suppose E = (0, E, 0) and B = (0, 0, B). Let us find the components
of the fields in the frame S

where E is zero. The velocity of this frame should be in the x-direction
(this keeps E
x
and B
x
zero); actually, one needs to choose v as
v =
EB
B
2
. (53)
14
Here v = (E/B, 0, 0) and γ = B/

B
2
−E
2
. Note that [v[ < 1, consistent with the second
postulate of STR. With such a choice, the fields in S

are
E

x
= E
x
= 0, E

y
= γ(E −uB) = 0, E

z
= 0,
B

x
= B
x
= 0, B

y
= 0, B

z
= γ(B −uE) =
B
γ
. (54)
Thus, E vanishes in S

and B gets scaled down by a factor γ while retaining the same direction.
The motion in S

is trivial: a circular motion around the lines of force. The direction in which the
spiral turns depends on the charge of the particle. When we come back to the frame S, the linear
motion along the positive x-axis is superimposed on the circular motion (this is called an E B
drift), so we get a true spiralling trajectory.
The particle always moves in the positive x direction, regardless of its charge, since its charge
entered nowhere in the calculation. The only difference will be the direction of its turn, whether it
is a right-handed screw or a left-handed one.
Such crossed fields can be effectively used to get monoenergetic particle beams. Suppose we get
a number of electrons from a cathode-ray tube, with a velocity distribution. If we pass them through
such crossed static fields, only those electrons with velocity equal to E/B will go undeflected; others
will bend towards or away from the electric field. If we know what energy is needed, we can set
E and B accordingly. Suitable entrance and exit slits with momentum selectors like deflecting
magnets can produce a very pure monoenergetic beam. This is particularly important in the high
and low-energy collider experiments.
If [E[ > [B[, there is no such spiralling motion. It is easy to show that with a velocity v =
(E B)/E
2
one can make B = 0 in S

while E

will be E/γ. This is the motion of a particle in
a constant electric field; it gets continuously accelerated. In S the overall velocity of S

will be
superimposed on it.
Q. Get the components as given in eq. (54). Show that E.B and E
2
− B
2
are indeed Lorentz
invariants.
Q. If E and B are perpendicular in a frame, show that they will remain perpendicular in all co-
moving frames.
Q. If E and B make an acute angle with each other in a frame, show that the angle between them
will always remain acute.
Q. What should be the form of v (eq. (53)) in the SI system?
5 Lagrangian and Equation of Motion
We can write the Maxwell’s equations in terms of E and B, or in terms of F
µν
and its dual, or
in terms of the four-potential A
µ
= (φ, A). All formulations are completely equivalent. The first
one uses the quantities you are familiar with, the second one is elegant and manifestly Lorentz
covariant, and the third one (which is just a longhanded way of the second; F
µν
= ∂
µ
A
ν
−∂
ν
A
µ
) is
suitable to get the equations of motion. What should the equations of motion be? Well, you know
the answer: the Maxwell’s equations! But we will try to get them from a different starting point.
From our lessons in classical mechanics we know that the Lagrangian L = T − V is a function
of the generalised position q and the generalised velocity ˙ q. It can even be an explicit function of
the time t, but let us consider only systems where L = L(q, ˙ q). The action S, defined as

L dt
between the initial time t
1
and the final time t
2
, must be an extremum for an allowed path. From
this follows the variational principle
δ

t
2
t
1
L dt = 0 (55)
15
and the equation of motion through the Euler-Lagrange prescription:
d
dt

∂L
∂ ˙ q


∂L
∂q
= 0. (56)
The Hamiltonian is H(q, p) = p ˙ q −L where p = ∂L/∂ ˙ q.
What happens for a classical field which can theoretically be extended over an infinite volume?
There is a chance that the quantities that we wish to evaluate may turn out to be infinite; in fact,
the total energy of the electromagnetic field,

1
2
(E
2
+ B
2
)dv is infinite if we consider an infinite
volume with a constant electric or magnetic field. But this is not a defect of the fields; the result is
infinite just because we integrated over an infinite volume! Thus it is better to talk about a density,
e.g., the energy density c, where the total energy is

c dv. Evidently, c is finite.
We can similarly talk about a Lagrangian density L, with

L dv = L. (57)
Apart from L being finite, this has an extra advantage; the action looks better from a relativistic
point of view:
S =

L d
4
x. (58)
What should L be a function of? The generalised coordinate is the field ϕ (do not confuse it with
the scalar potential φ) which in turn depends on the coordinates x
µ
, and the spatial and temporal
derivatives of ϕ, namely, ∂
µ
ϕ (again, from a relativistic standpoint, if we introduce a temporal
derivative we should at the same time introduce the spatial derivatives too). The Euler-Lagrange
equation is slightly more complicated, but analogous to eq. (56):
d
dt

∂L
∂ ˙ ϕ

+∇.

∂L
∂(∇ϕ)


∂L
∂ϕ
= 0. (59)
This can be expressed in a more compact notation:

µ
∂L
∂(∂
µ
ϕ)

∂L
∂ϕ
= 0. (60)
This gives the equation of motion of a field.
Before we go to the dynamics of the electromagnetic field, let us get some practice with some-
thing far easier: the motion of a charged particle in an electromagnetic field.
5.1 Charged Particle in an Electromagnetic Field: the Generalised Momentum
The ordinary 3-dimensional Lagrangian must be a scalar; we don’t want it to change under some
coordinate transformation. Similarly, a relativistic Lagrangian (we will talk about the Lagrangian
density only for the fields, not for individual particles; this is nothing but classical mechanics with
the force provided by the electromagnetic fields) must be a Lorentz scalar. I don’t know what the
Lagrangian is (Lagrangians are, as a rule, never derivable from first principles, they are not even
unique; their only test is the reproduction of the equation of motion) but in the nonrelativistic
limit it should look like −eφ (I am using e as the charge of the particle, to avoid confusion with
the generalised coordinate q). But φ is the zero-th component of A
µ
, so the actual 4-dimensional
Lagrangian must involve A
µ
. What should it be contracted with? I have only two four-vectors, x
µ
and the proper velocity u
µ
; but the Lagrangian must be translationally invariant too, apart from
being Lorentz invariant, so it cannot contain x
µ
. Thus, the interaction Lagrangian must be
L
int
= −
e
γ
u
µ
A
µ
= −eφ +ev.A, (61)
16
where the factor of γ has been introduced to get the nonrelativistic Lagrangian as −eφ and not as
−eγφ; this factor also changes u to the ordinary velocity v.
Eq. (61) modifies the canonical 3-momentum P from the mechanical 3-momentum p by the
addition of a term eA (remember that P = ∂L/∂v):
P = p +eA. (62)
The total energy is the mechanical energy,

p
2
+m
2
, plus the electrostatic energy eφ:
W =

(P−eA)
2
+m
2
+eφ. (63)
This is nothing but the usual energy-momentum relation p
µ
p
µ
= m
2
with
p = (E, p) = (W −eφ, P −eA) . (64)
Without the electromagnetic field, p = P and W = E, so the effect of the field is to make the
replacement
p
µ
=⇒p
µ
−eA
µ
. (65)
Eq. (65) is the most important equation in the theory of interaction of electromagnetic field with a
charged particle. Making this minimal substitution gives the entire dynamics; of course, one has
to rewrite A
µ
in terms of E and B.
In fact, eq. (65) has a much greater impact on physics. You know that quantum mechanics
stems from classical mechanics by the identification of the operator −i∇ with the 3-momentum p
and the identification of the operator i∂/∂t with the Hamiltonian or the energy E (we use ¯h = 1);
in short, by the identification of i∂
µ
with p
µ
. For an electron in an electromagnetic field, substitute
i∂
µ
with i∂
µ
+ eA
µ
(electron’s charge is negative) and operate this operator on the electron wave
function, and you get the entire quantum electrodynamics!
Q. We have not talked about the gauge invariance of the Lagrangian, but it is imperative to check
that, particularly when the Lagrangian is not a function of E and B but of A
µ
which is not a gauge
invariant quantity. First, check that if we add a total derivative df(x, t)/dt to the Lagrangian the
equation of motion remains unchanged.
Q. Consider a gauge transformation A
µ
→A
µ
+∂
µ
λ(x). Show that eq. (61) is changed by a total
time derivative. You will require
d
dt
=

∂t
+v.∇, (66)
so better be convinced that you know this! Anyway, the Lagrangian is not invariant, but the equa-
tion of motion is.
5.2 Lagrangian for the Electromagnetic Field
As we have mentioned earlier, in this case we will consider the Lagrangian density L and not L.
The Lagrangian density must be a Lorentz scalar, because we do not want it to get transformed
under a Lorentz transformation. This follows exactly the same logic that in ordinary 3-dimensional
case, the Lagrangian must be a scalar and can never be a vector. Moreover, electromagnetism
respects parity. This means that if you notice the motion of a particle in an electromagnetic field,
and then reverse all the spatial coordinates of the system, the new motion will also be allowed.
Thus, we expect L to be invariant under parity transformation x →−x too.
These two considerations severely limit the possible options. If we consider the Lagrangian
density for a free field, j
µ
= 0, and we have no other quantities at our disposal except F
µν
and
G
µν
. Thus, there are only three possible choices: F
µν
F
µν
, G
µν
G
µν
, and F
µν
G
µν
. Among them, the
17
third term, which is proportional to E.B, is not invariant under parity, since B is an axial vector
while E is a proper vector. The first two terms are equal, but they satisfy both the criteria for
being a valid term in the Lagrangian density. Thus, L ∝ F
µν
F
µν
.
Let us start with
L = −
1
4
F
µν
F
µν
. (67)
The factor of −1/4 is a matter of convention; any multiplicative factor would have given the same
equations of motion. Using the explicit form of F
µν
, this becomes
L = −
1
2
(∂
µ
A
ν

µ
A
ν
−∂
µ
A
ν

ν
A
µ
) . (68)
We treat A
µ
as the electromagnetic field, not E or B. This has some justification: ultimately we will
quantise A
µ
and get the photon as the excitation quantum of the field. Anyway, eq. (68) involves
only derivatives of A
µ
, not A
µ
itself, so eq. (60) becomes

µ
∂L
∂(∂
µ
A
ν
)
= 0. (69)
Let us now compute ∂L/∂(∂
µ
A
ν
). The first term on the right-hand side of eq. (68) is just the
square of ∂
µ
A
ν
, so this will yield −∂
µ
A
ν
(be careful about the position of the indices). For the
second term, we note that
∂L
2
∂(∂
ρ
A
τ
)
=
1
2
δ
ρ
µ
δ
τ
ν
η
νλ
η
µκ

λ
A
κ
+
1
2

µ
A
ν
η
νλ
η
µκ
δ
ρ
λ
δ
τ
κ
=
1
2
η
ρκ
η
τλ

λ
A
κ
+
1
2

µ
A
ν
η
νρ
η
µτ
= ∂
τ
A
ρ
(70)
so that
∂L
∂(∂
µ
A
ν
)
= −∂
µ
A
ν
+∂
ν
A
µ
= −F
µν
(71)
and the free-field Euler-Lagrange equations become

µ
F
µν
= 0. (72)
This set of four equations are nothing but two of the Maxwell’s equations, Gauss’ law and 3-
component Ampere’s law, written in the absence of any external charge or current densities. Ob-
viously, if we started with
L = −
1
4
G
µν
G
µν
(73)
we would have obtained the other two equations of Maxwell.
Why, then, we write the Lagrangian density in terms of F
µν
and not its dual? Suppose we have
a nonzero external four-current density j
µ
= (ρ, j). In this case, we can write another term L, of
the form j
µ
A
µ
, and
L = −
1
4
F
µν
F
µν
−j
µ
A
µ
(74)
yields the correct equations of motion, namely

µ
F
µν
= j
ν
. (75)
There is no such magnetic analogue of j
µ
. The factor of −1/4 is needed to get eq. (75).
Before we end this subsection, let us comment on the gauge invariance of eq. (75) (gauge
invariance of electromagnetism is such a sacred principle that we want to check it at all steps).
Consider the transformation A
µ
→A
µ
+ ∂
µ
λ. F
µν
is invariant by construction and so there is no
18
problem with eq. (67). In eq. (74) the term j
µ
A
µ
is apparently not invariant, but gets an extra
contribution of j
µ

µ
λ. However,
j
µ

µ
λ = ∂
µ
(j
µ
λ) −(∂
µ
j
µ
)λ. (76)
The first term is a four-divergence and vanishes when we compute S =

Ld
4
x. By an analogue
of the 3-dimensional divergence theorem, we can reduce the four-divergence integral to

j
µ
λ dv;
but nothing goes out of the total volume, so the contribution is zero. The second term is zero
only because the electric four-current is conserved by the continuity equation: ∂
µ
j
µ
= 0. Thus,
the Lagrangian density of the electromagnetic field in the presence of an external current is gauge-
invariant only because the current is conserved! The argument can of course be turned the other
way around: the current is conserved because we demand gauge invariance. In short,
Current conservation ↔Gauge invariance. (77)
This is such an important statement that we give it a separate equation number.
Why don’t we never talk about a term like A
µ
A
µ
? This would have been a perfectly valid term
in L, but unfortunately this does not respect gauge invariance. For the electromagnetic field this
does not matter, since such a term would give rise to the mass of the field quantum, and we know
that as far as experimental accuracy goes, the photon is indeed massless, so there is no harm if
the theory cannot accomodate a mass term. But there are theories where we need to have massive
photon-like objects without breaking the gauge invariance, and a consistent formulation of such
theories is indeed subtle. Unfortunately we don’t have time and space to discuss that here.
Q. Show that A
µ
A
µ
is not gauge invariant.
Q. From eq. (74), show that the momentum conjugate to A
0
is zero.
Q. Get the momentum conjugate to A. For this, compute ∂L/∂(∂
0
A
µ
). Show that this is equal to
the electric field E.
Q. If we start from L = −
1
4
G
µν
G
µν
, will you expect the momentum conjugate to A be the magnetic
field B? Explain.
Q. If A
µ
gives rise to the photon field, we expect the degrees of freedom of A
µ
and a real photon
to be equal. Are they equal?
5.3 Energy and Momentum of the Electromagnetic Field: Poynting’s Theorem
Consider a free electromagnetic field (i.e., j
µ
= 0). Let us construct a rank-2 contravariant tensor
T
αβ
as
T
αβ
=
∂L
∂(∂
α
A
λ
)

β
A
λ
−η
αβ
L, (78)
where L is given by eq. (67). From our knowledge of classical mechanics, T
00
is something like the
Hamiltonian density of the field. Using eq. (71), we recast eq. (78) as
T
αβ
= −F
αλ
η
νβ

ν
A
λ
−η
αβ
L. (79)
T
αβ
is called the canonical stress tensor. If you remember (if not, check) that L =
1
2
(E
2
−B
2
), it
is easy to calculate T
00
:
T
00
= −F


0
A
λ

1
2

E
2
−B
2

= −E.∂
0
A−
1
2

E
2
−B
2

. (80)
However, E = −∇φ −∂
0
A, so −E.∂
0
A = E
2
+E.∇φ. Putting this,
T
00
=
1
2

E
2
+B
2

+∇.(φE), (81)
19
where we have used the free-field equation ∇.E = 0 to get the last term on the right-hand side.
Similarly,
T
0i
= −E
j

i
A
j
= E
j

i
A
j
. (82)
Again, we note that
(EB)
i
= ǫ
kij
ǫ
klm
E
j

l
A
m
= (δ
il
δ
jm
−δ
im
δ
jl
) E
j

l
A
m
= (E
j

i
A
j
−E
j

j
A
i
) , (83)
where we have been a little cavalier in the positioning of the indices, but that does not matter since
they are all ordinary three-vectors. Now we can write eq. (82) as
T
0i
= (EB)
i
+∇. (A
i
E) . (84)
Integrating eqs. (81) and (84) over the whole volume, the three-divergence terms drop out due to
the divergence theorem, and we have

T
00
d
3
x =
1
2

E
2
+B
2

d
3
x = E,

T
0i
d
3
x =

(EB) d
3
x = P
i
. (85)
These are the usual expressions of the total energy and momentum of the electromagnetic field.
From eq. (85), it is easy to formulate a differential conservation law, a four-dimensional analogue
of Poynting’s theorem:

α
T
αβ
= 0. (86)
This is easy to prove. From the definition of T
αβ
,

α
T
αβ
= ∂
α

∂L
∂(∂
α
A
λ
)

β
A
λ

−∂
β
L
=


α
∂L
∂(∂
α
A
λ
)
.∂
β
A
λ
+
∂L
∂(∂
α
A
λ
)
.∂
α

β
A
λ

−∂
β
L
=

∂L
∂A
λ

β
A
λ
+
∂L
∂(∂
α
A
λ
)

β
(∂
α
A
λ
)

−∂
β
L
= ∂
β
L(A
λ
, ∂
α
A
λ
) −∂
β
L = 0. (87)
In the intermediate steps, we have used the Euler-Lagrange equation, the chain rule of differentia-
tion, and the fact that L is only a function of A
λ
and its derivative.
However, note that T
αβ
is not symmetric. It is also neither traceless (T
αα
= 0) nor gauge
invariant. The traceless property is needed for a massless photon to emerge after quantisation; the
symmetric nature is needed to conserve angular momentum of the field, and of course we would
prefer quantities to be gauge invariant. To this end, we write
T
αβ
= −F
αµ

β
A
µ
−η
αβ


1
4
F
µν
F
µν

=

F
αµ
η
βν
F
µν
+
1
4
η
αβ
F
µν
F
µν

−F
αµ
η
βν

µ
A
ν
. (88)
The quantity in parenthesis is symmetric. The last term can be written, with source-free Maxwell’s
equation ∂
µ
F
αµ
= 0, as ∂
µ

F
αµ
A
β

. This is a four-divergence and hence gives zero under integra-
tion. So we may neglect this term and define a symmetric stress tensor Θ
αβ
as
Θ
αβ
=

F
αµ
η
βν
F
µν
+
1
4
η
αβ
F
µν
F
µν

. (89)
Q. Satisfy yourself that you understand all steps that led to eqs. (81), (84), and (87).
Q. T
αβ
is not a symmetric tensor. Calculate T
i0
and show that it is different from T
0i
.
20
Q. Show that ∂
α
Θ
αβ
= 0.
Q. Show that
Θ
00
=
1
2
(E
2
+B
2
) ,
Θ
0i
= Θ
i0
= (EB)
i
,
Θ
ij
= −
¸
E
i
E
j
+B
i
B
j

1
2
δ
ij
(E
2
+B
2
)

. (90)
Identify Θ
ij
with the negative of Maxwell’s stress tensor.
6 Potential Formulation
The four-potential A
µ
is not unique; the source-free Lagrangian density is invariant under a gauge
transformation, and even the source term is if the current is conserved (and it is conserved).
If we want a unique solution for A
µ
, we must specify the gauge, and in relativistic electrody-
namics, the Lorentz gauge, ∂
µ
A
µ
= 0, is the most convenient one to use. In this gauge, the equation
of motion is
A
µ
= j
µ
. (91)
This is called the inhomogeneous wave equation. In the absence of any source, j
µ
= 0, the equation
reduces to the homogeneous wave equation.
The question we would like to ask: what is the potential at any point due to a moving charge?
Now that the gauge is specified, this is indeed a meaningful question with unambiguous answer.
Is our counting of the number of degrees of freedom correct? Apparently, A
µ
, being a Lorentz
vector, has four components, while a real photon has two (left and right circular polarisation).
Obviously, two degrees of freedom are just spurious; you may call it an artifact of our formalism.
The gauge condition gives one constraint and hence removes one degree of freedom, but what about
the second one? Well, if you look at the Lagrangian density of the field (even when a source term
is present) you will immediately see that the momentum conjugate to A
0
is zero, since there is no

0
A
0
term there. So the scalar potential is a cyclic coordinate and can be removed by equations of
motion; it does not represent a true degree of freedom.
6.1 Retarded Potential
This is really a very simple concept: electromagnetic wave, i.e., light, takes a certain amount of
time to reach the observer from the source. Thus, what the observer sees right now (say, at t) is
some configuration that was there at some earlier time (say, at t
0
). Obviously, t
0
< t; in fact, there
is no way to know what is happening to the source of the electromagnetic wave (let us just call it
the electron) at the present moment t.
If you happen to be familiar with the concept of the light cone, you know that points which are
inside the light cone can be causally connected (i.e., there is a cause, the signal transmits, and the
effect follows). If the signal travels at the velocity of light, it is the boundary of the light cone that
gives causally connected points. If I draw the world line of the electron, only at the point where it
intersects the light cone of the observer can I have any information about it. To be more specific,
if I start my light cone from (t
0
, x
0
) and it intersects the world line of the electron at (t, x), then at
the latter point I can have the information about the electron when it was at the position (t
0
, x
0
);
this is the retarded position (and the retarded time) of the electron. Since the separation between
these two points is light-like,
R
µ
R
µ
= 0, (92)
21
where R
µ
= (t −t
0
, x −x
0
). The point of intersection is, of course, unique. It is only the retarded
potential, A
µ
(t
0
, x
0
), that we can talk about and calculate.
There is another point of intersection on the backward light cone, but that is not of any physical
relevance, since that violates causality; effect precedes cause. These are called advanced time and
advanced position of the electron. However, they are not just esoteric concepts; it may be shown
that when we quantise the field, the advanced coordinate is related with the motion of antiparticles.
Q. Draw the light-cone diagram that we have discussed just now. Convince yourself that causality
is satisfied only on the forward light cone.
Q. How should the light-cone condition, R
µ
R
µ
= 0, look like if the velocity of the signal is some
v < 1 (remember, we take c = 1)?
6.2 Digression: Lorentz Invariance of Electric Charge
Why the electric charge e is Lorentz invariant? It would have been a total mess, both theoretically
and experimentally, if it were not; but that is hardly an answer. Note that the charge density ρ is
not Lorentz invariant, since it is the zero-th component of the four-vector j
µ
. In a frame where the
configuration is static (we neglect the internal motion of the configuration), j
µ
= (ρ
0
, 0), and in a
frame which is moving with a uniform velocity v with respect to the former, j
µ
= (ρ, ρv); current
is nothing but motion of charge. But j
µ
j
µ
is Lorentz invariant, so
ρ = γρ
0
. (93)
To get the total charge I have to integrate over the volume, but the volume element dV
0
gets
contracted to dV/γ (contraction occurs only in the direction of motion), so

ρdV is a constant.
6.3 The Li´enard-Wiechert Potential
Consider an electron of charge e (here e is negative) to be at a point x
µ
1
, and the observer at x
µ
2
. At
the proper frame, where the electron is at rest, we know the solution of the inhomogeneous wave
equation:
A
µ
=

e
4πr
0
, 0

, (94)
where r
0
is the spatial distance between the charge and the observer. This is nothing but Coulomb’s
law, but the retardation condition tells us that
R
µ
= (x
2
−x
1
)
µ
= (r
0
, r
0
) . (95)
Our task is to write the solution of A
µ
in some Lorentz covariant form which in the static limit
reduces to eq. (94). Since the proper velocity u
µ
is (1, 0) in the static frame, so that u.R = r
0
, we
can write
A
µ
=
e

u
µ
u.R
, (96)
subject to the condition R
2
= 0. In a moving frame, u
µ
= (γ, γv), and R
µ
= (r, r), so
A
µ
=
e
4π(r −r.v)
(1, v) . (97)
Eq. (97) is the Li´enard-Wiechert potential.
This is the simplest way to derive the form of the Li´enard-Wiechert potential, though a clever
guess is required. There are two other ways: one is the so-called method of ‘information-collecting
sphere’, which introduces the concept of the retarded time in a roundabout way and does not
22
contain an iota of more physics insight than this (see Panofsky and Phillips, or Griffiths, for this
method). The second method is through the use of Green functions (see Jackson for a detailed
analysis). This is no doubt more rigorous, but at the same time mathematically more complicated.
Q. What should be the form of eq. (97) in the SI and the Gaussian systems?
6.4 Fields due to a Moving Charge
Suppose a charge is moving along some trajectory. What should be the electric and the magnetic
fields at a certain distance? Well, to the observer, the present position is not known, so the fields
should come out as a function of the retarded position only. We have to use the usual definitions
of E and B, namely, E = −∇A
0
− ∂A/∂t and B = ∇ A, but we have to use eq. (97) for the
potentials. The extra complication comes from the r.v term in the denominator.
There is one exception: when the charge moves with uniform velocity. Only in this case, if we
know the retarded position and the velocity of the charge, the present position is also known, and
we can express the fields in terms of its present position. But this is precisely the same game that
we have played earlier! We have seen, from eqs. (45) and (46), that the fields are indeed directed
towards the present position of the charge
10
. Let us recall the main conclusions:
• A moving charge induces a magnetic field;
• For high velocity the field is like that of a plane transverse wave;
• The field does not radiate.
Q. Explicitly show how you get eq. (45) from eq. (97). If you cannot, see Panofsky-Phillips, section
19.2.
6.5 The Fate of the Potential
I have said enough to confuse you thoroughly. First, I said that the potential smells of action-
at-a-distance and hence can be a valid concept only in a nonrelativistic theory. This is true for
both scalar and vector potentials. When we bring in the Special Theory of Relativity, we also say
that nothing can move faster than light in vacuum, so no signal of electromagnetic disturbance can
travel to a far observer instantaneously. This kills such action-at-a-distance theories. The electric
and magnetic fields are the only measurable and relevant quantities. Indeed, we can write F
µν
in
terms of E and B alone. We can calculate the radiated energy in terms of E and B. Where, then,
is the place of the potential? Why should we waste so much time talking about A
µ
? Why should
we derive the form of the Li´enard-Wiechert potential? Isn’t it a useless concept in relativistic
electrodynamics?
To answer this apparent paradox, let us first understand what potential is. It is just a function
of space and time. Its main use is that it gives the correct E and B, when a particular prescription is
followed. It is not unique; electromagnetic gauge transformation tells us that. In the nonrelativistic
case, the introduction of the potential helps us to calculate the scattering amplitudes (we have not
shown how to calculate them in classical relativistic electrodynamics, and this is not an easy job).
That’s the end of the story. What we derive in the preceeding subsections is a function of space
and time from which F
µν
, and hence E and B, may be calculated.
But if that were the only case, we would scarcely spend so much time in deriving just another
function. The fact is that potential — by now I mean A
µ
, the four-potential — comes back with
10
Be careful about the notation. In eqs. (45) and (46), r denotes the present position of the charge. In eq. (97), r
denotes the retarded position. They are obviously not the same.
23
a vengeance in quantum electrodynamics. This is a quantum field theory which describes the
interaction of electrons and photons. In quantum mechanics, ψ is a wavefunction that describes a
particle, say an electron. In quantum field theory, ψ is upgraded to the status of an operator. This
operator, acting on a state, may create an electron (so that the number of electrons is increased by
one). It can also destroy a positron, but that is a separate issue. In short, ψ is an operator — this
we will call the electron field, and this is a function of x
µ
— whose excitations are the particles.
A
µ
is the field of the photon; you quantise it and get the photons as field excitations. That’s
where its importance lies. You can still call it a four-potential if you like, but remember what
Shakespeare said about roses.
7 Accelerated Charge
In this section, to avoid confusion, we will use the following convention:
r denotes the retarded position of the moving charge, and r
0
denotes its present position if it
would have moved with uniform velocity. Since it does not, the present position can be something
completely different, but that really does not matter, since any information about the present
position is inaccessible to us. The velocity and the acceleration of the charge is v and a respectively.
In general, p denotes the magnitude of the three-vector p. The angle between r and v is θ, while
that between r
0
and v is ψ.
x’
x
θ
ψ
Retarded pos.
Present pos.
Observation pt.
v v
r
r
0
r
Figure 2: Retarded and (virtual) present positions of a particle.
We call r
0
the virtual present position of the charge, and clearly r
0
= r − rv (remember that
v < 1 and light moves with unit velocity). The Li´enard-Wiechert denominator is
s = r −r.v = r
0

1 −v
2
sin
2
ψ, (98)
(To get this, use r
2
0
= r
2
+ r
2
v
2
− 2rr.v and s
2
= r
2
+ (r.v)
2
− 2rr.v, and subtract one from the
other; also use r v = r
0
v.) The expressions for E and B for a uniformly moving charge is
given in eqs. (45) and (46). In fact, we would like to write
B = v E =
1
r
r E, (99)
since r
0
r
0
= 0 (remember that only for uniformly moving charge, E is directed along the present
position r
0
). Both the fields go as 1/r
2
and the Poynting vector vanishes over the surface at infinity;
uniformly moving charge does not radiate.
For an accelerated charge, it can be shown that (the deduction is not difficult but involves a
lot of bookkeeping, and is done in any standard textbook; my favourite is Griffiths, which does a
24
brute force job)
E =
qr
0
4πγ
2
r
3
0
(1 −v
2
sin
2
ψ)
3/2
+
q
4πr
3
0
(1 −v
2
sin
2
ψ)
3/2
¦r (r
0
a)¦ ,
B =
1
r
r E. (100)
The second term is the component that carries energy away to infinity (both E and B go as 1/r
and hence the Poynting vector is nonzero even at infinity); this is the radiation field and exists only
if a = 0. Note that these fields are transverse:
r.E
rad
= r.B
rad
= 0. (101)
Therefore,
S = EB =
1
r

E
2
r −(r.E)E

=
1
r
E
2
r. (102)
7.1 Radiation from a Slow-moving Charge
For a nonrelativistic motion, r ≈ r
0
, v →0, so
E
rad
=
q
4πr
3
r (r a) =
q
4πr
[ˆr(ˆr.a) −a] (103)
and
S =
q
2
16π
2
r
2
ˆr[a
2
−(ˆr.a)
2
] =
q
2
16π
2
r
2
ˆra
2
sin
2
θ, (104)
where θ is the angle between r and a. Thus, (i) there is no radiation along the direction of
instantaneous acceleration; (ii) the radiation pattern is symmetric for θ →π −θ, π +θ and 2π −θ,
so shaped like a doughnut (an 8-shaped pattern) around the direction of a.
To get the total power, we have to integrate S over the surface of the sphere with radius r:
P =

S.da =
q
2
a
2
16π
2

sin
2
θd(cos θ)

dφ =
q
2
a
2
16π
2
.
4
3
.2π =
1

q
2
a
2
. (105)
This is known as the Larmor formula, and is valid only for nonrelativistic motion. The power
radiated does not depend on the sign of a; so whether the charge is accelerated or decelerated, we
will get a radiation.
Q. How should eq. (105) look in the SI and the Gaussian systems? [Hint: Don’t just put back
the factor of ǫ
0
; make a dimensional analysis to see how many powers of c you need.]
7.2 Relativistic Generalisation of Larmor’s Formula
How does eq. (105) change if the motion is relativistic? The deduction is due to Li´enard, but we
will follow a shorter path, using the covariance argument. Before that, let us just work out two
derivatives:

dt
=
d
dt
(1 −v
2
)
−1/2
= γ
3
v.a = γ
3
va cos α,
d(γv)
dt
= γ
3
(v.a)v +γa = γ
3
(va cos α)v +γa, (106)
where α is the angle between v and a, so that

d(γv)
dt

2


dt

2
= γ
2
a
2

4
v
2
a
2
cos
2
α = γ
2
a
2
(1+γ
2
v
2
)−γ
4
v
2
a
2
sin
2
α = γ
4

a
2
−v
2
a
2
sin
2
α

.
(107)
25
Writing eq. (105) as
P =
q
2
6πm
2

dp
dt
.
dp
dt

, (108)
we immediately see that the relativistic generalisation is
P = −
q
2
6πm
2

dp
µ

dp
µ

, (109)
where τ = t/γ is the proper time, i.e., the time kept by the clock of the charge. Note that only
dp
µ
/dτ, and not dp
µ
/dt, is a Lorentz vector. Now

dp
µ

dp
µ

=

dp

2

dE

2
, (110)
and in the nonrelativistic limit, τ →t, and the variation of E is negligible (since that is overwhelm-
ingly controlled by the rest mass m), so we get back eq. (105). However, in the relativistic case,
E = γm, p = γmv, and with eqs. (107) and (110), the expression for P reads
P =
q
2
6πm
2
γ
2
¸

dp
dt

2

dE
dt

2
¸
=
q
2

γ
6

a
2
−[v a[
2

. (111)
This is the relativistic generalisation of eq. (105). A comparison shows that there is a huge boost
factor of γ
6
, so the emitted power is tremendously enhanced. Is this enhancement isotropic or is
there a directional bias? This question we will address soon, but before that, let me just quote a
formula for the angular distribution of radiated power, without deriving:
dP
dΩ
=
q
2
16π
2
[n ¦(n −v) a¦[
2
(1 −n.v)
5
, (112)
where n is the unit vector along r. An angular integration recovers eq. (111).
Q. What should be the form of eq. (112) in the SI system?
7.3 Relativistic Motion: v | a
Suppose that at a particular retarded time t
r
the velocity was instantaneously parallel to the
acceleration. Then r
0
a = r a, and the electric field is given by eq. (103), except that the
denominator will now have s
3
instead of r
3
. The same is true for B. But
r
s
=
1
1 −v cos θ
, (113)
so we expect the Poynting vector to have the same form as in eq. (104) but an extra factor of
(1 −v cos θ)
6
in the denominator.
Now there is a catch. There will be a further correction of s/r on the energy loss, so that the
denominator contains (1 − v cos θ)
5
. Let us try to understand the origin of this correction, which
is analogous to the well-known Doppler shift.
The energy emitted by the electron in a time dt

is located in the volume between two spheres,
one of radius r centred at x

2
and the other of radius r +dt

centred at x

1
. Consider an infinitesimal
volume element dv of this asymmetrical shell; suppose this subtends a solid angle dΩ = dS/r
2
at
x

2
. Since dr = dt

−[r.v/r]dt

,
dv = dS dr = dS

1 −
v.r
r

dt

=
s
r
dS dt

. (114)
26
dv
dr
dS
r+cdt’
r
v dt’
r.v
r
dt’
1 2
Figure 3: Location of energy radiated by an electron as it moves from x

1
to x

2
.
Therefore the energy contained in this volume within the solid angle dΩ is
1
2
(E
2
+B
2
)(s/r)dSdt

=
E
2
(s/r)dSdt

.
Thus, the power radiated per unit solid angle, dP/dΩ, is given by
dP
dΩ
=
q
2
16π
2
a
2
sin
2
θ
(1 −v cos θ)
5
. (115)
To get the total power, we have to integrate over Ω. The integration over φ gives 2π, and the
integration over θ can be performed substituting x = cos θ and using

1
−1
(1 −x
2
) dx
(1 −vx)
5
=
4
3
1
(1 −v
2
)
3
, (116)
so that
P =
q
2
a
2

4
3
1
(1 −v
2
)
3
=
q
2
a
2
γ
6

, (117)
consistent with the Li´enard formula, eq. (111).
The power is not only enhanced, it is sharply peaked in the forward direction (see fig. 4).
Though the power at precisely θ = 0 is zero, most of it is concentrated within a narrow cone. The
angle θ
max
where the radiated power is maximum can be obtained by differentiating dP/dΩ with
27
4
3
2
1
0
1
2
3
4
2 0 2 4 6 8 10 12
Figure 4: Plot of dP/dΩ for v = 0 (the 8-shaped pattern), v = 0.5 and v = 0.98. Note that the
power has been scaled down by a factor of 10
−5
for v = 0.98!
respect to θ and setting the derivative to zero (of course, you have to check the second derivative
to ensure that this is a maximum). This gives 3v cos
2
θ
max
+ 2 cos θ
max
−5v = 0, so
θ
max
= cos
−1
¸√
15v
2
+ 1 −1
3v
¸
−→cos
−1

1 −
1

2

=
1

, (118)
where the last fraction is at the limit v →1.
This is, very crudely speaking, what happens when a fast-moving electron decelerates (maybe
within some material). The emitted radiation, called bremsstrahlung, is peaked in the forward
direction. There is a frequency distribution of the spectrum; the maximum frequency of the emitted
photon is obviously the change in the kinetic energy of the highest velocity electron,
1
2
mv
2
max
/¯h.
There is no lower limit on the frequency. There can be infinite number of zero-energy photons (the
total energy is still zero!) emitted along the direction of v.
Q. Show that at the limit v → 1, θ
max
= 1/2γ. (Hint: express v in terms of γ and do a binomial
expansion in terms of 1/γ
2
.)
Q. Find θ
max
for electrons with energy 1 GeV.
Q. Assume both v and a to be along the z-direction: v = v
ˆ
k, a = a
ˆ
k. Let n be the unit vector in
the spherical polar coordinate:
n = sin θ cos φ
ˆ
i + sinθ sinφ
ˆ
j + cos θ
ˆ
k. (119)
Get eq. (115) from eq. (112).
7.4 Frequency Distribution: Bremsstrahlung for Slow Electrons
Let us now go into a bit more detail of what has been said in the last paragraph of the preceding
subsection. For simplicity, we will discuss the case v | a but v ≪1, so that the retarded position
and time are almost identical with the present position and time (s ≈ r).
Recall that in this limit, the electric field that contributes to radiation is given by eq. (103) and
the Poynting vector by eq. (104). The amount of energy c flowing out per unit time, or the power
radiated P, into a solid angle dΩ, is given by
dc
dt dΩ
=
dP
dΩ
= [S[r
2
= [E
rad
[
2
r
2
=
q
2
a
2
sin
2
θ
16π
2
. (120)
28
What is the spectral composition of the radiation? In other words, what is the amount of energy
in the frequency band ω to ω +dω? For that, we first need to know how E
rad
varies with the time
t. Let us use the notation E
t
≡ [E
rad
[(t). We define the Fourier Transform (FT) of E
t
as
E
ω
=
1


−∞
E
t
e
iωt
dt, E
t
=
1


−∞
E
ω
e
−iωt
dω. (121)
We will also need Parseval’s theorem, which says


−∞
[E
t
[
2
dt =


−∞
[E
ω
[
2
dω = 2


0
[E
ω
[
2
dω, (122)
where we use the last step to avoid dealing with negative frequencies, which does not have any
meaning.
From eq. (120), we have
dc
dΩ
=

[E
t
[
2
r
2
dt = 2r
2

[E
ω
[
2
dω (123)
so that
dc
dω dΩ
= 2r
2
[E
ω
[
2
=⇒dc = 2dω

S
[E
ω
[
2
dS, (124)
where we have used r
2
dΩ = dS. To go further you have to know the form of E
t
. But we can say
something important even qualitatively. Recall that the FT of a narrow Gaussian (width∼ σ) is a
broad Gaussian (width∼ 1/σ) — this helps us to understand, for example, the uncertainty principle
in quantum mechanics. So, if the velocity change occurs for a small time τ (which defines, in some
sense, the width of E
t
), the frequency spectrum will have a high-frequency cutoff at ω
max
∼ 1/τ.
Suppose the velocity change ∆u of the charge takes place in a very short time interval ∆t. If
radiation takes place at t
0
, E
t
during this time interval is proportional to ∆u/∆t, which we can
write as ∆uδ(t −t
0
). This is nothing but making the velocity change instantaneous, and hence the
time for the charge to interact with the retarding electric field zero. Thus,
˙ u = δ(t −t
0
)∆u =⇒


−∞
˙ udt = ∆u. (125)
(Remember that all coordinates associated with the charge are actually retarded coordinates, but
since the velocity is small, this distinction is irrelevant.)
With such a simple ansatz, we have
E
ω
=
1


q sin θ
4πr
∆ue
iωt
0
, (126)
so that, from eq. (124),
dc = 2

S
[E
ω
[
2
dSdω
= 2
1

q
2
16π
2
(∆u)
2


0

π
0
sin
3
θdθ
=
q
2

2
dω(∆u)
2
. (127)
This means that the energy spectrum is uniform: equal amount of energy is emitted at every
frequency interval, and there is no cutoff on ω. This is clearly unnatural, and occurs only because
we have taken the interaction to be of zero duration. Actually, the energy of a bremsstarhlung
photon cannot be greater than the kinetic energy of the charge, so
¯hω
max
=
1
2
mu
2
. (128)
29
The number density of emitted photons in the frequency interval dω is given by
dN
ω
=
dc
¯hω
= α
2

(∆u)
2

ω
, (129)
where α = 1/137 is the fine structure constant. This tells you that as ω goes to zero (photons
becoming softer), the number density increases, and ultimately blows up for ω = 0. Thus, it is
possible to shake off infinite number of zero-energy photons. This, obviously, does not affect the
spectrum.
Q. Show that in the SI system, the last equation in (124) will have a multiplicative factor of ǫ
0
c
on the right-hand side.
Q. Show that in eq. (127), the right-hand side is to be multiplied by 1/(ǫ
0
c
3
).
Q. Using eq. (128), draw the ω versus c plot. Redraw the same, replacing ω by the wavelength λ.
Remember that dω ∝ dλ/λ
2
.
Q. Suppose that the velocity change is not instantaneous, but occurs for a time interval t
0
− τ/2
to t
0
+τ/2, during which the velocity changes uniformly and the total change is ∆u. Show that
[E
ω
[
2

sin
2
ξ
ξ
2
, (130)
where ξ = ωτ/2. Plot the functional form of [E
ω
[
2
. Do you have any idea why this looks like a
diffraction pattern?
7.5 Relativistic Motion: v ⊥ a
Suppose v and a are instantaneously perpendicular (for a circular motion, they are always perpen-
dicular). Taking v = v
ˆ
k, a = a
ˆ
i, and eq. (119), we get
n¦(n−v)a¦ = a

¦−sin
2
θ sin
2
φ + cos θ(v −cos θ)¦
ˆ
i + sin
2
θ sin φcos φ
ˆ
j −sin θ cos φ(v −cos θ)
ˆ
k

.
(131)
Squaring, we get (after some trigonometric bookkeeping), using eq. (112),
dP
dΩ
=
q
2
a
2
16π
2
[(1 −v cos θ)
2
−(1 −v
2
) sin
2
θ cos
2
φ]
(1 −v cos θ)
5
. (132)
An angular integration yields (do this)
P =
1

q
2
a
2
γ
4
, (133)
which you can get in a straightforward way from eq. (111).
The radiation is sharply peaked in the forward direction along v, i.e., θ = 0. Near θ = 0, the
φ-dependent term is small, so the pattern is almost uniform in φ. For an electron in a circular
orbit, the radiation sweeps like a beacon. This is known as the synchrotron radiation, after the
machines where it was first observed. A charged particle revolving in a circular orbit loses energy,
and as the rate of change of momentum dp/dt = γma, we have
P =
1

q
2
m
2
γ
2

dp
dt

2
, (134)
so that the energy loss is much more for an electron than a proton, for two reasons: m is smaller,
and for the same energy, γ is larger. This is precisely why the LEP is in all probability the last
circular e
+
e

collider, though protons will soon start running in the same machine with a much
higher energy, but with significantly smaller synchrotron loss. (This machine is an engineering
marvel: a perfectly circular pipe, deep underground, with a circumference of 27 km!)
30
The same phenomenon occurs in the sky. Pulsars are rotating neutron stars whose magnetic
axis does not coincide with the rotational axis (the same is true for earth too), and therefore relative
to the fixed axis of rotation, the magnetic dipole vector is changing, and the star gives off magnetic
dipole radiation. With a typical radius of 10 km, a rotational period of 10
−3
s, and a surface
magnetic field of 10
8
Tesla, pulsars radiate away a huge amount of power in a highly directional
beacon, and when that beacon intercepts the earth (it may not, in that case we won’t be able to
see it) we see the pulsar. (This is a naive picture of a pulsar. On and near the surface of the
pulsar, such radiations get damped because of the electromagnetic plasma present there, but the
plasma in turn radiates, so we get some beacon-like pulsed radiation based on more or less the
same principle.)
Even for a circular motion, and more so for arbitrary motions, the frequency spectrum of the
synchrotron radiation contains a large number of components. It can be shown (see Jackson, sec.
14.4) that the highest frequency in the synchrotron spectrum, ω
c
, is given approximately by
ω
c
≈ ω
0
γ
3
, (135)
where ω
0
is the frequency of rotation. Note the factor of γ
3
: in a typical 10 GeV machine,
γ
max
≈ 2 10
4
, ω
0
≈ 3 10
6
s
−1
, so ω
c
≈ 2.4 10
19
s
−1
, corresponding to 16 keV X-rays.
Q. Do the necessary vector algebra and trigonometry to get eq. (132) from eq. (112).
Q. The last run of LEP had 105 GeV electrons. What was the value of γ
max
? What should be
γ
max
for 7000 GeV protons?
Q. Calculate the ratio of synchrotron losses (power radiated) for LEP-II (electrons with 105 GeV)
and LHC (protons with 7000 GeV). Given, m
p
= 1836m
e
.
7.6 Thomson Scattering
Suppose a plane monochromatic wave falls on a free particle of charge q and mass m. This will
transfer some energy to the particle, so the particle will be accelerated and emit its own char-
acteristic radiation. This radiation will be emitted in all directions (but will not be spherically
symmetric). In the nonrelativistic limit (quantum-mechanically, where the energy of the photon is
much less than the rest energy of the particle), however, the frequency of the scattered wave is the
same as the incident wave. This process is known as Thomson scattering.
In the nonrelativistic limit, eq. (112) reads
dP
dΩ
=
q
2
16π
2
[n (n a)[
2
. (136)
The acceleration is provided by the incident plane wave. If its propagation vector be k
0
, and its
polarisation vector e
in
, the electric field can be written as
E = e
in
E
0
exp(−ik
µ
x
µ
) (137)
where k = (ω, k
0
). Remember that e
in
is a vector of unit magnitude and e
in
.k
0
= 0, since the
wave is transverse. For our subsequent analysis, we will take k
0
along the lab-fixed z-axis, so that
without any loss of generality, two linearly independent choices of e
in
may be taken along the x
and the y axes.
The force equation for nonrelativistic motion is ma(t) = qE, so
a(t) = e
in
q
m
E
0
exp(−ik
µ
x
µ
). (138)
The scattered wave propagates along n (see figure), which makes an angle θ with k
0
. If its polari-
sation be e

out
11
(which can have two possible values, e
1
and e
2
), a little vector algebra shows that
11
Since the polarisation vector is in general complex, we should complex conjugate the polarisation of the outgoing
wave compared to the incoming wave.
31
(convince yourself of this!)
[n (n a)[
2
= [e

out
.a[
2
, (139)
so that
dP
dΩ
=
q
2
16π
2
[e

out
.a[
2
. (140)
The electric field, and hence a, is a rapidly oscillating function of time, and we need an average
over a complete cycle of oscillation. For nonrelativistic motion, a, though a function of time, does
not change much over the oscillation period, and so the time-averaging is essentially computing
[e
in
.e
out
[
2
and replacing the harmonic part by 1/2:
k
0
n
e
e
2
1
θ
θ
φ
φ
Figure 5: Polarisation of the scattered wave.
'
dP
dΩ
` =
q
4
16π
2
m
2
1
2
[E
0
[
2
[e
in
.e
out
[
2
. (141)
The differential scattering cross-section, dσ/dΩ, is defined as the ratio of energy radiated per unit
time (i.e., power) per unit solid angle to the incident energy flux per unit area per unit time. But
the denominator is nothing but the time-averaged Poynting vector (the time-averaging brings a
factor of 1/2), which is [E
0
[
2
/2, so

dΩ
=

q
2
4πm

2
[e
in
.e
out
[
2
. (142)
The scattering geometry is shown in fig. 5. The polarisation vector e
1
lies in the plane containing
n and k
0
, and e
2
is perpendicular to it. In terms of unit vectors of the cartesian system i, j, and
k, we may write
e
1
= cos θ cos φi + cos θ sin φj −sin θk,
e
2
= −sinφi + cos φj. (143)
In general we do not see the individual final-state polarisations, so we have to sum over all possible
polarisation states. If the initial beam is polarised along the x-axis, the angular distribution,
32
summed over final-state polarisations, is (cos
2
θ cos
2
φ + sin
2
φ), while if it is polarised along the
y-axis, the angular distribution is (cos
2
θ sin
2
φ + cos
2
φ). If the incident beam is unpolarised, we
have to average over these contributions, so

dΩunpol
=

q
2
4πm

2
1
2

1 + cos
2
θ

. (144)
This is a procedure which is true for all scattering calculations, whether classical or quantum-
mechanical: to get the scattering amplitude for unpolarised incident beam, average over all initial-
state polarisations and if you do not look for a specific polarisation in the final state, sum over all
final-state polarisations.
Eq. (144) is the Thomson formula for scattering of electromagnetic radiation by a charged
particle. It is valid only if the energy of the incident radiation is much less than the rest energy of
the particle, e.g., scattering of waves upto X-ray (∼ keV) by electrons or upto γ-ray (∼ MeV) by
protons. The distribution is symmetrical around θ = π/2 (and does not have any φ dependence),
and by integrating, we get the Thomson cross-section:
σ
T
=

3

q
2
4πm

2
. (145)
For an electron (or a proton), q
2
/4π is the fine-structure constant α (in the natural system), which
is approximately 1/137. The ratio α/m, for an electron, is approximately 2.82 10
−15
m. This is
called the classical electron radius. Of course, electron is a point particle, but a classical distribution
of charge, which total the electronic charge, must have a radius of this order if its electrostatic self-
energy is equal to the electron mass. Note that the spin of the particle does not play any role in
the scattering
12
.
Q. In the natural system of units, α = 1/137, and electron mass m = 0.511 MeV. Also, 197 MeV-fm
≈ 1 (energy and length have opposite dimension, right?). From this, show that α
2
/m ≈ 2.82 fm
= 2.82 10
−15
m.
Q. What should be the formula for the classical electron radius in the SI system?
Q. Calculate the Thomson cross-section for electrons in barn (1 barn = 10
−24
cm
−2
).
Q. Suppose the electron is a uniformly charged sphere of radius r
0
, the classical electron radius.
If its angular momentum is ¯ h/2, calculate the maximum linear velocity of a point on its surface.
What’s wrong with this value?
Q. Suppose we look at only those scattered waves which are polarised along e
1
. Show how the differ-
ential Thomson cross-section, dσ/dΩ, will vary with θ for a fixed value of φ (take φ = 0, π/4, π/2, π)
and with φ for a fixed value of θ (take θ = 0, π/4, π/2, π). Repeat the exercise with waves polarised
along e
2
.
Q. If we sum over the final-state polarisations but start with polarised incident wave, show how
dσ/dΩ will look like as functions of θ and φ for (i) beam polarised along x-axis; (ii) beam polarised
along y-axis. Take the same values of θ and φ as given in the previous problem.
7.7 Modifications to the Thomson formula: Compton, Klein-Nishina
When the energy of the incident radiation is comparable to the rest energy of the particle, mod-
ifications to eq. (144) occur. The calculation is quantum-mechanical in the sense that it uses the
concept of photons, but the result can be translated into the language of classical electrodynamics.
12
This is expected since spin does not have a role in classical physics. Quantum-mechanically, it is rather subtle to
show that the same cross-section is obtained in the nonrelativistic limit for spin-0 and spin-1/2 particles. However,
there is a difference: spin-1/2 particles have a nonzero magnetic moment which interacts with the wave and contributes
to the scattering matrix element, see the next subsection.
33
Considering the scattering of a photon of energy E = ¯ hω by a particle of rest mass m and applying
the two-body relativistic kinematics, we get the Compton formula:
E

E
=
1
1 +
E
m
(1 −cos θ)
, (146)
where E

is the energy of the scattered photon (or the wave, if you prefer), and θ is the scattering
angle. This modifies the unpolarised differential cross-section:

dΩCompton
=

dΩThomson

E

E

2
. (147)
The modification factor comes entirely from the phase-space, and dσ/dΩ drops at large θ from its
Thomson value, the drop depending on the ratio E/m. To be precise, eqs. (146) and (147) are true
only for scattering from spin-0 particles. To get the total cross-section, one must remember that
E

/E is also a function of θ. At limiting cases, we get
σ
σ
T
= 1 −2
E
m
(E ≪m) ,
3
4
m
E
(m ≪E) . (148)
For spin-1/2 particles like electrons, not only the charge but also the magnetic moment is responsible
for scattering. This modifies the matrix-element squared part of eq. (147):
[e
in
.e

out
[
2
−→[e
in
.e

out
[
2
+
(k −k

)
2
4kk

, (149)
where k and k

are the wave numbers of the incident and the scattered waves respectively. This is
known as the Klein-Nishina formula.
Q. Plot the differential scattering cross-section as a function of θ for (i) Thomson; (ii) Compton
with E/m = 0.5, and (iii) Klein-Nishina with E/m = 0.5.
Q. Show that for E ≪m, the Compton cross-section is indeed given by σ = σ
T
(1 −2E/m).
7.8 Scattering from Bound Electrons
Consider an electron bound to an atom, with a characteristic frequency ω
0
. This is, of course,
a semiclassical description, where we assume the electron to be bound in a harmonic oscillator
potential, so that without any external force, the equation of motion is
m¨ x +mω
2
0
x = 0 . (150)
If we apply a harmonic electric field E = e
in
E
0
exp(−iωt) on this electron, the equation of motion
will be
m¨ x +mω
2
0
x = qE = qe
in
E
0
exp(−iωt) . (151)
The solution is (which should be familiar to you if you have studied optical dispersion)
x = e
in
q
m

1
ω
2
0
−ω
2

exp(−iωt) , (152)
which can be checked by direct substitution in eq. (151). The acceleration, in turn, is given by
a = ¨ x = e
in
q
m

−ω
2
ω
2
0
−ω
2

exp(−iωt) . (153)
For free electrons, ω
0
→0, which is the Thomson scattering that we have just studied. For strongly
bound electrons, ω
0
≫ω, and dσ/dΩ, which is proportional to a
2
, is scaled by a factor of (ω/ω
0
)
4
.
This is known as Rayleigh scattering; we can write
σ
Rayleigh
=

ω
ω
0

4
σ
Thomson
. (154)
34
Thus, the cross-section is large for higher frequencies. This explains why the rising and setting suns
are red; the rays have to cross a thicker layer of atmosphere and bluish components are scattered
away. This also explains why the sky is blue; if we are not in the line of the sun, there is more
chance to catch the scattered blue components than the red ones, which are less scattered. The
red range of the visible spectrum is at 4100 A while the violet end is at 6500 A; thus, we expect a
(41/65)
4
≈ 16% scattering of the violet end compared to the red end.
What happens when ω
0
= ω? This is when the electrons vibrate with their resonance frequency,
the displacement is maximum, and so is the absorption. The refractive index, instead of increasing
slowly with ω, drops sharply; this is known as the region of anomalous dispersion. In the next
section, we will see how the pole at ω = ω
0
is physically removed.
7.9 Cherenkov Radiation
We have said that a uniformly moving charge does not radiate. There is one exception. Note the
factor (1 − v cos θ)
3
in the expressions of E and B. Since v < 1, this factor is always positive.
But suppose a charged particle is moving in a medium where its velocity exceeds that of light
in that medium. The denominator will now read (1 − nv cos θ)
3
, where n is the refractive index
of the medium, and can be zero if θ = cos
−1
(1/nv). So there will be, theoretically, an infinite
field intensity in the direction specified by θ: a cone about the velocity vector. Experimentally,
one gets radiation inside this cone; this is known as Cherenkov radiation. This is nothing but an
electromagnetic shock wave
13
where the surface of the cone is the shock front. Outside the cone
the field intensities are zero; on the surface they are infinity and inside the cone they are finite but
nonzero. So if a highly energetic particle enters a medium, we may get a light signal at an angle θ.
This effect was first observed by Frank and Tamm, and the theoretical explanation was provided
by Cherenkov.
The Cherenkov effect is used extensively in high-energy particle detectors. Note that if you
can successfully track the emitted light (with photon detectors like photomultiplier tubes), you can
accurately determine the velocity, and hence the energy, of the particle. For cosmic-ray and neutrino
experiments (like SuperK), where the direction of the incoming particle is extremely important to
know, Cherenkov detectors are essential.
The frequency distribution of the emitted radiation is continuous but the intensity increases
with increasing frequency. That is why most of the Cherenkov radiation is in the ultraviolet; even
the visible radiation appears blue.
Q. Two electrons, with energies 5 MeV and 10 MeV respectively, enter water, whose n = 1.33.
Calculate the Cherenkov angles.
8 Radiation Reaction
An accelerated charge radiates, so it loses energy. Thus, we expect that under a given force, its
acceleration will be smaller than a particle with same mass but no charge. This is as if the emitted
radiation is exerting a back force on the charge. This is known as radiation reaction.
So far, we have never included this effect in our calculations. For example, in the calculation
of bremsstrahlung or synchrotron radiation, we have started from the expression of E
rad
, and
calculated the Poynting vector, and then integrated the flux over a surface at infinity. We have
never considered the fact that velocity and acceleration may change due to the reactive effects of
radiation.
13
An acoustic shock wave is characterised by abrupt change of pressure and temperature; they always increase
inside the shock, and the boundary is known as the shock front.
35
Hopefully, we can justify ourselves; the experiments match with calculations. Thus, the radia-
tion reaction should be small in such cases. It is better that this is so; because, as we will see soon,
this is one case which classical electrodynamics is just not competent enough to deal with, and so
is one of the major motivations for a quantum theory of electromagnetic fields.
8.1 When is the Radiation Reaction Important?
There are, in general, two types of motion that are physically interesting. First, the steady accel-
eration, where the particle accelerates uniformly from t = 0 to t = T, and its velocity changes from
zero to aT. Second, the periodic motion, where the particle vibrates with a characteristic frequency
ω
0
and an amplitude d. Let us consider both these cases one by one.
Case 1: Steady acceleration — The radiated power is given by the Larmor formula: P =
q
2
a
2
/6π. Note that this is a non-relativistic formula; everything that we derive now is non-
relativistic (a relativistic generalisation will be considered later). Thus, the total energy loss in
time T is of the order of
E
rad
=
q
2
a
2

T . (155)
The typical energy of the system, E
0
, is of the order of
E
0
∼ mv
2
= m(aT)
2
. (156)
If E
rad
(not to be confused with the radiation electric field) is to be comparable with E
0
, we should
have
q
2
a
2

T = ma
2
T
2
=⇒ T =
q
2
6πm
. (157)
This value of T is known as the characteristic time and will henceforth be denoted by τ. Thus,
radiation reaction is significant if T ≈ τ and negligible if T ≫τ.
Let us try to have an estimate of τ for electrons. We use α = q
2
/4π = 1/137, mass of the
electron m = 0.511 MeV, and 197 MeV-fm = 1. This gives τ = 1.876 10
−13
cm = 6.25 10
−24
s.
This is indeed a very small time; typically, light takes this much time to cross an atomic nucleus.
Case 2: Periodic motion — With an amplitude d and a characteristic frequency ω
0
, the typical
energy E
0
∼ mω
2
0
d
2
, while the acceleration a ∼ ω
2
0
d and the time period T ∼ 1/ω
0
. The energy
radiated in a full cycle is (q
2
a
2
/6π)T, which should be compared with the typical energy E
0
.
Putting the relevant quantities, we get the condition for radiation reaction to be significant over a
full cycle:
ω
0
τ ∼ 1 . (158)
Thus, we can neglect the radiation reaction if τ ≪ω
−1
0
. This is true, for example, even in the most
exotic cases like fast-rotating pulsars. However, the radiation effects can build up over time, and
the pulsar should fractionally slow down over a million of years.
Thus, we conclude that the radiation reaction is important only if (i) the motion changes
appreciably over a time scale τ, or (ii) over a length scale cτ (in our system, c = 1, so both should
be of the order of τ). This justifies the neglect of this effect in our earlier calculations.
What is the energy scale associated with the time scale τ? For electron, this is just 1/τ = 105
MeV. An electron with such an energy is highly relativistic, while all the derivations that we are
going to perform are non-relativistic in nature. Thus, the radiation reaction is always a small
perturbation, if it is at all significant.
36
8.2 The Abraham-Lorentz Formula
The rate at which the charged particle loses energy is given by the Larmor formula:
F
rad
.v = −
q
2
a
2

, (159)
where the minus sign indicates that the energy is being lost. However, this equation is incorrect.
The energy is carried away to infinity by the radiation fields; but there are other components of
E B, viz., those that go as r
−4
and r
−3
, which carry energy but do not take it far away from
the configuration. When the charge undergoes a periodic motion, energy is pumped back and forth
into these components, and they exert a back force as well
14
. All is not lost, however, since if
we take a time interval at the end of which the system returns to its previous state, the net force
exerted by the r
−4
and r
−3
components vanish. Thus, we would be safe if we write

t
2
t
1
F
rad
.v dt = −
q
2

t
2
t
1
a.a dt , (160)
with the proviso that v(t
1
) = v(t
2
), a(t
1
) = a(t
2
). One can integrate the right-hand side of eq.
(160) by parts:

t
2
t
1
a.a dt =

t
2
t
1

dv
dt

.

dv
dt

dt = v.
dv
dt
[
t
2
t
1

t
2
t
1
d
2
v
dt
2
.v dt , (161)
whose first term vanishes by the periodic boundary conditions, so that

t
2
t
1
¸
F
rad

q
2

˙ a
¸
.v dt = 0 . (162)
Thus, the simplest possible form of F
rad
can be written as
F
rad
=
q
2

˙ a. (163)
This is known as the Abraham-Lorentz formula. However, note that this is not a true derivation;
we do not know, for example, the component of F
rad
perpendicular to v.
An exact derivation is a bit cumbersome. One has to assume a charge distribution; this, in
itself, is a big objection, as for all we know the electron is a point charge. Without going into the
derivation, let us just note down the main features.
• The charge must be distributed over a small but finite region, as the field due to a point charge
blows up on the charge (E → ∞ as r → 0). In his original derivation, Lorentz assumed a
spherical charge distribution. One can have other distributions; e.g., Griffiths considers a
dumbbell-shaped configuration.
• The next step is to break the charge distribution into several fragments. For an accelerated
charge, the force exerted by fragment 1 on fragment 2 is not the same (and opposite) to that
of fragment 2 exerted on fragment 1. This is a breakdown of Newton’s third law, but can be
justified from the fact that the retarded position of 1 with respect to 2 is not the same as the
retarded position of 2 with respect to 1. The imbalances, summed up, give the net radiation
reaction, or the force of the charge on itself.
14
This is why Griffiths calls Radiation reaction a misnomer; according to him it should be called Field reaction.
However, the former term sticks.
37
• The first term of the expression for the self-force is a term that is proportional to acceleration,
and thus just adds up to the ‘bare’ mass of the charge; this is the mass that we will observe
in an experiment. There is a problem; in the limit when the distribution is shrunk to a
point, this term blows up. Thus, if we consider the electron as a point particle, the mass
would have been infinite! Clearly, this is a very fundamental problem, related to the exact
nature of electron. We need a quantum theory to address this issue, but let me mention that
the infinities haunt the quantum theories as well. Of course, there we use different tricks to
tackle this problem. And we could do that very successfully, as is evident from the success of
quantum electrodynamics.
• To circumvent this problem, Lorentz assumed the bare mass of the elctron to be zero; the
mass that we observe is the result of this self-force. It can be shown that the radius needed
for the charge distribution in this case is exactly the classical electron radius. However, we
have seen that such a classical picture is untenable; the velocity of a point on the ‘equator’
of electron is far above the velocity of light!
• The second term in the expression for the self-force is the Abraham-Lorentz term, proportional
to ¨ a. Rest of the terms drop out in the limit of a point charge. So we can work with the
Abraham-Lorentz formula in the classical limit. But, all in all, this is not a success story.
8.3 Problems with the Abraham-Lorentz Formula
Now for the paradoxes. Suppose the acceleration is constant: ˙ a = 0. The charge radiates, but the
radiation reaction is zero!
The opposite extreme, is a charge that is momentarily moving with constant velocity: a = 0
but ˙ a = 0. The charge does not radiate, but there is a self-force acting on it!
Consider the motion of a charge without any external force. The equation of motion is
F
rad
=
q
2

˙ a = ma, (164)
whose solution is
a(t) = a(t = 0) exp(t/τ) , (165)
where τ is the characteristic time for the charge. Unless a(t = 0) is zero, this is a runaway solution:
the acceleration increases spontaneously with time!
One can, however, try to circumvent the runaway solution. Suppose there is an external force,
which can be written as a function of time, so that one can write
m˙ v −mτ ¨ v = F
ext
(t) . (166)
Let us try a solution
˙ v(t) = exp(t/τ)u(t) , (167)
so that
¨ v(t) =
1
τ
exp(t/τ)u(t) + exp(t/τ) ˙ u(t) , (168)
and
m˙ u(t) = −
1
τ
exp(−t/τ)F
ext
(t) . (169)
Integrating this, we get
mu(t) = −
1
τ

t
C
exp(−t

/τ)F
ext
(t

) dt

, (170)
38
where C is the integration constant. This can be written as
m˙ v(t) = −
exp(t/τ)
τ

t
C
exp(−t

/τ)F
ext
(t

) dt

. (171)
Thus, acceleration at any time t depends not only on the instantaneous value of the external force
but sort of its weighted time-average. In particular, the acceleration at t will depend on the value
of the force at some earlier time t

! If t −t

< 0, the preacceleration is significant only for a time of
the order of τ, but it is there: even before the force is applied, the particle starts accelerating! This
violates causality, but at the microscopic scale: τ is of the order of 10
−24
s. Still, this is repugnant
from a theoretical point of view. One can circumvent either the runaway solution or the acausal
preacceleration, but not both.
Is there such a preacceleration? In other words, can an electron feel a force before it is applied?
This is difficult to answer from an experimental point of view: such preacceleration is significant
only for a very small time which is not accessible even to the present-day experiments, and thus
macroscopic causality is not violated. Theoretically, this would need, in contrast to the retarded
potentials discussed earlier, an advanced potential, a signal propagating in the backward light-cone.
Definitely, we will be more comfortable with a theory which does not involve this aesthetically
repelling component.
And these problems do not go away even in a relativistic classical theory.
8.4 Relativistic Motion: Dirac Formula
A Lorentz-covariant generalisation of the equation of motion would be
m
du
µ
ds
−F
µ
rad
= F
µ
ext
, (172)
and the covariant generalisation of (q
2
/6π) ˙ a can be written as
q
2

¸
d
2
u
µ
ds
2
+Su
µ
¸
, (173)
where S is some scalar. Now, the energy radiated per unit time, which is the zero-th component
of F
µ
u
µ
divided by γ, is equal to the work done per unit time, F
rad
.v, so that energy conservation
implies
F
µ
rad
u
µ
= 0 . (174)
But u
µ
u
µ
= 1, so this gives
S = −u
µ
d
2
u
µ
ds
2
= −
d
ds

u
µ
du
µ
ds

+
du
µ
ds
du
µ
ds
. (175)
The first term, which is
1
2
d(u
µ
u
µ
)/ds, is zero, as u
µ
u
µ
= 1, so
F
µ
rad
=
q
2

¸
d
2
u
µ
ds
2
+u
µ
du
ν
ds
du
ν
ds
¸
. (176)
This is known as the Dirac formula. Note that the last term is proportional to the expression for
power in the relativistic generalisation of Larmor’s formula:
P = −
q
2

du
ν
ds
du
ν
ds
. (177)
39
8.5 Radiation Reaction on a Charged Oscillator
Consider a charged particle, typically an electron, bound in a harmonic oscillator potential with a
characteristic frequency ω
0
. Since the charge is accelerating all the time, one may wish to include
the radiation reaction too for the equation of motion. Just for simplicity, let us consider the
one-dimensional case. The equation of motion is
m¨ x +mω
2
0
x −F
rad
= 0 , (178)
which can be written as
¨ x +ω
2
0
x −τ ¨ v = 0 , (179)
as F
rad
= mτ ¨ v. Without the last term, the solution would have been x(t) = x
0
exp(iω
0
t); the
particle vibrates with a frequency ω
0
. Let the solution be of the form
x(t) = x
0
exp(−αt) (180)
where α can be a complex number. One obtains, from eq. (179),
τα
3

2

2
0
= 0 , (181)
whose solution, for τ = 0, is the well-known α = ±iω
0
. The cubic equation can be exactly solved;
let us just note that there is one real solution, the so-called runaway solution, where the acceleration
goes on increasing, and hence the solution is unphysical. The remaining two solutions are complex,
conjugate to each other, since both τ and ω
0
are real. In the limit ω
0
τ ≪ 1, we just quote the
result from Jackson, which is true upto O(ω
2
0
τ
2
):
α =
1
2
Γ ±(ω
0
+ ∆ω) , Γ = ω
2
0
τ , ∆ω = −
5
8
ω
3
0
τ
2
. (182)
Note that the solution has an exponentially falling term exp(−
1
2
Γt) and an oscillatory term. This is
reminiscent of a damping force that is proportional to ˙ x; the result will be similar whenever we have
odd number of time derivatives of x. The frequency has been shifted by ∆ω ((this is known as the
level shift), which is indeed very small, unobservable to experiments, since ∆ω/ω
0
∼ (ω
0
τ)
2
≪ 1.
The decay constant Γ causes the energy of the oscillator, which goes as x
2
, to fall with an exponential
pattern exp(−Γt). Classically, this should be the width of a spectral line.
However, this classical treatment is not applicable to electronic states in an atom since the
spectral lines are quantum mechanical in origin. One can perform a more sophisticated calculation
in the ambit of quantum field theory. The qualitative results, the existence of a nonzero level shift
and a radiative broadening, are there, but their magnitudes differ from the classical counterparts.
In particular, the level shift can be much larger, even larger than the line width (classically, the
former is always smaller than the latter by a factor of ω
0
τ). One famous example is the shift of
hydrogen 2S
1/2
level; without any radiative corrections, this should be degenerate with the 2P
1/2
level, as we expect from the Dirac theory of electrons. Experimentally, the 2S
1/2
level is about
1000 MHz lower than 2P
1/2
; this is known as the Lamb shift after Willis Lamb, who discovered it
(with his student Retherford), and is one of the cornerstones of quantum electrodynamics. The
Lamb shift is entirely due to the radiative effects.
Q. Consider the equation of motion of a charged oscillator in the presence of a damping, so that
the equation of motion is
¨ x +γ ˙ x +ω
2
0
x = 0 . (183)
Assuming a solution x(t) = x
0
exp(−αt), show that α is indeed complex. Find the decay width
and the level shift if γ ≪ω
0
.
40
8.6 Scattering and Absorption of Radiation by a Charged Oscillator
Next consider a charge in a harmonic oscillator potential with characteristic frequency ω
0
being
driven by a driving force of frequency ω. Suppose this is an electric field; for one-dimensional case,
I can write
¨ x +ω
2
0
x −τ ¨ v =
q
m
[E
0
[ exp(−iωt) . (184)
The system oscillates with frequency ω, so d
3
x/dt
3
= −ω
2
˙ x, and
¨ x +γ ˙ x +ω
2
0
x =
q
m
[E
0
[ , (185)
where γ = ω
2
τ. The solution is similar to that of eq. (152):
x =
q
m
[E
0
[

1
ω
2
0
−ω
2
−iωγ

exp(−iωt) . (186)
Since γ ≪ω, ω
0
, we can neglect the radiation reaction both in the Thomson limit (ω
0
≫ω, almost
free electron) and in the Rayleigh limit (ω ≫ ω
0
, strongly bound electron). When ω ≈ ω
0
, the
energy does not have a pole any longer; it has been flattened out to a Gaussian shape by the ωγ
term (note that in this limit, γ = ω
2
τ = ω
2
0
τ = Γ of eq. (182))
15
. This is the range where the
dispersion is anomalous and the absorption is strong, just because the oscillator vibrates with a
large amplitude, and can take up lots of energy from the source.
15
A similar term occurs in the quantum theory when we consider the amplitude for a particle to move from some
point A to some other point B. The amplitude has a pole when the particle is real, i.e., it satisfies the energy-
momentum relationship p
µ
pµ = m
2
. This pole is removed by a similar term if the particle is unstable and can decay.
This is just a reflection of the uncertainty principle: for a particle which is not stable, the energy, or the rest mass,
cannot be measured with absolute precision. The form of the amplitude with such a pole-removing term is known as
the Breit-Wigner form.
41

7.9

Cherenkov Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 35

8 Radiation Reaction 8.1 8.2 8.3 8.4 8.5 8.6

When is the Radiation Reaction Important? . . . . . . . . . . . . . . . . . . . . . . . 36 The Abraham-Lorentz Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Problems with the Abraham-Lorentz Formula . . . . . . . . . . . . . . . . . . . . . . 38 Relativistic Motion: Dirac Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Radiation Reaction on a Charged Oscillator . . . . . . . . . . . . . . . . . . . . . . . 40 Scattering and Absorption of Radiation by a Charged Oscillator . . . . . . . . . . . 41

This note is based upon the following excellent textbooks: Jackson: Classical Electrodynamics Panofsky and Phillips: Classical Electricity and Magnetism Griffiths: Introduction to Electrodynamics Raychaudhuri: The Theory of Electricity and Magnetism Feynman Lectures, vol. 2 You are always advised to read the original textbooks. Remember that the supplementary problems form an integral part of the course. I will use the rationalised Lorentz-Heaviside system throughout. That system is explained in Section 1. I will also assume, as prerequisite, that you know (i) the ordinary noncovariant form of Maxwell’s equations; (ii) the basic postulates of the Special Theory of Relativity, and (iii) how to get the Lagrangian and the Hamiltonian of a particle. In fact, for a certain section, I have to use the Lagrangian formulation for a field, but probably you have already encountered that in your Classical Mechanics course. Even if you have not, the formulation is analogous. I will always consider electromagnetic fields in vacuum, ı.e., E = D and B = H. In fact, I will never use D and H and associated quantities like the electric polarisation P or magnetisation M. A four-dimensional vector will be labeled by a Greek index; all Greek indices will run from 0 to 3, 0 being the time component. I will use the flat space-time Minkowski metric ηµν = diag(1, −1, −1, −1). All repeated indices are implicitly summed over. This note is brief, and please remember that the problems are an integral part of the course; you must do them before proceeding to the next section. I could not discuss a number of interesting topics just for want of time. For example, I would like to discuss the physics of accelerated charges in more detail, including the frequency dependence of radiated power; also the theory of halfadvanced and half-retarded potentials. There may be numerous typos. Please feel free to inform me about them. You can contact me either at my office, or at akphy@caluniv.ac.in, where I am always available.

2

1

Maxwell’s Equations

Why should we take this course? Is something exciting happening in classical electrodynamics? The answer, unfortunately, is no; it is an age-old and established subject, like classical mechanics, and almost all the fundamental discoveries were made in the 19th century. Like classical mechanics, it is a tool (well, it is classical mechanics, only the force is provided by electric or magnetic fields, and you study the motion of charged particles in these fields, or the properties of the fields themselves), which is applied in many branches of physics, most notably in astrophysics, plasma physics (either man-made or astrophysical), and particle physics, and in a lot of applied branches as well. Let me give you just an example. You know that Special Relativity comes into play when the velocities are very large, comparable to that of light, and General Relativity is relevant for dealing with gravity. In astrophysics, you encounter both these situations. There are all sorts of radiation in the sky, and a proper knowledge of them can help us unravel the mystery of the origin of the universe, or at least the galaxies. You can map the sky not only with visible light but also with invisible electromagnetic waves, from microwaves to gamma-rays. So you must know how and where these radiations are generated. Relativistic classical electrodynamics helps you to do that. But electrodynamics has been quantised, and quantum electrodynamics is known to be one of the greatest intellectual achievements of mankind. No other theory, not even gravitation (Newtonian or Einsteinian), has been tested to such precision. Why, then, we still use the classical theory? Indeed, there are places (e.g., if you wish to calculate the scattering cross-section of an electron in the field of another electron, or the well-known Compton scattering) where a quantum calculation is easier and more precise. But these systems are microscopic; when you apply your theory to a macroscopic system (like a star), you have to use some kind of averaging over the ensemble, and come back to the classical regime, thanks to the correspondence principle. Also, the classical theory deals with concepts like electric and magnetic fields, which are easier to measure. A good point to start this course is the set of equations that tells you almost everything about classical electrodynamics (the force law of Lorentz completes the set). By the word classical we mean that there are no photons, only the electromagnetic wave. Electrons are the objects with which this classical field interacts. Sometimes the electrons are treated in a quantum-mechanical way; that is called a semiclassical approximation. Of course, we have no time here to go into a full quantum-mechanical theory of the interaction of electrons with photons. In most textbooks (e.g., Panofsky and Phillips) you will see the use of the SI, sometimes called the MKS, system. In that system Maxwell’s equations (in vacuum) read ∇.E = ρ , ε0 ∂E , ∂t

∇ × B = µ0 j + µ0 ε0 ∇×E = − ∇.B = 0. ∂B , ∂t

(1)

The electric charge is expressed in coulomb; the charge of an electron is −1.6 × 10−19 coulomb. In Raychaudhuri, as well as in the second edition of Jackson, you will encounter the Gaussian system, which, I must say, is gradually going out of use 1 : ∇.E = 4πρ,
1 Even though the system has become almost archaic, I will still recommend Raychaudhuri’s book to those who are not lucky to have him as a teacher — he died in 2005 — but would like to have a glimpse of the physics insight of a teacher sans pareil.

3

∇×B =

1 ∂E 4π j+ , c c ∂t 1 ∂B , ∇×E = − c ∂t ∇.B = 0. 1 F=q E+ v×B c

(2)

The Lorentz force law in these two systems is F = q(E + v × B) (SI), (Gaussian). (3)

In this note we will use the so-called rationalised Lorentz-Heaviside (RLH) system. This system is characterised by a scaling of the fields and charge and current densities of the SI system: √ √ 1 1 ESI = √ ERLH , BSI = µ0 BRLH , qSI = ε0 qRLH , jSI = √ jRLH . ε0 µ0 (4)

This seems to be a complicated scaling, and to top it all, we set c = 1. That is not a blasphemy. It just tells you that the units of length and time are related, and when you say the length of 1 second you actually mean the length light travels in 1 second (so the length of one year is actually one light-year). It also tells you that any velocity, apart from being a dimensionless quantity, must be a number whose magnitude is less than or equal to one. Some quantities get related too: energy and mass now have same dimension, so we can safely talk about a proton having a mass of 938 MeV 2 . But what do we gain? You may not believe it, but we have abolished all factors of µ0 , ε0 , and c from our subsequent discussions. E and B now stand on the same footing, so do ρ and j. Of course, physics does not change an iota, and if you like, at the end of your calculation, you can go back to the familiar SI result with the help of eq. (4). And after rescaling, it is better to make a dimensional analysis; you may have to introduce suitable powers of c in the expression 3 To verify this claim of elegance, let us look at the Maxwell’s equations, with eq. (4) and the relation µ0 ε0 = 1/c2 : ∂E , ∂t ∂B ∇×E = − , ∂t ∇.B = 0. ∇.E = ρ,

∇×B = j+

(5)

An important point to note is that if ρ = j = 0, the equations are symmetric under the interchange E → B, B → −E. We will come back to this later. Under the parity transformation x → −x, it is clear that E → −E, B → B, j → −j. Under time reversal t → −t, E does not change sign — it cannot if it is created by a static charge configuration — but B → −B, which is intuitively easy to understand: a magnetic field is created by a current and under time reversal, the current reverses its direction, so does B. Of course, there is no free lunch, but the price to pay is small. We cannot express the electric charge in the conventional unit of coulomb. Rather, the charge is something like 0.3 unit (this comes from the fine structure constant, something you will learn later, whose experimental value is
This is the system that is used by astrophysicists and particle physicists. They also use ¯ = 1 which relates mass h with length and time: [M ] = [L]−1 = [T ]−1 . We won’t need it for a discussion of classical electrodynamics. You may go further and use kB = 1; that will relate temperature with mass. 3 Our system only abolishes µ0 , ε0 and c, but not numerical factors like 4π, so there should not be any problem with dimensional analysis. An example will be given later.
2

4

Fourvectors have one temporal and three spatial components.approximately 1/137. every information must take some time to proceed from one point to another 4 .. one can write A = 2 B × r. This consequence rules out action-at-a-distance. one space and one time) dimension can be written as t′ = t cosh θ − x sinh θ. and their products are defined in a certain way. x′ = x cosh θ − t sinh θ. But there is one nice feature of Coulomb’s law: it tells you that our space (not space-time) must be 3-dimensional. (6) (7) where Λµ is the Lorentz transformation matrix. 0 0  0 1  (8) 5 . Check eq. so that action is not instantaneous. B is derivable from a vector potential A: B = ∇ × A. How does the Lorentz force law look like? Q. Lorentz force law. With c = 1. another electron at a distance of one light-year will feel it not before one year. With c = h = 1. the position four-vector. and the theoretical expression. Show that the energy density of the electromagnetic field and the Poynting vector are respectively given by 1 E2 + B2 and E × B in the RLH system. ¯ Q. They are written as Conventionally. (5). 2 2 Four-vectors There are two postulates of the Special Theory of Relativity (STR): physical laws are invariant in all frames which are mutually inertial. Now to the more mundane subject of four-vectors. The objects which carry these informations are called fields. in the RLH system. That it goes as inverse square tells you that d = 3. and is of an action-ata-distance form. and you will get a Coulomb-like law: the field goes as 1/r d−1 . For a boost along the x-direction. find how MeV is related to s−1 and fm−1 (1 fm= 10−13 m). x) (remember c = 1): µ x′ = Λν xν µ Aµ ≡ (A0 . How is θ related with v? Q. The zero-th component is the time component.e. and with ¯ = 1. We may even quantise these fields and get the corresponding field excitations. A). and gravitation resisted all attempts (and by the smartest minds on this planet!) of quantisation. and if the sun vanishes right now. A consequence of the second postulate is that nothing can travel faster than light (in vacuum). A three-vector has three spatial components. they satisfy certain transformation laws.e. and the velocity of light in vacuum is a constant in all inertial frames. the Greek indices run from 0 to 3 (and the Latin indices from 1 to 3. There are people who believe that this number is more handy ¯ than something like 10−19 ! Q. defined as xµ ≡ (t. Show that for a uniform magnetic field 1 B. is e2 /4π h — in the SI system it is e2 /4πε0 hc). Q. it looks like ν γ  −γv Λµ =  ν  0 0  −γv γ 0 0 0 0 1 0 4 That’s the problem with Coulomb’s law. show that the Lorentz transformation equations in 1 + 1 (i.. the particles. over the spatial components only). with v = 0 (electrostatics). the earth will fly off in a tangent only after eight minutes (assuming that the gravitational information travels at the speed of light). It is just a small matter that electrodynamics was quantised long ago (and some other forces which are analogous to electrodynamics were born quantised). the mediator of the force. and components 1 to 3 are the usual spatial components. Aµ will be a four-vector if and only if its components transform like the transformation of xµ . It just tells you the force between two charges. How? Try to write down in d-dimensions Gauss’ law of constancy of electric flux over concentric spheres enclosing a point charge q. i. contains more: the concept of a field E. So if I shake an electron here.

A note of caution here. 1. which is just the opposite of Griffiths. Griffiths. 6 . 1). ηµν is the simplest example of that. I will use the metric diag(1. −1. The invariant interval between two nearby space-time points in the Minkowski space is given by ds2 = dt2 − dx2 − dy 2 − dz 2 = ηµν dxµ dxν (10) where ηµν = diag(1. Thus. If ds2 is invariant. but they are (B 0 . One should write either Λµ ν or Λν µ to indicate clearly which one is the first index and which one is the second. the transformation law is A′ = Λµ Aν ν which can also be written as (∂x′ µ /∂xν )Aν . If in a product a Lorentz index occurs in pair. 5 (12) This statement depends on the choice of the metric tensor.B = Aµ Bµ = A0 B0 − A. once as a superscript (contravariant) and once as a subscript (covariant) that particular index is to be summed over from 0 to 3. This index becomes a dummy index. (10) can be written as dx′ µ dx′ = Λµ Λα dxν dxα . we must have Λν Λα = µ ν µ µ α δν . The reason is that the curvature of a sphere in this metric comes out to be positive. first check the metric it uses. This is called a covariant four-vector (also known as a 1-form) and transform as ∂xν ′ Bµ = Λν Bν = Bν . the transformations that satisfy these two 0 conditions are known as proper Lorentz transformations. This also has four components. −1. ν µ • Eq. Let’s see what we get. (7) defines a contravariant four-vector. timelike separations have ds2 > 0 and spacelike separations have ds2 < 0.B is a Lorentz scalar. −1. Note that repeated indices are summed over. we will hardly fall in such a trap ν ν during this course. for example. and will not bother about the sign of the curvature. which is confusing to say the least. (11) µ ∂x′ µ Note that Λν is just the inverse of Λµ . the process is called contraction. 1. We can define another four-vector Bµ = ηµν Aν . µ • Eq. Quantities with two such indices are called rank-2 tensors. the energy-momentum relation reads p2 = −m2 .with γ = 1 − v 2 . So. for an antisymmetric mixed tensor. only the zero-th component reverses sign. and in his case. −1. the time component is unchanged (that’s why whether you call it B 0 or B0 is absolutely immaterial) but the spatial components reverse sign 5 . Note that Aµ Bµ = Aµ B µ . (Is it correct to write Λµ Λν = 1?) ν µ µ −1/2 (9) One can also form a contravariant metric tensor η µν to raise the index: Aµ = η µν Aν . −B). For my case. before looking at a textbook. one may write. On the other hand. Anyway. −1) is called the metric tensor or simply the metric. One can construct higher rank tensors too. (10) also tells you something about forming a Lorentz scalar from two or more Lorentz vectors or tensors (vectors are nothing but tensors of rank 1). uses ηµν = diag(−1. People who deal with gravity and curved space-time call this flat space-time metric tensor and denote the full metric tensor by gµν . Note that Λ0 > 1 and detΛ = 1. −1) that keeps p2 = m2 . • The metric tensor can be used to lower the Lorentz index. A. Since I am a particle physicist. This is nothing but a dot product of two four-vectors. so can afford to be a little cavalier in the positioning of the indices. Aµ = −Aµ . This is generally the choice that people who work with gravity make. However. • There are quantities with more than one Greek index (also called Lorentz index). Eq. Otherwise. but let us not go into that.

7 . If gµν gµα = δν . antisymmetric (Aµν = −Aνµ ). −a2 /(1 − α kr 2 ). There can be tensors µ ν of mixed type. (iv) Aµ ≡ (φ. v2 (15) Q. Q. or with no such obvious property. How many independent equations does this tensor equation represent? Q. Q. show that Aαβ Bαβ = 0. Show that the Lorentz transformation equations for the coordinates when the velocity v of the moving frame is in an arbitrary direction is given by t′ = γ(t − v. the Einstein mass-energy relationship is just p µ p µ = m2 . Note that here the negative sign comes for the spatial part of the contravariant vector. where G is a constant and T αβ is a symmetric tensor satisfying ∂α T αβ = 0.B ′ = A. j) (charge and current densities). −∇)]. Q. 2. Consider the equation G αβ = 8πGT αβ . η µν ηµα = δα and η µν ηµν = 4. How should you insert factors of c in pµ and j µ in the SI system? Q. If Aαβ is a symmetric tensor and Bαβ is an antisymmetric one. From eq. All components of a four-vector must have the same dimension. find the general form of Λµ . right?) The contraction of ∂µ with any four-vector Aµ is known as the four-divergence 6 . with one (or more) contravariant and one (or more) covariant index. Examples of commonly used four-vector are (i) xµ ≡ (t.x)v − γvt. −a2 r 2 sin2 θ) where a is a function of time. Show that A. etc.. cA) is the correct four-vector in the SI system.) Q. −1. p). but since we are going to discuss relativistic electrodynamics. One can form tensors of rank 2 (or higher) by taking products (not contractions) of Lorentz vectors. Metric tensor is symmetric. Suppose that. but some texts use an overall minus sign in the metric. (That’s only natural: you expect ∂ µ xµ = 4. Obviously. the discussion will be incomplete without this. Use of fourvectors can make life more elegant. −1. (ii) pµ ≡ (E.B.1 The Relativistic Action and Distribution Functions This is a digression and has almost nothing to do with the material that follows. x′ = x + γ−1 (v. α β (13) The transformation law of a covariant tensor of rank 2 is Bµν = Λα Λβ Bαβ . x).g. The transformation law is obvious: A′ µν = Λµ Λν Aαβ . ν A rank-2 tensor has 16 components. what should be the µν ? (This metric is called the Friedmann-Robertson-Walker metric and is used to describe form of g the smooth expanding universe. it follows that η µν is also ν diag(1. and an antisymmetric rank-2 tensor has 6. −x).Since we can write ds2 = η µν dxµ dxν . an example is Λµ . 6 A contravariant vector in the denominator of a fraction is equivalent to a covariant vector in the numerator for contraction. ∇) [and ∂ µ ≡ (∂/∂t. You can easily formulate the transformation laws of higher rank tensors. Q. (iii) j µ ≡ (ρ. (15). e. the components of the four-potential scale when you go from RLH to SI system? Show that Aµ = (A0 . in a coordinate system. I can write Aµν = C µ Dν . (14) Another important four-vector is ∂µ = ∂/∂xµ ≡ (∂/∂t. we will encounter the most important antisymmetric tensor in physics shortly. −1). Show that a symmetric rank-2 tensor has 10 independent components. and as xµ ≡ (t. the metric tensor gµν is diag(1. It can be symmetric (Aµν = Aνµ ). and is generally written as a 4 × 4 matrix. ν Q. A) (scalar and vector potentials). −a2 r 2 .x). for example. How many components does the tensor Aαβ have? How does it transform? µ Q.B is a Lorentz scalar: A′ . How should A0 and A.

(18) ds where uµ = dxµ /ds is the proper 4-velocity. The total number of particles can be written as N= d4 pθ(p0 )δ(p2 − m2 )f (pµ ). so is L.In Newtonian mechanics. Another way to see this is to check that the Jacobian is unity (do this). all particles are assumed to be excitations of their corresponding fields. Anyway. (20) (19) The delta function ensures that all particles are real and the theta function tells us that the energies are positive. In the nonrelativistic limit (v ≪ 1) we have 1 L = −α 1 − v 2 ≈ −α + αv 2 + · · · .. and the geodesic equation looks more complicated. d4 p. it is a straight line. Let a t2 S = −α b ds = −α t1 1 − v 2 dt. I can write S = Ld4 x where L = Ld3 x. This can be seen from the fact that the Lorentz transformation is a ‘rotation’ in the 4-dimensional space. it shifts H. However. In relativistic theories the volume element d4 x is invariant. 2 (17) so by comparing with the nonrelativistic free Lagrangian ( 1 mv 2 ) we see that α = m and hence 2 S = −m ds 7 . For classical and relativistic single-particle dynamics. the equations of motion are obtained by minimising the action S = Ldt. analogous to the ordinary force-free motion. Thus. √ where ds2 = dt2 − dx2 − dy 2 − dz 2 = dt2 (1 − v 2 ) and hence ds = dt 1 − v 2 . described by a distribution function f (pµ ) at any given location in space. −m. One cannot make electromagnetism purely a property of the space-time because the charged particle trajectories depend on q/m and are different for different charged particles. ds which is a generalisation of force-free motion in relativistic mechanics 8 . the Hamiltonian. we vary the action and get b b δS = −m a δ(ds) = −m a dxµ δ(dxµ ) = −m ds b a uµ dδxµ = −muµ δxµ |b + m a b a δxµ duµ ds. or the equivalence between gravitational and inertial mass. If S is invariant. Maxwell’s equations) from the Lagrangian density of the electromagnetic field. (16) where α is some constant. Now we can write the delta function as δ(p2 − m2 ) = δ(p2 − E 2 ) = 0 7 1 [δ(p0 − E) + δ(p0 + E)] . In flat space-time.e. in the Lagrangian. But the Lagrangian is never unique and the equation of motion does not change by adding a constant term to L. In quantised versions. by +m: we now have the rest energy also apart from the standard kinetic energy. We do the same for relativistic motion. We will see later how to get the equations of motion (i. the metric depends on the coordinates. The reason for such an equivalence. 8 The trajectory is called a geodesic. L being the Lagrangian density. 8 . It can be shown (and this is the central point of the General Theory of Relativity) that a locally noninertial frame is equivalent to the presence of a gravitational field. so one only needs to know L there. ds2 = dsds = dxµ dxµ . Since N . If we wish to deal not with a single particle but with a large collection of them. Now. θ and δ(p2 − m2 ) are all Lorentz invariant. If δxµ vanishes at the end points. gravitational effects can be taken care of by studying the force-free trajectory in a non-inertial frame. If the space-time is curved. each of mass m. so is f . is that the gravitational field equally affects all particles. so dsδ(ds) = dxµ δ(dxµ ). we introduce a distribution function in nonrelativistic mechanics (remember the Maxwell-Boltzmann distribution). the action must be only a function of s. 2E (21) There is an extra term. This formulation is more relevant for field theories. Let us consider a set of N particles. To determine the dynamics. we obtain duµ = 0.

1 Relativistic Formulation of Maxwell’s Equations Field Tensor and Its Dual The electric and magnetic fields in eq. though none of the individual terms are. Taking the product. Let’s find them out and explicitly construct F µν . (5) can be derived from scalar and vector potentials: B = ∇ × A. F 13 . Q. − If you remember that ∂ µ ≡ ∂ ∂φ ∇. For the first term F 01 = ∂ 0 A1 − ∂ 1 A0 = −Ex . you can easily combine eqs. (26) (27) (28) or in an even more elegant way: ∂µ F µν = j ν where F µν = ∂ µ Aν − ∂ ν Aµ is the electromagnetic field tensor. From B = ∇ × A. 9 (30) Similarly. F 02 = −Ey and F 03 = −Ez . E = −∇φ − ∂A .A + ∂φ ∂t + ∂2 − ∇2 A = j. F 02 . ∂t2 (24) while Gauss’s law gives. E (22) As N and f are Lorentz invariant. a rank-2 tensor which is. This is by far the most important antisymmetric tensor in physics. The quantity in parenthesis is Lorentz invariant. 3 3. antisymmetric. but the other two equations give interesting results. we have . Q. ∂t (23) They automatically satisfy ∇. Finally. by construction.B = 0 and ∇ × E = −∂B/∂t (Faraday’s law). F 03 . so N= 1 2 d3 p f (p0 = E. Show that the Jacobian of Lorentz transformation is unity.where E 2 = p2 + m2 . and F 23 . after the addition and subtraction of ∂ 2 φ/∂t2 . −∇ . so is d3 p/E. (22). (24) and (25) into ∂µ ∂ µ Aν − ∂ ν ∂µ Aµ = j ν . note that E = γm and ds/dt = γ −1 . (29) F 12 = ∂ 1 A2 − ∂ 2 A1 = −Bz . Ampere’s law (with Maxwell’s correction) gives ∇ ∇. Multiplying the numerator and denominator by d3 x. so Eds/dt is Lorentz invariant. ∂t2 (25) ∂ ∂t . you know that this should have six independent components. p).A + + ∂t ∂t ∂2 − ∇2 φ = ρ. only the first delta function will be effective. Because of the theta function. Prove eq. F 12 . and the p0 integration will just replace p0 by E. The six independent components of F µν are F 01 . If you have done the exercises in the last section. so Ed3 x is also invariant. we get Ed3 x(ds/d4 x). we see that the phase space element d3 pd3 x is relativistically invariant.

show that eq. Q. where λ is some scalar function of x. Do you remember that we have talked about such a symmetry of the Maxwell’s equations earlier? So what does Gµν have to do with electrodynamics? It is quite straightforward to show that ∂µ Gµν = 0 (35) leads to the second pair of Maxwell’s equations. Gµν Gµν . Show that eq. = 0 otherwise and ǫµναβ = −ǫµναβ (this convention is opposite to identical). Convince yourself that ǫ3012 = −1. Given eq. by the substitution E → −E. whose product is −1. but the final results will be −Bx 0 −Ez Ey −By Ez 0 −Ex −Bz −Ey  . (33) that of Jackson. and F µν Gµν . Hence these two quantities are invariant under Lorentz transformations. (27) gives two of the Maxwell’s equations. Q. As F µν and Gµν are both antisymmetric rank-2 tensors. −Bx  0  (32) One can construct another rank-2 antisymmetric tensor from F µν . Why one does not construct a scalar like pµ pν F µν where p is some four-momentum (say that of the electromagnetic field)? Q. −Bx  0  (31) The covariant tensor Fµν is obtained by 0  −Ex =  −Ey −Ez  Fµν = ηµα ηνβ F αβ Ex 0 Bz −By Ey −Bz 0 Bx Ez By  . Q. Q. 10 . B → −E in eq. (31). Show that the transformation Aµ → Aµ + ∂ µ λ.B. Ex  0  (34) is called the dual tensor to F µν and is obtained by the substitution E → B. For the magnetic field components there are two −1s from the metric tensors. F µν 0  Ex =  Ey Ez  −Ex 0 Bz −By −Ey −Bz 0 Bx −Ez By  .e. (35) indeed gives the other two equations of Maxwell. ǫ0321 and ǫ2032 ? Q. The reason for this sign change is that we need one η00 and one ηii for the electric field components.Thus. this particular one is called the Lorentz gauge. one can construct three Lorentz scalars out of them: F µν Fµν . so there is no sign change. F µν Fµν is invariant but F µν Gµν is not. What are the values of ǫ2103 . Such transformations are generically called gauge transformations. Show that η µν ǫµναβ = 0. Define a rank-4 completely antisymmetric tensor ǫµναβ as ǫµναβ = +1 for 0123 or even permutations. Q. Q. Show that under the parity operation x → −x. keeps F µν invariant. Then  0  Bx 1 µναβ Fαβ =  Gµν = − ǫ  By 2 Bz i.. (31). = −1 for odd permutations of 0123. Show that the first two are proportional to E2 − B2 while the third one is proportional to E.

B → −E. 0 2 1 2 The complete set looks like ′ ′ ′ Ex = Ex . and boost back to the lab frame.3.3 Fields due to an Uniformly Moving Particle Eq. unless one always deals with nonrelativistic motions. 0 1 1 0 (37) Let us also work out a couple more: F 03 ′ = Λ0 Λ3 F αβ α β (38) ′ ⇒ −Ez = Λ0 Λ3 F 03 + Λ0 Λ3 F 13 = γ (−Ez − vBy ) . 0). b. By = γ(By + vEz ). what appears as an electric phenomena in one frame may appear to be a magnetic phenomena in another frame (this led Einstein to his celebrated 1905 paper On the Electrodynamics of Moving Bodies that introduced STR). γ+1 γ+1 (41) 3. (40) tells us how to get the fields in a comoving frame. Ey = γ(Ey − vBz ). 0. calculate the fields there. and F 10 = −F 01 . (40) is symmetric under the exchange of E → B. eq. The detector is located at the point (0. Thus. (36) α β 0 Consider the motion to be along the common x-axis. So you will get the same transformation laws if you start from the dual tensor Gµν .E) . (40) Eq. Q. if E (or B) is zero in a frame. (13): ′ F 01 = Λ0 Λ1 F αβ . (40). Λ1 and Λ1 1 0 1 00 = F 11 = 0. This gives are nonzero. We will derive the field transformation laws simply.B) . What we need is the reverse transformations of eq. 0 3 1 3 F 12 ′ = Λ1 Λ2 F αβ α β (39) ′ ⇒ −Bz = Λ1 Λ2 F 02 + Λ1 Λ2 F 12 = γ (vEy − Bz ) . how it can have a nonzero value in another frame. from the transformation law of a rank-2 tensor. in particular. B′ = γ (B − v × E) − v (v. ′ ′ ′ Bx = Bx . but F ′ −Ex = Λ0 Λ1 F 01 + Λ0 Λ1 F 10 = γ 2 − γ 2 v 2 F 01 = −Ex . it is better not to talk about E or B separately. call the nearest distance of approach b and let the y-axis pass through the detector). Let t = t′ = 0 11 . and the particle with charge q moves with a velocity v along the x-axis. Let us now look at the other side of the coin: suppose a particle is moving uniformly in the lab frame (say along the positive x direction). Suppose we are in the frame S. Show that for a general Lorentz transformation E′ = γ (E + v × B) − γ2 γ2 v (v. but to talk about F µν . Ez = γ(Ez + vBy ). Only Λ0 . 0) in the S frame (this can always be arranged. Bz = γ(Bz − vEy ). one can understand the field transormations. go to the frame where the particle is at rest. thus v = (v. Thus. Λ0 .2 Transformation of the Fields How does the electric and magnetic fields transform under Lorentz transformations? It has been beautifully demonstrated in Griffiths how. what should the fields be due to that particle? The approach is very easy. from a physical perspective.

In unprimed coordinates. γ ≫ 1). the fields in S ′ (I have not yet implemented the field transformations) look like qb qγvt ′ ′ ′ . so the two frames coincide at t = t′ = 0. of approximate height γq/b2 . Thus. E3 = 0. B2 . At t = 0.. and n. at the nearest distance of approach. (46) Thus the passing charge induces a magnetic field. (44) This we can write in a more compact way. E2 = . v x′ 3 At a time t in S and t′ in S ′ . let n be the unit vector along the line joining the instantaneous position of the charge (at the origin of S ′ ) and the detector. The electric field components in S ′ are ′ E1 = − qvt′ qb ′ ′ . 4π (b2 + γ 2 v 2 t2 )3/2 . E3 = 0. Also. E2 = . B3 are all zero. when the particle was at the origin of S. the denominator (b2 + γ 2 v 2 t2 )3/2 can be written as r 3 γ 3 (1− v 2 sin2 ψ)3/2 (replace b and vt by r and ψ and use γ 2 − 1 = γ 2 v 2 ). The distance is r ′ = (vt′ )2 + b2 . so E= and B = v × E. the coordinate of the detector in S ′ is x′ = −vt′ . ′3 4πr 4πr ′ 3 (42) ′ ′ ′ The magnetic field components B1 .S y’ y b x vt S’ x’ Figure 1: Fields in S due to a uniformly moving particle which is at rest in S ′ . in fact.. Also. transverse electric and magnetic fields E2 and B3 show a sharp peak. and 2 1 = 0. so E is always directed along n. b = r sin ψ and vt = −r cos ψ. and the peak falls off sharply as even for a small t the b2 term in the denominator will be utterly 12 qr 4πr 3 γ 2 1 − v 2 sin2 ψ 3/2 (45) .e. just as a static Coulomb field. x′ = b. i. (40).ˆ = cos ψ. for low velocities (γ ≈ 1) the expressions for E and B are identical to the ones that you get from the Coulomb and the Biot-Savart law. the unit radial vector joining the present position of the charge to the detector. That’s nothing new. More interesting is the high-velocity limit (v → 1. replacing v by −v: ′ E1 = E1 = − ′ E2 = γE2 = ′ B3 = γvE2 = qγvt 4π (b2 + γ 2 v 2 t2 )3/2 γqb . 4π (b2 + γ 2 v 2 t2 )3/2 γvqb . Note that E1 /E2 = −vt/b. (43) E1 = − 3/2 2 + γ 2 v 2 t2 ) 2 + γ 2 v 2 t2 )3/2 4π (b 4π (b Now we boost the fields back to the lab frame with the inverse of eq. Let S ′ be the frame where the particle is at rest. i.e.

13 . Wait. First. and whose zero-th component is dt/dτ = γ. Obviously. This is indistinguishable from the effect of a pulse of plane polarised radiation moving in the x-direction. Let us find the four-force K µ (sometimes called the Minkowski force) which behaves as a Lorentz vector. Here if you integrate the Poynting vector over a surface at infinity. K = γF and K 0 = dE/dτ is the power (call it proper power if you like). How do you write the form of j? [Hint: you need some Dirac δ-functions. and you will soon see that only if the charge is accelerated. Show that the full width at half maximum of the transverse pulse is of the order of b/γv. Rather. and the only option is uν F µν . On the other hand. so the field carries some energy. this is as watertight an argument (if not better) as the one given earlier. K = dp/dτ is. to get a radiation. since both the fields fall off as 1/r 2 (and hence the Poynting vector as 1/r 4 ) while the surface area of the sphere only grows as r 2 . Rather. ordinary force F = dp/dt is not the component of a four-vector. Lorentz force is an ordinary force. to get radiation one must have an accelerated charge. something which we don’t want. More about this 9 I call it hybrid since x is measured in S and τ in S ′ . a lot of relations can be deduced very quickly. There is no cheating here. the fields cannot fall off faster than 1/r. and theoretically there are good reasons why it should be a Lorentz scalar (there are effects that change the electric charge. a charge moving with constant velocity never radiates. But a charge moving with uniform velocity can be made static in another inertial frame moving with the same velocity. Let us note an important point right here. which we will call uµ . we would have ended up with q(B − v × E). this is a unique quantity and does not undergo any Lorentz transformation. we define a proper time τ which is the time that passes in a clock of the system where the particle is at rest. What should be the form of eq.negligible. the answer will be zero. and is actually zero at t = 0. the detector won’t feel its presence. why don’t we start with the dual tensor Gµν ? Well. Thus. has a zero time integral. A static charge does not radiate. (44) in the SI system? 4 Lorentz Force Equation and Its Generalisation We know that x is the spatial component of a four-vector xµ . Clearly. Q. since both x and t change under Lorentz transformation.] Show that for v ≪ 1 you indeed get the Biot-Savart law. there is a 1/r component in the field. E ∝ 1/r 2 and B = 0. The longitudinal field E1 varies rapidly from negative to positive. There should be the electric charge. so such charges do not radiate. Thus. experimentally we never saw a fast-moving electron to have any charge but −1. We will see later that radiation means the presence of energy at an infinite distance (as if the energy decouples itself from the charge configuration and moves to infinity). we wish to get the Lorentz force in the nonrelativistic limit. or the one that we will see later using retarded potentials. However. Q. Remember that it is wrong to talk about an ordinary four-velocity v µ . Physical laws must be invariant in all inertial frames. This is the four-velocity. If you know the full power of the Special Theory of Relativity. you will get a nonzero result. it does not radiate. There is another very elegant way to see that a charge moving with uniform velocity won’t radiate. but if we started from Gµν . The force should be constructed from the products of fields and velocity components. Similarly. so we throw q also. (45) and (46).6 × 10−19 coulomb. Q. is the electric charge Lorentz invariant? The answer is yes. v = dx/dt is the ordinary three-velocity. If you calculate the Poynting vector E × B from eqs. but that has nothing to do with Lorentz invariance). A charge e moves along the x-axis with a constant velocity v and passes the origin at t = 0. Second. and try with the product quν F µν . So the hybrid quantity dx/dτ behaves like the spatial component of a four-vector 9 . Is dx/dt the spatial component of a four-vector? The answer is no.

we will try to make life simpler by going to a frame where one of E and B is zero. E. (47) Thus the Lorentz force law is restored at the nonrelativistic limit γ ≈ 1. The variations for which δxµ vanishes at the end points lead to the equation of motion: m duµ = qFµν uν . The proper velocity is given by uν = (γ. as uµ uµ = 1. K0 = qγv. The velocity of this frame should be in the x-direction (this keeps Ex and Bx zero). They need not be parallel. if the fields are not perpendicular. B2 14 (53) . for c1√ since ultimately we must have the Lorentz force law. Moreover. Let us find the components of the fields in the frame S ′ where E is zero. Another way to get the equation of motion in an electromagetic field. uµ )ds.v as dx/dt = v. in fact. (49) where we have substituted q. Also suppose E = (0. In practical applications. actually.B is a Lorentz invariant quantity. 0) and B = (0. so the particle receives energy from the field and its velocity always changes. The scalar term can also be included in this by adding ηµν φ(x) to gµν . should not better be an explicit function of xµ ) as L(xµ . and Aµ d(δxµ ) = d(Aµ δxµ ) − ∂ν Aµ dxν δxµ .later. the charge of the particle. one needs to choose v as v= E×B . a ds (51) In the third term. Varying the action with respect to the trajectory xµ (s). since E2 − B2 is also Lorentz invariant. and dAµ = ∂ν Aµ dxν .E gives the rate of change of energy of the particle. analogous to that of eq. This tells us the total Lagrangian is L = −m 1 − v 2 −qA0 +qA. which is relevant for electrodynamics. ds (50) Using δAµ = (∂Aµ /∂xν )δxν = ∂ν Aµ δxν . The curved space-time metric tensor gµν is important for gravitational interaction. to maintain translational invariance.1 Motion in Combined Uniform and Static Electric and Magnetic Fields Consider a particle moving in constant uniform electric and magnetic fields. and the spatial components of K µ can be easily computed: K1 = q u0 F 10 + u2 F 12 + u3 F 13 = qγ (E + v × B)x . is this. which changes nothing. K 0 is not zero. −γv). only that field which is smaller in the starting frame can be made zero. where the argument x of the fields is a shorthand for the four-vector xν . we get b m a duµ µ δx + q(∂ν Aµ )uν δxµ − q(∂ν Aµ )uµ δxν ds − (muµ + qAµ ) δxµ |b = 0. (19). we get b δS = − m a dxµ δdxµ + qAµ d(δxµ ) + qδAµ dxµ = 0. we interchange the dummy indices µ and ν. B). terms upto c2 are enough. the dot product will always retain the same value in different frames and it is impossible to make one of them zero. 0. We are left with the c1 term. ds (52) 4. Let us write the action as S= a b (48) (−mds − qAµ dxµ ) . Still. Since E. here we will only discuss the case where they are perpendicular. uµ ) = 1 + c0 φ(x) + c1 Aµ (x)uµ + c2 gµν (x)uµ uν + · · · . The action can always be written as S = −m L(xµ . Suppose |E| < |B|. In such a combination. We can expand L (which.

All formulations are completely equivalent.B and E2 − B2 are indeed Lorentz invariants. (54). Ez = 0. it gets continuously accelerated. When we come back to the frame S. others will bend towards or away from the electric field. Q. The only difference will be the direction of its turn. The particle always moves in the positive x direction. From our lessons in classical mechanics we know that the Lagrangian L = T − V is a function of the generalised position q and the generalised velocity q. Note that |v| < 1. If E and B make an acute angle with each other in a frame. Such crossed fields can be effectively used to get monoenergetic particle beams. q). By = 0. regardless of its charge. since its charge entered nowhere in the calculation. there is no such spiralling motion. The motion in S ′ is trivial: a circular motion around the lines of force. The first one uses the quantities you are familiar with. It can even be an explicit function of ˙ the time t. or in terms of the four-potential Aµ = (φ. show that the angle between them will always remain acute. Show that E. you know the answer: the Maxwell’s equations! But we will try to get them from a different starting point.√ Here v = (E/B. Q. Suppose we get a number of electrons from a cathode-ray tube. If |E| > |B|. 0) and γ = B/ B 2 − E 2 . If we pass them through such crossed static fields. This is the motion of a particle in a constant electric field. Ey = γ(E − uB) = 0. F µν = ∂ µ Aν − ∂ ν Aµ ) is suitable to get the equations of motion. The direction in which the spiral turns depends on the charge of the particle. and the third one (which is just a longhanded way of the second. consistent with the second postulate of STR. or in terms of F µν and its dual. the second one is elegant and manifestly Lorentz covariant. What should be the form of v (eq. must be an extremum for an allowed path. If E and B are perpendicular in a frame. B ′ ′ ′ Bx = Bx = 0. we can set E and B accordingly. the fields in S ′ are ′ ′ ′ Ex = Ex = 0. In S the overall velocity of S ′ will be superimposed on it. Q. (53)) in the SI system? 5 Lagrangian and Equation of Motion We can write the Maxwell’s equations in terms of E and B. From this follows the variational principle t2 δ t1 L dt = 0 15 (55) . whether it is a right-handed screw or a left-handed one. If we know what energy is needed. only those electrons with velocity equal to E/B will go undeflected. It is easy to show that with a velocity v = (E × B)/E 2 one can make B = 0 in S ′ while E′ will be E/γ. Get the components as given in eq. With such a choice. defined as L dt ˙ between the initial time t1 and the final time t2 . the linear motion along the positive x-axis is superimposed on the circular motion (this is called an E × B drift). A). E vanishes in S ′ and B gets scaled down by a factor γ while retaining the same direction. The action S. γ (54) Thus. This is particularly important in the high and low-energy collider experiments. Q. Suitable entrance and exit slits with momentum selectors like deflecting magnets can produce a very pure monoenergetic beam. so we get a true spiralling trajectory. but let us consider only systems where L = L(q. What should the equations of motion be? Well. Bz = γ(B − uE) = . with a velocity distribution. show that they will remain perpendicular in all comoving frames. 0.

I don’t know what the Lagrangian is (Lagrangians are. apart from being Lorentz invariant.1 Charged Particle in an Electromagnetic Field: the Generalised Momentum The ordinary 3-dimensional Lagrangian must be a scalar. but the Lagrangian must be translationally invariant too. But φ is the zero-th component of Aµ . with L dv = L. The Euler-Lagrange equation is slightly more complicated. the total energy of the electromagnetic field. p) = pq − L where p = ∂L/∂ q. this is nothing but classical mechanics with the force provided by the electromagnetic fields) must be a Lorentz scalar. But this is not a defect of the fields. Before we go to the dynamics of the electromagnetic field. xµ and the proper velocity uµ . (58) What should L be a function of? The generalised coordinate is the field ϕ (do not confuse it with the scalar potential φ) which in turn depends on the coordinates xµ . e. so the actual 4-dimensional Lagrangian must involve Aµ . ∂(∂µ ϕ) ∂ϕ (60) This gives the equation of motion of a field.A.g.. if we introduce a temporal derivative we should at the same time introduce the spatial derivatives too). What should it be contracted with? I have only two four-vectors. their only test is the reproduction of the equation of motion) but in the nonrelativistic limit it should look like −eφ (I am using e as the charge of the particle. the energy density E. γ 16 (61) . We can similarly talk about a Lagrangian density L. the action looks better from a relativistic point of view: S= L d4 x.and the equation of motion through the Euler-Lagrange prescription: d dt ∂L ∂q ˙ − ∂L = 0. namely. where the total energy is E dv. the interaction Lagrangian must be e Lint = − uµ Aµ = −eφ + ev. ∂q (56) The Hamiltonian is H(q. Evidently. we don’t want it to change under some coordinate transformation. and the spatial and temporal derivatives of ϕ. (57) Apart from L being finite. Thus. as a rule. ˙ ˙ What happens for a classical field which can theoretically be extended over an infinite volume? There is a chance that the quantities that we wish to evaluate may turn out to be infinite. from a relativistic standpoint. they are not even unique. ∂ µ ϕ (again. ∂L ∂(∇ϕ) − ∂L = 0. E is finite. this has an extra advantage. let us get some practice with something far easier: the motion of a charged particle in an electromagnetic field. not for individual particles. but analogous to eq. 1 (E2 + B2 )dv is infinite if we consider an infinite 2 volume with a constant electric or magnetic field. never derivable from first principles. ∂ϕ (59) This can be expressed in a more compact notation: ∂µ ∂L ∂L − = 0. (56): d dt ∂L ∂ϕ ˙ + ∇. in fact. so it cannot contain xµ . to avoid confusion with the generalised coordinate q). Similarly. 5. a relativistic Lagrangian (we will talk about the Lagrangian density only for the fields. the result is infinite just because we integrated over an infinite volume! Thus it is better to talk about a density.

by the identification of i∂ i∂ µ with i∂ µ + eAµ (electron’s charge is negative) and operate this operator on the electron wave function. (64) Without the electromagnetic field. This means that if you notice the motion of a particle in an electromagnetic field. These two considerations severely limit the possible options.where the factor of γ has been introduced to get the nonrelativistic Lagrangian as −eφ and not as −eγφ. one has to rewrite Aµ in terms of E and B. the new motion will also be allowed. Thus. Among them. Gµν Gµν . but it is imperative to check that. You will require ∂ d = + v. First. Thus.2 Lagrangian for the Electromagnetic Field As we have mentioned earlier. and you get the entire quantum electrodynamics! Q. plus the electrostatic energy eφ: (63) (62) (P − eA)2 + m2 + eφ. substitute in short. 5. Eq. the Lagrangian must be a scalar and can never be a vector. Making this minimal substitution gives the entire dynamics. (61) is changed by a total time derivative. For an electron in an electromagnetic field. This follows exactly the same logic that in ordinary 3-dimensional case. particularly when the Lagrangian is not a function of E and B but of Aµ which is not a gauge invariant quantity. The total energy is the mechanical energy. of course. We have not talked about the gauge invariance of the Lagrangian. the Lagrangian is not invariant. This is nothing but the usual energy-momentum relation pµ pµ = m2 with p = (E. t)/dt to the Lagrangian the equation of motion remains unchanged. Show that eq. You know that quantum mechanics stems from classical mechanics by the identification of the operator −i∇ with the 3-momentum p and the identification of the operator i∂/∂t with the Hamiltonian or the energy E (we use ¯ = 1). (61) modifies the canonical 3-momentum P from the mechanical 3-momentum p by the addition of a term eA (remember that P = ∂L/∂v): P = p + eA. electromagnetism respects parity. the 17 . (65) is the most important equation in the theory of interaction of electromagnetic field with a charged particle. in this case we will consider the Lagrangian density L and not L. Moreover. The Lagrangian density must be a Lorentz scalar. h µ with pµ . In fact. j µ = 0. W = p2 + m2 . Q. eq. and we have no other quantities at our disposal except F µν and Gµν . p) = (W − eφ. Consider a gauge transformation Aµ → Aµ + ∂ µ λ(x). (66) dt ∂t so better be convinced that you know this! Anyway. (65) has a much greater impact on physics. If we consider the Lagrangian density for a free field. this factor also changes u to the ordinary velocity v. (65) Eq. check that if we add a total derivative df (x. we expect L to be invariant under parity transformation x → −x too. because we do not want it to get transformed under a Lorentz transformation. and F µν Gµν . there are only three possible choices: F µν Fµν . P − eA) . p = P and W = E.∇. but the equation of motion is. so the effect of the field is to make the replacement pµ =⇒ pµ − eAµ . and then reverse all the spatial coordinates of the system.

which is proportional to E. so this will yield −∂ µ Aν (be careful about the position of the indices). j). not Aµ itself. F µν is invariant by construction and so there is no 18 (75) . (67) 4 The factor of −1/4 is a matter of convention. In this case. not E or B.B. this becomes 1 L = − (∂ µ Aν ∂µ Aν − ∂ µ Aν ∂ν Aµ ) . L ∝ F µν Fµν . any multiplicative factor would have given the same equations of motion. Using the explicit form of F µν . (75). This has some justification: ultimately we will quantise Aµ and get the photon as the excitation quantum of the field. then. Thus. eq. 1 L = − F µν Fµν . since B is an axial vector while E is a proper vector. if we started with 1 (73) L = − Gµν Gµν 4 we would have obtained the other two equations of Maxwell. For the second term. namely ∂µ F µν = j ν . (68) involves only derivatives of Aµ . Anyway. Consider the transformation Aµ → Aµ + ∂ µ λ. Gauss’ law and 3component Ampere’s law. (60) becomes ∂µ ∂L = 0. ∂(∂µ Aν ) (69) Let us now compute ∂L/∂(∂µ Aν ). Before we end this subsection. The first term on the right-hand side of eq. we can write another term L. of the form j µ Aµ . let us comment on the gauge invariance of eq. (68) is just the square of ∂ µ Aν . and 1 (74) L = − F µν Fµν − j µ Aµ 4 yields the correct equations of motion. The factor of −1/4 is needed to get eq. so eq.third term. 2 (68) Let us start with We treat Aµ as the electromagnetic field. but they satisfy both the criteria for being a valid term in the Lagrangian density. (75) (gauge invariance of electromagnetism is such a sacred principle that we want to check it at all steps). Why. we write the Lagrangian density in terms of F µν and not its dual? Suppose we have a nonzero external four-current density j µ = (ρ. The first two terms are equal. There is no such magnetic analogue of j µ . written in the absence of any external charge or current densities. Obviously. (72) This set of four equations are nothing but two of the Maxwell’s equations. we note that ∂L2 ∂(∂ρ Aτ ) = = so that 1 ρ τ νλ µκ 1 ρ τ δµ δν η η ∂λ Aκ + ∂µ Aν η νλ η µκ δλ δκ 2 2 1 ρκ τ λ 1 η η ∂λ Aκ + ∂µ Aν η νρ η µτ = ∂ τ Aρ 2 2 (70) ∂L = −∂ µ Aν + ∂ ν Aµ = −F µν ∂(∂µ Aν ) (71) and the free-field Euler-Lagrange equations become ∂µ F µν = 0. is not invariant under parity.

E = −∇φ − ∂0 A.problem with eq. Unfortunately we don’t have time and space to discuss that here. From our knowledge of classical mechanics. From eq. we recast eq. In short. show that the momentum conjugate to A0 is zero. will you expect the momentum conjugate to A be the magnetic 4 field B? Explain.3 Energy and Momentum of the Electromagnetic Field: Poynting’s Theorem Consider a free electromagnetic field (i. (74). j µ = 0). (79) T αβ is called the canonical stress tensor. but gets an extra contribution of j µ ∂µ λ. check) that L = 1 (E2 − B2 ). If we start from L = − 1 Gµν Gµν . Show that Aµ Aµ is not gauge invariant. 2 2 (80) However.(φE). Show that this is equal to the electric field E. T 00 is something like the Hamiltonian density of the field.. Q. so there is no harm if the theory cannot accomodate a mass term. (67). Putting this.∂0 A − E2 − B2 . If you remember (if not. For the electromagnetic field this does not matter. (74) the term j µ Aµ is apparently not invariant. (67). so −E. 2 19 (81) . If Aµ gives rise to the photon field. By an analogue of the 3-dimensional divergence theorem.∂0 A = E2 + E. it 2 is easy to calculate T 00 : T 00 = −F 0λ ∂0 Aλ − 1 1 E2 − B2 = −E. compute ∂L/∂(∂ 0 Aµ ). In eq. Current conservation ↔ Gauge invariance. Let us construct a rank-2 contravariant tensor T αβ as ∂L ∂ β Aλ − η αβ L.∇φ. and a consistent formulation of such theories is indeed subtle. so the contribution is zero. Using eq. Get the momentum conjugate to A. Q. and we know that as far as experimental accuracy goes. (78) as T αβ = −F αλ η νβ ∂ν Aλ − η αβ L. we expect the degrees of freedom of Aµ and a real photon to be equal. T 00 = 1 E2 + B2 + ∇. For this. Q. but unfortunately this does not respect gauge invariance. Q. the Lagrangian density of the electromagnetic field in the presence of an external current is gaugeinvariant only because the current is conserved! The argument can of course be turned the other way around: the current is conserved because we demand gauge invariance. The second term is zero only because the electric four-current is conserved by the continuity equation: ∂µ j µ = 0. Thus. But there are theories where we need to have massive photon-like objects without breaking the gauge invariance. the photon is indeed massless. Q. However. since such a term would give rise to the mass of the field quantum. (76) The first term is a four-divergence and vanishes when we compute S = Ld4 x. This is such an important statement that we give it a separate equation number. but nothing goes out of the total volume. (71). we can reduce the four-divergence integral to j µ λ dv. Are they equal? (77) 5. j µ ∂µ λ = ∂µ (j µ λ) − (∂µ j µ )λ.e. (78) T αβ = ∂(∂α Aλ ) where L is given by eq. Why don’t we never talk about a term like Aµ Aµ ? This would have been a perfectly valid term in L.

the chain rule of differentiation. (81). However.∂α ∂ β Aλ − ∂ β L ∂(∂α Aλ ) ∂(∂α Aλ ) ∂L ∂L β ∂ Aλ + ∂ β (∂α Aλ ) − ∂ β L ∂Aλ ∂(∂α Aλ ) (87) = ∂ β L(Aλ . and (87). and we have T 00 d3 x = 1 2 E2 + B2 d3 x = E. (85) These are the usual expressions of the total energy and momentum of the electromagnetic field. This is a four-divergence and hence gives zero under integration. but that does not matter since they are all ordinary three-vectors. (88) The quantity in parenthesis is symmetric. ∂ α Aλ ) − ∂ β L = 0.E = 0 to get the last term on the right-hand side. as ∂µ F αµ Aβ . note that T αβ is not symmetric. T αβ is not a symmetric tensor. (81) and (84) over the whole volume. with source-free Maxwell’s equation ∂µ F αµ = 0. Similarly. the three-divergence terms drop out due to the divergence theorem. (84) Integrating eqs. (86) This is easy to prove.where we have used the free-field equation ∇. Calculate T i0 and show that it is different from T 0i . and the fact that L is only a function of Aλ and its derivative. (Ai E) . 20 . (83) where we have been a little cavalier in the positioning of the indices. and of course we would prefer quantities to be gauge invariant. The traceless property is needed for a massless photon to emerge after quantisation. Satisfy yourself that you understand all steps that led to eqs. (82) Again. From eq. 4 In the intermediate steps. It is also neither traceless (T αα = 0) nor gauge invariant. So we may neglect this term and define a symmetric stress tensor Θαβ as 1 Θαβ = F αµ η βν Fµν + η αβ F µν Fµν 4 . Now we can write eq. T 0i = −E j ∂ i Aj = E j ∂i Aj . The last term can be written. the symmetric nature is needed to conserve angular momentum of the field. To this end. Q.∂ β Aλ + . (89) Q. T 0i d3 x = (E × B) d3 x = Pi . we note that (E × B)i = ǫkij ǫklmEj ∂l Am = (δil δjm − δim δjl ) Ej ∂l Am = (Ej ∂i Aj − Ej ∂j Ai ) . ∂α T αβ = ∂α = = ∂L ∂ β Aλ − ∂ β L ∂(∂α Aλ ) ∂L ∂L ∂α . we write 1 T αβ = −F αµ ∂ β Aµ − η αβ − F µν Fµν 4 1 αβ µν = F αµ η βν Fµν + η F Fµν − F αµ η βν ∂µ Aν . (82) as T 0i = (E × B)i + ∇. it is easy to formulate a differential conservation law. (85). a four-dimensional analogue of Poynting’s theorem: ∂α T αβ = 0. we have used the Euler-Lagrange equation. (84). From the definition of T αβ .

while a real photon has two (left and right circular polarisation). Obviously. (92) 21 . ∂µ Aµ = 0. and the effect follows). 2 Identify Θij with the negative of Maxwell’s stress tensor. the Lorentz gauge. only at the point where it intersects the light cone of the observer can I have any information about it. 6. light. takes a certain amount of time to reach the observer from the source.. t0 < t. you may call it an artifact of our formalism. but what about the second one? Well. Obviously. then at the latter point I can have the information about the electron when it was at the position (t0 .1 Retarded Potential This is really a very simple concept: electromagnetic wave. (91) This is called the inhomogeneous wave equation.e. (90) 6 Potential Formulation The four-potential Aµ is not unique. 1 Θij = − Ei Ej + Bi Bj − δij (E2 + B2 ) . 2 = (E × B)i . the equation of motion is Aµ = j µ . Aµ . being a Lorentz vector. Rµ Rµ = 0. two degrees of freedom are just spurious. x0 ). there is no way to know what is happening to the source of the electromagnetic wave (let us just call it the electron) at the present moment t. If I draw the world line of the electron. The gauge condition gives one constraint and hence removes one degree of freedom.e. j µ = 0. the source-free Lagrangian density is invariant under a gauge transformation. Is our counting of the number of degrees of freedom correct? Apparently. If we want a unique solution for Aµ . Q. at t) is some configuration that was there at some earlier time (say. the signal transmits. in fact. this is indeed a meaningful question with unambiguous answer. x). what the observer sees right now (say. So the scalar potential is a cyclic coordinate and can be removed by equations of motion. If the signal travels at the velocity of light. i. x0 ) and it intersects the world line of the electron at (t. is the most convenient one to use. If you happen to be familiar with the concept of the light cone. if I start my light cone from (t0 . and even the source term is if the current is conserved (and it is conserved). it is the boundary of the light cone that gives causally connected points. you know that points which are inside the light cone can be causally connected (i. In the absence of any source. Since the separation between these two points is light-like. at t0 ). Thus. if you look at the Lagrangian density of the field (even when a source term is present) you will immediately see that the momentum conjugate to A0 is zero. Show that Θ00 = Θ0i = Θi0 1 2 (E + B2 ) ..Q. we must specify the gauge. the equation reduces to the homogeneous wave equation. and in relativistic electrodynamics. there is a cause. since there is no ∂0 A0 term there. this is the retarded position (and the retarded time) of the electron. has four components. To be more specific. it does not represent a true degree of freedom. The question we would like to ask: what is the potential at any point due to a moving charge? Now that the gauge is specified. Show that ∂α Θαβ = 0. In this gauge.

Convince yourself that causality is satisfied only on the forward light cone. γv). 6. (94).v) (97) . we take c = 1)? 6. of course. But j µ jµ is Lorentz invariant. j µ = (ρ0 . It is only the retarded potential. but the volume element dV0 gets contracted to dV /γ (contraction occurs only in the direction of motion). (94) Aµ = 4πr0 where r0 is the spatial distance between the charge and the observer. but that is hardly an answer. This is nothing but Coulomb’s law. There is another point of intersection on the backward light cone. 0) in the static frame. we know the solution of the inhomogeneous wave equation: e . so ρ = γρ0 . These are called advanced time and advanced position of the electron. How should the light-cone condition. In a frame where the configuration is static (we neglect the internal motion of the configuration). v) . so that u. j µ = (ρ. so ρdV is a constant. which introduces the concept of the retarded time in a roundabout way and does not 22 e (1. and the observer at xµ . we can write e uµ . Since the proper velocity uµ is (1. Q. though a clever e guess is required. In a moving frame. since that violates causality. that we can talk about and calculate. and Rµ = (r. current is nothing but motion of charge.R = r0 . the advanced coordinate is related with the motion of antiparticles. both theoretically and experimentally. Rµ Rµ = 0. uµ = (γ. There are two other ways: one is the so-called method of ‘information-collecting sphere’. unique.3 The Li´nard-Wiechert Potential e Consider an electron of charge e (here e is negative) to be at a point xµ . 0). Note that the charge density ρ is not Lorentz invariant. but the retardation condition tells us that Rµ = (x2 − x1 )µ = (r0 . x0 ). Draw the light-cone diagram that we have discussed just now. e This is the simplest way to derive the form of the Li´nard-Wiechert potential. but that is not of any physical relevance. effect precedes cause. r). Aµ (t0 . and in a frame which is moving with a uniform velocity v with respect to the former.R subject to the condition R2 = 0. The point of intersection is. if it were not. look like if the velocity of the signal is some v < 1 (remember. (96) Aµ = 4π u. (93) To get the total charge I have to integrate over the volume.2 Digression: Lorentz Invariance of Electric Charge Why the electric charge e is Lorentz invariant? It would have been a total mess.where Rµ = (t − t0 . (95) Our task is to write the solution of Aµ in some Lorentz covariant form which in the static limit reduces to eq. x − x0 ). However. r0 ) . At 1 2 the proper frame. where the electron is at rest. so Aµ = Eq. ρv). 4π(r − r. they are not just esoteric concepts. since it is the zero-th component of the four-vector j µ . Q. (97) is the Li´nard-Wiechert potential. it may be shown that when we quantise the field.0 .

What we derive in the preceeding subsections is a function of space and time from which F µν . then. 10 23 . What should be the form of eq. (97). so no signal of electromagnetic disturbance can travel to a far observer instantaneously. when a particular prescription is followed. Its main use is that it gives the correct E and B. the four-potential — comes back with Be careful about the notation. section 19.2.v term in the denominator. namely. we also say that nothing can move faster than light in vacuum. Where. see Panofsky-Phillips. so the fields should come out as a function of the retarded position only. It is just a function of space and time. • For high velocity the field is like that of a plane transverse wave. In eqs. the introduction of the potential helps us to calculate the scattering amplitudes (we have not shown how to calculate them in classical relativistic electrodynamics. Explicitly show how you get eq. may be calculated. When we bring in the Special Theory of Relativity. and we can express the fields in terms of its present position. (97) for the potentials. the present position is not known. is the place of the potential? Why should we waste so much time talking about Aµ ? Why should we derive the form of the Li´nard-Wiechert potential? Isn’t it a useless concept in relativistic e electrodynamics? To answer this apparent paradox. The second method is through the use of Green functions (see Jackson for a detailed analysis). (97) in the SI and the Gaussian systems? 6. The extra complication comes from the r. E = −∇A0 − ∂A/∂t and B = ∇ × A. Indeed. for this method). or Griffiths. We have to use the usual definitions of E and B. They are obviously not the same. This kills such action-at-a-distance theories. but at the same time mathematically more complicated. We can calculate the radiated energy in terms of E and B. if we know the retarded position and the velocity of the charge. we would scarcely spend so much time in deriving just another function. r denotes the present position of the charge. • The field does not radiate. In eq. What should be the electric and the magnetic fields at a certain distance? Well. Only in this case. This is no doubt more rigorous. to the observer. electromagnetic gauge transformation tells us that. and hence E and B. That’s the end of the story.5 The Fate of the Potential I have said enough to confuse you thoroughly. There is one exception: when the charge moves with uniform velocity. If you cannot. but we have to use eq. Q. The electric and magnetic fields are the only measurable and relevant quantities. But this is precisely the same game that we have played earlier! We have seen. It is not unique. that the fields are indeed directed towards the present position of the charge 10 . r denotes the retarded position. (45) from eq.contain an iota of more physics insight than this (see Panofsky and Phillips. This is true for both scalar and vector potentials. Let us recall the main conclusions: • A moving charge induces a magnetic field. (45) and (46). and this is not an easy job). Q. from eqs. let us first understand what potential is. we can write F µν in terms of E and B alone. But if that were the only case. In the nonrelativistic case. First. 6. The fact is that potential — by now I mean Aµ .4 Fields due to a Moving Charge Suppose a charge is moving along some trajectory. (97). I said that the potential smells of actionat-a-distance and hence can be a valid concept only in a nonrelativistic theory. (45) and (46). the present position is also known.

v = r0 1 − v 2 sin2 ψ. Both the fields go as 1/r 2 and the Poynting vector vanishes over the surface at infinity. This operator. p denotes the magnitude of the three-vector p. and is done in any standard textbook. In fact. We call r0 the virtual present position of the charge. In short. to avoid confusion. which does a 24 . my favourite is Griffiths.) The expressions for E and B for a uniformly moving charge is given in eqs. and r0 denotes its present position if it would have moved with uniform velocity. since any information about the present position is inaccessible to us. say an electron. E is directed along the present position r0 ). 7 Accelerated Charge In this section.v. The velocity and the acceleration of the charge is v and a respectively. ψ is an operator — this we will call the electron field. The angle between r and v is θ.v)2 − 2rr. ψ is a wavefunction that describes a particle. uniformly moving charge does not radiate. acting on a state. the present position can be something completely different. but remember what Shakespeare said about roses. That’s where its importance lies. (98) 2 (To get this. and clearly r0 = r − rv (remember that v < 1 and light moves with unit velocity). x Figure 2: Retarded and (virtual) present positions of a particle. and subtract one from the other. use r0 = r 2 + r 2 v 2 − 2rr. In quantum mechanics. it can be shown that (the deduction is not difficult but involves a lot of bookkeeping. Aµ is the field of the photon. In quantum field theory. you quantise it and get the photons as field excitations. vr x’ θ ψ v Present pos. For an accelerated charge. we will use the following convention: r denotes the retarded position of the moving charge.v and s2 = r 2 + (r. but that is a separate issue. The Li´nard-Wiechert denominator is e s = r − r. while that between r0 and v is ψ. You can still call it a four-potential if you like. also use r × v = r0 × v. Since it does not. It can also destroy a positron. and this is a function of xµ — whose excitations are the particles. In general. but that really does not matter. may create an electron (so that the number of electrons is increased by one). we would like to write B=v×E= 1 r × E. ψ is upgraded to the status of an operator.a vengeance in quantum electrodynamics. r (99) since r0 × r0 = 0 (remember that only for uniformly moving charge. This is a quantum field theory which describes the interaction of electrons and photons. Retarded pos. (45) and (46). r r0 Observation pt.

2π = 16π 2 3 6π (105) This is known as the Larmor formula. (ii) the radiation pattern is symmetric for θ → π − θ. The power radiated does not depend on the sign of a. π + θ and 2π − θ. but we e will follow a shorter path. we will get a radiation. (i) there is no radiation along the direction of instantaneous acceleration. this is the radiation field and exists only if a = 0. so whether the charge is accelerated or decelerated. (105) change if the motion is relativistic? The deduction is due to Li´nard. so Erad = and q q r × (r × a) = [ˆ(ˆ. v → 0.Brad = 0. r ≈ r0 . make a dimensional analysis to see how many powers of c you need.a)v + γa = γ 3 (va cos α)v + γa. S= To get the total power. How should eq.a = γ 3 va cos α. r r r (104) 16π 2 r 2 16π 2 r 2 where θ is the angle between r and a.] 7. dt (106) = γ 3 (v. Before that.a)2 ] = ˆa2 sin2 θ. r r (101) (102) 7.2 Relativistic Generalisation of Larmor’s Formula How does eq. Thus. Therefore. so that d(γv) dt − dγ dt 2 = γ 2 a2 +γ 4 v 2 a2 cos2 α = γ 2 a2 (1+γ 2 v 2 )−γ 4 v 2 a2 sin2 α = γ 4 a2 − v 2 a2 sin2 α . Note that these fields are transverse: r. r q qr0 + {r × (r0 × a)} .brute force job) E = B = 3 4πγ 2 r0 (1 1 r × E.1 Radiation from a Slow-moving Charge For a nonrelativistic motion. . and is valid only for nonrelativistic motion. (105) look in the SI and the Gaussian systems? [Hint: Don’t just put back the factor of ǫ0 . (107) 25 . where α is the angle between v and a. so shaped like a doughnut (an 8-shaped pattern) around the direction of a. we have to integrate S over the surface of the sphere with radius r: P = S. using the covariance argument. S=E×B= 1 1 E 2 r − (r. . 3 (1 − v 2 sin2 ψ)3/2 2 sin2 ψ)3/2 −v 4πr0 (100) The second term is the component that carries energy away to infinity (both E and B go as 1/r and hence the Poynting vector is nonzero even at infinity). Q. let us just work out two derivatives: dγ dt d(γv) dt 2 = d (1 − v 2 )−1/2 = γ 3 v.Erad = r.a) − a] rr 4πr 3 4πr (103) q2 q2 ˆ[a2 − (ˆ.E)E = E 2 r.da = q 2 a2 16π 2 sin2 θd(cos θ) dφ = 1 2 2 q 2 a2 4 q a .

except that the denominator will now have s3 instead of r 3 . Let us try to understand the origin of this correction. (103).. Is this enhancement isotropic or is there a directional bias? This question we will address soon. (111). dτ dτ (109) where τ = t/γ is the proper time. 2 dv = dS dr = dS 1 − 26 s v. so that the denominator contains (1 − v cos θ)5 . (105). one of radius r centred at x′ and the other of radius r + dt′ centred at x′ .Writing eq. the expression for P reads P = q2 2 γ 6πm2 dp dt 2 − dE dt 2 = q2 6 2 γ a − |v × a|2 . but before that. (104) but an extra factor of (1 − v cos θ)6 in the denominator. (105). (107) and (110). Since dr = dt′ − [r. in the relativistic case. The energy emitted by the electron in a time dt′ is located in the volume between two spheres. and the electric field is given by eq. (110) and in the nonrelativistic limit. suppose this subtends a solid angle dΩ = dS/r 2 at x′ . which is analogous to the well-known Doppler shift. s 1 − v cos θ (113) so we expect the Poynting vector to have the same form as in eq. A comparison shows that there is a huge boost factor of γ 6 .v)5 where n is the unit vector along r. However.e. An angular integration recovers eq. dΩ 16π 2 (1 − n. . Then r0 × a = r × a. E = γm. i. Now − dpµ dpµ = dτ dτ dp dτ 2 − dE dτ 2 . The same is true for B. so we get back eq.v/r]dt′ . What should be the form of eq. (112) in the SI system? (112) 7. without deriving: q 2 |n × {(n − v) × a}|2 dP = . r r (114) . Consider an infinitesimal 1 2 volume element dv of this asymmetrical shell. and with eqs.r dt′ = dS dt′ . and not dpµ /dt. Now there is a catch. But r 1 = . (105) as P = q2 6πm2 dp dp . the time kept by the clock of the charge. let me just quote a formula for the angular distribution of radiated power. and the variation of E is negligible (since that is overwhelmingly controlled by the rest mass m). is a Lorentz vector. There will be a further correction of s/r on the energy loss. Note that only dpµ /dτ .3 Relativistic Motion: v a Suppose that at a particular retarded time tr the velocity was instantaneously parallel to the acceleration. Q. 6π (111) This is the relativistic generalisation of eq. p = γmv. dt dt (108) we immediately see that the relativistic generalisation is P =− q2 6πm2 dpµ dpµ . so the emitted power is tremendously enhanced. τ → t.

it is sharply peaked in the forward direction (see fig. Though the power at precisely θ = 0 is zero. the power radiated per unit solid angle. (1 − vx)5 3 (1 − v 2 )3 1 q 2 a2 γ 6 q 2 a2 4 = . we have to integrate over Ω. and the integration over θ can be performed substituting x = cos θ and using 1 −1 1 4 (1 − x2 ) dx = . e The power is not only enhanced. The angle θmax where the radiated power is maximum can be obtained by differentiating dP/dΩ with 27 . dΩ 16π 2 (1 − v cos θ)5 (115) To get the total power. eq. 8π 3 (1 − v 2 )3 6π (116) so that P = (117) consistent with the Li´nard formula. (111). 1 Therefore the energy contained in this volume within the solid angle dΩ is 1 (E 2 + B 2 )(s/r)dSdt′ = 2 E 2 (s/r)dSdt′ .v r dt’ 1 r ′ Figure 3: Location of energy radiated by an electron as it moves from x′ to x2 .dv dS dr r+cdt’ 2 v dt’ r. 4). dP/dΩ. most of it is concentrated within a narrow cone. is given by q2 a2 sin2 θ dP = . Thus. The integration over φ gives 2π.

is given by dP q 2 a2 sin2 θ dE = = |S|r 2 = |Erad |2 r 2 = . a = ak. so √ 1 15v 2 + 1 − 1 1 −1 . we will discuss the case v a but v ≪ 1. Q. Find θmax for electrons with energy 1 GeV. very crudely speaking. There is a frequency distribution of the spectrum. you have to check the second derivative to ensure that this is a maximum). Let n be the unit vector in the spherical polar coordinate: ˆ n = sin θ cos φˆ + sin θ sin φˆ + cos θ k. (115) from eq. This is. ˆ ˆ Q.4 Frequency Distribution: Bremsstrahlung for Slow Electrons Let us now go into a bit more detail of what has been said in the last paragraph of the preceding subsection. (119) 7. called bremsstrahlung.98. into a solid angle dΩ. dt dΩ dΩ 16π 2 28 (120) . The amount of energy E flowing out per unit time.) Q. There is no lower limit on the frequency. Note that the power has been scaled down by a factor of 10−5 for v = 0. or the power radiated P . (104). For simplicity. what happens when a fast-moving electron decelerates (maybe within some material). The emitted radiation. v = 0.4 3 2 1 0 1 2 3 4 2 0 2 4 6 8 10 12 Figure 4: Plot of dP/dΩ for v = 0 (the 8-shaped pattern). 2 mvmax /¯ . so that the retarded position and time are almost identical with the present position and time (s ≈ r).5 and v = 0. the electric field that contributes to radiation is given by eq. (103) and the Poynting vector by eq. This gives 3v cos2 θmax + 2 cos θmax − 5v = 0. (Hint: express v in terms of γ and do a binomial expansion in terms of 1/γ 2 . the maximum frequency of the emitted 1 2 h photon is obviously the change in the kinetic energy of the highest velocity electron. Show that at the limit v → 1. i j Get eq.98! respect to θ and setting the derivative to zero (of course. (118) −→ cos−1 1 − 2 = θmax = cos 3v 8γ 2γ where the last fraction is at the limit v → 1. (112). is peaked in the forward direction. Recall that in this limit. Assume both v and a to be along the z-direction: v = v k. There can be infinite number of zero-energy photons (the total energy is still zero!) emitted along the direction of v. θmax = 1/2γ.

) With such a simple ansatz. the frequency spectrum will have a high-frequency cutoff at ωmax ∼ 1/τ . Eω = √ 4πr 2π so that. in some sense. ∞ u = δ(t − t0 )∆u =⇒ ˙ udt = ∆u. and there is no cutoff on ω. which does not have any meaning.What is the spectral composition of the radiation? In other words. we have 1 q sin θ ∆ueiωt0 . and occurs only because we have taken the interaction to be of zero duration. (124). So. this distinction is irrelevant. Et during this time interval is proportional to ∆u/∆t. dE = 2 S (126) |Eω |2 dSdω 2π π = 2 = 1 q2 (∆u)2 dω 2π 16π 2 q2 dω(∆u)2 . −∞ (121) We will also need Parseval’s theorem. the energy of a bremsstarhlung photon cannot be greater than the kinetic energy of the charge. for example. Actually. Let us use the notation Et ≡ |Erad |(t). This is nothing but making the velocity change instantaneous. We define the Fourier Transform (FT) of Et as 1 Eω = √ 2π 1 Et eiωt dt. (123) dE = 2r 2 |Eω |2 =⇒ dE = 2dω dω dΩ (124) S where we have used r 2 dΩ = dS. the width of Et ). To go further you have to know the form of Et . but since the velocity is small. what is the amount of energy in the frequency band ω to ω + dω? For that. ˙ −∞ (125) (Remember that all coordinates associated with the charge are actually retarded coordinates. if the velocity change occurs for a small time τ (which defines. This is clearly unnatural. we first need to know how Erad varies with the time t. (120). the uncertainty principle in quantum mechanics. (122) where we use the last step to avoid dealing with negative frequencies. Et = √ 2π −∞ ∞ ∞ Eω e−iωt dω. we have dE = dΩ so that |Et |2 r 2 dt = 2r 2 |Eω |2 dω |Eω |2 dS. From eq. which we can write as ∆uδ(t − t0 ). Thus. 2 (128) . which says ∞ −∞ ∞ ∞ |Et |2 dt = −∞ |Eω |2 dω = 2 0 |Eω |2 dω. Suppose the velocity change ∆u of the charge takes place in a very short time interval ∆t. from eq. Recall that the FT of a narrow Gaussian (width∼ σ) is a broad Gaussian (width∼ 1/σ) — this helps us to understand. so h ¯ ωmax = 29 1 mu2 . But we can say something important even qualitatively. If radiation takes place at t0 . 6π 2 dφ 0 0 sin3 θdθ (127) This means that the energy spectrum is uniform: equal amount of energy is emitted at every frequency interval. and hence the time for the charge to interact with the retarding electric field zero.

Thus. This is precisely why the LEP is in all probability the last circular e+ e− collider. Near θ = 0. they are always perpenˆ dicular). i j (131) Squaring. Do you have any idea why this looks like a diffraction pattern? 7. we get (after some trigonometric bookkeeping). so the pattern is almost uniform in φ.e. (127). during which the velocity changes uniformly and the total change is ∆u. dΩ 16π 2 (1 − v cos θ)5 An angular integration yields (do this) P = 1 2 2 4 q a γ . i. does not affect the spectrum. Suppose that the velocity change is not instantaneous. draw the ω versus E plot. and for the same energy. ξ2 (130) where ξ = ωτ /2. Plot the functional form of |Eω |2 . The radiation is sharply peaked in the forward direction along v. 6π (133) (132) which you can get in a straightforward way from eq. (128). but occurs for a time interval t0 − τ /2 to t0 + τ /2. θ = 0.5 Relativistic Motion: v ⊥ a Suppose v and a are instantaneously perpendicular (for a circular motion. For an electron in a circular orbit. This tells you that as ω goes to zero (photons becoming softer).The number density of emitted photons in the frequency interval dω is given by dNω = dE 2 dω = α (∆u)2 . we have P = 1 q 2 2 dp γ 6π m2 dt 2 . and as the rate of change of momentum dp/dt = γma. (112). after the machines where it was first observed. Show that in the SI system. for two reasons: m is smaller. Show that in eq. the φ-dependent term is small. ˆ n×{(n−v)×a} = a {− sin2 θ sin2 φ + cos θ(v − cos θ)}ˆ + sin2 θ sin φ cos φˆ − sin θ cos φ(v − cos θ)k . Q. Taking v = v k. Q. we get i. Remember that dω ∝ dλ/λ2 . with a circumference of 27 km!) 30 . though protons will soon start running in the same machine with a much higher energy. the number density increases. the radiation sweeps like a beacon. Using eq.. obviously. Show that |Eω |2 ∝ sin2 ξ . the last equation in (124) will have a multiplicative factor of ǫ0 c on the right-hand side. This is known as the synchrotron radiation. the right-hand side is to be multiplied by 1/(ǫ0 c3 ). Q. h ¯ω 3π ω (129) where α = 1/137 is the fine structure constant. it is possible to shake off infinite number of zero-energy photons. q 2 a2 [(1 − v cos θ)2 − (1 − v 2 ) sin2 θ cos2 φ] dP = . (119). Q. but with significantly smaller synchrotron loss. This. γ is larger. using eq. (111). and ultimately blows up for ω = 0. A charged particle revolving in a circular orbit loses energy. (134) so that the energy loss is much more for an electron than a proton. (This machine is an engineering marvel: a perfectly circular pipe. a = aˆ and eq. replacing ω by the wavelength λ. deep underground. Redraw the same.

and more so for arbitrary motions. we should complex conjugate the polarisation of the outgoing wave compared to the incoming wave. sec. a little vector algebra shows that out 11 Since the polarisation vector is in general complex. is given approximately by ωc ≈ ω0 γ 3 . corresponding to 16 keV X-rays. where the energy of the photon is much less than the rest energy of the particle). γmax ≈ 2 × 104 . In the nonrelativistic limit. For our subsequent analysis. ωc . This process is known as Thomson scattering. pulsars radiate away a huge amount of power in a highly directional beacon. which makes an angle θ with k0 . e1 and e2 ). Pulsars are rotating neutron stars whose magnetic axis does not coincide with the rotational axis (the same is true for earth too). With a typical radius of 10 km. the frequency spectrum of the synchrotron radiation contains a large number of components.The same phenomenon occurs in the sky. On and near the surface of the pulsar.6 Thomson Scattering Suppose a plane monochromatic wave falls on a free particle of charge q and mass m. (135) where ω0 is the frequency of rotation. 31 . ω0 ≈ 3 × 106 s−1 . This will transfer some energy to the particle. Remember that ein is a vector of unit magnitude and ein . m The scattered wave propagates along n (see figure). Given. Calculate the ratio of synchrotron losses (power radiated) for LEP-II (electrons with 105 GeV) and LHC (protons with 7000 GeV). so ωc ≈ 2. (112).4 × 1019 s−1 . k0 ). and a surface magnetic field of 108 Tesla. What was the value of γmax ? What should be γmax for 7000 GeV protons? Q. and therefore relative to the fixed axis of rotation. (This is a naive picture of a pulsar. the electric field can be written as E = ein E0 exp(−ikµ xµ ) (137) where k = (ω. however. we will take k0 along the lab-fixed z-axis. so the particle will be accelerated and emit its own characteristic radiation.k0 = 0. (132) from eq. so we get some beacon-like pulsed radiation based on more or less the same principle. the frequency of the scattered wave is the same as the incident wave. since the wave is transverse. in that case we won’t be able to see it) we see the pulsar.) Even for a circular motion. 14. The force equation for nonrelativistic motion is ma(t) = qE. (112) reads q2 dP = |n × (n × a)|2 . a rotational period of 10−3 s. and when that beacon intercepts the earth (it may not. Q. eq. but the plasma in turn radiates. 7. If its polarisation be e∗ 11 (which can have two possible values. The last run of LEP had 105 GeV electrons.4) that the highest frequency in the synchrotron spectrum. so that without any loss of generality. If its propagation vector be k0 . It can be shown (see Jackson. (136) dΩ 16π 2 The acceleration is provided by the incident plane wave. two linearly independent choices of ein may be taken along the x and the y axes. and its polarisation vector ein . This radiation will be emitted in all directions (but will not be spherically symmetric). and the star gives off magnetic dipole radiation. mp = 1836me . In the nonrelativistic limit (quantum-mechanically. the magnetic dipole vector is changing. Q. Note the factor of γ 3 : in a typical 10 GeV machine. so q (138) a(t) = ein E0 exp(−ikµ xµ ). Do the necessary vector algebra and trigonometry to get eq. such radiations get damped because of the electromagnetic plasma present there.

But the denominator is nothing but the time-averaged Poynting vector (the time-averaging brings a factor of 1/2). For nonrelativistic motion. (142) The scattering geometry is shown in fig. dP q4 1 = |E0 |2 |ein . 32 . In terms of unit vectors of the cartesian system i. (143) In general we do not see the individual final-state polarisations. e2 = − sin φi + cos φj. (140) dΩ 16π 2 out The electric field. though a function of time. and hence a.eout |2 .a|2 . we may write e1 = cos θ cos φi + cos θ sin φj − sin θk. j.(convince yourself of this!) ∗ |n × (n × a)|2 = |eout . The polarisation vector e1 lies in the plane containing n and k0 .e. and so the time-averaging is essentially computing |ein .eout |2 . which is |E0 |2 /2.a|2 . so we have to sum over all possible polarisation states. is defined as the ratio of energy radiated per unit time (i. (139) dP q2 = |e∗ . a. so dσ = dΩ q2 4πm 2 |ein . is a rapidly oscillating function of time. If the initial beam is polarised along the x-axis. 5. and k. the angular distribution. dσ/dΩ. does not change much over the oscillation period.. and e2 is perpendicular to it. and we need an average over a complete cycle of oscillation.eout |2 and replacing the harmonic part by 1/2: so that k 0 n θ e2 φ φ e θ 1 Figure 5: Polarisation of the scattered wave. dΩ 16π 2 m2 2 (141) The differential scattering cross-section. power) per unit solid angle to the incident energy flux per unit area per unit time.

average over all initialstate polarisations and if you do not look for a specific polarisation in the final state. calculate the maximum linear velocity of a point on its surface. The calculation is quantum-mechanical in the sense that it uses the concept of photons. π) and with φ for a fixed value of θ (take θ = 0. show that α2 /m ≈ 2.. Q. This is expected since spin does not have a role in classical physics.82 × 10−15 m. and by integrating. Quantum-mechanically. ¯ What’s wrong with this value? Q. Note that the spin of the particle does not play any role in the scattering 12 . Calculate the Thomson cross-section for electrons in barn (1 barn = 10−24 cm−2 ). The distribution is symmetrical around θ = π/2 (and does not have any φ dependence). while if it is polarised along the y-axis. Take the same values of θ and φ as given in the previous problem. whether classical or quantummechanical: to get the scattering amplitude for unpolarised incident beam. q 2 /4π is the fine-structure constant α (in the natural system). but the result can be translated into the language of classical electrodynamics. Also. it is rather subtle to show that the same cross-section is obtained in the nonrelativistic limit for spin-0 and spin-1/2 particles. π/4. Show how the differential Thomson cross-section. but a classical distribution of charge. see the next subsection. dσ/dΩ. π/4.7 Modifications to the Thomson formula: Compton. π). 197 MeV-fm ≈ 1 (energy and length have opposite dimension. scattering of waves upto X-ray (∼ keV) by electrons or upto γ-ray (∼ MeV) by protons. so dσ = dΩ unpol q2 4πm 2 1 1 + cos2 θ . If we sum over the final-state polarisations but start with polarised incident wave. π/2. Eq. 2 (144) This is a procedure which is true for all scattering calculations. If its angular momentum is h/2. and electron mass m = 0. which is approximately 1/137. will vary with θ for a fixed value of φ (take φ = 0.g. there is a difference: spin-1/2 particles have a nonzero magnetic moment which interacts with the wave and contributes to the scattering matrix element. modifications to eq. In the natural system of units. If the incident beam is unpolarised. we have to average over these contributions. electron is a point particle. The ratio α/m. e. (144) is the Thomson formula for scattering of electromagnetic radiation by a charged particle. the classical electron radius. Suppose we look at only those scattered waves which are polarised along e1 .82 × 10−15 m. must have a radius of this order if its electrostatic selfenergy is equal to the electron mass.82 fm = 2. Suppose the electron is a uniformly charged sphere of radius r0 . we get the Thomson cross-section: 8π σT = 3 q2 4πm 2 . (144) occur. the angular distribution is (cos2 θ sin2 φ + cos2 φ). However. Repeat the exercise with waves polarised along e2 . What should be the formula for the classical electron radius in the SI system? Q.511 MeV. (ii) beam polarised along y-axis. is (cos2 θ cos2 φ + sin2 φ). From this. Q. Q. (145) For an electron (or a proton). 7. Of course. which total the electronic charge. is approximately 2. This is called the classical electron radius. 12 33 . Klein-Nishina When the energy of the incident radiation is comparable to the rest energy of the particle. right?). π/2. for an electron. sum over all final-state polarisations. It is valid only if the energy of the incident radiation is much less than the rest energy of the particle.summed over final-state polarisations. α = 1/137. Q. show how dσ/dΩ will look like as functions of θ and φ for (i) beam polarised along x-axis.

a semiclassical description. At limiting cases. with a characteristic frequency ω0 . so that without any external force. 4E (148) For spin-1/2 particles like electrons. This is. and (iii) Klein-Nishina with E/m = 0. the equation of motion will be 2 m¨ + mω0 x = qE = qein E0 exp(−iωt) . To be precise. The acceleration. eqs. we can write σRayleigh = ω ω0 34 4 σT homson . (1 − cos θ) (146) where E ′ is the energy of the scattered photon (or the wave. This modifies the unpolarised differential cross-section: E′ dσ dσ = dΩ Compton dΩ T homson E 2 . the equation of motion is 2 m¨ + mω0 x = 0 . not only the charge but also the magnetic moment is responsible for scattering. where we assume the electron to be bound in a harmonic oscillator potential. This is known as Rayleigh scattering. and θ is the scattering angle. the Compton cross-section is indeed given by σ = σT (1 − 2E/m). in turn. 4kk′ (149) where k and k′ are the wave numbers of the incident and the scattered waves respectively. Q. which is the Thomson scattering that we have just studied. (151). This modifies the matrix-element squared part of eq. if you prefer). (154) . Q. Show that for E ≪ m. − ω2 (152) which can be checked by direct substitution in eq. x (151) The solution is (which should be familiar to you if you have studied optical dispersion) x = ein q m 2 ω0 1 exp(−iωt) . (153) For free electrons. is scaled by a factor of (ω/ω0 )4 . the drop depending on the ratio E/m. σT m 3m (m ≪ E) . For strongly bound electrons. (147): |ein . of course. and dσ/dΩ drops at large θ from its Thomson value. (146) and (147) are true only for scattering from spin-0 particles.e∗ |2 −→ |ein . To get the total cross-section. Plot the differential scattering cross-section as a function of θ for (i) Thomson. we get σ E =1−2 (E ≪ m) . 7.Considering the scattering of a photon of energy E = hω by a particle of rest mass m and applying ¯ the two-body relativistic kinematics. ω0 ≫ ω. (ii) Compton with E/m = 0. we get the Compton formula: E′ = E 1+ E m 1 . This is known as the Klein-Nishina formula. ω0 → 0. which is proportional to a2 .8 Scattering from Bound Electrons Consider an electron bound to an atom.5.e∗ |2 + out out (k − k′ )2 . (147) The modification factor comes entirely from the phase-space. one must remember that E ′ /E is also a function of θ. is given by ¨ a = x = ein q m −ω 2 2 ω0 − ω 2 exp(−iωt) . x (150) If we apply a harmonic electric field E = ein E0 exp(−iωt) on this electron. and dσ/dΩ.5.

they always increase inside the shock. Outside the cone the field intensities are zero. theoretically. This is as if the emitted radiation is exerting a back force on the charge. This explains why the rising and setting suns are red. one gets radiation inside this cone. 8 Radiation Reaction An accelerated charge radiates. with energies 5 MeV and 10 MeV respectively. so it loses energy. the rays have to cross a thicker layer of atmosphere and bluish components are scattered away. of the particle.33. its acceleration will be smaller than a particle with same mass but no charge. thus. where the direction of the incoming particle is extremely important to know. So far. This effect was first observed by Frank and Tamm. and the boundary is known as the shock front. and so is the absorption. you can accurately determine the velocity. we expect a (41/65)4 ≈ 16% scattering of the violet end compared to the red end. Experimentally. an infinite field intensity in the direction specified by θ: a cone about the velocity vector. we have started from the expression of Erad . We have never considered the fact that velocity and acceleration may change due to the reactive effects of radiation. enter water. whose n = 1. we expect that under a given force. There is one exception. Q. So there will be. and calculated the Poynting vector. Two electrons.9 Cherenkov Radiation We have said that a uniformly moving charge does not radiate. The frequency distribution of the emitted radiation is continuous but the intensity increases with increasing frequency. which are less scattered. there is more chance to catch the scattered blue components than the red ones. this is known as Cherenkov radiation. What happens when ω0 = ω? This is when the electrons vibrate with their resonance frequency. this factor is always positive. Since v < 1. on the surface they are infinity and inside the cone they are finite but nonzero. The refractive index. 7. The Cherenkov effect is used extensively in high-energy particle detectors. But suppose a charged particle is moving in a medium where its velocity exceeds that of light in that medium. This is nothing but an electromagnetic shock wave 13 where the surface of the cone is the shock front. even the visible radiation appears blue. 13 An acoustic shock wave is characterised by abrupt change of pressure and temperature. Calculate the Cherenkov angles. Note that if you can successfully track the emitted light (with photon detectors like photomultiplier tubes). For cosmic-ray and neutrino experiments (like SuperK). Note the factor (1 − v cos θ)3 in the expressions of E and B. In the next section. we may get a light signal at an angle θ. we have never included this effect in our calculations. and can be zero if θ = cos−1 (1/nv). Cherenkov detectors are essential. and hence the energy. the displacement is maximum. This also explains why the sky is blue. 35 . and then integrated the flux over a surface at infinity. this is known as the region of anomalous dispersion. That is why most of the Cherenkov radiation is in the ultraviolet. drops sharply. For example.Thus. This is known as radiation reaction. in the calculation of bremsstrahlung or synchrotron radiation. So if a highly energetic particle enters a medium. and the theoretical explanation was provided by Cherenkov. if we are not in the line of the sun. where n is the refractive index of the medium. instead of increasing slowly with ω. the cross-section is large for higher frequencies. Thus. The red range of the visible spectrum is at 4100 A while the violet end is at 6500 A. we will see how the pole at ω = ω0 is physically removed. The denominator will now read (1 − nv cos θ)3 .

E0 . E0 ∼ mv 2 = m(aT )2 .25 × 10−24 s. the radiation reaction is always a small perturbation.876 × 10−13 cm = 6. Let us consider both these cases one by one. c = 1. the experiments match with calculations. Putting the relevant quantities. This justifies the neglect of this effect in our earlier calculations. This is indeed a very small time. is of the order of q 2 a2 /6π. we can neglect the radiation reaction if τ ≪ ω0 . Case 1: Steady acceleration — The radiated power is given by the Larmor formula: P = Note that this is a non-relativistic formula. An electron with such an energy is highly relativistic. Thus. this is one case which classical electrodynamics is just not competent enough to deal with. Thus. we get the condition for radiation reaction to be significant over a full cycle: ω0 τ ∼ 1 . the total energy loss in time T is of the order of q 2 a2 Erad = T. where the particle accelerates uniformly from t = 0 to t = T . we should have q2 q 2 a2 T = ma2 T 2 =⇒ T = . everything that we derive now is nonrelativistic (a relativistic generalisation will be considered later). This is true. It is better that this is so. Case 2: Periodic motion — With an amplitude d and a characteristic frequency ω0 . typically. (157) 6π 6πm This value of T is known as the characteristic time and will henceforth be denoted by τ . radiation reaction is significant if T ≈ τ and negligible if T ≫ τ . the radiation reaction should be small in such cases. so both should be of the order of τ ). Thus. even in the most exotic cases like fast-rotating pulsars. (156) If Erad (not to be confused with the radiation electric field) is to be comparable with E0 . What is the energy scale associated with the time scale τ ? For electron. the steady acceleration. the periodic motion. while the acceleration a ∼ ω0 d and the time period T ∼ 1/ω0 . First. we conclude that the radiation reaction is important only if (i) the motion changes appreciably over a time scale τ . However. in general. we can justify ourselves. two types of motion that are physically interesting.Hopefully. The energy radiated in a full cycle is (q 2 a2 /6π)T . or (ii) over a length scale cτ (in our system. the radiation effects can build up over time. and 197 MeV-fm = 1. while all the derivations that we are going to perform are non-relativistic in nature. the typical 2 2 energy E0 ∼ mω0 d2 . and its velocity changes from zero to aT . Second. 36 . if it is at all significant. which should be compared with the typical energy E0 . mass of the electron m = 0. Thus. this is just 1/τ = 105 MeV. This gives τ = 1. light takes this much time to cross an atomic nucleus.1 When is the Radiation Reaction Important? There are. (155) 6π The typical energy of the system. and so is one of the major motivations for a quantum theory of electromagnetic fields. 8. where the particle vibrates with a characteristic frequency ω0 and an amplitude d. We use α = q 2 /4π = 1/137. as we will see soon. for example. because.511 MeV. Let us try to have an estimate of τ for electrons. (158) −1 Thus. Thus. and the pulsar should fractionally slow down over a million of years.

Lorentz assumed a spherical charge distribution. as the field due to a point charge blows up on the charge (E → ∞ as r → 0). however. The energy is carried away to infinity by the radiation fields. viz..g. One can have other distributions. Griffiths considers a dumbbell-shaped configuration. those that go as r −4 and r −3 . 6π (159) where the minus sign indicates that the energy is being lost. 14 37 . 6π (162) Thus. and they exert a back force as well 14 . dv t2 | − dt t1 t2 t1 d2 v . as for all we know the electron is a point charge. dt2 (161) whose first term vanishes by the periodic boundary conditions. dt dt dt = v. For an accelerated charge. is a big objection. we do not know. An exact derivation is a bit cumbersome. for example. The imbalances. let us just note down the main features. All is not lost. However. this. t1 (160) with the proviso that v(t1 ) = v(t2 ). • The next step is to break the charge distribution into several fragments. summed up.v = − q 2 a2 .8. (160) by parts: t2 t2 a. the former term sticks. In his original derivation. note that this is not a true derivation. the force exerted by fragment 1 on fragment 2 is not the same (and opposite) to that of fragment 2 exerted on fragment 1. energy is pumped back and forth into these components. the component of Frad perpendicular to v. but can be justified from the fact that the retarded position of 1 with respect to 2 is not the same as the retarded position of 2 with respect to 1. 6π (163) This is known as the Abraham-Lorentz formula. the simplest possible form of Frad can be written as Frad = q2 ˙ a. or the force of the charge on itself. the net force exerted by the r −4 and r −3 components vanish. which carry energy but do not take it far away from the configuration. this equation is incorrect.. Thus. in itself.v dt = 0 .v dt . One can integrate the right-hand side of eq. When the charge undergoes a periodic motion. but there are other components of E × B.v dt = − q2 6π t2 a. However. according to him it should be called Field reaction. so that t2 t1 Frad − q2 ˙ a . • The charge must be distributed over a small but finite region. However. This is why Griffiths calls Radiation reaction a misnomer. This is a breakdown of Newton’s third law. since if we take a time interval at the end of which the system returns to its previous state.2 The Abraham-Lorentz Formula The rate at which the charged particle loses energy is given by the Larmor formula: Frad . Without going into the derivation. a(t1 ) = a(t2 ). e. we would be safe if we write t2 t1 Frad .a dt . One has to assume a charge distribution. give the net radiation reaction.a dt = t1 t1 dv dv .

as is evident from the success of quantum electrodynamics. there we use different tricks to tackle this problem. if we consider the electron as a point particle. Rest of the terms drop out in the limit of a point charge. Thus. which can be written as a function of time. this is the mass that we will observe in an experiment. 8. is a charge that is momentarily moving with constant velocity: a = 0 ˙ but a = 0. this is not a success story. Let us try a solution ˙ v(t) = exp(t/τ )u(t) . proportional ¨ to a. However. (165) where τ is the characteristic time for the charge. so that ¨ v(t) = and Integrating this. Unless a(t = 0) is zero. τ (167) (168) (169) (166) q2 ˙ a = ma . We need a quantum theory to address this issue. The charge does not radiate. related to the exact nature of electron. we have seen that such a classical picture is untenable. The equation of motion is Frad = whose solution is a(t) = a(t = 0) exp(t/τ ) . τ t C exp(−t′ /τ )Fext (t′ ) dt′ . Of course. this is a very fundamental problem. but let me mention that the infinities haunt the quantum theories as well. But.3 Problems with the Abraham-Lorentz Formula ˙ Now for the paradoxes. Suppose the acceleration is constant: a = 0. this is a runaway solution: the acceleration increases spontaneously with time! One can. (170) 38 . so that one can write ˙ ¨ mv − mτ v = Fext (t) . And we could do that very successfully. in the limit when the distribution is shrunk to a point. this term blows up. 6π (164) 1 ˙ mu(t) = − exp(−t/τ )Fext (t) . So we can work with the Abraham-Lorentz formula in the classical limit. • To circumvent this problem. The charge radiates. but the radiation reaction is zero! The opposite extreme. the mass that we observe is the result of this self-force. try to circumvent the runaway solution. we get mu(t) = − 1 τ 1 ˙ exp(t/τ )u(t) + exp(t/τ )u(t) . Lorentz assumed the bare mass of the elctron to be zero. the mass would have been infinite! Clearly. the velocity of a point on the ‘equator’ of electron is far above the velocity of light! • The second term in the expression for the self-force is the Abraham-Lorentz term. There is a problem. all in all. It can be shown that the radius needed for the charge distribution in this case is exactly the classical electron radius. however.• The first term of the expression for the self-force is a term that is proportional to acceleration. and thus just adds up to the ‘bare’ mass of the charge. Suppose there is an external force. but there is a self-force acting on it! Consider the motion of a charge without any external force.

is equal to the work done per unit time. And these problems do not go away even in a relativistic classical theory. Frad . the energy radiated per unit time. 6π ds ds (177) 39 . Definitely. Theoretically. which is the zero-th component of F µ uµ divided by γ. so this gives S = −uµ d duµ d2 uµ =− uµ ds2 ds ds + duµ duµ . an advanced potential. ds (172) ˙ and the covariant generalisation of (q 2 /6π)a can be written as q 2 d2 uµ + Suµ . which is 2 d(uµ uµ )/ds. the acceleration at t will depend on the value of the force at some earlier time t′ ! If t − t′ < 0.4 Relativistic Motion: Dirac Formula A Lorentz-covariant generalisation of the equation of motion would be m duµ µ µ − Frad = Fext . Note that the last term is proportional to the expression for power in the relativistic generalisation of Larmor’s formula: P =− q 2 duν duν . But uµ uµ = 1. so that energy conservation implies µ (174) Frad uµ = 0 . Still. 6π ds2 (173) where S is some scalar. Is there such a preacceleration? In other words. so µ Frad = duν duν q 2 d2 uµ + uµ 2 6π ds ds ds . This can be written as ˙ mv(t) = − exp(t/τ ) τ t C exp(−t′ /τ )Fext (t′ ) dt′ . 8. a signal propagating in the backward light-cone. is zero. Now. In particular.where C is the integration constant. in contrast to the retarded potentials discussed earlier. but not both. One can circumvent either the runaway solution or the acausal preacceleration. but at the microscopic scale: τ is of the order of 10−24 s.v. (176) This is known as the Dirac formula. we will be more comfortable with a theory which does not involve this aesthetically repelling component. (171) Thus. this would need. this is repugnant from a theoretical point of view. the preacceleration is significant only for a time of the order of τ . can an electron feel a force before it is applied? This is difficult to answer from an experimental point of view: such preacceleration is significant only for a very small time which is not accessible even to the present-day experiments. the particle starts accelerating! This violates causality. but it is there: even before the force is applied. and thus macroscopic causality is not violated. ds ds (175) 1 The first term. as uµ uµ = 1. acceleration at any time t depends not only on the instantaneous value of the external force but sort of its weighted time-average.

which is true upto O(ω0 τ 2 ): 1 α = Γ ± (ω0 + ∆ω) . let us just note that there is one real solution. ¨ ˙ (183) Assuming a solution x(t) = x0 exp(−αt). The cubic equation can be exactly solved. conjugate to each other. where the acceleration goes on increasing. One obtains. (180) (181) whose solution. as we expect from the Dirac theory of electrons. ¨ as Frad = mτ v . but their magnitudes differ from the classical counterparts. even larger than the line width (classically. The frequency has been shifted by ∆ω ((this is known as the level shift). the former is always smaller than the latter by a factor of ω0 τ ). one may wish to include the radiation reaction too for the equation of motion. which goes as x2 .8. One famous example is the shift of hydrogen 2S1/2 level. the existence of a nonzero level shift and a radiative broadening. show that α is indeed complex. the ¨ particle vibrates with a frequency ω0 . this should be degenerate with the 2P1/2 level. Find the decay width and the level shift if γ ≪ ω0 . we just quote the 2 result from Jackson.5 Radiation Reaction on a Charged Oscillator Consider a charged particle. This is reminiscent of a damping force that is proportional to x. who discovered it (with his student Retherford). The Lamb shift is entirely due to the radiative effects. from eq. x (178) (179) which can be written as 2 ¨ x + ω0 x − τ v = 0 . the solution would have been x(t) = x0 exp(iω0 t). to fall with an exponential pattern exp(−Γt). However. unobservable to experiments. 40 . the 2S1/2 level is about 1000 MHz lower than 2P1/2 . so that the equation of motion is 2 x + γ x + ω0 x = 0 . In the limit ω0 τ ≪ 1. Experimentally. 8 (182) 1 Note that the solution has an exponentially falling term exp(− 2 Γt) and an oscillatory term. 2 τ α3 + α2 + ω0 = 0 . The decay constant Γ causes the energy of the oscillator. Consider the equation of motion of a charged oscillator in the presence of a damping. 5 3 ∆ω = − ω0 τ 2 . Q. let us consider the one-dimensional case. 2 2 Γ = ω0 τ . The equation of motion is 2 m¨ + mω0 x − Frad = 0 . The qualitative results. for τ = 0. since both τ and ω0 are real. since ∆ω/ω0 ∼ (ω0 τ )2 ≪ 1. and is one of the cornerstones of quantum electrodynamics. In particular. (179). are there. this is known as the Lamb shift after Willis Lamb. One can perform a more sophisticated calculation in the ambit of quantum field theory. Classically. the so-called runaway solution. Without the last term. The remaining two solutions are complex. the level shift can be much larger. Just for simplicity. this should be the width of a spectral line. Let the solution be of the form x(t) = x0 exp(−αt) where α can be a complex number. the result will be similar whenever we have ˙ odd number of time derivatives of x. and hence the solution is unphysical. typically an electron. Since the charge is accelerating all the time. without any radiative corrections. this classical treatment is not applicable to electronic states in an atom since the spectral lines are quantum mechanical in origin. which is indeed very small. is the well-known α = ±iω0 . bound in a harmonic oscillator potential with a characteristic frequency ω0 .

(152): x= 1 q |E0 | 2 − ω 2 − iωγ exp(−iωt) .e. just because the oscillator vibrates with a large amplitude. The solution is similar to that of eq. so d3 x/dt3 = −ω 2 x. it has been flattened out to a Gaussian shape by the ωγ 2 term (note that in this limit. ¨ ¨ m The system oscillates with frequency ω. γ = ω 2 τ = ω0 τ = Γ of eq. The amplitude has a pole when the particle is real. i. it satisfies the energymomentum relationship pµ pµ = m2 . strongly bound electron).8. When ω ≈ ω0 . we can neglect the radiation reaction both in the Thomson limit (ω0 ≫ ω. m (185) where γ = ω 2 τ . This is just a reflection of the uncertainty principle: for a particle which is not stable. cannot be measured with absolute precision. for one-dimensional case. and ˙ 2 x + γ x + ω0 x = ¨ ˙ q |E0 | . (182)) 15 . ω0 . I can write q 2 (184) x + ω0 x − τ v = |E0 | exp(−iωt) . or the rest mass. 15 A similar term occurs in the quantum theory when we consider the amplitude for a particle to move from some point A to some other point B. the energy does not have a pole any longer. almost free electron) and in the Rayleigh limit (ω ≫ ω0 . m ω0 (186) Since γ ≪ ω. The form of the amplitude with such a pole-removing term is known as the Breit-Wigner form.6 Scattering and Absorption of Radiation by a Charged Oscillator Next consider a charge in a harmonic oscillator potential with characteristic frequency ω0 being driven by a driving force of frequency ω. This pole is removed by a similar term if the particle is unstable and can decay. Suppose this is an electric field.. 41 . This is the range where the dispersion is anomalous and the absorption is strong. and can take up lots of energy from the source. the energy.

Sign up to vote on this title
UsefulNot useful