Chapter Fourteen

Radiation by Moving Charges

Sir Joseph Larmor

(1857 - 1942)

September 17, 2001

Contents
1 Liénard-Wiechert potentials 1

2 Radiation from an Accelerated Charge; the Larmor Formula 9

2.1 Relativistic Larmor Formula . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.1 Example: Synchrotron . . . . . . . . . . . . . . . . . . . . . . 14
2.1.2 Example: Linear Acceleration . . . . . . . . . . . . . . . . . . 15

3 Angular distribution of radiation 17

3.1 Example: Parallel acceleration and velocity . . . . . . . . . . . . . . . 17
3.2 Example: Acceleration Perpendicular to Velocity . . . . . . . . . . . 20
3.3 Comparison of Examples . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.4 Radiation of an Ultrarelativistic Charged Particle . . . . . . . . . . . 22

4 Frequency Distribution of the Radiated Energy 25

4.1 Continuous Frequency Distribution . . . . . . . . . . . . . . . . . . . 26
4.2 Discrete Frequency Distribution . . . . . . . . . . . . . . . . . . . . . 30

1
 4.3 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.3.1 A Particle in Instantaneous Circular Motion . . . . . . . . . . 32
4.3.2 A Particle in Circular Motion . . . . . . . . . . . . . . . . . . 37

5 Thomson Scattering; Blue Sky 42

6 Cherenkov Radiation Revisited 46

7 Cherenkov Radiation; Transition Radiation 51

7.1 Cherenkov Radiation in a Dilute Collisionless Plasma . . . . . . . . . 55

8 Example Problems 57
8.1 A Relativistic Particle in a Capacitor . . . . . . . . . . . . . . . . . . 57
8.2 Relativistic Electrons at SLAC . . . . . . . . . . . . . . . . . . . . . . 58
We have already calculated the radiation produced by some known charge and
current distribution. Now we are going to do it again. This time, however, we shall
consider that the source is a single charge moving in some fairly arbitrary, possibly
relativistic, fashion. Here the methods of chapter 9, e.g., multipole expansions, are
impractical and there are better ways to approach the problem.

1 Liénard-Wiechert potentials
The current and charge densities produced by a charge e in motion are

ρ(x, t) = eδ(x − x(t))

J(x, t) = ev(t)δ(x − x(t)) (1)

if x(t) is the position of the particle at time t and v(t) ≡ dx(t)/dt ≡ ẋ(t) is its
velocity. In four-vector notation,

J µ (x, t) = ecβ µ δ(x − x(t)) (2)

2
 where β µ ≡ (1, β) is not a four-vector; β = v/c.
From J and ρ, one finds A and Φ. We can do this in an infinite space by making
use of the retarded Green’s function G(x, t; x0 , t0 ) = δ(t − t0 − |x − x0 |/c)/|x − x0 |.
From here, we can evaluate the electromagnetic field as appropriate derivatives of
the potentials. All of these manipulations are straightforward. Furthermore, the
integrations are relatively easy because there are many delta functions. The problem
becomes interesting and unfamiliar, however, for highly relativistic particles which
produce large retardation effects.
Let us start from the integral expression for the potentials:
Z
Aµ (x, t) = d3 x0 dt0 G(x, t; x0 , t0 )J µ (x0 , t0 )
1 Z 3 0 0 δ(t − t0 − |x − x0 |/c)J µ (x0 , t0 )
= d x dt
c |x − x0 |
Z
δ(x − x(t0 ))δ(t − t0 − |x − x0 |/c)
0
= e d3 x0 dt0 β µ (t0 )
|x − x0 |
Z
δ(t − t − |x − x(t0 )|/c)
0
= e dt0 β µ (t0 ) (3)
|x − x(t0 )|
This form reflects the retarded nature of the problem.

The particle is not seen by the observer at its

present location x(t), but it appears to be at
x(t’) since light traveling from the particle x
at x(t’) arrives at the observer at time t. O
n (t’)
e

x (t’)
x(t)

3
The evaluation of this integral is not totally simple because the argument of the δ-
function is not a simple function of the time t0 . In general when one faces an integral
of this form, one invokes the rule
Z ,¯ ¯
¯ dg ¯
dt f (t )δ[g(t )] = f (t0 ) ¯¯ 0 ¯¯¯
0 0 0 ¯
(4)
dt t0

where t0 is the zero (there may be more than one) of g 1 , i.e., g(t0 ) = 0. Applying this
to the present case, we have

g(t0 ) = t0 + |x − x(t0 )|/c − t = t0 + [(x − x(t0 )) · (x − x(t0 ))]1/2 /c − t, (5)

so
à !
dg 1 dx(t0 ) (x − x(t0 )) v(t0 ) · (x − x(t0 ))
0
= 1 + −2 0
· 0
= 1 − 0
≡ 1 − β(t0 ) · n(t0 ). (6)
dt 2c dt |x − x(t )| c|x − x(t )|

The unit vector n points from the point x(t0 ) on the particle’s path toward the field
point x; it, and β, must be evaluated at a time t0 which is earlier than t, the time
at which the field is evaluated, by some amount which is determined by solving the
equation
t0 + |x − x(t0 )|/c = t. (7)

Combining Eqs. (3), (4), and (6) we find that the potentials are given simply by
" # " #
µ βµ eβ µ
A (x, t) = e = (8)
|x − x(t0 )|(1 − β · n) ret
Rκ ret

where R ≡ |x − x(t0 )|, κ = 1 − β · n, and the subscript ret means that the quantity
in brackets [...] must be evaluated at the retarded time t0 determined from Eq. (7).
Our potentials, Eq. (8), are known as the Liénard-Wieckert potentials. Probably
their most significant feature is the fact that they vary inversely as 1 − β · n or κ; this
factor can be very close to zero for n k β if β is close to one, i.e., for highly relativistic
P
1
More generally δ[g(x)] = j δ(x − xj )/|g 0 (xj )| where xj are the simple zeroes of g(x). If
g 0 (xj ) = 0 (a complex zero), then δ[g(x)] makes no sense.

4
particles, meaning that there is a strong maximum in the potentials produced by a
relativistic particle in the direction of the particle’s velocity (at some retarded time).
We wish next to find the electromagnetic field. One may do this in a variety of
ways. One is simply carefully to take derivatives of the Liénard-Wieckert potentials,
a procedure followed in, e.g., Landau and Lifshitz’ book, The Classical Theory of
Fields. Another, much more elegant, is to find the fields in the instantaneous rest
frame of the particle and to Lorentz-transform them to the frame interest. Another,
not very elegant at all, and about to be employed here, is to go back to the integrals,
Eq. (3), for the potentials and take derivatives of these expressions. Consider just the
vector potential,
Z
β(t0 )δ(t0 + R(t0 )/c − t)
0
A(x, t) = e dt . (9)
R(t0 )
If we wish to find B(x, t), we have to take derivatives of R with respect to various
components of x. Consider, for example,

df df R df
∇f (R) = ∇R = =n . (10)
dR dR R dR

Application of this simple rule gives (since ∇ × (ψa) = ∇ψ × a + ψ∇ × a, and

∇ × β(t0 ) = 0 due to the lack of an x dependence.)
Z Ã !
δ(t0 + R/c − t)
0
B(x, t) = ∇ × A(x, t) = e dt ∇ × β(t0 )
R
Z · ¸
0 1 0 1 0 0
= e dt (n × β) − 2 δ(t + R/c − t) + δ (t + R/c − t) (11)
R cR

where the prime on the delta function denotes differentiation with respect to the
argument. Hence
(" # Z Ã ! )
β×n 0 κ δ 0 (t0 + R/c − t)
B(x, t) = e + dt (n × β) (12)
κR2 ret
cR κ

Now, from Eqs. (5) and (6)

d(t0 + R/c − t)
=κ or κdt0 = d(t0 + R/c − t) (13)
dt0

5
and so
(" # Z Ã ! )
β×n 0 n×β 0 0
B(x, t) = e + d(t + R/c − t) δ (t + R/c − t)
κR2 ret
cκR
(" # Z )
β×n 0 ∂[(n × β)/cκR] 0
= e − d(t + R/c − t) δ(t + R/c − t)
κR2 ret
∂(t0 + R/c − t)
(" # " Ã !# )
β×n 1 ∂ n×β
= e −
κR2 ret
κ ∂t0 cRκ ret
(" # " Ã !# )
β×n 1 1 ∂ β×n
= e + . (14)
κR2 ret
c κ ∂t0 κR ret

The electric field can be found by similar manipulations:

1 ∂A(x, t)
E(x, t) = −∇Φ(x, t) −
cà ∂t !
Z
0 d δ(t0 + R/c − t) e Z 0 βδ 0 (t0 + R/c − t)
= −e dt n + dt
dR R c R
Z ( )
0
0 δ(t + R/c − t) 1 0 0
= e dt n + (β − n)δ (t + R/c − t)
R2 cR
· ¸ " Ã !#
n e 1 ∂ β−n
= e − . (15)
κR2 ret c κ ∂t0 κR ret

We now have expressions for E and B, but they involve time derivatives of retarded
quantities. We can work out each of these derivatives. First, we consider just the
derivative of n = R/R. One has
∂R
0
= −ẋ(t0 ) = −βc (16)
∂t
and
∂R ∂[(x − x(t0 )) · (x − x(t0 ))]1/2 1
0
= 0
= [−2(x − x(t0 )) · ẋ(t0 )] = −n · ẋ(t0 ) = −n · βc
∂t ∂t 2R
(17)
Hence,
1 dn 1 ∂R 1 ∂R 1
0
= 0
− 2 R 0 = − [β − (n · β)n] . (18)
c dt cR ∂t cR ∂t R
The quantity in brackets is can be written more concisely:

(n · β)n − β = (n · β)n − (n · n)β = n × (n × β); (19)

6
hence we find
1 dn 1
0
= n × (n × β). (20)
c dt R
Employing this useful result we have, for the electric field,
" µ ¶ Ã !#
n 1 n ∂ 1 1 ∂ β
E(x, t) = e + 2 2 n × (n × β) + −
κR 2 κR cκ ∂t0 κR cκ ∂t0 κR ret
" #
n(1 − n · β) 1
= e 2 2
+ 2 2 [(n · β)n − β]
κR κR ret
" µ ¶ Ã !#
n ∂ 1 1 ∂ β
+ − (21)
cκ ∂t0 κR cκ ∂t0 κR ret

or " µ ¶ Ã !#
n−β n ∂ 1 1 ∂ β
E(x, t) = e + − . (22)
κ2 R 2 cκ ∂t0 κR cκ ∂t0 κR ret
This expression may be related to that for the magnetic induction,
" # " Ã !#
β×n e 1 ∂ β×n
B(x, t) = e +
κR2 ret
c κ ∂t0 κR ret
" # " Ã ! #
β×n e 1 ∂ β
= e + ×n
κR2 ret
c κ ∂t0 κR ret
" µ ¶#
β 1
+e 2
× n × (n × β)
κR R
" # " Ã !ret #
β×n e 1 ∂ β
= e + ×n
κR2 ret c κ ∂t0 κR ret
· ¸
1
+e 2 2 [n ((n × β) · β) − (n × β)(n · β)] (23)
κR ret

Next, put all terms proportional to n × β over the same denominator, which is κ 2 R2 ,
and also make use of the fact that (n × β) · β = 0, to find
" # " Ã ! #
β×n e 1 ∂ β
B(x, t) = e + ×n (24)
κ2 R 2 ret
c κ ∂t0 κR ret

This is our final expression. Comparing it with Eq. (22) for E(x, t), we can see that

B(x, t) = [n]ret × E(x, t). (25)

7
Thus we can find B(x, t) quite easily provided we can find E(x, t). Henceforth, we
won’t spend more time on B but will examine only E from which B then follows
trivially.
To have an explicit expression for E with no time derivatives, we need to evaluate
µ ¶ " Ã ! #
∂ 1 1 ∂β c
= − 2 2 κ(−n · βc) + R −n · 0 − Rβ · [(n · β)n − β]
∂t0 κR κR ∂t R
· ¸
c R
= − 2 2 (1 − n · β)(−n · β) − (n · β̇) + β · [β − (n · β)n]
κR c
· ¸
c R
= − 2 2 β 2 − n · β − (n · β̇) (26)
κR c

and so
" µ ¶µ ¶ #
n−β n −c 2 R β̇
E(x, t) = e 2 2
+ β − n · β − n · β̇ − 2
κR cκ κ2 R2 c cκ R ret
" µ ¶µ ¶#
β −c R
−e β2 − n · β − n · β̇
cκ κ2 R2 c ret
" µ ¶ µ ¶ #
1 R β R 1
= e 3 2 n 1 − β 2 + n · β̇ + 3 2 −1 + β 2 − n · β̇ − 2 β̇
κR c κR c cκ R ret
" # ·
(n − β)(1 − β ) 2
1 ³ ´¸
= e +e (n − β)(n · β̇) − (1 − n · β)β̇ (27)
.
κ3 R 2 ret
cκ3 R ret

The second bracket in the final expression contains the quantity

(n − β)(n · β̇) − n · (n − β)β̇ = n × [(n − β) × β̇], (28)

so " # " #
(n − β)(1 − β 2 ) e n × [(n − β) × β̇]
E(x, t) = e + . (29)
κ3 R 2 ret
c κ3 R ret

If there is no acceleration, then β̇ = 0 and only the first term in E and the
corresponding term in B are finite. These terms fall off with distance as 1/R 2 , and
hence cannot give rise to a net flux of radiation to infinity. If there is an acceleration,
then the second term of E, and the corresponding term in B are finite. These fall off
as 1/R, and hence will give rise to radiation, meaning that the charged particle will

8
emit radiation only if it is accelerated. Hence the two terms are interpreted as the
non-radiation and radiation parts respectively.
From these results when β̇ = 0, we should be able to recover the results for
the fields of a uniformly moving charge which we derived in chapter 11 (Jackson
Eq. 11.152). With the particle moving along the x-direction with a constant velocity
v, the fields felt by an observer a distance b away are

Ek = −γqvt/[b2 + (γvt)2 ]3/2

(30)
E⊥ = γqb/[b2 + (γvt)2 ]3/2 .

where the observer and particle are closest at time t = 0. It is a reasonably simple
task to show that this is the same result as Eq. (29) above when β̇ = 0. Consider the
diagram below in which the path of a charged particle is along the abscissa. At time
t, the retarded and real location of the particle are P’ and P respectively, while O is
the observation point. The time required for light to travel from P’ to the observer
O is t = R/c, in which time the particle travels βR =PP’. Thus the distance P’Q is
βR cos θ = β · nR. From this it follows that the distance OQ is R − β · nR. Then
[(1 − β · n)R]2 = r2 − (P Q)2 = r2 − β 2 R2 sin2 (θ). Thus, as R sin θ = b
1 ³ 2 ´
[(1 − β · n)R]2 = b2 + v 2 t2 − β 2 b2 = b + γ 2 2 2
v t , (31)
γ2

and transverse component of Eq. (29) when β̇ = 0 is

" #
(R sin θ) ebγ
e 2 = (32)
γ (1 − β · n)3 R3 ret
[b2 + (γvt)2 ]3/2

O
R=OP’
)
- β. n b
(1
Q r
n v θ
P’ Rβ P -vt M

9
2 Radiation from an Accelerated Charge; the Lar-
mor Formula
Even given Eqs. (25) and (29), the problem of calculating the fields emitted from a
charge moving along an arbitrary trajectory is non-trivial. This is largely due to the
effect of retardation. The problem is greatly simplified if the particle is not moving
too fast.
In order to consider this limit, let’s consider the trajectory of a particle shown
below.

x(t) R
x

a
We will assume that the origin is located in the center of the region of interest of
the particles trajectory, and that this region of interest is of linear dimension a (an
example would be an electron bound to a classical atom, where a would be the Bohr
radius).
There are two ways in which the problem simplifies in the nonrelativistic limit,
β << 1. First, we may approximate

κ≈1 n−β ≈n 1 − β2 ≈ 1 . (33)

Second, and much more importantly, we can approximate functions of the retarded
time
f (t − |x − x(t0 )|/c) (34)

10
by making a Taylor series expansion around the origin

f (t − |x − x(t0 )|/c) ≈ f (t − r/c) + x(t0 ) · ∇f (t − r/c) + · · · (35)

where r = |x|. The ratio of the second term to the first is roughly f 0 x(t0 )/f c, where
the prime means differentiation of f with respect to t − r/c. If this is small, as it
must be for β ¿ 1, then the second term in the series may be neglected, then

f (t − |x − x(t0 )|/c) ≈ f (t − r/c) . (36)

This approximation is sometimes called the dipole approximation (we will see why
directly). Here the effects of retardation are simple but not insignificant. They are
simply due to the separation of the charge from the observer, not due to the motion of
the charge. Thus, the only way they are significant is if the acceleration of the charge
2
changes abruptly and it takes a while for the emitted fields to reach the observer.
With the approximations described above, the electric field becomes
· ¸ " #
n e n × (n × β̇)
Enr (x, t) = e 2 + . (37)
R ret c R ret

The Poynting vector is given by

c c
S = (E × B) = E × ([n]ret × E)
4π 4π
c
= {[n]ret E · E − E([n]ret · E)} . (38)
4π
If we take the limit of large R and keep just the radiation terms, then we find that
" # · ¸
e n(n · β̇) − β̇ e
E(x, t) = ; B(x, t) = [n]ret × E = − n × β̇ (39)
c R ret
cR ret

Note that n · E = n · B = 0 for these radiation fields, thus (as usual) the radiation is
transverse. The Poynting vector due to the radiation is then given by
¯· ¯2 ¸
ce2 ¯ ¯
S= 2 2
n ¯n × (n × β̇) ¯ (40)
4πR c ret
2
For example, suppose a particle starts to oscillate back and forth at time t = 0, there will be no
signal felt at the observer’s location x for what may be a very long time t = r/c.

11
The angular distribution of radiated power is

dP
= R2 (S · n) (41)
dΩ

which is easily evaluated by expanding the cross products. The result is

dP e2 2 2 e2
= (β̇ sin θ) = (v̇2 sin2 θ) (42)
dΩ 4πc 4πc3

where θ is the angle between n and β̇ (at the retarded time).

.
β

In these last expressions we have dropped the reminder that n and R must be evalu-
ated at the retarded time.
The radiation pattern is characteristic of dipole radiation with the reference z
axis parallel to the direction of the particle’s acceleration. The shape and magnitude
of the distribution are independent of the particle’s velocity and proportional to the
square of the acceleration. The total power radiated is the integral over all directions
of the distribution,
Z
dP 2 e2 v̇2
P= dΩ = (43)
dΩ 3 c3
which is known as the Larmor formula for the power radiated by an accelerated
particle.

12
2.1 Relativistic Larmor Formula
We turn now to the relativistic generalization of the Larmor formula. One can de-
termine this generalization by consideration of how power transforms under Lorentz
transformations. There are actually two ways (that I know of) to do this, of which
Jackson does one. One can also find it from direct computation and that is what we
shall do. The first step is to consider just how we shall define the power radiated by
a particle. The point is that the rate at which energy crosses a closed surface sur-
rounding the particle depends on that surface because of retardation. We are going
to calculate the power as a function not of the time t at which the fields are measured
on the surface but rather as a function of the retarded time t0 . Consider a surface
S which encloses the particle at all times during which it is radiating. The power
crossing unit area at x on S at time t is S(x, t) · n, where n is a unit outward normal,
and so the total energy crossing this unit area is
Z ∞
W = dt S(x, t) · n. (44)
−∞

Now let us transform to the retarded time t0 :

Z ∞ Z ∞
dt 0 0
W = dt 0 S(x, t(t )) · n = dt0 κ[S(x, t(t0 )) · n]. (45)
−∞ dt −∞

The integrand of this expression we identify as dW/dt0 , the rate at which the par-
ticle radiates what eventually passes through the unit area on S at x. This is the
instantaneous radiated power, and if we multiply it by R 2 we get dP(t0 )/dΩ:

dP(t0 )
= κR2 S · n. (46)
dΩ

If we suppose that R is large enough that only the radiation fields need be retained,
then we find, from our results for the fields and the definition of the Poynting vector,
that ¯
dP(t0 ) e2 [n × ((n − β) × β̇)]2 ¯¯
= ¯
¯0
(47)
dΩ 4πc (1 − n · β)5 t

13
 Of course, if one wants to know the radiated power per unit area at the time that it
actually gets there, then he will have to address the calculation of the retardation. In
other words, the radiated intensity at time t depends upon the behavior of the particle
at t0 , and the differential time elements are different as well: dt = dt0 (1 − n · β)ret .
For example, if a particle of velocity β is impulsively accelerated for a time τ , and
then brought to rest, a pulse of radiation will appear at the observer at time t = r/c,
of duration τ (1 − n · β)ret . The total energy lost of the particle, of course, must
eventually equal the energy radiated, but the energy lost by the charge per unit time
will differ from the energy radiated per unit time by the factor (1 − n · β) ret
However, by calculating the instantaneous power radiated, we have avoided having
to calculate the retardation. Thus we can proceed with a calculation of the energy
lost. Our result, Eq. (47), depends in a complicated way on both β and β̇ making
it possible for rather remarkable angular distributions to occur. The total power
radiated can also be computed by integrating the distribution over directions:
e2 Z dΩ
P(t0 ) = [n × ((n − β) × β̇)]2
4πc κ5
½ i2 ¾
e2 Z dΩ h
= (n − β)(n · β̇) − β̇(1 − β · n)
4πc κ5
e2 Z dΩ n
= (1 − 2n · β + β 2 )(n · β̇)2
4πc κ5 o
− 2(n · β̇ − β · β̇)(β̇ · n)(1 − n · β) + β̇ 2 (1 − 2β · n + (β · n)2 ) (48)

Let the angle between β and n be θ; further, let cos θ = u; also, let the angle between
β and β̇ be θ0 . Then
½ · ¸
e2 β̇ 2 Z 1 du 1
P(t0 ) = (β 2
− 1) u 2
cos 2
θ 0 + (1 − u2 ) sin2 θ0
2c −1 κ5 2 o
+2 cos2 θ0 uβ(1 − βu) + (1 − 2βu + β 2 u2 )
β̇ 2 e2 Z 1 du n
= −u2 (1 − β 2 ) + 2uβ(1 − βu) + 1 − 2βu + β 2 u2
2c −1 (1 − βu)5
h io
+ sin2 θ0 (1 − β 2 )u2 − (1 − u2 )(1 − β 2 )/2 − 2βu + 2β 2 u2 (49)
½ ·µ ¶ ¸¾
β̇ 2 e2 Z 1 du 3 1 2 2 1
= (1 − u2 ) + sin2 θ0 + β u − 2βu − (1 − β 2 ) .
2c −1 (1 − βu)5 2 2 2

14
Now introduce x ≡ 1 − βu; then

0 β̇ 2 e2 Z 1+β dx n 2
P(t ) = 3 5
(β − 1 + 2x − x2 )
2cβ 1−β x
h³ ´³ ´ ³ ´ io
+ sin2 θ0 3 + β2 1 − 2x + x2 /2 − 2β 2 (1 − x) − β 2 1 − β 2 /2
β̇ 2 e2 Z 1+β dx n³ 2 2
´
= β − 1 + 2x − x
2cβ 3 1−β x5
h³ ´ ³ ´ io
+ sin2 θ0 3 − 4β 2 + β 4 /2 + (−2 + β 2 )x + 3 + β 2 x2 /2
(Ã !
β̇ 2 e2 β2 − 1 2 1
= + −
2cβ 3 4x4 3x3 2x2
à !)
2 (3/2 − β 2 /2)(1 − β 2 ) β 2 − 3 3 + β 2
+ sin θ0 + +
4x4 3x3 4x2
β̇ 2 e2 γ 6 n³ 3 3 2
´
= −(β + β ) + 2(3β + β )/3 − β(1 − β )
cβ 3
h io
+ sin2 θ0 (3 − β 2 )(β + 3β 3 )/2 + (−3 + β 2 )(β + β 3 /3) + (3 + β 2 )(β − β 3 )/2
½ ¾
β̇ 2 e2 γ 6 2 3 2 5 2 2β̇ 2 e2 γ 6
= β − β sin θ0 = (1 − β 2 sin2 θ0 ). (50)
cβ 3 3 3 3c

Putting the result back in terms of vectors and their products, we have

2e2 γ 6 2
P= [β̇ − (β × β̇)2 ]. (51)
3c

This is the relativistic generalization of the Larmor formula.

It is instructive to look at some simple examples.

2.1.1 Example: Synchrotron

An electron moves in a circle of radius R at constant speed.

15
The acceleration is then entirely centripetal and has magnitude β̇ = cβ 2 /R. Then

c2 β 4 c2 β 4
β̇ 2 − (β × β̇)2 = (1 − β 2
) = , (52)
R2 R2 γ 2

and so
2e2 cγ 4 β 4
P= (53)
3R2
which can be rewritten in terms of the particle’s kinetic energy as
µ ¶4
2e2 c E
P= β 4. (54)
3R2 mc2

The energy emitted per cycle of the motion is ∆E = Pτ where τ is the period of the
motion, τ = 2πR/βc. Hence,
µ ¶4
4πe2 β 3 E
∆E = . (55)
3R mc2

An electron of energy E = 500 M ev in a synchrotron of radius R = 102 cm will radiate

in each cycle and energy ∆E ≈ 10−8 erg ∼ 104 ev which is a non-trivial amount if
one wants to increase the electron’s energy or even to maintain it. Basically, one
must apply an accelerating voltage of at least ten thousand volts during each cycle to
break even, i.e., to maintain the electron’s energy. This radiation is the reason why
circular very-high-energy electron accelerators don’t exist; however, the synchrotron
is a great Xray source.
But if one wants to produce high-energy protons in a circular accelerator, that is
a lot easier, especially if one is willing to make R rather large. For example, a 10 T ev
proton in a machine of radius R = 3 × 106 cm, which is about nineteen miles, radiates
away considerably less than 104 ev in one cycle.

2.1.2 Example: Linear Acceleration

An electron is accelerated in the direction of its velocity, β k β̇.

16
 e
. β
β

In this instance one trivially finds from Eq. (51) that

2e2 γ 6 2
P= β̇ . (56)
3c

This doesn’t look very encouraging (for an accelerator design) because of the factor of
γ 6 , but against this one has the fact that the only acceleration the particle feels is the
one produced by the fields acting to increase the particle’s kinetic energy; in the case
of the round accelerator, there is a large acceleration even for a particle of constant
energy. Thus, in the case of a linear accelerator, there is much less acceleration to
produce radiation. To make this point more clearly, let’s write the acceleration in
terms of the time rate of change of the particle’s momentum,

dv 1 dp
v̇ = = (57)
dt mγ 3 dt

and so à !2
2e2 dp
P= . (58)
3m2 c3 dt
Now relate the rate of change of momentum to the rate of change of energy of the
particle,
dp dE
=F = , (59)
dt dx
and so à !2
2e2 dE
P= (60)
3m2 c3 dx
or
P 2e2 dE/dx 2 (e2 /mc2 ) dE
= = , (61)
dE/dt 3m2 c3 dx/dt 3β mc2 dx

17
which says that the power radiated away is quite negligible in comparison with the
rate at which energy is being pumped into the particle unless one pumps the energy
in at a rate (in space) dE/dx comparable to the rest energy of the particle, mc 2 ,
in a distance rc = e2 /mc2 . This distance is rather small, being about 3 × 10−13 cm
for an electron, and it is difficult to put so much energy into the particle in such a
small distance (500, 000 volts in a distance of 10−13 cm!). Hence particles in linear
accelerators lose an insignificant amount of energy to radiation. The difficulty with
such accelerators is that, like all things excepting round ones, they must sooner or
later end.

3 Angular distribution of radiation

In this section we consider the angular distribution of the radiation emitted by an
accelerated charge. On the basis of what we said about the potentials, we expect to
find that it is strongly focussed in the forward direction, or parallel to the velocity,
in a frame where the particle has a speed close to c. As at the end of the preceding
section, we shall consider some particular examples. The basic equation is
n h io2
dP e 2 n × (n − β) × β̇
= ; (62)
dΩ 4πc (1 − n · β)5

the right-hand side of this expression is to be evaluated at the retarded time t 0 .

3.1 Example: Parallel acceleration and velocity

Let a particle have parallel velocity and acceleration. If these define the z direction,
and n points at an angle θ to the z axis,

18
 n
e θ
. β
β
then one finds

dP e2 β̇ 2 sin2 θ e2 v̇2 sin2 θ

= = , (63)
dΩ 4πc (1 − β cos θ)5 4πc3 (1 − β cos θ)5

and the total power is

2e2 v̇2 6
P= γ
3c3
β=0.05
β=0.5
dP/dΩ

For β close to one, this expression provides dramatic confirmation of the inadequacy
of the multipole expansion for describing radiation from accelerated charges. This
radiation pattern is sharply peaked close to the z axis and is actually zero precisely
on the axis. One would have to keep many multipole terms to produce such a distri-
bution.
We can find various basic properties which characterize the distribution. It has a

19
maximum at a value of θ or of u ≡ cos θ which we can find by differentiating:
à !
d 1 − u2 2u 5β(1 − u2 )
0= =− + , (64)
du (1 − βu)5 (1 − βu)5 (1 − βu)6
or
−2u(1 − βu) + 5β(1 − u2 ) = 0. (65)

This quadratic equation has the solution

" #
1 q 1 q
u= [ 1 + 15β 2 − 1] or θm = arccos ( 1 + 15β 2 − 1) . (66)
3β 3β
Further, β 2 = 1 − 1/γ 2 , so for β close to one, β ≈ 1 − 1/2γ 2 and
( Ã !" Ã ! #)
1 1 15
θm ≈ arccos 1+ 2 4 1− −1
3 2γ 32γ 2
à !
1 1
≈ arccos 1 − 2 ≈ . (67)
8γ 2γ
Also, the power distribution becomes, in the relativistic limit,
0.10
dP/dΩ

0.00
-2.0 -1.0 0.0 1.0 2.0
γθ
dP e2 v̇ 2 θ2 e2 v̇ 2 32θ2
= =
dΩ 4πc3 (1 − β + βθ 2 /2)5 4πc3 [2(1 − β) + βθ 2 ]5
e2 v̇ 2 32γ 5 θ2 8e2 v̇ 2 3 γ 2 θ2
= = γ (68)
4πc3 (1 + γ 2 θ2 )5 πc3 (1 + γ 2 θ2 )5
This may be integrated to find the rms angle
1 mc2
< θ2 >= = (69)
γ E

20
3.2 Example: Acceleration Perpendicular to Velocity
z
β
θ n

. y
β φ
x
A particle in instantaneously circular motion has its acceleration perpendicular to its
velocity. Let β = β²3 and β̇ = β̇²1 . Then, letting θ and φ specify the direction from
the particle (at time t0 ) to the observation point x, we have

²3 = cos θ n − sin θ θ and ²1 = sin θ cos φ n + cos θ cos φ θ − sin φ φ. (70)

Using these conventions we can work out the relevant vector products:

(n − β) × β̇ = β̇[cos θ cos φ φ + sin φ θ − β(sin θ sin φn + cos θ sin φθ + cos φφ)] (71)

and so

[n × ((n − β) × β̇)]2 = β̇ 2 [cos2 φ(β − cos θ)2 + sin2 φ(1 − β cos θ)2 ]
= β̇ 2 [β 2 cos2 φ − 2β cos θ cos2 φ + cos2 θ cos2 φ + sin2 φ
−2β cos θ sin2 φ + β 2 cos2 θ sin2 φ]
= β̇ 2 [cos2 φ(β 2 + cos2 θ − 1 − β 2 cos2 θ) + (1 − β cos θ)2 ] (72).

Hence
( )
dP e2 v̇ 2 cos2 φ(1 − β 2 ) sin2 θ 1
= 1 −
dΩ 4πc3 (1 − β cos θ)2 (1 − β cos θ)3
( )
e2 v̇ 2 cos2 φ sin2 θ
= 1− 2 . (73)
4πc3 (1 − β cos θ)3 γ (1 − β cos θ)2

21
In this case we find that there is radiation in the forward direction, or θ = 0. In
fact the main peak in the radiation is in this direction. If one makes a small angle
expansion of the power distribution in the case of highly relativistic particles, he will
find the result ( )
dP 2e2 v̇ 2 γ6 4γ 2 θ2 cos2 φ
= 1 − . (74)
dΩ πc3 (1 + γ 2 θ2 )3 (1 + γ 2 θ2 )2

β = 0.5

β = 0.05

Further, the total radiated power is

2e2 v̇ 2 4
P= γ . (75)
4πc3

3.3 Comparison of Examples

From the expressions for the total power produced in each of our two examples, we
see that for a given magnitude of acceleration, there is a factor of γ 2 more radiation
produced when the acceleration is parallel to the velocity than when it is perpendic-
ular. This is a misleading statement in some sense because what is actually applied

22
to a particle is not an acceleration but a force, and a given force will produce quite
different accelerations when applied perpendicular and parallel to the velocity. For a
force applied parallel to the velocity,

dp dv
F = = mγ 3 (76)
dt dt

and for a force applied perpendicular to v,

dp dv
F = = mγ . (77)
dt dt

Hence the powers produced in the two cases, expressed in terms of the forces or dp/dt,
are à !2 à !2
2e2 2 dp 2e2 dp
P⊥ = γ and Pk = . (78)
3m2 c3 dt 3m2 c3 dt
Thus, for a given applied force, γ 2 more radiation is produced if it is perpendicular
to v than if it is parallel. Of course, if the force comes from a magnetic field, then it
has to be perpendicular to the velocity.

3.4 Radiation of an Ultrarelativistic Charged Particle

The radiation emitted at any instant from an accelerating charged particle may be
decomposed into components coming from the parallel and perpendicular accelera-
tions of the particle. From the discussion above, it is clear that the radiation of an
ultrarelativistic γ À 1 particle is dominated by the perpendicular acceleration com-
ponent. Thus the radiation is approximately the same as that emitted by a particle
moving instantaneously in a circle of radius

v2
ρ= (79)
v̇⊥

Since the angular width of the pulse is ∼ 1/γ, the particle will travel a distance

ρ
d ∼ ρ∆θ ∼ (80)
γ

23
while illuminating the observer for a time
d ρ
∆t ∼ ∼ . (81)
v γv
If we assume that pattern of radiation is roughly that of a coiled beam of rectangular
cross-section then the front edge of the beam will travel

ρc ρ
D = c∆t ∼ = (82)
γv γβ
in time ∆t; whereas, the trailing edge will be a distance
à ! à !
β − β2 ρ 1 1 ρ ρ
L=D−d≈ ≈ 1− 2 −1+ 2 = 3 (83)
β2 γ 2γ γ γ 2γ
behind the front edge since the charge moves the distance d in the same time interval.
Thus the pulse width is roughly L in space or L/c in time.
This derivation (straight out of Jackson) raises about as many questions as it
answers. Perhaps the most fundamental is this: why is ∆t different than the observed
width of the pulse L/c. The difference is that ∆t is really a difference of retarded
times; whereas, L/c is in the time frame of the observer. Lets repeat this calculation
more carefully, distinguishing between the observer’s time and the particles (retarded)
time.

24
 The pulse width for a distant observer would be
L
= t2 − t1 = (t02 + R2 /c) − (t01 + R1 /c) = ∆t − (R1 − R2 ) /c . (84)
c

Point 1: Observer first begins to see pulse

Point2: Observer just stops seeing pulse.

1 2

Since R1 − R2 ≈ v 0 ∆t, we have

L ∆t ρ
= t2 − t1 = ∆t(1 − v/c) ≈ 2 = 3 , . (85)
c 2γ 2γ
where the right hand side is to be evaluated in the retarded time.
An observer would receive periodic pulses of width L/c in time.

P(t) T0 = L0 /c

T= L/c

t
From the principles of Fourier decomposition, we can also estimate the type of radia-
tion the observer would receive. From the uncertainty principle ∆t∆ω ∼ 1, the pulse
would contain components up to a cutoff3 of roughly
c c
ωc ∼ ∼ γ3 (86)
L ρ
3
Higher frequency components would yield a narrower pulse

25
However, the frequency ω0 of the circular motion is v/ρ ≈ c/ρ. Thus, a broad
spectrum of radiation is emitted up to γ 3 times the fundamental frequency of the
rotation. This kind of radiation, called synchrotron radiation provides a good, more
or less continuous, source of radiation in the range from visible and ultraviolet to soft
X-rays.
Thus in a 200 MeV electron synchrotron where γ ≈ 400 and the fundamental
frequency is ω0 ≈ 3 × 108 , the frequency of the emitted radiation extends to 2 × 1016 .
In a 10 GeV synchrotron, x-rays can be produced. Thus synchrotrons can be used as
high intensity radiation sources. Synchrotrons are now even being used in industry
as radiation sources for x-ray lithography4 .

4 Frequency Distribution of the Radiated Energy

The radiation produced by rotating charges is predominantly at the fundamental
frequency ω0 of the motion with much smaller amounts at integral multiples, or
harmonics, of this frequency. The expansion parameter in the problem is k0 a =
ω0 a/c ∼ v/c = β, and the energy emitted at higher frequencies than the fundamental
is proportional to some power of this parameter. For β << 1, this energy will be
relatively small. But if β ∼ 1, there will be a significant fraction of the total radiated
energy appearing at higher frequencies.
There is a simple and instructive way to see roughly how may harmonics will
contribute to the radiation. For a relativistic particle, dP/dΩ is a peaked distribution
which has a width in angle δθ ∼ 1/γ. If the particle is, e.g., travelling in a circle
with a frequency of motion ω0 , then the particle sweeps through an angle δθ in a time
δt0 ∼ δθ/ω0 = 1/γω0 . This is of the order of the duration of the pulse observed at
some fixed point in space, but measured in units of the time at the source. The time
4
Physics Today, October 1991.

26
which passes at the location of the observer is
dt
δt = δt0 = κδt0 . (87)
dt0
Further, during the pulse, n is parallel to β, n · β = β with corrections of order 1/γ 2
and so κ ∼ 1 − β[1 + O(1/γ 2 )] ∼ 1/γ 2 . Hence δt ∼ 1/γ 3 ω0 . Now, a pulse which lasts
a time δt at a point must contain in it frequencies of order5 1/δt. Thus the pulse
which our observer sees must have in particular frequencies of order ω ∼ γ 3 ω0 . If the
particle is highly relativistic, γ >> 1, then the typical frequencies in the pulse will be
much larger than the frequency of the particle’s motion, or ω0 ; they will be γ 3 times
as large, meaning that many harmonics must contribute to the pulse.

4.1 Continuous Frequency Distribution

To make our analysis of the radiation more quantitative we are going to consider the
Fourier transforms in time of the fields at an observation point x. Start from the
expression for the angular distribution of radiated power:
dP c
= (E(x, t) × B(x, t)) · [nR2 ]ret . (88)
dΩ 4π
Using just the radiation fields, noting that they are transverse to n far away from the
source, and making use the fact that B(x, t) = [n]ret × E(x, t), we can write
dP c
= {[R]ret E(x, t)}2 . (89)
dΩ 4π
For simplicity of notation, define
r
c
a(t) ≡ [R]ret E(x, t). (90)
4π
Then, in terms of a(t) the angular distribution of radiated power is
dP
= [a(t)]2 . (91)
dΩ
5
In quantum theory this statement would be called the uncertainty principle.

27
Further, from Eq. (29),
s " #
e2 n × [(n − β) × β̇]
a(t) = . (92)
4πc κ3 ret

Introduce also the Fourier transform6 a(ω) of a(t):

1 Z 1 Z
a(t) ≡ √ dωe−iωt a(ω); a(ω) = √ dteiωt a(t). (93)
2π 2π
Using the Fourier integral (and its complex conjugate) for a(t) in Eq. (91), we can
write the power distribution as
dP(t) 1 Z 0
= dωdω 0 a(ω)e−iωt · a∗ (ω 0 )eiω t . (94)
dΩ 2π
Integrating over all t, and thereby generating an expression for the angular distribu-
tion of the total radiated energy, we find
Z
dW dP(t) Z
∞
≡ dt = dωdω 0 δ(ω − ω 0 )a(ω) · a∗ (ω 0 )
dΩ −∞
Z ∞
dΩ Z ∞ n o
2
= dω |a(ω)| = dω |a(ω)|2 + |a(−ω)|2 . (95)
−∞ 0

The reality of a(t) demands that a∗ (−ω) = a(ω), and so the two terms in the inte-
grand are identical:
Z ∞ Z ∞
dW 2 dI(ω)
=2 dω |a(ω)| ≡ dω . (96)
dΩ 0 0 dΩ
The integrand, dI/dΩ, is interpreted as the total radiation received per unit frequency
per unit solid angle during the entire pulse of radiation. It is simply the square of
a(ω); further a(ω) is
s " #
e2 1 Z iωt n × [(n − β) × β̇]
a(ω) = √ dt e
4πc 2π κ3 ret
s " #
e2 Z 0 n × [(n − β) × β̇] iω(t0 +R(t0 )/c)
= dt e (97)
8π 2 c κ2
6
The field a(t) depends on x as well as on t; we suppress the former dependence as it is not of
interest at present.

28
It’s déja vu all over again. Recalling the typical equations generated in Chapter 9,
we can see that this one probably can be obtained without great difficulty from what
we did there. Proceeding in familiar fashion, then, let us look at this integral in the
far zone which means approximate R in the exponent by

R = |x − x(t0 )| ≈ r − n · x(t0 ) (98)

and so s " #
e2 Z 0 iωt0 iωr/c −iωn·x(t0 )/c n × [(n − β) × β̇]
a(ω) = dt e e e (99)
8π 2 c κ2
or s " #
iωr/c e2 Z 0 iω(t0 −n·x(t0 )/c) n × [(n − β) × β̇]
a(ω) = e dt e . (100)
8π 2 c κ2

n R(t’) P
x(t’)
O

If we confine the integration to times when the particle is accelerating, i.e. confine
it to the shaded region above, then we can put this integral into a form such that the
integrand involves β but not β̇; that can be done by, in essence, a parts integration.
First, consider
à !
d n × (n × β) n × (n × β) n × (n × β̇)
=− (−n · β̇) + (1 − n · β) (101)
dt0 κ κ 2 κ2
where we have not kept derivatives of n because they give corrections of relative order
|x(t0 )/R|. Now group terms as follows:
à ! n o
d n × (n × β) n × (n × β̇) n × n × [(n · β̇)β − (n · β)β̇]
= +
dt0 κ κ2 κ2

29
 n o
n × (n × β̇) n × n × [n × (β × β̇)]
= +
κ2 κ2
n × [(n − β) × β̇]
= (102)
κ2

where we have used the fact that

n h io
n × n × n × (β × β̇) = −n × (β × β̇). (103)

Using this identity to do a parts integration of the expression for a(ω) we find (again,
drop terms proportional to powers of 1/R)
s à !
e2 Z 0 n × (n × β) iω(t0 −n·x(t0 )/c)
a(ω) = −eiωr/c 2
dt iω(1 − n · β) e
8π c κ
s
Z
e2 0 0
= − 2
iωe iωr/c
dt0 [n × (n × β)]eiω(t −n·x(t )/c) . (104)
8π c

Combining this result and Eq. (96), we find for the radiated energy per unit frequency
per unit solid angle
¯ ¯
dI(ω) e 2 ω 2 ¯Z 0 0 ¯2
= 2 ¯¯ dt0 [n × (n × β)]eiω(t −n·x(t )/c) ¯¯ . (105)
dΩ 4π c

In this derivation we were not careful when doing the parts integration, meaning
that we did not worry about whether the terms involving the integrand evaluated at
the endpoints of the interval of integration contribute to the result. Consequently,
in any application of Eq. (105), one should check to see whether this assumption is
justified; there are occasions when it is not and then, naturally, the contributions from
the endpoint(s) must be included. Our formulation of dI/dΩ is most appropriate for a
source with an open orbit in which case the natural limits on the integration are ±∞,
and there is usually no difficulty in ignoring the contributions from the endpoints
where the particle is far away and unaccelerated; however, that is not enough to
guarantee that the endpoints contribute nothing.

30
 4.2 Discrete Frequency Distribution
For truly cyclic motion, it is more convenient and perhaps more sensible from a
physical point of view to set up a Fourier series to express the frequency distribution
of the radiation.

t’= 0 t’ = τ

x P
x(t’)
O

The point is that for non-cyclic motion, the distribution of radiation in frequency will
be continuous and the integral formula we derived above is appropriate for expressing
this distribution. But for cyclic motion, the radiation will be distributed in frequency
space only at harmonics or multiples of the fundamental frequency of the motion
and so a sum or series expansion of dI(ω)/dΩ is more appropriate for expressing the
distribution. We shall now set up this sum. To get started, suppose that the period
of the motion is τ 0 = 2π/ω0 . Then, letting the period measured by an observer at a
point x be τ , one can show that τ = τ 0 . That is, a time t for the observer and the
corresponding retarded time t0 are related by

t = t0 + R(t0 )/c. (106)

One period later in the life of the source, its time has increased to t0 + τ 0 and the
signal emitted by the source at this time will reach the observer at a time t + τ which
is
t + τ = t0 + τ 0 + R(t0 + τ 0 )/c. (107)

But R(t0 + τ 0 ) = R(t0 ) for the cyclic motion so

t + τ = t0 + τ 0 + R(t0 )/c. (108)

31
Comparing Eqs. (106) and (108), we see that τ = τ 0 .
The radiation fields produced by this cyclic motion of a charged particle will also
be periodic. Hence where we formerly had a Fourier integral for a(t), we now have
a Fourier series, a sum over frequencies which are integral multiples of the sources’s
frequency, r
X∞
c
a(t) = [RE(t)] ≡ an e−inω0 t (109)
4π n=−∞

with
ω0 Z 2π/ω0
an = dt a(t)einω0 t . (110)
2π 0
The energy received during one cycle, per unit solid angle, at some point x is the
integral of dP/dΩ over one period. We shall write it as dW/dΩ,
Z 2π/ω0 Z 2π/ω0
dW dP(t)
≡ dt = dt |a(t)|2
dΩ 0 dΩ 0
XZ 2π/ω0 2π X ∞
= dt (an · a∗m )e−i(n−m)ω0 t = |an |2
n,m 0 ω0 n=−∞
X∞
4π
= |an |2 . (111)
ω0 n=1

Notice that the n = 0 term has been discarded in the final expression; that is okay
because there is no radiation at zero frequency. The radiated energy at the frequency
nω0 is determined by an , which is
" # s
ω0 Z 2π/ω0 n × [(n − β) × β̇] e2 inω0 t
an = dt e
2π 0 κ3 ret
4πc
s
ω0 e2 Z 2π/ω0 0 n × [(n − β) × β̇] inω0 (t0 +R(t0 )/c)
= dt e . (112)
2π 4πc 0 κ2

We can now do the same integration by parts that led to Eq. (104) and find
s
Z 2π/ω0
ω0 e2 0 0
an = − inω0 dt0 [n × (n × β)]einω0 (t +R(t )/c) , (113)
2π 4πc 0

with no contribution ever from the end points of the interval as they correspond to
the same point on the periodic orbit of the particle. From Eq. (113), |an |2 follows

32
easily and hence dW/dΩ from Eq. (111). Notice also that the time-averaged power
distribution is just à ! Áµ ¶
dP dW 2π
= (114)
dΩ ave
dΩ ω0

4.3 Examples
4.3.1 A Particle in Instantaneous Circular Motion

Suppose we have a particle in instantaneous circular motion, meaning that its accel-
eration is, at least temporarily, perpendicular to its velocity. Any particle subjected
to a magnetic field but no electric field will satisfy this condition. We might do the
calculation by supposing that the motion is truly periodic and circular and using the
relatively easily applied Fourier series approach just developed. It is more difficult
and therefore more challenging to use the Fourier integral approach. Let’s try the
latter.
First, we want to characterize the orbit of the particle. A circular orbit at constant
speed can be described as

β = β[(cos ω0 t0 )²1 + (sin ω0 t0 )²2 ] (115)

and, for sufficiently small times t0 , meaning ω0 t0 << 1, we have

β ≈ β(²1 + ω0 t0 ²2 ) (116)

with corrections of order (ω0 t0 )2 . Let the observer be located in the x-z plane; then
n = cos θ ²1 + sin θ ²3 where θ << 1 if the observer is to experience the strong pulse
of radiation that the particle emits in the forward direction. We know that the times
of importance at the source for this pulse at the position of the observer are of order
t0 ∼ 1/ω0 γ. Consequently, in our approximation for β, we lose corrections of relative
order (ω0 t0 )2 ∼ 1/γ 2 in each of the components.
Let’s work out n × (n × β). Define a unit vector ²⊥ ≡ n × ²2 = cos θ ²3 − sin θ ²1 ;
it will prove to be useful.

33
 n
ε

ε
θ
v ρ vt/ρ
n × β = β[(n × ²1 ) + ω0 t0 (n × ²2 )] = β(sin θ ²2 + ω0 t0 ²⊥ ); (117)

and
n × (n × β) = β(sin θ ²⊥ − ω0 t0 ²2 ). (118)

Further, the interesting range of θ is of order 1/γ << 1, so let us replace sin θ by θ.
Further, let ²2 be designated ²k . Then, using this equation in Eq. (105) and setting
nω0 = ω, we find
¯ ¯
dI(ω) e2 ω 2 ¯¯Z 0 0 0 ¯2
= 2 ¯ dt (βθ²⊥ − ω0 t0 ²k )eiω(t −n·x(t )/c) ¯¯ . (119)
dΩ 4π c
We need to evaluate the exponent in order to complete the integral. Consider
cβ
n · x(t0 ) = cos θ (²1 · x(t0 )) =cos θ sin ω0 t0
ω0
à !µ ¶
cβ θ2 1
≈ 1− ω0 t0 − (ω0 t0 )3 ; (120)
ω0 2 6
Notice that we have kept the leading term and corrections to it of order 1/γ 2 . Basi-
cally, we have kept all phases of order unity when ω0 t0 ∼ 1/γ, θ ∼ 1/γ, and ω ∼ ω0 γ 3 ,
these being what we believe to be the important ranges of t0 , ω, and θ. Hence the
total phase is, to the order indicated,
à ! à !
0 0 ωβ 0 θ2 ωβ 1 θ2
ω[t − n · x(t )/c] = ωt − ω0 t0 1 − + (ω0 t0 )3 1 −
ω0 2 ω0 6 2
à ! à !
2 2
βθ ωβ θ
= ωt0 1 − β + + (ω0 t0 )3 1 −
2 6ω0 2
à !
1 θ2 ω
≈ ωt0 2
+ + (ω0 t0 )3
2γ 2 6ω0

34
 µ ¶" Ã ! #
ω 0 1 1
= ω0 t 2
+ θ2 + (ω0 t0 )3 . (121)
2ω0 γ 3

With the foregoing expression for the phase, we are able to write the frequency
and angle distribution of the intensity as
¯ h ³ ´ i ¯2
dI(ω) e2 ω 2 ¯¯Z ∞ 0 i ω ω0 t 0 1
+θ 2 + 13 (ω0 t0 )3 ¯
¯
= 2 ¯¯ dt (θ²⊥ − ω0 t0 ²k )e 2ω0 γ2
¯ .
¯
(122)
dΩ 4π c −∞

The integration has been extended to ±∞ even though the integrand is accurate only
for |ω0 t0 | ∼ 1/γ (or less). The extended range of integration is sensible only if there is
no important contribution coming from other regimes of t0 . That is the case because
the term in the phase proportional to t03 produces rapid oscillations of the integrand
at larger |t0 | which yield a very small net contribution to the integral.
Introduce x such that ω0 t0 ≡ x(1/γ 2 +θ2 )1/2 . The important range of t0 corresponds
to |x| ∼ 1 and the frequency distribution of the intensity is given by
¯ s " s # ³ ´3/2 ¯2
¯ Z ∞ ¯
dI(ω) e2 ω 2 ¯ 1 1 1 ω
i 2ω 1
+θ 2 (x+ 31 x3 ) ¯
= 2 ¯¯ + θ 2 dx θ²⊥ − + θ 2 x²
k e 0 γ2 ¯ .
¯
dΩ 4π c ¯ ω0 γ2 −∞ γ2 ¯
(123)
Let à !3/2
ω 1
η= + θ2 . (124)
3ω0 γ2
Then
¯Z "s à ! # ¯2
dI(ω) e2 ω 2 ¯ ∞ 1 1 ¯
¯ 2 θ² + 2 i 3
η(x+x 3 /3) ¯
= ¯ dx + θ ⊥ + θ x² k e 2 ¯
dΩ 4π 2 c ω02 ¯ −∞ γ2 γ2 ¯
¯s à ! ¯2
e2 ω 2 ¯¯ 1 2 θI ² +
1 2
¯
¯
≡ 2
¯
2 ¯ 2
+ θ ⊥ ⊥ 2
+ θ I k ² k ¯ .
¯
(125)
4π c ω0 γ γ

where
Z ∞ 3 )/2
I⊥ = dt ei(3ηt+ηt (126)
−∞

and
Z ∞ 3 )/2
Ik = dt tei(3ηt+ηt . (127)
−∞

35
These integrals may be expressed in terms of Airy functions which are modified Bessel
functions of order 1/3 and 2/3. An integral representation7 of the function Ai is
Z ∞
π 1/3 3 1Z ∞ 3

1/3
Ai[x/(3a) ] = dt cos(xt + at ) = dt ei(xt+at ) . (128)
(3a) 0 2 −∞
From this representation one can see that
2π
I⊥ = Ai[(3η/2)2/3 ]. (129)
(3η/2)1/3
As for Ik , it is
Z ∞ Z ∞
¯
i(3ηt+ηt3 )/2
¯
i(xt+ηt3 /2) ¯
Ik = dt te = dt te ¯
−∞ −∞ x=3η/2
µZ ¶¯ à !¯
1 d ∞ ¯ 1 d 2π ¯
i(xt+ηt3 /2) ¯
= dte ¯
¯ = Ai[x/(3η/2)1/3 ] ¯
¯
i dx −∞ x=3η/2 i dx (3η/2)1/3 x=3η/2
2π
= Ai0 [(3η/2)2/3 ]. (130)
i(3η/2)2/3
The prime on the Airy function denotes differentiation with respect to the argument.
The connection between Airy functions and modified Bessel functions is8
" #1/2
2/3 1 (3η/2)2/3
Ai[(3η/2) ]= K1/3 (η). (131)
π 3
Also9 ,
1 (3η/2)2/3
−Ai0 [(3η/2)2/3 ] =
√ K2/3 (η). (132)
π 3
Thus we may express the result for the frequency distribution of intensity in terms of
modified Bessel functions as
µ ¶2 ¯¯s à ! ¯2
¯
dI(ω) e2 ω ¯ 1 2
2θ √ K
1 2 2 ¯
= ¯ + θ 1/3 (η)² ⊥ + + θ √ K 2/3 (η)² k ¯
dΩ 2
4π c ω0 ¯ γ2 3 γ 2
i 3 ¯
µ ¶ Ã ! 2 " #
e2 ω 2 1 2 2 θ2 2
= 2 2
+θ K2/3 (η) + 2 2
K1/3 (η)
3π c ω0 γ 1/γ + θ
µ ¶ " #
3e2 γ 2 ω 2 2 2 2 2 γ 2 θ2 2
= (1 + γ θ ) K2/3 (η) + K (η) (133)
π 2 c ωc 1 + γ 2 θ2 1/3
7
Abramowitz and Stegun, 10.4.32.
8
Abramowitz and Stegun, 10.4.14.
9
Abramowitz and Stegun, 10.4.16.

36
with à !3/2
3 ω 1 ω
ωc = 3γ ω0 and η = 2
+ θ2 = (1 + θ2 γ 2 )3/2 . (134)
3ω0 γ ωc
This is a fairly transparent, if not simple, result. The variable η is proportional
to ω and is scaled by 3γ 3 ω0 ≡ ωc which we believe, on the basis of arguments given
earlier, to be the appropriate or characteristic scale of frequency in this radiating
system. For η << 1, Ã !ν
Γ(ν) 2
Kν (η) ∼ (135)
2 η
and for η >> 1, s
π −η
Kν (η) ∼ e . (136)
2η
Thus, for η << 1 and at θ = 0,
µ ¶2 Ã !2 Ã !4/3 Ã !2 µ ¶1/3 µ ¶2/3
dI(ω) 3e2 ω Γ(2/3) 2 e2 Γ(2/3) 3 ω
= 2 γ2 = .
dΩ π c ωc 2 ω/ωc c π 4 ω0
(137)
2
If θ 6= 0, we pick up a contribution proportional to K1/3 , leading to an additional
term which is proportional to ω 4/3 . Note also that the term we do have is produced
by waves with the electric field polarized in the plane of the particle’s orbit.
In the large frequency regime, ω >> ωc and θ = 0, we find
µ ¶2
dI(ω) e2 ω π e2 ω −2ω/3γ 3 ω0
= 2 3γ 2 e−2ω/ωc = e . (138)
dΩ π c ωc 2ω/ωc 2πcγ ωc

The accompanying figure shows the angular distribution of radiation at several values

37
 0.3 ω/ωc=3.0
ω/ωc=0.8
ω/ωc=0.1
0.2

dI/dΩ
0.1

0.0
0.0 0.5 1.0 1.5
of ω/ωc . γθ
This figure also very clearly makes the point that at a given frequency, except for the
very low ones, the intensity is greatest at the smallest angles.

4.3.2 A Particle in Circular Motion

Let’s do the same sort of calculation for truly circular motion. Then, from Eqs. (111)
and (113), we find that the energy emitted per cycle per unit solid angle is
X∞ ∞
dW dWn 4π X
= = |an |2 (139)
dΩ n=0 dΩ ω 0 n=1

with s ¯ ¯
e2 ¯Z 2π/ω0 ¯
2¯ 0 inω0 (t0 −n·x(t0 )/c) ¯
|an | = nω 0 ¯
¯ 0
dt (n × β)e ¯.
¯
(140)
16π 3 c
An individual term in the sum, dWn /dΩ, is the energy per cycle per unit solid angle
radiated at frequency ωn = nω0 . It can be written as
¯Z ¯2
dWn 4π e2 n2 ω04 ¯ 2π/ω0 ¯
¯ 0 inω0 (t0 −n·x(t0 )/c) ¯
= ¯
¯ 0
dt (n × β)e ¯ .
¯
(141)
dΩ ω0 16π 3 c

If we divide by the period, 2π/ω0 , we find the time-averaged power at frequency nω0
per unit solid angle,
¯Z ¯2
dPn e2 n2 ω04 ¯ 2π/ω0 ¯
¯ 0 inω0 (t −n·x(t )/c) ¯
0 0
= ¯
¯ 0
dt (n × β)e ¯ .
¯
(142)
dΩ 8π 3 c

38
Let the motion be in the x-y plane as before,

x(t0 ) = a[²1 cos(ω0 t0 ) + ²2 sin(ω0 t0 )] (143)

α = π/2 − θ
θ
n

ε2
a
x(t’)
ε1

so that
ω0 a
β(t0 ) = [−²1 sin(ω0 t0 ) + ²2 cos(ω0 t0 )]. (144)
c
Also, write n in terms of the usual10 spherical coordinates,

n = ²3 cos θ + ²1 sin θ cos φ + ²2 sin θ sin φ. (145)

Then

n · x(t0 ) = a[sin θ cos φ cos(ω0 t0 ) + sin θ sin φ sin(ω0 t0 )] = a sin θ cos(ω0 t0 − φ), (146)

and

aω0
n × β(t0 ) = [−²2 cos θ sin(ω0 t0 ) − ²1 cos θ cos(ω0 t0 )
c
+²3 sin θ(cos φ cos ω0 t0 + sin φ sin ω0 t0 )]
aω0
= [−²2 cos θ sin(ω0 t0 ) − ²1 cos θ cos(ω0 t0 ) + ²3 sin θ cos(ω0 t0 − φ)].
(147)
c
10
But be aware that the angle θ in this example is the polar angle and not the latitude as in the
previous example; the angle α introduced below is the latitude, i.e., the same as the θ of the previous
example.

39
We must do the integral
Z 2π/ω0 0 0 β
I= dt0 (n × β)einω0 (t −n·x(t )/c) ≡ [−²1 K cos θ − ²2 J cos θ + ²3 L0 sin θ] (148)
0 ω0
where ω0 t0 = y, and
Z 2π
K = dy cos(y + φ)ei(ny−nβ sin θ cos y) einφ
0
Z 2π
J = dy sin(y + φ)ei(ny−nβ sin θ cos y) einφ
0
Z 2π
0
L = dy cos yei(ny−nβ sin θ cos y) einφ (149)
0

with β = aω0 /c. Thus

   
 J 
  Z 2π  sin(y + φ) 
 

= dy  einφ einy e−inβ sin θ cos y
 K 
  cos(y + φ) 
0 
    
Z 2π  cos y 
  
 sin y 

= einφ dy 
sin φ

+ cos φ  i(ny−nβ sin θ cos y)
e
0  − sin y  
 cos y 

    
 L 
   M 
 
inφ  
≡ e sin φ  
+ cos φ  
 (150)
 M   L 

where
Z 2π
M = dy sin y einy−inβ sin θ cos y
0
Z 2π
1 d ³ −inβ sin θ cos y ´
= dy einy e
inβ sin θ 0 dy
1 Z 2π 2π Jn (nβ sin θ)
= − dy einy e−inβ sin θ cos y = − , (151)
β sin θ 0 β sin θ in
and
Z 2π
L = dy cos y einy e−inβ sin θ cos y
0
Z 2π
1 d
= dy ei(ny−nβ sin θ cos y)
−i d(nβ sin θ) 0
2π 1 dJn (nβ sin θ)
= − (152)
i in d(nβ sin θ)

40
Hence

β2 h 2 i
|I|2 = cos θ (|K| 2
+ |J| 2
) + sin 2
θ |L| 2
ω02
 Ã !2 Ã !2 
4π 2 β 2  2 Jn (nβ sin θ) dJn (nβ sin θ)
= 2
cos θ + (cos2 θ + sin2 θ) (153)
.
ω0 β sin θ d(nβ sin θ)

Also, Ã 
!2
dPn e2 n2 β 2 ω02  dJn (nβ sin θ) cot2 θ
= + (Jn (nβ sin θ))2  . (154)
dΩ 2πc d(nβ sin θ) β2
For better comparison of this result with things we already know, let us look at
some limiting cases. First, in the nonrelativistic limit, β << 1, we expect that only
the n = 1 term will contribute appreciably and, in addition, we can use the small
argument approximation to the Bessel function,

J1 (x) ≈ x/2 and J10 (x) ≈ 1/2 (155)

so that à !
dP1 e2 β 2 ω02 1 cot2 θ sin2 θ e2 ω04 a2
= + = (1 + cos2 θ) (156)
dΩ 2πc 4 4 8πc3
which may be compared with the result of a completely nonrelativistic calculation as
in, e.g., Jackson, Problem 14.2(b). In the highly relativistic limit, on the other hand,
we have β ≈ 1 and we also know that most of the radiation is close to the equatorial
plane or θ ≈ π/2. Let us introduce α ≡ π/2 − θ << 1. Then
Ã !2 ¯
¯
dPn e2 n2 β 2 ω02  dJn (x) tan2 α 2 ¯¯
= + (J n (x)) ¯ . (157)
dΩ 2πc dx β2 ¯
x=nβ cos α

The argument of the Bessel function Jn is comparable to n but is always less than n.
For this particular range of argument, it is the case11 that
s à √ !
1 2(n − x) 2 2(n − x)3/2
Jn (x) = K1/3 √ , (158)
π 3x 3 x
11
See, e.g., Watson, Bessel Functions, p. 249.

41
or √
1 y 1/3 2 2(n − x)3/2
Jn (x) = K1/3 (y) where y= √ . (159)
π 31/6 x1/3 3 x
Also, Ã !
dJn (x) 1 d y 1/3 K1/3 (y)
= 1/6 . (160)
dx π3 dx x1/3
As shown in the figure below, for the interesting values of x, K1/3 (y) varies much
more than x.
20

15 n=1
n=2
n=3
K1/3(y)

10

0
0 1 2 3
x

Thus, the preceding derivative may be approximated by

dJn (x) 1 d[y 1/3 K1/3 (y)] dy

≈
dx π31/6 x1/3 dy dx
1/3 1/3
√
y K2/3 (y) dy y K2/3 (y) 2(n − x)1/2
= − = (2x + n). (161)
π31/6 x1/3 dx 31/6 x1/3 3x3/2
Hence
" #
dPn e2 β 2 n2 ω02 y 2/3 n2 sin2 α 2 2(n − x)(2x + n)2 2
= K 1/3 (y) + K2/3 (y) . (162)
dΩ 2πcπ 2 31/3 x2/3 x2 9x3

Now let’s see what can be said about y and x. Expanding in powers of α and 1/γ,
we have à !3/2 à !
n 1 1 α2
y≈ + α2 and x≈n 1− 2 − , (163)
3 γ2 2γ 2

42
so à !2 " #
dPn e2 ω02 n2 1 α2 γ 2
= + α2 2
K2/3 (y) + K 2 (y) . (164)
dΩ 6π 3 c γ2 1 + α2 γ 2 1/3
This term describes radiation at the particular frequency ω = nω0 , so we can also
write it as
" #
dPn ω 2 e2 (ω 2 /ω02 ) α2 γ 2
= 0 (1 + α 2 2 2 2
γ ) γ K 2
2/3 (y) + K 2 (y) (165)
dΩ 2π 3π 2 cγ 6 1 + α2 γ 2 1/3

with
ω 2 2 3/2 ω
y= 3
(1 + α γ ) ≡ (1 + α2 γ 2 )3/2 . (166)
3ω0 γ ωc
By grouping terms appropriately in Eq. (165) we can write
µ ¶2 " #
dPn ω 2 3e2 ω α2 γ 2
= 0 2 2 2 2 2
(1 + α γ ) γ 2
K2/3 (y) + 2 2
2
K1/3 (y) . (167)
dΩ 2π π c ωc 1+α γ

This is the time-averaged power received at latitude α, at frequency ω = nω0 , per unit
solid angle. The energy/cycle received at this frequency is obtained if we multiply by
2π/ω0 , and, finally, the energy per unit frequency is found it we multiply by a factor
of the inverse spacing of the harmonics, or 1/ω0 . Hence we conclude that for a single
pulse, meaning one cycle of the particle,
µ ¶2 " #
dI(ω) 2π dPn 3e2 ω 2 2 2 2 2 α2 γ 2
= 2 = 2 (1 + α γ ) γ K2/3 (y) + K 2 (y) (168)
dΩ ω0 dΩ π c ω0 1 + α2 γ 2 1/3

which is precisely what we found when we calculated the radiated intensity from a
particle in instantaneous circular motion.

5 Thomson Scattering; Blue Sky

It’s time for a change of pace to something with less nineteenth-century analysis.
Thomson scattering provides just such a respite. It is the scattering of radiation by
a free charge. The mechanism involved, classically, is the coupling of the charge to
the incident electric field E0 ; this produces acceleration of the charge and consequent

43
radiation by the accelerated charge which becomes the “scattered” radiation. We saw
in Chapter 9 how one can describe scattering of electromagnetic radiation in this way.
We didn’t actually do Thomson scattering at that time, so we’ll do it now.
Thomson scattering is to be distinguished from Compton scattering in which the
same phenomenon is treated using quantum theory. One must use quantum theory,
when λ, the wavelength of the radiation, is comparable to the Compton wavelength of
the scatterer, λ = 2πc/ω ∼ h̄/mc ≡ λc ; λc is the Compton wavelength. This condition
may also be written as h̄ω ∼ 2πmc2 which says that if the photon energy h̄ω (∼ eV) is
comparable to the particle’s rest energy mc2 (∼ MeV), then the calculation must be
done using quantum theory. For, e.g., visible light the photon energy is much smaller
than an electron’s rest energy so the Thomson scattering, or classical, calculation is
quite adequate.
Assuming a nonrelativistic particle, we can make use of the Larmor formula for
the scattered radiation,

dP e2 2 2
= a sin Θ where Θ = 6 (n, a) (169)
dΩ 4πc3

for the instantaneous radiated power. To find the acceleration a of the particle, we
must solve the equation of motion of the scatterer. The force on it is provided by the
incident electric field,
E0 = E0 ²ei(k·x−ωt) (170)

where for convenience we shall let k = k²3 and assume linear polarization of the
incident plane wave ² = cos ψ ²1 + sin ψ ²2 .

44
 ε3
θ n
k
e ε2

ψ ε a
ε1
Similarly, we shall suppose for simplicity that the charged particle has no velocity
component along the direction of k and shall ignore the magnetic force which is
reasonable so long as the particle is nonrelativistic. Letting the particle be located in
the z = 0 plane, we find that it experiences an incident field

E0 = ²E0 e−iωt ; (171)

given that it has a mass m and charge e, the force acting on it is F = eE0 = ma, so

eE0 −iωt
a=² e , (172)
m

and
eE0
a·n= sin θ(cos ψ cos φ + sin ψ sin φ)e−iωt . (173)
m
where θ and φ specify, in spherical coordinates, the direction to the observation point.
The time average of the power emitted in some particular direction will be pro-
portional to

e2 E02
< a2 sin2 Θ >=< a2 (1 − cos2 Θ) >=< a2 − (a · n)2 >= [1 − sin2 θ cos2 (ψ − φ)]
2m2
(174)
where the brackets < ... > denote a time average. If the incident radiation is not
polarized, then we must average over ψ with the result that
µ ¶
2 2 e2 E02 1 2 e2 E02
< a sin Θ >= 2
1 − sin θ = 2
(1 + cos2 θ) (175)
2m 2 4m

45
and
dP e4
= |E0 |2 (1 + cos2 θ) (176)
dΩ 16πc3 m2
is the time-averaged power distribution when the incident wave is unpolarized.
We define the scattering cross-section in the usual way, i.e., the time-averaged
power per unit solid angle divided by the time-averaged incident power per unit area,
à !2
dσ dP/dΩ 1 e2
= 2
= (1 + cos2 θ). (177)
dΩ (c/8π)|E0 | 2 mc2
This is J. J. Thomson’s formula for the scattering of light by a charged particle. The
total cross-section is
Z Ã ! Ã !2 Ã !2
dσ 1 e2 8π e2 8π 2
σ= dΩ = 4π(1 + 1/3) = ≡ r (178)
dΩ 2 mc2 3 mc2 3 c

where rc is the classical radius of the particle. For an electron it is ∼ 3 × 10−13 cm

and σ is about 0.7 × 10−24 cm2 .
We may also calculate the scattering of radiation by a bound charge using the
model of a damped harmonic oscillator. Let a charge e with mass m be bound at the
origin of coordinates with a natural frequency of oscillation ω0 and damping constant
Γ. Then, given an applied electric field

E0 (x, t) = ²E0 ei(k·x−ωt) (179)

with k = k²3 , we know from earlier calculations that the charge will respond with a
displacement
eE0 /m
x(t) = x0 e−iωt where x0 = ² (180)
ω02 − ω 2 − iωΓ
if we approximate E0 (x, t) by E0 (0, t) as is reasonable when the particle’s displacement
from the origin is small compared to the wavelength of the incident radiation. That
is certainly true for visible light and an atomic electron.
From x(t) it is a simple matter to compute the acceleration,
eE0 ω 2 /m
a = −² e−iωt , (181)
ω02 − ω 2 − iωΓ

46
 and then to find the radiated, or scattered, time- and incident-polarization-averaged
power per unit solid angle,
µ ¶2 Ã !
dP e2 eE0 1 + cos2 θ
= ; (182)
dΩ 16πc3 m (1 − ω02 /ω 2 )2 + Γ2 /ω 2
the scattering cross-section follows:
à !
dσ 1 1 + cos2 θ
= rc2 . (183)
dΩ 2 (1 − ω02 /ω 2 )2 + Γ2 /ω 2
For sufficiently large ω, ω >> ω0 , Γ, the cross-section reduces to the Thomson result
as it should since for very large ω the particle will respond to the field as though it
were a free particle. Also, for ω << ω0 and ω02 /Γ, we find the Rayleigh scattering
result, µ ¶
dσ 1 ω 4
= rc2 (1 + cos2 θ) (184)
dΩ 2 ω0
4
with the characteristic ω behavior indicating dipole scattering.
Why is the sky blue? Why is the light polarized when we look at the sky perpen-
dicular to the line of sight from us to the sun?
sun sky
ε3

ε1 ε2 n

6 Cherenkov Radiation Revisited

While studying the energy loss of a charged particle traversing a material, we derived
an expression for the rate of energy loss through Cherenkov radiation. Specifically,

47
 for a charge q moving at speed β through a medium with a real dielectric function
²(ω), we found that the energy loss per unit path length is
dE q2 Z 1
= 2 dω ω(1 − ). (185)
dx c ²β >1
2 ²(ω)β 2
The form of this expression suggests that the radiation per unit length of path per
unit frequency is given by the integrand,
à !
dI(ω) q2 1
=ω 2 1− . (186)
dx c ²(ω)β 2
What we did not do is look at the form of the potentials and fields in real space and
time. That is interesting and revealing, so we are going to do it now using retarded
potentials.
Before starting, lets be sure that we understand the physical mechanism giving
rise to the radiation. It is not the incident particle, which may be reasonably de-
scribed as having constant velocity, that is doing the radiating. Rather, the incident
particle produces fields which act on the particles in the medium, causing them to
be accelerated in various ways. They then produce radiation fields which, when the
incident particle moves more rapidly than the speed of light in the medium, but not
when it moves more slowly, add in a coherent fashion to give Cherenkov radiation.
No Radiation Cherenkov Radiation
vt

θc
v < cε
v > cε
ct ct
ε ε
vt
48
It is not easy to see why the fields must cancel when the particle is moving slower
than the speed of light, but it is easy to see why they must not when it is moving
faster than light in the medium. Consider the right hand side of the figure above.
Before the wake of the radiation hits a particle in the medium, it does not feel the
incident particle. Once the wake hits a particular particle in the medium, not only it,
but all of its neighbors accelerate in a direction perpendicular to the wake (depending
upon the relative charge). Clearly these particles will radiate coherently.
If we are going to produce a calculation of Cherenkov radiation, then, we have to
find the total field produced by all of the particles and not just the incident particle.
We in fact did that in Chapter 13 in the space of k and ω. Here we want to determine
appropriate Liénard-Wieckert potentials for this field so as to find it in real space and
time. That turns out not to be very hard if we ignore the frequency-dependence of
the dielectric function.
Consider the Fourier-transformed potentials A(k, ω) and Φ(k, ω) produced by the
incident charge in the medium where it is taken to have constant velocity. As we
have seen, these obey the equations
à !
ω2 ρ(k, ω)
k 2 − 2 ² Φ(k, ω) = 4π
c ²
à !
2
ω 4π
k 2 − 2 ² A(k, ω) = J(k, ω) (187)
c c
where ρ and J are the macroscopic sources, i.e., the charge and current density of the
incident particle. Given that ² is independent of ω, then these are the same as the
equations obeyed by the potentials of a fictitious system consisting of a point particle
√
with charge q/ ² moving at constant velocity v in a ‘vacuum’ where the speed of
√
light is c0 = c/ ². Let’s rewrite them in such a way as to see this correspondence
more clearly:
à !
2 ω2 √ ρ(k, ω)
k − 02 [ ²Φ(k, ω)] = 4π √
c ²
à !
ω2 4π J(k, ω)
k 2 − 02 A(k, ω) = 0 √ . (188)
c c ²

49
 Now let’s think about these potentials in real space and time. Because they are
√
the same as a particle with a renormalized charge q/ ² moving in a vacuum with a
renormalized speed of light is c0 , we can write them as Liénard-Wieckert potentials
using the renormalized charge and velocity of signal propagation,
· ¸
√ q 1
²Φ(x, t) = √ (189)
² κR ret

and similarly for A(x, t). In this case the retardation means that κR should be
evaluated at the time t0 = t − R(t0 )/c0 . Further, if x(t0 ) = vt0 , v = v²3 , and x =
z²3 + ρ²⊥ , then
R(t0 ) = (z − vt0 )²3 + ρ²⊥ . (190)

Also, for this system, κ = 1 − n · v/c0 . Notice that κ can be negative; the absolute
value of κ should be employed in evaluating the potential because the potential really
involves |κR| as one may see by going back to its derivation, especially Eq. (4).
ε
z

x ρ
X
R
α
ε3
x (t’) v
It is useful to introduce a vector X ≡ x − vt which is the relative displacement of
the observation point and the particle at time t. Then

R = x − vt0 = x − vt + v(t − t0 ) = X + v(t − t0 ), (191)

and
t − t0 = R/c0 = |X + v(t − t0 )|/c0 . (192)

50
Square this relation to find
à !
X2 v2 2X · v
(t − t ) = 02 + 02 (t − t0 )2 +
0 2
02
(t − t0 ). (193)
c c c
This is a quadratic equation that can be solved for t − t0 ; there are two solutions
which are q
−X · v ± (X · v)2 − (v 2 − c02 )X 2
t − t0 = . (194)
v 2 − c02
Acceptable solutions must be real and positive. Given that v 2 > c02 , which we know
to be the regime where there is Cherenkov radiation, we find that there are either no
such solutions or there are two of them. The conditions under which there are two
are
X·v <0 and (X · v)2 > (v 2 − c02 )X 2 . (195)

Let the angle between X and v be α. Then we require, first, that α be larger than
π/2 and, second, that X 2 v 2 cos2 α > (v 2 − c02 )X 2 or cos2 α > 1 − c02 /v 2 . Hence there
is a cutoff angle α0 given by
q
α0 = arccos(− 1 − c02 /v 2 ) (196)

such that for α < α0 there is no potential. There can thus be potentials and fields
at time t only within a cone whose apex is the current position of the particle and
which has an apex angle of π − α0 . Within this cone the potential is the sum of two
terms, Φ = Φ1 + Φ2 , corresponding to the two allowed values of t − t0 . Making use of
Eq. (191), and the fact that R k n, we see that we can write, for either case,

[κR]ret = |(1−n·v/c)·R| = R−v·R/c = |X+v(t−t0 )|−X·v/c0 −v 2 (t−t0 )/c0 . (197)

Using Eqs. (189) and (197) with Eq. (194) for t − t0 , one finds that Φ1 ≡ Φ2 and that
µ ¶
2q 1
Φ(x, t) = q . (198)
² X 1 − (v 2 /c02 ) sin2 α

A similar expression can be found for A(x, t) and with a bit more work one can
compute the radiated power, recovering the same equations as found in chapter 13

51
for the particular case that ²(ω) is a constant. Such a dielectric function never exists,
of course, and so our conclusions are flawed in some respects. One of them has to do
with the form of Φ close to α = α0 ; Eq. (198) indicates that it in fact diverges here.
That would indeed happen if signals with all frequencies traveled at speed c0 so that
such a singular non-dispersing wave front could be built by superposing waves with
many different wavelengths including ones approaching zero. In reality, there is no
such singularity although the amplitude of the wave does have a strong maximum at
the leading edge.

7 Cherenkov Radiation; Transition Radiation

This time we will do a calculation using perturbation theory much the way we did
scattering via perturbation theory. We will learn a little bit more about the character
of Cherenkov radiation and will also derive a new (to us) phenomenon. The basic
requirement for validity of the calculation is to have |²(ω) − 1| << 1 so that we
can get away with calculating the true macroscopic fields as a correction to the fields
produced by the incident particle in vacuum. We have shown that, when expressed as
a function of x and ω, the electric field of a particle with charge q moving at constant
velocity v = v²3 on a trajectory x(t) = vt is, in vacuum,
s
2 qω iωz0 /v
Ei (x0 , ω) = e [K1 (ωρ0 /γv)²ρ0 − (i/γ)K0 (ωρ0 /γv)²3 ]. (199)
π γv 2

52
 v
k
θ ε3
P d 3x’
q ε2

ε1 ερ

This field produces a polarization in the medium which is

²(ω) − 1
P(x0 ω) = Ei (x0 , ω) (200)
4π
and a dipole moment P(x0 , ω)d3 x0 in a volume element d3 x0 . From chapter 9 (Eq. (43)),
we know that such a harmonic dipole moment gives rise to a radiation field
eikR 2
dErad (x, ω) = k [n × {P(x0 , ω)d3 x0 }] × n. (201)
R
In the radiation zone we can expand R = |x − x0 | as R = r − n · x0 , leading to
à !
eikr ²(ω) − 1 2 Z 3 0 −ikn·x0
Erad (x, ω) = k dx e {[n × Ei (x0 , ω)] × n} (202)
r 4π V
à !
eikr ²(ω) − 1 2 Z 2 0 0 −ikx0 sin θ
Z
0 0
= k d x⊥ {[n × Ei (x⊥ , 0, ω)] × n}e dz 0 ei(ωz /v−kz cos θ)
r 4π
(203)
where we have specified k = k(²3 cos θ + ²1 sin θ) without loss of generality since the
q
fields are invariant under rotation around the direction of v. Also, k = ω ²(ω)/c =
ω/vp where vp is the phase velocity of a wave with frequency ω. If the medium is
infinite in the z direction, we can complete the integration over z 0 with ease and find
à ! Z
eikr ²(ω) − 1 2 0
Erad (x, ω) = k δ(ω/v − k cos θ) d2 x0⊥ {[n × Ei (x0⊥ , ω)] × n}eikx sin θ .
r 4π
(204)

53
 q
Notice that k and ω are related by k = ω ²(ω)/c = ω/vp so that we find no radiation
field unless cos θ = vp /v which can only happen if v > vp . We seem to be on the
right track. The implication is that there is radiation coming out of this system in a
direction n which makes an angle α0 with the z axis where α0 = arccos(vp /v). What
we have shown in chapter 13 is that this is the direction of the outcoming radiation,
which is Cherenkov radiation, is perpendicular to the “bow wave” shown in the figure
earlier.
We can get something new out of this calculation also. Suppose that v < vp and
that the medium is finite in extent in the z direction. If it begins at z = 0 and ends
at some arbitrary, large, value of z, then the integral over z 0 is not from −∞ to +∞
but rather starts from zero and will be non-zero even though v < vp .

z=0
The corresponding radiation, known as transition radiation, comes about because
of the presence of the boundary, in this case between vacuum and dielectric. The
integral we must do can be evaluated with the use of a convergence factor (which
emulates the damping present in most materials):
Z ∞ Z ∞
0 0
dz 0 eiz (ω/v−k cos θ) = lim dz 0 e[−η+i(ω/v−k cos θ)]z
0 η→0 0
à !
−1 i
= lim = , (205)
η→0 −η + i(ω/v − k cos θ) ω/v − k cos θ

54
and so
à !Z
eikr ²(ω) − 1 2 i 0
Erad (x, ω) = k d2 x0⊥ {[n × Ei (x0⊥ , ω)] × n}e−ikx sin θ .
r 4π ω/v − k cos θ
(206)
More generally, a slab of material of finite thickness d produces an integral over z 0
which is ³ ´
Z d i 1 − ei(ω/v−k cos θ)d
0
dz 0 ei(ω/v−k cos θ)z =
. (207)
0 ω/v − k cos θ
Whether we treat a finite slab or no, we have to do the integral over the other two
coordinates. The integrand involves (n × Ei ) × n. For a given n = k/k, Ei , and x0 ,
and noting that ²ρ0 = cos φ0 ²1 + sin φ0 ²2 , we have

(n × Ei ) × n = (Eρ cos φ0 cos θ − Ez sin θ)(²2 × n) + Eρ sin φ0 ²2 . (208)

The second term is an odd function of φ0 and will give zero when integrated over x0⊥ .
Thus we must do the integral
Z
0
I= d2 x0⊥ (n × Ei ) × ne−ikx sin θ
(209)
s
Z
2 qω 0
= 2
(² 2 × n) dx0 dy 0 [K1 (ωρ0 /γv) cos θ cos φ0 + (i/γ)K0 (ωρ0 /γv) sin θ]e−ikx sin θ
π γv
(210)
Now, cos φ0 = x0 /ρ0 and
x0 γv ∂K0 (ωρ0 /γv)
cos φ0 K1 (ωρ0 /γv) = K 1 (ωρ 0
/γv) = − . (211)
ρ0 ω ∂x0
Using this in the expression for I, we can do an integration by parts and find
s " #
Z
2 qω ikγv sin θ i
I = 2
(² 2 × n) dx0 dy 0 − K0 (ωρ0 /γv) cos θ + K0 (ωρ0 /γv) sin θ
π γv ω γ
s à !
Z
2q ω 0
= (−i sin θ) k cos θ − 2 (²2 × n) dx0 dy 0 K0 (ωρ0 /γv)e−ikx sin θ . (212)
πv γ v
The integral over x0 can be done using the identity
Z ∞ √ π √
−|t| α2 +β 2
dz K0 (β z 2 + t2 ) cos(αz) = √ 2 e . (213)
0 2 α + β2

55
 Thus,
s à ! q
ω Z ∞ πdy 0 exp(−|y 0 | k 2 sin θ + ω 2 /γ 2 v 2 )
2
2q
I = ²2 × n(−i sin θ) k cos θ − 2 q
πv γ v −∞ k 2 sin2 θ + ω 2 /γ 2 v 2
s à !
2q ω 2π
= ²2 × n(−i sin θ) k cos θ − 2 2 . (214)
πv γ v k2 sin θ + ω 2 /γ 2 v 2
Hence
√ Ã !
−1
ikr ²(ω) k 2 sin θ 2 2π(q/v)²2 × n ω
Erad = e k cos θ − 2 . (215)
4πr ω/v − k cos θ k 2 sin2 θ + ω 2 /γ 2 v 2 γ v

7.1 Cherenkov Radiation in a Dilute Collisionless Plasma

This is a general result. Let us look at some specific example in a simple limit in
order to extract the salient qualitative features of transition radiation. First, we shall
choose a dielectric function. A very simple one, which has the added virtue that for
some frequencies ² < 1 so that it is impossible to have Cherenkov radiation, is that
for a dilute collisionless plasma, ²(ω) = 1 − ωp2 /ω 2 .

v Plasma
q
ε(ω) = 1 − ω2p / ω2

For ω large enough, this function is less than, but close to, unity so that it satisfies
the criteria for validity of the perturbation theory. Then

(²(ω) − 1)k 2 /4π = −ωp2 k 2 /4πω 2 = −ωp2 /4πc2 (216)

where we approximate ω by ck wherever it is not important.12 Suppose also that

12
This can be done in some places because ²(ω) ≈ 1; in other places the difference ²(ω) − 1 is
needed and here we cannot set ²(ω) equal to 1.

56
γ >> 1, i.e., that the particle is highly relativistic. These conditions allow us to
approximate as follows:
à !à !
ω ω ω ω ω√ ω ω ωp2 θ2
−k cos θ = − cos θ = − ² cos θ ≈ − 1 − 1−
v v vp βc c c(1 − 1/2γ 2 ) c 2ω 2 2
à ! à ! (217)
µ ¶
ω 1 ωp2 θ2 ω γ 2 ωp2 2 2 ω 1
≈ + + = 2 1+ 2 +γ θ ≡ 2 1+ 2 +η (218)
c 2γ 2 2ω 2 2 2γ c ω 2γ c ν
where
s
q √ ωp2 ωp2
ν ≡ ω/γωp η ≡ γ 2 θ2 β = 1 − γ −2 ≈ 1−1/2γ 2 and ²= 1− ≈ 1− .
ω2 2ω 2
(219)
Also,

ω2 2 2 ω2 ω 2 ² sin2 θ ω2 2 2 2 ω2
+ k sin θ = + = [1 + γ β ² sin θ] ≈ (1 + η) (220)
γ 2v2 γ 2 β 2 c2 c2 γ 2 β 2 c2 γ 2 c2

while
à !
ω ωq ω ω q 1 ω
k cos θ − 2 = ²(ω) cos θ − = ²(ω) cos θ − ≈ . (221)
vγ c βcγ 2 c βγ 2 c

Hence,
à ! √
eikr −ωp2 θ(2 2π)(q/c)(ω/c)²2 × n
Erad =
r 4πc2 (ω/2γ 2 c)(1 + η + 1/ν 2 )(ω 2 /γ 2 c2 )(1 + η)
√
eikr ωp2 2γ 4 θ2 2πq²2 × nω
= −
r 4π cω(1 + η + 1/ν 2 )ω 2 (1 + η)
s √
eikr 2 γ η q²2 × n
= − . (222)
r π ν c(1 + η + 1/ν 2 )(1 + η)
2

The radiated energy per unit frequency per unit solid angle is13

d2 I(ω) c q2γ 2 η
= 2 |rErad (x, ω)|2 = 2 . (223)
dΩdω 4π π c ν (1 + η + 1/ν 2 )2 (1 + η)2
4

13
Previously, we just called this dI/dΩ.

57
We can write dΩ = sin θdθdφ ≈ θdθdφ = dηdφ/2γ 2 and integrate over φ to find the
distribution per unit η. Also, let’s write the frequency in terms of ν, dν = dω/γω p ,
to find
Z
d2 I d2 I dω dΩ d2 I γωp π q 2 γωp η
= dφ = 2
= . (224)
dηdν dΩdω dν dη dΩdω γ πc ν (1 + η + 1/ν 2 )2 (1 + η)2
4

This expression fails at small ν (ω ∼ ωp ). It falls off as η −3 at large η. It peaks as a

function of η around η = 1 for small ν and at η = 1/3 for large ν. If one integrates
over η the result is

dI q 2 γωp
= [(1 + 2ν 2 ) ln(1 + 1/ν 2 ) − 2]. (225)
dν πc

Further, the total energy radiated is

Z ∞ dI q 2 γωp (q/e)2 γh̄ωp e2 (q/e)2
I= dν = = = (γh̄ωp ). (226)
0 dν 3c 3 h̄c 3(137)

A typical photon energy is γh̄ωp /3 (ν = 1/3), so the number of photons emitted

on average is (q/e)2 /137 which is quite a bit smaller than one. However, a sizable
amount of transition radiation can be obtained by employing a stack of thin slabs of
material with adjacent slabs having significantly different dielectric constants. Then
there is some radiation produced at each interface between different materials.

8 Example Problems

8.1 A Relativistic Particle in a Capacitor

A particle of charge e and mass m initially at rest is accelerated across a parallel
plate capacitor held at (stat) voltage V ; the distance between the plates is d. Assum-
ing nonrelativistic motion, find the total energy radiated by the particle during this
process. Then, without calculation, answer or estimate the following:

1. The angular distribution of the radiated energy of the particle,

58
 2. the order of magnitude of typical frequencies emitted from the charge, and

3. the angular distribution of radiated energy from the charge if it were highly
relativistic.

Solution. The field in a parallel plate capacitor is constant, thus so is the accel-
eration of the charged particle and its radiation.
Z
2 e2 2 2 e2
P = a W = P dt = t
3 c3 3 c3

where t is the duration of the pulse.

q
a = F/m = eE/m = eV /md and t = 2d/a

Hence, solving for W √

2 2e7/2 V 3/2
W =
3c3 m3/2 d
In the non relativistic limit, the angular distribution of the radiation is given by
sin2 (θ) where the angle is measured relative to the velocity vector. The typical fre-
quencies can be found using the Fourier uncertainty principle. Since the retarded
duration of the pulse is the same as the observer’s duration for a nonrelativistic par-
q
ticle, we have ω ∼ 1/t = eV /md2 . In the relativistic limit the angular distribution
of the radiation will be strongly pitched in the direction of the velocity/acceleration,
but will be zero along the axis. The maximum of the pulse of radiation will be at an
angle of θ ∼ 1/γ away from the axis defined by the velocity vector.

8.2 Relativistic Electrons at SLAC

At the Stanford linear accelerator, devices have been added at the end of the accel-
erator to guide electrons and positrons around roughly semicircular paths until they
collide head-on as shown in the sketch below. If each particle has a total energy
of 50GeV and rest energy of 0.5 MeV, while the circular paths have radii of about

59
 1km, roughly what is the fraction of the particles energy lost to radiation before the
collision takes place?

e-

accelerator beams
- e+ collision
e area
1 km

e+

Solution. To determine the power radiated, we must use the relativistic Larmor
formula. For β̇ ⊥ β, |β × β̇| = |β β̇|, so the power radiated is
· ¸
2e2 6 2 ³ ´2 2e2 6 2 ³ ´ 2e2 4 2
P = γ β̇ − β × β̇ = γ β̇ 1 − β 2 = γ β̇
3c 3c 3c

The acceleration is centripetal, so β̇ = v 2 /cr ≈ c/r, so

2e2 4 c2
P = γ
3c r2

Counting the initial deflection of the particles into the circular region, both the elec-
tron and the positron travel about 3/4 of a circle. The time it takes to do this is
roughly τ = 3πr/2c (assuming that the particles travel roughly at velocity c), so the
energy radiated is
πe2 4
∆E = γ
r
Since γ = E/mc2 , the relative energy loss is

∆E πe2 /r 3
= γ
E mc2

60
For E = 50 GeV and mc2 = 0.511 MeV, γ ≈ 105 , and the relative energy loss is

∆E π23 (10−20 ) (1015 )

= ≈ 0.9 × 10−2
E 0.911 (10−27 ) (9) (1020 ) (105 )

or roughly only one percent of the energy is lost.

61