Lecture 9 - Applications of 4–vectors,

and some examples

E. Daw

April 4, 2011

1 Review of invariants and 4–vectors

Last time we learned the formulae for the total energy and the
momentum of a particle in terms of its rest mass m0 and its
velocity ~v ,

E = γm0 c2
(1)
p~ = γmo~v .

We also learned a key relationship between the total energy E,

the momentum p~ and the rest mass m0 ,

E 2 = |~p|2 c2 + m20 c4 . (2)

Finally we learned that we could construct 4–vectors out of the

components of the spacetime coordinates of an event,

(x0 , x1 , x2 , x3 ) = (ct, x, y, z), (3)

and out of the energy and momentum components,

E
(p0 , p1 , p2 , p3 ) = ( , px , py , pz ). (4)
c

1
We noticed that minus the square of the 0th component, plus
the sum of the squares of the rest of the components equals an
invariant quantity. So, for example, −c2 t2 + x2 + y 2 + z 2 is an in-
variant, where (ct, x, y, z) are the coordinates of some event with
respect to the origin. If c2 t2 > (x2 + y 2 + z 2 ), then this quantity
is equal to −τ 2 , where τ is the proper time interval between
an event at the origin and the event having these coordinates.
This means, physically, that there is some observer for whom
both the events occur in the same place. For this observer, the
spatial displacement between the events is x = 0, y = 0, z = 0,
and hence the only displacement is a time displacement. But
this is the very observer for whom the time displacement is the
proper time interval, and we have just re-discovered that as long
as the events are time-like separated, there is an observer that
can get between the two events without exceeding the speed of
light, who will measure the proper time interval, the minimum
time interval between the events that any observer can mea-
sure. Furthermore, the existence of this observer implies that
there can be a causal chain of events connecting the two events,
so all observers therefore agree on which of the two events occurs
first.

If c2 t2 < (x2 + y 2 + z 2 ), there is no observer that can travel

between the events since to do so he would need to exceed the
speed of light. Events separated in this way are called space-like
separated, and it is not true that all observers agree on which
order space like separated events occurred.

A second invariant is formed from −E 2 /c2 +(px )2 +(py )2 +(pz )2 ,

which is equal to −m20 c2 . In both cases, the invariant quantity
can be written in matrix form as

  0 
−1 0 0 0 v
 0 +1 0 0   v1
  
( u0 u1 u2 u3 )  . (5)
 0 0 +1 0   v 2 
0 0 0 +1 v3

The square matrix in the middle on the right is known as the

Minkowski metric. Notice that as long as the minus sign is in
the right position, the two 4–vectors separated by the metric
need not be the same! So, for example, if you have a particle of
energy E, momentum (px , py , pz ), and position (x, y, z) at time
t, then −Et + xpx + ypy + zpz is an invariant quantity. You can

2
prove this to yourself by Lorentz transforming the components
of both position and momentum 4–vectors and verifying that
the transformation matrices multiply out to give a result that
depends neither on β nore on γ.

2 Applications of 4–momentum: con-

servation of the square of four-momentum

Now let us flex our new found muscles and see what can be done
with 4–vectors. One of the key applications of 4–vectors is new
conservation laws using invariants. Energy and momentum are
conserved as long as the same observer makes the measurement
before and after the process has taken place. The square of the
4 momentum (where by square I mean minus the square of the 0
component plus the sum of the squares of the other components)
is the same in one coordinate system before a process as it is in
a different non-accelerating frame of reference after the process.
This is invaluable, especially in problems where one is dealing
with the decay of a particle. For example:

Example 1 — A particle of rest mass M1 decays into a particle

of rest mass m2 and a photon. In the rest frame of the particle
produced in the decay, what is the energy Eγ of the photon, in
terms of the rest masses M1 and m2 and the speed of light c?

We could solve this problem using straight energy and mo-

mentum conservation, but it’s easier to use conservation of the
square of the 4–momentum. Consider the decaying particle in
its own rest frame. The total momentum is zero (it’s not mov-
ing), and the total energy is M1 c2 . Therefore the initial total
4–momentum is (E/c, px ) = (M1 c, 0), and the square of this 4–
momentum is −M12 c2 . Now, in the rest frame of the produced
particle, it’s energy is m2 c2 , since it is at rest, and the photon
has energy Eγ and momentum Eγ /c. Then the 4–momentum is
(m2 c + Eγ /c, Eγ /c). Equating the squares of the 4–momentum
of the decaying particle before the collision and the square of
the sum of the 4–momenta of the decay particle and emitted
photon after the collision, remembering that the square of the
0th component gets a minus sign due to the sign of the 00 metric
component, we get

3
  2  2
Eγ Eγ
−M12 c2 = − m2 c + c
+ c
 2  2
Eγ Eγ
= −m22 c2 − 2m2 Eγ − c
+ c
(6)
2m2 Eγ = M12 c2 − m22 c2

M12 c2 −m22 c2
Eγ = 2m2
.

Notice that even though the initial conditions and the final con-
ditions are specified in different frames of reference, there has
been no need for a fiddly Lorentz transformation to get from one
to the other. Instead, we exploited the fact that the square (in
the usual sense) of the 4–momentum is the same to all observers,
and conserved throughout the interaction.

Example 2 — A positron of total energy E strikes an electron

that is stationary in the lab. The two particles annihilate and
produce two photons, which in this particular instance (which
is a special case), move off at equal angles θ on opposite sides
of the axis along which the positron originally approached, as
illustrated in Figure 1. Find an exact expression for the angle θ
in terms of the energy E of the incident positron, the rest mass
me of both the electron ad the positron, and c.

Figure 1: The collision of a moving positron (e+ ) with an elec-

tron stationary in the laboratory. The two particles annihilate
yielding two photons.

We first apply conservation of energy, which gives

E + me c2 = 2Eγ . (7)

4
The momentum of the incident positron can be written in terms
of E and me c2 using E 2 = p2 c2 + m20 c4 , so that
p
pc = E 2 − m2e c4 . (8)

The target electron is at rest so it has no momentum. To con-

serve momentum, the y–components of the momenta of the two
photons must be equal and opposite. Furthermore, in this case,
we are told that each of the emitted photons is emitted down a
path that makes the same angle with the initial electron direc-
tion of incidence, as shown in the figure. This means that the
y–components of the momenta must obey

pγ1 sin θ = pγ2 sin θ, (9)

Therefore we must have

pγ1 = pγ2 , (10)

and therefore the photon energies are also equal, so Eγ1 = Eγ2 =
Eγ . So we again use conservation of momentum, this time along
the beam axis. We equate the positron momentum of Equation
8 with the x–components of the momenta of the two photons to
obtain p
E 2 − m2e c4 = 2Eγ cos θ. (11)
Dividing Equation 11 by Equation 7 eliminates the unknown Eγ
and obtain p
E 2 − m2e c4
cos θ = . (12)
E + me c2

Example 3 — An electron of total energy E = 10 GeV strikes

a proton at rest in the lab. Using appropriate approximations
to simplify the calculation where possible, what is the maximum
total rest mass that can be created in an inelastic collision where
both the electron and the proton are destroyed? The rest mass
of the proton is mp = (938 MeV)/c2 .

The easiest way to do this problem is using 4–vectors. In this

particular problem, the initial conditions are given in the lab
frame, so let’s write the total four momentum of the incident
electron plus the proton target in the lab frame. We start by
noticing that the incident electrons total energy is about 5000
times its rest energy, so we neglect the rest energy of the incident
electron. Therefore the momentum of the incident electron is
taken as E/c. The 4–momentum of the electron plus stationary
target proton is therefore

5
  
E/c + mp c
 E/c 
pi =  . (13)
 0 
0

Now, the maximum rest mass that can be created in the collision
is determined by the total energy in the frame in which the final
state particles are at rest. This is because any kinetic energy
the final state has could potentially have been used to generate
extra mass. The final state particles are only all at rest in the
centre of mass frame. In this frame, the rest energy is M c2 ,
where M is the sum of the masses of the particles that can be
created, and the momentum is zero. Therefore we write out the
4–momentum of the final state in the centre of mass frame.


Mc
 0 
pf = 
 0 .
 (14)
0

Now, we calculate the combination of the 4–momentum com-

ponents that is conserved in a frame independent manner. It is
most important to realize that this only works because this com-
bination of E and p is independent of the velocity of the inertial
frame. It is not true that total energy and momentum before a
collision in one frame are equal to total energy and momentum
after the collision in a different frame. Energy and momentum
conservation only apply where it is the same observer measuring
energy and momentum before and after the interaction.

2
E2

E
− + mp c + = −M 2 c2 . (15)
c c2

Notice how simple this procedure is, because in the centre of

mass frame at threshold where nothing is moving, the 4–momentum
only has one non-zero component. Now rearrange and simplify.

E 2 E2
−M 2 c2 = −
p c2
− 2Emp − m2p c2 + c2
M c2 = 2Em 2 2 2
q p c + (mp c ) (16)
= mp c2 m2E pc
2 + 1.

6
Using mp c2 = 938 MeV we obtain M c2 = 4.43 GeV.

3 The Greisen-Zatsepin-Kuzmin (GZK)

limit

Example 4 — Cosmic rays consist of high energy protons emitted

by astrophysical sources far from Earth. However, there is a
theoretical limit to the maximum energy of cosmic rays from deep
space caused by the cosmic microwave background radiation. If
a proton has sufficient energy, it can scatter off a microwave
background photon to form, for example, a neutron and a π–
meson.

p + γ → π + + n. (17)

Deduce the maximum proton energy that can travel through deep
space taking cosmic microwave background photons have ener-
gies of 6 × 10−4 eV, and using 140 MeV and 938 MeV for the
π–meson and proton rest energies, respectively. Assume that
the neutron and proton rest energies are the same.

You can again use 4–momentum to do this problem. First, write

down the total 4 momentum of the gamma and the proton. Use
pγ for the 4–momentum of the microwave photon, and pp for
the 4–momentum of the proton. The total 4–momentum of the
initial state is therefore (pγ + pp ). Now use a dot to represent
the operation of taking minus the square of the zeroth compo-
nent and adding to it the sum of the squares of the other three
components. With this notation the intial state 4 momentum
dotted with itself is:

(pγ + pp ) · (pγ + pp ) = pγ · pγ + 2pγ · pp + pp · pp . (18)

Now deal with these three terms on the right in order. pγ · pγ =

0, because it is equal to −Eγ2 /c2 + |p~γ |2 which is zero because
Eγ = |p~γ |c. Similarly, pp · pp = −m2p c2 because −Ep2 /c2 +
|p~p |2 = −m2p c2 . Therefore

7
 (pγ + pp ) · (pγ + pp ) = 2pp · pγ − m2p c2 . (19)

What about the 4–dot product between the four momenta pp

and pγ ? Do everything in one dimension, since at threshold
for the production of new particles, the collisions will be head
on. Then the incoming proton has energy Ep and momentum
+Ep /c, neglecting rest mass because the cosmic ray energy is
enormous. Similarly, the photon has energy Eγ and momentum
−Eγ /c. Forming the 4–dot product of these two we get

2Eγ Ep
pγ · pp = − . (20)
c2

Therefore Equation 19 becomes

4Eγ Ep
(pγ + pp ) · (pγ + pp ) = − − m2p c2 . (21)
c2

Now, as in example 2, at threshold the final state neutron and

proton will be produced at rest in the centre of mass frame.
Therefore the 4–momentum in the final state is

 
(mn + mπ )c
 0 
pf =  . (22)
 0 
0

Hence the 4–dot product of pf with itself is −(mn + mπ )2 c2 .

Equating this with the 4–dot product of the initial momentum
with itself, we obtain

4Eγ Ep
c2
+ m2p c2 = (mn + mπ )2 c2
(23)
= m2n c2 + 2mn mπ c2 + m2π c2

We assume m2p c2 = m2n c2 and rearrange to give

2(mn c2 )(mπ c2 ) + (mπ c2 )2

Ep = (24)
4Eγ

8
Plugging in some numbers, I get Ep = 1.2 × 1020 eV. A proton
having this energy would have a gamma factor of 1011 . To an ob-
server in the rest frame of this proton, our galaxy would appear
Lorentz contracted from a diameter of about 30 kPc to a diam-
eter of about 30, 000[pc] × 3.1 × 1016 [m/pc]/1011 = 9.3 × 109 m.
Such a proton, in its own reference frame, would traverse our
galaxy in 31 seconds.

Searches for ultra high energy cosmic ray protons have discov-
ered events close to the GZK threshold; discovery of a significant
population of protons having energies higher than this cutoff
would indicate that the source of the cosmic rays is very close to
us, something that we think is unlikely. Seeing whether cosmic
ray protons do indeed cut off at about 1020 eV is an interesting
current experimental problem.