Quantum Electrodynamics (QED), Winter 2015/16 Calculating cross sections

Calculating cross sections for different processes
Lepton production
The simplest possible process we can consider is lepton pair production, in which an electron and
position annihiliate to produce a different lepton-antilepton pair (µ or τ ) labelled by l±

e+ (p1 , r) + e− (p2 , s) → l+ (p01 , r0 ) + l− (p02 , s0 )

Only one Feynman diagram contributes at tree level, and gives

l+ (p01 , r0 ) e+ (p1 , r)

l− (p02 , s0 ) e− (p2 , s)

1
M = ie2 ūs0 (p02 )γ µ vr0 (p01 ) v̄r (p1 )γµ us (p2 )
(p1 + p2 )2
Averaging over incoming spins and summing over outgoing spins we have
1 X
X= |M|2
2·2 0 0
rsr s
e4 X 
ūs0 (p02 )γ µ vr0 (p01 )v̄r0 (p01 )γ ν us0 (p02 ) [v̄r (p1 )γµ us (p2 )ūs (p2 )γν vr (p1 )]

=
4(p1 + p2 )4
rsr0 s0
e4
≡ Aµν Bµν
4(p1 + p2 )4
where, given the completeness relations for u and v, we have

Aµν = Tr γ µ (p/01 − ml )γ ν (p/02 + ml )

 

B µν = Tr [γ µ (p/2 + me )γ ν (p/1 − me )]

These two terms are related by exchanging (p01 , p02 ) ↔ (−p2 , −p1 ).
Using the expression for Tr(γ µ1 · · · γ µn ) it is easy to show that

Aµν = 4 p0µ 0ν 0µ 0ν µν
p01 · p02 + m2l


1 p2 + p2 p1 − g

B µν = 4 pµ1 pν2 + pµ2 pν1 − g µν p1 · p2 + m2e




and hence
8e4 h
X= (p1 · p01 )(p2 · p02 ) + (p1 · p02 )(p2 · p01 )
(p1 + p2 )4
i
+ m2e (p01 · p02 ) + m2l (p1 · p2 ) + 2m2e m2l

1
Quantum Electrodynamics (QED), Winter 2015/16 Calculating cross sections

We now need to evaluate the cross-section. Let us work in the center of mass frame. We have

p1 = (E1 , p) p01 = (E10 , p0 )

p2 = (E2 , −p) p02 = (E20 , −p0 )

Since the incoming particles have the same mass, as do the outgoing particles, in the COM frame,
conservation of energy implies E1 = E1 = E10 = E20 ≡ E. Let θ be the angle between the incoming
and outgoing momenta, that is p · p0 = |p||p0 | cos θ. Finally, we note that the process can only occur
if E > ml . Since ml  me we have E  me and we may effectively neglect terms proportional to
me so that |p| ≈ E. We then find

α2 |p0 |
 
dσ
E 2 + |p0 |2 cos2 θ + m2l


= 4
dΩ COM 16E E
α2
1 + cos2 θ


≈ 2
16E

where α = e2 /4π. The second line is the ultra-relativistic E  ml limit where we can also neglect


terms proportional to ml and |p0 | ≈ E.

Bhabha scattering

The process
e+ (p1 , r) + e− (p2 , s) → e+ (p01 , r0 ) + e− (p02 , s0 )

is very similar to lepton pair production except that now two diagrams contribute to the amplitude.
Let us use the same labelling of momenta and spins. We have

e+ (p01 , r0 ) e+ (p1 , r)


e− (p02 , s0 ) e− (p2 , s)

1
Ma = ie2 ūs0 (p02 )γ µ vr0 (p01 ) v̄r (p1 )γµ us (p2 )
(p1 + p2 )2

e+ (p01 , r0 ) e+ (p1 , r)

e− (p02 , s0 ) e− (p2 , s)
2
Quantum Electrodynamics (QED), Winter 2015/16 Calculating cross sections

1
Mb = −ie2 ūs0 (p02 )γ µ us (p2 ) v̄r (p1 )γµ vr0 (p01 )
(p01 − p1 )2
where the first amplitude is the one that enters lepton pair production (with ml = me ). For simplicity,
in what follows, we consider the ultra-relativistic limit where the scattering energy is much larger than
me and we can effectively treat the incoming and outgoing particles as massless.
Calculating the spin sums we get four terms

1 X 1X 1X 1X 1X ∗
X= |M|2 = |Ma |2 + |Mb |2 + Ma M∗b + Ma Mb
2·2 0 0 4 4 4 4
rsr s
∗
≡ Xaa + Xbb + Xab + Xab

We already calculated the first term in the lepton production process. In the high-energy limit, we
have (p1 + p2 )2 = 2p1 · p2 and

2e4
(p1 · p01 )(p2 · p02 ) + (p1 · p02 )(p2 · p01 )
 
Xaa = 2
(p1 · p2 )

Since the two Feynman diagrams are related by exchanging p2 ↔ −p01 , we immediately have

2e4  0 0 0 0

Xbb = (p1 · p 2 )(p 1 · p2 ) + (p1 · p 2 )(p 2 · p1 )
(p1 · p01 )2

Finally we have
e4
Xab = Y
16(p1 · p2 )(p01 · p1 )
where
X
Y = ūs0 (p02 )γ µ vr0 (p01 )v̄r0 (p01 )γ ν vr (p1 )v̄r (p1 )γµ us (p2 )ūs (p2 )γν us0 (p02 )
rsr0 s0

/01 γ ν p
= Tr γ µ p /02


/1 γµ p
/ 2 γν p

Using the identities of the form γ µ · · · γµ , we find

/01 p
/1 γ ν p /02 = −8(p01 · p2 ) Tr p/1 p/02 = −32(p01 · p2 )(p1 · p02 )



Y = −2 Tr p / 2 γν p

To end, we substitute these results into the COM differential cross section formula. Since the final
and initial masses are equal (and effectively zero in our high energy limit), we have |p| = |p0 | = E.
Evaluating the four momentum products we find
" #
α2
1 + cos4 12 θ cos4 12 θ
 
dσ 1
= 1 + cos2 θ + − 2
dΩ COM 8E 2 2 sin4 21 θ sin2 12 θ

∗ respectively.
The three contributions are from Xaa , Xbb and Xab + Xab

3
 
Quantum Electrodynamics (QED), Winter 2015/16 Calculating cross sections

Compton scattering: unpolarized


Next we consider Compton scattering

γ(k, ) + e− (p, s) → γ(k 0 , 0 ) + e− (p0 , s0 )

with unpolarized photons and electrons. Again there are two Feynman diagrams which contribute
γ(k 0 , 0 ) γ(k, )

e− (p0 , s0 ) e− (p, s)
1
Ma = −ie2 ūs0 (p0 )/0 /us (p)
(p/ + k/) − m

γ(k 0 , 0 ) γ(k, )

e− (p0 , s0 ) e− (p, s)
1
Mb = −ie2 ūs0 (p0 )/ /0 us (p)
(p/ − k/0 ) − m
which are related by k ↔ −k 0 .
By momentum conservation we have p+k = p0 +k 0 and the mass-shell conditions p2 = p02 = m2
and k 2 = k 02 = 0. Thus we have (p + k)2 − m2 = 2p · k and (p − k 0 )2 − m2 = −2p · k. The two
diagrams give four contributions to the polarization sum
1 X 1X 1X 1X 1X ∗
X= |M|2 = |Ma |2 + |Mb |2 + Ma M∗b + Ma Mb
2·2 4  4 4 4
e2

Yaa Ybb Yab + Yba
≡ + 0
−
16 (p · k) 2 (p · k )2 (p · k)(p · k 0 )
where X
Yaa = ūs0 (p0 )
/0 (p
/ + k/ + m)/us (p)ūs (p)/(p/ + k/ + m)/0 us0 (p0 )
= Tr γ µ (p/ + k/ + m)γ ν (p/ + m)γν (p/ + k/ + m)γµ (p/0 + m)
 

and X
Yab = ūs0 (p0 )
/0 (p
/ + k/ + m)/us (p)ūs (p)/0 (p/ − k/0 + m)/us0 (p0 )
= Tr γ µ (p/ + k/ + m)γ ν (p/ + m)γµ (p/ − k/0 + m)γν (p/0 + m)
 

where Ybb and Yba are equal to Yaa and Yab with k ↔ −k 0 .
To calculate the traces we again repeated use identities of the form γ µ · · · γµ . We have, for in-
stance, given the cyclic property of traces,

/ + k/ + m)γ ν (p/ + m)γν (p/ + k/ + m)γµ (p/0 + m)γ µ

 
Yaa = Tr (p
/ + k/ + m)(4m − 2p/)(p/ + k/ + m)(4m − 2p/0 )
 
= Tr (p

4
Quantum Electrodynamics (QED), Winter 2015/16 Calculating cross sections

Then using the identities for Tr(γ µ1 · · · γ µn ), after some considerable algebra, we find

Yaa = 32 (p · k)(p · k 0 ) + m2 (p · k) + m4
 

Here we have used the fact that momentum conservation means the four-momentum products can be
written in terms of three invariants: p · k, p · k 0 and m2 . In particular

p2 = p02 = m2 k 2 = k 02 = 0
p0 · k = p · k 0 p0 · k 0 = p · k
p · p0 = m2 − p · k − p · k 0 k · k0 = p · k − p · k0

Recall Ybb is given by the same expression with k ↔ −k 0 and a similar calculation gives

Yab = Yba = 16m2 2m2 + p · k − p · k 0

 

Finally we would like to give an expression for the cross-section. Typically Compton scattering is
measured in the LAB frame. We have

p1 = (ω, k) p01 = (ω 0 , k0 )
p2 = (m, 0) p02 = (E 0 , p0 )

with |k| = ω and |k0 | = ω 0 . We define θ as the angle between in the incoming and outgoing photons,
that is k · k0 = ωω 0 cos θ. Since photons are massless we have v12 = 1. Subsituting into the general
(colinear frame) formula for the cross section we then have

ω0 ∂(E 0 + ω 0 ) −1
   
dσ 1
= X
dΩ LAB 64π 2 mωE ∂ω 0

To calculate the partial derivative we first note recall that p · k = k · k 0 + p · k 0 . This implies the
standard relation for the shift in Compton photon frequency
1 1 1
− 0 = (cos θ − 1)
ω ω m
√
We also have E 0 = |k − k0 |2 + m2 = ω 2 + ω 02 − 2ωω 0 cos θ + m2 . Thus
p

∂(E 0 + ω 0 ) ω 0 − ω cos θ ω(1 − cos θ) + m mω

0
= 0
+1= 0
= 0 0
∂ω E E Eω
where we have used energy conservation m + ω = E 0 + ω 0 and also the relation between ω and ω 0
just given. Gathering all these results together we finally have
2 
α2 ω0 ω0
   
dσ ω
= + − sin2 θ
dΩ LAB 2m2 ω ω0 ω