4

STANDARD MODEL
TREE-LEVEL PROCESSES

This chapter describes in some detail the tree-level computations of some

elementary standard model processes. Both leptons and quarks are treated
on the same footing, ignoring for the most part that quarks do not exist
as free states. Quantum corrections to these results are relegated to a later
chapter.

4.1 Tree-Level Parameters

In chapter 2 we described the classical standard model Lagrangian in terms
of eighteen parameters: three gauge coupling constants, three charged lep-
ton masses, six quark masses, three flavor-mixing angles, one CP-violating
phase, the Higgs mass and the vacuum value of the Higgs field which is fixed
by the masses of the W and Z gauge bosons. All these parameters have been
measured to varying degrees of accuracy, except the Higgs mass for which
there is an upper limit of 95 GeV (LEP, 1999).
All parameters of the standard model run (some only crawl) with the scale
at which they are measured. Lepton masses which hardly run with scale
can be measured directly through kinematics. To distinguish the running
from the physical fermion mass, we denote the latter in capital letters. The
experimental values of the charged lepton masses are

Me = .511 MeV , Mµ = 105.66 MeV , Mτ = 1777.05 MeV . (4.1)

Since quarks are subject to the strong QCD force, they do not exist as
asymptotic states. This makes the measurements of their masses at best
indirect, and subject to qualifications. The masses of the three lightest
quarks are extracted from the effective low energy chiral Lagrangian (see
next chapter), with the range of values

1
2 STANDARD MODEL TREE-LEVEL PROCESSES

Mu = 2 − 8 MeV , Md = 5 − 15 MeV , Ms = 100 − 300 MeV . (4.2)

For heavier quarks, one may adopt a universal operational definition for
their physical mass as say, half the energy needed for pair production of the
quark-antiquark bound state. The other three quarks have masses which
are all above the QCD scale, making their extraction from the data more
direct, with values ranging over

Mc = 1.0 − 1.6 GeV , Mb = 4.1 − 4.5 GeV , Mt = 175.5 ± 5.5 GeV . (4.3)

The masses and widths of the W and Z bosons have been measured to
remarkable accuracies

MW = 80.37 ± 0.09 GeV , ΓW = 2.06 ± 0.06 GeV , (4.4)

MZ = 91.187 ± 0.007 GeV , ΓZ = 2.496 ± 0.0027 GeV . (4.5)

The rest of the parameters appear in the Cabibbo-Kobayashi-Maskawa ma-

trix, traditionally represented as
 
Vud Vus Vub
U =  Vcd Vcs Vcb  . (4.6)
Vtu Vts Vtb
Not all of its nine matrix elements have been measured directly, but all can
be inferred from the unitarity relations

X
Vki∗ Vkj = δij , i, j = u, c, t . (4.7)
k=d,s,b

Since the rotation group is generated by the three family Gell-Mann matrices
lambdab2 , λ5 , λ7 , we can write the CKM matrix à la Euler

U = eia4 λ4 eia5 λ5 eia7 λ7 eia2 λ2 , (4.8)

where the CP-violating phase is set along λ4 . Following L. Wolfenstein, we

express its parameters as a power series in the Cabibbo angle,

a2 = λ ; a7 = Aλ2 ; a5 = Aρλ3 ; a4 = −Aηλ3 , (4.9)

 4.2 W-decay 3

where λ is the Cabibbo angle. The first three parametrize the rotations,
and Aη denotes the CP violating phase. In matrix form,
2
1 − λ2 λ Aλ3 (ρ − iη)
 
2
U = −λ 1 − λ2 Aλ2  , (4.10)
3
Aλ (1 − ρ − iη) −Aλ2 1
where A, ρ, and η are of order one. Their (1999) values

λ ≈ .2205 ± .0018 ; A ≈ 0.81 ± 0.06 , (4.11)

are obtained from direct measurements of Vus and Vcb , respectively. From
the branching ratio of B-meson decay into charm and up quarks, we get

p
ρ2 + η 2 = 0.36 ± 0.09 . (4.12)

The extraction of the phase from data is more indirect, as it necessarily

involves loops with the third family, since CP-violation in B-decay has not
yet been observed. One can at best constrain its value, either by deducing
Vtd from B − B mixing, or K − K mixing. One can quote the central values

ρ ≈ 0.05 , η ≈ 0.35 . (4.13)

Finally we note that since the CKM matrix is unitary, the first column times
the complex conjugate of the third column must vanish, yielding

Aλ3 (ρ + iη) − Aλ3 + Aλ3 (1 − ρ − iη) = 0 . (4.14)

With this parametrization, this is obviously tautological, but we may view

the three factors, ρ + iη, −1, and 1 − ρ − iη as the sides of a triangle in the
complex (ρ, η) plane, called the unitarity triangle. Its sides and its angles
(CP-violation) can be independently determined from experiments. If after
these measurements, the triangle does not close, there must be sources of
CP-violation beyond the standard model.

4.2 W-decay
As a first example of a tree-level application of the standard model, we
calculate the decay rate of the W -boson into an electron and its associated
antineutrino,
4 STANDARD MODEL TREE-LEVEL PROCESSES

W − → e− + ν̄e .

The decay proceeds at tree-level through the interaction

ig2
Lint = √ Wµ+ (x)J −µ (x) , (4.15)
2
using only the electron part of the charged current. This process is repre-
sented by the Feynman diagram

W
p
e

e

k
q

It corresponds to the matrix element

Z
ig2
Tf i =√ d4 xhW −(λ) (k)|Wµ+ (x)ν̄e (x)γ µ L e(x)|e(a) (p) ν̄ (b) (q)i , (4.16)
2
where we have rewritten the current in terms of four-component Dirac fields
to facilitate the computations to come; L is the projection operator onto
left-handed states

1 + γ5
L= . (4.17)
2
Assuming that the plane wave states are normalized in a box of volume V ,
we have

1 e−ik·x ∗(λ)
hW −(λ) (k)|Wµ+ (x)|0i = √ √ µ (k) , (4.18)
V 2EW
(λ)
where µ denote the plane wave solutions to the massive Klein-Gordon
equation with polarization λ = 1, 2, 3; they obey

k µ µ(λ) (k) = 0 . (4.19)

 4.2 W-decay 5

Similarly, the fermion fields acting on the plane wave states yield

1 1
h0|e(x)|e(a) (p)i = √ √ eip·x u(a) (p) , (4.20)
V 2Ee
1 1
h0|ν̄e (x)|ν̄e(b) (q)i = √ p eiq·x v̄ (b) (q) , (4.21)
V 2Eνe
where a, b are the fermion polarizations which run over two values. The
four-component covariant plane wave states, u(a) and v (a) , are normalized
to

ū(a) (p)u(b) (p) = 2me δab , v̄ (a) (p)v (b) (p) = −2mνe δ ab , (4.22)
where the bar denotes the Pauli adjoint. Putting it all together, and inte-
grating over x, we obtain the matrix element

ig2 (2π)4 δ (4) (k − p − q) ∗(λ)

Tf i = p µ (k)v̄ (a) (q)γ µ Lu(b) (p) . (4.23)
4V 3/2 EW Ee Eνe
Its absolute square gives the transition probability over all of space and
time. One of the two δ-functions (times (2π)4 ) is interpreted as the volume
of space-time, V T .
The differential decay rate is the transition probability per unit of time,
multiplied by the number of final states; with our normalization of the plane
wave states, for each final state with three momentum p~, the number of
available states is

V
d3 p , (4.24)
(2π)3
leading to the differential decay rate

g22 (4) d3 p d3 q
dΓ = δ (k − p − q) ×
16(2π)2 EW Ee Eνe
∗(λ) (λ)
µ (k)ρ (k)v̄
(a)
(q)γ µ Lu(b) (p)ū(b) (p)γ ρ Lv (a) (q) .(4.25)
The sums over the neutrino and electron polarizations, using

X X
v (a) (q)v̄ (a) (q) = /q −mνe , u(a) (p)ū(a) (p) = /p +me , (4.26)
a a

yield the trace

6 STANDARD MODEL TREE-LEVEL PROCESSES

Tr (γµ L[p / −mνe ]) = Tr(γµ /p γρ L /)

/ +me ]γρ L[q q , (4.27)

since Lγρ L = 0. To obtain the total decay rate, we average over the three
initial polarizations of the W -boson, using

X kµ kρ
∗(λ) (λ)
µ (k)ρ (k) = −(gµρ − 2 ) . (4.28)
MW
λ

Substituting in the differential decay rate, and dividing by the number of

polarization states of the W , we obtain

−g22 δ (4) (k − p − q) d3 p d3 q µρ k µ k ρ
dΓ = (g − 2 )Tr(γµ/γ
p ρ Lq
/) . (4.29)
48(2π)2 EW Ee Eνe MW
To evaluate the trace, we use the identity

γµ γν γρ = gµν γρ + gνρ γµ − gµρ γν + iµνρσ γ5 γ σ , (4.30)

where 0123 = 1, with the result

Tr(γµ γν γρ γσ L) = 2(gµν gρσ + gµσ gρν − gµρ gσν + iµνρσ ) . (4.31)

Then, by using the kinematical relations (q · p ≡ q0 p0 − ~q · p~)

1
p · k = m2e + p · q , q · k = p · q , p · q = (MW
2
− m2e ) , (4.32)
2
we find, in the rest frame of the W -boson,

g22 MW m2 m2e d3 p d3 q
  
dΓ = 1 − 2e 1+ 2 δ (4) (k − p − q) . (4.33)
24(2π)2 MW 2MW Ee Eνe
Use of the generic phase space integral formula

d3 p d3 q (4)
Z
I ≡ δ (k − p − q) ,
p0 q0
2π p 2
= [k − (m1 − m2 )2 ][k 2 − (m1 + m2 )2 ] , (4.34)
k2
where
 4.2 W-decay 7

q q
q0 = m21 + ~q · ~q , p0 = m22 + p~ · p~ .

Setting m1 = 0 and m2 = me , it yields the final result

2 
G M3 m2e m2e
 
− −
Γ(W → e + ν̄e ) = F √ W 1− 2 1+ 2 , (4.35)
6π 2 MW 2MW
written in terms of the Fermi coupling constant, GF ,

GF g22
√ = 2 , (4.36)
2 8MW
which parametrizes the strength of the current-current interaction in the
effective Hamiltonian

GF
Hint = 4 √ Jµ+ J −µ . (4.37)
2
The factor of 4 is to account for the L in the charged current (Fermi did not
envisage Parity violation). Note the weak dependence of the decay rate on
the outgoing particle masses, as long as they are small compared to the W
mass: the decay rate of the W into an electron, a muon, and a tau lepton are
nearly equal. This lepton universality is experimentally verified: the decay
leptonic branching ratios of the W into electron, muon, and tau leptons are
10.9 ± .04%, 10.2 ± 0.05%, and 11.3 ± 0.08%, respectively.
The W boson also decays into a quark and an antiquark of different flavors.
Since quarks do not exist as asymptotic states, the computation of the decay
width into a particular combination of quarks and antiquarks, say ūd, has
no direct physical meaning. As the quark and antiquark leave the point of
interaction and fly apart from one another, the strong QCD force grows with
their separation. It abates only when color neutral combinations of quarks
and antiquarks, such as pions, have been created. This is achieved by means
of sparking of the QCD vacuum through quark-antiquark pair creation. At
present, physicists do not have the tools necessary to compute in detail this
“hadronization” process. We only know is that it takes place 100% of the
time.
There is hope since as long as we do not seek too many details of the
hadronic final state, we can use unitarity to finesse these subtelties. For
example, the probability that a W decays into hadrons, without specifying
8 STANDARD MODEL TREE-LEVEL PROCESSES

anything about the hadrons, is the sum of the probability that it decays into
orthogonal quark-antiquark pairs

X X
Γ(W − → Hadrons) = Γ(W − → ūdi ) + Γ(W − → c̄di ) .
di =d,s,b di =d,s,b
(4.38)
Note that we have not included the top quark, since it is too heavy for
this decay to proceed, by which we mean that hadrons that contain the top
quark are more massive than the W .
We can compute these decay widths the same way as for the leptonic
decay. Since the quarks are all much lighter than the W , we can treat them
all as massless. In this case, the calculation has already been done. It is
easy to see that, for instance,

Γ(W − → ūd) = 3|Vud |2 Γ(W − → e− + ν̄e ) , (4.39)

where the factor of three is for the three colors of quarks, and Vud is the
relevant CKM matrix element. Thus, we find that

X
Γ(W − → Hadrons) = 3 |Vij |2 Γ(W − → e− + ν̄e ) ,
i=u,c
j=d,s,b
−
= 6Γ(W → e− + ν̄e ) . (4.40)

In the last we have used the unitarity of the CKM matrix. Thus we expect
that the W will decay into hadrons 2/3 of the time, and into each charged
lepton 1/9 of the time.

4.2.1 PROBLEMS
A. Verify the generic phase space formula Eq. (4.34).

B. Find the fraction of the time the W decays into strange hadrons.

C. Identify the main decay channel of the top quark, and calculate its life-
time.

D. 1-) Suppose there is a fourth chiral family of quarks and leptons, but that
only its charged lepton and the massless associated neutrino are lighter than
the W . Using the latest from the Particle Data Book, what is the lightest
possible value of the charged lepton mass?
 4.3 Z-decay 9

2-) Assuming that the fourth neutrino happens to have a (Dirac) mass,
calculate the fourth family leptonic contribution to the W decay rate. What
can you say about the possible values of the masses of these fourth family
leptons?

4.3 Z-decay
Using the results of the previous section, it is straightforward to compute
the width of the Z boson in the standard model. The only difference with
the previous section is that the interaction now proceeds via

ieZ µ
Lint = [J 3 − sin2 θw Jµem ] . (4.41)
sin θw cos θw µ
For the neutrino of each species, we just use the formula (4.22) with me = 0,
replace MW by MZ , and

g g2
√2 → .
2 cos θW
The result is

g22 MZ GF MZ3
 
0 1
Γ(Z → νe + ν̄e ) = = √ . (4.42)
4 24π cos2 θW 12π 2
Of course, this width can never be measured directly, since we do not have
the means to detect the neutrinos; it can only be measured by default, that
is, by independently measuring the width into visible products, and then
the total width.
The decay width into electron-positron pair is given by

GF MZ3
Γ(Z 0 → e+ + e− ) = (aL + aR ) √ , (4.43)
3π 2
where we have neglected the mass of the electron, and thus added incoher-
ently the left and right contributions, with

1
aL = − sin2 θW + sin4 θW ; aR = sin4 θW . (4.44)
4
By using unitarity arguments similar to those the previous section, we can
obtain the total rate into hadrons. The tree-level result is, neglecting the
masses of the quarks,
10 STANDARD MODEL TREE-LEVEL PROCESSES

GF MZ3
 
0 15 22
Γ(Z → Hadrons) = − 7 sin2 θW + sin4 θW √ . (4.45)
4 3 3π 2
These formulae can be used to extract the value of sin2 θW from experiments.
Finally, let us calculate polarized Z decay; we start from Eq. (4.41), but
without summing over the Z polarizations. It is easier to do the calculations
in the Z rest frame where the polarization vector is purely space-like with
components

µ(3) = (0, 0, 0, 1) , (4.46)

1
µ(±) = √ (0, 1, ∓i, 0) , (4.47)
2
Since we neglect the mass of the electron, the spin of the outgoing elec-
tron points away from its momentum (if it is left-handed), and that of the
positron along its momentum. By conservation of angular momentum, the
spin vector of the Z is aligned with the direction of the outgoing positron.
The contribution to the decay rate from a transversely polarized Z into a
left-handed electron-positron pair is then

aL g22 δ (4) (k − p − q) d3 p d3 q ∗(+) (+)

dΓT =  ρ Tr(γ µ/γ
p ρ Lq
/) . (4.48)
24(2π)2 cos2 θW MZ Ee Eē µ
A similar equation holds for the decay width into a right-handed pair, but
with the L projection operator replaced by

1 − γ5
R= , (4.49)
2
and aL replaced by aR . Using Eq. (4.47), and evaluating the trace, we find

aL g22 δ (4) (k − p − q) d3 p d3 q
dΓT = (q0 p0 − q3 p3 − q0 p3 + q3 p0 ) . (4.50)
24(2π)2 cos2 θW MZ Ee Eē
In the rest frame of the Z, we have (neglecting the electron mass),

MZ MZ
p0 = q0 = ; p~ = −~q = n̂ , (4.51)
2 2
where n̂ is a unit vector. After performing the δ-function integrations, we
find
 4.3 Z-decay 11

aL g22 MZ
dΩ 1 + n̂23 − 2n̂3 .


dΓT = 2 2
(4.52)
48(2π) cos θW
This leads to the decay angular distribution

dΓT G M3
= F √Z aL (1 − cos θ)2 , (4.53)
d cos θ 12π 2
where θ is the direction of the electron in the Z rest frame, relative its
polarization. This angular dependence is connected with the amplitude for
a Z polarized along one axis to be found with polarization at an angle θ
from it. It vanishes at θ = 0, when the electron-positron spin is anti-aligned
with that of the Z. A similar formula holds for the right handed electrons
with a change of sign in cos θ, and aL replaced by aR .
One can show that the decay of the longitudinal Z leads to the formula

dΓLong G M3
= F √Z 2aL sin2 θ . (4.54)
d cos θ 12π 2

4.3.1 PROBLEMS

A. Compute the branching ratio

Γ(Z 0 → bb̄)
Γ(Z 0 → hadrons)
and compare with the latest experimental value.

B. Compute the total Z decay rate into τ pair, without neglecting the τ
masses.

C. Verify Eq. (4.54) for the angular probability distribution for an electron
in the decay of a longitudinally polarized Z at rest. Discuss its physical
interpretation.

D. Consider the process e+ e− → µ+ µ− at the Z resonance with longitudi-

nally polarized electrons in the initial state. Derive the expression for the
longitudinal polarization asymmetry in this collision. Discuss how it is mea-
sured and its advantages in the extraction of the value of sin2 θW over other
methods. For reference, see B.W. Lynn and C. Verzegnassi, Phys. Rev. D
35, 3326(1987).
12 STANDARD MODEL TREE-LEVEL PROCESSES

E. Show that, using present data, the hypothesis of a fourth family with a
massless neutrino is ruled out. Then assume that the neutrino of the fourth
family has a mass. Together with the results of problem D of last section,
what are the allowed ranges of masses for the hypothetical fourth neutrino
and charged lepton?

4.4 Muon Decay

In the standard model, the muon decays into a muon neutrino, an electron
and its associated antineutrino

µ− → νµ + e− + ν̄e .

The interaction responsible for this decay is the same as that for W and Z
decays, namely

ig2 h
† †
i
Lint = √ W +α νeL σα eL + νµL σα µL + · · · . (4.55)
2
Since the W-boson is eight hundred times more massive than the muon, the
muon cannot decay into a physical W , but this is quantum field theory:
the W can appear as an intermediate state which can be off its mass-shell;
the emitted W can then in turn decay into an electron and its associated
neutrino. This is shown by the following Feynman diagram


l


p
k

W e
0
l

e

It corresponds to the effective interaction Hamiltonian

−g22
Z Z
† †
Hint = d4 x d4 y∆αβ
F (x − y)νµL (x)σα µL (x)eL (y)σβ νeL (y) , (4.56)
2
 4.4 Muon Decay 13

where

e−iq·(x−y)
Z
1
∆αβ
F (x − y) = g αβ
d4 q 2 + i , (4.57)
(2π)4 q 2 − MW
is the Feynman propagator of the W -boson in the Feynman gauge (ξ =
1). We are using the Minkowski representation of the propagator for this
problem. We drop the i term since it is only relevant when the W is on its
mass-shell, which is clearly not the case here.
We use the simple Fierz identity

† 1 †
νeL νµL = − σρ νµL σ ρ νeL , (4.58)
2
as well as

σ α σρ σα = −2σρ , (4.59)
to rewrite the interaction in the form

−g22
Z Z
†
Hint = 4
d x d4 y∆F (x − y)νµL (x)σ ρ νeL (y)e†L (y)σρ µL (x) . (4.60)
2
The matrix element between the states shown in the Feynman diagram yields
the transition amplitude over the whole of space-time

Tf i = hµ(a) (k)|Hint |e(b) (p); ν̄e(c) (l0 ); νµ(d) (l)i ,

where the states denote the plane wave solutions of the Dirac equation. The
superscripts a, b, c, d denote their two-valued polarizations. When evaluating
the action of the fermion fields on the plane wave states, we switch over to
the Dirac notation. Then using Eqs (4.6, 4.7), and integrating over x, y, and
q, we arrive at

Tf i = (2π)4 δ (4) (k − p − l − l0 )T̂f i , (4.61)

where

−g22 1 ū(d) (l)γ ρ Lv (c) (l0 )ū(b) (p)γρ Lu(a) (k)

T̂f i = 2 . (4.62)
8V 2 Eµ Ee Eνe Eνµ (k − l)2 − MW
p

By now the steps to obtain the differential decay rate should be familiar.
14 STANDARD MODEL TREE-LEVEL PROCESSES

First we take the absolute square of the transition amplitude over space-
time, interpret (2π)4 times one δ function as the volume of space-time, then
divide by the time interval, and multiply by the number of available final
states. The V ’s cancel, and we are left with

g24 δ (4) (k − p − l − l0 ) d3 p d3 l0 d3 l
dΓ =
64(2π)5 Eµ (k − l)2 − M 2 2 Ee Eνe Eνµ
 
W
×|ū(d) (l)γ ρ Lv (c) (l0 )ū(b) (p)γρ Lu(a) (k)|2 . (4.63)

The neutrinos are not detected, and the sum over their polarizations yields
the trace
 
Tr(γρ/l0 γσ Ll
/) = 2 lρ lσ0 + lρ0 lσ − gρσ l · l0 − iρσαβ lα l0β . (4.64)

In the limit of zero electron mass, the electron comes out purely left-handed.
Summing over the electron’s polarizations, we find

g24 δ (4) (k − p − l − l0 ) d3 p d3 l0 d3 l
dΓ =
32(2π)5 Eµ (k − l)2 − M 2 2 Ee Eνµ Eνµ
 
W
 
× lρ lσ + lρ lσ − gρσ l · l − iρσαβ lα l0β ū(a) (k)γρ/γ
0 0 0
p σ Lu(a) (k)(4.65)
.

If the polarization of the decaying muon is not measured, we average over

its polarization states to obtain the decay rate, obtaining

g24 δ (4) (k − p − l − l0 ) d3 p d3 l0 d3 l
dΓ =
32(2π) Eµ (k − l)2 − M 2 2 Ee Eνe Eνµ
5  
W
 
× lρ lσ0 + lρ0 lσ − gρσ l · l0 − iρσαβ lα l0β
 
× pρ k σ + k ρ pσ − g ρσ p · k − iρστ λ pτ k λ . (4.66)

With the help of the identity

ρσαβ ρστ λ = −2(gατ gβλ − gαλ gβτ ) , (4.67)

we arrive at the simpler expression

g24 δ (4) (k − p − l − l0 ) d3 p d3 l0 d3 l 0
dΓ =
8(2π)5 Eµ ((k − l)2 − MW 2 )2 E E E (p · l)(k · l ) .
e νe νµ
 4.4 Muon Decay 15

The phase space integral over the momenta of the two neutrinos

d3 l0 d3 l (4) lα lβ0
Z
Iαβ (k, p) ≡ δ (k − p − l − l0 )  2 ,
Eνe Eνµ (k − l)2 − M 2 W
m
can be simplified since, to O( Me,µ
W
) we can neglect the momentum transfer
in the propagator, leaving

d3 l0 d3 l (4)
Z
Iαβ (k − p) = δ (k − p − l − l0 )lα lβ0 .
Eνe Eνµ
We now use another generic phase space integral formula

d3 p d3 q
Z
Iαβ ≡ p p δ (4) (k − p − q)pα qβ ,
m22 + p~ · p~ m21 + ~q · ~q
I
= (k 2 [k 2 − (m1 − m2 )2 ][k 2 − (m1 + m2 )2 ]gαβ
12k 4
+ 2[k 4 + k 2 (m21 + m22 ) − 2(m21 − m22 )]kα kβ ) , (4.68)

where I was defined in the previous section. Applying it to our case, m1 = 0

and m2 = me ≈ 0, we find

π
(p − k)2 gαβ + 2(p − k)α (p − k)β .

Iαβ (k − p) ≈ (4.69)
6
Now we write

d3 p = Ee2 dEe dΩ , (4.70)

and introduce the dimensionless variable

2Ee
x= , (4.71)
mµ
which varies between zero and one. Integrating over the angles, we obtain
the differential decay rate in the rest frame of the muon

g24 m5µ
 
dΓ 2 2x
= 4 x 1 − . (4.72)
dx 32 · 25 π 3 MW 3
The total decay rate is
16 STANDARD MODEL TREE-LEVEL PROCESSES

g24 m5µ 1
Z  
2 2x
Γ = 4 x dx 1 − ,
32 · 25 π 3 MW 0 3
e4 m5µ
= , (4.73)
192 · 25 π 3 sin4 θW MW
4

that is

G2F µ5
Γ(µ− → νµ + e− + ν̄e ) = . (4.74)
192π 3

4.4.1 PROBLEMS

A. Find the maximum kinematically allowed value of (k − l)2 . Estimate the

error in the rate of τ decay from neglecting the momentum transfer term in
the propagator

B.Verify the phase space formula Eq. (4.68).

C. When the decaying muon is at rest and polarized, calculate the depen-
dence of the decay rate on the direction of the electron.