Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
Lecture Notes
Johan Bijnens, Leif L¨onnblad and Torbj¨orn Sj¨ostrand
9 Cross sections, Decay widths and Lifetimes
From now on we will be less formal. We will make many approximations, especially when
it comes to the treatment of spin. Keeping track of the spin factors is in principle straight
forward, but it is quite cumbersome and is better left to specialized computer programs,
such as CompHEP. Throughout we will use natural units where ¯ h = c = 1.
We describe the probability of a certain process in terms of its cross section. It is deﬁned
in analogy with the situation where a beam of pointlike particles with a given number
per unit area, f, is hitting an extended target of transverse size σ, where the probability
of something a particle hitting the target is σf.
To obtain cross sections we need to compute the scattering matrix, S, describing the
transition from an initial state, [i` to a ﬁnal state [f`:
[f` = S[i`,
giving the amplitude
S
fi
= 'f[S[i`.
We then deﬁne the normalization of the matrix element according to
S
fi
= δ
fi
+ (2π)
4
δ
(4)
(P
f
−P
i
)(−i´
fi
)
¸
j=f,i
1
2E
j
.
• δ
fi
is the Kronecker delta function which is unity if i = f and zero otherwise. In
this case it describes what happens if nothing happens.
• P
i
and P
f
are the sum of the momenta of all particles in the initial and ﬁnal state
respectively
• δ
(4)
(P) is the Dirac delta function, in this case in each of the four components of
the momentum argument. The delta function is deﬁned according to
∞
−∞
h(x)δ(x)dx ≡ h(0)
for any reasonable function h. This ensures that fourmomentum is conserved.
• The product runs over all initial and ﬁnal state particles and is a normalization con
vention based on the separation of the spacetime dependence of the wavefunction
1
ψ = ue
ip·x
⇒ ¯ uu = 2m.
1
Kane eq. 5.47.
1
The probability of a given transition is now
d{ ∝ [´
fi
[
2
Where we have averaged over all nonobserved degrees of freedom (DoF) in the initial
state and summed over all DoF in the ﬁnal state. Eg. if we have an unpolarized fermion
in the initial state, each spin state (up and down) contributes whit a factor a half each.
Similarly if we have a fermion in the ﬁnal state where we do not measure the ﬁnal spin,
the two spin states are simply added.
Introducing an arbitrary volume, V = L
3
, we get the density of states of a particle by
assuming periodic boundary conditions:
e
ip
x
L
= 1 ⇒ p
x
= 2πn/L.
Looking in three dimensions we get
dn =
V d
3
p
(2π)
3
.
We now get the transition probability assuming ﬁxed momenta in the initial state and
a continuum in the ﬁnal state by squaring the matrix element and multiplying with the
appropriate phase space volumes
dΓ = V (2π)
4
δ
(4)
(P
f
−P
i
)[´
fi
[
2
¸
j=f,i
1
2E
j
V
¸
j=f
V d
3
p
j
(2π)
3
From which we can get the cross section by dividing by the ﬂux: dσ = dΓ/ﬂux (where
the ﬂux will contain the additional volume factors to cancel the dependence).
If we consider the scattering of two particles, A and B, to a ﬁnal state f we get the
relative ﬂux factor by looking eg. in the rest frame of B: v
A
/V which gives
v
A
=
(p
A
p
B
)
2
−m
2
A
m
2
B
E
A
m
B
.
In general we get the relative ﬂux factor depending on the relative velocity
v
r
=
(p
A
p
B
)
2
−m
2
A
m
2
B
E
A
E
B
.
We now get the expression for the cross section
dσ =
(2π)
4
δ
(4)
(P
f
−p
A
−p
B
)E
A
E
B
(p
A
p
B
)
2
−m
2
A
m
2
B
[´
fi
[
2
1
2E
A
1
2E
B
¸
j=f
d
3
p
j
2E
j
(2π)
3
,
where the dependence on V is gone, and where the the factor E
A
E
B
will cancel.
The phase space element is, in fact, Lorentz invariant. It can be derived from the four
momentum phase space element for a free particle
δ(p
2
−m
2
)Θ(p
0
)d
4
p = δ(E
2
−p
2
−m
2
)Θ(E)dEd
3
p
2
where the δfunction ensures that the particle is onshell and the Θfunction ensures that
we have positive energy. The Dirac deltafunction is a bit tricky when it comes to variable
substitution. Formally
δ(f(x)) =
¸
x
0
;f(x
0
)=0
δ(x −x
0
)
[f
′
(x
0
)[
Take eg. a konstant factor k and a test function h, and make a substitution y = kx:
δ(k(x −x
0
))h(x) =
δ(y −kx
0
)h(
y
k
)
dy
[k[
= h(
kx
0
k
)
1
[k[
=
h(x
0
)
[k[
Taylor expanding any function f around f(x
0
) = 0 we need only take into account the
ﬁrst nonzero term since everything else is negligible, so f(x) ≈ f
′
(x
0
)(x−x
0
) and we get
δ(f(x)) =
δ(x −x
0
)
[f
′
(x
0
)[
In our example we then get
δ(E
2
−p
2
−m
2
)Θ(E)dEd
3
p = δ((E +
p
2
+m
2
)(E −
p
2
+ m
2
)Θ(E)dEd
3
p =
δ(E −
√
p
2
+ m
2
)
E +
√
p
2
+m
2
dEd
3
p =
d
3
p
2E
.
In a scattering we deﬁne s = (p
A
+p
B
)
2
= E
2
cm
. If m
A
, m
B
≪E
cm
we have
s = (p
A
+ p
B
)
2
= m
2
A
+ 2p
A
p
B
+ m
2
B
≈ 2p
A
p
B
and
4
(p
A
p
B
)
2
−m
2
A
m
2
B
≈ 4p
A
p
B
≈ 2s
And we get the standard form of the scattering cross section
dσ = (2π)
4
δ
(4)
(P
f
−p
A
−p
B
)
[´
fi
[
2
2s
¸
j=f
d
3
p
j
2E
j
(2π)
3
,
In a 2 → 2 scattering, A + B → C + D we have in the restframe p
A
+ p
B
= p
C
+ p
D
=
(E
cm
, 0, 0, 0) and
δ
(4)
(p
C
+ p
D
−p
A
−p
B
)
d
3
p
C
E
C
d
3
p
D
E
D
= δ
(3)
(p
C
+p
D
)δ(E
C
+ E
D
−E
cm
)
d
3
p
C
E
C
d
3
p
D
E
D
=
δ(E
C
+ E
D
−E
cm
)
d
3
p
C
E
C
E
D
Shifting to polar coordinates and introducing dΩ = d(cos θ)dφ we get
δ(
p
2
C
+ m
2
C
+
p
2
C
+ m
2
D
−E
cm
)
p
2
C
d[p
C
[dΩ
C
E
C
E
D
Again, we can transform the δfunction,
δ([p
C
[ −p(E
cm
, m
C
, m
D
))
p
C

√
p
2
C
+m
2
C
+
p
C

√
p
2
C
+m
2
C
p
2
C
d[p
C
[dΩ
C
E
C
E
D
3
and we can set p = [p
C
[ = p(E
cm
, m
C
, m
D
) and Ω
C
= Ω
p
2
dΩ
C
p
E
C
+
p
E
D
E
B
E
C
=
pdΩ
E
C
+ E
D
=
pdΩ
√
s
And the cross section becomes
dσ = (2π)
4
δ
(4)
(P
f
−p
A
−p
B
)
[´
fi
[
2
2s
d
3
p
C
2E
C
(2π)
3
d
3
p
D
2E
D
(2π)
3
=
[´
fi
[
2
(2π)
2
2s
1
4
p
√
s
dΩ = [´
fi
[
2
p
32π
2
s
3/2
dΩ
9.1 Lifetimes, Resonance widths and Decay probability
Now, look at the decay of an unstable particle R → C + D. We get the transition
probability
dΓ =
(2π)
4
δ
(4)
(P
f
−p
R
)
2E
R
[´
fi
[
2
¸
j=f
d
3
p
j
2E
j
(2π)
3
.
Again, looking at the rest frame of the decaying particle we have E
R
= m
R
=
√
s and the
same transformation as before of the ﬁnal state particles gives
dΓ = [´
fi
[
2
p
32π
2
m
2
R
dΩ
and integrating over the solid angle (assuming unpolarized R)
Γ = [´
fi
[
2
p
8πm
2
R
Assume an exponential decay for a particle formed at t = 0:
[ψ(t)[
2
= [ψ(0)[
2
e
−t/τ
Θ(t)
This will give the transition probability
Γ = −
1
[ψ(t)[
2
d[ψ(t)[
2
dt
=
1
τ
Now, the time dependence of the wave function of a free particle in its rest frame is
ψ(t) = ψ(0)e
−iE
ψ
t
so we now get
ψ(0)e
−iE
0
t
e
−tΓ/2
= ψ(0)e
−iE
ψ
t
and an unstable particle can be seen to have complex energy E
ψ
= E
0
−iΓ/2. Let’s do a
Fourier transformation
˜
ψ(E) =
1
√
2π
∞
dt e
iEt
ψ(t) =
ψ(0)
√
2π
∞
0
dt e
i(E−(E
0
−iΓ/2)
=
iψ(0)
√
2π
1
E −E
0
+ iΓ/2
⇒[
˜
ψ(E)[
2
=
[ψ(0)[
2
2π
1
(E −E
0
)
2
+ Γ
2
/4
Heissenberg says: undetermined lifetime means undetermined energy and the probability
of ﬁnding an unstable particle with a given energy is a distribution around E
0
.
4
9.2 Scattering through a resonance
We have the decay of a resonance R → C + D prepared by the fusion of A + B → R at
time t = 0. We ﬁrst remember the scattering in a spherically symmetric potential using
partial wave expansion
f(θ) =
¸
l
(2l + 1)
e
2iδ
l
−1
2iκ
P
l
(cos θ).
where l is the angular momentum, δ
l
is the corresponding phase shift, κ = [p
C
[ = [p
D
[
and P
l
are the Legendre polynomials. The latter have the property
dΩP
l
P
l
′ =
4π
2l + 1
δ
ll
′
so squaring and integrating we get the cross section
σ =
4π
κ
2
¸
l
(2l + 1) sin
2
δ
l
For δ
l
(E) = π/2 we have a large cross section, corresponding to a resonance in the l
channel. We can write
e
2iδ
l
−1
2i
= e
iδ
l
sin δ
l
=
sin δ
l
cos δ
l
−i sin δ
l
=
1
cot δ
l
−i
and Taylor expanding cot δ
l
around δ
l
(E
R
) = π/2 we have
cot δ
l
(E) = cot δ
l
(E
R
) + (E −E
R
)
d cot δ
l
dE
E=E
R
+
We will identify
d cot δ
l
dE
E=E
R
with −2/Γ and get for the dominating lterm
f(θ) ≈ f
l
(θ) ≈
2l + 1
κ
P
l
(cos θ)
(E −E
R
)(−2/Γ) −i
=
(2l + 1)P
l
(cos θ)
κ
Γ/2
(E −E
R
) −iΓ/2
Squaring again and we get the cross section
σ
R
(E) =
4π
κ
2
(2l + 1)
Γ
2
/4
(E −E
R
)
2
+ Γ
2
/4
where the last factor is called the nonrelativistic BreitWigner distribution and can be
compared to the exponential decay above. This distribution is peaked around E
R
and
falls to half the peak value at E = E
R
±Γ/2. Note that we can produce a resonance also
away from its nominal rest energy. Also note that the expression is not relativistically
invariant. We can making it invariant by by noting that E +E
R
≈ 2m
R
≈ 2m in the rest
frame:
Γ/2
(E −E
R
) −iΓ/2
=
E + E
R
E + E
R
Γ/2
(E −E
R
) −iΓ/2
≈
mΓ
(m
2
−m
2
R
) −im
R
Γ
and with m
2
= s we get the relativistic BreitWigner shape of the cross section
σ(s) =
4π
κ
2
(2l + 1)
sΓ
2
(s −m
2
R
)
2
+ m
2
R
Γ
2
= 16π(2l + 1)
Γ
2
(s −m
2
R
)
2
+m
2
R
Γ
2
5
where the last step is true for massless particles where κ =
√
s/2.
If we have more than one decay channel the widths are additive, Γ
tot
=
¸
f
Γ
f
, and since
γ ∝ 1/τ the more decay channels the faster decay. We get the cross section for a particular
ﬁnal state
σ(A + B →R →f) = σ(A +B →R)
Γ
f
Γ
tot
.
Now we reverse the time and get a factor
Γ
i
Γ
tot
there as well, but here we average rather than
sum. Eg. assuming we have spin multiplicities 2s
A
+ 1, 2s
B
+ 1 and colour multiplicities
c
A
and c
B
for the initial state and the corresponding 2s
R
+ 2 and c
R
for the resonance
decaying to a particular ﬁnal state, f, we get the cross section
σ(A + B →R →f) ≈ 16π
¸
(2s
R
+ 1)c
R
(2s
A
+ 1)(2s
B
+ 1)c
A
c
B
¸
Γ
AB
Γ
f
(s −m
2
R
)
2
+ m
2
R
Γ
2
tot
A rough estimate of the cross section can be obtained by
σ ≤ σ(s = m
2
R
) ≈ 16π [∼ 1]
Γ
AB
Γ
tot
m
2
R
Γ
2
tot
≤
16π
m
2
R
Important cross section formulae
Work in the rest frame of the process.
Assume incoming masses m
A
, m
B
≈ 0.
Deﬁne s = E
2
cm
= (p
A
+ p
B
)
2
= (p
C
+ p
D
)
2
.
A +B →C + D : dσ = [´
fi
[
2
p
C
32π
2
s
3/2
dΩ
C
R →C + D : dΓ = [´
fi
[
2
p
C
32π
2
m
2
R
dΩ
C
=⇒ Γ = [´
fi
[
2
p
C
8πm
2
R
if isotropic
A + B →R →C + D : σ = 16π
¸
2s
R
+ 1
(2s
A
+ 1)(2s
B
+ 1)
c
R
c
A
c
B
¸
Γ
AB
R
Γ
CD
R
(s −m
2
R
)
2
+m
2
R
Γ
2
R
Approximate Feynman rules
external fermion (pair)
uu = 2E ⇒ u ∼
√
2E
external gauge boson
ǫ
µ
≈ 1
6
f f
γ
photon vertex
eQ
f
γ
µ
≈ eQ
f
f f
′
W
±
W
±
vertex
g
2
√
2
V
CKM
ff
′ for lefthanded fermions
0 for righthanded fermions
f f
Z
0
Z
0
vertex
g
2
cos θ
W
(T
3
f
−Q
f
sin
2
θ
W
) for lefthanded fermions
g
2
cos θ
W
(−Q
f
sin
2
θ
W
) for righthanded fermions
f f
g
gluon vertex
g
3
γ
µ
≈ g
3
for quarks
0 for leptons
external scalar
1
f f
h
0
Higgs–fermion vertex
g
f
=
g
2
m
f
2m
W
W W
h
0
Higgs–W vertex
g
W
ǫǫ
′
≈ g
W
= g
2
m
W
(except if m
H
≫m
W
)
7
Boson resonance propagator
(R = γ, W
±
, Z
0
, g, . . .)
1
m
2
−m
2
R
+ im
R
Γ
R
Feynman rules and Lagrange densities
Looking at the Lagrange density we can identify diﬀerent kinds of propagators and ver
tices.
Terms quadratic in a ﬁeld corresponds to a propagator, i.e. a particle created at some
point (source) propagating and destroyed at another point (sink). E.g. for a real scalar
we have
L =
1
2
δ
µ
φδ
µ
φ −m
2
φ
2
which gives the requirement on the ﬁeld
δ
µ
δ
µ
φ + m
2
φ = 0 ⇒ −p
2
φ −m
2
φ = 0
and the propagator
1
p
2
−m
2
which is the Fourier transform of the socalled Yukawa potential
−
1
4π
e
−mr
r
Ternary terms (with three ﬁelds) corresponds to threevertices. E.g. the interaction term
in the electromagnetic Lagrange density with fermion ﬁelds:
L
int
= −J
µ
A
µ
= Q
f
¯
ψ
f
γ
µ
ψ
f
A
µ
where ψ
f
= u
f
(p)e
−ipx
destroys a fermion, f (or creates the corresponding antifermion)
with momentum p and
¯
ψ
f
= ¯ u
f
(p)e
ip
′
x
creates f (or destroys
¯
f) with momentum p
′
.
Similarly A
µ
= ε
µ
e
ikx
creates (or destroys) a photon with momentum k. Momentum
conservation means e
−ipx
e
ip
′
x
e
ikx
= 1, and we are left with the external wave functions
¯ u
f
(p
′
), u
f
(p), ε
µ
and the vertex factor Q
f
γ
µ
.
Quartic terms will give rise to fourparticle vertices, which we will not discuss here.
9.3 The W width
From Kane eq. (7.32) we get the part of the Lagrange density for the fermion–W
±
vertex
L = . . . +
g
2
√
2
[¯ u
L
γ
µ
d
L
+ ¯ ν
e
γ
µ
e
L
] W
+
µ
+ h.c. + . . .
8
repeated for the other two families. This gives us the decay modes of the W
+
:
W
+
→e
+
ν
e
, µ
+
ν
µ
, τ
+
ν
τ
, u
¯
d, c¯ s, (t
¯
b)
The last one is forbidden for real W
+
by energy conservation. Note that the quarks come
in three colours. We can write down the matrix element for the ﬁrst decay
´(W
+
→e
+
ν
e
) =
g
2
√
2
¯ ν
L
γ
µ
e
L
ε
µ
≈
g
2
√
2
m
W
where we have used the standard normalization ¯ νe ∼ 2E ∼ m
W
and the polarization
of the photon, ε
µ
= O(1). The squared matrix element then becomes [´[
2
= g
2
2
m
2
W
/2,
while a full calculation would give [´[
2
= g
2
2
m
2
W
/3. This will give us the width
dΓ
eν
W
= [´[
2
p
32π
2
m
2
W
dΩ
⇒Γ
eν
W
=
g
2
2
m
W
48π
=
α
2
m
W
12
Summing up to the total width we get a factor three from the three generations of leptons
and another factor three for each of the two ﬁrst generation of quarks:
Γ
tot
W
= (1 + 1 + 1 + 3 + 3)
α
2
m
W
12
=
3α
2
m
W
4
.
Inserting numbers (m
W
≈ 80 GeV, α
2
= 1/30) gives Γ
W
≈ 2 GeV. Recent experimental
results: m
W
= 80.425 ±0.038 GeV, Γ
W
= 2.138 ±0.044 GeV.
9.4 The Z
0
width
As for the W we have the relevant term in the Lagrange density
L = . . . +
g
2
cos θ
W
¸
f
¯
f
L
γ
µ
f
L
(T
3
f
−Q
f
sin
2
θ
W
) +
¯
f
R
γ
µ
f
R
(−Q
f
sin
2
θ
W
)
Z
µ
+ . . .
⇒[´[
2
=
g
2
2
cos
2
θ
W
m
2
Z
(T
3
f
−Q
f
sin
2
θ
W
)
2
+ (−Q
f
sin
2
θ
W
)
2
and with α
2
= g
2
2
/4π and a factor 2/3 from spin factors as in the W case we get
Γ
f
Z
=
α
2
m
Z
cos
2
θ
W
N
f
C
(T
3
f
)
2
−2T
3
f
Q
f
sin
2
θ
W
+ 2Q
2
f
sin
4
θ
W
We have very precise measurements from LEP1: m
Z
= 91.1876 ±0.0021 GeV and Γ
Z
=
2.4925 ±0.0023 GeV.
We can also measure the partial widths and branching fractions of Z
0
into quarks (BR(u¯ u) ≈
0.12, BR(d
¯
d) ≈ 0.15) and charged leptons (BR(l
+
l
−
) ≈ 0.033). From this we can infer the
partial width or branching fraction into neutrinos (BR(ν¯ ν) ≈ 0.07) — the invisible width
— which means we can measure the number of (massless) neutrino species and hence also
the number of families:
n
ν
=
Γ
invisible
Z
Γ
ν
Z
= 2.983 ±0.009
9
10 Production of W
±
and Z
0
The cleanest way to produce real and onshell W and Z is by lepton fusion. At LEP1
(1989–94) (and at the SLC ) they used e
+
e
−
→ Z
0
→ fermions. The corresponding
production of W, eg. e
−
¯ ν
e
→ W
−
→ fermions is not technically feasible since we cannot
create a beam of neutrinos which is well enough focused. Instead at LEP2 (1996–2000)
they used the processes e
+
e
−
→ W
+
W
−
(exchanging a neutrino) and e
+
e
−
→ Z
⋆
/γ
⋆
→
W
+
W
−
(via an oﬀshell Z
0
or γ). The latter uses the fact that SU(2) is a nonabelian
theory so that we have terms in the Lagrange density which couples the W ﬁeld to itself.
Alternatively we can use the fusion of quarks into W/Z. For this we can collide hadrons
which consist of quarks and use processes such as u¯ u →Z
0
→fermions and u
¯
d →W
−
→
fermions. At LEP the processes involving W/Z are dominating if we tune our collision
energy to hit the resonance peak. Here, we have controll of the hadronic collision energy
but not of the energy of the colliding quarks. In addition, quarks interact much more
strongly with gluons (QCD), so any electro–weak process will drown in a lot of QCD
processes. Therefore the dominant decay of Z
0
into quarks cannot be used as a singnal.
Neither can the decay into neutrinos. However the decay into chagred leptons can be
used (but remember the branching ratio is small). Similarly for the W the decay into
quarks will drown in the background of QCD processes. The only way is to look for a
charged lepton and then look for the fact that the nonobserved neutrino will induce a
imbalance of the transverse momentum of all the detected particles. The W and Z was
ﬁrst discovered at a proton–antiproton collider at CERN in 1983.
To calculate the cross section of W/Z at hadron colliders we have to try to describe the
quarks (and gluons) inside the hadron. We know that a proton consists of three valence
quarks (uud) together with a number of gluons and virtual quark–antiquark pairs (sea
quarks). We call the quarks and gluons inside the proton partons.
To calculate the cross section we use the concept of a parton density: f
i/p
(x) is the prob
ability to ﬁnd a parton i inside a proton carrying a fraction x of the protons momentum.
Unfortunately we cannot calculate the f
i/p
(however some of its behavior can be calcu
lated), but we can infer some sumrules such as
¸
i
1
0
xf
i/p
(x)dx = 1 and we can measure
them. It turns out that the (logarithmic) density of valence quarks peaks around 0.15
and the total fraction of the protons momentum carried by the valence quarks is about
50%. Also it turns out that the gluon density increases dramatically at small x as does
the seaquark density (although it is typically much smaller than the gluon density).
Let’s look at the reaction p¯ p →proton remnants + [u
¯
d →W
+
]. We deﬁne
s = (p
p
+ p
¯ p
)
2
and we look in the rest frame of the hadronic collision, so
p
p
≈
√
s
2
(1; 0, 0, 1), p
¯ p
≈
√
s
2
(1; 0, 0, −1),
and the momentua of the colliding quarks are
p
u
= x
1
p
p
, p¯
d
= x
2
p
¯ p
.
Then we deﬁne
ˆ s = (p
u
+p¯
d
)
2
≈ 2p
u
p¯
d
= x
1
x
2
p
p
p
¯ p
≈ x
1
x
2
(p
p
+ p
¯ p
)
2
= x
1
x
2
s
10
We can write the cross section
σ(W
+
) =
dx
1
dx
2
f
u/p
(x
1
)f¯
d/¯ p
(x
2
)ˆ σ
u
¯
d→W
+(ˆ s)
We have basically already calculated the partonic cross section
ˆ σ(ˆ s) = 16π
¸
3
2 2
1
3 3
Γ
u
¯
d
W
Γ
f
W
(ˆ s −m
2
W
)
2
+ m
2
W
Γ
2
W
.
To make life simple we use the narrowwidth approximation where we note that
∞
−∞
dˆ s
(ˆ s −m
2
W
)
2
+ m
2
W
Γ
2
W
=
∞
−∞
dˆ s
ˆ s
2
+ m
2
W
Γ
2
W
=
1
m
W
Γ
W
∞
−∞
dx
1 +x
2
=
π
m
W
Γ
W
so that if we want to replace the distribution in ˆ s with a deltafunction we use
dˆ s
(ˆ s −m
2
W
)
2
+ m
2
W
Γ
2
W
→
πδ(ˆ s −m
2
W
)
m
W
Γ
W
and we get
ˆ σ(ˆ s) ≈
4π
3
Γ
u
¯
d
W
Γ
f
W
π
m
W
Γ
W
δ(ˆ s −m
2
W
).
With ˆ s = x
1
x
2
s we get the total cross section
σ(W
+
) =
dx
1
dx
2
f
u/p
(x
1
)f¯
d/¯ p
(x
2
)
4π
2
3
Γ
u
¯
d
W
Γ
f
W
m
W
Γ
W
1
s
δ(x
1
x
2
−m
2
W
/s)
=
4π
2
3
Γ
W
sm
W
BR(W →u
¯
d)BR(W →f)
dx
1
dx
2
f
u/p
(x
1
)f¯
d/¯ p
(x
2
)
1
x
1
δ(x
2
−
m
2
W
sx
1
)
=
4π
2
3
Γ
W
sm
W
BR(W →u
¯
d)BR(W →f)
m
2
W
/s
dx
x
f
u/p
(x)f¯
d/¯ p
(
m
2
W
sx
1
).
In the same way we can get the cross section for u¯ u →Z
0
, etc.
Experimentally we ﬁnd the Z
0
by looking for two opositely charged leptons, e
+
e
−
or
µ
+
µ
−
(the tau decay hadronically and is more diﬃcult to detect). These can, of course,
come from other sources, mainly from the weak decay of heavy quarks, but if we in each
event where we ﬁnd eg. an e
+
e
−
pair, calculate the quantity m
2
+−
= (p
e
+ + p
e
−)
2
≈
2E
e
+E
e
−(1 − cos θ
+−
) and look at the distribution of these, we should ﬁnd a (Breit–
Wiegner) peak around m
+−
≈ m
Z
.
For the W, things are a bit more complicated since we cannot calculate the corresponding
mass m
2
lν
≈ 2E
l
E
ν
(1 − cos θ
lν
) as we do not know E
ν
. We can, however, measure the
transverse momentum components of the neutrino, assuming this was the only particle
not detected: p
ν⊥
≈ [
¸
i
p
i⊥
[, where the sum runs over all detected particles (energy) in
an event. If we look at the distribution of m
2
⊥
= 2p
l
p
ν⊥
and we assume that all Ws decay
in a plane transverse to the beam axis, we would expect to ﬁnd a Breit–Wigner peak
around m
⊥
≈ m
W
. The fact that the decay is more or less evenly distributed in azimuth
angle the peak will become smeared and slightly shifted towards lower m
⊥
(a Jacobian
peak).
Then we have the spin properties of the W which can be checked. We know that W
+
only
couples to lefthanded (spin oposite to the direction of motion) fermions and righthanded
11
(spin parallel to the direction of motion) antifermions. Hence if a proton comes along
the positive zaxis and the antiproton along the negative, a W
+
will normally have been
produced from a u in the proton and a
¯
d in the antiproton and its spin will then be along
the negative zaxis. This means that when it decays into a e
+
ν
e
the e
+
cannot go along
the positive zaxis since its spin must be paralell to its direction of motion. Hence if we
look at the distribution in azimuth angle of e
+
in events believed to be W
+
events we
expect it to be peaked in the direction opposite to the incoming proton. This is exactly
what was found.
11 Measuring Standard Model parameters
The Standard Model of particle physics contains of the order of 20 parameters: the
masses of the fundamental particles, a number of mixing angles (later), and the couplings
constants
α
1
α
2
¸
⇒
α
EM
=
α
1
α
2
α
1
+α
2
sin
2
θ
W
=
α
1
α
1
+α
2
α
3
⇔ α
S
α
EM
is best measured with the quantum hall eﬀect (solid state) giving an impressive
accuracy:α
EM
(0) = 1/137.03599911(46).
The precision of sin
2
θ
W
is not as good, but it can be measured in many ways to ensure
consistency. One of the best measurements is from the width of the Z at LEP1:
Γ
Z
=
α
EM
m
Z
sin
2
θ
W
cos
2
θ
W
f(sin
2
θ
W
).
We can also use the simpler sin
2
θ
W
= 1−m
2
W
/m
2
Z
. The current best value is sin
2
θ
W
(M
Z
) =
0.23122(15).
α
S
is even trickier to measure. The best measurement is from the ratio between the cross
sections σ(e
+
e
−
→ Z
0
→ q¯ qg) and σ(e
+
e
−
→ Z
0
→ q¯ q). But we cannot detect neither
quarks or gluons, all we see is jets of hadrons, and jets are nototiously diﬃcult to deﬁne
in a way so that we can calculate this ratio to a reasonable precision. Current best value
is α
S
(m
Z
) = 0.1176(20).
The place to ﬁnd the best measurements of coupling constants, masses, widths, mixing
angles and other stuﬀ is in the Particle Data book http://pdg.lbl.gov/.
The muon lifetime
Let’s try to calculate the lifetime of a muon. This is a useful exercise as most weak deacys
of particles are calculated in the same way. We can draw the decay with a Feynman
diagram where the muon ﬁrst decays into a muonneutrino and a W
−
immediately followed
by the decay of the W
−
into an electron and an antielectronneutrino. Denoting the
momenta of the diﬀerent particles, µ(p) →ν
µ
(k) +[W
−
(Q) →e
−
(q) + ¯ ν
e
(k
′
)], we use the
12
Fynman rules to write down the matrix element
´=
g
2
√
2
¯ ν
µ
γ
λ
P
L
µ
1
Q
2
−m
2
W
g
2
√
2
(¯ eγ
λ
P
L
ν
e
)
with P
L
= 1 −γ
5
. The ﬁrst parenthesis is the vertex factor for the µν
µ
W vertex and the
second the vertex factor for the eν
e
W vertex. Sandwiched inbetween is the propagator
factor corresponding to the intermediate W. Since the maximum value of Q
2
∼ m
2
µ
is
much smaller than m
2
W
, the propagator is to a very good approximation 1/m
2
W
. If we
ignore the spinor complication and remember that the wavefunctions are basically just
given by the square root of the energy we get m
µ
for each of the vertices and the full
matrix element is
´=
g
2
2
2
m
2
µ
m
2
W
= 2
√
2G
F
m
2
µ
.
Now we can write down the decay rate diﬀerential in the momenta of the incoming and
outgoing particles:
dΓ
µ
=
(2π)
4
δ
(4)
(k + q + k
′
−p)
2m
µ
[´[
2
d
3
k
2E
k
(2π)
3
d
3
q
2E
q
(2π)
3
d
3
k
′
2E
k
′ (2π)
3
=
1
(2π)
5
8G
2
F
m
4
µ
2m
µ
1
2
δ
(4)
(k + q + k
′
−p)
d
3
k
2E
k
d
3
q
2E
q
d
3
k
′
2E
k
′
,
where the extra factor one half comes from the fact that only lefthanded muons are
involved. We want to move to the rest frame of the muon, so p = (m
µ
; 0, 0, 0) so the
diﬀerential can be written
δ
(3)
(k +q +k
′
)δ(E
k
+ E
p
+ E
k
′ −m
µ
)
d
3
k
2E
k
d
3
q
2E
q
d
3
k
′
2E
k
′
,
and if we ignore the masses of the decay products we get
δ([k[ +[q[ +[k +q[ −m
µ
)
d
3
kd
3
q
8 [k[ [q[ [k +q[
.
We now go to polar coordinates (with k relative to a given axis and with q relative to k)
δ([k[ +[q[ +
k
2
+q
2
+ 2[k[[q[ cos θ
q
−m
µ
)
k
2
d[k[dΩ
k
q
2
d[q[dΩ
q
8[k[[q[
k
2
+q
2
+ 2[k[[q[ cos θ
q
Here we can always do the integral over Ω
k
(⇒4π) and over the azimuth angle of q around
the k direction (⇒2π), leaving us with
δ([k[ +[q[ +
k
2
+q
2
+ 2[k[[q[ cos θ
q
−m
µ
)
4π[k[d[k[ 2π[q[d[q[d cos θ
q
8
k
2
+q
2
+ 2[k[[q[ cos θ
q
Integrating also over cos θ
q
, the delta function gives
k
2
+q
2
+ 2[k[[q[ cos θ
q
= m
µ
−[k[ −[q[
and a jacobian
k
2
+q
2
+ 2[k[[q[ cos θ
q
[k[[q[
13
Simplifying miraculously to
π
2
d[k[d[q[
and we get the diﬀerential deacy width
dΓ
µ
=
2G
2
F
m
3
µ
(2π)
5
π
2
d[k[d[q[
which can be integrated ([k[ < m
µ
/2, [q[ < m
µ
/2, m
µ
−[k[ −[q[ < m
µ
/2) giving the total
width
Γ
µ
=
G
2
F
m
5
µ
128π
3
or, if you do the spin and kinematics correctly,
Γ
µ
=
G
2
F
m
5
µ
192π
3
.
So, by measuring Γ
µ
and m
µ
we can get a good value for g
2
(and, indirectly, sin
2
θ
W
).
25 Top and tauneutrino
The top quark was discovered in 1995 at Fermilab outside of Chicago in USA. The mass
has been measured to be m
t
= 174.2 ± 2.0 ± 2.6 GeV. The way they found it was to
use the fact that the top will decay (before it hadronizes) to a b and a W
+
. Top quarks
are pairproduced on p¯ p collisions through u¯ u → g
⋆
→ t
¯
t → bW
+
¯
bW
−
. The W can be
detected in the usual way with hightransversemomentum leptons and missing transverse
momentum (a bit trickier since the Ws have a substantial transverse momentum and the
missing momentum is the sun of the two neutrinos). The bquarks will hadronize into
B mesons which have a very long lifetime, giving rise to secondary vertices around a
millimeter from the collision vertex.
Even before the top quark was directly observed there was plenty of evidence that it must
exist (if the Standard Model is correct). One way is to consider the decay of the Z
0
into
b
¯
bpairs:
Γ
b
¯
b
Z
∝ (T
3
b
−Q
b
sin
2
θ
W
)
2
+ Q
2
b
sin
2
θ
W
The ﬁrst term is for lefthanded bquarks and here the T
3
b
is −1/2 if the b is part of a
doublet, but 0 if it is a singlet. Alternatively one may consider the forward–backward
asymmetry as in Kane.
In the same way there was indirect evidence for ν
τ
(also from Γ
invisible
Z
) before it was
directly observed at the DONUT experiment at Fermilab in 2000. At DONUT they
constructed a beam believed to consist of tauneutrinos. If so, a ν
τ
may exchange a W
with a nuclear target, producing a τ which is detected as it decays τ
−
→ν
τ
e
−
¯ ν
e
a couple
of tenths of a millimeter away from the primary collision vertex. They only observed four
such events, but that was enough.
14
12 Accelerators
We study the fundamental particles by colliding them at high energies. To get them
to high energies we use accelerators. These use electrical ﬁelds to accelerate beams of
charged particles. They also use magnetic ﬁelds to bend and focus the beams.
In a linear accelerator the energies achievable depends on the length, giving approximately
5 MeV/m (but new techniques are being developed to increase this).
Alternatively the beams are bent around in storage rings which enables repeated accel
eration. The only setback here is the synchrotron radiation which means that the beam
loses energy δE ∝ γ
4
/R
2
∝ E
4
/(m
4
R
2
) (larger ring or heavier particles give less energy
loss). Hence for e
±
beams this is a real problem. For (anti) proton beams there is, on
the other hand, the problem that we need very strong magnets to bend the beam, and
also that the proton is composite, which means that the quarks (gluons) which interacts
only carries around 10% of the energy.
Major laboratories
• CERN – Geneva, Switzerland, European organization including Sweden. www.cern.ch
• DESY – Hamburg, Germany, National facility with strong international component.
www.desy.de
• FNAL/Fermilab – Chicago, US, National laboratory owned by the DoE and oper
ated by a collection of Universities, with strong international component. www.fnal.gov
• SLAC – Stanford, US, National laboratory owned by the DoE and operated by
Stanford University. www.slac.stanford.edu
• KEK – Japan, National laboratory.
Major accelerators
name beams location period energy main
(GeV) physics
LEP1 e
+
e
−
CERN 89–95 91 Z
0
SLC e
+
e
−
SLAC 89–98 91 Z
0
LEP2 e
+
e
−
CERN 95–00 –209 W
±
HERA ep DESY 91–07(?) 30900 QCD
Tevatron p¯ p FNAL 85–07(?) 2000 top
RHIC Au Long Island 99– 2 200 A QGP
LHC pp CERN 07(?)– 14 000 H, BSM(?)
In the future:
name beams location period energy main
(GeV) physics
ILC e
+
e
−
DESY(?) 2012(?)– 500–1000 H, BSM(?)
CLIC e
+
e
−
CERN 2020(?)– –3000 H, BSM(?)
µµC µ
+
µ
−
FNAL(?) 2030(?)– –5000 BSM(?)
VLHC pp CERN/FNAL(?) 2030(?)– –200 000 BSM(?)
15
The performance of an accelerator is, besides the collision energy, determined by the
luminosity, L. The number of events we can expect for a given process with cross section,
σ, is determined by the integrated luminosity:
N = σ
dtL
We can use the accelerated beam to hit a stationary target. The luminosity is then given
by the number of particles hitting the target per unit time, n, and the density, ρ, and
length, l, of the target. A beam of transverse size, σ drills a hole in the target of volume
V = lσ where we can ﬁnd n
′
= V ρ particles in the target, giving us a luminosity
L = n n
′
/σ = nρl
For a ﬁxedtarget machine the (invariant) collision energy is approximately given by
(p
beam
+ p
target
)
2
≈
2E
beam
m
target
, so the energy increases only like the square root
of the beam energy.
For colliding beams, on the other hand we get a collision energy of
√
2E
+
E
−
, which is
much more economic, if you are able to focus the beams against eachother properly. If we
have a storage ring with n bunches of k particles per bunch going round at a frequency f
and the cross section of the interaction area is A we get a luminosity
L = fnk
2
/A
(note error in the book).
13 Detectors
At LHC we expect of the order of a hundred particles being produced per collision. We
wish to detect as many of them as possible so we want a good solidangle coverage of
detectors Ω →4π.
For charged particles it is typically possible to track them as they interact electrically with
detector elements. If we also have a magnetic ﬁeld we can determine their momentum
by the way they bend. For neutral particles we can use calorimeters to detect the energy
deposition in the detector parts. Typical examples of detectors are:
• Microvertex detectors: (CCD) silicon pixels (20 −50µm)
2
layered very close to the
interaction point. Detects charged particles.
• Wire chambers: ∼ 1m in radius, ﬁlled with gas and wires under high voltage. The
passing of a charged particle induces a discharge between the wires which can be
measured.
• Cherenkov counters: Particles which traverses a medium at a speed larger than the
speed of light in that medium will emit Cherenkov radiation (similar to a supersonic
bang). The photons are detected and gives information about the mass of the
particle.
16
• Electromagnetic calorimeter: Eg. lead glass and photon detectors. Also a neutral
particle will interact with nuclei in the detector giving ﬂashes of photons.
• Hadronic calorimeter: Hadron interacts strongly with matter (typically iron) and
we get ionization (kicking out charged hardons from the nucleus) of the detector
which can be detected.
• Muon Chambers: Any charged particle which can penetrate the calorimeters, typi
cally meters thick, is most likely to be a muon.
Nice pictures
No detector is perfect. For identiﬁed particles there is always uncertainties due to non
perfect eﬃciencies and purities. For measurements of energy and momenta there is always
an uncertainty, typically increasing with energy ∆E ∝
√
E.
14 Finding the Higgs
The discovery of Higgs would complete the Standard Model and has therefore been the
main objective of many accelerators in the past. So far noone has succeeded. LEP1 didn’t
see anything, but indirect measurements favoured a light Higgs. LEP2, as it was about to
close down, had a couple of events which could be Higgs, but the signiﬁcance turned out
to be too low. They did, however set the best limit so far on the mass: m
h
≥ 113 GeV
(95% conﬁdence). Even the Tevatron had something which looked like a Higgs event,
but the signiﬁcance was even lower. If the Higgs exists it will be discovered at the LHC
(although it may take some time as it is not easy).
To understand how we can produce and detect the Higgs, we start by looking at the
relevant terms in the Lagrange density and ﬁnd that there are fermion–fermion–Higgs,
W–W–Higgs and Z–Z–Higgs vertices, but no vertices containing gluons or photons.
For the fermion vertex we have the coupling
g
f
=
g
2
m
f
2m
W
so we can produce a Higgs by annihilating a ferminon–antifermion pair, and conversely
the Higgs can decay into a ferminon–antifermion pair. We can approximate the matrix
element by
[´[
2
≈ (g
f
m
h
)
2
c
f
2
where c
f
is three for quarks and 1 for leptons and 2 comes from the sum over spin states
times a fudge factor. From this we get the partial width
Γ
h
f
¯
f
≈
p
f
m
2
h
dΩ
32π
2
[´[
2
= α
2
c
f
8
m
2
f
m
2
W
m
h
.
Hence, we should use heavy fermions if we want to produce Higgs, and when the Higgs
decays it will predominately do so to heavy fermions. In particular, considering the
hierarchy of fermion masses, the Higgs will mainly decay to the heaviest possible fermion
17
(with m
f
< m
h
/2). Note that close to threshold there will be a kinematical damping of
the decay width
β
2
= (1 −
4m
2
f
m
2
h
)
3/2
For the W–W–Higgs vertex we have to calculate a bit more since we cannot make the
approximation u ∼
√
2E for the W wavefunctions, mainly because the W can be longi
tudinally polarized. The calculation in is performed in Kane (21.2) and the result is
[´[
2
≈
g
2
m
4
h
2m
2
W
giving the patial width
Γ
h
WW
≈
α
2
m
3
h
16m
2
W
.
Similarly we get
Γ
h
ZZ
≈
α
2
m
3
h
32m
2
W
.
So we get a width of the Higgs which is proportional to the mass
3
. This means that a
light Higgs is very narrow, while a heavy Higgs is extremely broad (in TeV units we get
Γ ∼ 0.5m
3
h
).
In e
+
e
−
→ Z
0
annihilation we can look for higgs using Z
0
→ Z
0
h, where either the
incoming (LEP2) or the outgoing (LEP1) Z
0
is oﬀshell (reduced cross section). Detecting
the ﬁnal Z
0
decaying into leptons we can calculate the missing mass, m
2
miss
= (p
e
+ +p
e
− −
p
l
+ −p
l
−)
2
, where we should ﬁnd a peak corresponding to the Higgs mass. Te higgs will
then mainly decay into b
¯
b (jets) and we can study the angular distribution to conﬁrm
the scalar nature of the produced resonance. At LEP2 they found some events, but not
signiﬁcan enough.
At the LHC we can produce the Higgs in the same way (u¯ u → Z
0
→ Z
0
h), but there
we do not have access to the missing mass method, so we have to see the Higgs directly.
However, there is a HUGE background of bottom jets at LHC to completely drown any
such Higgs signal.
If the Higgs is lighter than 2m
W
the dominant decay mode is to bottom quarks which are
diﬃcult to separate from the background. So we need to look at less likely decay channels
wich are easier to ﬁnd.
One such channel is h → γγ. Since the photon doesn’t couple directly to the Higgs this
decay goes through a fermion loop (typically top) or a W loop. The branching ratio is
tiny, ∼ 10
−3
, but the background of (isolated) high transverse momentum photons is even
smaller.
If the Higgs heavier than 2m
Z
we can look for the decay h → Z
0
Z
0
→ l
+
l
−
l
′+
l
′−
, the
goldplated signal. Again the combined branching ratio into Z and subsequently into
leptons is small, ∼ 10
−3
, but the background is virtually nonexistent.
At the LHC we can also produce Higgs via gluon–gluon fusion. As for the γγ decay,
this proces includes a topquark loop and the partial width is small. However, there is a
huge amount of gluons in the colliding protons, making this one of the main production
channels for light to mediummass Higgs.
18
We can also produce Higgs with W
+
W
−
fusion, where the Ws are radiated oﬀ quarks in
the protons.
Nice pictures
What if we don’t ﬁnd the Higgs at the LHC? There are some cross sections in the Standard
Model, in particular the longitudinal W
+
W
−
→W
+
W
−
cross section, which would grow
with energy in the absence of the Higgs and will eventually violate unitarity (∼ probability
of interaction larger than unity). So, if the Higgs does not exist, or is too heavy (m
h
≫
1 TeV) some other mechanism beyond the Standard Model must enter the scene. People
have made calculations “proving” that if we do not ﬁnd the Higgs at the LHC, we will
ﬁnd something else.
15 Quarks, Conﬁnement, . . .
Strong interactions are quite diﬀerent from electro–magnetic ones. In QED we have
the potential (of eg. an electron around a proton in an atom or around a positron in
positronium):
V (r) = −
α
EM
r
+ k
1
L S + k
2
µ
e
µ
p
Where the last two terms are the ﬁnestructure and hyperﬁne splitting terms respectively.
For strong interactions (QCD) we expect to have approximately the same thing,
V (r) = −
4
3
α
S
r
+k
′
1
L S +k
′
2
µ
q
µ
¯ q
where the Casimir factor 4/3 comes from group theory, the fact that we have three charges
(r, g, b) and an (in principle arbitrary) normalization. So we expect to have bound states
of quarks (hadrons) and, if enough energy is added, free quarks.
But because gluons selfinteract (contrary to photons) we get conﬁnement. While the
electromagnetic ﬁeld between two electrically charged particles extend out to inﬁnity,
the colourﬁeld between two colourcharged particles is compressed into a ﬂuxtube with
approximately constant transverse size d
⊥
≈ 0.7 fm – a string. This will give us
V (r) = −
4
3
α
S
r
+ κr,
where κ is a constant energy per unit length ∼ 1 GeV/fm ∼ 10 metric tons per meter.
Hence we can never give a quark enough energy to escape from a bound stata and become
free. What happens instead if a quark is given enough energy is that the ﬁeld (string) is
stretched to the point where enough energy is stored in the ﬁeld to make a virtual q¯ qpair
real. The colours of this pair may then screen the endpoint charges causing the string
to break. The original quark is still conﬁned but not to the same hadron as before. If
the energy is high enough, the string will break in several places giving many hadrons,
where the ones created close to the original quark in the string carries most of the quarks
momentum, while the other hadrons have a smaller fraction. In this way, hadrons are
typically emitted in a narrow cone around the original quark — a jet.
According to Heissenberg if a q¯ q is produced inside a transverse size d
⊥
will have an
uncertainty in transverse momenta of p
⊥
∼ 1/d
⊥
≈ 0.3 GeV.
19
Carful analysis indicates that diﬀerent breakups along the string are causually discon
nected, this means that the average “distance” between the breakups must remain the
same independent of the “length” of the string. It turns out that a reasonable deﬁnition
of “distance” is rapidity,
y =
1
2
ln
E +p
z
E −p
z
= ln
e + p
z
m
⊥
since the diﬀerence between two rapidities is invariants under Lorentz transformations
along the string direction (z). Hence we expect dn/dy to be constant, where n is the
number of breakups, or the number of hadrons. Now, if we have a string between a q
and a ¯ q in an e
+
e
−
annihilation with a collision energy
√
s, the maximum rapidity of a
hadron is y
max
∼ ln
√
s/m
⊥
and the average number of hadrons are
N
tot
∝ ln
√
s
'm
⊥
`
+ const
16 Light hadrons
All hadrons are coloursinglets. Note that it is not enough that a q¯ qpair in a meson
has opposite colour charge. Compare to the spin of a twoelectron system, where the
individual spins can combine in four diﬀerent ways, either as a spinone system
S = 1,
S
z
= 1 : [ ↑↑`
= 0 :
1
√
2
([ ↑↓` +[ ↓↑`)
= −1 : [ ↓↓`
or a spin 0 system
S = 0, S
z
= 0 :
1
√
2
([ ↑↓` −[ ↓↑`)
So there are two way of getting a “spin 0” system. In the same way a [r¯ r` is colourless,
but not a singlet as colour will appear under a phase rotation. Instead a meson is given
by
q¯ q =
1
√
3
([r¯ r` +[g¯ g` +[b
¯
b`)
(where the + is a phase convention)
Similarly we have baryons with
qqq =
1
√
6
([rgb` −[rbg` +[brg` −[bgr` +[gbr` −[grb`) =
1
√
6
ε
ijk
[q
i
q
j
q
k
`
Note that for any SU(N) group, we will get “baryons” with N “quarks”. In QCD we
can, in principle, have more complicated hadrons with quark content q¯ qq¯ q, q¯ qqqq etc. The
latter are usually called pentaquarks and have recently been discovered for the ﬁrst time
(uudd¯ s →nK
+
) corresponding to a nK
+
binding energy of O(100 MeV). This should be
compared to the typical binding energy in nuclear physics where the binding energies are
O(10 MeV).
Another option for coloursinglet states is the glueball gg, ggg, . . . This is allowed by the
gluon self interaction (we have no γγ bound states). However these states will mix with
ordinary meson states and are diﬃcult to observe. So far there are only indications.
20
In principal we can expect states with a valence gluon (eg. q¯ qg). Again there is mixing
with normal mesons and no clear evidence for these hermaphodites has been found.
We classify hadrons by:
• ﬂavour contents (gives charge, decay patterns etc.)
• mass
• internal spin S. Mesons S = 0, 1. Baryons S =
1
2
,
3
2
.
• internal angular momentum L = 0, 1, 2, . . .
• total spin J,
¯
J =
¯
S +
¯
L. (Often also called spin.)
• Radial excitation n.
• Relativistic corrections and mixing (makes everything quite messy).
It turns out to be very messy to calculate the masses of the hadrons, but we can try to
use the atom as an analogy to get
m =
¸
i
m
i
+ k
¸
i<j
'¯ µ
i
¯ µ
j
` =
¸
i
m
i
+ k
¸
i<j
'
¯
S
i
¯
S
j
`
m
i
m
j
where k ∝ [Ψ(0)[
2
and is ≈ m
2
u
640 MeV for mesons and ≈ m
2
u
200 MeV for baryons. This
formula does not include the binding energy, instead this is included by using constituent
quark masses: m
u
≈ m
d
≈ 330 MeV, m
s
≈ 500 MeV, m
c
≈ 1600 MeV and m
b
≈
5000 MeV. For the light quarks these masses are very diﬀerent from the mass of a freely
propagating quark.
Lets look at a π
+
= u
¯
d. It is a spin0 state so
¯
S
1
¯
S
2
=
1
2
(
¯
S
2
−
¯
S
2
1
−
¯
S
2
2
) =
1
2
(S(S + 1) −S
1
(S
1
+ 1) −S
2
(S
2
+ 1)) = −
3
4
so
m
π
= 2 330 −
3
4
640 = 180 MeV
(OK, it’s not perfect).
In Kane there are lists of meson and baryon states. For the mesons the main ones are the
vector mesons and the pseudoscalars (scalars since spin0, but pseudo, since odd parity).
Note that the ﬂavourdiagonal states are special since they can mix:
u¯ u, d
¯
d, s¯ s ⇒π
0
, η, η
′
If you want to see more states: pdg.lbl.gov
We always have the standard charge (C) and parity (P) symmetries. These can always
be applied to hadron state, although all states are not eigenfunctions of these symmerties.
The absolute parity of a state is not completely deﬁned, rather it is relative. If we have
a state of two particles a and b with a parities P
a
and P
b
with an engular momentu L
(assuming it is an eigenstate to L)
P[ab; L` = [P(a)P(b); L`(−1)
L
= P
a
P
b
(−1)
L
[ab; L`
21
where the (−1)
L
comes from the spherical harmonics, and P
a
P
b
= −1 for a fermion–anti
fermion pair and +1 for a boson–antiboson pair.
For charge conjugation the mesons are not in general eigenstates: C[a
¯
b; L, S` = ±[¯ ab; L, S`,
but for the ﬂavourdiagonal states we have
C[a¯ a; L, S` = [¯ aa; L, S` = (−1)(−1)
L
(−1)
S+1
[a¯ a; L, S` = (−1)
L+S
[a¯ a; L, S`
There are three diﬀerent ways a hadron can decay:
Strong Decay will always dominate if it is possible, ie. if the quark numbers and energy
can be conserved. eg. ∆
++
→pπ
+
(uuu →uud+
¯
du). The ∆
++
has a lifetime τ ≈ 10
−23
s
⇔Γ ≈ 100 MeV (diﬃcult to calculate).
Strong decays respect strong isospin conservation, which is not due to a gauge symmetry,
rather it is due to the fact that the strong force (gluon) cannot tell the diﬀerence between
a u and a d quark. It is only an approximate symmetry due to the (accidental?) fact that
m
u
≈ m
d
. Strong isospin symmetry would mandate that Γ(∆
++
→pπ
+
) = Γ(∆
+
→nπ
+
)
up to spin factors. Also we can predict the branching ratios of the K
⋆+
which can decay
into K
+
π
0
or K
0
π
+
(u¯ s →u¯ s +u¯ u and u¯ s →d¯ s +u
¯
d). If we describe the strong isospin
with the total spin and the projection on the u/d states we have
[
1
2
,
1
2
` =
1
√
3
[
1
2
,
1
2
`[1, 0` ±
2
√
3
[
1
2
, −
1
2
`[1, 1`
so BR(K
0
π
+
) = 2BR(K
+
π
0
) =
2
3
.
Electromagnetic decay Will dominate if possible and no strong decays are allowed.
Conserves quark number but not necessarily strong isospin eg. π
0
→γγ ([1, 0` →[0, 0`[0, 0`).
Also we have eg. Σ
0
→ Λγ which includes a spin ﬂip in the light diquarks ((ud)
1
s →
(ud)
0
s). The lifetimes are 10
−20
−10
−10
s and are typically approximately calculable.
Weak decay does not conserve quark numbers as in π
+
→µ
+
¯ ν
µ
(u
¯
d →W
+
→µ
+
¯ ν
µ
).
These are only important if no strong or electromagnetic decays are possible. The lifetimes
are 10
−10
−10
3
s and are typically approximately calculable.
17 Heavy Quarks
In November 1974 the October revolution occurred in particle physics. One found an
extremely narrow resonance in e
+
e
−
→µ
+
µ
−
and also in pp (u¯ u →µ
+
µ
−
). The resonance
was called J by one experiment and ψ by another. Today we call it J/ψ and it has a mass
of 3.1 GeV and a width of only 70 keV. The only possible explanation was the bound
state of a new heavier quark–antiquark ﬂavour  the charm quark.
It turns out that m(J/ψ) < 2m
D
(where D = c
¯
d, c¯ u) so strong decay is impossible and
the particle is longlived. Also the mass of the charm mass is ≈ m(J/ψ)/2 ≈ 1.6 GeV,
22
so we have nonrelativistic motion of the quark inside the meson and it is like a c¯ c atom
which means we can do “spectroscopy”.
Sure enough they found excited states
3
S
1
(1s) = J/ψ,
3
S
1
(2s) = ψ
′
,
3
S
1
(3s) = ψ
′′
,
3
P
J
= χ
c
. They eventually also found a pseudoscalar
1
S
0
= η
c
. Note that as soon as
the mass of these states bcome larger that 2m
D
(the eﬀective ionization energy), strong
decays are possible and the resonances become very wide.
The D mesons come in diﬀerent ﬂavours depending on the light quark: D
+
= c
¯
d, D
0
= c¯ u
and D
+
s
= c¯ s, and they can only decay weakly, ie. they are rather longlived. We can
calculate the decays in the same way as for the muon c → s[W
+
→ µ
+
¯ ν
µ
] and similarly
for W
+
→e
+
¯ ν
e
and W
+
→u
¯
d, and we get
Γ
c
∼
G
2
F
m
5
c
192π
3
(1 + 1 + 3)
In 1977 the b quark was discovered (more or less the same story, but it was not predicted in
the same way as the c quark). However, here the corresponding decay b →t[W
−
→ν
µ
µ
−
]
is kinematically forbidden due to the large mass of the top. Exchanging the top with
a charm quark works, however, but it is heavily suppressed by the CabbiboKobayashi
Mascawa matrix element and b is therefore much more longlived as compared to the
charm quark. This is why we can observe it with a secondary vertex, cτ ∼ 0.4 mm.
18 Deeply inelastic scattering
Rutherford found the atomic nuclei by scattering α particles oﬀ gold atoms. In 1968 at
SLAC the did a similar thing, ﬁring electrons with a ﬁxed energy at protons and measure
how they scattered: e
−
(k)p(P) → e
−
(k
′
)X(P
′
). We interpret this as an exchange of a
virtual photon with momentum q = k −k
′
. The ﬁnal momentum of the scattered proton
system is P
′
= P + q = P + k − k
′
. The virtuality of the exchanged photon propagator
is denoted
Q
2
= −q
2
= −(k −k
′
)
2
= 2kk
′
−2m
2
e
≈ 2EE
′
(1 −cos θ) = 4EE
′
sin
2
θ/2.
We also denote the invariant mass swuared of the hadronic ﬁnal state
W
2
= (P + q)
2
= m
2
p
+ 2Pq −Q
2
.
If W
2
= m
2
p
we have elastic scattering. If W
2
> m
2
p
we have inelastic scattering, the
proton “breaks up”. If W
2
≫m
2
p
we have deeply inelastic scattering.
We can sketch the matrix element of this scattering
´∝ e(
¯
k
′
γ
µ
k)
1
Q
2
e(
¯
P
′
. . . P)
we don’t measure the hadronic ﬁnal state so we write the last factor as e(
¯
P
′
γ
µ
P)F(Q
2
),
ie. as an elestic scattering, but with a form factor F(Q
2
) which can be identiﬁed with
the Fourier transform of the spatial charge distribution in the proton. We use the nor
malization F(0) = 1 and for a smeared charge distribution we expect F(Q
2
) → 0 as
Q
2
→∞.
23
Now, if the proton consists of pointlike constituents, we can interpret the results as if we
are elastically scattering on these partons. If the partons has a momentum fraction x of
the protons momentum,
ˆ
P = xP, we are looking at the process e
−
(k)q(
ˆ
P) →e
−
(k
′
)q(
ˆ
P
′
),
where
ˆ
P
2
=
ˆ
P
′2
≈ 0 for nearly massless partons. Hence we have
ˆ
P
′2
= (xP +q)
2
= x
2
P
2
+ 2xPq + q
2
≈ 2xPq −Q
2
= 0 ⇒ x =
Q
2
2Pq
≡ x
Bj
the Bjorken x ≈ Q
2
/(Q
2
+ W
2
).
Here we can write down the matrix element of the partonic scattering
´≈ e(
¯
k
′
γ
µ
k)
1
−q
2
e
i
(
¯
ˆ
P
′
γ
µ
ˆ
P) ≈ ee
′
ˆ s
−Q
2
where ˆ s = (k − xP)
2
≈ x(k + P)
2
= xs. In the rest frame of the partonic scattering we
have
Q
2
= −(k −k
′
)
2
≈ 2
√
ˆ s
2
√
ˆ s
2
(1 −cos
ˆ
θ)
so
dQ
2
=
ˆ s
2
d(cos
ˆ
θ) =
ˆ s
2 2π
d
ˆ
Ω
and with Q
i
= e
i
/e and α = e
2
/4π we get the cross section
dˆ σ
i
dQ
2
=
πα
2
Q
4
Q
2
i
.
Now we can do the same thing as when we calculated W production in p¯ p collisions, ie.
we use the parton density, f
i
(x), the probability of ﬁnding a parton, i, with momentum
fraction, x, and get the full diﬀerential cross section
dσ
i
dQ
2
dx
≈
πα
2
Q
4
Q
2
i
f
i
(x).
If we now deﬁne the structure function
F
2
(x) =
¸
i
Q
2
i
xf
i
(x)
we can write the cross section
dσ
dQ
2
dx
=
πα
2
Q
4
F
2
(x)
x
2(1 + (1 −y)
2
)
where the last factor is obtained from proper spin counting with y = Pq/Pk, the fraction
of the electron momentum taken by the photon.
In an alternative experiment they used a “beam” of muon neutrinos which scattered oﬀ
a proton by exchanging a W
+
producing a µ
−
which could be measured. Here we can do
the same calculation as before with some important modiﬁcations. First the propagator
particle is now massive, so we will get
1
Q
2
⇒
1
(m
2
W
−q
2
)
2
≈
1
m
4
W
24
Also, not all quark species can be involved, only d →u, ¯ u →
¯
d, s →c and ¯ c → ¯ s. So
¸
i
Q
2
i
xf
i
(x) ⇒xf
d
(x) +xf
¯ u
(x) +. . .
Hence, we can try to disentange the diﬀerent quark distributions by combining νp and ep
DIS in diﬀerent ways.
The picture of the proton becomes a bit more diﬃcult if we consider virtual ﬂuctuations.
A quark can emit a gluon and reabsorb it again. Also a gluon can emit a gluon, and a
gluon can split into a virtual q¯ q pair. The valence quarks are surrounded by a cloud of
virtual partons. On the other hand we have the virtual photon, it is shortlived with a
small wave function centered along the direction of the electron. We can approximate the
transverse size of the photon ∆r ∼ 1/∆p
⊥
∼ 1/Q. The higher the virtuality, the ﬁner
details in the proton we can see and also the more details of the cloud of virtual partons
around the valence quarks. This is called scaling violations. If we increase Q
2
we can see
that a parton at high x actually consist of two (or more) partons at lower x, so the xf(x)
is depleted at high x and increased at low x.
The end result is that the parton densities are Q
2
dependent, f
i
(x) ⇒ f
i
(x, Q
2
). The
‘Q
2
dependence is only logarithmic, and it is calculable
δf
(
x, Q
2
)
δ ln Q
2
∝
α
s
2π
¸
−P
i
f
i
(x, Q
2
) +
¸
j
1
x
dz
z
f
j
(z, Q
2
)P
j→i
(
x
z
)
¸
Here the ﬁrst term corresponds to the loss of a parton at x being resolved into two with
lower momentum fractions, while the second is the gain of a parton at a higher momentum
fraction being resolved into one at x. The P functions (where P
i
=
¸
j
dzP
i→j
(z)) are
called (Altarelli–Parisi) splitting functions.
19 e
+
e
−
physics
We have the process e
+
e
−
→ γ
⋆
→ f
¯
f (where the γ
⋆
→ Z
0
or a mixture of the two at
higher energies. The matrix element is easy to write down (for s << m
2
Z
)
´= e
e
(¯ eγ
µ
e)
1
s
e
f
(
¯
fγ
µ
f)
giving the cross section
σ =
4π
3
Q
2
f
α
2
s
One of the main successes of the quark model is the experimental conﬁrmation of the
ratio
R
had
=
σ(e
+
e
−
→hadrons)
σ(e
+
e
−
→µ
+
µ
−
)
= 3
¸
q;2m
q
<
√
s
Q
2
q
demonstrating the charges of the quarks and the fact that they come in three diﬀerent
colours. It is also possible to conﬁrm the spin1/2 nature of the quarks by looking at the
angular distribution of jets
dσ(e
+
e
−
→q¯ q)
dΩ
∝ 1 + cos
2
θ
25
At DESY they also managed to conﬁrm the existence and spin1 nature of gluons by
studyng threejet events e
+
e
−
→ q¯ qg. By stydying fourjet events e
+
e
−
→ q¯ qgg and
e
+
e
−
→q¯ qq
′
¯ q
′
it has been possible to conﬁrm the selfcoupling among gluons.
20 Running couplings
The couplings in the standard model Lagrangian are in principle impossible to measure,
since they are bare couplings, while in any measurement we never have eg. a completely
“bare” electron as it is always surrounded by an electromagnetic ﬁeld which will contain
virtual ﬂuctuations.
Hence, if we want to calculate an observable R (eg. a cross section or a decay width) we
get a perturbative series
R = α
n
(R
0
+ R
1
α + R
2
α
2
+ . . .)
R
0
is typically a ﬁnite number calculable with using a Feynman tree diagram. R
1
, however,
typically involves a divergent integral over a momentum in a loop diagram. But we know
we can make a measurement of R (at some scale q
2
) so we should be able to calculate
R(q
2
) = R
obs
(q
2
), even though we cannon directly measure α. The trick is to redeﬁne α
so that the divergencies in R
1
are cancelled, α →α(q
2
).
The way we redeﬁne α is easiest to see if we consider a eγ vertex e(k) →γ(q)e(k
′
). Here
the second order diagram contains a loop where the photon splits into an e
+
(p)e
−
(q −p)
pair which then annihilates into a photon again. In the matrix element we will get the
two terms
e¯ u(k
′
)γ
µ
u(k)ε
µ
−
d
4
p
(2π)
4
[e¯ u(k
′
)γ
µ
u(k)]
1
q
2
[e¯ u(p)γ
µ
u(p −q)][e¯ u(p −q)γ
λ
u(p)]
(p
2
−m
2
e
)((p −q)
2
−m
2
e
)
ε
λ
We here recognize the three propagators and we get a minus sign due to the fact that this
is a fermion loop (a boson loop would give a plus sign). We can simplify this to
e¯ u(k
′
)γ
µ
u(k)ε
µ
(1 −I(q
2
))
where the integral I(q
2
) is dimensionless an can be simpliﬁed to
I(q
2
) =
α
3π
∞
m
2
e
dp
2
p
2
−
2α
π
1
0
dx x(1 −x) ln
1 −
q
2
x(1 −x)
m
2
e
where the ﬁrst term is clearly divergent. To continue we need to introduce an ultraviolet
cutoﬀ, Λ
2
, for this integral (which we will send to inﬁnity later). If Q
2
= −q
2
≫ m
2
e
we
have
ln
1 −
q
2
x(1 −x)
m
2
e
≈ ln
Q
2
m
2
e
and we have
I(Q
2
) ≈
α
3π
ln
Λ
2
m
2
e
−
α
3π
ln
Q
2
m
2
e
=
α
3π
ln
Λ
2
Q
2
.
26
Now, if we can have one loop, we can have another, but that will just give the same
integral squared
´= α´
0
1 −
α
3π
ln
Λ
2
Q
2
+
α
3π
ln
Λ
2
Q
2
2
+ . . .
¸
¸
= ´
0
α
1 +
α
3π
ln
Λ
2
Q
2
→´
0
α
obs
(Q
2
)
Assume we have α
0
, some “original” α which we “normalize” by measuring α(µ
2
)
α(µ
2
) =
1
1
α
0
+
1
3π
ln
Λ
2
µ
2
⇒
1
α
0
=
1
α(µ
2
)
−
1
3π
ln
Λ
2
µ
2
then we can calculate
α(Q
2
) =
1
1
α(µ
2
)
−
1
3π
ln
Λ
2
µ
2
+
1
3π
ln
Λ
2
Q
2
=
1
1
α(µ
2
)
+
1
3π
ln
µ
2
Q
2
=
α(µ
2
)
1 +
α(µ
2
)
3π
ln
µ
2
Q
2
which is quite ﬁnite. We can now send Λ → ∞ (while the bare α
0
→ 0) and we can
express the coupling as measured at one scale in terms of the coupling at another scale,
and the coupling is running.
The physical interpretation is that the charge around an electron is screened, but for
larger Q
2
, we probe the electron at smaller distances where it is less screened and the
eﬀective coupling increases.
So far we only had electrons in the loop, but we can also have muons and quarks as long
as their squared mass is less than Q
2
(in principle we can also have heavier fermions in
the loop, but their contribution is tiny. It is easy to see that we now get
α(Q
2
) =
α(µ
2
)
1 −N
α(µ
2
)
3π
ln
Q
2
µ
2
with
N = n
l
+ 3
2
3
2
n
u
+ 3
1
3
2
n
d
The same procedure can be repeated for QCD, looking at a quark–gluon vertex and
inserting quarkloops. We then get
α
s
(Q
2
) =
α
s
(µ
2
)
1 −
n
f
2
α
s
(µ
2
)
3π
ln
Q
2
µ
2
but we will also have gluon loops in the gluon propagator, and since this is a boson loop
it comes with the oposite sign. Doing the math we get
α
s
(Q
2
) =
α
s
(µ
2
)
1 +
α
s
(µ
2
)
12π
(33 −2n
f
) ln
Q
2
µ
2
Again we have a coupling which runs, but in the other direction. For large Q
2
the coupling
decreases and we get asymptotic freedom. On the other hand for small Q
2
the coupling
diverges and we expect perturbation theory to break down. The picture is that if a red
quark emits a red–antiblue gluon and turns blue it can be viewed as if the original red
charge is smeared out. The closer we look the more smeared out it is.
27
Now let’s deﬁne Λ
QCD
as the scale where α
s
diverges ie. where
α
s
(µ
2
)
12π
= −(33 −2n
f
) ln
Λ
2
QCD
µ
2
We then get
α
s
(Q
2
) =
12π
(33 −2n
f
) ln
Q
2
Λ
2
QCD
As Q
2
increases, more and more quarks will contribute to the running, and to have a
continuous coupling one usually have Λ
QCD
(n
f
).
Conventionally we present measurements of α
s
at the scale m
2
Z
.
14 Low Energy and NonAccelerator Experiments
The Standard Model works extremely well. Although we have not found the Higgs (yet)
there is really no hard evidence of any physics Beyond the Standard Model (BSM). Of
course every particle physicists dream is to ﬁnd something which does not ﬁt into the
standard model (others, like myself, tries to improve our understanding of the SM so
that we have a chance to ﬁnd deviations). To do this one can either try to crank up the
energy in the collision experiments, hoping to ﬁnd new particles, which necessarily will
be associated with BSM physics. Building larger and larger accelerators is, however, very
expensive and diﬃcult. Alternatively one can use “natural” accelerators which seem to
exist in the universe, which constantly bombard us with high energy particles (these are,
however, rather diﬃcult to controll).
There is one thing one can do without high collision energies, and that is to search for
lowenergy processes which are very rare or even forbidden in th SM.
One example thing is to look for eﬀects of massive neutrinos in neclear βdecay (later).
Examples of decays which are forbidden in the standard model are leptonfamilynumber
violating processes such as µ
−
→ e
−
γ (current limit BR < 5 10
−11
) or µ
−
→ e
−
e
+
e
−
(BR < 10
−12
). Also some hadron decays are forbidden: K
0
→ µe (BR < 3 10
−11
),
K
+
→π
+
eµ (BR < 2 10
−10
). Especially forbidden is proton decay (τ
p
> 10
32
y).
22 Mixing angles
If we for a moment limit ourselves to two families, we can write the charged current in
the Lagrangian as
J
µ
ch
=
¯ u ¯ c
γ
µ
P
L
d
s
.
Here we assume that the quark ﬁelds are both mass eigenstates and weak eigenstates.
But his is not necessarily true. In fact it turns out to be false.
One analogy is to consider a light source which emits transversely polarized light, which
then traverses a medium where positive helicities travels faster than negative helicities,
28
and is ﬁnally detected after a tranverse polarizer. Hence the production and detection
eigenstates are the transverse polarizations (ε
x
, ε
y
) while the propagating eigenstates (the
“mass” eigenstates) are the circular polarizations (ε
+
, ε
−
).
Now, it turns out that the mass and weak eigenstates of quarks are not the same. What
enters into the quark–W vertex is the weak eigenstates, but what we want to have in
matrix element, when we calculate cross sections for “free” propagating quarks, is the
mass eigenstates. Let’s call the weak eigenstates of the downtype quarks
d
′
s
′
. These
can be tranformed by a unitary transformation to the mass eigenstates and vice versa.
d
′
s
′
L
= V
d
s
L
The unitary matrix can be written completely generally as follows
V =
cos θe
iα
sin θe
iγ
−sin θe
−iγ
cos θe
−iα
And we get the charged current
J
µ
ch
=
¯ u ¯ c
γ
µ
P
L
d
′
s
′
=
¯ u ¯ c
γ
µ
P
L
V
d
s
.
In principle we could do the same thing with the uptype quarks and get both a V
u
and
V
d
, but these would always show up together in any electroweak vertex as V
u†
V
d
which
itself is a unitary matrix, so we only need to consider the downtype quarks.
The phases in V above can be absorbed in the quark states and we are left with
V =
cos θ sin θ
−sin θ cos θ
Expressed in the mass eigenstates we now have the current
J
µ
ch
= ¯ uγ
µ
P
L
d cos θ + ¯ uγ
µ
P
L
s sin θ − ¯ cγ
µ
P
L
d sin θ + ¯ cγ
µ
P
L
s cos θ+
And we get the new couplings to the W by multiplying by g
2
/
√
2. Note that the “Cabbibo”
angle is a completely free parameter of the Standard Model. If nature has chosen the value
0, there would have been no mixing between families and the squark (and the lightest
strange hadrons) would be stable. Note that this does not change the neutral current at
all since, just looking at the lefthanded s and d part of current
J
µ
neu
=
¯
d ¯ s
V
†
T
L
3
−Q
f
sin
2
θ
W
V
d
s
=
¯
d ¯ s
T
L
3
−Q
f
sin
2
θ
W
d
s
it will be unchanged since V
†
V = 1.
We can generalize this to three generations
J
µ
ch
=
¯ u ¯ c
¯
t
γ
µ
P
L
V
¸
¸
d
s
b
¸
.
29
V is now a 3 3 unitary matrix which means that it can be described by nine diﬀerent
parameters. As for the twoﬂavour matrix we can get rid of some of these by redeﬁning
the phases of the quark wave functions. But we are left with four parameters which are all
completely free and need to be measured. Three of these can be expressed as real angles
similar to the Cabbibo angle, but there is stil one relative phase which we cannot get rid of
in this “CabbiboKobayashiMascawa” mixing matrix. This means that we get a camplex
current, and also complex terms in the Hamiltonian, W
µ
J
µ
ch
, which furthermore means
that the the theory is not completely invariant under timereversal transformations.
It turns out that the absolute values of the matrix elements coupling the third generation
to the others are fairly small ([V
cb
[ ≈ 0.03 and [V
ub
[ ≈ 0.003). Since the b quark cannot
decay into a top, the dominant decay mode is b → cW
−
where the width is suppressed
by [V
cb
[
2
. This is diﬀerent from the corresponding lepton, the τ which can simply decay
into a neutrino. One way of measuring [V
cb
[ is therefore to compare
Γ
b
Γ
τ
≈
9
5
[V
cb
[
2
m
5
b
m
5
τ
where the 9/5 comes from the number of decay modes available from the W.
24 CP Violation
There are three important discrete symmetries in nature, the charge (C), parity (P) and
time (T) reversal symmetries. Now if we look at the standard model we immediately ﬁnd
that P is explicitly broken in the coupling to the W, since P will change a righthanded
particle to a lefthanded one, and while the latter couples to the W, the former does not.
In the same way it is clear that C is violated, since a lefthanded particle couples to the
W while a lefthanded antparticle does not. However, we could expect the theory to be
invariant under the simultaneous CP transformation. There is a theorem saying that any
quantum ﬁeld theory is invariant under CPT transformation, but it turns out that the
standard model is not quite invariant under CP and therefor not under T transformations.
The reason for this “CP violation” is the remaining the phase in CKM matrix which we
couldn’t get rid of by redeﬁning the phases in the quark wavefunctions. (Note that this
could never happen if there was just two generations).
The ﬁrst place where CP violation was measured was in the decay of neutral kaons. If
we look at K
0
= d¯ s and
¯
K
0
= s
¯
d, neither of these are CP eigenstates. But if CP is
conserved we can create two combinations which are eigenstates: K
L
= K
0
−
¯
K
0
and
K
S
= K
0
+
¯
K
0
, where the former is odd and the latter is even under CP transformation.
Now, these combinations have diﬀerent decay modes (eg. K
L
→ πππ and K
S
→ ππ),
and it turns out that the K
L
is much more linglived that K
S
(hence the names long
(≈ 5 10
−8
s) and short (≈ 9 10
−11
s)).
But if CP is slightly violated, the K
L
and K
S
are not necessarily mass eigenstates. If
they were, we could have a “beam” of K
0
, where we expect all K
S
will decay and quickly
leaving a pure K
L
beam. But if K
L
is not a mass eigenstate, but made up from two states
with diﬀerent masses, travelling at diﬀerent velocity, after a while we will have diﬀerent
mixtures of these states, which will give us back a component corresponding to K
S
.
30
If CP was exactly conserved, the decay K
L
→ π
−
e
+
ν
e
would be exactly the same as
K
L
→π
+
e
−
¯ ν
e
. But it turns out that we can measure a small asymmetry:
Γ(K
L
→π
−
e
+
ν
e
) −Γ(K
L
→π
+
e
−
¯ ν
e
)
Γ(K
L
→π
−
e
+
ν
e
) −Γ(K
L
→π
+
e
−
¯ ν
e
)
= 0.00333 ±0.00014
So CP, and therefore also T, is violated in the Standard Model.
We can expect other sources of CP violation besides the phase in the CKM matrix. One
way of checking this is to measure also CP violation in the B
0
¯
B
0
, if that is inconsistent
with the K
0
¯
K
0
CP violation, there may be other sources.
If we study CP violation, we may eventually understand why there are so many more
particles tha antiparticles in the Universe.
24 Neutrino masses and oscillations
In the standard model the neutrinos are typically assumed to be massless. And upuntil a
few years ago they seemed to be. The experimental limits on their masses were, however,
not extremely strict. Eg. from betadecay of tritium we had m(ν
e
) < 3 eV, the decay
π
+
→ µ
+
ν
µ
gave the limit m(ν
µ
) < 0.19 MeV, and m(ν
τ
) < 18 MeV came from τ →
ν
τ
+ nπ.
There is, however, no `a priori reason we know about for the neutrinos to be massless (or
not). The masses could arise in (at least) two diﬀerent ways.
If there exists a righthanded neutrino (and lefthanded antineutrino) we would have a
coupling to the higgs ﬁeld: g
f
¯ νφν
R
and a (Dirac) mass term m
D
¯ νν
R
. So how do we see
a righthanded neutrino? Well, it has zero charge, it would have T
3
= 0 and no colour,
hence the only way it would interact is with the Higgs and via gravity. So we will never
see it directly.
The second way we could have neutrino masses is if the neutrino, just as the photon and
Z
0
, was its own antiparticle. In that case we could get terms in the Lagrangian looking
like m
M
¯ ν
L
ν
c
L
, where ν
c
L
is righthanded. But this would give us ν
L
↔ ¯ ν
R
, which violates
fermion number. This hasn’t been seen. Yet. One place where it could show up is in
neutrinoless doublebeta decay:
(A, Z) →(A, Z + 1) +e
−
+ ¯ ν
e
→(A, Z + 1) +e
−
+ ν
e
→(A, Z + 2) +e
−
e
−
So far, the nonobservation of this sets the limit m
M
< 0.2 eV.
If the neutrinos are massive, there is a possibility that the masseigenstates are not the
same as the weak eigenstates, just like what happened for their cousins, the downtype
quarks. Denoting the mass eigenstates ν
i
we have
¸
¸
ν
e
ν
µ
ν
τ
¸
= U
¸
¸
ν
1
ν
2
ν
3
¸
Now a neutrino would be produced (eg. in π
+
→µ
+
ν
µ
) in a weak eigenstate at time 0:
ν
µ
(0) = U
µ,1
ν
1
(0) +U
µ,2
ν
2
(0) +U
µ,3
ν
3
(0)
31
but then we will have a propagation of the mass eigenstates, so that
ν
µ
(t) = U
µ,1
ν
1
(0)e
iE
1
t
+ U
µ,2
ν
2
(0)e
iE
2
t
+U
µ,3
ν
3
(0)e
iE
3
t
Let’s simplify and only use two generations, which means that we can describe the mixing
matrix with only one angle, α
ν
e
ν
µ
=
cos α sin α
−sin α cos α
ν
1
ν
2
Now, if we start out with a pure ν
e
we have
ν
e
(0) = ν
1
(0) cos α + ν
2
(0) sin α
and
ν
µ
(0) = −ν
1
(0) sin α + ν
2
(0) cos α = 0
but it is the mass eigenstates which propagates, so after some time t we will have
ν
µ
(t) = −ν
1
(0)e
−iE
1
t
sin α + ν
2
(0)e
−iE
2
t
cos α
and there is a probability that the ν
e
has turned into a ν
µ
['ν
e
(0)[ν
µ
(t)`[
2
= sin
2
α cos
2
α
e
−iE
1
t
−e
−iE
2
t
= sin
2
2α sin
2
(t(E
2
−E
1
)/2)
where
E
2
−E
1
=
E
2
2
−E
2
1
E
1
+E
2
≈
m
2
2
−m
2
1
2E
=
∆m
2
2E
≈
2m∆m
2E
=
∆m
γ
where the latter can be seen as a boosted version of the fact that also a neutrino at rest
has no deﬁnite weak eigenstates.
Including also the third family we can write the probability for oscillation from family α
to β
P(α →β) = δ
αβ
− 4
¸
i>j
ℜ(U
⋆
αi
U
βi
U
αj
U
⋆
βj
) sin
2
1.27∆m
2
ij
E/L
+ 2
¸
i>j
ℑ(U
⋆
αi
U
βi
U
αj
U
⋆
βj
) sin
2
2.54∆m
2
ij
E/L
where we have scaled things so that L should be given in km, E in GeV and ∆m
2
ij
in eV
2
.
Note that there is also some oscillations induced when neutrinos traverses matter. This
is because neutrinos interacts, however weakly, with matter with Z
0
exchange. While Z
0
exchange is exactly the same for electron and muonneutrinos, electron neutrinos can
also react with the electrons in matter through W exchange, ν
e
e
−
→e
−
ν
e
. This will give
an eﬀective mass of the neutrino which is diﬀerent from the mass in vacuum. We can
write
i
d
dt
ν
e
ν
µ
= U
E
1
0
0 E
2
U
†
ν
e
ν
µ
+
√
2G
F
N
e
0
0 0
ν
e
ν
µ
which will give us
P(ν
e
→ν
µ
) =
sin
2
2θ
W
2
sin
2
1.27
W∆m
2
L
E
32
with
W
2
= sin
2
2θ +
√
2G
F
N
e
2E
∆m
2
−cos 2θ
2
The signal for this is eg. the appearance of a ν
µ
where we only expect ν
e
or the disappear
ance of ν
e
. The latter has been suspected for a long time from the fact that only half of
the expected number of neutrinos from the sun were found – but nobody really trusted
the models for neutrino production in the sun enough to make a deﬁnite statement.
Now there is ample evidence for neutrino oscillations and hence also for neutrino masses (or
rather mass diﬀerences). The evidence comes both from solar and atmospheric neutrinos
as well as from ’man made’ neutrino sources (reactors and accelerators).
We can study atmospheric neutrinos. These are produced when charged particles from
space hits the atmosphere and produces mainly pions which, in turn, produces mainly
muons (together with ν
µ
) which in turn decays (giving another ν
µ
and a ν
e
). These are
typically much more energetic than solar neutrinos. Now if we detect these and look at the
direction they come from, if they come from above, they have traveled a short distance
in the atmosphere, while if they come from below they have been traveling ∼ 10
4
km
through the earth. Hence we can vary L by varying the polar angle and we should see
oscillations. And we do. By looking at both ν
e
and ν
µ
it is found that there is a big
eﬀect of ν
µ
disappearing, not to ν
e
but to ν
τ
. From this one gets a mass diﬀerence which
is larger than the mass diﬀerence responsible for the ν
e
↔ ν
µ
oscillations. The current
numbers are approximately
∆m
2
sol
≈ 8 10
−5
eV
2
∆m
2
atm
≈ 3 10
−3
eV
2
θ
sol
≈ 34
◦
θ
atm
≈ 37 −53
◦
26 Open issues
So there you have it. The standard model. It works very well and we hate it! Why?
• There are too many free parameters. Twelve fermion masses, eight mixing param
eters, three couplings and the higgs ﬁeld parameters (µ, λ) ⇔ (m
h
, m
Z
). Wouldn’t
it be much nicer if we had a theory where these could be predicted?
• Many things seems arbitrary. Why are there three generations. Why are the left
handed fermions in SU(2) doublets and the righthanded in singlets?
• Why do we just have SU(3) SU(2) U(1)?
• Why is there charge quantization? In principle Y in U(1) could be anything.
Then there is the obvious problem that gravity does not ﬁt in anywhere. And the fact that
we have such huge ratios of similar parameters, eg. m
W
/m
Pl
∼ 10
−17
, m
e
/m
W
∼ 10
−5
,
m
ν
/m
e
∼ 10
−7
. There is the fact that baryonic matter only makes up a fraction of the
33
matter of the universe and that matter only makes up a fraction of the energy of the
universe (the cosmological constant).
As if this isn’t enough, there is a problem with the higgs mass. Masses, just as couplings
depend on the scale due to loop corrections and need to be renormailized (looking just at
QED)
m →m
1 +
α
3π
∞
m
2
e
dp
2
p
2
→m
1 +
α
3π
ln
Λ
2
m
2
e
For the higgs the renormalization turns out to be worse
δm
2
h
∝
∞
m
2
h
∝ Λ
2
If there is no additional structure for particles until we reach the Planck scale, the Higgs
mass we observe at electroweak scales must be the tiny diﬀerence between two huge
numbers (the hierarchy problem). We don’t like that.
27 Grand Uniﬁcation
Just as the electroweak symmetry is spontaneously broken by the Higgs ﬁeld, U(1)
Y
SU(2)
L
→ U(1)
EM
, we could try to imagine that the whole structure of the standard
model is a spontaneously broken version of a higher symmetry.
The simplest such Grand Uniﬁed Theory (GUT) was invented by Georgi and Glashow
and is now excluded by date. It is however instructive to look at it anyway. It assumes
that there is only one symmetry group which is spontaneously broken into the standard
model,
SU(5) →SU(3) SU(2) U(1)
We then have the fermions in a quintet
¸
¸
¸
¸
¸
¸
¸
¸
¸
¯
d
r
¯
d
b
¯
d
g
¸
e
−
ν
e
¸
L
where now
¯
d
L
=
¯
d
R
is an weak SU(2) singlet. The quarks are still associated with
the SU(3)t colourtriplet and the leptons are still in the weak SU(2) doublet. In the
fundamental representation of an SU(N) group the generators are traceless N N ma
trices, where the diagonal generator will give the charge, which reuires
¸
Q
i
= 0, so that
3Q
d
+ Q
e
= 0 and Q
d
= −1/3 (and also Q
u
= 2/3) and we have charge quantization,
Q
p
+ Q
e
= 0.
So where are the other quarks and leptons? Well the quintet above is actually the anti
quintet of the group. We also have the antisymmetric 5 5 decuplet (cf. SU(3) have
34
triplet quarks, but also octet gluons)
1
√
2
¸
¸
¸
¸
¸
¸
¸
¸
¸
0 ¯ u
b
−¯ u
g
−¯ u
b
0 ¯ u
r
¯ u
g
−¯ u
r
0
¸
−u
r
−d
r
−u
g
−d
g
−u
b
−d
b
u
r
u
g
u
b
d
r
d
g
d
b
0 −e
+
e
+
0
¸
L
Note that there are ten independent ﬁelds here. Then we have the quintet of the right
handed ﬁelds
(d
r
, d
g
, d
b
, e
+
, ¯ ν
e
)
R
and a corrsponding antidecuplet for the righthanded uptype quarks and electrons. For
the decuplets we have Q
ij
= Q
i
+ Q
j
so eg. Q
u
= Q
4,1
= Q
4
+ Q
1
= Q
d
+ Q
e
.
The SU(5) group has 5
2
− 1 = 24 generators, which will give us 24 gauge bosons, lets
denote them
¸
¸
¸
¸
¸
¸
¸
¸
g
ij
−
2B
√
30
δ
ij
¯
X
r
¯
Y
r
¯
X
g
¯
Y
g
¯
X
b
¯
Y
b
X
r
X
g
X
b
Y
r
Y
g
Y
b
¸
W
3
√
2
+
3B
√
30
W
+
W
−
−
W
3
√
2
+
3B
√
30
¸
¸
We can write the running of the couplings in the following way
1
α
i
(M
2
)
=
1
α
i
(µ
2
)
+
b
i
4π
ln
M
2
µ
2
where we have b
3
= 11 −4n
F
/3, b
2
= 22/3 −4n
F
/3 and b
′
1
= −4n
F
/3 (b
′
1
is related with
the hypercharge rather than the electric charge). We assume that there is actually only
one coupling at som large uniﬁcation scale, M
2
GUT
, but the broken symmetry gives them
diﬀerent values at lower scales. Then we would have eg.
1
α
3
(µ
2
)
+
b
3
4π
ln
M
2
GUT
µ
2
=
1
α
2
(µ
2
)
+
b
2
4π
ln
M
2
GUT
µ
2
So we can in principle calculate this scale
ln
M
2
GUT
µ
2
=
6π
11
1
α
2
(µ
2
)
−
1
α
3
(µ
2
)
Inserting experimental data we will ﬁnd M
GUT
∼ 10
18
GeV, but things do not exactly
match up.
If SU(5) is the true symmetry group at high scales the coupling g
5
would be the only
copuling in the theory. All other couplings, g
1
, g
2
, g
3
, would be derived from the breaking
of SU(5) into the SU(3) SU(2) U(1). In particular we should be able to derive the
ratio between g
1
and g
2
which is governed by θ
W
.
Let’s look at the covariant derivative of SU(5)
D
µ
= ∂
µ
−ig
5
T
a
U
µ
a
35
And Pick out the parts relevant to the electroweak sector using
B
µ
= A
µ
cos θ
W
+ Z
µ
sin θ
W
W
µ
3
= −A
µ
sin θ
W
+ Z
µ
cos θ
W
D
µ
= ∂
µ
−ig
5
(T
3
W
µ
3
+T
1
B
µ
+. . .) = ∂
µ
−ig
5
sin θ
W
(T
3
+cot θ
W
T
1
)A
µ
+. . . = ∂
µ
−ieQA
µ
+. . .
From this we can identify e = g
5
and Q = T
3
−cot θ
W
T
1
≡ T
3
+cT
1
. Now, for any repre
sentation, R, of a group we have orthogonality and equal normalization of the generators
T
a
, so that
Tr
R
T
a
T
b
= N
R
δ
ab
So
TrQ
2
= Tr(T
3
+cT
1
)
2
= (1 −c
2
)TrT
2
3
since TrT
2
3
= TrT
2
1
and
sin
2
θ
W
=
1
1 +c
2
=
TrT
2
3
TrQ
2
=
0 + 0 + 0 +
1
2
2
+
1
2
2
1
3
2
+
1
3
2
+
1
3
2
+ 1 + 0
=
3
8
Now, this is fairly far from 0.23, but remember sin
2
θ
W
is a ratio between couplings, and
couplings are running and, hence, so is sin
2
θ
W
. And evolving down to the weak scale we
are not too far oﬀ.
The 24 gauge bosons included the eight gluons and the W
±
, Z
0
and γ. The other twelve
are X,
¯
X, Y and
¯
Y which all come in three diﬀerent colours. Sandwiching the gauge
boson matrix between the fundamental representations,
¯
5A5,
¯
10A5, etc. we get terms in
the Lagrangian corresponsing to new vertices, eg.
X →uu, X →e
+
¯
d, Y →ud, Y →
¯
d ¯ ν
e
, Y →e
+
¯ u
and we conclude that X and Y has charges 4/3 and 1/3 respectively. The problem with
this is that the proton can decay:
p = uud →uY →u ¯ ue
+
→π
0
e
+
The anology with weak muon decay
Γ
µ
=
G
2
F
m
5
µ
192π
2
=
α
2
2
384π
m
5
µ
m
4
W
we get something like
Γ
p
∝ α
2
5
m
5
p
m
4
Y
A proper calculation using m
Y
∼ m
GUT
∼ 10
15
GeV will give τ
p
∼ 10
31±2
years. The
current limit p →π
0
e
+
is τ
p
> 10
33
years.
Careful analysis tells us that the simplest GUT, based on SU(5) is excluded by date, but
one could easily imagine higher symmetry groups. An example is SO(10) which includes
a righthanded neutrino an extra Z boson and much more. Most of these suﬀer from
baryonnumber nonconservation and other maladies, and currently GUTs are not very
popular.
36
28 Sypersymmetry
Postulate there being a symmetry between fermions and bosons, with an operator Q
changing one into the other
[b` = Q[f` and [f` =
¯
Q[b`
but leaving any other quantum number as it is. The transformation is actually deﬁned in
terms of an algebra where
¦Q,
¯
Q¦ = Q
¯
Q+
¯
QQ = 2σ
µ
P
µ
where P
µ
= i∂
µ
is a translation. In some sense the transformation of a boson into a fermion
is the square root of a translation. (Remember that
√
−1 = i and
∂
µ
∂
µ
= γ
µ
∂
µ
.) We
also have
¦Q, Q¦ = ¦
¯
Q,
¯
Q¦ = 0 and [Q, P] = [
¯
Q, P] = 0.
There are many ways of deﬁning supersymmetry, we will only look at the simplest minimal
supersymmetric extension of the standard model, the MSSM. Basically sypersymmetry
means that all particles we know must have a supersymmetric partner with exactly the
same quantum numbers but diﬀerent spin.
normal partner spin
q
L
˜ q
L
0 squarks
q
R
˜ q
R
0 can mix
l
L
˜
l
L
0 sleptons
l
R
˜
l
R
0 also mix
ν
L
˜ ν
L
0 sneutrinos
γ ˜ γ 1/2 (photino zino higgsino)
Z
0
˜
Z 1/2 all mix together into
h
0
1/2 neutralinos ˜ χ
0
i
H
0
˜
H 1/2 2 + 3 + 1 + 1 + 1 = 8 spin states for the bosons
A
0
1/2 gives four neutralinos with two spin states each.
W
±
˜
W
±
1/2 (wino higgsino) mix together
H
±
˜
H
±
1/2 charginos ˜ χ
±
i
g ˜ g 1/2 gluino
It turns out we need we need two Higgs doublets, one to generate mass for u, ν and one for
d, l. Two doublets give eight degrees of freedom, three are “eaten” by the W/Z masses,
leaving us with ﬁve Higgs particles. Their partners have the same spin and mixes into
neutralinos and charginos. It is possible to derive mass relations between the Higgses and
the W/Z, in particular m
h
< m
Z
which would be excluded by experiments if it wasn’t for
some radiative corrections giving m
h
< 103 GeV.
If SUSY was unbroken the masses of any particle would be the same as its superpartner.
But this is not the case, so SUSY must be broken. Exactly how it is broken, nobody
knows, but there are some suggestions.
37
We can deﬁne a “parity” relating to SYSU, called Rparity
R = (−1)
L+3B+2S
where L is lepton number, B is baryon number and S is spin. So all ordinary parti
cles have R = +1 and their superpartners (sparticles) have R = −1. If Rparity is
conserved, sparticles can only be produced in pairs. This also means that the lightest
supersymmetric particle (LSP) is stable.
Even if the masses are not the same, the couplings do not care if we have particles or
sparticles. Hence we have that vertices such as eg.
W
−
→e
−
ν
e
, W
−
→ ˜ e
−
˜ ν
e
,
˜
W
−
→ ˜ e
−
ν
e
,
˜
W
−
→e
−
˜ ν
e
all have the same coupling (g
2
). So as soon as we come above the massthreshold for
producing sparticles, they will be produced at the same rate as their ordinary particle
equivalents.
Now it was said before that the loopcorrections to the higgs mass, eg. from a toploop,
gave δm
2
h
∝ −
∞
m
2
t
dp
2
, leading to the hierarchy problem. Now we will also have stoploops
which come in with oposite sign since they are bosons (but the same coupling). The end
result is that δm
2
h
∝ m
2
˜
t
−m
2
t
which is ﬁnite. This solves the hierarchy problem and many
other problems with the standard model.
Another thing that SUSY can explain is dark matter. If there is a stable LSP, which
only interacts weakly (eg. ˜ χ
0
) it would be produced copiously at the big bang and would
basically still be around. The interaction with matter would be weak since σ( ˜ χ
0
p) ∝ m
−2
˜ q
(cf. σ(νp) ∝ m
−2
W
). However, if Rparity is not conserved we could have decays like ˜ γ →νγ
– it could still contribute to dark matter if the decay is slow enough. Note that we can
get Rparity violation by either lepton number or baryon number violation, but if we have
both we will get proton decay.
With Rparity conserved we would get characteristic decay chains of sparticles according
to their mass hierarchy., eg.
˜ u →d + [ ˜ χ
+
1
→ν
τ
+ [˜ τ
+
→τ
+
˜ χ
0
1
]]
This comes with a price though. There are double the amount of particles and hence
particle masses which are free parameters. There are also a bunch of extra mixing param
eters for which there are no predictions. In total we have over hundred free parameters
which need to be measured. If SUSY is discovered it will give a lot of hard (and boring)
work to check its consistency.
String theory
How do we include gravity in the standard model. Naively we would try to make a
quantum ﬁeld theory of general relativity and we would get a spin2 graviton. But there
would also be nonrenormalizable divergencies. The problem is that in normal QFT we
consider particles to be pointlike. The higher the scale the closer we get, and also the
38
higher the gravitational potential, in addition the higher the scale, the higher energy i.e.
mass and, hence, the higher the force. The result is that no known renormalization works.
A possible solution started to emerge 1984 when string theory suggested as a viable
explanation to what may be going on. The idea is that elementary particles, rather than
being pointlike are actually modes of vibrations on small pieces of string. This means that
particles are intrinsically extended, which also means that the divergencies are muﬄed.
While a particle will describe a world line, x
µ
(τ), the fourposition at any given proper
time, a string will describe a world sheet, x
µ
(τ, σ), where σ is an additional coordinate
along the string. The strings can either be open or closed (x
µ
(τ, σ + 2π) = x
µ
(τ, σ)).
Just as particles are not pointlike, nor are the vertices. Rather the vertices are smeared
out in a way such that several diﬀerent particle vertices (eg. s, t and uchannel diagrams)
correspond to one and the same string vertex.
To construct a string theory which is consistent with QFT is a bit tricky. To avoid
tachyons (particles with m
2
< 0 which travel and convay information faster than light,
not to be confused with normal oﬀshell virtual particles), it turns out that we need 26
dimensions. Things improve if we allow for supersymmetry, in which case we only need
nine space and one time dimension as long as the basic QFT Lagrangian is invariant
under SO(32) or E
8
E
8
. Now since
E
8
⊃ E
6
⊃ SO(10) ⊃ SU(5) ⊃ SU(3) SU(2) U(1)
we can live with this restriction, even though we have to live with a whole new zoo of
particles living in the other E
8
which only interacts gravitationally.
It turns out you can construct exactly ﬁve such (super) string theories, called typeI,
typeIIA, typeIIB, heterotic SO(32) and heterotic E
8
E
8
(heterotic has closed strings).
So, which one has nature chosen? Well, recently it has been realized that they may all just
be diﬀerent aspect of the same underlying theory, Mtheory, and that they are related to
eachother through duality transforms (cf. solid state physics where we can describe the
same phenomena either by atomic vibrations or by phonons).
It turns out that M theory requires eleven dimensions. But let’s go back to ten dimensions
where we can construct ﬁve theories which all lack unphysical divergencies and all include
SU(3) SU(2) U(1) together with supersymmtry, a spin2 graviton and a spin3/2
gravitino. How do you explain that there are ten dimensions when we only live in four?
Well the extention of the universe in the other six dimensions is extremely small. The
universe is curled up into tubes with circumference L ∼ m
−1
Pl
.
Recently it was realized that we haven’t really tested gravity at small distances. This
means that we can imagine that all standard model ﬁelds live on a fourdimensional brade
in a higher dimensional universe. What if gravity was allowed to propagate throughout
all dimensions, what would then experiments tell us about the maximum size of these
dimensions? Well, not much. Gravity has only been studied down to millimeter scales.
Lets look at the gravitational potential around a particle with mass m generalized to 4+n
dimensions:
V (r) ∼
m
m
n+2
Pl
1
r
n+1
,
Now, if the n extra dimensions are of size R, for distances much larger than that the
39
potential would look like
V (r) ∼
m
m
n+2
Pl
1
R
n
1
r
But this is the distances we are used to, where we have measured the potential to be
V (r) ∼
m
m
2
Pl4
1
r
where we have found m
Pl4
≈ 10
19
GeV. But if the extra dimensions are large this means
that the true Planck scale is much smaller
m
Pl
∼ R
−
n
n+2
m
2
n+2
Pl4
In fact, for suitable values of n and R the Planck scale may be almost as small as the
weak scale. Not only would this cure the hierarchy problem, it also means that it could
be accessible at the LHC. Contrary to standard super string theory, this means that we
could enter into the realm of quantum gravity at the LHC (with the production of mini
black holes and lots of other fun stuﬀ) — truly an exciting prospect.
Particle Physics and the Universe
This is a rather active research area right now and we have to change numbers rather
often. An example here is that the Hubble constant which was only know to within a
factor of two is now quite well known. So keeping these notes up to date is always a
bit of a problem, espescially since the new developments often involve a large technical
background.
There are many areas of contact, there exists in fact a whole ﬁeld called astroparticle
physics for which by now many books and reviews have appeared.
The most exciting recent developments have been in the contact with cosmology but even
this relation goes back quite some time.
Astrophysics constraints on Particle Physics
These consists of knowing the physics between a variety of astronomical objects and then
using these to obtain limits on extra eﬀects not included normally.
These typically appear in the radiation of particles in stellar (or their successors) which
can have sizable eﬀects. Here are some
• In stars the thermonuclear process is rather well understood. If there were more
weakly interacting particles that would get radiated at a reasonable rate they would
carry of heat from the thermonuclear fusion engine running in the core of the star.
This provides limits on axions and all light weakly interacting particles. They have
to be weakly interacting to get out the core. The best limits tend to come from red
giants since these have the highest temperatures in the core.
40
• A celebrated example is the fact that neutrino oscillations were really ﬁrst discovered
from the solar neutrinos.
• When a star ﬁnally collapses and produces a neutron star, very many neutrinos are
radiated. Detection of these and their arrival times have allowed many limits on
neutrino properties.
Nuclear Formation in the Big Bang
In the early universe the neutrons were caught into (mainly)
4
He nuclei. The precise
amount of neutrons that get caught in various light nuclei versus the total number of
protons and neutrons available is very sensitive to many conditions around. This has
allowed to put some limits on the numbers of light neutrinos before limits from LEP
appeared. It also allows limits on the numbers of relativistically moving particles (usually
called radiation) in general at the time of nuclear synthesis.
The dark matter problem
It is known that there exists a lot more matter in the universe than there is in stars and
other normal objects. This was ﬁrst discovered in the velocity curves of distant galaxies.
Particle physics has many candidates for this in beyond the standard model scenarios.
• Massive neutrinos
• Axions
• Various supersymmetric particles, these come in a class known as WIMPs (Weakly
Interacting Massive Particles)
There are basicallythree types, hot dark matter (neutrinos), cold dark matter (WIMPs)
and the ones where various ﬁelds get expectation values (which may vary with time) and
as such produce the dark matter content.
Baryon Number Generation
One of the mysteries is why is there any baryon number left at all. In the beginning,
pick your favourite energy, there are very many quarks and antiquarks around, as well
as leptons. When the temperature drops below that required to have creation reactions
running the annihilation of quarks and leptons starts. But they do not annihilate fully,
some of them are left and the universe now exists mainly of quarks and leptons, not of
antiquarks and antileptons.
This tiny left over is rather puzzling and in order to solve it many conditions need to be
satisﬁed (CPT conservation assumed here):
1. There have to be baryon number violating processes
41
2. Some things need to go out of thermal equilibrium, otherwise the equal balance
principle would have baryon and antibaryon number equal.
3. CP violation needs to be present.
In fact the standard model has all these three properties but not in a way suﬃcient to
create the observed asymmetry (but it makes it worse for other higher energy mechanisms
to produce the asymmetry). Grand Uniﬁed Theories provide one option for producing
this imbalance.
Inﬂation
There are many problems associated with the fact that the universe is very big and looks
rather homogenuous and isostropic on very large scales, as well as being very ﬂat. These
problems are:
• Flatness problem: this arises because the universe now is rather ﬂat but the evolu
tion from general relativity is such that the universe moves away from ﬂatness with
time. So in order to be as ﬂat as observed now it had to be extraordinarily ﬂat at
early times.
• Horizon problem: Many of the areas of the universe can not have exchanged signals
with each other because of the expansion of the universe. But still when we see
those they look remarkably alike. How could this be?
• Monopoles, Cosmic Strings and Domain Walls: a problem: similar to the previous
one. We expect several cases of spontaneous symmetry breaking to have occurred
in the history of the universe. If the ﬁrst breaking happens in diﬀerent directions
in diﬀerent areas of the universe they start relaxing together but then socalled
topological defects (similar to the domains in a ferromagnet) can occur and we
should have observed these if they existed.
All of these are solved (at least in principle) by inﬂation. Note that the absolute simplest
version of inﬂation does not work and no simple nice working model is known but many
general features are well supported.
A really short Big Bang primer from General Relativity
We only discuss here the simple FriedmanLemaˆıtre universes. Assume the universe is ho
mogeneous and isostropic. That means that the energymomentum tensor can be written
in the simple form
T
µν
=
¸
¸
¸
¸
ρ 0 0 0
0 p 0 0
0 0 p 0
0 0 0 p
¸
(1)
42
and ρ(t) and p(t) are only functions of time. The matter is supposed to be at rest and
behave like a socalled perfect ﬂuid. The metric can be written in the general form
ds
2
= dt
2
−R(t)
2
dr
2
1 −kr
2
+ r
2
dθ
2
+ sin
2
θdφ
2
. (2)
The parameter k can take the values ±1, 0 leading to the following general geometry of
the universe
k = +1 Finite in space universe, closed, spherical geometry
k = 0 Inﬁnite in space, open, ﬂat geometry
k = −1 Inﬁnite in space, open, hyperbolic geometry.
Notice that I have not said anything about whether the universe has a ﬁnite lifetime.
From Einstein’s equations we then get
H(t)
2
≡
˙
R(t)
R(t)
2
=
8πG
N
ρ(t)
3
−
k
R(t)
2
. (3)
The pressure p satisﬁes
d
ρR
3
= −pd
R
3
. (4)
This has to be supplemented with an equation coupling ρ and p, known as the equation
of state. The latter is in general complicated but for most purposes some simpliﬁcations
are used. Usually it is assumed that
p(t) = wρ(t) . (5)
The following cases are of interest
• w = 0 corresponding to particles at rest, known as a matter dominated universe.
ρ ∼ R
−3
• w = 1 corresponding to particles moving at the speed of light whose energies is
dominated by the kinetic energy and thus gets redshifted, known as a radiation
dominated universe. ρ ∼ R
−4
• w = −1 Or a cosmological constant. ρ is constant.
• −1 ≤ w ≤ 0 known as quintessence.
Note that in general relativity energy does not need to be conserved.
The later cases are normally realized as being produced by the values of a ﬁeld φ which
then obeys the equation of motion
¨
φ + 3H
˙
φ = −
∂V (φ)
∂φ
, . (6)
For k = 0 these have the behaviour
43
matter ρ(t) = ρ
0
/R(t)
3
R(t) =
√
6πG
N
ρ
0
t
2/3
radiation ρ(t) = ρ
1
/R(t)
4
R(t) =
32πG
N
ρ
1
/3 t
1/2
cosmological constant ρ = ρ
2
R(t) = e
√
8πG
N
ρ
2
/3 t
So a possibility of exponential expansion exists if at some point we have a constant density.
This is known as inﬂation.
A short history of the universe can now be described as.
1. For very small times all particles are relativistic and they dominate by far the energy
density. The universe expands as a radiation dominated one and the particles slowly
redshift.
2. The the energies of the φ ﬁeld becomes low enough that the eﬀect of the V (φ)
becomes important and we shift to an exponentially expanding universe. The large
inﬂation here means that the entire observable universe was reall a very small part
beforehand.
3. The φ has slowly reached the minimum of the potential such that inﬂation stops
and the universe is again radiation dominated.
4. Finally all relativistic particles get redshifted to such low energies that their rest
mass dominates and we have the present (almost) matter dominated universe.
5. A newer observation is that we actually again have something like a (now much
smaller one) cosmological constant around. This is known as the dark energy.
How does this solve the various problems. Basically all in the same way. Since the entire
universe came from a very small preinﬂation area it had ample time to get homogenuous
and organize itself. During the inﬂation any curvature present would also have gotten
stretched away so that the universe afterwards will be ﬂat with an incredible precision.
Cosmic Microwave Background
The cosmic microwave background has many interesting properties. First its existence
by itself was a major success for the Big Bang model and was in fact what really made
it universally accepted. Now the ﬁne inhomogeneities are of major interest see e.g. [2].
In 2003 there was a major step forward with the results from the WMAP satellite, the
successor of the COBE satellite, the results of which led to the Nobel prize in Physics
2006.
Structure Formation
This is the question how the present structures seen in the universe at scales from galaxies
and up are formed from the initial small perturbation caused by quantum eﬀects and/or
44
whatever in very early universe. This is done by galaxy surveys and complements the
structures seen in the microwave background by covering diﬀerent time scales and size
scales. Here come names like “the great wall, the gigantic bubble,. . . ”
Other
There are many more speculative connections involving strings, extra dimensions gravity
and gauge interactions operating in a diﬀerent number of dimensions, quantum gravity
and gravitational radiation,. . . .
References
[1] L. Bergstr¨om and A. Goobar, Cosmology and Particle Astrophysics
[2] M. Kamionkowski, http://arXiv.org/abs/hepph/0210370
45
The probability of a given transition is now dP ∝ Mf i 2 Where we have averaged over all nonobserved degrees of freedom (DoF) in the initial state and summed over all DoF in the ﬁnal state. Eg. if we have an unpolarized fermion in the initial state, each spin state (up and down) contributes whit a factor a half each. Similarly if we have a fermion in the ﬁnal state where we do not measure the ﬁnal spin, the two spin states are simply added. Introducing an arbitrary volume, V = L3 , we get the density of states of a particle by assuming periodic boundary conditions: eipx L = 1 Looking in three dimensions we get dn = V d3 p . (2π)3 ⇒ px = 2πn/L.
We now get the transition probability assuming ﬁxed momenta in the initial state and a continuum in the ﬁnal state by squaring the matrix element and multiplying with the appropriate phase space volumes dΓ = V (2π)4 δ (4) (Pf − Pi )Mf i 2 1 j=f,i 2Ej V V d3 pj (2π)3
j=f
From which we can get the cross section by dividing by the ﬂux: dσ = dΓ/ﬂux (where the ﬂux will contain the additional volume factors to cancel the dependence). If we consider the scattering of two particles, A and B, to a ﬁnal state f we get the relative ﬂux factor by looking eg. in the rest frame of B: vA /V which gives vA = (pA · pB )2 − m2 m2 A B EA mB .
In general we get the relative ﬂux factor depending on the relative velocity vr = (pA · pB )2 − m2 m2 A B EA EB .
We now get the expression for the cross section dσ = (2π)4 δ (4) (Pf − pA − pB )EA EB (pA · pB )2 − m2 m2 A B Mf i 2 1 1 2EA 2EB d3 pj , 2Ej (2π)3
j=f
where the dependence on V is gone, and where the the factor EA EB will cancel. The phase space element is, in fact, Lorentz invariant. It can be derived from the fourmomentum phase space element for a free particle δ(p2 − m2 )Θ(p0 )d4 p = δ(E 2 − p2 − m2 )Θ(E)dEd3 p 2
where the δfunction ensures that the particle is onshell and the Θfunction ensures that we have positive energy. The Dirac deltafunction is a bit tricky when it comes to variable substitution. Formally δ(x − x0 ) δ(f (x)) = f ′ (x0 ) x0 ;f (x0 )=0 Take eg. a konstant factor k and a test function h, and make a substitution y = kx: δ(k(x − x0 ))h(x) = kx0 1 h(x0 ) y dy = h( ) = δ(y − kx0 )h( ) k k k k k
Taylor expanding any function f around f (x0 ) = 0 we need only take into account the ﬁrst nonzero term since everything else is negligible, so f (x) ≈ f ′ (x0 )(x − x0 ) and we get δ(f (x)) = In our example we then get δ(E 2 − p2 − m2 )Θ(E)dEd3 p = δ((E + p2 + m2 )(E − p2 + m2 )Θ(E)dEd3 p = √ δ(E − p2 + m2 ) d3 p √ 2 . dEd3 p = 2E E + p + m2
2 In a scattering we deﬁne s = (pA + pB )2 = Ecm . If mA , mB ≪ Ecm we have
δ(x − x0 ) f ′ (x0 )
s = (pA + pB )2 = m2 + 2pA · pB + m2 ≈ 2pA · pB A B and And we get the standard form of the scattering cross section dσ = (2π)4 δ (4) (Pf − pA − pB ) Mf i 2 2s 4 (pA · pB )2 − m2 m2 ≈ 4pA · pB ≈ 2s A B d3 pj , 2Ej (2π)3
j=f
In a 2 → 2 scattering, A + B → C + D we have in the restframe pA + pB = pC + pD = (Ecm , 0, 0, 0) and δ (4) (pC + pD − pA − pB ) d3 pC d3 pD d3 pC d3 pD = δ (3) (pC + pD )δ(EC + ED − Ecm ) = EC ED EC ED δ(EC + ED − Ecm )
d3 pC EC ED Shifting to polar coordinates and introducing dΩ = d(cos θ)dφ we get
2 δ( p2 + mC + C
p2 + m2 − Ecm ) C D
p2 dpC dΩC C EC ED
Again, we can transform the δfunction, δ(pC  − p(Ecm , mC , mD )) p2 dpC dΩC C pC  pC  EC ED √ 2 2 +√ 2 2 pC +mC pC +mC 3
and we can set p = pC  = p(Ecm , mC , mD ) and ΩC = Ω p2 dΩC
p ED p EC
+
EB EC
=
pdΩ pdΩ = √ EC + ED s
And the cross section becomes dσ = (2π)4 δ (4) (Pf − pA − pB ) Mf i 2 d3 pC d3 pD = 2s 2EC (2π)3 2ED (2π)3
p Mf i 2 1 p √ dΩ = Mf i 2 dΩ 2 2s 4 (2π) s 32π 2 s3/2
9.1
Lifetimes, Resonance widths and Decay probability
Now, look at the decay of an unstable particle R → C + D. We get the transition probability d3 pj (2π)4 δ (4) (Pf − pR ) . Mf i 2 dΓ = 3 2ER j=f 2Ej (2π) √ Again, looking at the rest frame of the decaying particle we have ER = mR = s and the same transformation as before of the ﬁnal state particles gives p dΓ = Mf i 2 dΩ 32π 2 m2 R and integrating over the solid angle (assuming unpolarized R) p Γ = Mf i 2 8πm2 R Assume an exponential decay for a particle formed at t = 0: ψ(t)2 = ψ(0)2 e−t/τ Θ(t) This will give the transition probability Γ=− 1 1 dψ(t)2 = 2 ψ(t) dt τ
Now, the time dependence of the wave function of a free particle in its rest frame is ψ(t) = ψ(0)e−iEψ t so we now get ψ(0)e−iE0 t e−tΓ/2 = ψ(0)e−iEψ t and an unstable particle can be seen to have complex energy Eψ = E0 − iΓ/2. Let’s do a Fourier transformation iψ(0) 1 1 ψ(0) ∞ ˜ dt ei(E−(E0 −iΓ/2) = √ ψ(E) = √ dt eiEt ψ(t) = √ 2π 0 2π E − E0 + iΓ/2 2π ∞
1 ψ(0)2 ˜ ⇒ ψ(E)2 = 2π (E − E0 )2 + Γ2 /4 Heissenberg says: undetermined lifetime means undetermined energy and the probability of ﬁnding an unstable particle with a given energy is a distribution around E0 . 4
The latter have the property dΩPl Pl′ = 4π δll′ 2l + 1 so squaring and integrating we get the cross section σ= 4π κ2 (2l + 1) sin2 δl l For δl (E) = π/2 we have a large cross section.9. We ﬁrst remember the scattering in a spherically symmetric potential using partial wave expansion f (θ) = l (2l + 1) e2iδl − 1 Pl (cos θ). Note that we can produce a resonance also away from its nominal rest energy. 2iκ where l is the angular momentum. κ = pC  = pD  and Pl are the Legendre polynomials. Also note that the expression is not relativistically invariant.2 Scattering through a resonance We have the decay of a resonance R → C + D prepared by the fusion of A + B → R at time t = 0. corresponding to a resonance in the lchannel. We can write e2iδl − 1 1 sin δl = eiδl sin δl = = 2i cos δl − i sin δl cot δl − i and Taylor expanding cot δl around δl (ER ) = π/2 we have cot δl (E) = cot δl (ER ) + (E − ER ) We will identify d cot δl dE E=ER d cot δl dE E=ER + ··· with −2/Γ and get for the dominating lterm f (θ) ≈ fl (θ) ≈ (2l + 1)Pl (cos θ) Pl (cos θ) Γ/2 2l + 1 = κ (E − ER )(−2/Γ) − i κ (E − ER ) − iΓ/2 4π Γ2 /4 (2l + 1) κ2 (E − ER )2 + Γ2 /4 Squaring again and we get the cross section σR (E) = where the last factor is called the nonrelativistic BreitWigner distribution and can be compared to the exponential decay above. δl is the corresponding phase shift. We can making it invariant by by noting that E + ER ≈ 2mR ≈ 2m in the rest frame: Γ/2 Γ/2 E + ER mΓ = ≈ 2 − m2 ) − im Γ (E − ER ) − iΓ/2 E + ER (E − ER ) − iΓ/2 (m R R and with m2 = s we get the relativistic BreitWigner shape of the cross section σ(s) = Γ2 sΓ2 4π = 16π(2l + 1) (2l + 1) 2 κ2 (s − m2 )2 + m2 Γ2 (s − m2 )2 + mR Γ2 R R R 5 . This distribution is peaked around ER and falls to half the peak value at E = ER ± Γ/2.
and since γ ∝ 1/τ the more decay channels the faster decay. We get the cross section for a particular ﬁnal state Γf σ(A + B → R → f ) = σ(A + B → R) . A+B →C +D : R→C +D : =⇒ A+B →R→C +D : pC dΩC 32π 2 s3/2 pC dΓ = Mf i 2 dΩC 32π 2 m2 R pC if isotropic Γ = Mf i 2 2 8πmR cR 2sR + 1 σ = 16π (2sA + 1)(2sB + 1) cA cB dσ = Mf i 2 ΓAB ΓCD R R 2 2 (s − mR ) + m2 Γ2 R R Approximate Feynman rules external fermion (pair) √ uu = 2E ⇒ u ∼ 2E external gauge boson ǫµ ≈ 1 6 . If we have more than one decay channel the widths are additive. 2sB + 1 and colour multiplicities cA and cB for the initial state and the corresponding 2sR + 2 and cR for the resonance decaying to a particular ﬁnal state. assuming we have spin multiplicities 2sA + 1. but here we average rather than tot sum. Eg. Assume incoming masses mA . 2 Deﬁne s = Ecm = (pA + pB )2 = (pC + pD )2 . we get the cross section σ(A + B → R → f ) ≈ 16π (2sR + 1)cR ΓAB Γf (2sA + 1)(2sB + 1)cA cB (s − m2 )2 + m2 Γ2 R R tot A rough estimate of the cross section can be obtained by σ ≤ σ(s = m2 ) ≈ 16π [∼ 1] R ΓAB Γtot 16π ≤ 2 2 2 mR Γtot mR Important cross section formulae Work in the rest frame of the process. f . Γtot = f Γf . mB ≈ 0. Γtot Now we reverse the time and get a factor ΓΓi there as well.where the last step is true for massless particles where κ = √ s/2.
f f photon vertex eQf γ µ ≈ eQf γ f f′ W ± vertex g √2 VfCKM for lefthanded fermions ′ 2 f 0 for righthanded fermions W± f f Z 0 vertex g2 2 3 cos θW (Tf − Qf sin θW ) for lefthanded fermions g2 2 cos θ (−Qf sin θW ) for righthanded fermions W Z0 f f gluon vertex g3 γ µ ≈ g3 for quarks 0 for leptons g external scalar 1 f f Higgs–fermion vertex g2 m gf = 2m f W h0 W W Higgs–W vertex (except if mH ≫ mW ) gW ǫǫ′ ≈ gW = g2 mW h0 7 .
. a particle created at some point (source) propagating and destroyed at another point (sink). Momentum ′ conservation means e−ipx eip x eikx = 1.3 The W width From Kane eq. for a real scalar we have 1 δµ φδ µ φ − m2 φ2 L= 2 which gives the requirement on the ﬁeld δ µ δµ φ + m2 φ = 0 and the propagator ⇒ −p2 φ − m2 φ = 0 1 − m2 which is the Fourier transform of the socalled Yukawa potential p2 − 1 e−mr 4π r Ternary terms (with three ﬁelds) corresponds to threevertices. .g. i.e. 2 8 . g. and we are left with the external wave functions uf (p′ ). + . ¯ and the vertex factor Qf γµ . . W ± .c. Terms quadratic in a ﬁeld corresponds to a propagator.) 1 m2 − m2 + imR ΓR R Feynman rules and Lagrange densities Looking at the Lagrange density we can identify diﬀerent kinds of propagators and vertices. (7.Boson resonance propagator (R = γ. εµ 9. the interaction term in the electromagnetic Lagrange density with fermion ﬁelds: ¯ Lint = −Jµ Aµ = Qf ψf γµ ψf Aµ where ψf = uf (p)e−ipx destroys a fermion. + √ [¯L γ µ dL + νe γ µ eL ] Wµ + h. . Z 0 . E.g. which we will not discuss here. uf (p). Quartic terms will give rise to fourparticle vertices. ¯ Similarly Aµ = εµ eikx creates (or destroys) a photon with momentum k. . E.32) we get the part of the Lagrange density for the fermion–W± vertex g2 + u ¯ L = . . f (or creates the corresponding antifermion) ′ ¯ ¯ with momentum p and ψf = uf (p)eip x creates f (or destroys f ) with momentum p′ . .
α2 = 1/30) gives ΓW ≈ 2 GeV. εµ = O(1). 2 g2 3 m2 (Tf − Qf sin2 θW )2 + (−Qf sin2 θW )2 cos2 θW Z f ⇒ M2 = 2 and with α2 = g2 /4π and a factor 2/3 from spin factors as in the W case we get Γf = Z α2 mZ f 3 3 2 NC (Tf )2 − 2Tf Qf sin2 θW + 2Qf sin4 θW 2θ cos W We have very precise measurements from LEP1: mZ = 91. The squared matrix element then becomes M2 = g2 m2 /2.038 GeV.009 nν = Z ν ΓZ 9 . Note that the quarks come in three colours. s (t¯ b) The last one is forbidden for real W+ by energy conservation. This will give us the width W dΓeν = M2 W ⇒ Γeν = W p dΩ 32π 2 m2 W 2 α2 mW g2 m W = 48π 12 Summing up to the total width we get a factor three from the three generations of leptons and another factor three for each of the two ﬁrst generation of quarks: Γtot = (1 + 1 + 1 + 3 + 3) W 3α2 mW α2 mW = . τ + ντ .repeated for the other two families. We can also measure the partial widths and branching fractions of Z 0 into quarks (BR(u¯) ≈ u ¯ 0. ΓW = 2.044 GeV.4925 ± 0. BR(dd) ≈ 0.033).983 ± 0..1876 ± 0. We can write down the matrix element for the ﬁrst decay g2 g2 M(W + → e+ νe ) = √ νL γ µ eL εµ ≈ √ mW ¯ 2 2 where we have used the standard normalization ν e ∼ 2E ∼ mW and the polarization ¯ 2 of the photon. + g2 cos θW 3 ¯ ¯ fL γ µ fL (Tf − Qf sin2 θW ) + fR γ µ fR (−Qf sin2 θW ) Zµ + . This gives us the decay modes of the W+ : W + → e+ νe .12. c¯. Recent experimental results: mW = 80. µ + νµ . From this we can infer the partial width or branching fraction into neutrinos (BR(ν ν ) ≈ 0. 9.0021 GeV and ΓZ = 2.0023 GeV. W 2 while a full calculation would give M2 = g2 m2 /3.. . .15) and charged leptons (BR(l+ l− ) ≈ 0.138 ± 0. ¯ ud.4 The Z0 width As for the W we have the relevant term in the Lagrange density L = .425 ± 0. 12 4 Inserting numbers (mW ≈ 80 GeV.07) — the invisible width ¯ — which means we can measure the number of (massless) neutrino species and hence also the number of families: Γinvisible = 2.
1). Unfortunately we cannot calculate the fi/p (however some of its behavior can be calculated). We call the quarks and gluons inside the proton partons. so any electro–weak process will drown in a lot of QCD processes. so √ √ s s pp ≈ (1. Alternatively we can use the fusion of quarks into W/Z. Here.10 Production of W ± and Z 0 The cleanest way to produce real and onshell W and Z is by lepton fusion. 0. 0. We deﬁne p s = (pp + pp )2 ¯ and we look in the rest frame of the hadronic collision. For this we can collide hadrons ¯ which consist of quarks and use processes such as u¯ → Z 0 → fermions and ud → W − → u fermions. we have controll of the hadronic collision energy but not of the energy of the colliding quarks. Similarly for the W the decay into quarks will drown in the background of QCD processes. The W and Z was ﬁrst discovered at a proton–antiproton collider at CERN in 1983. 0. In addition. ¯ ¯ . −1). eg. At LEP the processes involving W/Z are dominating if we tune our collision energy to hit the resonance peak. 0. Therefore the dominant decay of Z 0 into quarks cannot be used as a singnal. We know that a proton consists of three valence quarks (uud) together with a number of gluons and virtual quark–antiquark pairs (sea quarks). The only way is to look for a charged lepton and then look for the fact that the nonobserved neutrino will induce a imbalance of the transverse momentum of all the detected particles. At LEP1 (1989–94) (and at the SLC ) they used e+ e− → Z 0 → fermions. but we can infer some sumrules such as i 01 xfi/p (x)dx = 1 and we can measure them. To calculate the cross section of W/Z at hadron colliders we have to try to describe the quarks (and gluons) inside the hadron.15 and the total fraction of the protons momentum carried by the valence quarks is about 50%. The latter uses the fact that SU (2) is a nonabelian theory so that we have terms in the Lagrange density which couples the W ﬁeld to itself. Instead at LEP2 (1996–2000) they used the processes e+ e− → W + W − (exchanging a neutrino) and e+ e− → Z ⋆ /γ ⋆ → W + W − (via an oﬀshell Z 0 or γ). quarks interact much more strongly with gluons (QCD). However the decay into chagred leptons can be used (but remember the branching ratio is small). Then we deﬁne s = (pu + pd )2 ≈ 2pu pd = x1 x2 pp pp ≈ x1 x2 (pp + pp )2 = x1 x2 s ˆ ¯ ¯ ¯ ¯ 10 p d = x2 p p . ¯ 2 2 and the momentua of the colliding quarks are p u = x1 p p . It turns out that the (logarithmic) density of valence quarks peaks around 0. e− νe → W − → fermions is not technically feasible since we cannot ¯ create a beam of neutrinos which is well enough focused. ¯ Let’s look at the reaction p¯ → proton remnants + [ud → W + ]. pp ≈ (1. Also it turns out that the gluon density increases dramatically at small x as does the seaquark density (although it is typically much smaller than the gluon density). The corresponding production of W . Neither can the decay into neutrinos. To calculate the cross section we use the concept of a parton density: fi/p (x) is the probability to ﬁnd a parton i inside a proton carrying a fraction x of the protons momentum.
The fact that the decay is more or less evenly distributed in azimuth angle the peak will become smeared and slightly shifted towards lower m⊥ (a Jacobian peak). measure the lν transverse momentum components of the neutrino. come from other sources. assuming this was the only particle not detected: pν⊥ ≈  i pi⊥ . We know that W + only couples to lefthanded (spin oposite to the direction of motion) fermions and righthanded 11 . of course. we should ﬁnd a (Breit– Wiegner) peak around m+− ≈ mZ . These can. but if we in each event where we ﬁnd eg. however. For the W . Then we have the spin properties of the W which can be checked. s W 3 m W ΓW With s = x1 x2 s we get the total cross section ˆ σ (ˆ) ≈ ˆ s 4π 2 Γud Γf 1 W W δ(x1 x2 − m2 /s) σ(W ) = dx1 dx2 fu/p (x1 )fd/¯(x2 ) ¯p W 3 m W ΓW s 4π 2 ΓW 1 m2 ¯ = BR(W → ud)BR(W → f ) dx1 dx2 fu/p (x1 )fd/¯(x2 ) δ(x2 − W ) ¯p 3 smW x1 sx1 2 2 4π ΓW dx m ¯ = BR(W → ud)BR(W → f ) 2 fu/p (x)fd/¯( W ).We can write the cross section σ(W + ) = s dx1 dx2 fu/p (x1 )fd/¯(x2 )ˆud→W + (ˆ) σ ¯ ¯p We have basically already calculated the partonic cross section σ (ˆ) = 16π ˆ s 3 Γu d Γf 1 W W · . u Experimentally we ﬁnd the Z 0 by looking for two opositely charged leptons. 2 · 2 3 · 3 (ˆ − m2 )2 + m2 Γ2 s W W W ∞ −∞ ¯ To make life simple we use the narrowwidth approximation where we note that ∞ −∞ dˆ s = 2 2 (ˆ − mW ) + m2 Γ2 s W W dˆ s 1 = 2 + m 2 Γ2 s ˆ m W ΓW W W ∞ −∞ π dx = 2 1+x m W ΓW so that if we want to replace the distribution in s with a deltafunction we use ˆ πδ(ˆ − m2 ) s dˆ s W → 2 2 2 2 (ˆ − mW ) + mW ΓW s m W ΓW 4π ud f π ¯ ΓW ΓW δ(ˆ − m2 ). We can. we would expect to ﬁnd a Breit–Wigner peak around m⊥ ≈ mW . If we look at the distribution of m2 = 2pl pν⊥ and we assume that all W s decay ⊥ in a plane transverse to the beam axis. things are a bit more complicated since we cannot calculate the corresponding mass m2 ≈ 2El Eν (1 − cos θlν ) as we do not know Eν . etc. an e+ e− pair. where the sum runs over all detected particles (energy) in an event. calculate the quantity m2 = (pe+ + pe− )2 ≈ +− 2Ee+ Ee− (1 − cos θ+− ) and look at the distribution of these. mainly from the weak decay of heavy quarks. e+ e− or µ+ µ− (the tau decay hadronically and is more diﬃcult to detect). ¯p 3 smW sx1 mW /s x + ¯ and we get In the same way we can get the cross section for u¯ → Z 0 .
and jets are nototiously diﬃcult to deﬁne in a way so that we can calculate this ratio to a reasonable precision. widths. This is exactly what was found. The best measurement is from the ratio between the cross sections σ(e+ e− → Z 0 → q q g) and σ(e+ e− → Z 0 → q q ). One of the best measurements is from the width of the Z at LEP1: ΓZ = αEM mZ f (sin2 θW ).23122(15). 11 Measuring Standard Model parameters The Standard Model of particle physics contains of the order of 20 parameters: the masses of the fundamental particles.gov/. masses. αS is even trickier to measure. a number of mixing angles (later).1176(20). µ(p) → νµ (k) + [W − (Q) → e− (q) + νe (k ′ )]. 2θ sin θW cos W 2 We can also use the simpler sin2 θW = 1−m2 /m2 . all we see is jets of hadrons. We can draw the decay with a Feynman diagram where the muon ﬁrst decays into a muonneutrino and a W − immediately followed by the decay of the W − into an electron and an antielectronneutrino. This means that when it decays into a e+ νe the e+ cannot go along the positive zaxis since its spin must be paralell to its direction of motion. Current best value is αS (mZ ) = 0. a W + will normally have been ¯ produced from a u in the proton and a d in the antiproton and its spin will then be along the negative zaxis. The muon lifetime Let’s try to calculate the lifetime of a muon.lbl. The precision of sin2 θW is not as good. and the couplings constants α1 α2 α3 ⇒ ⇔ αEM = αα1 α22 1 +α sin2 θW = α1α1 2 +α αS αEM is best measured with the quantum hall eﬀect (solid state) giving an impressive accuracy:αEM (0) = 1/137. we use the ¯ 12 .(spin parallel to the direction of motion) antifermions. but it can be measured in many ways to ensure consistency. The place to ﬁnd the best measurements of coupling constants. The current best value is sin2 θW (MZ ) = W Z 0. Hence if we look at the distribution in azimuth angle of e+ in events believed to be W + events we expect it to be peaked in the direction opposite to the incoming proton. Hence if a proton comes along the positive zaxis and the antiproton along the negative.03599911(46). Denoting the momenta of the diﬀerent particles. mixing angles and other stuﬀ is in the Particle Data book http://pdg. But we cannot detect neither ¯ ¯ quarks or gluons. This is a useful exercise as most weak deacys of particles are calculated in the same way.
2Ek 2Eq 2Ek′ and if we ignore the masses of the decay products we get d3 kd3 q δ(k + q + k + q − mµ ) . If we W W ignore the spinor complication and remember that the wavefunctions are basically just given by the square root of the energy we get mµ for each of the vertices and the full matrix element is 2 √ g2 m 2 µ = 2 2GF m2 . Since the maximum value of Q2 ∼ mµ is much smaller than m2 . M= µ 2 m2 W Now we can write down the decay rate diﬀerential in the momenta of the incoming and outgoing particles: dΓµ = d3 k d3 q d3 k ′ (2π)4 δ (4) (k + q + k ′ − p) M2 2mµ 2Ek (2π)3 2Eq (2π)3 2Ek′ (2π)3 8G2 m4 1 (4) d3 k d3 q d3 k ′ 1 F µ · . The ﬁrst parenthesis is the vertex factor for the µνµ W vertex and the second the vertex factor for the eνe W vertex.Fynman rules to write down the matrix element g2 g 1 √2 (¯γλ PL νe ) M = √ νµ γ λ PL µ e ¯ Q2 − m2 2 2 W with PL = 1 − γ5 . the delta function gives k2 + q 2 + 2kq cos θq = mµ − k − q and a jacobian k2 + q 2 + 2kq cos θq kq 13 . 0. so p = (mµ . the propagator is to a very good approximation 1/m2 . 0. Sandwiched inbetween is the propagator 2 factor corresponding to the intermediate W . · δ (k + q + k ′ − p) = (2π)5 2mµ 2 2Ek 2Eq 2Ek′ where the extra factor one half comes from the fact that only lefthanded muons are involved. leaving us with δ(k + q + k2 + q 2 + 2kq cos θq − mµ ) 4πkdk · 2πqdqd cos θq 8 k2 + q 2 + 2kq cos θq Integrating also over cos θq . 0) so the diﬀerential can be written δ (3) (k + q + k′ )δ(Ek + Ep + Ek′ − mµ ) d3 k d3 q d3 k ′ . We want to move to the rest frame of the muon. 8 k · q · k + q We now go to polar coordinates (with k relative to a given axis and with q relative to k) δ(k + q + k2 + q 2 + 2kq cos θq − mµ ) 8kq k2 + q 2 + 2kq cos θq k2 dkdΩk · q 2 dqdΩq Here we can always do the integral over Ωk (⇒ 4π) and over the azimuth angle of q around the k direction (⇒ 2π).
a ντ may exchange a W with a nuclear target. sin2 θW ). indirectly.Simplifying miraculously to π 2 dkdq and we get the diﬀerential deacy width dΓµ = 2G2 m3 2 F µ π dkdq (2π)5 which can be integrated (k < mµ /2. If so. 192π So.2 ± 2. 25 Top and tauneutrino The top quark was discovered in 1995 at Fermilab outside of Chicago in USA. G2 m 5 µ Γµ = F 3 . Alternatively one may consider the forward–backward asymmetry as in Kane. by measuring Γµ and mµ we can get a good value for g2 (and. The W can be p u bW detected in the usual way with hightransversemomentum leptons and missing transverse momentum (a bit trickier since the W s have a substantial transverse momentum and the missing momentum is the sun of the two neutrinos). The way they found it was to use the fact that the top will decay (before it hadronizes) to a b and a W + . The bquarks will hadronize into B mesons which have a very long lifetime. q < mµ /2. but 0 if it is a singlet. Top quarks ¯ are pairproduced on p¯ collisions through u¯ → g ⋆ → tt → bW +¯ − . They only observed four such events. 14 . but that was enough.6 GeV. In the same way there was indirect evidence for ντ (also from Γinvisible ) before it was Z directly observed at the DONUT experiment at Fermilab in 2000. One way is to consider the decay of the Z 0 into b¯ bpairs: ¯ Γbb ∝ (Tb3 − Qb sin2 θW )2 + Q2 sin2 θW Z b The ﬁrst term is for lefthanded bquarks and here the Tb3 is −1/2 if the b is part of a doublet. if you do the spin and kinematics correctly. At DONUT they constructed a beam believed to consist of tauneutrinos.0 ± 2. producing a τ which is detected as it decays τ − → ντ e− νe a couple ¯ of tenths of a millimeter away from the primary collision vertex. The mass has been measured to be mt = 174. Even before the top quark was directly observed there was plenty of evidence that it must exist (if the Standard Model is correct). mµ − k − q < mµ /2) giving the total width G2 m 5 µ Γµ = F 3 128π or. giving rise to secondary vertices around a millimeter from the collision vertex.
the problem that we need very strong magnets to bend the beam. US. Alternatively the beams are bent around in storage rings which enables repeated acceleration.gov • SLAC – Stanford. These use electrical ﬁelds to accelerate beams of charged particles. National laboratory owned by the DoE and operated by Stanford University. BSM(?) –3000 H. US. www. Switzerland. The only setback here is the synchrotron radiation which means that the beam loses energy δE ∝ γ 4 /R2 ∝ E 4 /(m4 R2 ) (larger ring or heavier particles give less energy loss). www. on the other hand.cern. and also that the proton is composite. Major accelerators name LEP1 SLC LEP2 HERA Tevatron RHIC LHC name ILC CLIC µµC VLHC beams e+ e− e+ e− e+ e− ep p¯ p Au pp location period energy (GeV) 91 91 –209 30×900 2000 2 × 200 × A 14 000 main physics Z0 Z0 W± QCD top QGP H. European organization including Sweden. www.slac. They also use magnetic ﬁelds to bend and focus the beams. Hence for e± beams this is a real problem.stanford. www. National facility with strong international component. National laboratory owned by the DoE and operated by a collection of Universities.de • FNAL/Fermilab – Chicago.ch • DESY – Hamburg. BSM(?) –5000 BSM(?) –200 000 BSM(?) DESY(?) 2012(?)– CERN 2020(?)– FNAL(?) 2030(?)– CERN/FNAL(?) 2030(?)– 15 . Germany. giving approximately 5 MeV/m (but new techniques are being developed to increase this).fnal. National laboratory.desy. Major laboratories • CERN – Geneva. For (anti) proton beams there is. with strong international component.12 Accelerators We study the fundamental particles by colliding them at high energies. To get them to high energies we use accelerators.edu • KEK – Japan. In a linear accelerator the energies achievable depends on the length. which means that the quarks (gluons) which interacts only carries around 10% of the energy. BSM(?) CERN 89–95 SLAC 89–98 CERN 95–00 DESY 91–07(?) FNAL 85–07(?) Long Island 99– CERN 07(?)– location period In the future: beams e+ e− e+ e− µ+ µ− pp energy main (GeV) physics 500–1000 H.
n. • Wire chambers: ∼ 1m in radius. besides the collision energy. ρ. The number of events we can expect for a given process with cross section. 16 . Detects charged particles. and length. σ. A beam of transverse size. We wish to detect as many of them as possible so we want a good solidangle coverage of detectors Ω → 4π. If we have a storage ring with n bunches of k particles per bunch going round at a frequency f and the cross section of the interaction area is A we get a luminosity L = f nk 2 /A (note error in the book). ﬁlled with gas and wires under high voltage. so the energy increases only like the square root of the beam energy. of the target. The luminosity is then given by the number of particles hitting the target per unit time. if you are able to focus the beams against eachother properly. σ drills a hole in the target of volume V = lσ where we can ﬁnd n′ = V ρ particles in the target. L. giving us a luminosity L = n · n′ /σ = nρl For a ﬁxedtarget machine the (invariant) collision energy is approximately given by (pbeam + ptarget )2 ≈ 2Ebeam mtarget . If we also have a magnetic ﬁeld we can determine their momentum by the way they bend. For charged particles it is typically possible to track them as they interact electrically with detector elements. • Cherenkov counters: Particles which traverses a medium at a speed larger than the speed of light in that medium will emit Cherenkov radiation (similar to a supersonic bang). is determined by the integrated luminosity: N =σ dtL We can use the accelerated beam to hit a stationary target. √ For colliding beams. 13 Detectors At LHC we expect of the order of a hundred particles being produced per collision.The performance of an accelerator is. The photons are detected and gives information about the mass of the particle. determined by the luminosity. For neutral particles we can use calorimeters to detect the energy deposition in the detector parts. which is much more economic. The passing of a charged particle induces a discharge between the wires which can be measured. on the other hand we get a collision energy of 2E+ E− . l. Typical examples of detectors are: • Microvertex detectors: (CCD) silicon pixels (20 − 50µm)2 layered very close to the interaction point. and the density.
however set the best limit so far on the mass: mh ≥ 113 GeV (95% conﬁdence). In particular. as it was about to close down. m2 32π 2 8 m2 W h pf Hence. LEP2. typically increasing with energy ∆E ∝ E. had a couple of events which could be Higgs. • Hadronic calorimeter: Hadron interacts strongly with matter (typically iron) and we get ionization (kicking out charged hardons from the nucleus) of the detector which can be detected. LEP1 didn’t see anything. W –W –Higgs and Z–Z–Higgs vertices. lead glass and photon detectors. Nice pictures No detector is perfect. • Muon Chambers: Any charged particle which can penetrate the calorimeters. If the Higgs exists it will be discovered at the LHC (although it may take some time as it is not easy). So far noone has succeeded. They did.• Electromagnetic calorimeter: Eg. we start by looking at the relevant terms in the Lagrange density and ﬁnd that there are fermion–fermion–Higgs. and conversely the Higgs can decay into a ferminon–antifermion pair. but the signiﬁcance turned out to be too low. For the fermion vertex we have the coupling gf = g2 m f 2mW so we can produce a Higgs by annihilating a ferminon–antifermion pair. considering the hierarchy of fermion masses. typically meters thick. From this we get the partial width Γhf¯ ≈ f cf m2 dΩ f M2 = α2 mh . To understand how we can produce and detect the Higgs. the Higgs will mainly decay to the heaviest possible fermion 17 . but the signiﬁcance was even lower. We can approximate the matrix element by M2 ≈ (gf mh )2 cf 2 where cf is three for quarks and 1 for leptons and 2 comes from the sum over spin states times a fudge factor. is most likely to be a muon. Even the Tevatron had something which looked like a Higgs event. For measurements of energy and momenta there is always √ an uncertainty. Also a neutral particle will interact with nuclei in the detector giving ﬂashes of photons. and when the Higgs decays it will predominately do so to heavy fermions. For identiﬁed particles there is always uncertainties due to nonperfect eﬃciencies and purities. we should use heavy fermions if we want to produce Higgs. 14 Finding the Higgs The discovery of Higgs would complete the Standard Model and has therefore been the main objective of many accelerators in the past. but no vertices containing gluons or photons. but indirect measurements favoured a light Higgs.
where we should ﬁnd a peak corresponding to the Higgs mass. but the background is virtually nonexistent. Again the combined branching ratio into Z and subsequently into leptons is small. but the background of (isolated) high transverse momentum photons is even smaller. ∼ 10−3 . One such channel is h → γγ. there is a HUGE background of bottom jets at LHC to completely drown any such Higgs signal. but there we do not have access to the missing mass method. where either the incoming (LEP2) or the outgoing (LEP1) Z 0 is oﬀshell (reduced cross section). However. this proces includes a topquark loop and the partial width is small. Note that close to threshold there will be a kinematical damping of the decay width 4m2 β 2 = (1 − 2f )3/2 mh For the W –W –Higgs vertex we have to calculate a bit more since we cannot make the √ approximation u ∼ 2E for the W wavefunctions. So we need to look at less likely decay channels wich are easier to ﬁnd. The branching ratio is tiny. making this one of the main production channels for light to mediummass Higgs. If the Higgs is lighter than 2mW the dominant decay mode is to bottom quarks which are diﬃcult to separate from the background. 16m2 W α2 m3 h . but not signiﬁcan enough. The calculation in is performed in Kane (21. u At the LHC we can produce the Higgs in the same way (u¯ → Z 0 → Z 0 h). If the Higgs heavier than 2mZ we can look for the decay h → Z 0 Z 0 → l+ l− l′+ l′− . However. there is a huge amount of gluons in the colliding protons. At LEP2 they found some events. while a heavy Higgs is extremely broad (in TeV units we get Γ ∼ 0. As for the γγ decay. mainly because the W can be longitudinally polarized. Te higgs will b then mainly decay into b¯ (jets) and we can study the angular distribution to conﬁrm the scalar nature of the produced resonance. so we have to see the Higgs directly. This means that a light Higgs is very narrow. Detecting the ﬁnal Z 0 decaying into leptons we can calculate the missing mass.2) and the result is M2 ≈ giving the patial width Γh W ≈ W Similarly we get Γh ≈ ZZ g2 m 4 h 2 2mW α2 m3 h . At the LHC we can also produce Higgs via gluon–gluon fusion.(with mf < mh /2). 18 . ∼ 10−3 . 2 32mW So we get a width of the Higgs which is proportional to the mass3 . the goldplated signal. Since the photon doesn’t couple directly to the Higgs this decay goes through a fermion loop (typically top) or a W loop.5m3 ). h In e+ e− → Z 0 annihilation we can look for higgs using Z 0 → Z 0 h. m2 = (pe+ + pe− − miss pl+ − pl− )2 .
7 fm – a string. the string will break in several places giving many hadrons. People have made calculations “proving” that if we do not ﬁnd the Higgs at the LHC. For strong interactions (QCD) we expect to have approximately the same thing. 19 . This will give us V (r) = − 4 αS + κr. If the energy is high enough. hadrons are typically emitted in a narrow cone around the original quark — a jet. free quarks. . while the other hadrons have a smaller fraction. What happens instead if a quark is given enough energy is that the ﬁeld (string) is stretched to the point where enough energy is stored in the ﬁeld to make a virtual q q pair ¯ real. In QED we have the potential (of eg. The colours of this pair may then screen the endpoint charges causing the string to break.We can also produce Higgs with W + W − fusion. which would grow with energy in the absence of the Higgs and will eventually violate unitarity (∼ probability of interaction larger than unity). While the electromagnetic ﬁeld between two electrically charged particles extend out to inﬁnity. But because gluons selfinteract (contrary to photons) we get conﬁnement. . ¯ According to Heissenberg if a q q is produced inside a transverse size d⊥ will have an uncertainty in transverse momenta of p⊥ ∼ 1/d⊥ ≈ 0. Hence we can never give a quark enough energy to escape from a bound stata and become free. 15 Quarks.3 GeV. The original quark is still conﬁned but not to the same hadron as before. b) and an (in principle arbitrary) normalization. 3 r where κ is a constant energy per unit length ∼ 1 GeV/fm ∼ 10 metric tons per meter. or is too heavy (mh ≫ 1 TeV) some other mechanism beyond the Standard Model must enter the scene. if the Higgs does not exist. the fact that we have three charges (r. where the W s are radiated oﬀ quarks in the protons. Strong interactions are quite diﬀerent from electro–magnetic ones. g. Nice pictures What if we don’t ﬁnd the Higgs at the LHC? There are some cross sections in the Standard Model. So we expect to have bound states of quarks (hadrons) and. V (r) = − 4 αS ′ ′ + k1 L · S + k2 µq µq ¯ 3 r where the Casimir factor 4/3 comes from group theory. an electron around a proton in an atom or around a positron in positronium): αEM V (r) = − + k1 L · S + k2 µe µp r Where the last two terms are the ﬁnestructure and hyperﬁne splitting terms respectively. In this way. where the ones created close to the original quark in the string carries most of the quarks momentum. if enough energy is added. Conﬁnement. . So. we will ﬁnd something else. the colourﬁeld between two colourcharged particles is compressed into a ﬂuxtube with approximately constant transverse size d⊥ ≈ 0. in particular the longitudinal W + W − → W + W − cross section.
.Carful analysis indicates that diﬀerent breakups along the string are causually disconnected. In the same way a r¯ is colourless. This is allowed by the gluon self interaction (we have no γγ bound states). It turns out that a reasonable deﬁnition of “distance” is rapidity. 20 . have more complicated hadrons with quark content q q q q . in principle. we will get “baryons” with N “quarks”. the maximum rapidity of a ¯ √ hadron is ymax ∼ ln s/m⊥ and the average number of hadrons are √ s + const Ntot ∝ ln m⊥ 16 Light hadrons All hadrons are coloursinglets. or a spin 0 system Sz = 1 :  ↑↑ 1 = 0 : √2 ( ↑↓ +  ↓↑ ) = −1 :  ↓↓ 1 Sz = 0 : √ ( ↑↓ −  ↓↑ ) 2 So there are two way of getting a “spin 0” system. In QCD we can. Now. Note that it is not enough that a q q pair in a meson ¯ has opposite colour charge. this means that the average “distance” between the breakups must remain the same independent of the “length” of the string. This should be s compared to the typical binding energy in nuclear physics where the binding energies are O(10 MeV). Similarly we have baryons with 1 1 qqq = √ (rgb − rbg + brg − bgr + gbr − grb ) = √ εijk qi qj qk 6 6 Note that for any SU (N ) group. However these states will mix with ordinary meson states and are diﬃcult to observe. Another option for coloursinglet states is the glueball gg. . q q qqq etc. r but not a singlet as colour will appear under a phase rotation. So far there are only indications. Hence we expect dn/dy to be constant. either as a spinone system S = 1. where n is the number of breakups. . ggg. or the number of hadrons. √ we have a string between a q if + − and a q in an e e annihilation with a collision energy s. Instead a meson is given by 1 q q = √ (r¯ + g¯ + b¯ ) ¯ r g b 3 (where the + is a phase convention) S = 0. 1 E + pz e + pz y = ln = ln 2 E − pz m⊥ since the diﬀerence between two rapidities is invariants under Lorentz transformations along the string direction (z). The ¯ ¯ ¯ latter are usually called pentaquarks and have recently been discovered for the ﬁrst time (uudd¯ → nK + ) corresponding to a nK + binding energy of O(100 MeV). Compare to the spin of a twoelectron system. where the individual spins can combine in four diﬀerent ways.
rather it is relative. Again there is mixing ¯ with normal mesons and no clear evidence for these hermaphodites has been found.) • mass 3 • internal spin S. ms ≈ 500 MeV. η ′ u ¯ s If you want to see more states: pdg. Baryons S = 1 . but we can try to use the atom as an analogy to get m= i mi + k i<j µi µj = ¯¯ i mi + k i<j ¯¯ Si Sj mi mj where k ∝ Ψ(0)2 and is ≈ m2 ·640 MeV for mesons and ≈ m2 ·200 MeV for baryons. It turns out to be very messy to calculate the masses of the hadrons. 1. J = S + L. (Often also called spin. The absolute parity of a state is not completely deﬁned. instead this is included by using constituent quark masses: mu ≈ md ≈ 330 MeV. 2 • internal angular momentum L = 0. η. decay patterns etc. 2. Mesons S = 0. q q g). mc ≈ 1600 MeV and mb ≈ 5000 MeV. ¯ ¯ ¯ • total spin J. dd.gov We always have the standard charge (C) and parity (P ) symmetries. although all states are not eigenfunctions of these symmerties. These can always be applied to hadron state. s¯ ⇒ π 0 . 2 . Note that the ﬂavourdiagonal states are special since they can mix: u¯. L = P (a)P (b). since odd parity). L (−1)L = Pa Pb (−1)L ab. . It is a spin0 state so 1 ¯ 1 3 ¯ ¯ ¯2 ¯2 S1 · S2 = (S 2 − S1 − S2 ) = (S(S + 1) − S1 (S1 + 1) − S2 (S2 + 1)) = − 2 2 4 so mπ = 2 · 330 − (OK. . 1. L 21 3 · 640 = 180 MeV 4 . If we have a state of two particles a and b with a parities Pa and Pb with an engular momentu L (assuming it is an eigenstate to L) P ab. ¯ Lets look at a π + = ud.lbl. . • Radial excitation n. it’s not perfect). For the mesons the main ones are the vector mesons and the pseudoscalars (scalars since spin0. We classify hadrons by: • ﬂavour contents (gives charge.) • Relativistic corrections and mixing (makes everything quite messy). In Kane there are lists of meson and baryon states. For the light quarks these masses are very diﬀerent from the mass of a freely propagating quark.In principal we can expect states with a valence gluon (eg. but pseudo. This u u formula does not include the binding energy.
6 GeV. Also we have eg. One found an extremely narrow resonance in e+ e− → µ+ µ− and also in pp (u¯ → µ+ µ− ). S = (−1)L+S a¯. if the quark numbers and energy ¯ can be conserved. a but for the ﬂavourdiagonal states we have Ca¯.the charm quark. and Pa Pb = −1 for a fermion–antifermion pair and +1 for a boson–antiboson pair. S a a a a There are three diﬀerent ways a hadron can decay: Strong Decay will always dominate if it is possible. Today we call it J/ψ and it has a mass of 3. The resonance u was called J by one experiment and ψ by another. π 0 → γγ (1. For charge conjugation the mesons are not in general eigenstates: Ca¯ L.1 GeV and a width of only 70 keV. Strong isospin symmetry would mandate that Γ(∆++ → pπ + ) = Γ(∆+ → nπ + ) up to spin factors. The ∆++ has a lifetime τ ≈ 10−23 s ⇔ Γ ≈ 100 MeV (diﬃcult to calculate). Also we can predict the branching ratios of the K ⋆+ which can decay ¯ into K + π 0 or K 0 π + (u¯ → u¯ + u¯ and u¯ → d¯ + ud). eg. L. rather it is due to the fact that the strong force (gluon) cannot tell the diﬀerence between a u and a d quark. ¯ Weak decay does not conserve quark numbers as in π + → µ+ νµ (ud → W + → µ+ νµ ). 0 → 0. The lifetimes are 10−20 − 10−10 s and are typically approximately calculable. S = ¯a. 0 ). Strong decays respect strong isospin conservation. Σ0 → Λγ which includes a spin ﬂip in the light diquarks ((ud)1 s → (ud)0 s). It is only an approximate symmetry due to the (accidental?) fact that mu ≈ md . which is not due to a gauge symmetry. Conserves quark number but not necessarily strong isospin eg. 1. b. = √  . − 1. c¯) so strong decay is impossible and the particle is longlived. L. S = (−1)(−1)L (−1)S+1 a¯. Also the mass of the charm mass is ≈ m(J/ψ)/2 ≈ 1. 22 . ¯ u It turns out that m(J/ψ) < 2mD (where D = cd. S . Electromagnetic decay Will dominate if possible and no strong decays are allowed. 0 0. L. L. ¯ ¯ These are only important if no strong or electromagnetic decays are possible.where the (−1)L comes from the spherical harmonics. ∆++ → pπ + (uuu → uud+ du). The lifetimes are 10−10 − 103 s and are typically approximately calculable. The only possible explanation was the bound state of a new heavier quark–antiquark ﬂavour . L. 17 Heavy Quarks In November 1974 the October revolution occurred in particle physics. 0 ± √  . 1 2 2 3 2 2 3 2 2 2 so BR(K 0 π + ) = 2BR(K + π 0 ) = 3 . S = ±¯b. If we describe the strong isospin s s u s s with the total spin and the projection on the u/d states we have 1 1 1 1 1 2 1 1  . ie.
and they can only decay weakly. We use the normalization F (0) = 1 and for a smeared charge distribution we expect F (Q2 ) → 0 as Q2 → ∞. cτ ∼ 0. They eventually also found a pseudoscalar 1 S0 = ηc . 3 S1 (3s) = ψ ′′ . strong decays are possible and the resonances become very wide. however. However. D0 = c¯ u + and Ds = c¯. The virtuality of the exchanged photon propagator is denoted Q2 = −q 2 = −(k − k ′ )2 = 2kk ′ − 2m2 ≈ 2EE ′ (1 − cos θ) = 4EE ′ sin2 θ/2. . 23 . We interpret this as an exchange of a virtual photon with momentum q = k − k ′ . . p If W 2 = m2 we have elastic scattering. This is why we can observe it with a secondary vertex. ¯ Γc ∼ G2 m 5 F c (1 + 1 + 3) 192π 3 In 1977 the b quark was discovered (more or less the same story. Sure enough they found excited states 3 S1 (1s) = J/ψ. 3 PJ = χc .4 mm. here the corresponding decay b → t[W − → νµ µ− ] is kinematically forbidden due to the large mass of the top. If W 2 > m2 we have inelastic scattering. P ) Q2 ¯ we don’t measure the hadronic ﬁnal state so we write the last factor as e(P ′ γµ P )F (Q2 ). but it was not predicted in the same way as the c quark). ie. If W 2 ≫ m2 we have deeply inelastic scattering. ﬁring electrons with a ﬁxed energy at protons and measure how they scattered: e− (k)p(P ) → e− (k ′ )X(P ′ ). ¯ The D mesons come in diﬀerent ﬂavours depending on the light quark: D+ = cd. but it is heavily suppressed by the CabbiboKobayashiMascawa matrix element and b is therefore much more longlived as compared to the charm quark. 18 Deeply inelastic scattering Rutherford found the atomic nuclei by scattering α particles oﬀ gold atoms. e We also denote the invariant mass swuared of the hadronic ﬁnal state W 2 = (P + q)2 = m2 + 2P q − Q2 . We can s calculate the decays in the same way as for the muon c → s[W + → µ+ νµ ] and similarly ¯ + + + ¯ and we get for W → e νe and W → ud. they are rather longlived. the p p proton “breaks up”. In 1968 at SLAC the did a similar thing. 3 S1 (2s) = ψ ′ .so we have nonrelativistic motion of the quark inside the meson and it is like a c¯ atom c which means we can do “spectroscopy”. but with a form factor F (Q ) which can be identiﬁed with the Fourier transform of the spatial charge distribution in the proton. Note that as soon as the mass of these states bcome larger that 2mD (the eﬀective ionization energy). as an elestic scattering. Exchanging the top with a charm quark works. 2 ie. p We can sketch the matrix element of this scattering ¯ M ∝ e(k ′ γ µ k) 1 ¯′ e(P . The ﬁnal momentum of the scattered proton system is P ′ = P + q = P + k − k ′ .
dQ2 dx Q i If we now deﬁne the structure function F2 (x) = i ⇒ x= Q2 ≡ xBj 2P q Q2 xfi (x) i we can write the cross section πα2 F2 (x) dσ = 4 2(1 + (1 − y)2 ) dQ2 dx Q x where the last factor is obtained from proper spin counting with y = P q/P k. so we will get 1 1 1 ≈ 4 ⇒ 2 Q2 (mW − q 2 )2 mW 24 . x. If the partons has a momentum fraction x of ˆ ˆ ˆ the protons momentum. Here we can do the same calculation as before with some important modiﬁcations. dQ2 Q i Now we can do the same thing as when we calculated W production in p¯ collisions. Here we can write down the matrix element of the partonic scattering 1 s ˆ ¯′ ˆ ¯ ˆ M ≈ e(k ′ γ µ k) 2 ei (P γµ P ) ≈ ee′ −q −Q2 where s = (k − xP )2 ≈ x(k + P )2 = xs. Hence we have ˆ where P ˆ P ′2 = (xP + q)2 = x2 P 2 + 2xP q + q 2 ≈ 2xP q − Q2 = 0 the Bjorken x ≈ Q2 /(Q2 + W 2 ). if the proton consists of pointlike constituents. i. the probability of ﬁnding a parton. p we use the parton density. First the propagator particle is now massive. ie. we are looking at the process e− (k)q(P ) → e− (k ′ )q(P ′ ). and get the full diﬀerential cross section dσi πα2 ≈ 4 Q2 fi (x). the fraction of the electron momentum taken by the photon. P = xP . fi (x).Now. we can interpret the results as if we are elastically scattering on these partons. In the rest frame of the partonic scattering we ˆ have √ √ s s ˆ ˆ ˆ Q2 = −(k − k ′ )2 ≈ 2 (1 − cos θ) 2 2 so s ˆ s ˆ ˆ ˆ dQ2 = d(cos θ) = dΩ 2 2 · 2π and with Qi = ei /e and α = e2 /4π we get the cross section πα2 σ dˆi = 4 Q2 . with momentum fraction. ˆ 2 = P ′2 ≈ 0 for nearly massless partons. In an alternative experiment they used a “beam” of muon neutrinos which scattered oﬀ a proton by exchanging a W + producing a µ− which could be measured.
Also a gluon can emit a gluon.2mq < s demonstrating the charges of the quarks and the fact that they come in three diﬀerent colours. The higher the virtuality. and it is calculable δf( x. s → c and c → s. This is called scaling violations. The P functions (where Pi = j dzPi→j (z)) are called (Altarelli–Parisi) splitting functions. not all quark species can be involved. we can try to disentange the diﬀerent quark distributions by combining νp and ep DIS in diﬀerent ways. . . so the xf (x) is depleted at high x and increased at low x. the ﬁner details in the proton we can see and also the more details of the cloud of virtual partons around the valence quarks. fi (x) ⇒ fi (x. only d → u. Q2 ) αs ∝ −Pi fi (x. If we increase Q2 we can see that a parton at high x actually consist of two (or more) partons at lower x. while the second is the gain of a parton at a higher momentum fraction being resolved into one at x. On the other hand we have the virtual photon. The valence quarks are surrounded by a cloud of ¯ virtual partons. and a gluon can split into a virtual q q pair. Q2 )Pj→i ( ) z z 19 e+e− physics ¯ We have the process e+ e− → γ ⋆ → f f (where the γ ⋆ → Z 0 or a mixture of the two at higher energies. ¯ i Hence. A quark can emit a gluon and reabsorb it again.¯ Also. It is also possible to conﬁrm the spin1/2 nature of the quarks by looking at the angular distribution of jets dσ(e+ e− → q q ) ¯ ∝ 1 + cos2 θ dΩ 25 giving the cross section . Q2 ). We can approximate the transverse size of the photon ∆r ∼ 1/∆p⊥ ∼ 1/Q. The ‘Q2 dependence is only logarithmic. So ¯ ¯ ¯ i Q2 xfi (x) ⇒ xfd (x) + xfu (x) + . The picture of the proton becomes a bit more diﬃcult if we consider virtual ﬂuctuations. The matrix element is easy to write down (for s << m2 ) Z 1 ¯ M = ee (¯γµ e) ef (f γ µ f ) e s 4π Q2 α2 f σ= 3 s One of the main successes of the quark model is the experimental conﬁrmation of the ratio σ(e+ e− → hadrons) Q2 =3 Rhad = q + e− → µ+ µ− ) √ σ(e q. Q2 ) + 2 δ ln Q 2π 1 j x Here the ﬁrst term corresponds to the loss of a parton at x being resolved into two with lower momentum fractions. The end result is that the parton densities are Q2 dependent. it is shortlived with a small wave function centered along the direction of the electron. x dz fj (z. u → d.
At DESY they also managed to conﬁrm the existence and spin1 nature of gluons by studyng threejet events e+ e− → q q g. The way we redeﬁne α is easiest to see if we consider a eγ vertex e(k) → γ(q)e(k ′ ). . while in any measurement we never have eg. The trick is to redeﬁne α so that the divergencies in R1 are cancelled. By stydying fourjet events e+ e− → q q gg and ¯ ¯ e+ e− → q q q ′ q ′ it has been possible to conﬁrm the selfcoupling among gluons. if we want to calculate an observable R (eg. even though we cannon directly measure α. Here the second order diagram contains a loop where the photon splits into an e+ (p)e− (q − p) pair which then annihilates into a photon again. for this integral (which we will send to inﬁnity later). since they are bare couplings. To continue we need to introduce an ultraviolet cutoﬀ.) R0 is typically a ﬁnite number calculable with using a Feynman tree diagram. however. typically involves a divergent integral over a momentum in a loop diagram. ¯ ¯ 20 Running couplings The couplings in the standard model Lagrangian are in principle impossible to measure. a completely “bare” electron as it is always surrounded by an electromagnetic ﬁeld which will contain virtual ﬂuctuations. In the matrix element we will get the two terms e¯(k ′ )γ µ u(k)εµ − u d4 p u u 1 [e¯(p)γµ u(p − q)][e¯(p − q)γ λ u(p)] [e¯(k ′ )γ µ u(k)] 2 u ελ (2π)4 q (p2 − m2 )((p − q)2 − m2 ) e e We here recognize the three propagators and we get a minus sign due to the fact that this is a fermion loop (a boson loop would give a plus sign). R1 . 3π me 3π me 3π Q 26 . But we know we can make a measurement of R (at some scale q 2 ) so we should be able to calculate R(q 2 ) = Robs (q 2 ). Λ2 . Hence. α → α(q 2 ). a cross section or a decay width) we get a perturbative series R = αn (R0 + R1 α + R2 α2 + . If Q2 = −q 2 ≫ m2 we e have q 2 x(1 − x) Q2 ln 1 − ≈ ln 2 m2 me e and we have I(Q2 ) ≈ α α Λ2 Q2 Λ2 α ln 2 − ln 2 = ln 2 . . We can simplify this to e¯(k ′ )γ µ u(k)εµ (1 − I(q 2 )) u where the integral I(q 2 ) is dimensionless an can be simpliﬁed to I(q 2 ) = α 3π ∞ m2 e dp2 2α − p2 π 1 0 dx · x(1 − x) ln 1 − q 2 x(1 − x) m2 e where the ﬁrst term is clearly divergent.
We can now send Λ → ∞ (while the bare α0 → 0) and we can express the coupling as measured at one scale in terms of the coupling at another scale. It is easy to see that we now get α(Q2 ) = with N = nl + 3 α(µ2 ) 1 − N α(µ ) ln Q2 3π µ 2 3 2 2 2 nu + 3 1 3 2 nd The same procedure can be repeated for QCD. but we can also have muons and quarks as long as their squared mass is less than Q2 (in principle we can also have heavier fermions in the loop.Now. but in the other direction. The closer we look the more smeared out it is. but their contribution is tiny. For large Q2 the coupling decreases and we get asymptotic freedom. The picture is that if a red quark emits a red–antiblue gluon and turns blue it can be viewed as if the original red charge is smeared out. some “original” α which we “normalize” by measuring α(µ2 ) α(µ2 ) = then we can calculate α(Q2 ) = 1 α(µ2 ) + . . looking at a quark–gluon vertex and inserting quarkloops. = M0 α 1+ α 3π ln Λ2 Q2 → M0 αobs (Q2 ) 1 1 α0 + 1 3π ln Λ2 µ2 ⇒ 1 1 Λ2 1 − ln 2 = α0 α(µ2 ) 3π µ 1 2 1 − 3π ln Λ2 + µ 1 3π ln Λ2 Q2 = 1 1 α(µ2 ) + 1 3π ln Q2 µ2 = α(µ2 ) 1+ α(µ2 ) 3π µ ln Q2 2 which is quite ﬁnite. we can have another. We then get αs (Q2 ) = αs (µ2 ) 1− nf αs (µ2 ) 2 3π ln Q2 µ 2 but we will also have gluon loops in the gluon propagator. So far we only had electrons in the loop. and since this is a boson loop it comes with the oposite sign. if we can have one loop. and the coupling is running. but for larger Q2 . . The physical interpretation is that the charge around an electron is screened. but that will just give the same integral squared M = αM0 1 − α Λ Λ α ln 2 + ln 2 3π Q 3π Q 2 2 2 Assume we have α0 . 27 . On the other hand for small Q2 the coupling diverges and we expect perturbation theory to break down. we probe the electron at smaller distances where it is less screened and the eﬀective coupling increases. Doing the math we get αs (Q2 ) = αs (µ2 ) 1+ αs (µ2 ) (33 12π − 2nf ) ln Q2 µ 2 Again we have a coupling which runs.
where 2 ΛQCD αs (µ2 ) = −(33 − 2nf ) ln 12π µ2 We then get αs (Q2 ) = 12π Q2 (33 − 2nf ) ln Λ2 QCD As Q2 increases. 22 Mixing angles If we for a moment limit ourselves to two families. One example thing is to look for eﬀects of massive neutrinos in neclear βdecay (later). and to have a continuous coupling one usually have ΛQCD (nf ). hoping to ﬁnd new particles. Although we have not found the Higgs (yet) there is really no hard evidence of any physics Beyond the Standard Model (BSM). like myself. Of course every particle physicists dream is to ﬁnd something which does not ﬁt into the standard model (others. K + → π + eµ (BR < 2 · 10−10 ). very expensive and diﬃcult. s Here we assume that the quark ﬁelds are both mass eigenstates and weak eigenstates. But his is not necessarily true. Z 14 Low Energy and NonAccelerator Experiments The Standard Model works extremely well. Conventionally we present measurements of αs at the scale m2 . we can write the charged current in the Lagrangian as d µ ¯ ¯ Jch = u c γ µ PL . 28 . however. more and more quarks will contribute to the running. tries to improve our understanding of the SM so that we have a chance to ﬁnd deviations). which necessarily will be associated with BSM physics. which constantly bombard us with high energy particles (these are. and that is to search for lowenergy processes which are very rare or even forbidden in th SM. Examples of decays which are forbidden in the standard model are leptonfamilynumberviolating processes such as µ− → e− γ (current limit BR < 5 · 10−11 ) or µ− → e− e+ e− (BR < 10−12 ). There is one thing one can do without high collision energies. Also some hadron decays are forbidden: K 0 → µe (BR < 3 · 10−11 ). Especially forbidden is proton decay (τp > 1032 y). rather diﬃcult to controll). however.Now let’s deﬁne ΛQCD as the scale where αs diverges ie. which then traverses a medium where positive helicities travels faster than negative helicities. Building larger and larger accelerators is. One analogy is to consider a light source which emits transversely polarized light. In fact it turns out to be false. Alternatively one can use “natural” accelerators which seem to exist in the universe. To do this one can either try to crank up the energy in the collision experiments.
ε− ). there would have been no mixing between families and the squark (and the lightest strange hadrons) would be stable. when we calculate cross sections for “free” propagating quarks. What enters into the quark–W vertex is the weak eigenstates. Now. u c t ¯ ¯ b 29 . We can generalize this to three generations µ Jch = d ¯ γ µ PL V s . These mass eigenstates. is the d′ . The phases in V above can be absorbed in the quark states and we are left with V = cos θ sin θ − sin θ cos θ Expressed in the mass eigenstates we now have the current µ Jch = uγ µ PL d cos θ + uγ µ PL s sin θ − cγ µ PL d sin θ + cγ µ PL s cos θ+ ¯ ¯ ¯ ¯ √ And we get the new couplings to the W by multiplying by g2 / 2.and is ﬁnally detected after a tranverse polarizer. Note that this does not change the neutral current at all since. but these would always show up together in any electroweak vertex as V u† V d which itself is a unitary matrix. εy ) while the propagating eigenstates (the “mass” eigenstates) are the circular polarizations (ε+ . In principle we could do the same thing with the uptype quarks and get both a V u and V d . but what we want to have in matrix element. Let’s call the weak eigenstates of the downtype quarks s′ can be tranformed by a unitary transformation to the mass eigenstates and vice versa. it turns out that the mass and weak eigenstates of quarks are not the same. Note that the “Cabbibo” angle is a completely free parameter of the Standard Model. d′ s′ =V L d s L The unitary matrix can be written completely generally as follows V = And we get the charged current µ Jch = cos θeiα sin θeiγ −iγ − sin θe cos θe−iα d′ s′ u c γ µ PL ¯ ¯ = u c γ µ PL V ¯ ¯ d s . If nature has chosen the value 0. just looking at the lefthanded s and d part of current µ Jneu = ¯ ¯ d s L V † T3 − Qf sin2 θW V d s = ¯ ¯ d s L T3 − Qf sin2 θW d s it will be unchanged since V † V = 1. Hence the production and detection eigenstates are the transverse polarizations (εx . so we only need to consider the downtype quarks.
where we expect all KS will decay and quickly leaving a pure KL beam. since P will change a righthanded particle to a lefthanded one. which furthermore means that the the theory is not completely invariant under timereversal transformations. 30 . we could expect the theory to be invariant under the simultaneous CP transformation. which will give us back a component corresponding to KS . It turns out that the absolute values of the matrix elements coupling the third generation to the others are fairly small (Vcb  ≈ 0. these combinations have diﬀerent decay modes (eg. and while the latter couples to the W .003).V is now a 3 × 3 unitary matrix which means that it can be described by nine diﬀerent parameters. after a while we will have diﬀerent mixtures of these states. where the former is odd and the latter is even under CP transformation. But if CP is s ¯ conserved we can create two combinations which are eigenstates: KL = K0 − K 0 and 0 ¯ KS = K0 + K . the KL and KS are not necessarily mass eigenstates. and it turns out that the KL is much more linglived that KS (hence the names long (≈ 5 · 10−8 s) and short (≈ 9 · 10−11 s)). If ¯ ¯ we look at K 0 = d¯ and K 0 = sd. But if KL is not a mass eigenstate. But we are left with four parameters which are all completely free and need to be measured. As for the twoﬂavour matrix we can get rid of some of these by redeﬁning the phases of the quark wave functions. the charge (C). but it turns out that the standard model is not quite invariant under CP and therefor not under T transformations. The reason for this “CP violation” is the remaining the phase in CKM matrix which we couldn’t get rid of by redeﬁning the phases in the quark wavefunctions.03 and Vub  ≈ 0. the former does not. and also complex terms in the Hamiltonian. Wµ Jch . If they were. This means that we get a camplex µ current. Three of these can be expressed as real angles similar to the Cabbibo angle. the τ which can simply decay into a neutrino. 24 CP Violation There are three important discrete symmetries in nature. This is diﬀerent from the corresponding lepton. neither of these are CP eigenstates. but there is stil one relative phase which we cannot get rid of in this “CabbiboKobayashiMascawa” mixing matrix. Now if we look at the standard model we immediately ﬁnd that P is explicitly broken in the coupling to the W . Now. In the same way it is clear that C is violated. There is a theorem saying that any quantum ﬁeld theory is invariant under CP T transformation. (Note that this could never happen if there was just two generations). travelling at diﬀerent velocity. parity (P ) and time (T ) reversal symmetries. But if CP is slightly violated. since a lefthanded particle couples to the W while a lefthanded antparticle does not. but made up from two states with diﬀerent masses. Since the b quark cannot decay into a top. However. the dominant decay mode is b → cW − where the width is suppressed by Vcb 2 . One way of measuring Vcb  is therefore to compare Γb 9 m5 b ≈ Vcb 2 5 Γτ 5 mτ where the 9/5 comes from the number of decay modes available from the W . KL → πππ and KS → ππ). we could have a “beam” of K 0 . The ﬁrst place where CP violation was measured was in the decay of neutral kaons.
If the neutrinos are massive. the decay KL → π − e+ νe would be exactly the same as KL → π + e− νe . Z + 1) + e− + νe → (A. So we will never see it directly. if that is inconsistent 0 ¯0 with the K K CP violation. just like what happened for their cousins.00333 ± 0.If CP was exactly conserved. however. One place where it could show up is in neutrinoless doublebeta decay: (A. the nonobservation of this sets the limit mM < 0.2 eV. Z + 1) + e− + νe → (A.1 ν1 (0) + Uµ. We can expect other sources of CP violation besides the phase in the CKM matrix. is violated in the Standard Model. the downtype quarks. it would have T3 = 0 and no colour. And upuntil a few years ago they seemed to be. it has zero charge. Z + 2) + e− e− ¯ So far. The second way we could have neutrino masses is if the neutrino. But it turns out that we can measure a small asymmetry: ¯ Γ(KL → π − e+ νe ) − Γ(KL → π + e− νe ) ¯ = 0. hence the only way it would interact is with the Higgs and via gravity.3 ν3 (0) 31 νe ν1 νµ = U ν2 ντ ν3 . was its own antiparticle. where νL is righthanded. there may be other sources. no ` priori reason we know about for the neutrinos to be massless (or a not).19 MeV. which violates ¯ c ¯ fermion number. in π + → µ+ νµ ) in a weak eigenstate at time 0: νµ (0) = Uµ. however. The masses could arise in (at least) two diﬀerent ways. there is a possibility that the masseigenstates are not the same as the weak eigenstates. Z) → (A. the decay π + → µ+ νµ gave the limit m(νµ ) < 0. In that case we could get terms in the Lagrangian looking c like mM νL νL . just as the photon and Z 0 . If there exists a righthanded neutrino (and lefthanded antineutrino) we would have a coupling to the higgs ﬁeld: gf ν φνR and a (Dirac) mass term mD ν νR . Yet.00014 − e+ ν ) − Γ(K → π + e− ν ) Γ(KL → π ¯e e L So CP . 24 Neutrino masses and oscillations In the standard model the neutrinos are typically assumed to be massless.2 ν2 (0) + Uµ. Eg. So how do we see ¯ ¯ a righthanded neutrino? Well. not extremely strict. There is. If we study CP violation. and therefore also T . The experimental limits on their masses were. from betadecay of tritium we had m(νe ) < 3 eV. and m(ντ ) < 18 MeV came from τ → ντ + nπ. But this would give us νL ↔ νR . we may eventually understand why there are so many more particles tha antiparticles in the Universe. This hasn’t been seen. Denoting the mass eigenstates νi we have Now a neutrino would be produced (eg. One ¯ way of checking this is to measure also CP violation in the B 0 B 0 .
ij Note that there is also some oscillations induced when neutrinos traverses matter.27 W2 E 32 .and muonneutrinos. if we start out with a pure νe we have νe (0) = ν1 (0) cos α + ν2 (0) sin α and νµ (0) = −ν1 (0) sin α + ν2 (0) cos α = 0 but it is the mass eigenstates which propagates. νe e− → e− νe . E in GeV and ∆m2 in eV2 . however weakly. This is because neutrinos interacts. This will give an eﬀective mass of the neutrino which is diﬀerent from the mass in vacuum.3 ν3 (0)eiE3 t Let’s simplify and only use two generations. α νe νµ = cos α sin α − sin α cos α ν1 ν2 Now. with matter with Z 0 exchange. electron neutrinos can also react with the electrons in matter through W exchange.27∆m2 · E/L ij ⋆ ⋆ ℑ(Uαi Uβi Uαj Uβj ) sin2 2. Including also the third family we can write the probability for oscillation from family α to β P (α → β) = δαβ − 4 + 2 i>j i>j ⋆ ⋆ ℜ(Uαi Uβi Uαj Uβj ) sin2 1.1 ν1 (0)eiE1 t + Uµ. While Z 0 exchange is exactly the same for electron. which means that we can describe the mixing matrix with only one angle. so that νµ (t) = Uµ.but then we will have a propagation of the mass eigenstates. We can write √ d νe νe νe E1 0 2GF Ne 0 † i =U U + νµ 0 E2 νµ 0 0 dt νµ which will give us P (νe → νµ ) = sin2 2θ 2 W ∆m2 L sin 1.54∆m2 · E/L ij where we have scaled things so that L should be given in km. so after some time t we will have νµ (t) = −ν1 (0)e−iE1 t sin α + ν2 (0)e−iE2 t cos α and there is a probability that the νe has turned into a νµ  νe (0)νµ (t) 2 = sin2 α cos2 α e−iE1 t − e−iE2 t = sin2 2α sin2 (t(E2 − E1 )/2) where E2 − E1 = 2 2 ∆m2 2m∆m ∆m m2 − m2 E2 − E1 1 = ≈ = ≈ 2 E1 + E2 2E 2E 2E γ where the latter can be seen as a boosted version of the fact that also a neutrino at rest has no deﬁnite weak eigenstates.2 ν2 (0)eiE2 t + Uµ.
while if they come from below they have been traveling ∼ 104 km through the earth. And we do. produces mainly muons (together with νµ ) which in turn decays (giving another νµ and a νe ). mν /me ∼ 10−7 . they have traveled a short distance in the atmosphere. not to νe but to ντ . me /mW ∼ 10−5 . eight mixing parameters. Wouldn’t it be much nicer if we had a theory where these could be predicted? • Many things seems arbitrary. There is the fact that baryonic matter only makes up a fraction of the 33 . The latter has been suspected for a long time from the fact that only half of the expected number of neutrinos from the sun were found – but nobody really trusted the models for neutrino production in the sun enough to make a deﬁnite statement. These are produced when charged particles from space hits the atmosphere and produces mainly pions which. mZ ). It works very well and we hate it! Why? • There are too many free parameters. eg. The standard model. Why are the lefthanded fermions in SU (2) doublets and the righthanded in singlets? • Why do we just have SU (3) × SU (2) × U (1)? • Why is there charge quantization? In principle Y in U (1) could be anything. By looking at both νe and νµ it is found that there is a big eﬀect of νµ disappearing. Now if we detect these and look at the direction they come from. if they come from above. three couplings and the higgs ﬁeld parameters (µ.with W 2 = sin2 2θ + √ 2GF Ne 2E − cos 2θ ∆m2 2 The signal for this is eg. We can study atmospheric neutrinos. λ) ⇔ (mh . From this one gets a mass diﬀerence which is larger than the mass diﬀerence responsible for the νe ↔ νµ oscillations. And the fact that we have such huge ratios of similar parameters. Now there is ample evidence for neutrino oscillations and hence also for neutrino masses (or rather mass diﬀerences). These are typically much more energetic than solar neutrinos. in turn. Then there is the obvious problem that gravity does not ﬁt in anywhere. Twelve fermion masses. the appearance of a νµ where we only expect νe or the disappearance of νe . mW /mPl ∼ 10−17 . The evidence comes both from solar and atmospheric neutrinos as well as from ’man made’ neutrino sources (reactors and accelerators). Hence we can vary L by varying the polar angle and we should see oscillations. The current numbers are approximately ∆m2 sol 2 ∆matm θsol θatm ≈ ≈ ≈ ≈ 8 · 10−5 eV2 3 · 10−3 eV2 34◦ 37 − 53◦ 26 Open issues So there you have it. Why are there three generations.
So where are the other quarks and leptons? Well the quintet above is actually the antiquintet of the group. 27 Grand Uniﬁcation Just as the electroweak symmetry is spontaneously broken by the Higgs ﬁeld. In the fundamental representation of an SU (N ) group the generators are traceless N × N matrices. It is however instructive to look at it anyway. The simplest such Grand Uniﬁed Theory (GUT) was invented by Georgi and Glashow and is now excluded by date. U (1)Y × SU (2)L → U (1)EM . It assumes that there is only one symmetry group which is spontaneously broken into the standard model. we could try to imagine that the whole structure of the standard model is a spontaneously broken version of a higher symmetry. SU (5) → SU (3) × SU (2) × U (1) We then have the fermions in a quintet ¯ dr ¯ db ¯ dg e− νe L ¯ ¯ where now dL = dR is an weak SU (2) singlet. so that 3Qd + Qe = 0 and Qd = −1/3 (and also Qu = 2/3) and we have charge quantization. Qp + Qe = 0. where the diagonal generator will give the charge. We also have the antisymmetric 5 × 5 decuplet (cf. there is a problem with the higgs mass. just as couplings depend on the scale due to loop corrections and need to be renormailized (looking just at QED) α ∞ dp2 α Λ2 m→m 1+ →m 1+ ln 2 3π m2 p2 3π me e For the higgs the renormalization turns out to be worse δm2 ∝ h ∞ m2 h ∝ Λ2 If there is no additional structure for particles until we reach the Planck scale. SU (3) have 34 . Masses. the Higgs mass we observe at electroweak scales must be the tiny diﬀerence between two huge numbers (the hierarchy problem). The quarks are still associated with the SU (3)t colourtriplet and the leptons are still in the weak SU (2) doublet. which reuires Qi = 0. As if this isn’t enough. We don’t like that.matter of the universe and that matter only makes up a fraction of the energy of the universe (the cosmological constant).
b2 = 22/3 − 4nF /3 and b′1 = −4nF /3 (b′1 is related with the hypercharge rather than the electric charge). g1 . In particular we should be able to derive the ratio between g1 and g2 which is governed by θW . would be derived from the breaking of SU (5) into the SU (3) × SU (2) × U (1). MGUT . If SU (5) is the true symmetry group at high scales the coupling g5 would be the only copuling in the theory. e+ . g3 . db . Qu = Q4. We assume that there is actually only 2 one coupling at som large uniﬁcation scale. dg . The SU (5) group has 52 − 1 = 24 generators. For the decuplets we have Qij = Qi + Qj so eg. which will give us 24 gauge bosons. Then we would have eg. νe )R ¯ and a corrsponding antidecuplet for the righthanded uptype quarks and electrons. 1 b3 M2 1 b2 M2 + ln GUT = + ln GUT α3 (µ2 ) 4π µ2 α2 (µ2 ) 4π µ2 So we can in principle calculate this scale ln 2 6π MGUT = µ2 11 1 1 − α2 (µ2 ) α3 (µ2 ) Inserting experimental data we will ﬁnd MGUT ∼ 1018 GeV. but the broken symmetry gives them diﬀerent values at lower scales. Let’s look at the covariant derivative of SU (5) µ Dµ = ∂ µ − ig5 Ta Ua 35 . but things do not exactly match up.1 = Q4 + Q1 = Qd + Qe . g2 . All other couplings. but also octet gluons) 0 u −¯b 1 √ ug ¯ 2 ur dr ub −¯g ¯ u 0 ur ¯ −¯r u 0 ug ub dg db −ur −ug −ub 0 e+ −dr −dg −db −e+ 0 L Note that there are ten independent ﬁelds here. Then we have the quintet of the righthanded ﬁelds (dr . lets denote them ¯ ¯ Xr Yr 2B ¯ ¯ Xg Yg gij − √30 δij ¯ ¯ Xb Yb Xr Xg Xb Yr Yg Yb W3 √ 2 3B + √30 W+ 3 3B √ W− − W2 + √30 We can write the running of the couplings in the following way 1 bi M2 1 = + ln 2 αi (M 2 ) αi (µ2 ) 4π µ where we have b3 = 11 − 4nF /3.triplet quarks.
Z 0 and γ. And evolving down to the weak scale we are not too far oﬀ. X → uu. 10A5. eg. hence. Y → ud. and couplings are running and. but remember sin2 θW is a ratio between couplings. we get terms in the Lagrangian corresponsing to new vertices. The 24 gauge bosons included the eight gluons and the W ± . The current limit p → π 0 e+ is τp > 1033 years. . R. etc. Careful analysis tells us that the simplest GUT. . . The problem with this is that the proton can decay: p = u ud → u Y → u ue+ → π 0 e+ ¯ The anology with weak muon decay 2 G2 m 5 α2 m5 F µ µ Γµ = = 4 2 192π 384π mW we get something like 2 Γp ∝ α5 m5 p 4 mY A proper calculation using mY ∼ mGUT ∼ 1015 GeV will give τp ∼ 1031±2 years. Y → e+ u ¯ and we conclude that X and Y has charges 4/3 and 1/3 respectively.23. From this we can identify e = g5 and Q = T3 − cot θW T1 ≡ T3 + cT1 . .And Pick out the parts relevant to the electroweak sector using B µ = Aµ cos θW + Z µ sin θW µ W3 = −Aµ sin θW + Z µ cos θW µ Dµ = ∂ µ −ig5 (T3 W3 +T1 B µ +. based on SU (5) is excluded by date. X. so is sin2 θW . Most of these suﬀer from baryonnumber nonconservation and other maladies. = ∂ µ −ieQAµ +. for any representation. and currently GUTs are not very popular. but one could easily imagine higher symmetry groups. ¯ 5A5. ¯ X → e+ d.) = ∂ µ −ig5 sin θW (T3 +cot θW T1 )Aµ +. Now. ¯¯ Y → dνe . . An example is SO(10) which includes a righthanded neutrino an extra Z boson and much more. Sandwiching the gauge ¯ boson matrix between the fundamental representations. Y and Y which all come in three diﬀerent colours. of a group we have orthogonality and equal normalization of the generators Ta . so that TrR Ta Tb = NR δab So 2 2 since TrT3 = TrT1 and 2 TrQ2 = Tr(T3 + cT1 )2 = (1 − c2 )TrT3 sin θW 2 2 1 TrT3 = = = 1 + c2 TrQ2 0+0+0+ 1 3 2 1 2 1 3 2 + 2 1 2 2 + 1 3 2 = + +1+0 3 8 Now. 36 . The other twelve ¯ ¯ are X. this is fairly far from 0. .
Q} = {Q. The transformation is actually deﬁned in terms of an algebra where ¯ ¯ ¯ {Q. with an operator Q changing one into the other b = Qf ¯ and f = Qb but leaving any other quantum number as it is. There are many ways of deﬁning supersymmetry. In some sense the transformation of a boson into a fermion √ is the square root of a translation. ν and one for d. But this is not the case. (wino higgsino) mix together charginos χ± ˜i gluino It turns out we need we need two Higgs doublets. (Remember that −1 = i and ∂ µ ∂µ = γ µ ∂µ . Two doublets give eight degrees of freedom. P ] = 0. l.28 Sypersymmetry Postulate there being a symmetry between fermions and bosons. Exactly how it is broken. Basically sypersymmetry means that all particles we know must have a supersymmetric partner with exactly the same quantum numbers but diﬀerent spin. 37 . Q} = 0 and [Q. normal qL qR lL lR νL γ Z0 h0 H0 A0 W± H± g partner spin qL ˜ 0 squarks qR ˜ 0 can mix ˜L l 0 sleptons ˜R l 0 also mix νL ˜ 0 sneutrinos γ ˜ ˜ Z ˜ H ˜ W± ˜ H± g ˜ 1/2 1/2 1/2 1/2 1/2 1/2 1/2 1/2 (photino zino higgsino) all mix together into neutralinos χ0 ˜i 2 + 3 + 1 + 1 + 1 = 8 spin states for the bosons gives four neutralinos with two spin states each. so SUSY must be broken. P ] = [Q. It is possible to derive mass relations between the Higgses and the W/Z. but there are some suggestions. the MSSM. leaving us with ﬁve Higgs particles. Their partners have the same spin and mixes into neutralinos and charginos. we will only look at the simplest minimal supersymmetric extension of the standard model. Q} = QQ + QQ = 2σ µ Pµ where Pµ = i∂µ is a translation. nobody knows. three are “eaten” by the W/Z masses.) We also have ¯ ¯ ¯ {Q. If SUSY was unbroken the masses of any particle would be the same as its superpartner. one to generate mass for u. in particular mh < mZ which would be excluded by experiments if it wasn’t for some radiative corrections giving mh < 103 GeV.
We can deﬁne a “parity” relating to SYSU. they will be produced at the same rate as their ordinary particle equivalents. W − → e− νe . ˜ ˜ ˜ W − → e− νe . Naively we would try to make a quantum ﬁeld theory of general relativity and we would get a spin2 graviton. ∞ gave δm2 ∝ − m2 dp2 . Note that we can get Rparity violation by either lepton number or baryon number violation. ˜ ˜ W − → e− νe ˜ all have the same coupling (g2 ). This also means that the lightest supersymmetric particle (LSP) is stable. called Rparity R = (−1)L+3B+2S where L is lepton number. B is baryon number and S is spin. but if we have both we will get proton decay. χ0 ) it would be produced copiously at the big bang and would ˜ −2 basically still be around. If SUSY is discovered it will give a lot of hard (and boring) work to check its consistency. the couplings do not care if we have particles or sparticles. This solves the hierarchy problem and many ˜ t h t other problems with the standard model. The higher the scale the closer we get. Even if the masses are not the same. In total we have over hundred free parameters which need to be measured. if Rparity is not conserved we could have decays like γ → νγ ˜ W – it could still contribute to dark matter if the decay is slow enough. With Rparity conserved we would get characteristic decay chains of sparticles according to their mass hierarchy. Now we will also have stoploops h t which come in with oposite sign since they are bosons (but the same coupling). The end result is that δm2 ∝ m2 − m2 which is ﬁnite. W − → e− νe . sparticles can only be produced in pairs. u → d + [χ+ → ντ + [˜+ → τ + χ0 ]] ˜ ˜1 τ ˜1 This comes with a price though. There are double the amount of particles and hence particle masses which are free parameters. Now it was said before that the loopcorrections to the higgs mass. String theory How do we include gravity in the standard model. So all ordinary particles have R = +1 and their superpartners (sparticles) have R = −1. leading to the hierarchy problem. and also the 38 . The interaction with matter would be weak since σ(χ0 p) ∝ mq ˜ ˜ (cf. from a toploop. The problem is that in normal QFT we consider particles to be pointlike. Another thing that SUSY can explain is dark matter. However. If Rparity is conserved. But there would also be nonrenormalizable divergencies. There are also a bunch of extra mixing parameters for which there are no predictions. Hence we have that vertices such as eg. σ(νp) ∝ m−2 ). If there is a stable LSP. which only interacts weakly (eg.. So as soon as we come above the massthreshold for producing sparticles. eg. eg.
a spin2 graviton and a spin3/2 gravitino. So. which one has nature chosen? Well. it turns out that we need 26 dimensions.higher the gravitational potential. Things improve if we allow for supersymmetry. The result is that no known renormalization works. What if gravity was allowed to propagate throughout all dimensions. even though we have to live with a whole new zoo of particles living in the other E8 which only interacts gravitationally. called typeI. M theory. rather than being pointlike are actually modes of vibrations on small pieces of string. mass and. Lets look at the gravitational potential around a particle with mass m generalized to 4+n dimensions: m 1 V (r) ∼ n+2 n+1 . what would then experiments tell us about the maximum size of these dimensions? Well. a string will describe a world sheet. recently it has been realized that they may all just be diﬀerent aspect of the same underlying theory. σ). s. the higher the force. It turns out you can construct exactly ﬁve such (super) string theories. the higher energy i. While a particle will describe a world line. for distances much larger than that the 39 . σ)).e. t.and uchannel diagrams) correspond to one and the same string vertex. typeIIB. But let’s go back to ten dimensions where we can construct ﬁve theories which all lack unphysical divergencies and all include SU (3) × SU (2) × U (1) together with supersymmtry. Now since E8 ⊃ E6 ⊃ SO(10) ⊃ SU (5) ⊃ SU (3) × SU (2) × U (1) we can live with this restriction. To avoid tachyons (particles with m2 < 0 which travel and convay information faster than light. where σ is an additional coordinate along the string. How do you explain that there are ten dimensions when we only live in four? Well the extention of the universe in the other six dimensions is extremely small. heterotic SO(32) and heterotic E8 × E8 (heterotic has closed strings). not much.and one time dimension as long as the basic QFT Lagrangian is invariant under SO(32) or E8 × E8 . nor are the vertices. mPl r Now. The idea is that elementary particles. This means that particles are intrinsically extended. xµ (τ. This means that we can imagine that all standard model ﬁelds live on a fourdimensional brade in a higher dimensional universe. Just as particles are not pointlike. which also means that the divergencies are muﬄed. It turns out that M theory requires eleven dimensions. hence. σ + 2π) = xµ (τ. and that they are related to eachother through duality transforms (cf. To construct a string theory which is consistent with QFT is a bit tricky. Rather the vertices are smeared out in a way such that several diﬀerent particle vertices (eg. the fourposition at any given proper time. in which case we only need nine space. xµ (τ ). in addition the higher the scale. The universe is curled up into tubes with circumference L ∼ m−1 . The strings can either be open or closed (xµ (τ. A possible solution started to emerge 1984 when string theory suggested as a viable explanation to what may be going on. Gravity has only been studied down to millimeter scales. if the n extra dimensions are of size R. Pl Recently it was realized that we haven’t really tested gravity at small distances. not to be confused with normal oﬀshell virtual particles). solid state physics where we can describe the same phenomena either by atomic vibrations or by phonons). typeIIA.
Contrary to standard super string theory. The most exciting recent developments have been in the contact with cosmology but even this relation goes back quite some time. They have to be weakly interacting to get out the core. But if the extra dimensions are large this means that the true Planck scale is much smaller n+2 mPl ∼ R− n+2 mPl4 n 2 In fact. Not only would this cure the hierarchy problem. it also means that it could be accessible at the LHC. espescially since the new developments often involve a large technical background.potential would look like V (r) ∼ m 1 1 mn+2 Rn r Pl m 1 m2 r Pl4 But this is the distances we are used to. An example here is that the Hubble constant which was only know to within a factor of two is now quite well known. Astrophysics constraints on Particle Physics These consists of knowing the physics between a variety of astronomical objects and then using these to obtain limits on extra eﬀects not included normally. Particle Physics and the Universe This is a rather active research area right now and we have to change numbers rather often. These typically appear in the radiation of particles in stellar (or their successors) which can have sizable eﬀects. This provides limits on axions and all light weakly interacting particles. for suitable values of n and R the Planck scale may be almost as small as the weak scale. where we have measured the potential to be V (r) ∼ where we have found mPl4 ≈ 1019 GeV. this means that we could enter into the realm of quantum gravity at the LHC (with the production of mini black holes and lots of other fun stuﬀ) — truly an exciting prospect. there exists in fact a whole ﬁeld called astroparticle physics for which by now many books and reviews have appeared. The best limits tend to come from red giants since these have the highest temperatures in the core. Here are some • In stars the thermonuclear process is rather well understood. 40 . There are many areas of contact. If there were more weakly interacting particles that would get radiated at a reasonable rate they would carry of heat from the thermonuclear fusion engine running in the core of the star. So keeping these notes up to date is always a bit of a problem.
The dark matter problem It is known that there exists a lot more matter in the universe than there is in stars and other normal objects. But they do not annihilate fully. This was ﬁrst discovered in the velocity curves of distant galaxies. some of them are left and the universe now exists mainly of quarks and leptons. these come in a class known as WIMPs (Weakly Interacting Massive Particles) There are basicallythree types. • When a star ﬁnally collapses and produces a neutron star. When the temperature drops below that required to have creation reactions running the annihilation of quarks and leptons starts. Nuclear Formation in the Big Bang In the early universe the neutrons were caught into (mainly) 4 He nuclei. It also allows limits on the numbers of relativistically moving particles (usually called radiation) in general at the time of nuclear synthesis. This has allowed to put some limits on the numbers of light neutrinos before limits from LEP appeared. not of antiquarks and antileptons. In the beginning. hot dark matter (neutrinos). • Massive neutrinos • Axions • Various supersymmetric particles.• A celebrated example is the fact that neutrino oscillations were really ﬁrst discovered from the solar neutrinos. pick your favourite energy. Detection of these and their arrival times have allowed many limits on neutrino properties. There have to be baryon number violating processes 41 . Particle physics has many candidates for this in beyond the standard model scenarios. as well as leptons. This tiny left over is rather puzzling and in order to solve it many conditions need to be satisﬁed (CPT conservation assumed here): 1. The precise amount of neutrons that get caught in various light nuclei versus the total number of protons and neutrons available is very sensitive to many conditions around. cold dark matter (WIMPs) and the ones where various ﬁelds get expectation values (which may vary with time) and as such produce the dark matter content. Baryon Number Generation One of the mysteries is why is there any baryon number left at all. very many neutrinos are radiated. there are very many quarks and antiquarks around.
That means that the energymomentum tensor can be written in the simple form ρ 0 0 0 0 p 0 0 (1) T µν = 0 0 p 0 0 0 0 p 42 . Inﬂation There are many problems associated with the fact that the universe is very big and looks rather homogenuous and isostropic on very large scales.2. Note that the absolute simplest version of inﬂation does not work and no simple nice working model is known but many general features are well supported. We expect several cases of spontaneous symmetry breaking to have occurred in the history of the universe. But still when we see those they look remarkably alike. How could this be? • Monopoles. 3. These problems are: • Flatness problem: this arises because the universe now is rather ﬂat but the evolution from general relativity is such that the universe moves away from ﬂatness with time. All of these are solved (at least in principle) by inﬂation. If the ﬁrst breaking happens in diﬀerent directions in diﬀerent areas of the universe they start relaxing together but then socalled topological defects (similar to the domains in a ferromagnet) can occur and we should have observed these if they existed. otherwise the equal balance principle would have baryon and antibaryon number equal. Cosmic Strings and Domain Walls: a problem: similar to the previous one. A really short Big Bang primer from General Relativity We only discuss here the simple FriedmanLemaˆ universes. In fact the standard model has all these three properties but not in a way suﬃcient to create the observed asymmetry (but it makes it worse for other higher energy mechanisms to produce the asymmetry). as well as being very ﬂat. Assume the universe is hoıtre mogeneous and isostropic. CP violation needs to be present. So in order to be as ﬂat as observed now it had to be extraordinarily ﬂat at early times. Grand Uniﬁed Theories provide one option for producing this imbalance. • Horizon problem: Many of the areas of the universe can not have exchanged signals with each other because of the expansion of the universe. Some things need to go out of thermal equilibrium.
and ρ(t) and p(t) are only functions of time. open. ﬂat geometry k = −1 Inﬁnite in space. The metric can be written in the general form ds2 = dt2 − R(t)2 dr2 + r2 dθ2 + sin2 θdφ2 1 − kr2 . ρ ∼ R−4 • w = −1 Or a cosmological constant. The matter is supposed to be at rest and behave like a socalled perfect ﬂuid. Usually it is assumed that p(t) = wρ(t) . From Einstein’s equations we then get H(t) ≡ The pressure p satisﬁes d ρR3 = −pd R3 . Note that in general relativity energy does not need to be conserved. φ + 3H φ = − ∂φ For k = 0 these have the behaviour 43 (6) (5) 2 ˙ R(t) R(t) 2 = k 8πGN ρ(t) − . hyperbolic geometry. closed. open. 0 leading to the following general geometry of the universe k = +1 Finite in space universe. ρ is constant. known as a matter dominated universe. The later cases are normally realized as being produced by the values of a ﬁeld φ which then obeys the equation of motion ∂V (φ) ˙ ¨ . (2) The parameter k can take the values ±1. ρ ∼ R−3 • w = 1 corresponding to particles moving at the speed of light whose energies is dominated by the kinetic energy and thus gets redshifted.. spherical geometry k = 0 Inﬁnite in space. Notice that I have not said anything about whether the universe has a ﬁnite lifetime. known as the equation of state. The following cases are of interest • w = 0 corresponding to particles at rest. (4) This has to be supplemented with an equation coupling ρ and p. • −1 ≤ w ≤ 0 known as quintessence. 3 R(t)2 (3) . known as a radiation dominated universe. The latter is in general complicated but for most purposes some simpliﬁcations are used.
During the inﬂation any curvature present would also have gotten stretched away so that the universe afterwards will be ﬂat with an incredible precision. This is known as the dark energy. the successor of the COBE satellite. the results of which led to the Nobel prize in Physics 2006. 2. 1. This is known as inﬂation. 5. A short history of the universe can now be described as. Since the entire universe came from a very small preinﬂation area it had ample time to get homogenuous and organize itself. First its existence by itself was a major success for the Big Bang model and was in fact what really made it universally accepted. Basically all in the same way. 4. Finally all relativistic particles get redshifted to such low energies that their rest mass dominates and we have the present (almost) matter dominated universe. Now the ﬁne inhomogeneities are of major interest see e. A newer observation is that we actually again have something like a (now much smaller one) cosmological constant around. The large inﬂation here means that the entire observable universe was reall a very small part beforehand. [2]. Structure Formation This is the question how the present structures seen in the universe at scales from galaxies and up are formed from the initial small perturbation caused by quantum eﬀects and/or 44 . The the energies of the φ ﬁeld becomes low enough that the eﬀect of the V (φ) becomes important and we shift to an exponentially expanding universe. The φ has slowly reached the minimum of the potential such that inﬂation stops and the universe is again radiation dominated.g. 3.matter radiation cosmological constant ρ(t) = ρ0 /R(t)3 ρ(t) = ρ1 /R(t)4 ρ = ρ2 R(t) = R(t) = √ 6πGN ρ0 t 2/3 1/2 32πGN ρ1 /3 t √ R(t) = e 8πGN ρ2 /3 t So a possibility of exponential expansion exists if at some point we have a constant density. The universe expands as a radiation dominated one and the particles slowly redshift. In 2003 there was a major step forward with the results from the WMAP satellite. How does this solve the various problems. Cosmic Microwave Background The cosmic microwave background has many interesting properties. For very small times all particles are relativistic and they dominate by far the energy density.
org/abs/hepph/0210370 45 . Kamionkowski. quantum gravity and gravitational radiation. . Cosmology and Particle Astrophysics o [2] M. Goobar. the gigantic bubble.. http://arXiv. This is done by galaxy surveys and complements the structures seen in the microwave background by covering diﬀerent time scales and size scales. .whatever in very early universe. ” Other There are many more speculative connections involving strings. . References [1] L. . extra dimensions gravity and gauge interactions operating in a diﬀerent number of dimensions. .. Here come names like “the great wall. Bergstr¨m and A.
This action might not be possible to undo. Are you sure you want to continue?