Lecture Notes

PHAS0040: Nuclear and Particle Physics
Prof. Emily Nurse, Prof. Andreas Korn, Prof. Matthew Wing

University College London
2023
Contents
1 Basic Ideas of Particle Physics 5

1.1 Fundamental Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.1 The structure of the atom, a brief history . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.2 Neutrinos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.3 The muon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.1.4 Antimatter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.1.5 The Standard Model content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 Units and scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2.1 Natural units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 Particle Physics is High Energy Physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.4 Particle interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.4.1 Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.4.2 Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4.3 Feynman Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.5 Four-vectors and Invariants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.5.1 Centre-of-mass frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.6 Range of forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.6.1 Virtual Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.7 Yukawa Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.8 The scattering amplitude . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.9 Cross sections and luminosities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.9.1 Differential cross sections: what we measure . . . . . . . . . . . . . . . . . . . . . . . . 19
1.9.2 Differential cross sections: predicting the rate . . . . . . . . . . . . . . . . . . . . . . . 19
1.10 Unstable Particles and the Breit-Wigner formula . . . . . . . . . . . . . . . . . . . . . . . . . 21
2 Leptons & Quarks 23

2.1 Leptons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1.1 Lepton Number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1.2 Lepton decays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.3 Neutrino Oscillations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.4 Absolute Neutrino mass scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.1.5 Number of Neutrinos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.1.6 Charged Lepton Number Violation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.2 Quarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1
PHAS0040: Nuclear and Particle Physics 2019
2.2.1 Hadrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.2.2 Hadron Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3 Experimental methods 35
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2 Accelerators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2.1 Centre-of-mass energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2.2 Luminosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.3 Interaction of Particles with matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.3.1 Ionisation Energy loss: the Bethe-Bloch Formula . . . . . . . . . . . . . . . . . . . . . 37
3.3.2 Interactions of Photons and Electrons . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.3.3 Interactions of Hadrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.4 Particle Detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.4.1 Position measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.4.2 Tracking Detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.4.3 Calorimeters: Energy Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.4.4 Particle Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.4.5 Multi-purpose Detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.4.6 HEP Experiments at UCL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4 The Strong Interaction 50

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.2 QCD versus QED . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.3 Hadrons pre-1960s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.4 Deep Inelastic Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.4.1 Evidence for quarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.4.2 Proton momentum fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.4.3 Parton Distribution Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.5 Colour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.5.1 Colour saves the Pauli principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.5.2 Colour charges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.5.3 Colour singlets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.5.4 Gluons and colour flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.5.5 Evidence for NC = 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.6 Asymptotic freedom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.7 Colour confinement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.8 Jets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.8.1 Quark jets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.8.2 Gluon jets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.9 QCD today . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5 The Electroweak Interaction: Part I 64

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.2 Symmetries and conservation laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.2.1 Parity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.2.2 Charge conjugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2
5.2.3 The spin structure of weak interactions . . . . . . . . . . . . . . . . . . . . . . . . . . 67
6 Electroweak Interactions: Part II: Quark Mixing 70

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
6.2 Quark Mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
6.2.1 Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
6.2.2 The Cabbibo Angle and GIM mechanism . . . . . . . . . . . . . . . . . . . . . . . . . 70
6.2.3 The CKM mixing matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
6.2.4 Kaon states and CP violation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
7 The Electroweak Interaction: Part III 77

7.1 The unification of electromagnetic and weak interactions . . . . . . . . . . . . . . . . . . . . . 77
7.1.1 Problems with the theory of weak interactions . . . . . . . . . . . . . . . . . . . . . . 77
7.1.2 The unified electroweak theory (very briefly!) . . . . . . . . . . . . . . . . . . . . . . . 78
7.1.3 The low energy approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
7.1.4 The discovery of neutral currents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
7.1.5 The discovery of the W and Z bosons . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
7.2 The Higgs boson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
7.2.1 Gauge invariance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
7.2.2 The Higgs mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
7.2.3 Discovery of the Higgs boson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
8 Nuclear Phenomenology 88
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
8.2 Shape of the Nucleus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
8.3 Nuclear Mass and Binding energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.4 Semi Empirical Mass Formula (SEMF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.5 Nuclear Stability and Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
8.6 β-Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
8.6.1 β − -Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
8.6.2 β + -Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
8.6.3 Electron Capture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
8.7 α-Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
8.8 γ-Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
9 Nuclear Structure 96
9.1 Fermi Gas Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
9.1.1 Fermi momentum and energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
9.1.2 Origin of the asymmetry term . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
9.2 Shell Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
9.2.1 Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
9.2.2 First Models: Nuclear Potentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
9.2.3 Spin Orbit Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
9.2.4 Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
9.2.5 Pairing Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
9.2.6 Predictions: Spins in the Shell Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
3
9.2.7 Predictions: Parity in the Shell Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

9.2.8 Predictions: Magnetic Moments in the Shell Model . . . . . . . . . . . . . . . . . . . . 104
9.2.9 Excited States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
9.3 Collective Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
9.4 Theory of Nuclear β-Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
9.4.1 Screening Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
9.4.2 Kurie Plot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
10 Fission and Fusion 108

10.1 Fission and Fusion: Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
10.2 Fission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
10.2.1 Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
10.2.2 Spontaneous and Induced Fission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
10.2.3 Nucleus–neutron cross sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
10.2.4 Fission chain reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
10.2.5 Nuclear fission bombs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
10.2.6 Nuclear fission power reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
10.3 Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
10.3.1 Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
10.3.2 Coulomb barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
10.3.3 Stellar fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
10.3.4 Fusion reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
11 Literature Guide 117

11.1 Basic Ideas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
11.2 Leptons, Quarks & Hadrons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
11.3 Experimental Methods I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
11.4 Experimental Methods II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
11.5 Strong Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
11.6 Electroweak Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
11.6.1 Electroweak Interactions II Quark Mixing . . . . . . . . . . . . . . . . . . . . . . . . . 118
11.7 Nuclear Phenomenology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
11.8 Structure of Nuclei . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
11.8.1 Shell Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
11.8.2 Beta decay revisited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
11.9 Fission & Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
4
Chapter 1
Basic Ideas of Particle Physics
Our current understanding of the building blocks of nature is described by a theory known as the Standard
Model of particle physics. It describes the fundamental particles and the interactions between them using
relativistic quantum mechanics. In this course we will discuss some of the fundamentals of the theory,
highlighting the historical developments such as the discoveries of the particles it describes. This module of
the course introduces some of the basic principles and tools in particle physics, many of which will be useful
later in the course.
1.1 Fundamental Particles

1.1.1 The structure of the atom, a brief history
Throughout history our understanding of the building blocks of nature has improved as technology and
insight has allowed us to search deeper and deeper. In the 19th century it was believed that atoms were
the fundamental building blocks. However, the existence of close to 100 elements with periodically recurring
properties was an indication of an underlying structure. In 1897 British physicist JJ Thomson discovered
the electron. He proposed that the mysterious glows observed when a high voltage was applied in a gas
volume (cathode rays) were fundamental particles that were inside atoms: he was right, he had discovered
the electron. He then proposed the plum pudding model, where the negatively charged electrons were like
raisins stuck in a positively charged pudding. It was not until 1911 when his former PhD student, Ernest
Rutherford, made a surprise discovery (by firing alpha particles at gold atoms, and realising that occasionally
some of them fired straight back) that atoms in fact consisted of a small and dense positive nucleus with
orbiting electrons. His work also led to the discovery of protons. Neutrons were discovered in 1932 by James
Chadwick, working under Rutherford, and it was realised that the atomic nucleus consists of protons and
neutrons. It was not until the 1960s that it was realised that protons and neutrons each consist of constitute
quarks with fractional charge. As far as we know today (although this may change in the future) quarks and
electrons are the most fundamental particles that make up the matter we see around us.
1.1.2 Neutrinos
The final fundamental particle involved in nuclear interactions is the electron–neutrino. This particle will be
studied in detail later in the course, for now we just note that it was postulated in 1930 by Wolfgang Pauli in
5
order to conserve momentum and energy in β–decay. It is an extremely light and weakly interacting particle
(with zero electric charge), which was not discovered until 1956.
1.1.3 The muon

But this is not the whole story. In 1936 another exciting discovery was made. A new heavier partner to the
electron was detected in cosmic ray showers. It was just like the electron but it penetrated further through
matter. The muon had been discovered. A particle just like the electron but with a mass 207 times as large
(this is why is interacts less with matter – more on this in “Experimental Methods”). This discovery was a
complete surprise and was the start of a whole new chapter in the discovery of new heavier unstable particles,
which we will summarised in Section 1.1.5.
1.1.4 Antimatter
It is not only matter particles that exist in our Universe. In 1932, Anderson discovered a particle from cosmic
rays in his cloud chamber experiment that behaved just like an electron, but bent in the opposite direction
in a magnetic field. The sign of the electric charge can be determined since the direction of curvature of
a charged particle moving through a magnetic field is dependent on the sign of the electric charge. This
was evidence for a new particle known as the positron, which is the antiparticle partner of the electron.
The existence of antiparticles had actually already been predicted as they were needed to interpret negative
energy solutions for particles. Particles are described by relativistic quantum mechanics. The energy of a
relativistic particle is given by the relativistic energy (E), momentum (p) and mass (m) relation equation:
E 2 = p2 c2 + m2 c4 , (1.1)
p
which has solutions: E = ± p2 c2 + m2 c4 . We can easily see that one of the solutions gives us a negative
energy. In quantum mechanics a complete set of basis states is required, and we cannot simply disregard
these negative energy solutions: we must find a physical interpretation for them. It turns out that we can
interpret these negative energy particle solutions as positive energy particles with opposite electrical charge
moving backwards in time (the maths is the same so this interpretation works). For each particle there
exists an antiparticle with equal mass and spin but opposite electrical charge and other quantum numbers
(e.g. lepton number, which will be covered later in the course). There are some particles (the photon and Z
boson) that are their own antiparticle. It is not yet known if neutrinos are their own antiparticles, and this
is an active area of research.
1.1.5 The Standard Model content

The particle content of the Standard Model is very briefly introduced in this section, with many more details
about the particles and their interactions given throughout the rest of this course.
The fermions (matter particles): The fundamental matter particles as we know them today are shown
in the top table in Figure 1.1. They are all fermions (they have a half integer spin). They can be sub-
divided into quarks and leptons, and they are arranged into three generations (or families) with increasing
mass. The particles that are the building blocks of atoms belong to the 1st (lowest mass) generation. Their
electrical charges are also shown in Figure 1.1. The lower table in the Figure shows the antiparticles, which
have opposite electrical charges (the other properties like mass and spin are the same as the corresponding
particle).
6
The force carrying bosons: The interactions between the fermions is governed by the exchange of another
type of particle known as bosons. These have integer spin. All of the particles interact via the weak
interaction (via the exchange of the massive W + , W − and Z bosons). All but the neutrinos interact via the
electromagnetic (EM) interaction (via the exchange of massless photons). The quarks also interact via the
strong interaction (via the exchange of massless gluons).
The Higgs boson: The Higgs particle is a newly discovered neutral boson with zero spin, that interacts
with all particles with mass, with a strength of interaction dependent on the mass.
Figure 1.1: The Standard Model matter particles and their antiparticles
The Standard Model bosons are summarised in Table 1.1.
Boson Electrical charge spin

gluon 0 1
photon 0 1
W+ +1 1
W− −1 1
Z 0 1
Higgs 0 0
Table 1.1: The Standard Model bosons and their electrical charge and spin
7
There is a huge variation in the masses of the fundamental particles, and the reason for this is not known.
The masses of the particles are shown in Table 1.2. The mass units will be introduced in Section 1.2, but
the point here is to highlight the huge variation. It is the lightest particles that are stable, as we shall see in
Section 1.4.1. The heaviest particle, the top quark, is as massive as a gold atom!
Fermion Mass [MeV/c2 ]

Neutrinos < 0.0001
Electron 0.5
Boson Mass [MeV/c2 ]
up-quark 3
photon 0
down-quark 5
gluon 0
strange-quark 100
W boson 80385
muon 106
Z boson 91188
charm-quark 1300
Higgs boson 125000
tau 1780
bottom-quark 4500
top-quark 173000
Table 1.2: The masses of the fundamental particles
1.2 Units and scales

In particle physics we deal with very small distance scales and masses. A useful and common distance unit
is therefore the Fermi: 1 fm = 10−15 m. For example, the radius of the proton is 0.8 fm and nuclei have radii
in the range 0.8 to 10 fm. Areas are given in terms of the Barn: 1 b = 10−28 m2 . Interaction cross sections,
which give the likelihood of the interaction occurring and will be discussed in Section 1.9, are often given in
terms of millibarns (1 mb = 10−3 b), microbarns (1 µb = 10−6 b), nanobarns (1 nb = 10−9 b), picobarns (1
pb = 10−12 b) or femtobarns (1 fb = 10−15 b).
1.2.1 Natural units

Natural units are often used in place of SI units. Natural units are based on the fundamental constants of
quantum mechanics and special relativity: GeV, c and h̄, where:
• 1 GeV = 109 eV = 109 × 1.6 × 10−19 J= 1.6 × 10−10 J (here 1 electronvolt, eV, is the energy gained
by the charge of a single electron moved across an electric potential difference of one volt).
• c = 3 × 108 ms−1 is the speed of light in a vacuum.
• h̄ = 1.055 × 10−34 Js is the unit of action in quantum mechanics.
In order to simplify equations, we often choose to set h̄ = c = 1. To get back to SI units the factors of h̄ and
c must be reinserted.
Energy, Momentum and Mass

Energy: The Large Hadron Collider (LHC) has an energy of 13 TeV (13 × 103 GeV). In order to convert
this into Joules we just multiply by the conversion factor 1.6 × 10−10 J/GeV to give 2.1 × 10−6 J.
8
Momentum: Recall that for a massless particle E = pc. Therefore a massless particle with E = 6.5 TeV
(the energy of one of the LHC beams) has a momentum p = 6.5 TeV/c = 6.5 TeV, if we set c = 1. To
convert this into SI units for momentum [kg ms−1 ] we reinsert the factor of c and use the conversion factor
from GeV to Joules:
6.5 × 103 × 1.6 × 10−10 J
= 3.5 × 10−15 kg ms−1 .
3 × 108 ms−1
Mass: Recall that E = mc2 . The electron mass is me = 0.511 MeV/c2 = 0.511 × 10−3 GeV/c2 . To convert
this into SI units for mass [kg] we reinsert the factor of c2 and use the conversion factor from GeV to Joules:
0.511 × 10−3 × 1.6 × 10−10 J

2 = 9.1 × 10−31 kg.
(3 × 108 ms−1 )
Length
The conversion from SI units to natural units is found by a dimensional analysis. The natural units for length
(metres in SI units) are h̄c GeV−1 . This can be seen by substituting in the SI units for h̄ [Js], c [ms−1 ] and
GeV [J], then: Js · ms−1 · J−1 = m.
The radius of the proton is 4.1 GeV−1 in natural units where h̄ = c = 1. To convert this into SI units we
first reinsert the factors of h̄ and c:
4.1 × 1.055 × 10−34 × 3 × 108 m
4.1h̄c GeV−1 = = 4.1 × 0.197 × 10−15 m = 0.8 fm.
1.6 × 10−10
A very useful conversion factor to remember is h̄c = 0.197 GeV fm.
Time
Again, the conversion from SI units to natural units is found by a dimensional analysis. The SI unit of time
is seconds (s), so we need to combine factors of h̄ [Js], c [ms−1 ] and GeV [J] to get s. This is possible with
h̄GeV−1 = [Js · J−1 ] = s.
1.3 Particle Physics is High Energy Physics

As you will see later in the course, when studying particle physics we need high-energy accelerators or cosmic
rays. There are two reasons for this:
1. Recall from your quantum mechanics courses that the de Broglie equation relates the wavelength of a
particle to its momentum via: λ = hp . If we want to resolve the smallest distances associated with the
fundamental constituents of matter we need the highest possible momenta and hence energies. As an
example consider the HERA electron–proton collider, which was built to probe deep into the structure
of the proton. The particles collide with a momenta of 150 GeV/c in the centre-of-mass frame (more
on the centre-of-mass frame in Section 1.5.1). We can therefore work out the distance scales probed at
this collider by calculating the associated wavelength.
h 2πh̄ 2πh̄ 2πh̄c 2π0.197 fm
λ= = = = = = 0.008 fm,
p p 150 GeV/c 150 GeV 150
9
which is one hundred times smaller than the radius of the proton. Note that the actual distance scales
probed vary from interaction to interaction as it depends on the kinematics of the exchanged particle.
2. Recall from your relativity course Einstein’s famous equation: E = mc2 . The factor of c2 is extremely
large, and therefore creating massive particles (such as the top quark or the Higgs boson) requires large
energies.
1.4 Particle interactions

1.4.1 Decay
If a particle can decay to other particles then it will. Therefore, most particles in the Standard Model are
unstable. There are often different allowed decay modes and the rate of each decay mode depends on the
strength of the interaction. The strong force is the strongest and therefore the decays will be fastest. The
weak force is the weakest and therefore decays will be slowest. There are a few stable particles, that cannot
decay due to conservation laws, e.g. the conservation of energy, momentum and electric charge. There are
other conservation laws that will be discussed throughout this course.1
The stable particles are the electron, the positron, the lightest neutrino and the photon. Quarks and
gluons are never found in isolation, but are bound into states known as hadrons (more on this in the module
on “Strong interactions” later in the course). As far as we know, the proton is the only stable hadron.
The following conservation rules ensure that these particles are stable (and limit the possible decay
channels of the other particles):
• Conservation of energy and momentum ensure that a particle must always decay to two or more particles
and the mass of the decaying particle must exceed the sum of the masses of the particles produced in
the decay. This is simple to see by considering the decay of a particle in its rest frame. The total energy
before the decay is the rest-mass energy of the particle and the total momentum is zero. In order to
conserve energy the sum of the energies of the produced particles must be equal to the rest mass of
the decaying particle. In order to conserve momentum, the sum of the particle momenta must be zero.
There must be more than one particle in the decay so that their momenta can cancel to zero. If only
one particle were produced it would have to be at rest to conserve energy. But in this case it must
have exactly the same mass as the decaying particle. If this process is possible then the reverse is also
possible, and the particles could each decay into each other and they would keep switching between
each other. This is not a decay at all but a mixing of two particles. If the mass of a single produced
particle is less than the decaying particle it would need to be moving so that energy can be conserved.
However it is not possible to conserve momentum in this case. If the sum of the masses of the produced
particles is exactly the mass of the decaying particle, they would be produced at rest. Otherwise the
sum of the masses must be less than the mass of the decaying particle and some energy in the final
state comes from the kinetic energy.
• Conservation of electric charge means that the sum of electric charges before an interaction must equal
the sum after an interaction. Therefore the decay products of a charged particle must include at least
one charged particle. Since the electron is the lightest charged particle it is stable.
1 These conservation laws are due to properties of the world. For example, the fact that the laws of physics do not change
with time leads to the law of the conservation of energy (thanks to the theorem of Emmy Noether).
10
There are other conservation laws to be discussed later in this course that explain the stability of the proton
and lightest neutrino and the allowed decay channels of all other particle decays.
1.4.2 Scattering
Particles interact with each other either via elastic scattering (where the initial and final state particles are
the same) or via inelastic scattering (where the initial and final state particles are different). Which processes
are allowed is governed by the conservation laws introduced in Section 1.4.1 (some of which are discussed later
in the course). An example of elastic scattering is e+ + e− → e+ + e− . An example of inelastic scattering
is e+ + e− → µ+ + µ− . Particles interact by the exchange of other particles, which act as intermediate
particles in the interaction. In elastic scattering, momentum is transferred from one particle to another by
the exchange of a boson. In inelastic scattering as well as the momentum transfer, the particle types change.
1.4.3 Feynman Diagrams

Particle decay and scattering interactions can be represented by Feynman diagrams. They are extremely
useful tools that help us to visualise and aid in the calculation of the cross section of any process. Figure 1.2(a)
shows an example of a simple Feynman diagram describing the elastic scattering of two electrons. The
convention in this course is to have time running from left-to-right (some authors use bottom-to-top), so in
this diagram there are two incoming electrons. The bottom electron emits a photon which is subsequently
absorbed by the top electron. This exchange of a photon causes the momenta of the electrons to change.
A very similar process is shown in Figure 1.2(b), where the photon is emitted from the top electron and
absorbed by the bottom one. The Feynman diagram in Figure 1.2(c) depicts the sum of the other two
figures. It is convention to draw Feynman diagrams like Figure 1.2(c) to represent both time-orderings of
the process. In these diagrams the incoming and outgoing particles are real particles, and the intermediate
particles are virtual particles (more on virtual particles later). Feynman diagrams in electromagnetism all
e− e− e− e− e− e−
γ γ γ
e− (a)
e− e− (b)
e− e− (c)
e−
Figure 1.2: Feynman diagram for e− + e− elastic scattering. (a) and (b) show different time orderings and
(c) depicts the sum of (a) and (b): (c) = (a) + (b).
involve photons and electrically charged particles. Concentrating on electrons and positrons for now:
• Figure 1.3 shows the elastic scattering e+ +e+ → e+ +e+ . Note the direction of arrows has changed with
respect to Figure 1.2(c). This is because the arrow directions do not indicate the direction of motion,
11
but rather they denote particles flowing in the direction of time and antiparticles flowing opposite to the
direction of time (see Section 1.1.4 where the idea of antiparticles being particles travelling backwards
in time was introduced).
• Figure 1.4 shows the two Feynman diagrams for the elastic scattering e+ + e− → e+ + e− . There is now
an additional diagram that would not have been allowed for the processes with two same-sign charged
particles in the initial and final states, because charge must be conserved at each vertex.
• Figure 1.5 shows the Feynman diagram for the inelastic scattering process e+ + e− → γγ.
e+ e+
e+ e+
Figure 1.3: Feynman diagram for e+ + e+ elastic scattering.
e+ e+
e− e−
γ
γ
e− (a)
e− e+ (b) e+
Figure 1.4: Feynman diagrams for e+ e− elastic scattering.
Each vertex (point where the particle lines come together) in the above diagrams represent the coupling
of a photon to the charge of the electron/positron, and the coupling is proportional to e. The probability
of the process occurring is proportional to e2n where n is the number of vertices in the diagram. In all the
diagrams in the above figures, n = 2. These are all examples of the lowest-order diagram for a particular
process. For each diagram, it is also possible for the same process to occur in much more complicated ways.
As an example Figure 1.6 shows the e− + e− → e− + e− process with additional intermediate particles. In
this diagram you can see there are four vertices, so the probability is proportional to e8 .
Inspecting a Feynman diagram gives a good indication on whether the process is allowed or forbidden as
each vertex must obey conservation laws.
12
e− γ
e+ γ
Figure 1.5: Feynman diagram for the inelastic scattering process e+ + e− → γγ.
− −
e e
e−
e−
e− e−
Figure 1.6: A higher-order Feynman diagram for e− e− elastic scattering.
1.5 Four-vectors and Invariants

An extremely useful concept in particle physics is that of four-vectors. Recall from your special relativity
lectures the time-space four-vector: (ct, ⃗x) = (ct, x, y, z).
In particle physics we make extensive use of the energy-momentum four-vector (known as the four-
momentum): P = (E, c⃗ p) = (E, cpx , cpy , cpz ). As you will have learnt in your relativity courses, under a
change of reference frame these four-vectors undergo Lorentz transformations (see the lecture slides for the
matrix form of the Lorentz transformations).
It turns out that the inner product of a pair of four-vectors is invariant under a change of reference frame
(just like the inner product of two spatial vectors is invariant under coordinate system rotations). The inner
product of a four-vector A is defined as:
⃗·A
A2 = A · A = A0 A0 − A ⃗
For an energy-momentum four-vector we have:

P 2 = P · P = E 2 − p⃗2 c2 .
You can convince yourselves that this inner product will not change under a Lorentz transformation, by
substituting the explicit form of the Lorentz transformation into the above. Since this is true in any frame,
13
it can be conveniently evaluated in the rest frame where p⃗ = 0. This really simplifies things as
P 2 = E2
and we know that E 2 = m2 c4 when a particle is at rest, therefore:
P 2 = E 2 − p⃗2 c2 = m2 c4
is true in any frame, which is just the Einstein energy-momentum relationship. This is encouraging, it tells
us that the rest mass of a particle is independent of the frame it is in.
We can define the invariant mass, W = P/c2 . For a system of N particles, the total four-momenta is the
sum of the individual four-momenta, so we have:
N
!2 N
!2
X X
2 4
W c = Ei − c p⃗i , (1.2)
i i
which is an invariant quantity (i.e. it is the same in any frame of reference). It is also a conserved quantity.
For one particle the invariant mass is the same as the rest mass. A very important use of the invariant mass
is in determining the mass of a short-lived particle. If a particle decays to a set of N particles its mass can
be determined from the invariant mass of the decay products, since it is conserved.
1.5.1 Centre-of-mass frame

P
For a system of particles the centre-of-mass (or centre-of-momentum) frame is the frame in which i p⃗i is
zero. In a collider experiment the energy in the centre-of-mass frame is a very useful quantity, as it gives
the maximum energy available to create a massive particle: E = mc2 . In any other frame (including the
laboratory frame) some of the energy is tied up in the conserving the momentum. The square of the centre-
2
of-mass energy, ECM , is just equal to the invariant mass squared, in the centre-of-mass frame. Since the
invariant mass squared is frame invariant, it can be evaluated from beam energies given in the laboratory
(or any other) frame, and this will give us the centre-of-mass energy squared. Below are a couple of worked
examples for evaluating the centre-of-mass energy:
Example 1 Consider a simple case in which two beams are colliding with equal energy in opposite direc-
tions. The centre-of-mass frame is the same as the lab frame (i.e. the frame in which we observe the beams).
In order to determine the centre-of-mass energy we just evaluate the invariant mass in the centre-of-mass
frame (for convenience I am dropping factors of c), using Equation 1.2 we have:
2 2
W 2 = (2 × Ebeam ) − (0) = 4Ebeam
2
giving:
ECM = 2Ebeam
where Ebeam is the beam energies in the lab frame and ECM is the centre-of-mass energy.
14
Example 2 A slightly more complicated example is for a fixed-target experiment, where there is one moving
beam of particles colliding with a stationary target. The mass of the particles in the target and the beam
can usually be assumed to be much less than the energy of the beam. We evaluate the invariant mass in the
lab-frame:
2 2
W 2 = (Ebeam + mtarget ) − (⃗ 2
pbeam ) = Ebeam + m2target + 2Ebeam mtarget − p⃗2beam
2
substituting in Ebeam = p⃗2beam + m2beam we obtain:
W 2 = m2beam + m2target + 2Ebeam mtarget
we can neglect the first two terms in the limit that the beam energy is much larger than the masses of the
particles in the beams, giving:
W 2 = 2Ebeam mtarget
and therefore: √ p
ECM = W2 = 2Ebeam mtarget .
See Problem Sheet 1 for more examples of determining the centre-of-mass energy for different collider
experiments.
1.6 Range of forces

Consider two particles A and B, scattering by means of the exchange of another particle, X, as shown in
Figure 1.7. Consider the lower vertex in the rest frame of the incident particle A. The four-momentum before
B B
A A
Figure 1.7: Feynman diagram for particle scattering via the exchange of another particle.

the interaction is MA c2 , 0 and the four-momentum after the interaction is (EA , p⃗) + (EX , −⃗ p), where to
conserve momentum before and after the interaction we have set the momenta of the final-state particles
to be equal in magnitude
p p and opposite in sign. The energy after the interaction is therefore EA + EX =
MA2 c4 + p2 c2 + MX 2 c4 + p2 c2 , which is clearly larger than the energy before the interaction (M c2 ).
A
It appears that energy conservation has been violated! This is understood by considering the uncertainty
principle. The energy difference before and after the interaction can be borrowed for a time given by
h̄
Heisenberg’s uncertainty principle: ∆t ≈ ∆E . The difference in energy depends on p⃗, but is always ≥ MX c2 .
h̄
Therefore ∆t ≤ MX c2 . The maximum distance a particle can travel in this time is determined by the speed
15
of light to give: r ≤ ∆t × c ≤ Mh̄c Xc

2 . This is the maximum distance particle X can propagate before being
absorbed by particle B and therefore defines the maximum range, RX , of the force mediated by particle X.
For electromagnetism the massless photon leads to an infinite range. For the weak interaction, mediated by
the massive W and Z bosons the range is of order RW = Mh̄W c ≈ Mh̄Z c ≈ 0.002 fm.
As we will see later in the course, even though the gluon is massless, the strong force is relatively short
range (∼ 1 fm), but this is for different reasons that will be explained in the module on “Strong Interactions”.
1.6.1 Virtual Particles

As mentioned previously, these exchanged particles are known as virtual particles, and they only exist for
a short time. In contrast, particles that exist for long periods (those in the external lines of a Feynman
diagram) are known as real particles. These virtual particles do not satisfy the Einstein energy-momentum
relationship: E 2 − p⃗2 c2 = m2 c4 . We still define the four-momentum squared as P 2 = E 2 − p⃗2 c2 , but P 2 is
not equal to the mass of the particle. The particles are known to be “off mass-shell”. So e.g. photons can
have a non-zero P 2 and in fact P 2 can even be negative (see the module on “Strong Interactions”). Virtual
particles are often indicated with a asterisk, e.g. γ ∗ for a virtual photon.
1.7 Yukawa Potential

We have been considering forces to be mediated by particles (X in Figure 1.7). It is useful to relate this
back to the classical idea of a particle in a potential (e.g. in the case of the electromagnetic interaction:
an electrically charged particle being scattered by a static potential of which another electrically charged
particle is the source). The wavefunction of the force carrying particle is assumed to satisfy the Klein-
Gordan equation, which is a relativistic wave equation. First, as an aside, we derive the Klein-Gordan
equation.
Deriving the Klein-Gordan equation We make quantum mechanical substitutions to the relativistic
2 4
equation E 2 = p2 c2 + MX c . In order to do this we substitute the following operators acting
on a wavefunc-

∂ ∂ ∂ ∂
tion, ϕ (⃗x, t), for the energy and momentum terms: E → ih̄ ∂t and p⃗ → ih̄∇, where ∇ = ∂x , ∂y , ∂z is the
vector differential operator. This leads to:
∂2
−h̄2 ϕ (⃗x, t) = −h̄2 c2 ∇2 ϕ (⃗x, t) + MX
2 4
c ϕ (⃗x, t) . (1.3)
∂2t
Static solution to the Klein-Gordan equation The next step is to relate this back to the classical idea
of a potential from a source. To do this we interpret the wavefunction, ϕ (⃗x, t), as a potential coming from
the source that is particle A (from Figure 1.7). In some sense the potential can be related to the density of
force-carrying particles.
We consider the source to be independent of time. We can then consider static solutions (i.e. the spatial
part of the wavefunction) to give:
M 2 c2
∇2 ϕ (⃗x) = X2 ϕ (⃗x) . (1.4)
h̄
If we consider ϕ (⃗x) to be a spherically symmetric potential then we can work in spherical co-ordinates
∂2 2 ∂
and set ϕ (⃗x) → V (r). In this case ∇2 = ∂r 2 + r ∂r .
16
Considering electromagnetism, where X is a photon and MX = 0, we obtain ∇2 V (r) = 0. You can easily
show that a solution to this is given by V (r) = Cr , where C is a constant, by substituting in this form of
V (r):
∂2V ∂ −C 2 −C

2 ∂V 2C 2C
+ = + = 3 − 3 =0
∂r2 r ∂r ∂r r2 r r2 r r
2 2
e e 1
as required. For electromagnetism C = − 4πϵ 0
, so that V (r) = − 4πϵ 0 r
, where e is the electric charge of the
source and ϵ0 is the permittivity of free space. This is the familiar Coulomb form of a potential, and the
e2
strength of the interaction is given by the coupling: − 4πϵ 0
.
2 −r/R
g e
The general solution, for MX > 0, can be shown (see Problem Sheet 1) to be V (r) = 4π r , where
h̄
R = MX c is the range of the interaction, introduced in Section 1.6, and g is a coupling constant, at each vertex
of a Feynman diagram, which is analogous to e in electromagnetism. For MX = 0 we obtain e−r/R = e0 = 1
and we obtain the Coulomb form. For very large masses the potential becomes very small for large r and
only becomes Coulomb like again if r ≤ R. This is known as the Yukawa potential (after Hideki Yukawa who
introduced the idea of forces due to massive particle exchange in 1935).
We express couplings, and hence the strength of the interaction, in terms of dimensionless parameters, so
that the values do not depend on the system of units, which is obtained by inserting a factor of h̄c into the
denominator:
e2 1
α= ∼
4πϵ0 h̄c 137
and
g2
αX = .
4πh̄c
1.8 The scattering amplitude

The probability for a particle to be scattered by a potential V (r) can be calculated by first calculating the
scattering amplitude, M. Then the probability is proportional to |M|2 (in quantum mechanics probabilities
are always calculated from the square of the modulus of a probability amplitude). It turns out that for small
coupling constants, g 2 much smaller than 4πh̄c, the amplitude is given by:
Z
M = ei⃗pi ·⃗x/h̄ V (x) e−i⃗pf ·⃗x/h̄ d3 ⃗x, (1.5)
where ei⃗pi ·⃗x/h̄ is the initial-state wavefunction, ei⃗pf ·⃗x/h̄ is the final-state wavefunction and V (x) is the scat-
tering potential. If we write p⃗ = p⃗f − p⃗i and p⃗ · ⃗x = |⃗ p| r cos θ, you can do this integral over all space for the
Yukawa potential by substituting d3 ⃗x = r2 sin θdθdϕdr to obtain (see Problem Sheet 1):
−g 2 h̄2
M p⃗2 =

2 2 c2
. (1.6)
|⃗
p | + MX
If you do a full relativistic calculation (beyond the scope of this course) you arrive at a very similar result:
g 2 h̄2 c2
M q2 =

q 2 − MX 2 c4 , (1.7)
17
2 2
where q 2 is the four-momentum exchange: q 2 = (Ef − Ei ) − (⃗
pf c − p⃗i c) . In the limit that MX is very large
compared to the four-momentum transferred, this becomes:
g 2 h̄2
M∼− 2 c2 , (1.8)
MX
and the strength of the interaction is independent of the four-momentum transferred. In the limit that MX
is very small compared to the four-momentum transferred, this becomes:
g 2 h̄2 c2
M∼ , (1.9)
q2
such that the interaction is independent of MX . We will see in the module on “Electroweak Interactions”
that this means that the strength of the weak interaction is actually similar to that in electromagnetism at
very high four-momentum transfers.
1.9 Cross sections and luminosities

In the previous section we saw the form of the scattering amplitude. The probability of a process occurring
is then proportional to the square of the modulus of the amplitude. In this section we will show how the
amplitude is used to predict the rates of a certain process occurring. First we will introduce the idea of cross
sections and luminosities.
When calculating the interaction rate for a certain process (e.g. the interaction between two particles)
we must take into account the flux of the initial-state particles (number of particles passing each other per
unit area per unit time) as well as the likelihood that the process will occur. The rate of a certain process
occurring is given by R = σ × L, where σ is the cross section and L is the instantaneous luminosity.
The cross section characterises the probability for a certain process (e.g. the production of a Higgs at the
LHC). In accelerators it is independent of all beam parameters apart from the energy. All the fundamental
physics is contained in σ, which is an expression of the underlying quantum mechanical probability that an
interaction will occur. The instantaneous luminosity tells us how many incident particles there are per unit
area per unit time and is expressed in units of m−2 s−1 . The cross section has units of area (such that σ × L
gives us the particle rate) and is typically expressed in millibarns to femtobarns, as discussed
R in Section 1.2.
The integrated luminosity L is just the time integral of the instantaneous luminosity: Ldt, and is usually
given in units of inverse barns (usually inverse picobarns or inverse femtobarns). Accelerators will run for a
certain amountR of time, although not continuously, and the number of events of a certain type will be given
by: N = σ × Ldt = σ × L. As an example, the cross section to produce a pair of Z bosons at the LHC with
a centre-of-mass energy of 13 TeV is 16.7 pb. In 2015 the ATLAS detector at the LHC (more on this later
in the course) recorded 3.2 fb−1 of proton–proton interactions. This gives 16.7 × 103 fb ×3.2 fb−1 = 53,440
ZZ events produced in 2015.
Modern colliders employ bunched beams, where each beam contains a number of bunches of particles
that circulate in the beam in both directions. In this case L depends on the number of colliding bunches,
the number of particles in each beam, the cross sectional area of the beams and the frequency with which
the bunches circulate the ring (more on this in the “Experimental Methods” part of the course).
For the simpler example of a fixed target experiment, where a beam of particles collides with a stationary
target we have:
L = nb × vi × N, (1.10)
18
where nb is the number density of particles in the beam, vi is the velocity of the beam, and N is the number
of illuminated particles in the target. We will look at this again in a bit more detail in section 3.2.2
When two particles of a certain type interact, many different processes can occur, and the total cross
section is the sum of all possibilities. For example, when two protons interact at the LHC, it is possible (but
rare) that a Z boson is produced. It is also possible (but even rarer) that a Higgs boson is produced. The
most likely interaction is a scattering between two constituents of the protons via the exchange of a gluon.
The total proton–proton cross section is then the sum of all the cross sections for the different processes.
1.9.1 Differential cross sections: what we measure

In practice an experiment will often measure differential cross sections, which give the dependence of the
cross section on some kinematic variable. For example, the ATLAS detector at the LHC surrounds the beam
and detects particles that are created when the beams interact. The position of the final state particles will
be measured and the cross section depends on these positions. Similarly the cross section depends on the
energy and momentum of the produced particles, and on the combined kinematics of all the particles in the
final state. In a simple scattering where the initial state energies and momenta are known, with a 2-body
final state, the direction of only one particle needs to be defined as conservation of momentum and energy
will constrain the direction of the other particle. The direction of a particle is specified by a polar angle,
θ, and an azimuthal angle, ϕ, in a reference frame where the z-axis coincides with the beam direction. The
measured rate for the particle to be emitted into an element of solid angle dΩ = d (cos θ) dϕ is then given by:
dσ (θ, ϕ)
dR = L × dΩ. (1.11)
dΩ
If we want to obtain the total cross section from such a measurement we simply integrate over all solid angles:
2π +1
dσ (θ, ϕ)
Z Z
σ= dϕ d (cos θ) . (1.12)
0 −1 dΩ
1.9.2 Differential cross sections: predicting the rate

Note: this subsection goes into a fairly detailed derivation of the differential cross section for the scattering
of two particles. It is here for the interested reader, but is non-examinable. The idea of the cross section
depending on the density of states, which is discussed below, is however an important concept.
We now relate the experimentally defined cross sections to theoretical calculations of the probability of
the transition between an initial state i and a final state f . We shall consider a simple case where a single
particle beam is interacting with a single target particle. The whole system is confined to an arbitrary volume
V , which we define as a cube with three sides of length L (we will see that the final result does not depend
on the value of V ).
The rate is given by σ × L. In this simple example we can use Equation 1.10. The number of target
particles is one (N = 1) and we have one incident particle per unit volume (nb = V1 ), giving: L = vVi , where
vi is the velocity of the beam particle. We can therefore use Equation 1.11 and write the measured rate for
the particle to be emitted into an element of solid angle dΩ in the direction (θ, ϕ) as:
vi dσ (θ, ϕ)
dR = dΩ. (1.13)
V dΩ
19
We now switch gears and obtain the theoretical expression for dR. In Quantum Mechanics (QM) the
transition rate to go from an initial state i to a final state f is given by Fermi’s Second Golden Rule2 :
2
2π
Z
dR QM
= ψf∗ V (x)ψi d3 x ρ (Ef ) , (1.14)
h̄
where ψi is the initial-state wavefunction (describing the initial particles) and ψf is the final-state wavefunc-
tion (describing the final particles), V (x) is a potential (not to be confused with the volume V ) and ρ (Ef ) is
the density of states and will be specified below. The wavefunctions are given by ψi = Ai ei⃗pi ·⃗x/h̄ and ψf =
RLRLRL ∗
Af ei⃗pf ·⃗x/h̄ , where Ai,f are normalisation factors. In our example the integral 0 0 0 ψi,f ψi,f dxdydz = 1,
as the particles are confined to the box so the probability to find them somewhere inside the box is 1. This
gives A2i,f = L13 and Ai,f = √1V , for both ψf and ψi . We can therefore write:
2π 2
dRQM = |Mi→f | ρ (Ef ) , (1.15)
h̄V 2
where Mi→f is the amplitude for the transition i → f , which we defined in Equation 1.5.
Now lets turn back to ρ (Ef ), which we introduced as the density of states. What this means is that
ρ (Ef ) dEf is the number of possible quantum states of the final-state particles where the total energy of the
particles is between Ef and Ef + dEf . We will find this by first evaluating ρ (p), where p = |⃗ p|, and ρ (p) dp
is the number of possible quantum states of a particle where the momentum is between p and p + dp. We
will then change variables and convert back to ρ (Ef ) dEf .
Because our particles are confined to the box, they must satisfy periodic boundary conditions,3 which
means that ψ(x + L, y, z) = ψ(x, y, z), and similar for the y and z co-ordinates. This means an integer
number of wavelengths must fit into the box in each dimension. We can use this condition to determine
the allowed values for the momentum components px , py and pz . Considering first the px component we
have eipx x/h̄ = eipx (x+L)/h̄ , which implies eipx L/h̄ = 1 and therefore px = 2πh̄
L nx , where nx is an integer.
2πh̄ 2πh̄

Similarly py = L ny and pz = L nz , where nz and ny are integers. What this means is that there
L L 3 3

are 2πh̄ states per unit px (similarly for py and pz ) and therefore there are 2πh̄ d p⃗ momentum states
corresponding to momenta between p and p + dp. Using d3 p⃗ = p2 dpdΩ, we can write this as:
V 2
ρ (p) dp = 3p dpdΩ (1.16)
(2πh̄)
momentum states pointing into a solid angle dΩ with a momenta with magnitude between p and p + dp.
Ok, stay with me, we are almost there. The next step is to change variables to obtain ρ (Ef ) dEf . First lets
change variables from p to E for a single particle. We use:
dp
ρ (E) dE = ρ (p) dE, (1.17)
dE
dn
which is perhaps easier to explicitly see if we write the density of states as ρ (p) = dp (where n is the number
dn dn dp dp
of states) and similarly for ρ (E) such that dE dE = dp dE dE. We can evaluate dE :
2Aderivation of this is beyond the scope of this course.

3 see
Appendix A in Nuclear and Particle Physics, B. R. Martin or section 3 (p 60) in the Modern Particle Physics, M.
Thomson.
20
1 2 1 2 dE
(a) non-relativistically: Here E = 2 mv + mc2 and p = mv such that E = 2 p /m and dp = p/m = v.
dp
Therefore dE = 1/v. And
pc2
p dE 1
(b) relativistically: Here E = p2 c2 + m2 c4 and dp = 2 × 2pc2 × E1 = E . Using the relativistic expressions
dE dp
for energy and momentum: E = γmc2 and p = γmv, we obtain dp = v such that dE = 1/v again.
dp
It can be shown that dE = 1/v holds for a 2 → 2 scattering as long as p and v are interpreted as magnitudes
of the the relative momentum (pf ) and relative velocity (vf ) of the final-state particles and E is the total
energy of the final-state particles (Ef ). We therefore use Equation 1.16 and Equation 1.17 and obtain
V p2f
ρ (Ef ) = 3 dΩ. (1.18)
(2πh̄) vf
So, from Equation 1.15, we now have:
2π 2 V p2f
dRQM = |M i→f | 3 dΩ. (1.19)
h̄V 2 (2πh̄) vf
Equating this to dR from Equation 1.13, we get:
2π 2 V p2f vi dσ (θ, ϕ)
|Mi→f | 3 dΩ = Ω, (1.20)
h̄V 2 (2πh̄) fv V dΩ
which rearranges to give:

dσ (θ, ϕ) 1 p2f 2
= |Mi→f | . (1.21)
dΩ 4π 2 h̄4 vi vf
You can see that the terms for the volume, V have cancelled out and our final result does not depend on our
arbitrary volume.
This expression is valid for general two-body relativistic scattering processes:
A (⃗ pi ) → C (⃗
pi ) + B (−⃗ pf ) + D (−⃗
pf ) (1.22)
as long as spin is neglected.
1.10 Unstable Particles and the Breit-Wigner formula

Recall from Equation 1.7 the form of the scattering amplitude. It appears to have a divergence for q 2 = MX 2 2
c .
This apparent problem arises because this form of the propagator does not account for the particle being
2
unstable. The time dependence of a particle in its rest frame is given by ψ ∼ e−iMX c t/h̄ (since E =
2 −i(Et−c⃗p·⃗
x)/h̄
MX c in the rest frame and generally the wavefunction has the form ψ ∼ e ). However, for an
unstable particle we want the probability density to obey exponential decay such that ψψ ∗ ∝ e−tΓ = e−t/τ ,
where τ is the mean lifetime of the particle and Γ is the decay rate constant. This amounts to setting
ψ ∼ e−it(MX c /h̄−iΓ/2) , which you can check yourself by evaluating ψψ ∗ . This suggests making the change
2
MX c2 → MX c2 − iΓh̄/2 in Equation 1.7, which leads to:
g 2 h̄2 c2
M q2 =

2 c4 , (1.23)
q2 − MX + iMX c2 Γh̄
21
where we have assumed that MX c2 ≫ Γh̄ and therefore neglected the Γ2 term in the expression. The cross
2
section is then proportional to |M| giving:
g 4 h̄4 c4
σ∝ 2 .
2 c4 ) + M 2 c4 Γ2 h̄2
(q 2 − MX X
The dependence of the cross section on p2 , which is also the square of the centre-of-mass energy, is known
as a Breit-Wigner resonance, and leads to a characteristic distribution when measuring the cross section
versus the centre-of-mass energy (see the lecture slides for plots). The distribution peaks at the mass of the
decaying particle and has an intrinsic width dependent on the decay rate of the particle. As we will see later
in the course, particles are discovered by looking for such peaks in the measured data. And this is also often
how we measure their masses, as we know that the peak will be at the mass of the decaying particle. For
a particle that can decay to different final states, the total decay rate is equal to the sum of the individual
decay rates. Consequently, the total width in the lineshape is the sum of the partial widths that are obtained
for each individual decay. This means that the number of decay modes can be extracted by measuring the
width of the distribution, you will see in the module on “leptons and quarks” (section 2) that this is how we
have determined the number of neutrino flavours.
22
Chapter 2
Leptons & Quarks
We will now have a more systematic look at the matter particles that make up particle physics. We will see
that there are composite particles that are made up of other particles, as well, as far as we know, fundamental
elementary particles. Fundamental matter particles come in three generations and two main types: Leptons
and Quarks. The fundamental particles that build up everyday matter are all from the first generation: two
Leptons (the electron and the electron-neutrino) and two Quarks (the up and down quark).
One property we use to distinguish particles is the spin. An intrinsic property that has some interesting
quantum mechanical consequences. Particles with integer spin are called Bosons (following Bose-Einstein
statistics) and particles with half integer spin are called Fermions, following so called Fermi-Dirac statistics.
This leads to the Pauli exclusion principle, which states Fermions can not occupy the exact same state.
2.1 Leptons
Leptons are fundamental Fermions that do not interact through the strong force. They come in three gener-
ations, where each generation consists of a doublet of an electrically charged lepton ℓ− and a corresponding
neutrino νℓ . The electrically charged lepton ℓ− interacts through the electromagnetic and weak force, while
the neutrino can only interact through the weak force. The corresponding anti-particles have the same masses
but opposite charges and quantum numbers.

νe νµ ντ ν̄e ν̄µ ν̄τ
(2.1)
e− µ− τ− e+ µ+ τ+
2.1.1 Lepton Number

From experimental observations as well as symmetries of the underlying theory, we can deduce a conserved
quantity called Lepton number. Negatively charged Leptons and the corresponding neutrino carry a Lepton
number of 1. Antiparticles carry the opposite Lepton Number of −1. For states with more than one Lepton,
the Lepton number is additive.
L = N (l− ) − N (l+ ) + N (νl ) − N (ν¯l ); l = e, µ, τ (2.2)
The lepton number L as well as the individual lepton numbers (Le , Lµ , Lτ ) are conserved in all interactions.
23
Figure 2.1: An overview of the elementary matter particles. Three generations of Quark and Lepton doublets
exist.
2.1.2 Lepton decays

As neutrinos are the lightest particles of a generation, they are stable. The lightest electrically charged
lepton, the electron, is stable. The heavier leptons can decay through the exchange of a weak W boson.
Lepton Universality means that the coupling of the W boson to any Lepton Doublet (a charged lepton
and a neutrino) are the same for all three generations. However decay rates etc. will also depend on the
“phase space”, e.g. the available energy and hence we need to take the mass differences between lepton
species into account. We can make a simple dimensional argument to understand this dependence. The
decay width Γ has units of [mass]. The Feynman diagram for charged lepton decay (and hence the decay
amplitude) contains a W propagator m12 , which has to be squared to calculate the decay rate, proving a
W
term with dimensions ∼ [mass]−4 . In order to arrive at units of [mass] for the decay width Γ, a term with
dimensions ∼ [mass]5 is needed. The relevant Q-value is the mass of the decaying Lepton, hence:
2
1 1 5
= Γ(µ− ) ∼ Γ(µ− → e− νµ ν̄e ) ∼ (mµ ) → [mass] (2.3)
τµ m2W
When comparing the decay rate of τ Leptons and muons, we also need to take into account that the
heavier τ Lepton has more options to decay into. We hence have to correct with the branching fraction
Br(τ − → e− ντ ν̄e ):
Γ(τ − → e− ντ ν̄e ) Γ(τ − )Br(τ − → e− ντ ν̄e )

= (2.4)
Γ(µ− → e− νµ ν̄e ) Γ(µ− → e− νµ ν̄e )
We can estimate the value for the branching fraction Br(τ − → e− ντ ν̄e ), by considering the possible decay
modes for the τ Lepton:
24
• τ − → e− ντ ν̄e
• τ − → µ− ντ ν̄µ
• τ − → ντ dū, as we will see later, this decay can appear in three different colours for the up and anti-down
quarks
So we get a value for the branching fraction Br(τ − → e− ντ ν̄e ) of 1/(2+3*1) = 1/5 = 0.2 The measured
value of Br(τ − → e− ντ ν̄e ) = 0.17 takes into account that the decay probabilities into these 5 modes are not
exactly equal.
5
Γ(τ − ) × 0.17

mτ
∼ (2.5)
Γ(µ− ) mµ
Putting in measured values for the lifetimes and masses of the two Leptons, it can be seen that the
relation holds to decent accuracy.
Γ(τ − ) τµ 2.2µs
= ∼ ∼ 7.58 × 106 (2.6)
Γ(µ− ) ττ 2.9 × 10−13 s
5 5
1 mτ 1 1.777 GeV
∼ ∼ ∼ 7.79 × 106 (2.7)
0.17 mµ 0.17 0.106 GeV
2.1.3 Neutrino Oscillations

In the SM neutrinos are massless. However we now know that this description is not adequate and indeed
neutrinos do posses mass. This has some important implications. In particular the mass eigenstates could
be different from the flavour eigenstates of the weak interaction. The flavour eigenstates would than be
superpositions of the mass eigenstates as shown in Eq.2.9.
|νe ⟩ = |ν1 ⟩ cos θ + |ν2 ⟩ sin θ (2.8)

|νµ ⟩ = −|ν1 ⟩ sin θ + |ν2 ⟩ cos θ (2.9)
This essentially corresponds to a rotation between two sets of basis state vectors.

νe cos θ sin θ ν1
= (2.10)
νµ − sin θ cos θ ν2
Due to the different time evolution this leads to an oscillation between neutrino states and hence lepton
flavour violation:
|ν, t = 0⟩ = |νe ⟩ = |ν1 ⟩ cos θ + |ν2 ⟩ sin θ (2.11)
−iE1 t −iE2 t
|ν, t⟩ = |ν1 ⟩e h̄ cos θ + |ν2 ⟩e h̄ sin θ (2.12)
Meaning, a neutrino, that started out as, e.g., a νe can at a later time be identified as νµ . We can use Eq.2.12
to calculate the time dependent transition probability:
25
−iE1 t −iE2 t 2
P (νe → νµ ) = |⟨νµ |ν, t⟩|2 = − sin θ cos θe h̄ + sin θ cos θe h̄
−iE2 t −iE1 t 2 sin2 2θ −iE2 t −iE1 t
−iE2 t −iE1 t
∗
= sin2 θ cos2 θ e h̄ − e h̄ = e h̄ − e h̄ e h̄ − e h̄
4
2 sin2 2θ
sin 2θ −iE2 t −iE1 t
iE 2 t iE1 t −i(E1 −E2 )t i(E1 −E2)t

= e h̄ − e h̄ e h̄ − e h̄ = 1−e h̄ +1−e h̄
4 4 h i
 (E1 −E2 )t
−i(E1 −E2 )t i(E1 −E2)t
1 − cos
!
2
sin 2θ e h̄ +e h̄ h̄
= 1− = sin2 2θ  
2 2 2
2 (E1 − E2 )t

2
= sin 2θ sin
2h̄
A more convenient form makes use of the following approximation:

s
m2 c4 1 m2 c4
p
E = p c + m c = pc 1 + 2 2 ∼ pc 1 +
2 2 2 4 (2.13)
p c 2 p2 c2
L
We will also replace time t with the flight distance L assuming speed of light: t = c. Which leads to:
26
(Eα − Eβ )t

2 2
P (να → νβ ) = sin (2θ) sin (2.14)
2h̄
" #
(m2α − m2β )c3 1 L
= sin2 (2θ) sin2 (2.15)
2p 2h̄ c
" #
2 2 3
(m α − m β )c L
= sin2 (2θ) sin2 (2.16)
E 4h̄
" #
2 2
(m2α − m2β )c4 L
= sin (2θ) sin (2.17)
E 4ch̄
" #
2 2 4
(m α − m β )c L
= sin2 (2θ) sin2 (2.18)
E 4 ∗ 197[ MeV][ fm]
" #
2 2
(m2α − m2β )c4 L
= sin (2θ) sin (2.19)
E 788 · 10−9 [ eV][ m]
" #
2 2 4
(m α − m β )c L
= sin2 (2θ) sin2 (2.20)
E 0.788 · 10−9 [ eV][ km]
" #
2 2
(m2α − m2β )c4 L
= sin (2θ) sin ·1· (2.21)
E 0.788 · 10−9 [ eV][ km]
" #
2 2
(m2α − m2β )c4 [ eV] L
= sin (2θ) sin (2.22)
E [ eV] 0.788 · 10−9 [ eV][ km]
" #
2 2
(m2α − m2β )c4 10−9 [ GeV] L
= sin (2θ) sin (2.23)
E [ eV] 0.788 · 10−9 [ eV][ km]
" #
2 2
(m2α − m2β )c4 L[ GeV]
= sin (2θ) sin (2.24)
E 0.788 · [ eV2 ][ km]
" #
2 2
(m2α − m2β )c4 [ GeV] L
= sin (2θ) sin 1.27 (2.25)
[ eV2 ] E [ km]
2 4

2 2 ∆m c [ GeV] L
= sin (2θ) sin 1.27 (2.26)
[ eV2 ] E [ km]
Where we also used h̄c ∼ 197 MeV · fm.

The mixing formalism (Eq. 2.10 ) can be extended to three neutrino families:
+  
νe Ue1 Ue2 Ue3 ν1
+
νµ =  Uµ1 Uµ2 Uµ3  ν2 (2.27)
ντ Uτ 1 Uτ 2 Uτ 3 ν3
27
The neutrino mixing matrix is known as the PMNS or Pontecorvo–Maki–Nakagawa–Sakata matrix:
0 sin θ13 e−iδ

     
Ue1 Ue2 Ue3 1 0 0 cos θ13 cos θ12 sin θ12 0
U =  Uµ1 Uµ2 Uµ3  =  0 cos θ23 sin θ23   0 1 0   sin θ12 cos θ12 0 
Uτ 1 Uτ 2 Uτ 3 0 − sin θ23 cos θ23 − sin θ13 eiδ 0 cos θ13 0 0 1
(2.28)
Evidence for Neutrino Oscillation

The first hints for Neutrino oscillations came from the so called Solar Neutrino deficit. The nuclear reactions
that power the sun, produce electron neutrinos. Ray Davis performed a long standing pioneering experiment
in the Homestake mine in South Dakota. The flux recorded on earth showed a deficit with respect to
theoretical calculations. For a long time it was not clear if improved calculations or a new physics mechanism
was needed. The Sudbury Neutrino Observatory (SNO) finally settled the issue. Solar neutrino experiments
measure sin2 (2θ12 ).
Atmospheric Neutrinos originate from the decay of Pions and Muons produced by impinging cosmic rays
in the atmosphere. The SuperKamiokande Experiment provided the first conclusive evidence for Neutrino
Oscillations, by measuring the zenith angle dependence of the atmospheric muon neutrino flux. Measurements
of atmospheric neutrinos yield sin2 (2θ23 ). The 2015 Physics Nobel Prize went to Takaaki Kajita and Arthur
McDonald, leading figures of the Superkamiokande and SNO experiments. Ray Davis received the Prize in
2002 for his work.
Neutrino oscillations are now also studied at nuclear reactors. Reactor experiments usually are made up
of several detectors at different distances from the source, a so called Near and Far detector. With that
technique the uncertainty on the incoming neutrino flux can be reduced by the measurement of the nearly
unmixed flux at the Near detector end, close to the reactor. Reactor neutrinos are currently used to determine
the third mixing angle sin2 (2θ13 ).
sin2 (2θ12 ) = 0.851 ± 0.019 (2.29)

2
sin (2θ23 ) = 0.992+0.006
−0.009 (2.30)
2
sin (2θ13 ) = 0.086 ± 0.003 (2.31)
2.1.4 Absolute Neutrino mass scale

Neutrino oscillation measurements only tell us that neutrinos have mass and what the mass difference between
the neutrino species is, they do not provide a measure of an absolute mass scale.
∆m2solar = ∆m221 = (7.53 ± 0.18) × 10−5 eV2 (2.32)

∆m2atm = ∆m232 = (2.453 ± 0.034) × 10−3 eV2 (2.33)
As the sign of the mass difference is not always known, this leads to two possible neutrino mass hierarchies.
The normal hierarchy m1 < m2 < m3 , shows an increase in mass for every generation, as observed for the
charged leptons and quarks.
There are three common ways to measure the absolute neutrino mass scale, the end point of the β−decay
spectrum, measurements of neutrino-less double beta decay and limits from cosmology.
28
The end point of the β−decay spectrum shifts to lower energies with increasing mass of the electron
neutrino. This is exploited for example in the Katrin experiment, which measures the electron spectrum
from Tritium decays. Tritium is particular suitable as the energy available to the electron is very small
E ∼ 18.6keV. Further details on the deformation of the spectrum are discussed in section 9.4.2 on Kurie
plots.
The structure formation in the early universe depends on the gravitational pull of the mass density
of relic neutrinos. Hence measurements of density fluctuations allow to put limits on the sum of neutrino
masses.
Neutrino-less double beta decay experiments

Double beta decay experiments, like SuperNemo, can not only measure the absolute neutrino mass, but can
also make an inference about the nature of the neutrino. Fermions in the Standard Model have a distinct
anti-particle that carries opposite charges. Particles with a distinct anti-particle are called Dirac particles. As
neutrinos carry no charge, they could be their own anti-particle. We call such particles Majorana particles.
Figure 2.2: Feyman diagram of double beta decay.
Double beta decay is allowed in the Standard Model and occurs in several nuclei. As we will see in
detail in section 8.6, β−decay transforms a neutron into a proton and emits an electron and an anti-electron
neutrino. In double beta decay two such transitions occur simultaneously. For a nucleus of mass A and
charge Z (A, Z) → (A, Z + 2) + 2e− + 2ν¯e . If neutrinos are of Majorana type and are their own anti-particle,
they can annihilate and neutrino-less double beta decay can occur (A, Z) → (A, Z + 2) + 2e− . Double beta
decay is already a rare process and the rate for neutrino-less double beta decay is even smaller. The decay
rate for neutrino-less double beta decay is proportional to the neutrino mass squared and can be used to
infer the neutrino mass scale. Neutrino-less double beta decay violates Lepton number and is forbidden in
the Standard Model.
29
2.1.5 Number of Neutrinos

Three neutrino species (νe , νµ , ντ ) have been observed so far. From measurements of the decay width of the
Z boson at LEP, we know that there are only three neutrinos with a mass below (m < 45 GeV), half the
mass of the Z boson. The Z boson can decay into neutrinos (Z → ν ν̄) and hence another neutrino species
would increase its decay width. The resulting fit to the number of neutrinos can be seen in Figure 2.3.
Figure 2.3: The fit of the Z boson lineshape with the number of neutrinos as parameter. The filled
circles are the data and three theoretical curves for 2,3 and 4 neutrino species are shown. Taken from
doi:10.1016/j.physrep.2005.12.006.
2.1.6 Charged Lepton Number Violation

Neutrino oscillation violates Lepton number, but on a macroscopic scale. How can the Lepton number
of charged Leptons still be conserved in microscopic interactions and decays? The answer lies in the vast
difference of scales. Neutrino oscillations occur over kilometres, while the length scale of interactions is much
much smaller.
The probability for the Lepton number violating decay µ− → e− γ can be related to the known branching
ratio of the muon decay µ → e− ν¯e νµ , the neutrino oscillation probability P (νµ → νe ) and the probability
of emitting a photon. The probability of emitting a photon is simply given by the electromagnetic coupling
constant α = 1/137.
Br(µ → e− νµ ν¯e γ) = αBr(µ → e− νµ ν¯e ) (2.34)
The branching ratio of the muon decay Br(µ → e− ν¯e νµ ) ∼ 1.0, is essentially one, as all other decay pos-
sibilities are forbidden or suppressed. This yields the following equations for the probability of the Lepton
30
Figure 2.4: Feynman diagram for the process µ− → e− γ
number violating decay µ− → e− γ:
Br(µ → e− γ) = αBr(µ → e− νµ ν¯e )P (νµ → νe )

1
= 1.0P (νµ → νe )
137
∆m212 L

1 2 2
∼ sin (2θ12 ) sin 1.27
137 Eν
−5 2
7 × 10 eV L

0.9
∼ sin2 1.27
137 Eν
× −5
eV2 L

0.9 7 10
∼ sin2 1.27
137 Eν
2
7 × 10 eV2 L
−5

−3
∼ 10 1.27
Eν
Where the small angle approximation sin θ ∼ θ was used in the last step. We have inserted the measured
values for ∆m2 . In the above formula, L is in [m] and Eν in [MeV] The length L can be estimated using
Heisenbergs uncertainty relation for emitting a virtual W-Boson of mass mW ∼ 80 GeV.
ch̄ ch̄ 197 MeVfm
L = c∆t = = ∼ ∼ 2.5 × 10−3 × 10−15 m
∆E mW 80 GeV
This is the known range of the weak force. The energy of the neutrino will be roughly Eν < mµ − me , leading
to a probability of < 10−48 , which is negligible. Hence microscopically Lepton number remains conserved in
decays and interactions!
2.2 Quarks
Quarks are strongly interacting fundamental Fermions. Quarks also form doublets and occur in three gen-
erations. While we have never observed free quarks, their existence is very well established. The first
experimental evidence came from Lepton-Nucleon scattering, the SLAC-MIT experiment shot electrons on
protons and found three constituents, in a fashion similar to the famous Rutherford experiment. Friedman,
31
Kendal and Taylor were honoured with the Nobel Prize for their discovery. The notion of different quark
types comes from the increase in cross section at e− e+ machines. When passing the energy threshold for a
+ −
e →hadrons)
new type of quark-anti-quark pair, the ratio R = σ(e 2
σ(e+ e− →µ+ µ− ) is increased by a step of size NC ∗ Qquark
in electron-positron colliders. Quarks come in three different colour charges, hence we have to multiply final
states by a colour factor NC = 3. We first encountered this when discussing the decay of the τ lepton. The
quark and anti-quark doublets are:

u c t ū c̄ t̄
(2.35)
d s b d¯ s̄ b̄
2.2.1 Hadrons
Hadrons are strongly interacting composite particles made up of quarks. We will later see in the discussion
of the strong force or Quantum Chromo Dynamics (QCD) that quarks carry a colour charge, red, blue or
green. Due to quark confinement, that will be discussed further in the lecture on strong interactions, quarks
can not be observed freely but form colourless hadrons. The reason the colour charge has its name, is that it
has properties similar to real colour. There are two possibilities to form colourless objects: combine a colour
with its anti-colour (blue and anti-blue, e.g yellow) or combine three colours (like r,g,b). This leads to the
two types of hadrons: Mesons (q q̄) and Baryons (qqq) and Anti-Baryons (q̄ q̄ q̄).
The number of quarks Nq = N (q) − N (q̄) is conserved. This leads to the Baryon number B = Number
of quarks/3 =Nq/3 being conserved as well. Baryons have Baryon number B = 1, for Anti-Baryons B = -1
and Mesons have B = 0.
Only strong and electromagnetic interactions conserve also the individual quark flavour number:
Nf = N (f ) − N (f¯); (f = u, d, s, c, b, t) (2.36)
New quantum numbers Strangeness S, charm C, Beauty B were introduced. These are conserved in strong
and electromagnetic interactions. → If flavour number changes, we deal with a weak decay!
Examples of Hadrons
¯ made up of up and down type
Mesons The three lightest mesons are the π mesons π − (dū), π 0 (uū), π + (ud),
¯ The quark contents of π mesons
quarks. Actually the π 0 is more realistically a superposition √12 (uū + dd).
are easily derived, knowing that a u-quark has +2/3 and a down quark −1/3 electric charge.
Further examples of Mesons can be found in Tab. 2.1. A complete list can be obtained from:
http://pdg.lbl.gov/2017/listings/contents listings.html.
Baryons Protons (uud) and neutrons (udd) are examples of Baryons made up of quarks of the first gen-
eration. Again their quark content can be easily derived from the electric charges. Both the proton and
neutron have a mass of roughly 1 GeV.
Further examples of Baryons can be found in Tab. 2.2 A complete list can be obtained from:
http://pdg.lbl.gov/2017/listings/contents listings.html.
2.2.2 Hadron Decay

The proton is the lightest Baryon and hence stable (τ > 1033 years), the heavier neutron can decay. All other
known hadrons (Mesons and Baryons) decay. The lifetime varies substantially and depends on the allowed
32
particle mass [MeV] Lifetime [s] main decays

¯
π + (ud) 140 2.6 × 10−8 π + → µ+ νµ (∼ 100%)
0 √1 ¯
π ( 2 (uū + dd)) 135 8.4 × 10−17 π 0 → γγ(∼ 100%)
K + (us̄) 494 1.2 × 10−8 K+ → µ+ νµ (∼ 64%); K + → π + π 0 (∼ 21%)
K ∗+ (us̄) 892 1.3 × 10−23 K + → Kπ(∼ 100%)
D− (dc̄) 1869 1.1 × 10−12 Several
B − (bū) 5278 1.6 × 10−12 Several
Table 2.1: Some examples of Mesons
particle mass [MeV] Lifetime [s] main decays

p(uud) 938 Stable None
n(udd) 940 887
Λ0 (uds) 1116 2.6 × 10−10 Λ0 → pπ − (64%); Λ0 → nπ 0 (36%)
++
∆ (uuu) 1232 ∼ 0.6 × 10−23 ∆ → pπ + (100%)
Ω− (sss) 1672 0.8 × 10−10 Ω → Λ K (68%); Ω → Ξ0 π − (24%)
0 −
Λ+c (udc) 2285 2.1 × 10−13 Several
Table 2.2: Some examples of Baryons
decay mechanism. As can be seen in Fig. 2.5 decays through strong interactions lead to a very short lifetime,
where as particles that can only decay weakly have long lifetimes.
Lets look at some examples: The ρ0 has the same quark content as a π 0 but it is heavy enough that it
can decay strongly via gluon emission into two Pions:
¯ → ππ
ρ0 (uū + dd) m(ρ) = 775 MeV
The π 0 is too light (there are no lighter hadrons it could decay into) and has to decay electromagnetically
into two photons:
¯ → γγ m(π 0 ) = 135 MeV
π 0 (uū + dd)
As we can see in Fig. 2.5 the ρ has a much shorter lifetime than the π 0 . The ρ can also undergo electromagnetic
decays, but the much faster strong decay dominates. The charged Pion can only decay via the weak force,
exchanging a W-Boson:
¯ → µ+ νµ
π + (ud)
Hence it has a much longer lifetime as seen in Fig. 2.5.
The strong and electromagnetic interactions conserve flavour, while W exchange of the weak interaction
changes flavour. That’s why flavour transitions in general involve a charge change. The exchange of a weak
Z Boson also preserves flavour and so called neutral flavour changing currents are suppressed. This will be
revisited in more detail in the lectures on electroweak interactions, section 6.
The lifetime also depends on the available energy states. So for example the neutron decay n → p+e− + ν¯e
exhibits a very low Q-value, Q = mn −mp −me ∼ 0.79 MeV and hence has a rather long lifetime of τ ∼ 887 s.
We will encounter another example of this in the decay of neutral Kaons K 0 (ds̄) in section 6.2.4. Neutral
Kaons can decay into two K 0 → ππ or three Pion states K 0 → πππ. Both are flavour changing weak decays
33
Figure 2.5: Lifetimes of various particles and the underlying dominant decay mechanism. (from G. Venkatara-
man, The Big and the Small: v. 1: Journey into the Microcosm - The Story of Elementary Particles, Sangam
Books Ltd, 2002)
(a s → d transition). The K 0 → πππ decay is suppressed by an additional gluon emission and a largely
reduced Q-value (Q = mK − 3mπ < mK − 2mπ ).
34
Chapter 3
Experimental methods
3.1 Introduction
After introducing the basic building blocks of matter, it is now time to learn about the experimental methods
employed. There are a number of sources of particles: In the early days of particle physics, many particles
were discovered in cosmic rays. Cosmic-ray particles have been observed with energies in excess of 1019 eV.
However the rate of such extremely energetic particles is very low (∼ 1 per km2 per century) and very
unpredictable. While experiments continue to investigate cosmic rays to this day, the advent of accelerators
has dramatically advanced research, providing high intensity sources of known particles with a known energy.
3.2 Accelerators
All particle accelerators employ the Coulomb force to increase particle energy. Early accelerators used a
static electric field, while modern accelerators employ electromagnetic waves for acceleration. Accelerators
can be divided into linear and circular machines. Circular accelerators use magnets to bend the charged
particles into an orbit. They have the advantage that the energy can be increased in several revolutions,
hence reducing the size required to achieve the same energy compared to linear accelerators. However circular
accelerators suffer from energy losses due to Synchrotron radiation. As the effect is proportional to 1/m4
this limits in particular the acceleration of electrons, which are light. The first particle accelerator was the
Cyclotron build by Ernest Lawrence, that used a fixed magnetic field. Now mostly synchrotrons are used,
where the magnetic field strength is increased synchronously with energy, allowing for a constant bending
radius.
3.2.1 Centre-of-mass energy

Experiments at particle accelerators can be divided into two categories: collider experiments and fixed target
experiments. In a collider experiment two particle beams are brought into collision. While asymmetric setups
exist, most commonly the two colliding beams have the same energy E and collide head on. Then the initial
four vectors are P1 = (E, p) and P2 = (E, −p) and the centre of mass energy squared s = (P1 + P2 )2 =
P12 + 2 ∗ P1 · P2 + P22 .
35
Neglecting masses: P12 = P22 = 0 and the centre of mass energy

√ p p
s = 2P1 · P2 = 2EE − 2|⃗ p||⃗
p| cos θ
.
For a head on collision → cos θ = −1, neglecting masses |⃗p| ∼ E, hence
√ p √
s = 2EE − 2|⃗ p||⃗
p| cos θ = 2EE + 2EE = 2E
In a fixed target experiment, a beam of energy E hits not to surprisingly a fixed target of mass m.
The initial four vectors are P1 = (E, p) and P2 = (m, 0), then centre of mass energy squared s = (P1 + P2 )2 =
P12 + 2 ∗ P1 · P2 + P22 . Neglecting masses of the beam particle again: P12 = 0 but P22 = m2 and the centre of
mass energy becomes
√ p p
s = 2P1 · P2 + m2 = 2Em + m2 .
√ √
For large energies we can also neglect the target mass s ∼ 2Em.
3.2.2 Luminosity
An important quantity that characterizes the performance of an accelerator is the instantaneous luminosity
L. The luminosity relates to the rate R of a certain process via:
dN
R= = Lσ (3.1)
dt
where σ is the cross section for the process. The integrated luminosity L that gives the number of produced
events: Z
N = σL = σ Ldt (3.2)
The instantaneous luminosity is related to the flux Φ1,2 of the two particles:
dN1,2 1
Φ1,2 = = n1,2 v1,2 (3.3)
dt A
where n1,2 is the number density of particles, A the area they pass through and v1,2 their velocity. The
collision rate is given by the incoming flux and the number and size of targets.
dN
= Φ1 Ntarget σ (3.4)
dt
Considering one beam, the number of targets is given by
Φ2
N2 = n 2 A = A (3.5)
v2
So that for a circular collider the instantaneous luminosity is given by:
nN1 N2 f
L= (3.6)
A
where n number of colliding bunches (particle packets), N1,2 = number of particles in each bunch for beam
1,2, f = frequency with which bunches circulate the ring A = 4πσx σy cross-sectional area of the beams.
36
Note that the common notation uses σ both for the cross section, as well as σx, for the extend of the beam
profile in each direction.
In order to increase the luminosity, we can make the beam area smaller (reduce A, σx , σy ), increase the
number of particles (increase N1,2 ) or increase the number of bunches. In principle the revolution frequency
v
f = 2πR can also be used to increase the luminosity, but in practice particles are already highly relativistic,
moving very close to the speed of light (0.9999c), that any increase is negligible.
3.3 Interaction of Particles with matter

In order to observe particles, they need to interact. The mechanisms of particle interaction, depends on the
particle type. Hadrons will interact strongly with atomic nuclei. Charged Leptons and Hadrons ionise via
the electromagnetic interactions. Photons, being the carriers of the electromagnetic force, interact with the
charges inside the atom. Neutrinos can only interact weakly and hence are the most difficult particles to
detect.
3.3.1 Ionisation Energy loss: the Bethe-Bloch Formula

Particles heavier than electrons lose energy through electromagnetic interactions with the electrons of the
atom. This energy loss was calculated by Bethe and Bloch. Ionisation is the most important mechanism
for particle detection. The Bethe-Bloch curve, shown in Fig. 3.1, has a number of important characteristic
features. The energy loss of a particle inside a given material, depends only on the speed and charge of
the particle (∼ βγ). At low energies dE/dX ∼ 1/β 2 and slower particles lose more energy. Particles which
lose the least energy around the minimum of βγ ∼ 4 are called minimum ionising particles (MIPs). At high
energies, there is a relativistic rise in the energy loss.
We can derive the basic features using a simple model of a particle moving a long a straight line in the
electric field of an electron. From geometry we can define see Figure. 3.2.
x = b tan θ (3.7)
b = r cos θ (3.8)
∆x
∆t = (3.9)
v
b
dx = d(b tan θ) = dθ (3.10)
cos2 θ
The Force on the particle is given by Coulomb’s law. Then from integrating Newton’s law F = ma =
dp/dt, we can derive an expression for the momentum transfer. This is the momentum transferred from the
37
dE 1
Figure 3.1: Shown is the energy loss dX ρ versus γβ in different media (hydrogen H2 to lead P b). For particles
with different masses (shown muon, Pion and proton), the same momentum corresponds to different βγ. For
a fixed momentum a heavier particle has a lower βγ. (Taken from https://pdg.lbl.gov/2016/reviews/rpp2016-
rev-passage-particles-matter.pdf)
particle to one electron in the material:

ze2 ze2
F⊥ = FCoulomb · cos θ = 2
· cos θ = · cos θ (3.11)
r (b/ cos θ)2
Z∞ Z∞
dx ze2 dx
Z
∆p = F⊥ dt = F⊥ = · cos θ (3.12)
v (b/ cos θ)2 v
−∞ −∞
π π
Z2 2 3 Z2
ze cos θ b ze2 cos θ 2ze2
= dθ = dθ = (3.13)
vb2 cos2 θ vb vb
−π
2 −π
2
38
Figure 3.2: Schema of a particle of charge ze approaching an electron.
In summary we derived the following expressions for the momentum transfer and energy loss for the
interaction with one electron:
2ze2
∆p = (3.14)
vb
me v 2 ∆p2 2z 2 e4
∆E = = = (3.15)
2 2me me v 2 b2
When passing through material the particle will of course encounter many electrons. The total number
of electrons to be considered is given by Eq. 3.16. Here we we used that every atom has Z electrons and the
density of atoms is NAA ρ :
NA ρ
dNe = ne dV = ne 2πb db dx = Z 2πb db dx (3.16)
A
In the above formula, ne is the Electron density, NA Avogadros number (#atoms), Z atomic number
(#electrons per atom), A atomic weight and ρ the mass density.
So for each electron encounter an energy amount, ∆E as defined in Eq. 3.14, is going to be lost. So the
total energy loss amounts to:
NA ρ
−dE = ∆EdNe = ∆EZ 2πb db dx (3.17)
A
2z 2 e4 NA ρ
= Z 2πb db dx (3.18)
me v 2 b2 A
4πz 2 e4 NA ρ db
= Z dx (3.19)
me v 2 A b
Hence the energy loss per length can be calculated as:
bZ
max
dE 4πz 2 e4 NA ρ db
→− = Z (3.20)
dx me v 2 A b
bmin
4πz 2 e4 Z bmax
= 2
NA ρ ln (3.21)
me v A bmin
39
Now we need to know what the smallest bmin and largest bmax allowed impact parameters are. From
Eq. 3.14 we see that these are related to the largest momentum transfer and the smallest possible energy
transfer.
2ze2 ze2 2 ze2 2

bmin = = = (3.22)
v∆pmax v ∆pmax v 2me v
s r
2z 2 e4 ze2 2
bmax = = (3.23)
me v 2 ∆Emin v Ime
Where the maximum momentum transfer is ∆pmax = 2me v in analogue to a ball bouncing of a wall
and I is the Ionisation potential, which corresponds to the smallest amount of energy required to kick of an
electron from the atom.
Inserting Eq. 3.23 into Eq. 3.21 yields:
dE 4πz 2 e4 Z bmax
− = NA ρ ln (3.24)
dx me v 2 A bmin
2 4
r
4πz e Z ze2 2 v 2me v
= NA ρ ln (3.25)
me v 2 A v Ime ze2 2
s
4πz 2 e4 Z (2me v)2
= N A ρ ln (3.26)
me v 2 A 2Ime
4πz 2 e4 Z 1 (2me v)2
= 2
NA ρ ln (3.27)
me v A2 2Ime
4πz 2 e4 ne 1 2me v 2
= ln (3.28)
me v 2 2 I
In the last step Eq. 3.16 was used.
Bragg Peak
From the features of the Bethe-Bloch curve and Fig. 3.1 it can be seen that a particle loses most energy when
it is slowest. As a particle gets stopped in a medium it will deposit most of its energy at the end of its path.
This has important applications in medical physics as proton beam therapy (PBT) radiating cancer cells.
3.3.2 Interactions of Photons and Electrons

Photons can lose energy through three mechanisms:
Photoelectric effect This mechanism was first explained by Albert Einstein. A photon hits an electron in
the orbit around the nucleus hard enough to liberate it from its shell, while the photon gets absorbed in the
3
process. The probability for the occurrence of the photoelectric effect is roughly proportional to ∼ Z 5 /E 2
(as can be seen in Fig. 3.3), hence it is important at low energies and for materials with large Z.
The probability for a photon to lose energy through the occurrence of photoelectric effect, strongly depends
on material and energy, as dictated by the shell structure of the atom.
40
Compton scattering In Compton scattering a photon scatters of an electron, gets deflected and transfers
some of its energy. The probability for Compton scattering is roughly ∼ Z ln E/E (as shown in Fig. 3.3).
Pair production Pair production is the dominant method of energy loss at high energies , once photons
have enough energy (E > 2me ) to produce an e− e+ pair (as indicated in Fig. 3.3). Its probability above
threshold is roughly constant in energy and increases with ∼ Z 2 . For this mechanism to conserve energy and
momentum, the electric field of a nucleus is needed.
Figure 3.3: The contributions of the three main mechanisms for Photon energy loss vs energy.
The relative contributions of these mechanisms at different energies is shown in Fig. 3.3. The photon
intensity is reduced exponentially with depth in a material:
I = I0 e−µx (3.29)
where µ is the attenuation coefficient.
Electron Energy loss through Bremsstrahlung

Electrons are very light particles and it is easy to deflect them in the electric field of the nucleus. When looking
at the Feynman diagrams for pair creation and Bremsstrahlung, we notice that the underlying mechanism is
the same.
Radiation length
The quantity that describes energy loss through bremsstrahlung or pair creation is the radiation length X0
or LR .
α3 Z 2

1 1 183
= = 4NA 2 ρ ln 1/3 (3.30)
X0 LR m A Z
41
The energy loss is then given by:

dE E
= (3.31)
dX X0
which leads to: x
E = E0 e− X0 (3.32)
Electromagnetic Shower
When a high energetic photon or electron enters a material an electromagnetic shower develops through
continuous repetition of bremsstrahlung and pair creation. We can understand this process using an extremely
simplified model as shown in Fig. 3.4. We make the assumption that every step of X0 electrons radiate a
photon and photons create an e+ e− pair with the energy equally distributed among the resulting particles.
In reality there is a continuous range of energies and steps.
After t steps, the number of particles N and their energy is given by:
E0 E0
N (t) = 2t ; E(t) = = t (3.33)
N 2
The shower process stops, once bremsstrahlung is not the dominating process any more. For electrons
this is the critical energy Ec at which the energy loss contributions from ionisation (aka Bethe-Bloch) and
bremsstrahlung are equal:
E0
Nmax = = 2tmax (3.34)
Ec
Figure 3.4: Schema for the development of an electromagnetic shower.
42
3.3.3 Interactions of Hadrons

In addition to ionisation, hadrons also undergo strong interactions with nuclei. In these interactions new
hadrons are produced and the average energy per hadron is reduced. The mechanisms for hadrons are more
complex than for photons and electrons.
Hadronic Shower
When hadrons are stopped in a material, they develop showers very similar to the electromagnetic showers
described in Section 3.3.2. The relevant quantity is the nuclear absorption length:
1 A
la = λa = = (3.35)
nσinel NA ρσinel
were σinel is the inelastic scattering cross section, NA is Avogadro’s constant, ρ the density and A the mass
number.
However the shower development is more complex. In particular neutral Pions, π 0 , are produced, that
decay into two photons, leading to an electromagnetic component of hadronic showers.
Cosmic ray air showers Primary cosmic rays hitting atomic nuclei in the top of the atmosphere create
hadronic showers, known as cosmic ray air showers. Due to the larger depth, many of the produced particles,
e.g. neutral and charged Pions and Kaons, can decay. These decays produce electromagnetic showers from
π 0 → γγ as well as muons from π ± → µ± νµ and K ± → π 0 µ± νµ . At ground level mainly muons survive.
During the development of the shower in the atmosphere, fluorescence and Cherenkov light is emitted from
the shower.
3.4 Particle Detectors

Particle detectors use the interaction of particles to measure there properties. The properties of interest are
the position of the particle trajectory, the particle momentum and energy.
For most particles these properties are reconstructed from their decay products.
The particles that pass a typical detector without decaying are:
• Electrons
• Muons
• Photons
• charged Pions
• Kaons
• Protons
• Neutrons
• Neutrinos
Most of these particles can be distinguished in an experiment by their interactions in specific sub-detectors.
Many experiments also employ detectors for particle identification (PID) as described in Section 3.4.4.
43
3.4.1 Position measurement

Position measurements indicate that a particle passed through a given location. In the early days of particle
physics, cloud, bubble and spark chambers where used to display the ionisation tracks of particles, much like
the trails of airplanes in the sky. For modern high rate experiments detectors with an electronic readout are
needed.
Scintillators
Scintillators indicate the passing of a charged particle by emitting light. Passing particles excite atoms and
molecules, light is emitted during the decay of these excited states. A large variety of scintillators exist and
they can be roughly divided into inorganic crystal and organic scintillators. Organic scintillators can be clear
plastic or liquid.
In order to readout scintillators electronically they need to be coupled to a light sensitive detector. The
traditional choice is a photomultiplier tube or PMT. A PMT consists of a photo cathode, where light frees
electrons via the photoelectric effect. These electrons are then accelerated with the help of an applied high
voltage towards a dynode, where the impact frees more secondary electrons. Through a chain or cascade of
dynodes along a potential difference more and more electrons are freed and finally create an amplified electric
pulse on the anode. More modern photo-detectors (SiPMT) employ principles of semiconductor detectors
described in section 3.4.1.
Gas Detectors
These are very basic counters that record particle passing. In their simplest from, they consist of a wire
inside a gas volume. A high voltage is applied to create a potential difference between the wire and the outer
wall.
The ionisation energy depends on the gas, on average an energy of ∼ 30eV is required to ionise. A number
of different gas detectors can be distinguished. The main difference is the so called gas multiplication created
by the acceleration of the electrons in the applied electric field.
Ionisation Counters At lower voltage only the charge carriers initially created in the ionisation process
drift to the anode and cathode.
Proportional Counters At a moderate voltages the electrons created in the initial ionisation process
are accelerated enough so they themselves can ionise the medium. The resulting avalanche leads to gas
multiplication and the initial pulse is amplified. In the proportional range, the resulting pulse height is
proportional to the initial ionisation pulse.
Geiger-Müller Counters At high voltage the gas multiplication saturates and the created pulse becomes
independent of the initial ionisation.
Semiconductor Detectors
Semiconductors like Germanium and Silicon have an energy gap between the valence and conduction bands
of the electrons. They can be doped with other atoms, so that an excess of n or p carriers is created. A
typical semiconductor detector operates as a reverse biased diode, creating a depletion zone on the border
44
between n- and p-doped material. Particles passing through the depletion zone, create charge carriers through
ionisation. Very low energy (2.8 eV for Ge and 3.6 eV for Si) is needed to create a pair of charge carriers.
This low energy makes especially Germanium a great detector for energy measurements. The possibility to
create very finely segmented semiconductor detectors has led to the creation of very high precision position
and tracking detectors.
3.4.2 Tracking Detectors

Tracking detectors use several position measurements to allow reconstruction of the trajectory of a particle.
They usually operate inside a magnetic field. Charged particles exhibit a curved trajectory in the magnetic
field, that depends on their momentum. If the magnetic field is known, the measurement of the curvature
radius can be used to determine the particle’s momentum. The deflection is given by the Lorentz force:
mv 2 mv p
F = qv × B = →R= = (3.36)
R qB qB
p/[GeV ]
A useful formula is R/[m] = 3.3× B/[Tesla] . Tracking detectors employ a number of different position
measurement technologies as described in section 3.4.1. Multi-wire Proportional Chambers (MWPC) use an
array of wires inside a gas volume. In 1992 George Charpak received the Nobel Prize in physics for this
invention. Drift Chambers allow a more precise position measurement by measuring the time between the
passing of the particles and the arrival of the drifting ionisation electrons. If the drift velocity is known the
distance to the wire can be calculated.
Semiconductor detectors as described in section 3.4.1 can be finely segmented and make excellent trackers.
Semiconductor microstrip detectors are segmented only in one dimension and consist of strips separated by
a few (∼ 50)µm. Pixel detectors are very similar to a digital camera and use pixel dimensions as small
as 25 × 100µm2 . These structures are etched and doped using processes very similar to those used in the
production of microprocessor chips. A tracker consists of several layers of these detectors. A passing particle
leaves a hit in each layer. Special algorithms are needed to reconstruct the trajectory of the particles from
these hits. Decay vertices can be reconstructed by intersecting tracks.
A scintillating fibre tracker utilises thin fibres made of plastic scintillator grouped together in bundles.
3.4.3 Calorimeters: Energy Measurement

Calorimeters stop particles and measure their energy. As we have seen in Sections 3.3.2 and 3.3.3, when
particles enter material, they shower. In general two kinds of calorimeters are used: electromagnetic and
hadronic.
Calorimeters that actively sample only parts of the shower are called sampling, sandwich or shashlik
calorimeters. The names come from the fact that a dense passive absorber layer (often iron or lead) is
interspersed with an active layer (plastic scintillator, liquid argon, etc.).
The main contribution to the resolution of calorimeters are statistical fluctuations in the shower devel-
opment. The measured energy is proportional to the number of final √ particles recorded. The number of
particles N is governed by Poisson statistics and has an uncertainty N .
√
σ(E) N 1 1
∼ =√ ∼√ (3.37)
E N N E
45
Electromagnetic Calorimeter
As the radiation length depends on the Z of the material for electromagnetic calorimeters (ECAL) lead
glass or inorganic scintillators, Bismuth Germanate (BGO) and lead tungstate (PbWO4), are used. ATLAS
uses a sampling calorimeter with liquid argon as active material and lead as absorber. The electromagnetic
calorimeter stops photons and electrons. Other charged particles, like charged hadrons, deposit some energy
but pass through the ECAL. Electromagnetic calorimeters can reach resolutions of σ(E) √1
E ∼ 1 − 10% E .
Hadronic Calorimeter
Hadronic calorimeters (HCAL) follow after the electromagnetic calorimeter. For hadronic calorimeters the
nuclear interaction length and hence density is the important parameter. Hadrons shower and get stopped
in the hadronic calorimeter. Due to the complexity of hadronic interactions and large fluctuations, hadronic
calorimeters only reach a resolution of σ(E)
E ∼ 35 − 100% E
√1
3.4.4 Particle Identification

In a normal detector it is impossible to distinguish the different types of semi-stable charged hadrons (pions,
kaons and protons). It can also be very hard to distinguish pions from electrons. In order to help with this,
specific detectors are employed for PID. The general principle is to find the mass of the particle in question
by measuring its momentum and velocity simultaneously.
mβ v
p = βγm = p ; β= (3.38)
1− β2 c
Time of Flight An easy way to measure the velocity β is to measure the time ∆t a particle takes to cross
a known distance L. Then v = cβ = L/∆t. Either the particle crossing is measured at start and end or
the beam crossing is taken when considering particles from a collision. The required time resolutions limit
the particle separation capabilities of time of flight. TOF is effective in the region βγ < 3 and often used to
separate pions, kaons and protons.
dEdX Energy loss varies with βγ and hence can be used for particle identification. Especially the rise at
low velocities. The dEdX method is used to separate kaons and pions in the region βγ < 3. The relativistic
rise allows PID also for 100 < βγ < 1000.
Cherenkov Radiation Cherenkov radiation is emitted if the speed of a charged particle passing through
a medium with refractive index n is greater than the speed of light in that medium β = n1 . This leads to
threshold Cherenkov detectors, which are employed for PID in the range 3 < βγ < 14. The Cherenkov
1
light is emitted in a cone with an opening angle cos θC = nβ around the direction of the particle. This is
exploited in Ring Imaging Cherenkov Detectors (RICH). In a RICH the light of the emitted Cherenkov cone
is reflected and recorded as a projected ring. The diameter of the ring is proportional to the opening angle
θ. RICH detectors operate in the regime 14 < βγ < 140. Large scale neutrino detectors like Icecube, can
distinguish between electrons and muons, by the fact that the much lighter electrons scatter more and create
fuzzy edges on the recorded Cherenkov rings.
46
Transition Radiation Transition radiation detectors work similarly to threshold Cherenkov counters.
Transition radiation is emitted when particles above a certain velocity pass through the boundary of two
materials with different dielectric constant. Transition radiation detectors can be used up to βγ < 1000 and
usually help identify electrons.
3.4.5 Multi-purpose Detectors

Almost all general purpose detectors at colliders follow a typical shell structure as can be seen in Fig.3.5.
From the inside, closest to the collisions outwards we find:
• vertex and tracking detectors (inside a magnetic field)
• electromagnetic calorimeter
• hadronic calorimeter
• muon detectors
This setup allows measurement of momentum and energy of the particles produced and some basic object
identification to be made:
• photons deposit energy in the electromagnetic calorimeter, but leave no trace in the tracker
• electrons deposit energy in the electromagnetic calorimeter with an associated track in the tracker
• positrons the same as electrons, but the track is curved in the opposite direction
• neutral hadrons deposit energy in the hadronic calorimeter, but leave no track
• charged hadrons deposit energy in the hadronic calorimeter and leave a track
• jets quarks tend to produce jets of particles that contain many hadrons, they are identified by the
tracks and the energy deposits in the calorimeters
• b-jets jets originating from a bottom-quark contain longer lived b-hadrons. With a cτ ∼ 450µm they
decay a few mm away from the collision point, due to boost. B−jets can be identified by reconstructing
the displaced secondary vertex (a track crossing away from the collision point).
• muons muons can traverse the whole detector and are measured in the outer most muon spectrometer
• neutrinos escape the detector, they can be identified employing the fact that the sum of all momenta
in the plane transverse to the beam needs to add up to zero. Any “missing momentum” points to an
escaped particle!
Short lived particles can be identified on a statistical basis, by reconstructing the invariant mass of their
decay products. So for example a sample of Z-Bosons can be selected by finding oppositely charged muon
pairs with momenta and directions that result in an invariant mass around 91 GeV. The Higgs Boson was
discovered as an excess in the di-photon invariant mass around 125 GeV.
47
Figure 3.5: A cross section through the general purpose LHC detector ATLAS. The small circle at the bottom
is the beam pipe and the two colliding proton beams pass through the paper plane.
Trigger
The purpose of a trigger is to signal an interesting event. At the LHC a fast trigger system is essential to
cope with the data rate. Proton bunches collide 40 million times every second, yet only about 1000 events
can be stored in that time. In the first trigger stage, fast electronics signals for example a muon candidate
that passes through the muon chambers. Similarly events with large deposits in the calorimeter or with large
missing energy are selected. In later trigger stages the information from all sub detectors is readout, passed
to a computer and further reconstructed.
3.4.6 HEP Experiments at UCL

The UCL HEP group is involved in a large variety of particle physics experiments. Below is an overview of
current experiments.
• ATLAS, A Large Toroidal LHC Apparatus is a multi-purpose detector at the Large Hadron Collider.
Work at UCL includes measuring the newly discovered Higgs as well as searches for Dark Matter and
48
New Physics!
• SuperNemo, an upgrade of the Neutrino Ettore Majorana Observatory (NEMO experiment) looks
for neutrino-less double beta decay, discussed in section 2.1.4
• ANITA, the Antarctic Impulsive Transient Antenna is a balloon experiment looking for very high
energetic cosmic muons via radio waves emitted by neutrinos passing through the ice of the Arctic
• ARA Askaryan Radio Array looks for radio emissions of very high energy neutrinos interacting in the
Antarctic ice
• LUX, the Large Underground Xenon dark matter experiment is a direct dark-matter detection exper-
iment looking for recoils of nuclei being hit by a dark-matter particle
• LZ:LUX-Zeppelin is an extension and upgrade of the LUX and Zeppelin experiments

• g-2 is a precision (0.14 parts per million!) measurement of the muon anomalous magnetic moment,
sensitive to physics beyond the Standard Model
• mu2e a Muon-to-Electron Conversion Experiment at Fermilab near Chicago, looks for Lepton flavour
violation in muon decays
• mu3e at PSI in Switzerland, searches for lepton flavour violation through the forbidden µ+ → e+ e+ e−
decay
• MINOS, the Main Injector Neutrino Oscillation Search investigates neutrino oscillations in the muon
neutrino beam at FermiLab in Chicago. It consists of a near and far detector
• CHIPS, consists of CHerenkov detectors In mine PitS and is built inside a lake to measure neutrinos
at the far side of the FermiLab neutrino beam. It is a new concept to build large detectors cheaply (as
cheap as Chips)!
• Noνa, investigates in the Neutrino beam at the Main Injector Off-Axis νe Appearance. It is a neutrino
oscillation experiment.
• DUNE Deep Underground Neutrino Experiment at Fermilab plans to measure CP violation in neu-
trinos.
• AWAKE is investigating a new accelerator technology, using a plasma wake field.
• LUXE Laser Und X-ray free electron laser Experiment aims to investigate QED in extreme conditions,
e.g. strong fields.
• LEGEND Large Enriched Germanium Experiment for Neutrinoless ββ Decay, a follow up experiment
to SuperNEMO.
• SBND The Short-Baseline Near Detector will be one of three liquid argon neutrino detectors sitting
in the Booster Neutrino Beam (BNB) at Fermilab.
49
Chapter 4
The Strong Interaction
4.1 Introduction
The strong interaction is the strongest of the four fundamental forces. It is the interaction between quarks and
gluons and is responsible for holding atomic nuclei together. The strong interaction is described by a theory
termed Quantum ChromoDynamics (QCD). Before discussing some of the details of QCD, a comparison with
the simpler theory describing electromagnetism, termed Quantum ElectroDynamics (QED) will be given.
4.2 QCD versus QED

In QCD, interactions are mediated by gluons: spin-1, massless, electrically neutral bosons. These are analo-
gous to the photons in QED, which are also spin-1, massless, electrically neutral bosons. However we shall
see that gluons are self-interacting which leads to some fundamental differences between the theories. Colour
charge is the source of the strong force, just as electric charge is the source of the electromagnetic force.
Only particles with non-zero colour charge couple to gluons, so leptons do not “feel” the strong force. The
gluon coupling to colour charge is larger than the photon coupling to electric charge, and as we shall see in
Section 4.6 it gets stronger and stronger at longer distance scales.
4.3 Hadrons pre-1960s

After the discovery of the proton (1919) and the neutron (1932) a series of many more seemingly “funda-
mental” particles were discovered. These were the various hadrons (mesons and baryons) that are in fact
still being discovered today. It seemed quite a mess, although there were some underlying patterns in the
quantum numbers of the observed particles (see Sec. 2.2). In 1964 Gell-Mann and Zweig suggested that
the hadrons were in fact made up of constituent “quarks” with fractional charge. Their model was termed
the quark model and met some resistance until evidence for constituents of the proton was found in Deep
Inelastic Scattering experiments in 1968.
50
4.4 Deep Inelastic Scattering

4.4.1 Evidence for quarks
The basic idea of Deep Inelastic Scattering (DIS) is to scatter high-energy leptons (electrons, muons and
neutrinos have all been used) off nucleons (neutrons or protons). The directions and energies of the scattered
leptons depend on what they interact with within the nucleon, allowing us to “see” the nucleon structure at
very small distance scales. It is “deep” because the momentum transfer of the scattering is high, so distances
much smaller than the size of a nucleon (0.8 fm) are probed. It is “inelastic” because such high momentum
transfers lead to the break-up of the nucleon, so the final-state particles are not identical to the initial state
ones. It is a similar principle to Rutherford’s famous scattering experiment that allowed him to see the
structure of the atom.
In 1968 DIS experiments at SLAC led to the first evidence for the substructure of the proton. A 20 GeV
electron beam was fired at stationary protons, and detectors measured differential
cross-sections versus
Q2 = −q 2 , where q is the four-momentum transfer of the interaction. If k = E, ⃗k is the four-momentum

of the incoming electron and k ′ = E ′ , k⃗′ is the four-momentum of the outgoing electron, we can derive
2
an expression for q 2 = (k − k ′ ) in terms of the incoming and outgoing electron energies and the electron
scattering angle:
2 2
q 2 = (E − E ′ ) · (E − E ′ ) − ⃗kc − ⃗k ′ c · ⃗kc − ⃗k ′ c = E 2 + E ′2 − 2EE ′ − ⃗kc − ⃗k ′ c + 2⃗k · ⃗k ′ c2 .
If we neglect the electron mass compared to the large momenta, we have E = ⃗k c and E ′ = ⃗k ′ c and
therefore:
q 2 = 2EE ′ (cos θ − 1) .
You can see that the four-momentum squared is always ≤ 0, which may seem unintuitive but is perfectly
allowed. We refer to this type of interaction as space-like, as the momentum or space part of the four-vector
is larger, leading to a negative four-momentum squared.
The experiments made measurements of
Q2 = −q 2 = 2EE ′ (1 − cos θ) (4.1)
2
and found many more high-Q scatterings than expected for a model with diffuse charge throughout the
proton. Looking at Equation 4.1, more high-Q2 interaction means more large scattering angle interactions.
This is the sort of behaviour expected for a proton consisting of smaller constituents, and provided the first
evidence for the substructure of the proton. The constituents were named “partons” as they had not yet
been equated to the quarks in the proposed quark model. Note that parton is now used as a word to describe
both quarks and gluons, and we will see later that gluons are an important constituent of protons.
There have been many DIS experiments since this discovery, giving very precise data on the structure of
nucleons. The most precise comes from the HERA electron–proton collider in Hamburg, that ran from 1992
until 2007. It had a maximum electron beam energy of 27.5 GeV and proton beam energy of 920 GeV, and
hence a centre-of-mass energy of 318 GeV.
4.4.2 Proton momentum fraction

Figure 4.1 shows a diagram of the most likely type of interaction in electron–proton scattering, where a
virtual photon is exchanged between the electron and one of the partons within the proton. In addition to
51
Figure 4.1: Deep Inelastic Scattering: diagram of virtual photon exchange between a lepton and a proton.
The four-momenta of the interacting particles are labelled, as defined in the text.
Q2 , another very important Lorentz invariant quantity is defined as
−q 2
x= , (4.2)
2p.q
where p is the four-momentum of the incoming proton and q = k − k ′ , as defined above. The physical
interpretation of x is the fraction of the incoming proton’s momentum carried by the struck parton (defined
in any frame in which the proton has a very large momentum). Equation 4.2 can be derived by assuming
that the electron scatters off a massless parton carrying a fraction, x, of the proton’s momentum. In this
case p = (Ep , p⃗p c) is the four-momentum of the proton, xp = (xEp , x⃗ pp c) is the four-momentum of the
parton before interacting with the exchanged photon. Now, from the lectures on “Basic ideas in particle
2
physics” we know that the four-momentum squared is an invariant and therefore: (xp) = mc2 , where we
are considering the rest frame of the parton. Since we are considering the parton to be massless we have:
2
(xp) = 0. Similarly the four-momentum squared of the interacting constituent after the interaction is 0. If
we equate the four-momentum squared before and after the interaction on the lower vertex in Figure 4.1 we
obtain:
(xp + q) · (xp + q) = (mparton c2 )2 ∼ 0
which leads to:
2
(xp) + q 2 + 2xp.q = 0
2
and since (xp) = 0 we obtain Equation 4.2. x is an incredibly important variable. It gives us the fraction
of the proton’s momentum carried by a given parton within the proton, which tells us about the structure
of the proton. Detailed analysis of the DIS data showed that it is consistent with the partons being spin-1/2
particles with fractional charges: exactly as predicted by Gell-Mann and Zweig’s quark model.1
1 We shall see later that the partons also include neutral gluons within the proton.
52
4.4.3 Parton Distribution Functions

The fundamental insight of the DIS data is that the electron–proton cross-section, σep , can be described as
i
the product of the electron–quark cross-sections for each quark flavour i, σeq , with the Parton Distribution
Functions (PDFs) inside the proton:
X
i
σep (x) dx = σeq × fi (x) dx, (4.3)
quark flavours,i
where the PDFs, fi (x), are defined such that fi (x) dx is the number of partons of a certain type in the proton
with momentum fraction between x and x + dx. For example the up-quark PDF for the proton fup (x) is
defined such that fup (x) dx is the number of up-quarks within the proton with momentum fraction between
x and x + dx. For an electromagnetic interaction between the quark and electron (as shown in Figure 4.1)
the cross-section for each quark flavour is proportional to the square of the quark charge.
The PDFs are not known from first principles and must be measured experimentally. However, they are
universal, so can be measured at one collider (e.g. HERA) and used to predict cross sections at another
collider (e.g. the LHC). The actual measurements by the experiments at HERA were of the so called proton
structure functions, for example: X
F2 (x) = e2i xfi (x) ,
i
where ei is the charge of the quark. The PDFs can then be extracted from these measurements.
Scaling violations
In the simple parton or quark model we have two up-quarks and one down-quark in the proton. The structure
functions should not be dependent on the four-momentum transfer of the measurement, as the number of
partons is just three, irrespective of how energetically we probe (as long as the probe is energetic enough
to resolve the proton structure, which it is at HERA). The HERA (and other) experiments measured the
structure functions versus x and Q2 , for example see Figure 4.2. For x ∼ 0.1 there appears to be very little
dependence of the structure functions on Q2 , as expected. This is known as “Bjorken scaling” and implied
independence of the results on the resolution at which the proton was probed. The first measurements of
the structure functions were in this region and provided evidence for the parton model. However, in later
years, measurements were made at more and more x and Q2 values. For small values of x a clear increase of
the structure functions with Q2 can be observed in Figure 4.2: it appears that the number of partons in the
proton depends on how energetically it is probed. This phenomenon is known as “scaling violations”.
These scaling violations are understood to be due to the fact that the simple model in which a nucleon is
composed of three static quarks is not the entire story. In fact, the quarks within a hadron are continuously
interacting by exchanging gluons, which can split into a q q̄ pair momentarily then recombine to a gluon
again. This is known as a quantum fluctuation and is allowed provided the time and the implied violation of
energy conservation are compatible with the uncertainty principle. This is pictured in Figure 4.3. The single
purple blobs that are not associated with a green blob represent the three quarks that describe the quark
content of the baryons (e.g. uud for the proton). These are called the “valence quarks” of the baryon, and
they make up its quantum numbers (e.g. charge, strangeness etc). Emitted from these quarks are gluons
and the gluons can split into a quark–antiquark pair, shown as the green and purple blobs. These are called
the “sea quarks” of the baryon. Since they are always produced in a quark–antiquark pair, their charges etc
will cancel. In terms of the DIS data, this means that the electron could be interacting with one of the sea
quarks, and larger Q2 values mean that the structure of the proton is probed to a higher resolution and more
53
Figure 4.2: Proton structure function versus x for different Q2 values.
Figure 4.3: The parton content of a baryon.
sea quarks are seen. In general, the amount of gluon radiation from the quarks reshuffles the momenta and
introduces a Q2 dependence. A full QCD analysis of the structure of the nucleon successfully describes these
scaling violations to high accuracy.
54
PDF shapes
Lets take a moment to think about what the PDFs look like versus x. Figure 4.4 shows x multiplied by the
PDFs plotted versus x for all the quark PDFs combined: xQ(x) (Q here is not related to the four-momentum
transfer!). These PDFs will include contributions from both the valence and the sea quarks. It also shows
Figure 4.4: Proton PDFs.
the same thing for the antiquarks: xQ̄(x). Since the sea quarks are always produced in pairs, we would
expect xQ(x) = xQ̄(x). We can therefore determine the valence quark contribution by: x Q(x) − Q̄(x) .
This is also shown in the figure. We can see that the distribution for the valence quarks peaks at about
x = 0.2. The valence quarks dominate and carry most of the protons momentum, but at lower x values the
sea quarks become important. The results in this figureRreveal a very important point about the gluons in
1
the proton. If we integrate the sum of the two curves: 0 xQ(x) + xQ̄(x) dx, we might expect to obtain
1, which would indicate that the entire proton momentum is carried by the quarks. In fact we only obtain
roughly 50%. This tells us that 50% of the proton’s momentum is in fact carried by the gluons. While they
do not interact directly with the leptons in DIS experiments, the gluon PDFs can be directly determined at
the LHC. Figure 4.5 shows PDFs at two Q2 values. The up and down valence PDFs are shown (uv and dv )
as well as the sea quark and gluon PDFs. The latter two are divided by 10 so that they can fit onto the
plots. You can see that as Q2 increases the contribution from the sea quarks and the gluons get very large
at low x values. These regions are very important in LHC interactions, as we will see below.
LHC cross-sections
Analogous to Equation 4.3 for DIS, at the LHC cross-sections are determined by convoluting the PDFs with
the parton–parton cross-sections. As an example, consider the production of a Z boson at the LHC, i.e.
55
(a) (b)
Figure 4.5: Proton PDFs at different Q2 values.
pp → Z. The cross-section is given by:

XZ 1 Z 1
fi x1 , Q2 fj x2 , Q2 σij→Z dx1 dx2 ,

σpp→Z =
ij 0 0
where x1 is the momentum fraction carried by the interacting quark in the first proton, x2 is the momentum
fraction carried by the interacting quark in the second proton and fi,j x1 , Q2 is the PDF for a given quark
flavour. The integrals are over all possible x1 and x2 values and the sums are over all possible quark flavours
contributing to the interaction. If the Z boson is produced with a mass of 91 GeV/c2 , and the beam energies
of the protons are 6500 GeV, we can determine the x values of the interacting protons. For simplicity we
assume that x1 = x2 = x so that the momenta of the quarks exactly cancel2 and we write:
2
MZ2 c4 = (E1 + E2 ) · (E1 + E2 ) − (⃗
p1 c + p⃗2 c) · (⃗
p1 c + p⃗2 c) = (2 × x × Ebeam )
and
91
x= = 0.007
2 × 6500
where we are neglecting the masses of the interacting quarks. From Figure 4.5(b) we can see that in this
region the sea quarks dominate. So it becomes clear that sea quarks are very important in LHC interactions
(which is why processes like uū → Z can occur even though there are no valence antiquarks in the proton).
Gluons are also very important in LHC interactions as their contribution becomes very large at low-x.
2 Note however that this is not true in general, so that if quarks with different x values were to interact, the lab frame would
not be equal to the centre-of-mass frame for the interacting quarks (even if it is for the protons).
56
4.5 Colour
4.5.1 Colour saves the Pauli principle
According to the Pauli exclusion principle, the wavefunction of a system must be antisymmetric (i.e. must
change sign) under the exchange of two identical fermions. Consequently, identical fermions cannot be in the
same quantum state, meaning they cannot have identical quantum numbers (see a text book on Quantum
Mechanics such as “Quantum Mechanics”, Alastair I M Rae for a discussion on this). When the quark model
was first introduced, it appeared to be in violation with this fundamental principle of quantum mechanics.
In order to demonstrate this, consider the Ω− baryon, which consists of three strange quarks and has a spin
of 3/2, meaning that the spin of all three quarks are aligned. It appears that the three quarks are in the
same quantum state. If any two of the three quarks were exchanged, the wavefunction would be symmetric.
But this is in clear violation of the Pauli principle! In order to solve this apparent paradox, it was proposed
by Greenberg in 1964 that each quark has an additional intrinsic property termed “colour”. Just as two
electrons can occupy the same energy level of an atom as long as they have different spin values, quarks can
exist in a baryon with the same spin values as long as they have different colours. In Section 4.5.5 we will
see that it has been experimentally confirmed that there are three different colour states that the quarks can
be in.
4.5.2 Colour charges

Any quark can exist in one of the three colour states which are called red (r), green (g) and blue (b) and
any antiquark can exist in either the antired (r̄), antigreen (ḡ) or antiblue (b̄) states. These colour states are
independent of the flavour states (u, d, s, c, t, b). Each state is characterised by two conserved colour charges
(which are quantum numbers) labelled: I3C (colour isospin charge) and Y C (colour hypercharge). These are
the strong interaction analogues of the electric charge in electromagnetism. The values of the colour charges
depend on the colour state. The values of the colour charges are given in Table 4.1 for the three colour states
of quarks and the three colour states of antiquarks. These are additive quantum numbers.
I3C YC I3C YC
r 1/2 1/3 r̄ −1/2 −1/3
g −1/2 1/3 ḡ 1/2 −1/3
b 0 −2/3 b̄ 0 2/3
Table 4.1: Colour charges for quarks (left) and antiquarks (right)
4.5.3 Colour singlets

As we will see in Section 4.7, hadrons are only ever observed as colour singlets in which the total colour
charge is zero, i.e. I3C = Y C = 0. From Table 4.1 we can see that this can be achieved by combining a colour
with its anticolour, as is the case in mesons, or by combining a red, a green and a blue quark, as is the case
in baryons. The colour wavefunction for a meson is given by the following superposition of states:
1
ψ = √ rr̄ + gḡ + bb̄
3
57
where r indicates a quark in a red colour state and so on. This is clearly colourless. The colour wavefunction
for a baryon is given by the following superposition of states:
1
ψ = √ (r1 g2 b3 − g1 r2 b3 − r1 b2 g3 + g1 b2 r3 + b1 r2 g3 − b1 g2 r3 )
6
where r1 indicates that quark 1 is in the red colour state, and so on. You can easily check that by swapping
the indices of any two of the quarks (e.g. 1 ↔ 2) the wavefunction is antisymmetric as required by the Pauli
exclusion principle.
It is easy to see that it is only possible to produce colourless hadrons by including a quark–antiquark pair
(mesons), or a combination of three quarks or three antiquarks (baryons or antibaryons). Free quarks, or
combinations like qq or qq q̄ are not allowed. More unusual combinations like qqqq q̄, known as pentaquarks,
are colourless and are therefore allowed, however they have only been observed very recently. The first
evidence for pentaquarks was announced in July 2015, by the LHCb experiment at the LHC. Specifically
they discovered a uudcc̄ state.
4.5.4 Gluons and colour flow

By considering the conservation of colour charge at the vertex of a quark–quark scattering diagram via the
exchange of a gluon, it is clear that gluons themselves are manifestly colour charged and that they must
carry both a colour and an anti-colour (see the lecture slides for details).
4.5.5 Evidence for NC = 3

One way to determine the number of possible colour states for a quark, NC , is by realising that if a process
involves quarks in the final state, then the rate of that process occurring is really a sum of the possible quark
colour states. This means that the rate a process would have if there were no colour degree of freedom must
be multiplied by the number of colour states, NC .
There are a number of pieces of evidence for NC = 3. We will discuss one here, namely the ratio of
the cross-section for e+ e− → Hadrons to the cross-section for e+ e− → µ+ µ− (another are the W boson
branching ratios). We consider the annihilation of e+ e− to a virtual photon, where the photon decays to a
pair of opposite electric-charge sign fermions. We are working at a centre-of-mass energy of around 30 GeV,
low enough that photon exchange is dominant, high enough that the masses of all fermions apart from the
top quark are negligible. The Feynman diagram for e+ e− → µ+ µ− is shown in Figure 4.6(a). For e+ e− →
Hadrons we must consider e+ e− → q q̄ for all quark–antiquark final states, such as the Feynman diagram
shown in Figure 4.6(b). The possible q q̄ final states are the following:
¯ cc̄, ss̄, bb̄
uū, dd,
where we have not included the tt̄ final state as the centre-of-mass energy is not high enough. It is the sum
of these that gives us the e+ e− → Hadrons cross-section.
The coupling of these particles to the photon is proportional to the electrical charge, and therefore the
cross-section for each final state is proportional to the square of the electric charge. If we neglect the masses
of particles, the cross-section for each final state is the same. However, the quarks have an additional degree
of freedom as compared to the leptons, which is the colour degree of freedom. In effect this means there are a
factor of NC more ways the process can occur for each quark and so the cross-section for quarks is multiplied
58
e− µ− e− q
γ∗ γ∗
e+ (a) µ+ e+ (b) q̄
Figure 4.6: Feynman diagrams for (a) e+ e− → µ+ µ− and (b) e+ e− → q q̄
by NC . The ratio of the the cross-sections can therefore be found by:
¯ + σ(cc̄) + σ(ss̄) + σ(bb̄) 2 2 1 2 2 2 1 2 1 2

σ(uū) + σ(dd) 3 + 3 + 3 + 3 + 3 11NC
= NC × = .
σ(µ+ µ− ) 12 9
33
The experimental data shows that the ratio of cross-sections is very close to 9 , indicating that there are
three colour charge states.
4.6 Asymptotic freedom

Asymptotic freedom is the statement that the strong interaction gets weaker at high energies (and therefore
short distances). In order to explain this we need to first consider what we mean by the strength of an
interaction. We define the strength of an interaction in terms of the coupling of the force carrying particles
to the charge of the interaction (the photon to electrical charge in QED and the gluon to colour charge in
QCD). These are parameters that we measure experimentally by considering the rates of certain processes
such as those shown in Figure 4.7. Figure 4.7(a) shows electron–electron scattering via photon exchange,
the strength of the interaction for each vertex is given by the familiar electromagnetic coupling constant,
2
α = 4πϵe0 h̄c ∼ 137
1
. Figure 4.7(b) shows quark–quark scattering via gluon exchange, the strength of the
interaction for each vertex is given by the strong coupling constant αs ∼ 0.1. However, in turns out that
these coupling “constants” are not really constant at all. In fact they depend on the four-momentum transfer
of the process used to measure them! We will first consider the simpler case of electromagnetism. Consider
again the electron–electron scattering process. As well as the lowest order diagram shown in Figure 4.7(a),
there are many other digrams, involving more vertices, that must be considered. An example is shown in
Figure 4.8(a). A detailed calculation reveals that including these higher-order diagrams has the effect of
reducing the effective coupling by an amount that increases for lower momentum transfers, or consequently
1
larger distances. It turns out that α ∼ 137 is only valid at momentum transfers approaching zero. The
1
coupling increases to α ∼ 129 for a large momentum transfer of 100 GeV. While we do not provide a
detailed derivation of this so-called “running” of the coupling constant in this course, it can be understood
by considering the charge of an electron (or any charged particle). Imagine measuring the charge of an
electron by placing another charge a distance from it and measuring the force acting on it. The electron will
be continuously emitting and reabsorbing photons (quantum fluctuations). The photon itself can in turn
59
e− e− q q
γ∗ gluon
e− (a)
e− q (b) q
Figure 4.7: Feynman diagrams for (a) e− e− and (b) qq scattering at the lowest order.
e− e− q q q q
γ∗
e− e+ q q̄
γ∗
e− (a)
e− q (b) q q (c) q
Figure 4.8: Feynman diagrams for (a) e− e− scattering with a fermion loop, (b) qq scattering with a fermion
loop and (c) qq scattering with a gluon loop.
fluctuate momentarily into e+ e− pairs. The positrons in these pairs will be attracted to the original electron
and the electrons will be repelled, making the positrons closer on average. This has the effect of shielding
(and reducing) the electron charge by an amount dependent on the distance from the charge (the smaller
the distance, the less the shielding and the closer to the true “bare” electron charge the effective measured
charge will be). This effect is termed the “vacuum polarisation effect” due to similarity to the polarisation
of a dielectric medium by a charge (here the polar molecules take the role of our e+ e− pairs). Note that this
is a separate effect to the fall off of the electromagnetic force with distance due to the inverse square law,
due to the form of the Coulomb potential.
Now let’s turn to QCD. Just like for QED, the quark–quark scattering process shown in Figure 4.7(b)
can also occur via different diagrams including loops. There is a diagram analogous to that in Figure 4.8(a)
containing a quark loop in place of the electron–positron loop. This is shown in Figure 4.8(b). This type of
diagram has a similar screening effect to that in Figure 4.8(a), effectively reducing the size of the coupling
at lower momentum transfers or larger distances. Due to the fact that gluons carry colour charge, there is
another diagram with a gluon loop instead of a quark loop, as shown in Figure 4.8(c). It turns out that this
60
type of diagram has the opposite effect and causes an antiscreening, where the size of the coupling increases
at lower momentum transfers or larger distances. The dependence of the strong coupling constant on Q2
depends on the combination of these two types of diagram and is given approximately by:

2
αs µ2
αs Q = 2 , (4.4)
1 + Bαs (µ2 ) ln Qµ2
33−2N 2
where B = 12π f , Nf is the number of quark flavours with 2Mq c2 < Q2 , and µ is some reference four-
momentum value at which αs is measured.3 Equation 4.4 tells us how αs varies but the reference value must
be measured experimentally. In fact αs has been measured at various Q2 values and the dependence with
Q2 agrees exceptionally well with the QCD predictions.
We can see that for 6 quark flavours 33 − 2Nf is a positive number, which leads to a decrease of αs with
Q2 . And we can see that the larger the number of quark flavours (and hence more corrections of the form
in Figure 4.8(b)) the smaller this number. In a world where 2Nf > 33, αs would increase with Q2 , just as α
does in QED.
The strong coupling constant gets larger and larger at lower energies or larger distances. In this region
it is not possible to do reliable calculations in QCD. This is because if the coupling is not a small number,
it is not possible to calculate the effect of the infinitely possible ways a certain process can occur with many
additional vertices involving the strong interaction. We are in a region where perturbation theory does not
work and QCD essentially breaks down. We often have to rely on empirical models to describe the data.
4.7 Colour confinement

As we pointed out in Section 4.5, we never observe free coloured objects in nature. The coloured quarks
and gluons are confined to colour-singlet hadrons. It has been postulated (and so far this is backed up by
experiment) that the strong interaction is confining, making it impossible for coloured objects to be free.
This is described by an empirical model as we are in a region where QCD is not predictive (as we cannot
apply perturbation theory at low energies). However the idea is that the strong interaction continues to get
stronger and stronger at larger and larger distances. This leads to a potential energy between two coloured
objects that increases as the distance between them increases, leading to a constant force between them. The
potential energy is described by the form:
αs h̄c
V (r) = − + λr, (4.5)
r
where λ is a constant experimentally determined to be λ ∼ 1 GeV fm−1 . At r ≤ 0.1 fm interactions are
thought to be dominated by one-gluon exchange and we see the familiar Coulomb-like potential (the first
term on the right hand side of Equation 4.5). At larger distances, one-gluon exchange no longer dominates.
It is believed that the exchanged gluons between coloured objects are attracted together and squeezed into a
tube, with the stored energy proportional to the distance between them (the second term on the right-hand
side of Equation 4.5). Imagine two quarks being pulled apart. As they separate the potential energy between
them increases linearly and it would require an infinite amount of energy to separate them to infinity. As a
result, coloured objects arrange themselves into bound colourless hadronic states and quarks and gluons are
always bound to hadrons.
3 Note that some text books have a simpler form for Equation 4.4 which makes some approximations.
61
4.8 Jets
4.8.1 Quark jets
As we have seen, quarks are never observed in isolation. In scattering processes involving quarks in the
final state, what we actually observe experimentally are streams of collinear hadrons known as “jets”. For
example, at an e+ e− collider, in the process e+ e− → q q̄ (which could go via an intermediate virtual photon
or Z boson), the high-energy quarks are produced travelling back-to-back in the centre-of-mass frame. They
move apart from each other at high velocities. As they move apart the force from the colour field between
them is constant so the energy in the colour field increases linearly with the distance they get pulled apart.
At some point it is energetically more favourable for another quark–antiquark pair to form in between them.
This process continues and more quark–antiquark pairs are formed. Many hadrons are formed so that two
jets of hadrons travelling back-to-back in the directions of the original quark and antiquark are produced. At
high enough energies, the relative momenta of the hadrons in jets is small compared to the momenta of the
original quarks. In this case, to a good approximation, the sum of the momenta of the constituent hadrons
within a jet of particles is equal to the momentum of the original quark from the interaction. Thus, when we
measure the properties of jets we are really studying the properties of the original quarks. The first jets were
observed in 1975 at the SPEAR e+ e− collider that had a centre-of-mass energy ranging between 3 GeV and
7.4 GeV. The angular distributions of the two-jet events were in agreement with that expected from spin-1/2
quarks.
4.8.2 Gluon jets

While the majority of jet events in e+ e− events involve two jets, occasionally there are events with three
distinct jets. These come from events where a high momentum, non-collinear gluon is emitted by one of the
quarks, as shown in Figure 4.9, where the gluon forms a distinct jet. The first three jet event was observed in
1979 by the TASSO experiment at the PETRA e+ e− collider, which had a centre-of-mass energy of 30 GeV.
Studies of the angular distributions of these events were consistent with a spin-1 particle. This was the first
direct evidence for the gluon.
e− q
γ∗
e+ q̄
Figure 4.9: Feynman diagram for e+ e− → q q̄g.
62
4.9 QCD today

Today QCD is used very successfully to predict the rate of interactions at colliders. At the LHC a typical
interaction is very complicated as there is QCD radiation from the initial state partons as well as possible
final state partons. At high-momentum it is incredibly successful as predicting the rates of jets. But at low
momentum, where perturbation theory does not work, it cannot be used. Instead empirical models are used
to model processes, with parameters obtained from fits to the data.
While as far as we know quarks are the fundamental particles of the strong interaction, we are searching
for possible quark substructure at the LHC. So far no evidence has been found.
63
Chapter 5
The Electroweak Interaction: Part I
Dr Emily Nurse 2017
5.1 Introduction
In this module we will discuss the electroweak interaction, which is a merged theory describing both the
theory of electromagnetic interactions and the weak interaction of nuclear β decay. We will learn that the
separate weak and electromagnetic interactions are low energy manifestations of the electroweak theory. We
will go on to discuss the Higgs mechanism that provides mass to the fundamental particles, and the associated
Higgs boson, with particular emphasis on its recent discovery at the LHC.
Before all this we will learn about some unexpected asymmetries (parity and charge conjugation violation)
that were observed in the weak interaction in the 1950s, which were eventually explained by the spin structure
in the weak interaction. This will lead to a discussion of another subtle asymmetry in the weak interaction,
that is still being understood today (CP violation).
5.2 Symmetries and conservation laws

As described by the famous theorem of mathematician Emmy Noether, conservation laws in physics come
from symmetries seen in nature. For example, conservation of momentum comes from a symmetry of the
laws of physics under a spatial translation. Or put another way, from the fact that the laws of physics do not
change if we perform a spatial translation (e.g. you would get the same results if you performed an experiment
in London or in Chicago). Similarly, the conservation of angular momentum comes from symmetry of physical
laws under a rotation and the conservation of energy comes from symmetry of physical laws under a translation
in time (you would get the same results if you performed an experiment today or this time next year). These
are remarkable results that can be derived from Quantum Mechanics (see e.g. Section 4 in “Particle Physics”,
B.R. Martin and G. Shaw.) Richard Feynman discusses this principle in his usual accessible and enlightening
way in the Feynman lectures: http:www.feynmanlectures.caltech.edu/I_52.html
64
5.2.1 Parity
There is a conserved quantity known as parity that comes from a symmetry under space-inversion1 . This
means that the physical laws of a system do not change if the spatial co-ordinates are inverted: x → −x,
y → −y and z → −z, which is equivalent to a mirror reflection (x → −x and y → −y) together with a
rotation through 180◦ (z → −z). We can establish the effect of space-inversion on the following kinematic
properties of a system:
1. The position vector ⃗r changes to −⃗r.
2. The momentum vector p⃗ changes to −⃗
p.
⃗ (= ⃗r × p⃗) remains unchanged as both ⃗r and p⃗ change sign.
3. The angular momentum vector L
Parity is a discrete conserved quantum number and is equal to the eigenvalue of any wavefunction that is an
eigenstate of the space-inversion transformation, P̂ . We define the intrinsic parity of a particle by considering
the space-inversion operator acting on a particle at rest:
P̂ ψa = Pa ψa , (5.1)
where Pa is the particle’s intrinsic parity. Two successive space-inversion transformations take us back to
where we started:
P̂ P̂ ψa = P̂ Pa ψa = Pa2 ψa = ψa , (5.2)
so we can deduce that Pa = ±1. Intrinsic parities of particles are generally determined by experiment,
assuming that parity is conserved in interactions, together with the fact that a fermion and its antiparticle
must have opposite parities and a boson and its antiparticle must have the same parity.2 By considering a
multiparticle state with a wavefunction that is the product of single-particle wavefunctions3 , it is clear that
parity must be a multiplicative quantum number and the total parity of a system of particles a, b, ... with
L
orbital angular momentum is given by: Pa Pb ... (−1) . The contribution to parity from the orbital angular
momentum comes from the consideration of the form of a wavefunction of a particle with definite angular
momentum and noting the fact that in spherical polar coordinates a space-inversion transformation implies:
r → r, θ → π − θ, ϕ → π + ϕ.4
Conservation of parity in an interaction can be considered in one of two ways, which amount to the same
question:
1. Is the interaction invariant under the space-inversion transformation?
2. Is the total parity before the interaction equal to the total parity after the interaction?
Until the late 1950s it was believed, as we might intuitively expect, that all physical laws were invariant
under space-inversion and therefore conserved parity.
1 Often “parity” is used to mean the symmetry as well as the conserved quantity. I prefer to call the symmetry space-inversion
to avoid this confusion (note that in the previous examples this would be like e.g. energy and time having the same name).
2 If interested, a proof of why this is in e.g. “Modern Particle Physics”, M. Thomson.
3 This is only true if the particles are distinguishable.
4 For details see section 1.3.1 in “Nuclear and Particle Physics”, B.R. Martin.
65
The τ − θ puzzle
In 1949 Powell discovered two weakly decaying particles in cosmic ray showers that were known as the θ+
and τ + mesons. They had the same mass, charge and spin, so they looked much like the same particle. They
decayed weakly to two different final states: θ+ → π + π 0 and τ + → π + π 0 π 0 . While it is perfectly fine for one
particle to decay via different decay modes, these two final states have different parities. In order to see this
we use the experimentally determined fact that the intrinsic parity of both the charged and neutral pions
is −1, together with the fact that the orbital angular momentum in both cases is zero.5 The two particles
cannot then be the same particle if parity is conserved in the decay.
Inspired by this puzzle, in 1956 Lee and Yang surveyed the experimental evidence and suggested that
the weak interaction does not necessarily conserve parity (even though it had been a firm assumption by
physicists until this time).
Parity violation in β decay

The first evidence for parity violation in the weak interaction came one year later in 1957 by Wu and
collaborators. She studied the β decay of polarised Cobalt-60: 60 Co → 60 Ni* + e− + ν̄e . In this reaction
a neutron from within the 60 Co nucleus is converted into a proton and an electron and antineutrino are
emitted. Polarised means that the 60 Co spins are all aligned. Wu measured the direction of the emitted
electrons with respect to the 60 Co spin direction and found that they were more likely to be travelling
in the direction opposite the nuclear spin. What does this tell us? Consider what a space-inversion (or
parity transformation) of this reaction would do. It would leave the spin of the nucleus unchanged (since
spin is an angular momentum), but it would reverse the direction of the emitted electrons. Space-inversion
symmetry (and consequently parity conversion) of this process would then imply that the electrons should be
emitted with equal probability aligned and anti-aligned with the nuclear spin. It is clear then that the results
of Wu’s experiments proved that parity violation exists in the weak interaction. This was a monumental
breakthrough in the understanding of the weak interaction.
These results solved the τ − θ puzzle. What we now know is that the θ+ and τ + mesons are in fact the
same particle (known as the charged kaon: K + ) which can decay to either π + π 0 or π + π 0 π 0 since parity is
violated in these weak decays. Note that the strong and electromagnetic interactions do conserve parity.
5.2.2 Charge conjugation

Charge conjugation is another important discrete symmetry. A transformation under the charge conjugation
operator, Ĉ, changes a particle, a, into an antiparticle, ā, together with a phase factor6 , Ca , specific to that
particle:
Ĉψa = Ca ψā . (5.3)
We can see from Equation 5.3 that ψa is only a eigenstate of Ĉ if the particle is its own antiparticle: a = ā,
e.g. for the photon and the neutral pion. In these cases the eigenvalue, Ca , is known as the intrinsic C-parity
of the particle. If we apply the operator twice we should be back where we began, so if a = ā we obtain:
Ĉ Ĉψa = ĈCa ψa = Ca2 ψa = ψa leading to Ca = ±1. For interactions involving particles that are eigenstates
of the charge conjugation operator, the total C-parity before and after the interaction must be conserved
5 Ifinterested you can see a discussion of why this is in e.g. Section 6.6.1 of “Nuclear and Particle Physics”, B.R. Martin.
6A phase factor in quantum mechanics is defined as eiθ , where θ is known as the phase. The absolute value of a phase factor
is 1. A phase factor therefore does not have any affect on probabilities.
66
if the interaction is symmetric under a charge conjugation transformation. Said another way: the total C-
parity of an interaction is conserved if the interaction is symmetric under particle ↔ antiparticle exchange.
The C-parity of particles can be measured by assuming charge conjugation conservation in interactions. For
particles that are not their own antiparticles eigenstates of charge
conjugation can only be formed by linear
combinations of a particle–antiparticle pair, e.g. √12 K 0 + K¯0 where K 0 = ds̄ and K¯0 = sd. ¯ It is worth
noting that in the case that a ̸= ā the phase factor Ca in Equation 5.3 cannot be measured and can therefore
be set to 1. Just like parity, C-parity is a multiplicative quantum number. It turns out, as we shall see in the
next section, that the weak interaction is not invariant under a charge conjugation transformation, however
both the electromagnetic and strong interactions are.
Parity and charge conjugation violation in muon decay

In 1957 various experiments measured the angular distribution of electrons (positrons) from the decay of
polarised (anti)muons: µ− → e− ν̄e νµ or µ+ → e+ νe ν̄µ . These measurements showed it to be of the form:

1 1
Re± (θ) = Γµ± 1 ± cos θ (5.4)
2 3
where Re− (θ) is the rate of electrons from muon decay as a function of the angle between the muon spin and
the electron direction, θ, Re+ (θ) is the same thing for positrons in antimuon decay, Γµ− is the muon decay
rate constant and Γµ+ is the antimuon decay rate constant. The first thing to note from Equation 5.4 is the
difference between the form of Re− (θ) and Re+ (θ), seen immediately from the sign in front of the 13 cos θ
term (positive for e+ and negative for e− ). If charge conjugation were conserved, we would expect the form
of the two to be identical (as a charge conjugation transformation would just change the µ− decay into a µ+
decay and vice versa). By inspecting Equation 5.4 we can also see that parity is violated in this decay. The
parity transformation implies that θ → π − θ and ϕ → π + ϕ in spherical polar coordinates. If we consider
the muon spin to be aligned with the z-axis so that for a parity transformation, θ → π − θ in Equation 5.4,
such that 1 + 31 cos θ → 1 − 13 cos θ and vice versa. We therefore conclude that parity is also violated in muon
and antimuon decay.
However, an important thing to note is that if we consider a charge conjugation transformation (C)
together with a parity transformation (P), we are back where we started, as the sign in front of 13 cos θ
changes twice. This is telling us that even though the weak interaction violates C and P symmetries, it is
invariant under a combined CP transformation. While this is true to a good approximation, we will see later
that CP violation does exist at a small level in the weak interaction, and this is intimately linked to the
dominance of matter over antimatter in our Universe. We should also reiterate here that the individual C
and P symmetries are conserved in the electromagnetic and strong interactions.
5.2.3 The spin structure of weak interactions

The C and P violation that was observed experimentally in the 1950s is now explained within the Standard
Model by the spin structure of weak interactions. In order to understand this we will first introduce helicity,
which is defined as the projection of a particle’s spin on its direction of motion: H = ⃗s|⃗p.⃗p| , where ⃗s is the
particle spin and p⃗ is the particle momentum. Recall that the projection of a fermion’s spin on any chosen axis
can take two possible values: + 12 or − 12 . Therefore, a fermion can either have a positive helicity: H = + 21 or
a negative helicity: H = − 12 . A negative helicity means that the particle is left-handed. A positive helicity
means it is right-handed. It turns out that only left-handed neutrinos and right-handed antineutrinos interact
67
weakly. This is the origin of the observed parity violations in weak interactions, as we shall see. And in fact,
since neutrinos only interact via the weak interaction, it turns out that only left-handed neutrinos (νL ) and
right-handed antineutrinos (ν̄R ) are observed in nature! This was proved in 1958 when Goldhaber measured
the helicity of the neutrino (https://doi.org/10.1103/PhysRev.109.1015).
For any particle with non-negligible mass, such that v < c, it is always possible for an observer to
travel faster and overtake the particle. A left-handed particle would then appear right-handed. Therefore,
for fermions other than neutrinos we cannot say that they are 100% left- or right-handed. For relativistic
2 2
fermions with mass, the forbidden helicity states are suppressed by an approximate factor: ∼ mc E , where
m is the fermion mass and E is the fermion energy. This is clearly a small number for relativistic fermions.
We can see that a parity transformation on a left-handed neutrino brings us to a right-handed neutrino
(the momentum changes direction and the spin remains unchanged). Since right-handed neutrinos do not
exist it is clear that parity is violated. Similarly, charge conjugation transformation on a left-handed neutrino
turns it into a left-handed antineutrino, which also does not exist, demonstrating that charge conjugation
is also violated. However a combined CP transformation would change a left-handed neutrino into a right-
handed antineutrino, which does exist, so it appears CP can be conserved. We will see in the next sections
explicitly how the lack of the existence of right-handed neutrinos and left-handed antineutrinos led to the
observations in β-decay and muon-decay that we have discussed.
β decay revisited
We now revisit the parity violation observed in β decay for Cobalt-60, where we saw that electrons were
preferentially emitted in the direction opposite to the direction of the nuclear spin. 60 Co has a spin of 5
and 60 Ni* has a spin of 4. In order to conserve angular momentum, the spin of the produced electron and
antineutrino (both spin-1/2) must both be aligned with the nuclear spin. In order to conserve momentum
the electron and antineutrino will be travelling in opposite directions. In the case that the antineutrino is
in the direction of the nuclear spin and the electron is in the opposite direction, the antineutrino will be
right-handed and the electron will be left-handed in the massless limit, which is perfectly allowed. However,
the case where the electron is in the direction of the nuclear spin, and the antineutrino is in the opposite
direction, we are dealing with a left-handed antineutrino and a right-handed electron in the massless limit.
This configuration is forbidden by the spin structure of the weak interaction and leads to the observed
experimental results, where the electron is preferentially emitted in the direction opposite the nuclear spin.
Muon decay revisited

We now revisit the parity violation observed in muon decay. This is a three body decay so the orientation
of the particles is not as simple, but if we look at the decays with the highest possible electron (or positron)
energies, then we are in a configuration where the electron (positron) is travelling back-to-back with the
neutrino and the antineutrino, which would then travel very close to each other. In this case, ensuring
momentum conservation, the energy of the electron/positron will be Ee ∼ mµ /2. If we consider the two
possible cases where: (1) the electron travels in the direction opposite the spin of the decaying muon and (2)
the electron travels in the same direction of the decaying spin, we can convince ourselves that there is only
one possible spin alignment for each that does not involve either a right-handed neutrino or a left-handed
antineutrino. In order to conserve angular momentum, in case (1) the electron must have negative helicity
and in case (2) the electron must have positive helicity. In the relativistic limit case (2) is suppressed.
Since the mass of the electron is very small compared to the decaying muon, the relativistic limit is a good
approximation, and we come to the experimentally observed conclusion that electrons are preferentially
68
emitted in the direction opposite the muon spin. If we work through the same logic for antimuon decay, we
reach the opposite conclusion for the direction of the positrons, just as was observed experimentally.
Pion decay
Consider the charged pion decay process:
π + → ℓ+ + νℓ
where ℓ+ is a positron or an antimuon. Ignoring helicity suppression, we would expect these rates to be
similar, due to lepton universality (e.g. the coupling of the W + to the e+ νe is the same as that to the µ+ νµ ).
There are some expected differences in the rates due to kinematic effects. The smaller mass of the electron
with respect to the muon results in it having a larger momentum, and due to the density of states factor of
the reaction rate, we would expect the rate to be just over two times that for the muons. Experimentally,
however, the branching ratio to µ+ νµ is (99.98770 ± 0.00004)% and that to e+ νe is (0.01230 ± 0.00004)%.
This huge difference in the branching ratios can be understood by helicity suppression resulting from the
spin structure of the weak interaction.
If you consider the process in the rest frame of the decaying π + , the ℓ+ and the νℓ would be produced back-
to-back, travelling in opposite directions in order to conserve momentum. Now, the π + is a spin-0 particle
and the ℓ+ and the νℓ fermions are both spin-1/2 particles. So, in order to conserve angular momentum, the
spins of the two particles must be anti-aligned. This gives two possible scenarios:
1. a left-handed νℓ and a left-handed ℓ+
2. a right-handed νℓ and a right-handed ℓ+

We know that 2. is forbidden, as right-handed neutrinos are not observed in nature. So the only way the
process can occur is via 1. But this involves a left-handed antiparticle, the ℓ+ . If the ℓ+ were massless,
this decay would not be possible. Since the charged leptons are not massless the decay is possible, but it is
2 2
suppressed by the factor mc E , where m is the charged lepton mass and E is its energy. We can determine
the charged lepton’s energy in the rest frame of the π by conserving momentum and energy before and after
the decay to give: mπ c2 = Eℓ + Eν and |pℓ |c = |pν |c. The neutrino mass is negligible compared to the
2
energies giving: Eν = |pν |c = |pℓ |c. Now write the expression: Eℓ2 = p2ℓ c2 + m2ℓ c4 = mπ c2 − Eℓ + m2ℓ c4 =
Eℓ2 + m2π c4 − 2mπ c2 Eℓ + m2ℓ c4 and rearrange to give:

m2ℓ + m2π c2
Eℓ = .
2mπ
2
2mπ mℓ
This leads to a suppression factor of m 2 +m2 . Using mπ = 139.57 MeV/c2 , mµ = 105.7 MeV/c2 and
ℓ π
me = 0.51 MeV/c2 we can straight away see that the muon is not highly relativistic in this decay, but the
2
me (m2 +m2 )
electron is. The ratio of the suppression factor for electrons to that for muons is mµ (mµ2 +mπ2 ) = 5.7×10−5 .
e π
A detailed calculation, combined with the density of states factor mentioned above, leads to a ratio that is
in very good agreement with the experimentally determined branching ratios.
69
Chapter 6
Electroweak Interactions: Part II:

Quark Mixing
6.1 Introduction
In section 2.2 the different quark flavours u, d, s, c, b, t were introduced. The strong and electromagnetic
interactions conserve flavour, while W boson exchange of the weak interaction changes flavour. That is
why flavour transitions in general involve a change of electric charge (the W ± boson carries electric charge).
Both leptons and quarks come in three generations of doublets (see Eq. 2.1 and 2.35). For quarks to first
order the flavour transitions via W boson exchange couple an “up-type” quark (u, c, t) with electric charge
+2/3 to a “down-type” (d, s, b) quark with electric charge −1/3. Couplings to the anti-quarks are as always
implied. The exchange of a weak Z Boson preserves flavour and so called neutral flavour changing currents
are suppressed. This can be easily seen from the associated Feynman diagrams.
6.2 Quark Mixing

6.2.1 Evidence
The decays of hadrons showed that for quarks a mixing across generations is possible. For example kaon
decays (K − → µ− ν¯µ , K 0 → π + π − ) show a transition from the second generation into the first, a s → u
transition. Experiments found, however, that the decay across generations is suppressed with respect to
decays within a doublet of a generation.
6.2.2 The Cabbibo Angle and GIM mechanism

Niccolo Cabbibo found an elegant solution to the problem. At the time only three quarks (u, d, s) where
known. He proposed that the weak eigenstates of the two known “down-type” quarks (d′ s′ ) were actually a
linear superposition of the quark mass eigenstates (ds):
′
d cos θC sin θC d
= (6.1)
s′ − sin θC cos θC s
70
Figure 6.1: An application of the GIM mechanism. The two diagrams pick up an opposite sign and cancel
out.
d′ = +d cos θC + s sin θC
′
s = −d sin θC + s cos θC
As for the neutrino mixing we can think of this as a rotation into a different basis using a rotation matrix.
That “down-type” quarks are rotated and not “up-type” is pure convention, as Cabbibo’s solution also helped
to solve another problem. Certain kaon decays were not observed. These decays proceeded through both an
u − s and s − d transition. The suppression of these decays could be explained by a cancellation if a fourth
quark was involved (with a resulting u − c and c − d transition). This was proposed in 1970 by Sheldon Lee
Glashow, John Iliopoulos and Luciano Maiani and is known as the GIM mechanism. The charm quark was
eventually discovered in 1974.
6.2.3 The CKM mixing matrix

The Cabbibo-Kobayashi-Maskawa (CKM) matrix is an extension of the mixing formalism to three families
or generations.
 ′    
d Vud Vus Vub d
 s′  =  Vcd Vcs Vcb   s  (6.2)
b′ Vtd Vts Vtb b
Unitary conditions
∗ ∗
Vtd∗
    
Vud Vcd Vud Vus Vub 1 0 0
† † ∗ ∗
VCKM VCKM = VCKM VCKM =  Vus Vcs Vts∗   Vcd Vcs Vcb  =  0 1 0  (6.3)
∗
Vub Vcb∗ Vtb∗ Vtd Vts Vtb 0 0 1
The matrix equation 6.3 leads to 9 constraints. In particular the diagonals of the identity matrix yield:
|Vud |2 + |Vus |2 + |Vub |2 = 1 (6.4)
2 2 2
|Vcd | + |Vcs | + |Vcb | = 1 (6.5)
2 2 2
|Vtd | + |Vts | + |Vtb | = 1 (6.6)
71
This guarantees that the total size of the coupling is kept the same. The zeros in the identity matrix lead
to constrains that can be expressed as a triangle in the complex plane. For example the bottom left corner
reads:
∗
Vud Vub + Vcd Vcb∗ + Vtd Vtb∗ = 0 (6.7)
Equation 6.7 can be visualised as the triangle in the complex plane as seen in Fig. 6.2.
A popular parametrisation of the CKM matrix was given by Wolfenstein:
2
1 − λ2 Aλ3 (ρ − iη)
 
λ
2
 −λ 1 − λ2 Aλ2  (6.8)
Aλ3 (1 − ρ − iη) −Aλ2 1
Figure 6.2: The unitarity triangle in the complex plane, visualising equation 6.7.
Where λ = sin θC ∼ 0.2 and A ∼ 0.8. It can be seen, that transitions within the same generation are
favoured, while transitions across generations are highly suppressed. Or in other words Vud ∼ Vcs ∼ Vtb ∼
O(1) ≫ Vus ∼ Vcd ∼ Vcb ∼ Vts ≫ Vub ∼ Vtd . The 3x3 case with three generations is special as it is the
smallest number of generations where the mixing parameters contain a non-removable complex phase. All
possible transitions between quark flavours are summarised in Fig. 6.3. Each transition amplitude has a
vertex factor of the corresponding CKM matrix element. We can see that for example charm decays into a
down quark are less likely than those into a strange quark. The ratio of these decays is given by ratio of the
squared amplitudes |Vcd |2 /|Vcs |2 = | − λ|2 /|1 − λ2 |2 ∼ |0.2|2 /|1 − 0.04|2 ∼ 0.04.
6.2.4 Kaon states and CP violation

The neutral kaon K 0 , a s̄d state, and its anti-particle the K¯0 , a sd¯ state, form an interesting system. Charge
conjugation turns each particle into its anti-particle, hence neither K 0 nor K¯0 are eigenstates of the charge
conjugation operator Ĉ nor the combined operator of parity and charge conjugation operator CP ˆ :
Ĉ|K 0 ⟩ = −|K¯0 ⟩; Ĉ|K¯0 ⟩ = −|K 0 ⟩ (6.9)

0
P̂ |K ⟩ = −|K 0 ⟩; P̂ |K¯0 ⟩ = −|K¯0 ⟩ (6.10)
ˆ
→ CP |K 0 ⟩ = +|K¯0 ⟩; CP
ˆ |K¯0 ⟩ = +|K 0 ⟩ (6.11)
72
Figure 6.3: The possible transitions and their vertex factor.
Using the usual conventions. It turns out that a process involving the so called box diagram shown in
Fig.6.4 can turn a K 0 into a K¯0 and vice versa. This means that the observed particles can be superpositions
of K 0 and K¯0 states. We can chose these superpositions to be CP ˆ eigenstates as shown in Eq. 6.15. The
signs become obvious, if we note, that CP simply turns a K into a K¯0 and vice versa.
ˆ 0
Figure 6.4: The box diagram exchanging two W bosons that convert a K 0 into a K¯0 and vice versa.
73
Figure 6.5: The Feynman diagram for the decay K 0 → ππ.
Figure 6.6: The Feynman diagram for the decay K 0 → πππ.
1
√ |K 0 ⟩ + |K¯0 ⟩
|K10 ⟩ =

(6.12)
2
1
|K2 ⟩ = √ |K 0 ⟩ − |K¯0 ⟩
0

(6.13)
2
1
ˆ |K20 ⟩ = √ CP

ˆ |K¯0 ⟩ = √1 |K¯0 ⟩ − |K 0 ⟩
ˆ |K 0 ⟩ − CP

CP (6.14)
2 2
ˆ |K10 ⟩ = +|K10 ⟩; CP
→ CP ˆ |K20 ⟩ = −|K20 ⟩ (6.15)
Let us now look at possible decays of these two states K10 and K20 . Both K 0 and K¯0 can decay into ππ
ˆ even: CP
states as shown in Fig. 6.5. Such two pion states are CP ˆ (ππ) = +1.
CP (π 0 π 0 ) = (Pπ0 )2 (−1)L (Cπ0 )2 = (−1)2 (−1)0 (1)2 = +1 (6.16)

Where we have used the parity (P (π 0 ) = −1) and charge conjugation C(π 0 ) = 1 values for the π 0 as derived
in the θ − τ puzzle. For the π + π − state one can also use the observation that both P̂ and Ĉ interchange the
π − with the π + and hence the combination CP ˆ leaves the π − π + state unchanged.
74
ˆ is conserved on the CP
If CP ˆ even K 0 state should be able to decay into two pions. The CP ˆ odd K 0
1 2
ˆ odd final state. A state formed of three pions, πππ, is such a state.
state will have to decay into a CP
CP (π 0 π 0 π 0 ) = (Pπ0 )3 (−1)L (Cπ0 )3 = (−1)3 (−1)0 (1)3 = −1 (6.17)
The decay into πππ is suppressed by phase space (Q = m(K 0 ) − 3 × m(π) ∼ 498 MeV − 3 × (135 MeV) =
94 MeV) (see Fig. 6.6). We hence expect the K20 state to live longer than the K10 state. The observed states
0
are aptly named KL(ong) ∼ K20 with τl ∼ 5.2 × 10−8 s and Ks(hort)
0
∼ K10 with τs ∼ 8.9 × 10−11 s
The different lifetimes of these states mean that an originally created K 0 will with time oscillate into a
¯
K . This can be seen if we use Eq. 6.15 to show K 0 as a superposition of K20 and K10 states:
0
1 1
|K 0 ⟩ = √ |K10 ⟩ + |K20 ⟩ ∼ √ |Ks0 ⟩(t) + |KL0 ⟩(t)

(6.18)
2 2
1 0 −ims − 2τt t
0 −iml − 2τl

∼ √ |Ks ⟩e s + |K ⟩e
L (6.19)
2
(6.20)
The time evolution will progress in a similar fashion as for neutrino oscillations, but now the lifetime
difference has to be also taken into account. Meson oscillation phenomena have been observed in many
neutral mesons:
K 0 = |ds̄⟩ D0 = |ūc⟩ Bd0 = |b̄d⟩ Bs0 = |b̄s⟩
¯ D̄0 = |uc̄⟩ B¯0 = |bd⟩
K¯0 = |ds⟩ ¯ B¯0 = |bs̄⟩
d s
The corresponding mass and lifetime configurations for these systems are shown in Fig. 6.7.
So far we have assumed CP ˆ is conserved. E.g that the Ks0 ∼ K 0 can only decay into the CP ˆ even
1
0 0 ˆ
ππ state and KL ∼ K2 only into the CP odd πππ state. However Chronin and Fitch found that a small
fraction of the longer lived KL0 do decay via two pions. This was a clear sign of CP violation. There are two
possible explanations for this. First the physical states |Ks0 ⟩ and |KL0 ⟩ might not be perfectly equal to the
CP eigenstates K10 and K20 but contain a small contamination, ϵ, of the “wrong” CP state. This possibility
is shown in Eq. 6.22.
1
|KL0 ⟩ = |ϵK10 ⟩ + |K20 ⟩

√ (6.21)
1+ϵ 2
1
|Ks0 ⟩ = |K10 ⟩ − ϵ|K20 ⟩

√ (6.22)
1 + ϵ2
The second possibility is that CP is not conserved in the decay of the CP eigenstates K10 and K20 . This
is shown in Eq. 6.23 and known as direct CP violation.
⟨ππ|K20 ⟩ = ϵ′ ̸= 0 (6.23)
Measurements
′ show that both effects exist, but are small. It was found that ϵ = (2.232 ± 0.007) × 10−3
and ϵϵ = (14.7 ± 2.2) × 10−4 .
CP violation effects have also been found in bottom mesons and baryons. Precision measurements of B-
hadron decays have greatly improved the understanding of CP violation and the CKM matrix. Unfortunately
75
Figure 6.7: Shown are the mass and lifetime configurations in form of a Breit–Wigner resonance for K 0 = |ds̄⟩,
D0 = |ūc⟩, Bd0 = |b̄d⟩ and Bs0 = |b̄s⟩ mesons. The large lifetime difference between |Ks0 ⟩ and |KL0 ⟩ is visible in
the top left. For B-mesons in the bottom row, the mass difference is more important and the corresponding
states are referred to as BLight and BHeavy . (Taken from doi:10.1142/S0217751X99001391)
all CP violating effects found so far are too small to explain the baryon asymmetry of the universe. The
origin of the known CP violation is understood to be the complex phase ρ − iη of the CKM matrix. A look
at the box diagrams for K 0 and K¯0 (Fig. 6.4) reveals a difference between the transition matrix elements
for these particles if V ̸= V ∗ . The Jarlskog invariant J = Im(Vij Vkl Vij∗ Vkl∗ ), named after Swedish scientist
Cecilia Jarlskog, which is a phase-convention-independent measure of CP violation is given by the area of
the CKM triangles.
76
Chapter 7
The Electroweak Interaction: Part III
7.1 The unification of electromagnetic and weak interactions

Before the 1960s the weak and electromagnetic interactions were thought to be completely distinct inter-
actions as they behaved so differently. In particular, the strength of the weak interaction is much weaker,
which is why particles that decay via the weak interaction have a longer lifetime.
7.1.1 Problems with the theory of weak interactions

Initially weak decays were treated as a four-point interaction with one vertex, for example for β-decay:
n → pe− ν̄e (here n is a neutron and p is a proton), as shown in Fig. 7.1(a). While this was a perfectly
good description at low energies, the calculations led to cross sections that increase linearly with the energy
of the interaction and resulted in infinities in the calculations. This is clearly non-nonsensical as it does
not make sense for a theory to give unphysical predictions for cross sections. It was clear that the four-
point interaction was a low-energy approximation to an underlying theory. In order to solve this problem,
the W boson propagator was introduced, analogous to the photon in the electromagnetic interaction, as
shown in Fig. 7.1(b). This solved the above problems, however the new theory also produced infinities in
p p
n e− n e−
−
W
(a) ν̄e (b) ν̄e
Figure 7.1: Feynman diagrams for nuclear β-decay described by (a) a four-point interactions and (b) the
exchange of a W − boson.
the calculation of the cross sections of processes with more than one W boson! Something was clearly not
right. Inspired by hints that the electromagnetic and weak interactions had some connections (e.g. the W
77
boson carries an electric charge and so must be connected in some way to the electromagnetic interaction)
physicists in the 1960s developed a unified electroweak theory that solved the above problems of the theory
just involving W bosons. This new theory predicted the existence of a new neutral boson, which led to an
exact cancellation of the problematic infinities. As an example of how this works consider the higher order
correction to the e+ e− → µ+ µ− process involving two W bosons shown in Fig. 7.2(a). This is an example of a
process that led to infinities in calculations. However, if similar diagrams involving a new neutral boson (the
Z boson) and the photon in the loops, see Fig. 7.2(b) and 7.2(c) for examples, are included in calculations,
the infinities exactly cancel, leading to sensible results. This mathematical trick may seem like a cheat, but
the new particle they predicted to exist was later discovered (see Sections 7.1.4 and 7.1.5) and is now an
integral part of the Standard Model.
e+ µ+ e+ µ+ e+ µ+
W+ Z0 γ
νe νµ e− µ− e− µ−
W− Z0 Z0
e− µ− e− µ− e− µ−
(a) (b) (c)
Figure 7.2: Feynman diagrams for higher-order corrections to e+ e− → µ+ µ− .
7.1.2 The unified electroweak theory (very briefly!)

A detailed description of the unified theory is beyond the scope of this course. Here we just give an indication
of what it means. We can see that the mathematical trick that led to cancellations in the previous section
requires a connection between the W and Z bosons and the photon couplings to fermions in order to ensure
an exact cancellation of the infinities. In fact, it turns out that the Standard Model describes electroweak
interactions in a high-energy Universe (e.g. in the period before 10−12 seconds after the Big Bang) as
occurring via the exchange of four massless bosons known as the W1 , W2 , W3 and B. The strength of
the interactions due to these four bosons is very similar. Of course, this is not what we see at everyday
energies today. And it turns out that a process known as electroweak spontaneous symmetry breaking via
the Higgs mechanism (to be discussed in Section 7.2.2) occurs when we go from high to low energies (as
the Universe cooled down). When this process occurs the four bosons mix into the four physical states that
we see in interactions today: three massive bosons, the W + = √12 (W1 + iW2 ), W − = √12 (W1 − iW2 ) and
Z = sin θW W3 + cos θW B and one massless boson, the γ = cos θW W3 − sin θW B, where θW is known as the
weak mixing angle (or the Weinberg angle) and is defined by cos θW = M MZ . We now have one particle that is
W
massless and three that are massive, leading to large differences between the interactions that they mediate.
An explicit calculation of the couplings of the physical W and Z bosons and the photon leads to: 2√e2ϵ =
√ 0
gW sin θW = gZ cos θW , where e/ ϵ0 is the strength of the photon coupling to charged fermions with charge
78
e, gW is the strength of the W boson coupling to fermions, and gZ characterises the strength of vertices
involving the Z boson (although the strength of the interaction is not the same for all fermion pairs so gZ is
multiplied by different factors depending on which fermions are involved). Note that the intrinsic strengths
of the weak and electromagnetic interactions are actually very similar, but the difference in the masses of
the bosons leads to different interaction strengths as we shall see in Section 7.1.3.
7.1.3 The low energy approximation

The four-point interactions were extremely successful at describing low-energy weak interactions with inter-
action strengths much smaller than electromagnetic interactions. We have seen that the intrinsic strength of
the two types of interaction are in fact quite similar. The fact that the weak interactions appear so weak is
due to the large masses of the exchange particles. Recall from the “The scattering amplitude” section in the
g 2 h̄2
“Basic Ideas of Particle Physics” module that the amplitude of an interaction has the form: M ∼ − MW2 c2 ,
W
for an interaction involving a W boson exchange with a four-momentum exchange much less than MW .
Because of the large mass of the W boson, the interaction appears weak at low four-momentum transfers.
At four-momentum transfers much larger than MW , the interaction is no longer dependent on the mass of
the W boson and depends only on the intrinsic strength of the interaction. At these energies the weak and
electromagnetic interactions have similar strengths.
7.1.4 The discovery of neutral currents

So, the existence of a new massive Z boson had been predicted in the 1960s, but so far no interactions
involving this particle had been explicitly observed (although it was of course used to calculate the higher-
order corrections to processes such as e+ e− → µ+ µ− discussed before). All observed weak interactions
involved the exchange of an electrically charged particle, such as the well-known weak decays of particles.
In 1973 a beam of neutrinos was delivered to CERN’s Gargamelle bubble chamber experiment. As
was expected, interactions involving the exchange of the charged W boson, known as “charged current
interactions”, were observed. The explicit process is shown in Fig. 7.3(a), where a muon-neutrino scatters
off nucleons in the atoms of the detector via the exchange of a charged particle (a W boson), changing the
muon-neutrino into a muon. But in addition to this expected type of process, another was observed involving
the exchange of a neutral boson, known as a “neutral current interaction”. The explicit process is shown in
Fig. 7.3(b). Here the scattering does not turn the muon-neutrino into a muon, it remains a muon-neutrino.
This is only possible if the exchange particle is neutral (and recall that the photon couples to electric charge
and so cannot be responsible for this interaction). This was exactly the type of process predicted by the unified
electroweak theory, which had predicted the existence of a massive weakly interacting, electrically neutral
boson. What is more, the rates of the interactions were compatible with the expectations of the electroweak
theory and led to predictions of the W and Z masses of MW = 78.3 GeV/c2 and MZ = 89 GeV/c2 , not
far from the current measurements of these parameters. This discovery was a huge success for the unified
electroweak theory and led to a need to directly discover the W and Z bosons.
7.1.5 The discovery of the W and Z bosons

The W and Z bosons were discovered at the Super proton–antiproton Synchrotron (SppS) at CERN. In pp̄
collisions it was thought that the W and Z bosons could be produced by quark–antiquark (q q̄) annihilations,
′
e.g. via q + q̄ → Z and q + q̄ → W ± .
79
νµ µ− νµ νµ
W+ Z0
Hadrons Hadrons
N N
(a) (b)
Figure 7.3: Feynman diagrams for (a) charged current and (b) neutral current interactions.
First let us recall from Section 1.10, that the cross section for an interaction involving the exchange of an
unstable particle has the form:
g 4 h̄4 c4
σ∝ , (7.1)
(q 2 − MX2 c4 )2 + M 2 c4 Γ2 h̄2
X
leading to a distinctive Breit–Wigner distribution when the cross section is plotted versus q 2 , which is the
four-momentum squared of the propagator and is equal to the invariant mass squared (and also the centre-of-
mass energy squared) of the interaction. There is a clear enhancement in the cross section when p approaches
MX c2 , where MX is the mass of the mediating particle. So if we want to produce the W and Z bosons then
the centre-of-mass energy√must be very close to the mass of those particles. The SppS was built with a
centre-of-mass energy of s = 540 GeV. Of course this is much larger than the predicted masses of the W
and Z bosons; recall, however, that the interacting quarks will only carry a fraction of the (anti)proton’s
momentum and each q q̄ interaction will have a different centre-of-mass energy.
The idea was to search for W and Z bosons produced at the SppS and, since they decay almost instan-
taneously, to observe their decay products in the detectors surrounding the interaction point. The W and
Z bosons will decay to a pair of fermions, those involving electrons and muons are the easiest to observe
experimentally. When searching for pp̄ interactions (events) creating a Z boson, the idea is to look for events
with a e+ e− or µ+ µ− pair from the decay of the Z. If a Z boson is created, the invariant mass of the e+ e−
or µ+ µ− pair should peak at MZ due to the form of the cross section in Equation 7.1. This is exactly what
was observed, as can be seen in Fig. 7.4. The Z boson had been discovered with a mass completely consistent
with that predicted by the rates of charged and neutral currents.
For the W boson the situation is more complicated. The eνe and µνµ decays involve the weakly interacting
neutrinos that cannot be detected directly by these experiments. Their existence is inferred by an imbalance
in momentum in the plane transverse to the colliding beams and kinematic variables related to MW were
measured, where a peak was also observed. The W boson had also been discovered with a mass completely
consistent with that predicted by the rates of charged and neutral currents.
Since their discovery the W and Z bosons have been studied in great detail. In particular the LEP collider
at CERN collided electrons with positrons at a centre-of-mass energy of 91 GeV, exactly the rest mass of
the Z boson1 , meaning that many millions of Z bosons were produced and studied in great detail. Today,
1 Note that because electrons and positrons are fundamental particles, the individual interactions will have a centre-of-mass
80
Figure 7.4: Invariant mass of the lepton-pair in candidate Z → e+ e− and Z → µ+ µ− events.
at the LHC, many millions of W and Z bosons are produced and studied in great detail.
7.2 The Higgs boson

7.2.1 Gauge invariance
The Standard Model is known as a “gauge theory”. A fundamental requirement of all gauge theories is that
the equations of the theory are invariant under local gauge transformations of the wavefunctions describing
the particles (this is a gauge symmetry). A local gauge transformation transforms the wavefunction like:
′
Ψ (⃗x, t) → Ψ (⃗x, t) = e−if (⃗x,t) Ψ (⃗x, t)
where local means that the function f depends on ⃗x and t (in a global gauge transformation it would
be a constant). We can immediately see that the free particle equations are not invariant under such a
transformation, because they contain time and space derivatives that lead to troublesome terms leading to
differences in the equations before and after the transformation. For example, consider a time derivative on
the wavefunction: ′
∂Ψ (⃗x, t) ∂f (⃗x, t) ′ ∂Ψ (⃗x, t)
= −i Ψ (⃗x, t) + e−if (⃗x,t) ,
∂t ∂t ∂t
and it is clear that the first term on the right hand side will be present in the equations after the transfor-
mation, but not before. To solve this problem additional terms are added to the equations that ensure an
exact cancellation with the troublesome terms, ensuring that the equations are gauge invariant. The amazing
thing about this trick is that the additional terms (those that are added to ensure gauge invariance) are the
terms that describe the particle interactions via bosons (often referred to as gauge bosons for this reason).
So by ensuring gauge invariance we have gone from a non-interacting theory to an interacting theory. This is
how the mathematical equations for interactions in the Standard Model are derived. As an explicit example,
in QED the photon is introduced when insisting that the free electron equations are invariant under gauge
transformations.2 After imposing this mathematical trick we obtain a theory that agrees spectacularly well
with experiment.
energy of 91 GeV, unlike in hadron collisions.
2 You can see this explicitly for the case of the photon in QED in Section 10.1 in “Modern Particle Physics”, M. Thomson.
81
This is all well and good for QED (and for QCD), but in the electroweak theory a problem arises. If
you allow a mass term in the equations of motion for the gauge bosons in the theory, the gauge invariance
is destroyed. In other words, for this trick to work it is a requirement that the gauge bosons are massless.
But we know that the W and Z bosons are massive. You might ask whether we could just say that this
theory is not a gauge invariant, but it turns out that gauge invariance is required in order to ensure the exact
cancellations of the infinities in diagrams such as those in Fig. 7.2.
7.2.2 The Higgs mechanism

After searching in a number of ways to overcome this problem in the electroweak theory, the Higgs mechanism
was proposed in 1964 by Englert, Brout, Higgs, Hagen, Guralnik and Kibble. We do not go into the details
of the Higgs mechanism in this course, but just give a hint of what it does. The basic idea is to break the
symmetry in a spontaneous way, which allows masses to be added to the gauge bosons, but does not destroy
the theory. The idea is that a field, known as the Higgs field, exists at all points in the Universe. It is the
interactions with this field that give the particles their mass in such a way that the theory describing the
interactions remains gauge invariant. The gauge symmetry must be broken in some way to give the particles
masses, but this happens because the expectation value of the Higgs field in the vacuum is non-zero (all the
other fields have zero expectation values in the vacuum).
We can visualise how a field would have a zero expectation value if we consider the potential of the field
to have a shape that does not minimise at zero. Consider the potential shown in Fig. 7.5, where the potential
actually has a local maximum when the field is zero. The minimum of the potential can take one of any of
Figure 7.5: The potential of the Higgs field versus the real and imaginary components of the field.
the values at the base of the rim of this “Mexican hat” shape. Our Universe chooses a given point in the
minimum for the value of the field in the vacuum. This choice spontaneously breaks the symmetry. The
masses then arise from the interactions of the gauge bosons with the non-zero vacuum expectation value of
the Higgs field. So the value of the field in the vacuum is not gauge invariant, but the interactions are. This
allows the particles to have mass without breaking the theory.
Peter Higgs pointed out that if this mechanism is correct then a new particle must exist which is the
excitations of the Higgs field (just as photons are excitations of an EM field). The Higgs boson was predicted
to couple to all particles with mass, with a coupling that increases with the mass of that particle. Because
of the success of the electroweak theory in all other aspects, the possible existence of this particle was taken
82
very seriously and in order to verify it the hunt for the Higgs boson began. More than 50 years since its
prediction the Higgs boson was finally discovered in July 2012 by the LHC experiments, ATLAS and CMS.
7.2.3 Discovery of the Higgs boson

Higgs production at the LHC
The Standard Model predicted that Higgs bosons will be created at the LHC. There are a number of different
production mechanisms. Since the coupling of the Higgs boson to other particles depends on the particle’s
mass, the dominant production mechanisms involve the heaviest particles, which are the top quark, bottom
quark and the W and Z bosons. The dominant production mechanism is known as gluon–gluon fusion.
Gluons have zero mass and so do not couple directly with the Higgs, so this process occurs via a quark loop,
as shown in Fig. 7.6(a). This loop will usually be the top quark as it is the heaviest. The next most likely
production mechanism is known as weak boson fusion (or vector boson fusion). In this case either a W ± or
a Z boson is radiated from a quark from each proton. Then either a W + W − or a ZZ pair fuse to form a
Higgs. This process is shown in Fig. 7.6(b). The next most likely production mechanism is known as W ± or
Z boson associated production. In this case a virtual W ± or Z boson is produced, e.g. by quark–antiquark
annihilation, and the boson radiates a Higgs boson, as shown in Fig. 7.6(c). It should be noted that the
initial W ± or Z boson will be virtual with an invariant mass larger than the W ± or Z boson mass, so that
there is enough energy for the Higgs and the W ± or Z boson to be created. While the gluon–gluon fusion
process dominates, the other two processes are very important as they provide additional final-state particles
that help distinguish the events from other “background” processes, as we shall see in a moment.
′
g q q( ) q V
V
H
t
H V
V
g (a) q (b)
′
q( )
′
q̄ ( ) (c) H
Figure 7.6: Partial Feynman diagram for (a) gluon–gluon fusion, (b) weak boson fusion and (c) associated
W/Z Higgs production. In these diagrams V indicates a W ± or a Z boson. Note that these are not complete
diagrams as the Higgs bosons and W ± and Z bosons also decay.
Higgs decay
A Higgs boson produced at the LHC will decay almost instantaneously. Any unstable particles produced in
the decay will then go on to decay further. As its coupling depends on the mass of the particle, it is most
likely to decay to the heaviest particles that it is kinematically allowed to decay to. Since the mass of the
Higgs boson is MH = 125 GeV/c2 , if a Higgs is produced with this invariant mass, there would not be enough
energy to create two top quarks, each with a mass of Mt = 175 GeV/c2 . The heaviest particle with a mass
83
less than half the Higgs mass is the bottom quark. For that reason the dominant decay of the Higgs boson
is to a bb̄ pair. The next most likely decay is to a W + W − pair. Even though MH < 2MW , if one of the W
bosons is produced with a mass less than MW (which is allowed for virtual particles) this is possible, although
suppressed. Other important decays are to the other heavy particles: a ZZ pair, a τ + τ − pair. And while it
is not immediately obvious, the decay to a γγ pair is one of the most important channels. As the photon is
massless this occurs via a particle loop, just like gluon–gluon fusion, where the most likely particles in the
loop are a top quark or a W boson (see Fig. 7.8 for partial Feynman diagrams for the dominant Higgs decays
including the γγ loop). Now that the Higgs has been discovered we know its mass and we can determine
the branching fractions to various particles. However, the Higgs mass was not known before its discovery,
and as can be seen in the plot in Fig. 7.7, the branching fractions depend very strongly on the mass of the
Higgs boson (in particular, at higher masses decays to the heavier particles become more important). It was
therefore essential to search in all accessible decay channels, keeping an open mind about the possible mass
of the particle. We will learn about some of the decay channels in the next sections.
Figure 7.7: Theoretical branching fraction of the Higgs Boson depending on the Higgs mass. The observed
Higgs Boson was found at a mass of 125 GeV/c2 and the relevant branching fractions are indicated by the
line at that mass point.
The challenge of the Higgs discovery

The LHC collides bunches of protons 40 million times every second. But the most likely type of interaction
is an uninteresting low-energy QCD interaction. A Higgs boson is produced only once in every 10 billion
proton–proton interactions. The challenge we face is finding the needle of Higgs interactions in the haystack
of proton–proton interactions. We therefore need to look in Higgs decay channels that have low backgrounds.
A background is another type of process that mimics the Higgs production as it has the same, or similar,
final-state particles. We cannot know in a given proton–proton interaction (event) if a Higgs was produced,
even if it looks very much like it was. Instead, we search for a statistically significant excess of events over
84
Figure 7.8: Leading order Feynman diagrams for important Higgs decay modes.
the expected contribution from background processes. Since the invariant mass of the Higgs decay particles
will peak at the mass of the Higgs boson, we can search for a peak in the invariant mass distribution of
decay products. This is typically how we search for new particles (e.g. recall the discovery of the Z boson in
Section 7.1.5). In order to do this we need a good resolution of the energy and momenta measurements of
the decay particles so that the peak, which is likely to sit on a continuum distribution, does not get smeared
away. Electrons, muons and photons all have very good resolutions so decays of the Higgs that end up with
these particles in the final-state are much easier to deal with than those with neutrinos or hadrons. Despite
57% of Higgs bosons decaying to a bb̄ pair, it was the much rarer ZZ and γγ decay channels in which it was
initially discovered. We will go through each of the main decay channels in turn.
H → ZZ (discovery channel)
In this decay channel the two Z bosons will decay almost instantaneously to a pair of fermions. As usual,
the easiest to detect are the e+ e− and µ+ µ− channels, so for a H → ZZ event that would be: e+ e− e+ e− ,
µ+ µ− µ+ µ− or e+ e− µ+ µ− . It is possible to produce a pair of Z bosons through other processes other than
from a Higgs decay. However, if we plot the invariant mass of the four leptons in the final state there should
be a peak at the mass of the Higgs for the events coming from a Higgs. A peak at 125 GeV/c2 was observed
and this contributed to the discovery of the Higgs boson. See Fig. 7.11 for the Feynman diagrams for the
background processes. The invariant mass plot similar to that used to make the discovery is shown in Fig. 7.9.
H → γγ (discovery channel)
The photons from this decay channel will be identified in the calorimeters of the detectors. There is a large
background of diphoton production at the LHC, however the resolution of the photons is good so a little
bump at the Higgs mass can be observed on the continuum of the background if the invariant mass of the
two photons is plotted. A peak at 125 GeV/c2 was observed and this also contributed to the discovery of
the Higgs boson. The invariant mass plot used to make the discovery is shown in Fig. 7.10. See Fig. 7.11 for
the Feynman diagrams for the background processes.
85
Figure 7.9: Histogram of the invariant mass of 4 muons. At a mass of 125 GeV/c2 a clear peak can be
seen from H → ZZ → µ+ µ− µ+ µ− decays. The peak at 90 GeV/c2 comes from Z → µ+ µ− µ+ µ− decays.
Collision data are shown as black dots and the expectations from simulation are indicated in colour.
Figure 7.10: Histogram of the invariant mass of two photons. At a mass of 125 GeV/c2 a clear peak from
H → γγ decays can be seen over the continuum background.
H → W +W −
This decay channel is the second largest (with a branching fraction of 21%), however it was much harder
to discover the Higgs this way than in the ZZ and γγ channels. This is because the leptonic decays of the
W involve neutrinos that cannot be detected at the LHC detectors. The hadronic decays do not have this
problem, but they suffer form huge backgrounds and poor resolution. Since a neutrino cannot be detected
its momentum and energy cannot be determined so the invariant mass of the decay products cannot be
reconstructed. A quantity related to the invariant mass, using the kinematics in the plane transverse to the
beam only is plotted instead, which does not have such a clear peak. Nonetheless it has been used to discover
the Higgs boson in this channel too (although this happened after the initial discovery of the Higgs in July
86
2012 when more data had been collected, with more time to analyse it).
H → τ +τ −
This process was also discovered a long time after the ZZ and γγ channels. This is because the taus will
decay very quickly, and the decay products always contain neutrinos and very often hadrons. So, again, the
invariant mass cannot be fully reconstructed. A variable as close as possible to the invariant mass is plotted,
but this has a much worse resolution than in the ZZ and γγ channels. The main background in this case is
Z → τ + τ − events. While the invariant mass in this case would peak at 91 GeV/c2 , rather than 125 GeV/c2 ,
the resolution on the mass-like variable is so large that the two peaks get merged into each other. We are
therefore looking for a hump on the shoulder of the distribution rather than a distinctive peak.
In order to enhance the Higgs events over the Z and other background processes, the weak boson fusion
process is used as this leads to additional identifying features (e.g. two high-energy jets from the final-state
quarks seen in Fig. 7.6(b)). Utilising these additional features the Higgs has been discovered in this decay
channel.
H → bb̄
Despite the fact that 57% of Higgs bosons decay to a bb̄, this decay channel has only been very recently
discovered. It suffers from an extremely large background, which is the QCD production of a bb̄ final state,
which is 10 million times more likely to occur than a H → bb̄ event. Also important is the poor resolution of
the jets coming from the bottom quarks, making a peak in the dijet invariant mass difficult to resolve. For
this reason this decay channel is traditionally searched for via the associated W ± or Z production mechanism.
The distinctive features of the leptonic decays of the W ± and Z hugely reduce the background. Nonetheless,
the backgrounds are still large and although this process has been seen in 2017, uncertainties remain relatively
large.
Figure 7.11: Leading order Feynman diagrams for important background processes in Higgs searches, that
are leading to a similar signature.
87
Chapter 8
Nuclear Phenomenology
8.1 Introduction
Atoms are made of a nucleus orbited by a cloud of electrons. This picture followed from the famous scattering
experiments by Rutherford. The nucleus is composed of protons and neutrons. The chemical properties, e.g.
which element we are dealing with is determined by the electron cloud. The number of electrons balances
the number of protons. The number of protons is hence known as atomic number Z. The mass number
A = N + Z is the sum of the number of neutrons N and protons Z. For nuclides the notation A Z X is used,
where X is the usual symbol for the chemical element. Nuclides with common properties can be classified as
follows:
Isotopes have the same atomic number Z, hence the same number of protons and hence the same chemistry,
e.g. 11 14
6 C and 6 C.
11 11
Isobars have the atomic mass, e.g. 6 C and 5 B.
Isotones have the same number of neutrons N , e.g. 157 N and

14
6 C which are both made of 8 neutrons.
The mass of the neutron and proton is rather well known:
Mp = 938.27208816 ± 0.00000029 MeV (8.1)

Mn = 939.56542052 ± 0.00000054 MeV (8.2)
12
Often the unified atomic mass u, a twelfth of the mass of a carbon-12 6 C is used:
u = 931.49410242 ± 0.00000028 MeV
8.2 Shape of the Nucleus

The Rutherford experiment established that there is a positively charged nucleus at the centre of the atom.
This was done by scattering an α-particle, a helium nucleus, on larger atoms. The corresponding cross
88
section, assuming scattering of pointlike, spinless objects, can be derived classically and yields:
Z 2 α2 (h̄c)2

dσ
= (8.3)
4E 2 sin4 θ2

dΩ Rutherford
If electrons, which are spin 1/2 Fermions, are scattered, their spin has to be taken into account. This was
calculated by Mott:
dσ dσ θ
= 1 − β 2 sin2 (8.4)
dΩ Mott dΩ Rutherford 2
Scattering experiments with electrons probe the charge distribution of the nucleus. The scattering cross
section on a composite particle with an extended spatial charge distribution f (x) has to be modified with
the form factor F (q 2 ) as shown in Eq. 8.6. The form factor F (q 2 ) is the Fourier transform of the charge
distribution f (x):
1
Z
q ·⃗
i⃗ x
F (q 2 ) = f (⃗x)e h̄ d3 ⃗x (8.5)
Ze

dσ dσ 2
= F (q 2 ) (8.6)
dΩ experimental dΩ Mott
The first to perform these electron scattering experiments was Hofstaedter. From empirical fits to the
measured form factor, the size of the nucleus could be determined. For the nuclear radius R it was found:
1 1
R = R0 × A 3 = 1.2 fm × A 3 (8.7)
8.3 Nuclear Mass and Binding energy

The mass of the nucleus is an important parameter. Isotopes of an element have different mass (e.g. carbon-
12 12 14
6 C and carbon-14 6 C). The mass number A only specifies the total number of nucleons. As can be seen
from the definition of the unified atomic mass u in section 8.1 nuclei are actually lighter than the sum of the
mass of the constituting nucleons.
M (Z, A) < Np × Mp + Nn × Mn
The origin of this mass deficit is the binding energy B(Z, A):
B(Z, A) = −∆M (Z, A) = − (M (Z, A) − Np × (Mp + me ) − Nn × Mn ) = − (M (Z, A) − Z(Mp + me ) − (A − Z)Mn )

(8.8)
Often the binding energy per nucleon B/A is quoted.
8.4 Semi Empirical Mass Formula (SEMF)

The Semi Empirical Mass Formula (SEMF) incorporates elements of the liquid drop model. The SEMF
makes an attempt at describing the mass of nuclei depending on mass number A and atomic number Z. The
binding energy B(A, Z) can be calculated from the SEMF using Eq. 8.8.
2
2 Z2 Z − A2 1
M (Z, A) = Z(Mp + me ) + (A − Z)Mn − av A + as A + aC 1 + aa
3 ± δap A− 2 (8.9)
A3 A
89
often used notation → VSCAP:

a1 = aV Volume; a2 = as Surface; a3 = aC Coulomb; a4 = aa Asymmetry; a5 = ap Pairing
aV = 15.67, as = 17.23, ac = 0.714, aa = 93.15, ap = 11.1, all in MeV
Constituent mass term The 0th term simply adds the masses of all constituents f0 (Z, A) = Z(Mp +
me ) + (A − Z)Mn
Volume term The 1st term assumes that nucleons bind to each neighbour and the binding energy is
1
proportional to the volume of the nucleus. In the simple drop model we can assume a sphere of radius R ∼ A 3 .
1 3
Binding energy reduces the mass of the nucleus and hence is negative: f1 (Z, a) = −a1 A 3 = −a1 A
Surface term Nucleons on the surface lack binding partners, hence the binding energy is reduced by an
2
amount proportional to the surface area: f2 (Z, A) = +a2 A 3
Coulomb term Protons carry positive electric charge and hence repel each other according to the Coulomb
law, reducing the binding energy.
Asymmetry term This will be further investigated in Section 9.1.2.
Pairing term Nucleons in the same spatial state, but opposite spin directions can combine to give a wave-
function of spin-0. This bosonic state follows Bose-Einstein statistics and hence the wave-functions overlap
very heavily. This means the nucleons are tighter bound if they can be paired and the binding energy is
1
maximised when both Z and N are even f5 (Z, A) = −f (A) = −a5 A 2 . In an odd–even or even–odd nucleus
the effects cancel and f5 (Z, A) = 0. Finally for an odd-odd nucleus with both an unpaired proton and an
1
unpaired neutron the binding energy is reduced by f5 (Z, A) = f (A) = a5 A 2 .
8.5 Nuclear Stability and Decay

Only a fraction of possible neutron and proton combinations are stable. Nuclei decay into energetically more
favourable stables. The number of decays in a given time is called the activity A and follows a well known
law:
dN N
A=− = (8.10)
dt τ
where τ is the lifetime and N the number of nuclei. This means the number of nuclei remaining follows
t
N (t) = N0 e− τ . Often the half-life T 21 is quoted, the time after which half of the nuclei have decayed. It is
related to the lifetime:
90
T1
2
0.5N0 = N0 e − τ (8.11)
T1
− τ2
0.5 = e (8.12)
T 1
ln 0.5 = − 2
(8.13)
τ
T 12
− ln 2 = − (8.14)
τ
τ ln 2 = T 12 (8.15)
The units used to measure activity are

• Becquerel 1 Bq = 1 decay/s
• Curie 1 Ci = 3.7 × 1010 decay/s
In practise the exposure, measured in Roentgen 1 R = 2.58 C/Kg or the absorbed dose in rad 1 rad =
0.01 J/Kg or Gray 1 Gy = 1 J/Kg are a better to classify sources.
Finally, the biologically equivalent dose, measured in rem (Roentgen equivalent man) or Sievert Sv, takes
the varying biological effects of α, β and γ radiation into account.
8.6 β-Decay
Figure 8.1: For a fixed mass number A, the SEMF can be shown to follow a parabola in atomic number Z
as shown here. The lowest mass point has the largest binding energy and corresponds to the most stable
isobar. In the N vs Z plane this creates a valley of stability as shown on the right.
A close inspection of the SEMF Eq. 8.9 reveals that for a fixed mass number A, the mass depends
quadratically on the number of protons Z. That means that for every odd mass number A a parabola is
formed as shown in Eq. 8.20 and Fig. 8.1. The most tightly bound and hence most stable nucleus is found at
the bottom of the parabola. An odd mass number leads to either an odd-even or even-odd nucleus for which
91
the pairing effects cancel and hence only a single parabola is formed. A detailed look at Eq. 8.20 reveals that
for even mass numbers the situation is more complex. Due to the pairing term nuclei with an even number
of protons Z and even number of neutrons√ are more tightly bound than odd–odd nuclei. Hence the SEMF
shows two parabolas separated by 2ap / A. The more stable nucleus is usually found on the lower parabola
with even-even nuclei. This is shown in Fig. 8.1. The resulting valley of stability can also be seen in Fig. 8.3.
1

M (Z, A) = αA ± λA− 2 − βZ + γZ 2 (8.16)
1 aa
α = Mn − av + as A− 3 + (8.17)
4
β = aa + Mn − Mp − me (8.18)
aa 1
γ = + a c A− 3 (8.19)
A
λ = ap (8.20)
8.6.1 β − -Decay
Elements on the left of the parabola minimum shown in Fig. 8.1 are neutron unstable, e.g. it is energetically
favourable to turn a neutron into a proton. This is β − decay:
n → p + e− + ν¯e
M (Z, A) → M (Z + 1, A) + e− + ν¯e
Such a move will increase Z by 1 while the mass number A stays the same, hence the nucleus moves closer
to the minimum of the SEMF parabola. This decay is allowed if:
M (Z, A) > M (Z + 1, A)
8.6.2 β + -Decay
Elements on the right of the parabola minimum shown in Fig. 8.1 are proton unstable, e.g. it is energetically
favourable to turn a proton into a neutron. This is β + decay:
p → n + e+ + νe
M (Z, A) → M (Z − 1, A) + e+ + νe
Such a move will decrease Z by 1 while the mass number A stays the same, hence again the nucleus moves
closer to the minimum of the SEMF parabola, with the most stable isobar. This decay is allowed if:
M (Z, A) > M (Z − 1, A) + 2me
Note that if we use atomic masses, not only do we need to take the created positron into account, but also
the excess electron when going from Z to Z − 1.
92
Figure 8.2: Binding energy per nucleon. The blue line represents Eq. 8.27.
8.6.3 Electron Capture

Another way to turn a proton into a neutron in the nucleus is electron capture.
p + e− → n + νe
M (Z, A) + e− → M (Z − 1, A) + νe
Electron capture is allowed if:
M (Z, A) > M (Z − 1, A) + ϵ
where ϵ is the excitation energy of the atomic shell of the captured electron.
8.7 α-Decay
An alpha particle is a helium-4 4 He nucleus. This nucleus is especially tightly bound and the binding energy
has been measured to be B(2, 4) = 28.3 MeV. Hence emitting an α-particle is energetically favoured if
B(2, 4) > B(Z, A) − B(Z − 2, A − 4). Using the SEMF Eq. 8.9 and approximating the stability line with
Z = N = A/2 an estimate of the binding energy difference, depending only the mass number A can be
obtained:
dB
B(Z, A) − B(Z − 2, A − 4) = B(A/2, A) − B(A/2 − 2, A − 4) ∼ ∆A (8.21)
dA
In the last step we made a linear approximation for a small change in A. For an alpha particle (a helium-4
4
He nucleus), ∆A = 4. In order to evaluate the slope of the binding energy dB dA , we could go back to SEMF
93
or use the plot of the measured binding energy per nucleon B/A shown in Fig. 8.2. The later turns out to
be a bit more straight forward.

B 1 B
d = dB − dA (8.22)
A A A2

dB d(B/A) B
→ = A + (8.23)
dA dA A
The slope d(B/A)

dA can be approximated by the slope of black line in Fig. 8.2 and hence d(B/A)
dA ∼ −7.7 ×
−3
10 MeV. From Eq. 8.21, decay by α emission is allowed if:

dB d(B/A) B
B(2, 4) = 28.3 MeV > B(Z, A) − B(Z − 2, A − 4) ∼ ∆A = A + ∆A (8.24)
dA dA A

d(B/A) B B
28.3 MeV > A + ∆A = −A × 7.7 × 10−3 MeV + 4 (8.25)
dA A A
28.3 MeV B
> + − A × 7.7 × 10−3 MeV (8.26)
4 A
B 28.3 MeV
< + A × 7.7 × 10−3 MeV = 7 MeV + A × 7.7 × 10−3 MeV (8.27)
A 4
The last equation 8.27 can be thought of as the blue line in Fig. 8.2, which crosses the binding energy per
nucleon curve at a mass number of A ∼ 153.
8.8 γ-Decay
Following a decay process, nuclei are often left in an excited state. They de-excite emitting a photon via
γ-decay.
94
Figure 8.3: Nuclide chart indicating the main decay options of the nuclei. The valley of stability is shown
in black. Note that high mass elements have an excess of neutrons over protons.
95
Chapter 9
Nuclear Structure
The SEMF discussed in Section 8.4 reasonably describes the broad features of the nuclear binding energy.
The underlying liquid drop model provided a qualitative explanation for spontaneous fission. However some
detailed variations in nuclear binding energy are not described. In addition some of the terms were not
motivated by the liquid drop model, but just added adhoc into the SEMF.
9.1 Fermi Gas Model

The Fermi gas model can explain some of these added terms. The model assumes a 3D potential well in
which the nucleons can move freely, like a gas. The potential that nucleons experience is created by the
superposition of the other nucleons. The potential well is different for protons and neutrons. Heavy nuclei
have a surplus of neutrons due to Coulomb repulsion of protons reducing the binding energy. Hence the
potential well for neutrons has to be deeper than that for protons and protons are on average less strongly
bound.
4πV 2
n(p)dp = dn = p dp (9.1)
(2πh̄)3
The number density is found by integrating up to the Fermi momentum pF :
V (pF )3
n= (9.2)
6π 2 h̄3
For protons and neutrons there are two spin states to consider and hence:
V (ppF )3 V (pnF )3
Z= 3 N = (9.3)
3π 2 h̄ 3π 2 h̄3
4 3 4 1 4
V = πR = π(R0 A 3 )3 = πR03 A (9.4)
3 3 3
1
Where the important result from Eq. 8.7 R ∼ R0 A 3 with R0 ∼ 1.2 fm has been used.
96
9.1.1 Fermi momentum and energy

If the stability line is approximated with Z = N = A/2 a relation of the Fermi momentum and energy to the
measured radius can be derived:
V (ppF )3 V (pnF )3
A/2 = Z = 3 =N = (9.5)
2
3π h̄ 3π 2 h̄3
13
h̄ 9π
→ pF = ppF = pnF = (9.6)
R0 8
In reality the Fermi momenta for protons and neutrons differ due to the difference in their potentials caused
by the Coulomb repulsion felt by protons. Assuming non-relativistic kinematics, Eq. 8.7 and an approximate
nucleon mass of M ∼ 1 GeV it is found:
p2F
EF = ∼ 33 MeV (9.7)
2M
9.1.2 Origin of the asymmetry term

The asymmetry term in the SEMF (Eq. 8.9) could not be derived from the liquid drop model but was added
adhoc. The Fermi gas model can motivate this term.
The average kinetic energy per nucleon is given by:
R pf R pf p2 2
E p2 dp
o R kin
p dp 3 pf 2 3
⟨Ekin ⟩ = pf 2 = oR pf2M2 = = pf 2 ∼ 20 MeV (9.8)
o
p dp o
p dp 5 2M 10M
For the average total kinetic energy of a nucleus the kinetic energies of all nucleons have to be added:
3
N · (pnF )2 + Z · (ppF )2

Ekin (N, Z) = N · ⟨En ⟩ + Z · ⟨Ep ⟩ = (9.9)
10M
Using Eq. 9.3 and Eq. 9.4 this yields:
2 23 5 5
!
3 h̄ 9π N3 +Z3
Ekin (N, Z) = 2 (9.10)
10M R0 4 A3
The total kinetic energy of a nucleus with fixed mass number A is given by:
2 23 5 5
!
(A − Z) 3 + Z 3

3 h̄ 9π
Ekin (N, Z) = 2 (9.11)
10M R0 4 A3
Here N = A − Z was used. This formula has a minimum at Z = A/2 = N . Taylor expanding around the
minimum Z = A/2 = N , we find:
2 23 2 !
3 h̄ 9π 5 A2 − Z
Ekin (A, Z) = A+ + ··· (9.12)
10M R0 8 9 A
97
And substituting back N = A − Z

2 23 2
!
5 (N − Z)

3 h̄ 9π
Ekin (N, Z) = A+ + ··· (9.13)
10M R0 8 9 A
The first term gets absorbed in the volume term and the second term reproduces the asymmetry term
of the SEMF. The asymmetry term prefers nuclei with an equal number of protons and neutrons. However
Coulomb repulsion reduces binding energy for protons and dominates for heavier nuclei. Hence due to the
Coulomb term heavy nuclei show an excess of neutrons over protons, despite the asymmetry term.
9.2 Shell Model

The Shell Model is based very closely on ideas from atomic physics : e.g. the orbital structure of atomic
electrons. Atomic energy levels are given by the main quantum number n = 1, 2, 3, . . .. In atomic physics
there are energy-degenerate levels with orbital angular momentum l = 0, 1, 2, . . . , (n − 1) for a given level n.
For any l there are (2l + 1) sub-states with different values of the magnetic quantum number – the projection
of l along any chosen axis: magnetic quantum number ml = −l, −l + 1, . . . , 0, 1, . . . , l − 1, l. Due to the
rotational symmetry of the Coulomb potential these sub-states will be degenerate in energy. Since electrons
have spin 1/2 each state can be occupied by 2 electrons, corresponding to spin projection ms = ±1/2. So in
summary, atomic levels are described by the quantum numbers (n, l, ml , ms ) and have nd degenerate states:
n−1
(2l + 1) = 2n2 . The description of energy levels in the nuclear potential will have some analogies
P
nd = 2
l=0
in the structure of atomic orbits, but in nuclear physics we are not dealing with the same simple Coulomb
potential.
9.2.1 Evidence
Magic Numbers
Nuclides with a certain number of neutrons or protons are particularly tightly bound. These numbers (2,
8, 20, 28, 50, 82, 126, 184) are known as magic numbers. Magic nuclei are made of a magic number
of protons or neutrons. Magic nuclei have a larger nuclear excitation energy (see Fig. 9.1) and for magic
neutron numbers a smaller neutron capture cross section. As Fig. 9.2 shows, magic nuclei also tend to show
the largest disagreements with the SEMF.
For doubly magic nuclei both the number of protons and neutrons is magic. Examples are: 42 He2 , 16 8 O8 ,
40 48 208
20 Ca 20 , 20 Ca 28 , 82 Pb126 . The existence of such magic numbers suggests a shell structure of energy levels.
9.2.2 First Models: Nuclear Potentials

The spherical potential well is a good starting point. The symmetry of the problem allows us to solve the
Schrödinger equation by separation of variables:
ψ(r, θ, ϕ) = R(r)Θ(θ)Φ(ϕ) (9.14)
As only R(r) depends on the potential V (r) a universal solution for the angular part can be found:
d2 Φ 1
+ m2 Φ = 0 → ϕm = √ eimϕ (9.15)
dϕ2 2π
98
Figure 9.1: Nuclear excitation energies for different numbers of neutrons and protons.
m2

1 d dΘ
sin θ + l(l + 1) − Θ=0 (9.16)
sin θ dθ dθ sin2 θ
where l = 0, 1, 2, ... and m = −l, ..., −2, −1, 0, 1, 2, ..., l. The general solutions Θlm (θ)Φm (ϕ) = Ylm (θ, ϕ)
combine to form the well known spherical harmonics Ylm (θ, ϕ), which are, for example, responsible for
spatial properties of atomic orbitals. This general solution in the form of spherical harmonics Ylm (θ, ϕ) stays
valid for various forms of the nuclear potential V (r). Calculations using the simple square potential well or
the harmonic oscillator potential do not lead to energy levels consistent with the observed magic numbers.
A more realistic potential is given by the Woods–Saxon potential, an intermediate from between a simple
square potential well and the harmonic oscillator potential.
V0
VWS (r) = (9.17)
1 + exp( r−a
b )
The form and parameters of this potential are obtained from the form factor measured in scattering experi-
ments (see Section 8.2). While somewhat better, this potential alone still does not describe observations.
9.2.3 Spin Orbit Coupling

The last missing piece is the addition of the spin-orbit coupling. This calculation was first done by Maria
Goeppert-Mayer. She received the Nobel prize in 1963 for her work on the nuclear shell model. The potential
has to be modified:
⃗ · S⟩
⟨L ⃗
V (r) = VWS + Vls (r) 2
h̄
99
Figure 9.2: A comparison of the SEMF predicted binding energy per nucleon B/A with measured data.
Using:
⃗ + L)
J 2 = (S ⃗ 2 ⃗ ·S
= S 2 + 2L ⃗ + L2 (9.18)
2
J −S −L 2 2 ⃗ ·S
= 2L ⃗ (9.19)
j(j + 1) − s(s + 1) − l(l + 1) = ⃗ ·S
2L ⃗ (9.20)
an expression for the split can be found:
⃗ · S⟩
⟨L ⃗ j(j + 1) − l(l + 1) − s(s + 1)
l
for j = l + 1
= = 2 2 (9.21)
2 (l+1) 1
h̄ 2 − 2 for j = l − 2
The resulting energy levels can be seen in Fig. 9.3. There is a strong splitting of levels with different total
angular momentum j and the splitting increases with orbital angular momentum l.
2l + 1
∆Els = ⟨Vls (r)⟩
2
It is important to note, that levels with larger j have lower energy (e.g. in the notation of Section 9.2.4
the level 1p 32 is lower than 1p 12 ). Each state of total angular momentum J has a degeneracy of 2J + 1
given by the number of projections of J along any chosen axis (from −J to +J). So for example 1d 52 can
accommodate a total of 6 nucleons.
100
Figure 9.3: Energy levels predicted by a nuclear shell model using Woods–Saxon potential with spin-orbit
coupling.
9.2.4 Nomenclature
A shell configuration in nuclear physics uses a slightly different notation than atomic physics. A level filled
with N nucleons is written as:
N
(nlJ ) (9.22)
where n is the main quantum number, l is the orbital angular momentum and J is the total angular momen-
tum of the level. The orbital momentum uses letters to indicate l values:
l = 0: s; l = 1: p; l = 2: d; l = 3: f ; l = 4: g; l = 5: h; l = 6: i
This “spdf ” notation has its origin in atomic spectroscopy, where sharp, principle, diffuse and fundamental
spectroscopy lines were distinguished.
2 4
2 6 1
N : 1s 21 1p 32 1p 21
1d 25 2s 12
2 4 2 6 1
Z : 1s 12 1p 32 1p 21 1d 25 2s 12
101
40 17 16
Figure 9.4: Examples of shell configurations for 18 Ar, 7 N and 7 N
17
Figure 9.5: Examples of shell configurations for the excited states of 7 N.
The symmetries and degeneracy from nuclear potentials are different from Coulomb potential used to
describe atomic levels. The meaning of the quantum numbers is not exactly the same. In particular states
like 1d with n < l can exist, that are forbidden in the Coulomb case. Here 1d is defined as the lowest lying
state with l =“d”= 2.
Each state of total angular momentum J has a degeneracy of 2J + 1 given by the number of projections
of J along any chosen axis (from −J to +J). So for example 1d 52 can accommodate a total of 6 nucleons.
102
9.2.5 Pairing Hypothesis

In a fully filled shell, for every nucleon with a projection of J along a chosen axis mJ a nucleon with the
opposite projection −mJ exists and the two nucleons pair up to combine to zero total angular momentum.
Empirically the same is true for all even pairings of nucleons. This is the pairing hypothesis.
9.2.6 Predictions: Spins in the Shell Model

The spin of a nucleus is given by the angular momentum contributions of the constituent nucleons. According
to the pairing hypothesis all nuclei with fully filled shells as well as all even–even nuclei will have spin 0.
For odd–even and even–odd nuclei the spin will be given by the total angular momentum of the unpaired
nucleon. In odd–odd nuclei with both an unpaired proton and neutron, the angular momenta have to be
added. This means an exact prediction is not possible, but the spin sN of the nucleon will be in the range
|Jp − Jn | <= sN <= Jp + Jn . Where Jp is the total angular momentum of the unpaired proton and Jn the
total angular momentum of the unpaired neutron. Examples are given in Fig. 9.4: For 40 18 Ar all nucleons are
paired and the spin of the nucleus is predicted to be 0. For 17 7 N the unpaired proton occupies level 1p1/2
hence the spin is given by J = 1/2. In 16 7 N an unpaired proton occupies level 1p1/2 while an unpaired
,
neutron is found in level 1d5/2 , hence the spin will be in the range from |1/2 − 5/2| = 2 to 1/2 + 5/2 = 3.
9.2.7 Predictions: Parity in the Shell Model

The intrinsic parities of proton and neutron are +1 and parity due to orbital angular momentum is given by:
l
P = (−1) (9.23)
Parity is a multiplicative quantum number:
Y
P = Pi (9.24)
Hence a pair of particles with the same orbital angular momentum l will have even (+1) parity:
l l 2
P = (−1) × (−1) = (−1) l = +1 (9.25)
So again, a pairing hypothesis can be formulated: even–even nuclei have even parity. For even–odd and
odd–even nuclei, the last unpaired nucleon determines parity according to Eq. 9.23.
In an odd–odd nucleus with both an unpaired proton and neutron, one has to remember that parity is a
multiplicative quantum number:
Pnucleus = Plast−proton × Plast−neutron (9.26)
Examples are again given in Fig. 9.4: For 40 18 Ar all nucleons are paired and the parity of the nucleus is
predicted to be even. For 17 7 N the unpaired proton occupies level 1p1/2 hence l =“p”= 1 and the parity
P = (−1)1 is odd. In 16 7 N an unpaired proton occupies level 1p1/2 , while an unpaired neutron is found in
level 1d5/2 , hence for the proton we find again lp =“p”= 1 and Pp = (−1)1 , while for the neutron ln =“d”= 2
and Pn = (−1)2 = 1. So 16 7 N has odd parity P = Pp × Pn = −1 × 1 = −1.
103
9.2.8 Predictions: Magnetic Moments in the Shell Model

The magnetic moment of a nucleus is given by:
A
⟨⃗j⟩ µN X ⃗
µ
⃗ = gj µN = {gl li + gs s⃗i } (9.27)
h̄ h̄ i=0
Where µN = 2m eh̄
p
is the nuclear magneton and gj is the Lande g-factor, s⃗i the spins and ⃗li the orbital angular
momenta of the contributing nucleons. As protons and neutrons are composite particles their g-factors are
non-trivial to compute. Unfortunately predictions of the shell model for nuclear magnetic moments are not
very reliable, this is likely due to deformations that have to be taking into account, leading to the collective
model, briefly discussion in Section 9.3.
9.2.9 Excited States

Excitations from the ground state are possible by promoting nucleons from their ground state level into
higher levels. An examples of this can be seen in Fig. 9.5. Comparing the level differences with Fig. 9.4 the
first excited state is likely promoting the proton from 1p 23 to 1p 12 :
2 4 2
N : 1s 21 1p 32 1p 12
2 3 2
Z : 1s 12 1p 32 1p 12
However predicting the correct order of these excited states proves to be difficult.
9.3 Collective Model

Comparing predictions of excited states with measurements shows that expected excited states do exist but
not always in precisely the order anticipated by the shell model. Especially the calculations of higher excited
states is much more complicated.
The collective model is an attempt to bring together elements of the shell and liquid drop models. The
interactions between nucleons leads to permanent deformation of the potential. Deformation represents
collective motion of nucleons and are related to liquid drop model. There are two major types of collective
motion:
• Vibrations: Surface oscillations

• Rotations : Rotation of a deformed shape
Recent encouraging developments in nuclear calculations are due to progress in computing power.
9.4 Theory of Nuclear β-Decay

To understand the β-decay momentum spectrum the Fermi Theory of β-decay can be used. Fermi treated
the decay as a 4-point interaction, while it is now known that a weak W -boson is exchanged. The Fermi
104
coupling constant GF of the 4-point interaction is given by

√ 4π(h̄c)3 αW
GF = 2 = 1.166 × 10−5 GeV−2 (h̄c)3
(MW c2 )2
Fermi’s second Golden rule gives the transition rate ω as:
2π
ω= |Mf i |2 n(E)
h̄
Where Mf i is the matrix element going from initial state i to final state f and n(E) is the density of
states. In the case of a simple 4-point interaction with coupling GF the matrix element can be written as
Mf i ∼ GVF .
Energy conservation in β-decay yields:
E0 = Ee + Eν + Erecoil ∼ Ee + Eν (9.28)
Fermi’s Golden rule can be written as:

2π 2π
dω = |Mf i |2 nν (Eν )ne (Ee )dEe = |Mf i |2 nν (E0 − Ee )ne (Ee )dEe (9.29)
h̄ h̄
The density of states is given as usual:
4πV 2
n(p)dp = dn = p dp (9.30)
(2πh̄)3
4πV
dn = n(E)dE = pEdE (9.31)
(2πh̄)3 c2
Actually the density of states depends on two energies Ee and Eν : n(Ee , Eν )dEe dEν leading to a double
differential d2 n. The two energies are not independent and we need to enforce energy conservation using a
delta function.
4πV 4πV
d2 n = n(Ee , Eν )dEe dEν = 3 2
pe Ee dEe pν Eν dEν δ(E0 − Eν − Ee ) (9.32)
(2πh̄) c (2πh̄)3 c2
Z
4πV 4πV
dn = n(Ee , Eν )dEν dEe = 3 2
pe Ee dEe pν (E0 − Ee ) (9.33)
(2πh̄) c (2πh̄)3 c2
Inserting the results from Eq. 9.33 into Eq. 9.29 yields:
dω 2π 4πV 4πV
= |Mf i |2 pe Ee pν (E0 − Ee ) (9.34)
dEe h̄ (2πh̄)3 c2 (2πh̄)3 c2
2
2π GF 4πV 4πV
= 3 2
pe Ee 3 pν (E0 − Ee ) (9.35)
h̄ V (2πh̄) c (2πh̄) c2
4
= G2F pe Ee pν (E0 − Ee ) (9.36)
(2π)3 h̄7 c4
(9.37)
Differentiating the usual energy momentum relation E 2 = p2 c2 + m2 c4 yields EdE = pc2 dp which can be
used to translate dp into dE.
105
dω dEe dω 4
= = G2F p2e pν (E0 − Ee ) (9.38)
dpe dpe dEe (2π)3 h̄7 c2
GF 2 2
= pe pν (E0 − Ee ) (9.39)
2π 3 h̄7 c2
Eν (E0 −Ee )
For massless neutrinos pν = c = c
dω G2F 2 G2F 2 G2F

= 7 2 pe pν (E0 − Ee ) ∼ 7 3 pe Eν (E0 − Ee ) = p2e (E0 − Ee )2 (9.40)
dpe 3
2π h̄ c 3
2π h̄ c 2π 3 h̄7 c3
9.4.1 Screening Factor

For the final spectrum result the effect of the Coulomb potential needs to be taken into account. This is
done using the screening factor F (Z, pe ):
2πη Ze2
F (Z, pe ) ∼ where η = ± for β ± (9.41)
1 − e−2πη 4πh̄βe
As shown in Fig. 9.6 positrons are repelled by the positive charges of the protons, while electrons are
attracted. The β − spectrum is shifted towards lower momenta and the β + spectrum towards higher momenta.
The effect increases with the charge and hence the number of protons Z in the nucleus.
Figure 9.6: The screening factor F (Z, Ee ) for electrons β − and positrons β + . Taken from Hyperphysics.
9.4.2 Kurie Plot

Combining Eq. 9.41 with the original calculation Eq. 9.40 yields the final β-spectrum:
dω F (Z, Ee )G2F p2e (E − Ee )2
= (9.42)
dpe 2π 3 h̄7 c4
106
The Kurie plot, proposed by Franz Kurie, plots the following H(Ee ) function versus the kinetic energy
of the β particles. s
dω 1
H(Ee ) = ∼ (E − Ee ) (9.43)
dpe pe 2 F (Z, pe )
Hence for massless neutrinos the Kurie plot shows a simple line.
Non-zero neutrino mass

For massive neutrinos the relation Eν = cpν does not hold, instead Eν2 = p2ν c2 +m2 c4 . Going back to Eq. 9.39
s
dω 1 p q p
H(Ee ) = 2
∼ pν (E0 − Ee ) ∼ (E0 − Ee ) (E0 − Ee )2 − m2ν c4 (9.44)
dpe pe F (Z, pe )
Figure 9.7: Kurie plot for different non-zero neutrino masses.
107
Chapter 10
Fission and Fusion
10.1 Fission and Fusion: Principles

As you have learnt in the module on “Nuclear Phenomenology”, the nucleons in nuclei are bound together
with an attractive force (the nuclear strong force), therefore it takes energy to split the nucleus up into
individual nucleons. This energy is known as the binding energy of the nucleus. Since the energy of the
bound nucleus plus the binding energy is equal to the energy of the individual nucleons, the total energy of
the nucleus is the energy of the free nucleons minus the binding energy. And since energy is equal to mass
(times c2 ) we say that the mass of a nucleus is smaller than the mass of the individual nucleons, and the
difference in mass is due to the negative binding energy.
The basic principle behind nuclear fission and fusion is to take advantage of the difference in the binding
energy per nucleon between nuclei with different nucleon numbers. This is seen in Fig. 10.1, which shows that
the binding energy per nucleon versus the nucleon number increases rapidly for the lowest nucleon numbers
(with some spikes seen for magic nuclei), reaches a maximum at 56 Fe then slowly decreases again. It can be
seen that splitting one of the heaviest nuclei into two lighter nuclei results in a larger total binding energy,
and since this energy is negative it results in a smaller total energy (or mass). The leftover energy can then be
released as kinetic energy, which can be utilised, essentially unlocking some of the energy inside the nucleus.
This is the basic principle behind nuclear fission. It is also clear that if two of the lightest nuclei can be fused
together to form a heavier nucleus, the total (negative) binding energy will be larger and hence the total
energy will be smaller, releasing some kinetic energy. This is the basic principle behind nuclear fusion.
10.2 Fission
10.2.1 Basics
For a nucleon number larger than 100 it can be energetically favourable for massive nuclei to split into lighter
daughter nuclei, with the extra energy carried off as kinetic energy of the fission products. An example is:
235
92
U → 92
37
Rb + 140
55
Cs + 3n + X MeV, (10.1)
although it should be noted that this is not the only fission process that 235 92
U undergoes. The decay
products of fission will themselves decay, for example via β-decay and the resulting nuclei will often give off
108
Figure 10.1: Binding energy per nucleon versus nucleon number.
additional neutrons. These neutrons are known as delayed neutrons as they will be released much later than
the prompt neutrons of Equation 10.1. We will see in Section 10.2.4 that these delayed neutrons are vital for
the safe running of nuclear fission reactors.
Let us breakdown the energy released using the Semi-Empirical Mass Formula (SEMF). The difference in
binding energies between the initial and final states is given by the difference in energies for each correction
term of the SEMF.1 We will now go through each correction term in turn (although we ignore the very small
pairing term):
1. Volume term: −av (AU − ARb − ACs ) = −46.7 MeV

2/3 2/3 2/3
2. Surface term: +as AU − ARb − ACs = −160 MeV

2 2 2
ZU ZRb ZCs
3. Coulomb term: +ac 1/3 − 1/3 − 1/3 = +339 MeV
AU ARb ACs

(ZU −AU /2)2 (ZRb −ARb /2)2 (ZCs −ACs /2)2
4. Asymmetry term: +aa AU − ARb − ACs = +26.1 MeV
where we have used av = 15.56, as = 17.23, ac = 0.697 and aa = 93.14. So the total difference in energy
released at this stage is 158 MeV. In fact the daughter nuclei also release some energy, due to the subsequent
decays discussed above, so the total energy released is actually ≈ 200 MeV per fission.
1 The total number of nucleons is the same before and after the decay, so it is only the correction terms we need to consider.
109
10.2.2 Spontaneous and Induced Fission

Spontaneous fission is when the fission process occurs without external action. In the SEMF we have assumed
that the nucleus is spherical, because this minimises the surface area and hence the energy from the surface
term. If we imagine the process of fission starting with the deformation of a spherical nucleus into a prolate
shape, leading to the eventual splitting into two nuclei, then it is important to consider how the binding
energy changes as the nucleus deforms. As the nucleus stretches into a prolate shape, the surface term will
increase and the Coulomb term will decrease (assuming the volume remains the same). If the change in
Coulomb term is larger than the change in the surface term then the deformed shape will be energetically
favourable and the nucleus is unstable, meaning that spontaneous fission can occur. It turns out that this is
2
only the case for ZA ≥ 2aac ≈ 49, which is only the case for very heavy nuclei with Z > 116 and A ≥ 270.
s
Fission can also occur in lighter nuclei, but this requires penetrating a potential barrier. Since the energy
of the nucleus increases as it deforms (but then decreases again once it has split) there is a barrier that
must be overcome. This is shown in Figure 10.2, which shows as a solid line the energy of a nucleus with
Z2 2as
A < ac . It can be seen that the energy increases as the nucleus deforms then falls off again as it splits (note
also the dashed line, which shows the energy decreasing through the process in the case of the very heavy
2
unstable nuclei with ZA ≥ 2a ac ). The difference in the energies for the spherical nucleus and the maximum of
s
the curve is known as the activation energy (or fission barrier), which is required in order to induce fission.
Spontaneous fission can occur in this case via a quantum mechanical tunnelling through the barrier, however
the probability is extremely small. For example for 238 92
U the transition rate for spontaneous fission is
−24 −1 −18 −1
3 × 10 s compared to about 5 × 10 s for α decay.
Figure 10.2: Energy of a nucleus during the fission process.
Another possibility is to supply the energy needed to overcome the barrier. This is known as induced
fission and is achieved by supplying a flow of neutrons. Since neutrons are electrically neutral they can
approach the nuclei and be attracted by the strong nuclear force. For heavy nuclei, such as uranium, the
activation energy is only about 6 MeV. This energy can be supplied with a flow of low energy neutrons that
induce neutron capture reactions. They push the nucleus into an excited state above the fission barrier and
it then splits up. When a nucleus absorbs a neutron some energy is released due to the binding energy of
that neutron, if this energy is as large as the activation energy then fission can be induced. The binding
energy of the last added neutron is not the same as the average binding energy per nucleon (this is explained
by the shell model). For uranium it is a little less, and it depends on how adding another neutron affects the
110
pairing term. For example, the capture of a neutron by 235 92

U (which leads to 236 92
U ) converts an even–odd
nucleus to a more tightly bound even–even nucleus due to the larger pairing term. Conversely, the capture of
a neutron by 23892
U (which leads to 23992
U ) converts an even–even nucleus to a less tightly bound even–odd
one. The binding energy of the last neutron is 6.5 MeV for 236 92
U and 4.8 MeV for 239 92
U (the difference
between them is due to the difference in the change in the pairing term). The activation energy for fission in
235
92
U is 5 MeV (less than the energy supplied by neutron capture) and that in 238 92
U is 6 MeV (more than
the energy supplied by neutron capture). For this reason fission can be induced in 235 92
U by slow neutrons,
in fact the kinetic energy of the neutrons can be zero as a neutron capture will always provide enough energy
to induce fission. Conversely, for fission to be induced in 238 92
U neutrons with at least 1.2 MeV of kinetic
energy (fast neutrons) are required. Materials such as 23592
U that require only slow neutrons to induce fission
are known as fissile materials, other examples include 233 92
U , 239
94
Pu and 24194
Pu . Materials such as 238 92
U
232
that require fast neutrons to induce fission are known as non-fissile materials. Other examples include 90 Th
, 240
94
Pu and 24294
Pu .
10.2.3 Nucleus–neutron cross sections

Since neutrons are used to induce fission it is useful to look at the nucleus–neutron cross section for uranium,
the most commonly used element in fission reactors. Figure 10.3 shows the total and fission nucleus–neutron
cross sections for 235 92
U and for 238 92
U . The total cross section includes contributions from elastic and
inelastic scattering, and from neutron absorption. If the neutron is absorbed an excited state is formed that
will either emit one or more particles, e.g. neutrons, protons, α particles; de-excite by emitting photons;
or undergo fission. There are resonant regions in the neutron absorption cross section where the compound
nucleus produced after the neutron absorption is in a region of excitation where its energy levels are well
separated.
There are a number of things to note about Figure 10.3. Firstly we can see that the fission cross section
for 238
92
U is only relevant if the neutron has a kinetic energy above the activation energy of 1.2 MeV, whereas
for fissile 235
92
U the fission cross section is largest for small energy neutrons. In fact the fission cross section
is 84% of the total cross section at 0.1 eV. For 238 92
U the total cross section is dominated by elastic and
inelastic scattering and shows little dependence on energy except in the resonant region between 1 eV and
1 keV, where is becomes likely that a neutron will be captured into a resonance with radiative decay of the
excited state.
For 23592
U the total cross section at low energies is dominated by fission, with a sizeable contribution from
radiative capture of the neutron with the formation of the excited state 236 92
U plus one or more photons.
There is a clear resonance region, where in this case the resonance will decay via fission. At higher energies
the total cross section is dominated by elastic scattering and inelastic excitation of the nucleus.
10.2.4 Fission chain reactions

We have seen that neutrons can induce fission and also that fission produces neutrons. This leads to the idea
of a chain reaction in which one fission reaction leads to one or more fission reactions due to the produced
neutrons. If the probability that a given neutron induces fission is q and each fission reaction produces an
average of n neutrons, then each neutron will lead to (nq − 1) additional neutrons in a time tp , where tp is the
average time before absorption of the neutron occurs. If there are N (t) neutrons present at time t, then at
N (t+δt)−N (t)
time t+δt there will be N (t)+N (t) (nq − 1) δt/tp neutrons present and therefore dNdt = limt→0 =
dN (nq−1)dt R N (t) dN
R δt
t (nq−1)dt
N (t) (nq − 1) /tp giving N (t) = tp . Integrating between t = 0 and t = t gives: N (0) N (t) = 0 tp
111
(a) (b)
Figure 10.3: Nucleus–neutron cross sections versus the neutron kinetic energy for (a) uranium-235 and (b)
uranium-238.
(nq−1)t
giving ln [N (t)] − ln [N (0)] = tp leading to:
N (t) = N (0) e(nq−1)t/tp . (10.2)
This leads to three possible scenarios, depending on the exact value of nq:
1. For nq < 1, N (t) decreases exponentially, the process is said to be subcritical and the reactions will
soon die out.
2. For nq = 1, N (t) remains constant, the process is said to be critical and the conditions are right for a
sustained, controlled reaction as is required in a nuclear power plant.
3. For nq > 1, N (t) increases exponentially and the process is said to be supercritical. The energy will
grow very rapidly, leading to an explosion, exactly what is needed in a nuclear fission bomb.
10.2.5 Nuclear fission bombs

Before starting this section I would like to note that the topic of nuclear fission bombs is obviously extremely
sensitive. They have had devastating effects on peoples lives and the teaching of the topic in no way condones
their use, it is simply here to demonstrate how nuclear fission has been utilised and show the physics behind
the weapons.
First we consider the conditions required to create a supercritical reaction. For simplicity we consider
pure 23592
U , which has an average n ≈ 2.5 and tp ≈ 10−8 seconds (note that the process in Equation 10.1
is only one way that fission can occur). For supercritical conditions we require q > 0.4. One of the key
determining factors in the design of a bomb is the size of the metal. If it is small enough that neutrons are
likely to reach the edge of the volume before tp , then the reaction will die out. A neutron with 2 MeV kinetic
energy (which is the average energy of a neutron coming from fission) will move approximately 20 cm in a
straight line in 10−8 seconds. Due to the relatively small kinetic energy compared to the neutron rest mass
(mn = 939.57 MeV/c2 ) we can estimate this using the classical expression for kinetic energy: 12 mn v 2 , so that
2
we have: 2 MeV = 12 × 939.57 MeV/c2 × v 2 giving: 0.0043 = vc2 which gives v = 2 × 107 ms−1 and hence
112
the distance travelled is 20 cm. However, the neutron will not travel in a straight line because an average of
6 collisions is expected before inducing fission, and if we assume that the direction of the neutron changes
randomly after each collision, the average distance travelled is actually 7 cm. However, not all these neutrons
will induce fission, as some will escape the material and some will be captured in nuclei without inducing
fission. This means that a larger sphere is required and it turns out that a radius of 9 cm, corresponding to
a critical mass of 50 kg, is required to ensure a supercritical reaction. A bomb is usually made by starting
with a subcritical mass of material then using a mechanism to increase its size when the bomb is to go off.
10.2.6 Nuclear fission power reactors

In order to produce power from nuclear fission, a critical reaction is required, with nq = 1. As we will see,
any small deviations above unity would lead to a huge increase in the power output in a very short time.
We will consider a thermal reactor, which uses uranium as the fuel and low-energy neutrons to establish a
chain reaction. Natural uranium contains only 0.7% 235 92
U and 99.3% 238 92
U , so a neutron is much more likely
238
to interact with a nucleus of 92 U . A 2 MeV neutron has little chance of inducing fission in natural uranium.
It is more likely to scatter inelastically with a 238
92
U nucleus, losing energy in the process (see Figure 10.3(b)).
After a couple of such interactions the neutron energy will be below the threshold of 1.2 MeV for inducing
fission in 238
92
U . Such a neutron is much more likely to be captured into one of the 238 92
U resonances, with
the emission of photons, before inducing fission in one of the rare 235 92
U nuclei. There are two possible ways
to ensure a critical reaction in uranium:
235
1. Enrich the uranium so that it contains 2–3% of 92
U , making it more likely that a neutron will induce
fission in a 235
92
U nucleus.
2. Surround the natural uranium fuel in a large volume of moderator material, which has the job of slowing
down the fast neutrons produced in fission via elastic collisions. The slow neutrons are not energetic
enough to be absorbed into a 23892
U resonance. It will therefore be more likely to eventually meet and
induce fission in a 235
92
U nucleus, the cross section for which is high at low energies (see Figure 10.3).
Usually heavy water (D2 O) or graphite are used as moderator materials.
Vital to the safe running of a nuclear fission reactor is that it operates with precisely nq = 1. This is
achieved with control rods that are mechanically inserted into the reactor whenever the reaction needs to be
reduced (and removed if the reaction needs to be increased). They are usually made of cadmium, which has
a very high absorption cross section for neutrons. The problem with this mechanism is that tiny increases to
nq, from e.g. fluctuating temperatures, would very rapidly lead to an explosive reaction. In order to see this
we consider Equation 10.2. Typically tp ≈ 10−3 seconds in natural uranium (note it is much longer than the
equivalent time in pure 23592
U due to the smaller cross sections), so for a modest 0.1% increase in nq such
that nq = 1.001, the reactor flux would increase by e60 ≈ 1026 in only one minute, which is clearly a highly
unstable system. Mechanical insertion of the control rods is not possible on the timescales required to control
such rapid increases in flux. It is therefore necessary to have a much smaller rate of increase in a safe nuclear
reactor. The key to this is to consider the affect of delayed neutrons, which appear on average ≈ 13 seconds
after the fission. These delayed neutrons come from the fission of decay products that have undergone a
series of β decays. Taking account of delayed neutrons we need to replace nq with (nprompt + ndelayed ), where
nprompt is the number of prompt neutrons per fission and ndelayed is the number of delayed neutrons per
fission. To ensure that reactions remain constant we have the requirement (nprompt + ndelayed ) q = 1, where
ndelayed ≈ 0.02, giving a roughly 1% correction to nprompt . The idea is to keep nprompt q far enough below 1
113
that small variations in q cannot lead to an unstable system. However (nprompt + ndelayed ) q is kept very close
to 1, and it is the time scale of the delayed neutrons (≈ 13 seconds) that dictate the growth of the neutron
flux in the event of a deviation from criticality. These timescales are manageable for mechanical control of
the reactor via the insertion of neutron absorbing control rods.
10.3 Fusion
10.3.1 Basics
Fusion is the opposite process to fission, where two very light nuclei fuse to form a heavier nucleus, releasing
energy due to the fact that the (negative) binding energy per nucleon is larger for the fused nucleus. The
energy released in fusion is less than that in fission, but the light nuclei are much more abundant in nature,
making it an attractive source for possible future power generation.
10.3.2 Coulomb barrier

The practical problem with fusion comes from the fact that two positively electrically charged nuclei will
repel each other due to the Coulomb repulsion, stopping them getting close enough for the strong nuclear
force to take over allowing them to fuse. The Coulomb potential between two nuclei gives the amount of
energy required to overcome this Coulomb barrier and is given by:
′
1 ZZ e2
VC = ,
4πϵ0 R + R′
′ ′ 1
where Z and Z are the two atomic numbers and R and R are their effective radii. If we take R = 1.2A 3 fm
(see Eq. 8.7) then we obtain:
′ ′
e2 h̄cZZ ZZ
VC = h i = 1.198 h i MeV,
4πϵ0 h̄c 1.2 A1/3 + (A )
′ 1/3
fm A1/3 + (A′ )
1/3
2
where we have used α = 4πϵe0 h̄c = 1371
and h̄c = 0.197 GeV fm (see the “Basic Ideas in Particle Physics”
′ ′
notes). For simplicity we set A ≈ A ≈ 2Z ≈ 2Z , then VC ≈ 0.15A5/3 MeV and for A ≈ 8, VC ≈ 4.8 MeV.
This would be the amount of energy required to overcome the Coulomb barrier, and would be smaller for
lighter nuclei.
In order to overcome this barrier a confined mixture of nuclei can be heated enough so that the kinetic
energy supplies the required energy. Using the relationship between particle energy and temperature in a
medium, E = kB T , with Boltzmann’s constant kB = 8.6 × 10−5 eV K−1 , we obtain T ≈ 5.610 K for a particle
energy of 4.8 MeV. This is a very high temperature! Actually, fusion will occur at lower temperatures than
this for two reasons:
1. Quantum tunnelling: similar to α-decay, discussed in the “Nuclear Phenomenology” module, the prob-
−G
ability to penetrate an
penergy barrier via quantum tunnelling is e , where G is the energy dependent
Gamow factor: G = EG /E, where EG increases as the barrier increases and E is the energy of the
particles.
2. A collection of nuclei with a given mean energy will have a Maxwellian distribution of energies about
the mean and the distribution of the energies will have the form e−E/kB T .
114
So the effect from quantum mechanical tunnelling increases with energy and that from the energy distribution
of the particles decreases with energies. When these two effects are combined we end up with a reaction rate
that peaks at a certain energy and there is a small range of energies within which fusion takes place at a
non-negligible rate (see the lecture slides for a distribution of the reaction rate versus energy).
10.3.3 Stellar fusion

The energy of the sun comes from nuclear fusion reactions, and mostly from the proton–proton chain. The
temperature of the sun is 107 K and all the material is fully ionised and is referred to as a plasma. The
nuclei of hydrogen atoms (which are just protons) are the starting point for the process, which has the three
following steps:
1. Two hydrogen nuclei (protons) fuse to form a deuterium nucleus:
1
1
H + 11 H → 21 H + e+ + νe + 0.42 MeV.
1
2. The deuterium fuses with more hydrogen to produce helium and a photon: 1
H + 21 H → 32 He + γ +
5.49 MeV.
3. Two helium nuclei (from the above two steps occurring twice) fuse to form an α particle plus some
extra hydrogen nuclei: 32 He + 32 He → 42 He + 11 H + 11 H + 12.86 MeV.
In each step the energy released is the difference in binding energies
between the initial and
final states. The
relevant binding energies are: B 21 H = 2.224 MeV, B 32 He = 7.718 MeV and B 42 He = 28.3 MeV. For
the first step the differences in the masses of the particles in the initial and final states must also be taken
into account: me+ = 0.511MeV/c2 , mn = 939.566 MeV/c2 and mp = 938.272 MeV/c2 . You can convince
yourselves that accounting for these gives the energy differences seen above. The extra energy is in the kinetic
energy of the particles. The first step is slow as it involves the weak interaction and it therefore sets the
scale for the long lifetime of the sun. For a complete chain the first two steps occur twice providing the two
helium nuclei for the third step. The total energy released is: 2 × (0.42 + 5.49) + 12.86 = 24.68 MeV and the
overall interaction is given by:
4 11 H → 42 He + 2e+ + 2νe + 2γ + 24.68 MeV.

The produced positrons will annihilate with electrons in the plasma providing an extra 1.02 MeV of energy
per positron so the total energy is 26.72 MeV. However, each neutrino will carry off an average of 0.26 MeV
of energy, which will be lost in space. The rest of the energy is transported to the surface of the sun and
emitted as photons heloor ejected high-energy particles. The total energy radiated from the sun from the
proton–proton chain is therefore 6.55 MeV per proton. The proton–proton chain is the dominant fusion
process powering the sun, but other processes involving the fusion of heavier nuclei also occur producing
additional energy and changing the composition of elements in the sun to include the heavier elements.
10.3.4 Fusion reactors

Due to the abundance of light nuclei the idea of utilising the energy from fusion as a source of power is
a very attractive one. The proton–proton interactions that power the sun are very slow (due to the weak
interaction) so cannot be used. Possible candidates for fusion reactor processes are deuterium–deuterium
fusion and tritium–deuterium fusion. The former proceeds via: 21 H + 21 H → 32 He + n + 3.27 MeV or 21 H + 21 H →
3
1
H + p + 4.03 MeV and the latter via: 21 H + 31 H → 42 He + n + 17.62 MeV, which gives off more energy and also
115
has the advantage of having a larger cross section than deuterium–deuterium fusion. In all these reactions the
energy produced comes from the difference in binding energies before and after the reaction. The problem
with tritium–deuterium fusion is the low abundance of natural tritium, which has a half-life of only 17.7
years. By contrast deuterium is found in huge quantities in sea water.
Nonetheless, due to its advantages, tritium–deuterium fusion is considered as a candidate for fusion reac-
tors. The cross section for this process is reasonable when the nuclei have an energy of 20 keV, corresponding
2×104
to a temperature of T = kEB = 8.6×10 −5 = 2 × 10
8
K. The main problem with practical fusion is how to
contain plasmas at such high temperatures. Any material container would vaporize at these temperatures.
The solution is to contain the plasma using either:
1. magnetic confinement where the charged particles in the plasma follow a helical path as they curve
round a magnetic field with a direction that points in a circle around a doughnut shape or
2. inertial confinement where pulsed lasers bombard small pellets of a tritium–deuterium mixture in
many directions at the same time at very high energies.
Fusion research aims to try and make the ratio of energy output to energy input (required to reach such high
temperatures) greater than one so that fusion is feasible for a source of power. This is known as the Lawson
criteria. This ratio is larger for longer confinement times and higher particle densities. A value greater than
one has just been reached and research is ongoing to further improve this.
116
Chapter 11
Literature Guide
11.1 Basic Ideas

The book [4] has a good review of units in Chapter 2.1. The Yukawa potential is treated in Section 1.5 of
Ref. [1]. Cross section and decay can be found in Section 1.6 of Ref. [1].
11.2 Leptons, Quarks & Hadrons

This followed quite closely Ref. [1]:
• Chapter 3 Particle Phenomenology
• Section 3.1: Leptons
• Section 3.1.4: Neutrino Mixing and Oscillations
• Section 3.1.5: Oscillation Experiments and Neutrino Masses
• Section 3.1.6: Lepton Numbers revisited
• Section 3.2: Quarks
• Section 3.3.2: Quark Model Spectroscopy
For the Lepton part, some ideas are described in Section 10.1 (”Properties of Leptons”) & 10.2 (”The
Types of Weak Interactions”) of Ref. [2]. For the Quark section, some ideas are described in Section 8.1
(”Quarks in Hadrons”), 8.2 (”The Quark-Gluon interaction”) & 10.2 of Ref. [2]. An advanced treatise on
neutrino oscillations can be found here: [5]
Some useful information can be found online here: [3].
11.3 Experimental Methods I

The accelerator part can be found in Section 4.1 (”Experimental Methods: Overview”) and 4.2 (”Accelerators
and Beams”) of Ref. [1]. Ionisation energy loss and the Bethe–Bloch formula are discussed in Section 4.3.2
(”Ionisation Energy Loss”) of Ref. [1].
117
11.4 Experimental Methods II

Interactions of photons and electrons are described in Section 4.3.3 (”Radiation Energy Loss”), 4.3.4 (”In-
teractions of Photons in Matter”) and 4.4.5.1 (”Electromagnetic Showers”) in Ref.[1]. The Interactions of
hadrons is explained in Section 4.3.1 of Ref.[1]. The detector parts are rather weak in both books (Section
4.4 ”Particle Detectors” in Ref.[1] and A.2 ”Detectors” in Ref. [2]). A good resource is [7].
11.5 Strong Interactions

DIS is discussed in section 5.8 of Ref.[1] and Chapters 7.1 & 7.3 of [2]. The concept of colour is investigated
in 5.1,5.2 & 5.7 of Ref.[1] and Chapter 8 of [2]. The asymptotic freedom and the running of coupling constants
is discussed 5.4 & 5.6 of Ref.[1] and Chapter 8 of [2].
11.6 Electroweak Interactions

The relevant chapters for EWK interactions are Sections 6.1-6.4 and 9.3 in Ref. [1] and Chapter 10.5 & 12
in Ref. [2]. The Higgs discovery papers make a good read [8, 9].
11.6.1 Electroweak Interactions II Quark Mixing

The relevant chapters for quark mixing and the CKM matrix are Section 6.5.1 and 6.6 in Ref. [1] and Section
10.4 and 15.5 in Ref. [2]. A detailed but advanced discussion can be found here [10]
11.7 Nuclear Phenomenology

The relevant chapters of Ref. [1] are:
• Section 2.2 (”Nuclear Shapes and Sizes”)
• Section 2.3 (”Semi Empirical Mass Formula”)
• Section 2.5 (”Radioactive Decay”)
• Section 2.6 (”beta-decay phenomenology”)
• Section 2.7 (”Fission”).
A far more rigorous description of the shape of Nuclei is found in Chapter 5 (”Geometric Shape of Nuclei”) of
Ref. [2]. The discussions on alpha-decay are based on Section 7.6 of Ref. [1] and Section 3.2 (”Alpha Decay”)
of Ref. [2]. Spontaneaus fission follows Section 3.3 (”Nuclear Fission”) of Ref. [2] and Section 2.7 (”Fission”)
of Ref. [1].
11.8 Structure of Nuclei

The Fermi Gas model is described in Section 7.2 (”Fermi Gas Model”) of Ref. [1] and Section 18.1. (”The
Fermi Gas Model”) of Ref. [2].
118
11.8.1 Shell Model

The shell model followed partially Section 7.3 (”Shell Model”) in Ref. [1] and Section 18.3. (”The Shell
Model”) of Ref. [2]. A very good resource is also [6].
11.8.2 Beta decay revisited

The treatment of the beta-decay spectrum and Kurie plots followed Section 7.7 (”Beta Decay”) of Ref. [1],
with some parts from Chapter 16.6 of Ref. [2].
11.9 Fission & Fusion

Fission is discussed in Section 8.1 and Fusion in Section 8.2 of Ref. [1] with some elements discussed in
Chapter 20.5 of Ref. [2].
119
Bibliography
[1] B.R. Martin, Nuclear and Particle Physics, 2nd Edition, Wiley, West Sussex, 2009.
[2] B. Povh, K. Rith, Ch. Scholz, F. Zetsche, W. Rodejohann, Particles and Nuclei, 7th Edition, Springer,
Heidelberg, 2015.
[3] Particle Data Group, http://pdg.lbl.gov/.
[4] M. Thomson, Modern Particle Physics, Cambridge University Press, Cambridge, 2013.
[5] http://pdg.lbl.gov/2015/reviews/rpp2015-rev-neutrino-mixing.pdf
[6] M. Goeppert Mayer, Nobel Lecture,
http://www.nobelprize.org/nobel prizes/physics/laureates/1963/mayer-lecture.html
[7] F. Hartmann, A. Sharma, Multipurpose detectors for high energy physics, an introduction, Nucl. Inst.
Meth. A666 (2012) 1-9, http://www.sciencedirect.com/science/article/pii/S0168900211020626
[8] ATLAS Collaboration,Observation of a new particle in the search for the Standard
Model Higgs boson with the ATLAS detector at the LHC, PLB 716, 1, (2012), 1-29,
https://www.sciencedirect.com/science/article/pii/S037026931200857X
[9] CMS Collaboration, Observation of a new boson at a mass of 125 GeV with the CMS experiment at the
LHC, PLB 716, 1, (2012), 30-61, https://www.sciencedirect.com/science/article/pii/S0370269312008581
[10] Particle Data Group, http://pdg.lbl.gov/2017/reviews/rpp2017-rev-ckm-matrix.pdf
120

Lecture Notes

Uploaded by

Lecture Notes

Uploaded by

PHAS0040: Nuclear and Particle Physics

Prof. Emily Nurse, Prof. Andreas Korn, Prof. Matthew Wing

1 Basic Ideas of Particle Physics 5

2 Leptons & Quarks 23

4 The Strong Interaction 50

5 The Electroweak Interaction: Part I 64

5.2.3 The spin structure of weak interactions . . . . . . . . . . . . . . . . . . . . . . . . . . 67

6 Electroweak Interactions: Part II: Quark Mixing 70

7 The Electroweak Interaction: Part III 77

9.2.7 Predictions: Parity in the Shell Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

10 Fission and Fusion 108

11 Literature Guide 117

Basic Ideas of Particle Physics

1.1 Fundamental Particles

1.1.3 The muon

1.1.5 The Standard Model content

The Standard Model bosons are summarised in Table 1.1.

Boson Electrical charge spin

Fermion Mass [MeV/c2 ]

Table 1.2: The masses of the fundamental particles

1.2 Units and scales

1.2.1 Natural units

Energy, Momentum and Mass

0.511 × 10−3 × 1.6 × 10−10 J

1.3 Particle Physics is High Energy Physics

1.4 Particle interactions

1.4.3 Feynman Diagrams

Figure 1.4: Feynman diagrams for e+ e− elastic scattering.

1.5 Four-vectors and Invariants

For an energy-momentum four-vector we have:

and we know that E 2 = m2 c4 when a particle is at rest, therefore:

1.5.1 Centre-of-mass frame

W 2 = m2beam + m2target + 2Ebeam mtarget

1.6 Range of forces

of light to give: r ≤ ∆t × c ≤ Mh̄c Xc

1.6.1 Virtual Particles

1.7 Yukawa Potential

1.8 The scattering amplitude

1.9 Cross sections and luminosities

1.9.1 Differential cross sections: what we measure

1.9.2 Differential cross sections: predicting the rate

2Aderivation of this is beyond the scope of this course.

which rearranges to give:

as long as spin is neglected.

1.10 Unstable Particles and the Breit-Wigner formula

MX c2 → MX c2 − iΓh̄/2 in Equation 1.7, which leads to:

Leptons & Quarks

2.1.1 Lepton Number

L = N (l− ) − N (l+ ) + N (νl ) − N (ν¯l ); l = e, µ, τ (2.2)

2.1.2 Lepton decays

Γ(τ − → e− ντ ν̄e ) Γ(τ − )Br(τ − → e− ντ ν̄e )

2.1.3 Neutrino Oscillations

|νe ⟩ = |ν1 ⟩ cos θ + |ν2 ⟩ sin θ (2.8)

A more convenient form makes use of the following approximation:

Where we also used h̄c ∼ 197 MeV · fm.

The neutrino mixing matrix is known as the PMNS or Pontecorvo–Maki–Nakagawa–Sakata matrix:

0 sin θ13 e−iδ

Evidence for Neutrino Oscillation

sin2 (2θ12 ) = 0.851 ± 0.019 (2.29)

2.1.4 Absolute Neutrino mass scale

∆m2solar = ∆m221 = (7.53 ± 0.18) × 10−5 eV2 (2.32)

Neutrino-less double beta decay experiments

Figure 2.2: Feyman diagram of double beta decay.

2.1.5 Number of Neutrinos

2.1.6 Charged Lepton Number Violation

Figure 2.4: Feynman diagram for the process µ− → e− γ

number violating decay µ− → e− γ:

Br(µ → e− γ) = αBr(µ → e− νµ ν¯e )P (νµ → νe )

2.2.2 Hadron Decay

particle mass [MeV] Lifetime [s] main decays

Table 2.1: Some examples of Mesons

particle mass [MeV] Lifetime [s] main decays

Λ+c (udc) 2285 2.1 × 10−13 Several

Table 2.2: Some examples of Baryons

3.2.1 Centre-of-mass energy

Neglecting masses: P12 = P22 = 0 and the centre of mass energy