Jets, Hadronization, and Testing QCD

Confinement forbids a free quark, yet high-energy collisions manifestly produce quarks and gluons. The resolution is that a struck parton does not travel far before its color flux tube breaks into hadrons, and because the fragmentation is soft it barely deflects the parton's momentum. The result is a jet: a narrow cone of hadrons whose total momentum reconstructs the parton that produced it. Jets make the quarks and gluons of the Lagrangian visible as collimated sprays, and their rates and shapes are among the most precise tests of QCD. This lesson reads the jet structure of electron-positron annihilation, identifies the gluon in three-jet events, and extracts $α_{s}$ from the pattern.

Throughout, $ℏ = c = 1$ .

Two-jet events

The reference reaction is $e^{+} e^{-} \to γ^{*} \to q \overset{q}{ˉ}$ , the quark version of the muon-pair process of the QED tree lesson. The quark and antiquark emerge back to back in the center-of-momentum frame, each fragmenting into a jet. Two observations confirm the partonic origin. First, the two jets are collinear and opposite, as required by momentum conservation for a two-body partonic final state. Second, the jet axis follows the angular distribution

\frac{d σ}{d Ω} \propto 1 + cos^{2} θ,

the distribution for producing a spin- $\frac{1}{2}$ pair from a virtual photon. A spin- $0$ pair would give $sin^{2} θ$ instead. The measured $1 + cos^{2} θ$ confirms that the jets originate from spin- $\frac{1}{2}$ quarks.¹

The total annihilation rate into hadrons, summed over the accessible flavors and colors, is the $R$ -ratio measured in the quark model; the two-jet angular shape confirms the spin of the constituents that $R$ counts.

A two-jet event in electron-positron annihilation. The virtual photon produces a quark and antiquark back to back; each fragments into a collimated spray of hadrons along the original parton direction. The jet axis follows the one-plus-cosine-squared distribution characteristic of a spin-half pair.

Three-jet events and the gluon

A quark can radiate a hard gluon before hadronizing, $e^{+} e^{-} \to q \overset{q}{ˉ} g$ , and when the gluon is energetic and well separated it fragments into a jet of its own. The event then has three jets in a plane. Three-jet events were observed at the PETRA collider at DESY in 1979 and provided the first direct evidence for the gluon: their rate is fixed by $α_{s}$ , and their angular correlations match a spin- $1$ radiated quantum rather than a spin- $0$ one.² The fraction of three-jet to two-jet events measures the coupling directly,

\frac{σ _{3 -jet}}{σ _{2 -jet}} \sim \frac{α _{s}}{π},

the QCD analogue of hard-photon bremsstrahlung in QED. The gluon jet is distinguishable statistically: gluons carry the adjoint color charge $C_{A} = 3$ against the quark's $C_{F} = \frac{4}{3}$ , so gluon jets radiate more and are broader with higher hadron multiplicity.

A three-jet event. One of the quarks radiates a hard gluon before hadronizing, and the gluon fragments into its own jet. The three jets lie in a plane; the rate relative to two-jet events is proportional to alpha_s, and the observation at PETRA in 1979 was the first direct evidence for the gluon.

Fragmentation and hadronization

Between the hard parton and the observed hadrons lies hadronization, the nonperturbative step in which color-charged partons convert into color-singlet hadrons. It cannot be computed in perturbation theory, but its gross features follow from the flux-tube picture. As the quark and antiquark separate, the tube between them stretches and breaks by pair creation, again and again, until all the energy is stored in low-momentum hadrons. Each break produces a new $q \overset{q}{ˉ}$ pair, and the fragments combine into mesons and baryons streaming along the original parton direction.

The distribution of hadron momenta is summarized by a fragmentation function $D_{i}^{h} (z)$ , the probability that parton $i$ yields a hadron $h$ carrying a fraction $z$ of the parton momentum. Fragmentation functions are measured, not calculated, and like the parton distributions they evolve with scale by DGLAP equations. The guiding empirical fact is local parton-hadron duality: the flow of hadrons closely tracks the flow of the underlying partons, so the jet momentum reproduces the parton momentum up to soft corrections.

From parton to hadrons. A hard quark radiates a cascade of gluons, which split into quark-antiquark pairs (the parton shower); at the confinement scale the partons combine into color-singlet hadrons flowing along the original direction. The perturbative shower is calculable; the final hadronization is not, but it preserves the parton's momentum and direction.

Jet algorithms and event shapes

Turning a spray of hadrons into a definite jet requires a rule. A jet algorithm clusters particles into jets by a distance measure; the modern sequential-recombination algorithms merge the closest pair repeatedly until well-separated jets remain. The requirement that a jet definition be infrared and collinear safe — unchanged when a particle emits a soft gluon or splits collinearly — is what makes the jet cross section calculable in perturbation theory.

An alternative to counting jets is to characterize the whole event by a continuous event-shape variable. The most-used is the thrust,

T = \overset{n}{^} max \frac{\sum _{i} ∣ p _{i} \cdot n ^ ∣}{\sum _{i} ∣ p _{i} ∣},

maximized over the axis $\overset{n}{^}$ . A perfect two-jet (pencil-like) event has $T \to 1$ ; a spherical multi-jet event has $T \to \frac{1}{2}$ . The distribution of $T$ is calculable in QCD, and its shape depends on $α_{s}$ : more gluon radiation broadens events and pushes $T$ away from $1$ .

The thrust distribution. Most events are close to two-jet, piling up near thrust equal to one; gluon radiation produces the tail toward lower thrust. The shape of the tail is proportional to alpha_s, so fitting it measures the strong coupling. A larger coupling shifts weight toward broader, lower-thrust events.

Precision alpha_s

Jet rates and event shapes give some of the most precise determinations of the strong coupling. The three-jet rate, the thrust distribution, and related observables are each proportional to $α_{s}$ at leading order, computed to higher orders in perturbation theory, and fit to the data at the $Z$ mass. Combined with the deep-inelastic and $τ$ -decay determinations at other scales, they populate the running-coupling curve and yield the world average

α_{s} (M_{Z}^{2}) = 0.1180 \pm 0.0009.

That a single coupling, extracted from jet shapes at $M_{Z}$ , from DIS at a few GeV, and from $τ$ decays below that, all lie on the one asymptotic-freedom curve is the quantitative triumph of QCD as the theory of the strong force.³

Two-jet production in $e^{+} e^{-}$ annihilation, the $1 + cos^{2} θ$ angular distribution as evidence for spin- $\frac{1}{2}$ quarks, and the jet picture of hadronization are treated in Griffiths, Introduction to Elementary Particles, §9.5, and Thomson, Modern Particle Physics, Ch. 10. ↩
The discovery of three-jet events and the gluon at PETRA (TASSO, MARK-J, PLUTO, JADE) in 1979, and the gluon spin determination from three-jet angular correlations, are described in Thomson, Ch. 10, and Perkins, Introduction to High Energy Physics, Ch. 6. The relative rate $σ_{3 -jet} / σ_{2 -jet} \sim α_{s} / π$ follows from the $q \overset{q}{ˉ} g$ matrix element. ↩
The extraction of $α_{s}$ from jet rates and event shapes, the thrust variable, and infrared-collinear safety are discussed in Thomson, Ch. 10, and Perkins, Ch. 6. The world average $α_{s} (M_{Z}^{2}) = 0.1180 \pm 0.0009$ is from the Particle Data Group review of QCD, pdg.lbl.gov. ↩

Two-jet events

Three-jet events and the gluon

Fragmentation and hadronization

Jet algorithms and event shapes

Precision alpha_s

Footnotes