On this page
In 1979, at the PETRA collider of DESY in Hamburg, Sau Lan Wu of the University of Wisconsin was hunting for one shape in her data: three sprays of particles flying out in a flat star. She had worked out in advance how much energy that needed, and a method to find it. When the energy arrived, so did the shape: the first direct evidence of the gluon, as Wu describes in her 2018 talk "Discovery of the Gluon".
Wu was born in Hong Kong, went to Vassar College on a full scholarship in 1960, did her PhD at Harvard, and joined Wisconsin as an assistant professor in 1977, according to the American Institute of Physics.
What is a gluon, and why was it hard to see?
A gluon is the particle that carries the strong force between quarks, as a photon carries the electromagnetic force (Wu's slides). Before 1979, the only force-carrying particle anyone had observed was the photon.
Quarks and gluons never travel alone. Each turns into a narrow spray of ordinary particles, called a jet, so you can only look for a gluon's jet. Experiments at SLAC had already found evidence for quarks inside protons (see how deep inelastic scattering found evidence for quarks), but the gluon was still only inferred.
How did three-jet events prove the gluon existed?
In 1976, John Ellis, Mary Gaillard and Graham Ross proposed searching for gluons in electron-positron collisions (Ellis's history). An electron and a positron annihilate into a quark and an antiquark, which fly apart as two jets. Sometimes the quark should shake off a gluon, giving a third jet. Drawn out, it is one of Feynman's diagrams with an extra curly line.
Why must the third jet be new? The electron and positron are two fermions, which cannot turn into three fermions, so the three jets cannot all be quarks or antiquarks (Wu, slide 33). The third must be a boson that, like a quark, makes a jet: the gluon.
How did Sau Lan Wu find the three-jet events?
Wu used that plane. With Georg Zobernig, her first postdoc, she published "A method of three-jet analysis in e+e− annihilation" in 1979. It finds the plane that best fits all the particles, projects every track onto it, and sorts tracks by angle into three groups. Working in two dimensions cut computer time sharply on 1970s machines (Wu, slides 21-25).
The program was ready in autumn 1978. At 13 GeV and 17 GeV the team found no three-jet events. Then, in June 1979, Wu and Zobernig found one clear three-jet event out of 40 hadronic events at 27.4 GeV (Wu, slide 27). Björn Wiik showed it at a conference in Bergen, Paul Söding showed four TASSO events in Geneva within two weeks, and TASSO published "Evidence for planar events in e+e− annihilation at high energies". By late August 1979, all four PETRA experiments had data confirming the result.
In 1995, Wu, Söding, Wiik and Günter Wolf received the European Physical Society High Energy and Particle Physics Prize "for the first evidence for three-jet events in e+e− collisions at PETRA" (Wu, slide 38).
Worked example: redo Wu's energy estimate
Before PETRA switched on, Wu asked whether it would have enough energy to show three jets clearly. Her rough estimate (slides 18-20) needs only vectors and a cosine:
SLAC's SPEAR collider, at 7.4 GeV, showed two clean back-to-back jets. So a jet pair needs a mass of about 7.4 GeV to look clean.
At PETRA, take the most even case: three jets of equal energy E, 120 degrees apart.
For two massless jets of energy E at 120 degrees, the pair's mass squared is 2E²(1 − cos 120°) = 3E². So the pair mass is √3 E.
Set √3 E = 7.4 GeV. Then E is about 4.3 GeV, and three jets need about 13 GeV in total.
Each jet has 120 degrees of room instead of 180, so it must be narrower. Multiply by 180/120 = 1.5 to get about 20 GeV.
Phase-space considerations pushed Wu's estimate to about 22 GeV, roughly three times SPEAR's energy.
PETRA passed that line in spring 1979, and the events appeared.
If step 3 trips you up, the Fermi Problem Library has worked problems with a step-by-step solver that explains the reasoning rather than printing an answer.
What can you learn from how Wu worked?
Estimate before you measure. Wu's estimate put clear three-jet events near 22 GeV, so empty runs at 13 and 17 GeV were expected (Wu, slides 20 and 27).
Let a conservation law simplify the problem. Momentum conservation turned a 3D search into a 2D one.
For more on quarks, read Murray Gell-Mann and the quarks inside protons and neutrons. And if vectors adding to zero still feel abstract, the Fermi Simulation Lab has interactive simulations for ideas that are easier to see than to read.
