Revisiting the b revolution

19 May 2017

A look back at the Fermilab experiment that unearthed the b quark.


Scientists summoned from all parts of Fermilab had gathered in the auditorium on the afternoon of 30 June 1977. Murmurs of speculation ran through the crowd about the reason for the hastily scheduled colloquium. In fact, word of a discovery had begun to leak out [long before the age of blogs], but no one had yet made an official announcement. Then, Steve Herb, a postdoc from Columbia University, stepped to the microphone and ended the speculation: Herb announced that scientists at Fermilab Experiment 288 had discovered the upsilon particle. A new generation of quarks was born. The upsilon had made its first and famous appearance at the Proton Center at Fermilab. The particle, a b quark and an anti-b quark bound together, meant that the collaboration had made Fermilab’s first major discovery. Leon Lederman, spokesman for the original experiment, described the upsilon discovery as “one of the most expected surprises in particle physics”.

The story had begun in 1970, when the Standard Model of particle interactions was a much thinner version of its later form. Four leptons had been discovered, while only three quarks had been observed – up, down and strange. The charm quark had been predicted, but was yet to be discovered, and the top and bottom quarks were not much more than a jotting on a theorist’s bedside table.

In June of that year, Lederman and a group of scientists proposed an experiment at Fermilab (then the National Accelerator Laboratory) to measure lepton production in a series of experimental phases that began with the study of single leptons emitted in proton collisions. This experiment, E70, laid the groundwork for what would become the collaboration that discovered the upsilon.


The original E70 detector design included a two-arm spectrometer [for the detection of lepton pairs, or di-leptons], but the group first experimented with a single arm [searching for single leptons that could come, for example, from the decay of the W, which was still to be discovered]. E70 began running in March of 1973, pursuing direct lepton production. Fermilab director Robert Wilson asked for an update from the experiment, so the collaborators extended their ambitions, planned for the addition of the second spectrometer arm and submitted a new proposal, number 288, in February 1974 – a single-page, six-point paper in which the group promised to get results, “publish these and become famous”. This two-arm experiment would be called E288.

The charm dimension

Meanwhile, experiments at Brookhaven National Laboratory and at the Stanford Linear Accelerator Center were searching for the charm quark. These two experiments led to what is known as the “November Revolution” in physics. In November of 1974, both groups announced they had found a new particle, which was later proven to be a bound state of the charm quark: the J/psi particle.

Some semblance of symmetry had returned to the Standard Model with the discovery of charm. But in 1975, an experiment at SLAC revealed the existence of a new lepton, called tau. This brought a third generation of matter to the Standard Model, and was a solid indication that there were more third-generation particles to be found.

The Fermilab experiment E288 continued the work of E70 so much of the hardware was already in place waiting for upgrades. By the summer of 1975, collaborators completed construction on the detector. Lederman invited a group from the State University of New York at Stony Brook to join the project, which began taking data in the autumn of 1975.

One of the many legends in the saga of the b quark describes a false peak in E288’s data. In the process of taking data, several events at an energy level between 5.8 and 6.2 GeV were observed, suggesting the existence of a new particle. The name upsilon was suggested for this new particle. Unfortunately, the signals at that particular energy turned out to be mere fluctuations, and the eagerly anticipated upsilon became known as “oopsLeon”.


What happened next is perhaps best described in a 1977 issue of The Village Crier (FermiNews’s predecessor): “After what Dr R R Wilson jocularly refers to as ‘horsing around,’ the group tightened its goals in the spring of 1977.” The tightening of goals came with a more specific proposal for E288 and a revamping of the detector. The collaborators, honed by their experiences with the Fermilab beam, used the detectors and electronics from E70 and the early days of E288, and added two steel magnets and two wire-chamber detectors borrowed from the Brookhaven J/psi experiment.

The simultaneous detection of two muons from upsilon decay characterised the particle’s expected signature. To improve the experiment’s muon-detection capability, collaborators called for the addition to their detector of 12 cubic feet – about two metric tonnes – of beryllium, a light element that would act as an absorber for particles such as protons and pions, but would have little effect on the sought-for muons. When the collaborators had problems finding enough of the scarce and expensive material, an almost forgotten supply of beryllium in a warehouse at Oak Ridge National Laboratory came to the rescue. By April 1977, construction was complete.

Six weeks to fame

The experiment began taking data on 15 May 1977, and saw quick results. After one week of taking data, a “bump” appeared at 9.5 GeV. John Yoh, sure but not overconfident, put a bottle of champagne labelled “9.5” in the Proton Center’s refrigerator.

But champagne corks did not fly right away. On 21 May, fire broke out in a device that measures current in a magnet, and the fire spread to the wiring. The electrical fire created chlorine gas, which when doused with water to put out the fire, created acid. The acid began to eat away at the electronics, threatening the future of E288. At 2.00 a.m. Lederman was on the phone searching for a salvage expert. He found his expert: a Dutchman who lived in Spain and worked for a German company. The expert agreed to come, but needed 10 days to get a US visa. Lederman called the US embassy, asking for an exception. Not possible, said the embassy official. Just as it began to look hopeless, Lederman mentioned that he was a Columbia University professor. The official turned out to be a Columbia graduate, class of 1956. The salvage expert was at Fermilab two days later. Collaborators used the expert’s “secret formulas” to treat some 900 electronic circuit boards, and E288 was back online by 27 May.

By 15 June, the collaborators had collected enough data to prove the existence of the bump at 9.5 GeV – evidence for a new particle, the upsilon. On 30 June, Steve Herb gave the official announcement of the discovery at the seminar at Fermilab, and on 1 July the collaborators submitted a paper to Physics Review Letters. It was published without review on 1 August.

Since the discovery of the upsilon, physicists have found several levels of upsilon states. Not only was the upsilon the first major discovery for Fermilab, it was also the first indication of a third generation of quarks. A bottom quark meant there ought to be a top quark. Sure enough, Fermilab found the top quark in 1995.

Bumps on the particle-physics road

The story of “bumps” in particle physics dates back to an experiment at the Chicago Cyclotron in 1952, when Herbert Anderson, Enrico Fermi and colleagues found that the πp cross-section rose rapidly at pion energies of 80–150 MeV, with the effect about three times larger in π+p than in πp. This observation had all the hallmarks of the resonance phenomena that was well known in nuclear physics, and could be explained by a state with spin 3/2, isospin 3/2. With higher energies available at the Carnegie Synchro-cyclotron, in 1954 Julius Ashkin and colleagues were able to report that the πp cross-section fell above about 180 MeV, revealing a characteristic resonance peak. Through the uncertainty principle, the peak’s width of some 100 MeV is consistent with a lifetime of around 10–23 s. Further studies confirmed the resonance, later called Δ, with a mass of 1232 MeV in four charge states: Δ++, Δ+, Δ0 and Δ.

The Δ remained an isolated case until 1960, when a team led by Luis Alvarez began studies of Kp interactions using the 15 inch hydrogen bubble chamber at the Berkeley Bevatron. Graduate students Stan Wojcicki and Bill Marciano studied plots of the invariant mass of pairs of particles produced, and found bumps corresponding to three resonances now known as the Σ(1385), the Λ(1405) and the K*(892). These discoveries opened a golden age for bubble chambers, and set in motion the industry of “bump hunting” and the field of hadron spectroscopy. Four years later, the Δ, together with Σ and Ξ resonances, figured in the famous decuplet of spin-3/2 particles in Murray Gell-Mann’s quark model. These resonances and others could now be understood as excited states of the constituents – quarks – of more familiar longer-lived particles.

By the early 1970s, the number of broad resonances had grown into the hundreds. Then came the shock of the “November Revolution” of 1974. Teams at Brookhaven and SLAC discovered a new, much narrower resonance in experiments studying, respectively, pBe  e+eX and e+e annihilation. This was the famous J/psi, which after the dust had settled was recognised as the bound state of a predicted fourth quark, charm, and its antiquark. The discovery of the upsilon, again as a narrow resonance formed from a bottom quark and antiquark, followed three years later (see main article). By the end of the decade, bumps in appropriate channels were revealing a new spectroscopy of charm and bottom particles at energies around 4 GeV and 10 GeV, respectively.

This left the predicted top quark, and in the absence of any clear idea of its mass, over the following years searches at increasingly high energies looked for a bump that could indicate its quark–antiquark bound state. The effort moved from e+e colliders to the higher energies of p–p machines, and it was experimental groups at Fermilab’s Tevatron that eventually claimed the first observation of top quarks in 1995, not in a resonance, but through their individual decays.

However, important bumps did appear in p-p collisions, this time at CERN’s SPS, in the experiments that discovered the W and Z bosons in 1983. The bumps allowed the first precise measurements of the masses of the bosons. The Z later became famous as an e+e resonance, in particular at CERN’s LEP collider. The most precisely measured resonance yet, the Z has a mass of 91.1876±0.0021 GeV and a width of 2.4952±0.0023 GeV.

However, a more recent bump is probably still more famous – the Higgs boson as observed in 2012. In data from the ATLAS and CMS experiments, small bumps around 125 GeV in the mass spectrum in the four-lepton and two-photon channels, respectively, revealed the long-sought scalar boson (CERN Courier September 2012 p43 and p49).

Today, bump-hunting continues at machines spanning a huge range in energy, from the BEPC-II e+e collider, with a beam energy of 1–2.3 GeV in China, to the CERN’s LHC, operating at 6.5 TeV per beam. Only recently, LHC experiments spotted a modest excess of events at an energy of 750 GeV;  although the researchers cautioned that it was not statistically significant, it still prompted hundreds of publications on the arXiv preprint server. Alas indeed, on this occasion as on others over the decades, the bump faded away once larger data sets were recorded.

Nevertheless, with the continuing searches for new high-mass particles, now as messengers for physics beyond the Standard Model, and searches at lower energies providing many intriguing bumps, who knows where the next exciting “bump” might appear?

• Christine Sutton, formerly CERN.



bright-rec iop pub iop-science physcis connect