Inside a locked cabinet, an envelope held a number that was poised to rock the physics community, regardless of its contents. The value, a clock measurement deliberately hidden to keep physicists’ data analysis unbiased, would be used in a calculation that could either bring one of particle physics’ foremost mysteries to an end—or take it even deeper. On a recent video call, 170 scientists gathered to watch the envelope opening.
“When we saw the number on the screen, it was a feeling of great relief, excitement, pride, and joy,” said Sudeshna Ganguly, associate scientist at Fermilab. “We had to unmute in order to shout.”
Today, the global team of scientists based at Fermilab’s Muon g-2 Experiment released one of particle physics’ most highly anticipated measurements, describing the behavior of the electron’s heavier cousin, the muon, in a magnetic field. If further work confirms their results, then researchers will be dealing with an unexplained discrepancy between their experimental work and the backbone theory of particle physics, the Standard Model.
Physicists carry out high-precision measurements of particles’ properties as a means to probe the universe’s underpinnings. Magnetic moment is one such property, describing how particles precess, or wobble like a faltering spinning top, in the presence of a magnetic field. Previous measurements at Brookhaven National Laboratory in New York have hinted that measurements of the muon’s magnetic moment might not agree with the Standard Model’s predictions. Today, the Muon g-2 Collaboration announced that they’d measured, to extreme precision, a value that agrees closely with the Brookhaven value—still in disagreement with the Standard Model. Taken together, experimental measurements now disagree with Standard Model calculations by 4.2 standard deviations.
This discrepancy is not strong enough to meet the five standard deviation (called “five sigma”) statistical milestone that particle physicists use to signify whether they’ve truly discovered something. Five sigma means that the odds of particles following the rules of the Standard Model producing the measured value after the duration of the experiment are around one in 3.5 million. In other words, if the experimenters reached the five sigma confidence level, then they could be pretty sure that the Standard Model is missing a piece to explain the g-2 (pronounced gee-minus-two) value.
The g-2 measurement’s story dates as far back as 1928, when physicist Paul Dirac calculated the electron’s magnetic moment, g, to be exactly 2. However, deeper searches into the property by physicists such as Julian Schwinger brought about smaller corrections, calculated as the difference between g and 2. Experiments to measure g-2 for the muon followed, first at Columbia’s Nevis Laboratories and CERN, then at Brookhaven National Lab, whose collaboration finished taking measurements in 2001 and released their final results—and evidence of a discrepancy—in 2004.
A decade later, national lab scientists coordinated to continue the effort and move the fragile 50-foot ring of Brookhaven’s g-2 experiment from Long Island, New York to Fermilab in Illinois—first on a barge down the Atlantic and up the Mississippi, and then on the back of a truck. Scientists have been running the experiment at Fermilab since 2017.
The experiment begins with protons from Fermilab’s accelerators, smashed into a fixed target, producing more protons plus the antiparticle partner of the muon called the antimuon (basically, a muon behaving like its own mirror image with the opposite charge) and another particle called the pion that decays into antimuons. This muon beam races around the experiment’s electromagnetic ring at almost the speed of light. All the while, antimuons begin to decay out of the beam as anti-electrons (aka positrons) which strike the detectors. The measurement of these positrons allows the scientists to determine how the antimuons were behaving in the magnetic field, and thus what the g-2 value of the antimuons was. Antimuons are simply easier to produce than muons, but the g-2 value would be the same for both.
The cause of the discrepancy isn’t clear. “Maybe it’s a particle that’s hard to produce,” Joe Lykken, theoretical physicist and deputy director of research at Fermilab, told Gizmodo on a video call. If so, “it should show up somewhere else, like cosmic observations. Or maybe it’s already shown up in our data and we need to tease it out.” There are plenty of unsolved cosmic mysteries and discrepancies—dark matter, the Hubble tension, or recent results from CERN’s LHCb experiment, to name a few—that theorists might try to link to the muon g-2 discrepancy.
Researchers across the particle physics community are excited about the news. “The fact that two independent experiments both see well over three standard deviations from the established prediction means that, as far as I am concerned, the experimental measurement is sound,” said Freya Blekman, professor in elementary particle physics at the Vrije Universiteit Brussel, in a Twitter direct message. “But the ball is now really in the court of the theoretical physicists who are calculating the value that is being compared to.”
Theorists have spent the past two decades trying to understand this discrepancy and what might be causing it. Some wonder whether the discrepancy is real at all; one controversial paper calculated the muon g-2 value with a different method that seemed to explain away the discrepancy. But even that calculation, should it be verified and accepted by the broader theory community, would introduce another set of discrepancies, said Chris Polly, co-spokesperson for the Muon g-2 Experiment. Polly explained that the Fermilab experiment would be valuable whether or not the discrepancy was real; any physics theory trying to solve the universe’s mysteries must also agree with the team’s precise measurements.
Of course, the ongoing covid-19 pandemic challenged the scientists as well. Covid-19 precautions limited the number of researchers who could sit inside the experiment’s control room. “When you’re coordinating a run during normal times, you’re always in the control room and experimental hall telling people what to do,” said Ganguly. But the past year cut down the number of people who could enter the experimental hall and fix things when they broke. “To do that over a Zoom was a challenge. But in the end, we had to make it work.”
The team operated what they could from afar, such as the the trolley that runs inside of the experiment in order to map out its magnetic field. And social distancing had little impact on the data analysis; the team represents scientists from countries around the world, already used to carrying out much of their work remotely, explained Polly.
Today, the Muon g-2 Collaboration researchers are celebrating—but they have their work cut out for them. Through further analysis, refinements, and experimental runs, the team plans to continue reducing the experimental error in order to calculate an even more precise muon g-2 value.
“So far, we’ve only analyzed 6% of the data, and when we combine the results of all of the runs, we’ll get an even better measurement,” said Ganguly. “It’s super exciting to be a part of this.”