by Harry Cliff
* * *
—
Tuesday, March 30, 2010, was a big day for everyone at CERN. It marked the start of the LHC’s physics program and with it the beginning of the search for answers to some of the deepest questions it is possible to ask about our universe. I still remember those first few weeks vividly: the excitement of seeing the first event displays showing particles flying through the LHCb detector, the sense of responsibility and pressure that came with helping to keep one small part of the experiment running smoothly, the thrill of seeing the first data showing that real particles were being made in our detector.
As the days and weeks passed, each of the four LHC experiments recorded more and more collisions at an ever-increasing rate, as the engineers driving the LHC from the CERN Control Centre gradually learned how to operate the impossibly complex machine that they had built. Having set a new world record for the highest energy collisions ever created in a lab, their task was now to figure out how to increase the rate of collisions, so that more and more data could be recorded each day by the four experiments.
LHCb, the experiment I work on, is a specialized detector designed primarily to study particles known as “bottom quarks.” These are heavy companions of the ordinary down quarks that make up protons and neutrons, and by measuring their properties in detail we can learn a great deal about both the standard model and potentially new quantum fields that we haven’t yet discovered. However, one thing LHCb is not designed to do is to search for the Higgs boson.
The Higgs was the target of the two largest experiments at the LHC: ATLAS and CMS.
These two behemoths are “general purpose” detectors, which means that they’re designed to search for as wide a variety of different particles as possible. They sit on opposite sides of the ring; ATLAS is just next to the main CERN site, and therefore close to all the handy amenities like the restaurants, while CMS is several kilometers away in the middle of the French countryside, which is very picturesque but must be a bit of a pain if you fancy popping back to CERN for lunch.
Each of these two experiments is staffed by a team of more than three thousand physicists, engineers, and computer scientists from across the world, each making a small, or sometimes not so small, contribution to the ultimate goal of discovering new features of the subatomic landscape. There are the senior physicists, many of whom were involved in the early planning of the experiments and now have managerial roles in the collaboration, setting strategy and leading the not inconsiderable task of building consensus among three thousand sometimes oversized egos. Then there are the hardware and software experts who designed and built the experimental equipment and software stack, who are absolutely critical to running and maintaining the experiment, while planning and delivering future upgrades. And then there is a veritable army of physicists, many of them young PhD students and postdoctoral researchers, whose job is to analyze the relentless flood of data and search for signs of new particles.
The experiments are only possible thanks to this enormous international group effort, and yet remarkably, the search for the Higgs boson ultimately came down to just a handful of young people who had the awesome responsibility of bringing a fifty-year search to its dramatic conclusion, and with it, providing the answer to the question, why do particles have mass? One of these fortunate few was Matt Kenzie, a former colleague of mine at Cambridge who back in those early days was a PhD student at Imperial College London just starting out on CMS.
Matt first arrived at CERN in the spring of 2011, just as the LHC was firing up for its second year of collisions. He had started his physics career with aspirations of being a theorist and perhaps doing something small like discovering a quantum theory of gravity. However, after a year studying for his master’s, he realized the life of a theorist wasn’t for him and ended up submitting a last-minute application for a PhD in experimental particle physics. So last-minute in fact that the deadline for applications had already passed, but by a stroke of luck Imperial had a spare place going after one of the candidates they’d made an offer to had dropped out. After a hastily arranged interview he soon found himself embarking on a PhD, and before he knew it, he was heading to CERN to work on the search for the Higgs boson.
Of the three thousand people on CMS, hundreds were involved in the Higgs search to greater or lesser degrees, but in reality, almost all of the day-to-day work was done by a relatively small group of a few dozen researchers. By Matt’s own admission it was dumb luck that he ended up one of the select few who had the incredible privilege of analyzing the data that would settle one of the longest-running questions in physics. “It was really a complete fluke that I landed in the middle of this huge scoop,” he told me.
Physicists at ATLAS and CMS knew that if Higgs bosons got created in the collisions at the LHC then they would decay almost instantaneously, within just 10-22 seconds, far too short a time for them to reach any of the sensitive parts of the detectors. That meant that the only way to find evidence for the Higgs would be to try to catch the particles it decayed into as they zipped out through the detector, a bit like trying to figure out the make and model of a car by filling it with dynamite and photographing the various bits of shrapnel as they fly past.
However, unlike a car, which always falls apart into the same basic bits, the Higgs can decay in a variety of different ways. Some of these are easier to spot than others. For example, its decay into a bottom quark and an anti-bottom quark is by far the most common, with more than half of all Higgs bosons decaying this way, but nonetheless it is incredibly hard to spot thanks to the vast numbers of bottom quarks that get produced by the LHC collisions. Searching for the Higgs decaying to a bottom–anti-bottom pair is less like looking for a needle in a haystack than looking for a needle in a pile of other very-similar-looking needles.
Luckily, there are far better ways to spy a Higgs. The most promising decays of all are when a Higgs turns into either two high-energy photons or two Z bosons. That doesn’t mean seeing these decays is easy, but it is at least more like looking for a needle in a haystack (or perhaps a needle in a field full of haystacks).
Matt’s task on arriving at CERN was to help write the computer code that would search for signs of a Higgs boson decaying into two photons. He remembers getting a bit of a shock the first time he presented his work to one of CMS’s working groups, receiving a particularly hostile roasting from a postdoctoral researcher from MIT. It turned out that there was quite a bit of rivalry between the MIT and Imperial groups, who were both vying to lead the search, and Matt had unwittingly found himself in the middle of this battle. “It was a bit of a baptism of fire,” as he put it. “I came away feeling this is quite a tough environment.”
At times the pressure was intense. Not only was Matt’s team, which included physicists from Imperial, San Diego, CERN, and Italy, competing against the MIT group in the search for the Higgs decay to two photons, there were other groups at CMS working on the Higgs decay into two Z bosons, not to mention three thousand other researchers on their giant rival ATLAS, who were racing toward the same goal. Matt worked in a tight-knit team with a couple of other PhD students and two postdoctoral researchers. They’d meet at least twice a week, often talking through their latest results over breakfast or lunch in CERN’s Restaurant 1, the bustling cafeteria that is the hub of social activity at the lab. There were a lot of long evenings working in the office after hours, and one or two all-nighters, particularly around the timings of public updates on the Higgs hunt.
So how do you actually hunt for the Higgs? Well, it all begins with the collisions produced by the LHC. Let’s say that the Higgs field exists, and that when protons smash into each other at the LHC they have enough energy to set the Higgs field wobbling, creating a new particle, the Higgs boson. The first challenge is that the chances of producing a Higgs in a single collision are extremely low. Thanks to the probabilistic nature of quantum mechani
cs, when two protons collide you cannot say ahead of time what particles will be produced. A collision is like rolling a die with an enormous number of faces, with each face corresponding to a different potential outcome.
Most of the time a collision will just produce particles we already know about: quarks, gluons, photons, maybe a W or Z boson. The odds of making a Higgs are fantastically slim—just two in a billion—so to have any chance of seeing a decent number the LHC had to create a stupendously large numbers of collisions. In fact, it is able to produce around a billion collisions every second inside ATLAS and CMS, and the aim is to do this twenty-four hours a day, seven days a week, for about nine months of the year, minus a few technical stops and the time it takes to set up for a new run. By the end of 2012 this incredible collision rate meant the LHC had produced 6 quadrillion (thousand trillion) collisions.
This vast number of collisions implies a corresponding vast amount of data, so much in fact that the LHC could fill every hard drive on Earth in a matter of days. The only way to cope with this digital tsunami is to throw away most of the collisions before they are even recorded. This is done by a set of very fast computer algorithms called “triggers,” which look at each collision in real time—once every 25 nanoseconds—and decide whether it looks like something interesting may have happened or whether it’s just a load of boring old quarks and gluons that we’ve seen before. On the rare occasions it does look interesting, the data is recorded, and analysts like Matt can start their search.
Once the data is stored, the way you search for Higgs bosons that have decayed into two photons is actually rather straightforward, at least in principle. The teams at CMS and ATLAS used tailor-made algorithms to sift through their huge datasets in search of collisions containing a pair of high-energy photons. However, the hard part is that there is an enormous number of other ways that photons can get made when two protons smash into each other, and most of them don’t involve a Higgs boson at all. Matt and his colleagues needed to find a way of sifting out the real Higgs bosons from this background of random photons that didn’t come from a Higgs, a bit like panning for gold in a fast-flowing stream.
Even after the data is sifted, there will still be a lot more background than real Higgses, but here one final trick comes into play. If you add up the energies of the two photons, that directly tells you the mass of the particle they came from. For random background photons, their energies can add up to more or less any number, which means that if you plot them on a graph they will be spread out over a wide range of masses. However, photons that came from a Higgs will always add up to give the same mass—the mass of the Higgs boson—which means that they cluster around the same value, creating a little bump on top of the flat background. If this bump is prominent enough, then that is the smoking gun of a new particle being produced in your experiment.
Just before Christmas in 2011, rumors began to swirl around the corridors of CERN that ATLAS and CMS had spotted something. After a year of intense data collecting and analysis, a special seminar was arranged on December 13 to update the scientific community on the search for the most elusive particle in history. Hundreds of people queued to get a seat in CERN’s main auditorium to hear the presentations, forcing Matt to watch the proceedings via live webcast from his office. The results were tantalizing but inconclusive. Both ATLAS and CMS had seen a small bump at a mass of 125 GeV (for reference a proton has a mass of around 1 GeV and a Z boson has a mass of 90 GeV). But it was too small to be sure that it wasn’t just a statistical fluctuation. However, small as the bump might have been, the fact that two independent experiments had both seen hints of a new particle at the same mass set many pulses racing in the particle physics community.
When data collection began again in the spring of 2012, everything suddenly became much more serious. Sensing that both they and ATLAS were hot on the trail of the Higgs, Matt and the team had to work even harder to make sure they kept up with their competitors. To ensure complete secrecy and avoid the danger that they might consciously or unconsciously massage their methods and bias the result, the analysis was performed blind. This meant that it was impossible to look at the final data until everything had been approved by the collaboration, leading to a dramatic final moment of unblinding. Only then would the result be revealed, and they’d know whether or not they’d found the Higgs.
By late June 2012, ATLAS and CMS had recorded as much data as they had in the whole of 2011 and it was time to take another look. On the morning of the unblinding, Matt and the small team gathered around a laptop amid the hubbub of Restaurant 1, bleary-eyed after several weeks of grueling hours, and finally peered at the crucial graph. There on the screen was a bump in exactly the same place they had seen it in the 2011 data.
Being so new to particle physics, Matt didn’t realize the full significance of what they’d seen until their results were presented to the whole CMS collaboration later that afternoon. Hundreds of physicists squeezed into the seminar room to listen to Mingming Yang, an energetic Chinese PhD student on the MIT team, present the results of the search for the Higgs boson decaying into two photons. She teased the audience by slowly revealing the results in stages, first 2011, then 2012. As she came to the last slide showing the combined result she gave a final rhetorical flourish: “I hope you remember this moment for the rest of your lives.” A click later and there on the projector screen was a graph with a clear spike at 125 GeV. When the team searching for the Higgs boson’s decay into two Z bosons revealed a spike in exactly the same place a few minutes later, Matt said, “All hell broke loose.”
The next day he got an email from Joe Incandela, the spokesperson of CMS, inviting him and around fifty other physicists to help him prepare for a public announcement that had been scheduled for the Fourth of July 2012 (Higgsdependence Day as it jokingly became known at CERN). Joe had the awesome responsibility of revealing CMS’s results to the world, alongside Fabiola Gianotti, who would be presenting the results on behalf of ATLAS. The gang of fifty spent several days locked away in that seminar room, discussing how best to present the results and polishing the final presentation.
While this was going on, Matt was flying back and forth to the United Kingdom each weekend to see his father, who was seriously ill. He did his best to explain to his parents what had been found at CERN, and while they understood it was a big deal, he got the sense they didn’t fully grasp what this Higgs business was all about. “Their response was sort of ‘That’s lovely, dear.’ They had more important things to worry about at the time.”
The upshot of this was that Matt hadn’t expected to be at CERN for the announcement and so had passed up the offer of a reserved seat in the main auditorium. However, as things turned out, he was back at CERN on the crucial day and managed to sneak in with a few of the other members of the CMS Higgs team.
The atmosphere in the room was electric, with one physicist describing it as like being at a football match. Now I’m not sure he’d ever been at a football match, but it was certainly very lively by particle physics standards. People had camped outside the auditorium overnight for a chance to get a seat, with hundreds having to be turned away. Meanwhile, in London, I was at a huge live webcast event near Parliament with a bunch of sciencey types, journalists, and members of government, all waiting for the presentations to start.
A couple of minutes before kickoff, the director general of CERN, Rolf Heuer, entered the auditorium flanked by Peter Higgs and François Englert, whose original research had set off this incredible chain of events almost half a century earlier. It was then that Matt realized he was involved in something big.
First came the presentation from CMS, as Joe Incandela revealed the same peaks that they’d all seen a couple of weeks earlier. Now everyone was waiting with bated breath to see if ATLAS had seen the same thing. When Fabiola Gianotti revealed a graph showing a bump at exactly the same mass as CMS, the room erupted.
The assembled physicists hollered and cheered the tremendous joint achievement. Peter Higgs, now well into his eighties, could be seen wiping away a tear. He later said that he had never expected the particle he had first predicted as a young man in 1964 to be discovered in his lifetime. It was only when CERN’s director general made his final declaration at the end of the event that Matt realized just what he had been involved in: “As a layman, I would now say, I think we have it.”
They had found the Higgs boson.
Skip Notes
*1 To be fair, we did learn how to measure the strength of gravity by rolling a ball down a slope, which actually turns out to be excellent training for a career in physics (in the seventeenth century).
*2 Compact, I guess, is a relative term. Its rival experiment, ATLAS, on the opposite side of the ring, is almost twice the height, width, and length of CMS.
*3 We now know that neutrons and protons are made of quarks, up-down-down in the case of a neutron, up-up-down in the case of a proton, so at the fundamental level what happens is a down quark turns into an up quark, an electron, and an antineutrino.
*4 Of course, this wouldn’t be true if you were close to a massive body like the Earth. The gravity of the Earth defines a preferred direction in space, which breaks the rotational symmetry.
