‘If nature is kind to us and the Higgs particle has a mass within the current range of the LHC,’ said CERN Director-General Rolf Heuer of the decision, ‘we could have enough data in 2011 to see hints, but not enough for a discovery. Running through 2012 will give us the data needed to turn such hints into discovery.’4
The stage was set.
____________
Einstein’s secretary Helen Dukas once asked if he could provide a simple explanation of relativity that she could use in response to the many queries she received from reporters. He thought about it for a while and suggested: ‘An hour sitting with a pretty girl on a park bench passes like a minute, but a minute sitting on a hot stove seems like an hour.’5
Among the thousands of scientists involved in the collaborations at Fermilab and CERN, the tension and excitement was now palpable. There had been no particle discovery in over a decade. Nearly eleven years had passed since the Higgs had been ‘glimpsed’ by the LEP collider. And now the promise of new physics was desperately, agonisingly close. What was it? Six months? A year? Two years? This was definitely ‘hot stove’ territory.
It was perhaps inevitable that the dam would burst.
Columbia University mathematical physicist Peter Woit had maintained a blog on high-energy physics since the 2006 publication of his successful book Not Even Wrong, a critique of contemporary string theory. On 21 April 2011 he received an anonymous posting which contained the abstract of an internal ATLAS discussion paper. The paper claimed to have found four-sigma evidence for a Higgs boson with a mass of 115 GeV.
This was not a hoax. The paper was authored by a small team of ATLAS physicists at the University of Wisconsin-Madison under the leadership of Sau Lan Wu, who had been part of the Aleph collaboration that had ‘glimpsed’ the Higgs in 2000 towards the end of the LEP’s lifetime. It was therefore no coincidence that Wu had gone back to look at the energy range where she believed she had seen those earlier hints.
There were two problems, however. The first was physical. The particle had been observed in the so-called di-photon mass distribution from a combined total of about 64 inverse picobarns of data gathered during 2010 and early 2011.
At an energy of 7 TeV, the proton-proton collisions occurring in the LHC actually involve quark-quark collisions and the fusion of gluons which, in theory, may produce Higgs bosons. The decay channels open to the Higgs depends on its mass. For a large-mass Higgs decay channels involving the production of two W particles or two Z particles would be available. But for a low-mass 115 GeV Higgs there is insufficient energy to reach these channels. Instead, the Higgs decays through alternative routes. One of these involves the production of two high-energy photons, a process written as H →γγ.
The problem was that the observed resonance was about 30 times larger than the Standard Model prediction for this particular decay channel.
Decay of the Higgs into two photons is dominated in the Standard Model by so-called W-boson ‘loops’ involving the production and subsequent annihilation of W bosons. The upshot is that this decay route is predicted to happen very infrequently, accounting for only about 0.2 per cent of all possible decay pathways. If this really was the Higgs, then its decay into two photons was for some reason being greatly enhanced. Other new particles, such as fourth or even fifth generation quarks and leptons, might have to be invoked to explain this.
The second problem concerned the status of the finding. The leaked document was an internal, so-called ATLAS-Communications or ‘COM’ note – designed for the rapid distribution of un-vetted and un-approved results for discussion within the collaboration. There was no sense in which this could be construed as an ‘official’ view from the ATLAS group. Subsequent review and re-analysis could eliminate this result entirely, long before any formal paper could be written.
The news of the leaked COM note hit the ‘blogosphere’ just before the long Easter weekend and, for a few days, the discussion was contained among high-energy physics bloggers and their followers. In 2009 Dorigo had predicted that news of the discovery of the Higgs would appear first in a blog post. He felt that his prediction had been validated, but he was nevertheless highly doubtful that this was the Higgs and offered $1000 against $500 that further data would reveal no new 115 GeV particle in the di-photon decay channel.
The story was picked up by the British mainstream media on Easter Sunday, 24 April. Jon Butterworth, an ATLAS physicist based at University College London, gave a balanced report to the UK’s Channel 4 News. He said: ‘What’s happened here is a bunch of people have spent four nights without sleep. They’ve made some plots and got rather over-excited, [and] sent them in an internal note around the collaboration. Which is fine. Everyone’s excited out there but unfortunately it’s leaked out. It’s a very exciting place at the moment.’6 The story was reported widely in the newspapers the following day.
In his blog for the Guardian newspaper, Butterworth expanded on this theme: ‘Retaining a detached scientific approach is sometimes difficult. And if we can’t always keep clear heads ourselves, it’s not surprising people outside get excited too. This is why we have internal scrutiny, separate teams working on the same analysis, external peer review, repeat experiments, and so on.’7
Counter-rumours appeared shortly afterwards. A French high-energy physics blog claimed on 28 April that, on examining more data, ATLAS physicists were finding that the evidence for the Higgs was quickly disappearing. On 4 May New Scientist staff reporter David Shiga posted an online news item claiming to have seen a document leaked from the CMS collaboration which indicated that a search through their data had ‘come up empty’.8 Through such leaks the interested observer caught glimpses of the chaotic to-ing and fro-ing now going on within the ATLAS and CMS collaborations.
On 8 May, the ATLAS collaboration released an official update. Further analysis of a total of 132 inverse picobarns of data from 2010 and 2011 had indeed come up empty; the diphoton mass distribution showing no excess of events. In a subsequent blog posting, Butterworth explained that this null result has hardly surprising: even the Standard Model predictions suggested that there should be nothing to see yet, but that something could be expected ‘soonish’. ‘So stay interested in the di-photon mass spectrum,’ he wrote, ‘but wait for solid results before opening the champagne.’9
It seemed that we wouldn’t have to wait too long. At midnight on 22 April, the LHC set a new world record for instantaneous luminosity, of 4.67 × 1032 cm−2 s−1, or 467 inverse microbarns (millionths of a barn) per second. The engineer-in-charge that evening, Laurette Ponce, had visited CERN as a child and had joined the laboratory in 1999 to study for her PhD. ‘I never imagined then that one day it would be me pressing the button to fill the LHC,’ she said.10
As it was midnight, there were few in the CERN Control Room to witness the moment. Ponce shouted and danced, waving her arms in the air like a teenager.
This dramatic increase in luminosity had been achieved by injecting more and more proton bunches from the SPS into each beam circulating around the LHC. By 3 May, the peak luminosity had been increased further, to 880 inverse microbarns per second with 768 bunches per beam. Towards the end of May, a peak luminosity of 1260 inverse microbarns per second was recorded.
To put this in perspective, the cross-section for inelastic proton-proton collisions at 7 TeV is about 60 millibarns, or 0.06 barns. An instantaneous luminosity of 1260 inverse microbarns per second therefore implies 1260 × 106 × 0.06 = over 75 million collisions per second. If we take the cross-section for Higgs boson production at 7 TeV to be 9 picobarns,* then this instantaneous luminosity implies 1260 × 106 × 9 × 10−12 = 0.011 Higgs bosons per second, or one Higgs boson produced on average every 90 seconds.
The furore over the leak had sparked interest in the process by which a ‘blockbuster’ result would be officially announced. James Gillies, CERN’s Head of Communications, explained to New Scientist that any such result would first be discussed and agreed within
the collaboration (ATLAS or CMS) that found it, before being communicated to the CERN Director-General. It would then be communicated to the second collaboration, so that the result could be corroborated. Then the heads of other laboratories and the individual member states that contribute to CERN’s funding would be advised. The announcement would then take the form of a seminar organised at CERN. By this time, many thousands of people would be in the know. A further leak appeared not only very possible, but almost inevitable.
So, where would the dam break next?
By 17 June, the LHC had already delivered its milestone 1 inverse femtobarn of data to each of the detector collaborations, an objective that had been set for the whole of 2011. ‘I don’t think we set the targets too low,’ Heuer explained in his mid-year talk to staff. ‘I think we set the targets realistically, but not optimistically. And I must say for me, as a born optimist, the machine was running better than expected.’11
But for Lyn Evans, this was no real surprise. ‘The LHC is working much better than anybody expected, except me,’ he declared. ‘I’m very happy.’12 Evans had joined CERN in 1969 and had been part of the LHC project from its inception at the Lausanne workshop in 1984. He had led the project since 1993. It had been an emotional journey.
With this much data now delivered to both ATLAS and CMS, expectations rose to an all-time high. This should be enough data to provide three-sigma evidence for a Higgs boson in the mass range 135–475 GeV. Or it should be enough data to exclude the Higgs with 95 per cent confidence from 120–530 GeV. Projecting forward to the end of 2012, it seemed that the matter would be definitively settled, one way or another.
‘To my mind, the Shakespeare question for the Higgs – to be, or not to be – will be answered at the end of next year,’ said Heuer.13
All attention now turned to the European Physical Society (EPS) high-energy physics conference in Grenoble, France, scheduled to start on 21 July.
____________
The EPS meeting was to be the first opportunity for the ATLAS and CMS collaborations to share what each had found with over 1 inverse femtobarns of data. That the collaborations could present results literally within weeks of the data being gathered was testimony to the hard work and commitment of hundreds of physicists who had worked tirelessly – and with little sleep – on the analysis.
It became apparent that the Higgs boson (or bosons) – if such existed – was not about to be ‘found’, as such. Instead, possible Higgs mass ranges would first be eliminated from enquiries, narrowing the search to smaller and smaller mass ranges until, finally, the Higgs would be left with nowhere to hide.
The ATLAS collaboration could now exclude a Standard Model Higgs boson with a mass between 155–190 GeV and 295–450 GeV, with 95 per cent confidence. In itself, this was already a powerful result. Finding nothing across such a broad energy range put the cat among several different theoretical pigeons; most of them concerned with physics beyond the Standard Model.
But there was more. The ATLAS data also showed an excess of events above background between 120–145 GeV. This could be due to a number of things, such as errors in the analysis, fluctuations in the background of events that hadn’t been properly anticipated or calculated, or systematic uncertainties in the detector. Or this could be the first intimation that something like a Standard Model Higgs boson, or perhaps even multiple Higgs bosons, were lurking in this energy range.
The excess was dominated by events which could be ascribed to two different Higgs decay channels. These involved a Higgs decaying to two W particles, and thence to two charged leptons and two neutrinos (written H → W+W−→ +ν −ν)*, with a smaller contribution from the channel in which a Higgs decays to two Z0 particles and thence to four charged leptons (written H → Z0Z0 → +−+−).† The former was expected to be one of the most dominant decay channels for a Standard Model Higgs of sufficient mass but, of course, the neutrinos and anti-neutrinos so produced had to be inferred because they can’t be detected, and it is notoriously difficult to distinguish genuine Higgs events from the background. Consequently, only a range of Higgs masses could be implied from data for this channel.
The second channel is much cleaner. In fact, this is the ‘golden’ channel, so called because it is almost completely free from background events and so gives a potentially very precise measure of the Higgs mass. It is also very rare, with only about one in every thousand Higgs bosons decaying this way.
The observed excess of events in the combined ATLAS data was just 2.8 standard deviations, or 2.8-sigma, above background. This was not quite three-sigma ‘evidence’ and far from the five-sigma needed to declare a discovery. It was, nevertheless, a strong hint. What had CMS found?
The CMS collaboration announced 95 per cent confidence limit exclusions over the ranges 149–206 GeV, much of the region 200–300 GeV and 300–440 GeV. The combined CMS data also showed an interesting excess of events between 120–145 GeV, with a statistical significance that had proved difficult to evaluate but which was slightly less than ATLAS had claimed.
This was electrifying. ATLAS and CMS, which before the conference had worked separately, secretively and in competition, had both found much the same thing.
There was still a very long way to go. After the presentations some members of the ATLAS and CMS collaborations gathered together to celebrate with champagne and discuss the next steps. A small working group would be convened to combine the results from the two collaborations and update them with new data to provide a more definitive assessment.
The LHC had continued to break its own records. On 30 July it reached a peak luminosity of 2030 inverse microbarns per second (more than 120 million proton-proton collisions per second). Despite some stability problems, by 7 August the collider had delivered more than 2 inverse femtobarns of data to both ATLAS and CMS. This was already twice the amount of data that had been analysed and shown at the EPS conference.
The combined and updated results would be ready in time for the next big summer conference, the Fifteenth International Symposium on Lepton-Photon Interactions at High Energies, scheduled to commence on 22 August at the Tata Institute in Mumbai, India.
It seemed that the answer to the Shakespeare question could be available within months.
____________
Einstein once declared: ‘The Lord is subtle, but he is not malicious’.* Although the next chapter in the saga of the Higgs search might not betray the existence of an overly malicious deity, as the events unfolded it would become reasonable to accuse the Lord of possessing a certain mischievous wit.
In the weeks before the Mumbai conference, rumours began to circulate in the blogosphere that the combined ATLAS and CMS data now spoke much less ambiguously of a Higgs boson with an energy around 135 GeV. It seemed that the combined ATLAS and CMS data suggested an excess of Higgs decay events with much greater than three-sigma significance. Expectations built in intensity. Although three-sigma evidence would not represent a ‘discovery’, it would be possible to judge from the confidence of the physicists closest to the results whether they believed that this indeed was ‘it’.
I met with Peter Higgs on a wet Thursday afternoon in Edinburgh, a few days before the Mumbai conference was due to commence. Higgs had retired in 1996 but had remained in Edinburgh close to the University department where he had first become a lecturer in mathematical physics in 1960. He was now a sprightly 82 years old. We sat in a coffee shop with his colleague and friend Alan Walker, and talked about his experiences and his hopes for the near future.
Higgs had published the paper that was to bind him forever to the particle that bears his name in 1964.* He had waited 47 years for some kind of vindication. We talked about the prospects for the Mumbai conference, and the grounds we had for optimism that something momentous may be about to be reported. ‘It’s difficult for me now to connect with the person I was then [in 1964],’ he explained, ‘But I’m relieved it’s coming to an end. It will be nice after all this time to be
proved right.’14
Finding the Higgs boson would inevitably be rewarded with a Nobel Prize for the Higgs mechanism, and debates had raged about precisely who of those involved would be recognised by the Nobel Committee: Englert, Higgs, Guralnik, Hagen and Kibble.† We talked about the outbreak of publicity likely to surround a strongly positive announcement from Mumbai and any subsequent announcement from the Swedish Academy. The Press Office at Edinburgh University would be heavily involved. And if things got out of hand Higgs would simply unplug his phone and refuse to answer his doorbell.
But, it seemed, there would be no call for these extreme measures just yet. As the Mumbai conference got underway the following Monday, 22 August, James Gillies at CERN issued a press release. There was no mention of the combined ATLAS and CMS data that had been promised in Grenoble. When updated with another inverse femtobarn or more of collision data gathered in the time between the conferences, the excess events observed by both ATLAS and CMS in the low mass region around 135 GeV had actually declined in significance. ‘Now, with additional data analysed, the significance of those fluctuations has slightly decreased,’ declared the press release, rather solemly.15
It was hard not be disappointed. The hints that had surfaced in the results presented in Grenoble had become less significant in results presented in Mumbai. The outstanding performance of the LHC in delivering more than two inverse femtobarns of data to each detector facility by August had helped build expectations that the ‘Shakespeare question’ might be answered sooner rather than later. The Lord had obviously decided to be mischievous – it wasn’t going to be that easy.
Higgs:The invention and discovery of the 'God Particle' Page 16
