CERN has welcomed more than 120 delegates to an online kick-off workshop for a new collaboration on high-performance computing (HPC). CERN, SKAO (the organisation leading the development of the Square Kilometre Array), GÉANT (the pan-European network and services provider for research and education) and PRACE (the Partnership for Advanced Computing in Europe) will work together to realise the full potential of the coming generation of HPC technology for data-intensive science.
It is an exascale project for an exascale problem
Maria Girone
“It is an exascale project for an exascale problem,” said Maria Girone, CERN coordinator of the collaboration and CERN openlab CTO, in opening remarks at the workshop. “HPC is at the intersection of several important R&D activities: the expansion of computing resources for important data-intensive science projects like the HL-LHC and the SKA, the adoption of new techniques such as artificial intelligence and machine learning, and the evolution of software to maximise the potential of heterogeneous hardware architectures.”
The 29 September workshop, which was organised with the support of CERN openlab, saw participants establish the collaboration’s foundations, outline initial challenges and begin to define the technical programme. Four main initial areas of work were discussed at the event: training and centres of expertise, benchmarking, data access, and authorisation and authentication.
One of the largest challenges in using new HPC technology is the need to adapt to heterogeneous hardware. This involves the development and dissemination of new programming skills, which is at the core of the new HPC collaboration’s plan. A number of examples showing the potential of heterogeneous systems were discussed. One is the EU-funded DEEP-EST project, which is developing a modular supercomputing prototype for exascale computing. DEEP-EST has already contributed to the re-engineering of high-energy physics algorithms for accelerated architectures, highlighting the significant mutual benefits of collaboration across fields when it comes to HPC. PRACE’s excellent record of providing support and training will also be critical to the success of the collaboration.
Benchmarking press
Establishing a common benchmark suite will help the organisations to measure and compare the performance of different types of computing resources for data-analysis workflows from astronomy and particle physics. The suite will include applications representative of the HEP and astrophysics communities – reflecting today’s needs, as well as those of the future – and augment the existing Unified European Applications Benchmark Suite.
Access is another challenge when using HPC resources. Data from the HL-LHC and the SKA will be globally distributed and will be moved over high-capacity networks, staged and cached to reduce latency, and eventually processed, analysed and redistributed. Accessing the HPC resources themselves involves adherence to strict cyber-security protocols. A technical area devoted to authorisation and authentication infrastructure is defining demonstrators to enable large scientific communities to securely access protected resources.
The collaboration will now move forward with its ambitious technical programme. Working groups are forming around specific challenges, with the partner organisations providing access to appropriate testbed resources. Important activities are already taking place in all four areas of work, and a second collaboration workshop will soon be organised.
The AEgIS collaboration at CERN’s Antiproton Decelerator (AD) has reported a milestone in its bid to measure the gravitational free-fall of antimatter – a fundamental test of the weak equivalence principle. Using a series of techniques developed in 2018, the team demonstrated the first pulsed production of antihydrogen, which allows the time at which the antiatoms are formed to be known with high accuracy. This is a key step in determining “g” for antimatter.
“This is the first time that pulsed formation of antihydrogen has been established on timescales that open the door to simultaneous manipulation, by lasers or external fields, of the formed atoms, as well as to the possibility of applying the same method to pulsed formation of other antiprotonic atoms,” says AEgIS spokesperson Michael Doser of CERN. “Knowing the moment of antihydrogen formation is a powerful tool.”
General relativity’s weak equivalence principle holds that all particles with the same initial position and velocity should follow the same trajectories in a gravitational field. It has been verified for matter with an accuracy approaching 10–14. Since theories beyond the Standard Model such as supersymmetry, or the existence of Lorentz-symmetry violating terms, do not necessarily lead to an equivalent force on matter and antimatter, finding even the slightest difference in g would suggest the presence of quantum effects in the gravitational arena. Indirect arguments constrain possible differences to below 10–6g, but no direct measurement for antimatter has yet been performed due to the difficulty in producing and containing large quantities of it.
ALPHA, AEgIS and GBAR are all targeting a measurement of g at the 1% level in the coming years.
Antihydrogen’s neutrality and long lifetime make it an ideal system in which to test this and other fundamental laws, such as CPT invariance. The first production of low-energy antihydrogen, reported in 2002 by the ATHENA and ATRAP collaborations at the AD, involved a three-body recombination reaction (e++e++p → H+e+) involving clouds of antiprotons and positrons. Since then, steady progress by the AD’s ALPHA collaboration in producing, manipulating and trapping ever larger quantities of antihydrogen has enabled spectroscopic and other properties of antimatter to be determined in exquisite detail.
Whereas three-body recombination results in an almost continuous antihydrogen source, in which it is not possible to tag the time of the antiatom formation, AEgIS has employed an alternative charge-exchange process between trapped and cooled antiprotons and positronium (e+e– bound system). Bursts of positrons are accelerated and then implanted into a nano-channelled silicon target above an electromagnetic trap containing cold antiprotons, where, with the aid of laser pulses, they produce a cloud of excited positronium a few millimetres across. This can lead to the formation of antihydrogen within sub-μs timescales, the moment of production being defined by the wellknown laser firing time and the transit time of positronium toward the antiproton cloud. Since the antihydrogen is not trapped in the apparatus, it drifts in all directions until it annihilates on the surrounding material, producing pions and photons that are detected by a scintillating array read out by photomultipliers. The scheme allows the time at which 90% of the atoms are produced to be determined with an uncertainty of around 100 ns.
Further steps are required before the measurement of g can begin, explains Doser. These include the formation of a pulsed beam, greater quantities of antihydrogen, and the ability to make it colder. “With only three months of beam time this year, and lots of new equipment to commission, most likely 2022 will be the year in which we establish pulsed beam formation, which is a prerequisite for us to perform a gravity measurement.”
Targeted approach
Following a proof-of-principle measurement of g for antihydrogen by the ALPHA collaboration in 2013, ALPHA, AEgIS and a third AD experiment, GBAR, are all targeting a measurement of g at the 1% level in the coming years. In contrast to AEgIS’s approach, whereby the vertical deviation of a pulsed horizontal beam of cold antihydrogen atoms will be measured in an approximately 1 m-long flight tube, GBAR will take advantage of advances in ion-cooling techniques to measure ultraslow antihydrogen atoms as they fall from a height of 20 cm. ALPHA, meanwhile, will release antihydrogen atoms from a vertical magnetic trap and measure the distribution of annihilation positions when they hit the wall – ramping the trap down slowly so that the coldest atoms, which are most sensitive to gravity, come out last. All three experiments have recently been hooked up to the AD’s ELENA synchrotron, which enables the production of very low-energy antiprotons.
Given that most of the mass of antinuclei comes from massless gluons that bind their constituent quarks, physicists think it unlikely that antimatter experiences an opposite gravitational force to matter and therefore “falls up”. Nevertheless, precise measurements of the free fall of antiatoms could reveal subtle differences that would open an important crack in current understanding.
Popular representations of the Standard Model (SM) often hide its beautiful weirdness, for example slotting quarks and leptons into boxes and arranging them like a low-grade Mendeleev, or contriving a dartboard arrangement. The “double simplex” scheme invented in 2005 by US theorist Chris Quigg, which was recently given a flashy makeover by Quanta magazine (see image), is much richer (arXiv:hep-ph/0509037).
Jogesh Pati and Abdus Salam’s suggestion, in their 1974 shot at a grand unified theory, that lepton number be regarded as a fourth colour, inspired Quigg to place the leptons at the fourth point of an SU(4) tetrahedron. The additional edges therefore represent possible leptoquark transitions. Left-handed fermion doublets (left) are reflected in the broken mirror of parity to reveal right-handed fermion singlets (right), though Quanta, unlike Quigg perhaps favouring a purely left-handed Majorana mass term, omit possible right-handed neutrinos.
A final distinction is that Quigg chooses to superimpose the left and right simplexes – a term for a generalised triangle or tetrahedron in an arbitrary number of dimensions – while Quanta elects to separate the tetrahedra, and label couplings to the Higgs boson with sweeping loops. This obscures a beautiful feature of Quigg’s design, whereby the Yukawa couplings hypothesised by the SM, which couple the left- and right-handed incarnations of massive fermions in interactions with the Higgs field, link opposite corners of the superimposed double simplex, placing the Higgs boson at the centre of the picture. Quigg, who intended that the double simplex precipitate questions, also points out that the corners of the superimposed tetrahedra define a cube, whose edges suggest a possible new category of feeble interactions yet to be discovered.
Mait Müntel came to CERN as a summer student in 2004 and quickly became hooked on particle physics, completing a PhD in the CMS collaboration in 2008 with a thesis devoted to signatures of double-charged Higgs bosons. Continuing in the field, he was one of the first to do shifts in the CMS control room when the LHC ramped up. It was then that he realised that the real LHC data looked nothing like the Monte Carlo simulations of his student days. Many things had to be rectified, but Mait admits he was none too fond of coding and didn’t have any formal training. “I thought I would simply ‘learn by doing’,” he says. “However, with hindsight, I should probably have been more systematic in my approach.” Little did he know that, within a few years, he would be running a company with around 40 staff developing advanced language-learning algorithms.
Memory models
Despite spending long periods in the Geneva region, Mait had not found the time to pick up French. Frustrated, he began to take an interest in the use of computers to help humans learn languages at an accelerated speed. “I wanted to analyse from a statistical point of view the language people were actually speaking, which, having spent several years learning both Russian and English, I was convinced was very different to what is found in academic books and courses,” he says. Over the course of one weekend, he wrote a software crawler that enabled him to download a collection of French subtitles from a film database. His next step was to study memory models to understand how one acquires new knowledge, calculating that, if a computer program could intelligently decide what would be optimal to learn in the next moment, it would be possible to learn a language in only 200 hours. He started building some software using ROOT (the object-oriented program and library developed by CERN for data analysis) and, within two weeks, was able to read a proper book in French. “I had included a huge book library in the software and as the computer knew my level of vocabulary, it could recommend books for me. This was immensely gratifying and pushed me to progress even further.” Two months later, he passed the national French language exam in Estonia.
Mait became convinced that he had to do something with his idea. So he went on holiday, and hired two software developers to develop his code so it would work on the web. Whilst on holiday, he happened to meet a friend of a friend, who helped him set up Lingvist as a company. Estonia, he says, has a fantastic start-up and software-development culture thanks to Skype, which was invented there. Later, Mait met the technical co-founder of Skype at a conference, who coincidentally had been working on software to accelerate human learning. He dropped his attempts and became Lingvist’s first investor.
Short-term memory capabilities can differ between five minutes and two seconds!
Mait Müntel
The pair secured a generous grant from the European Union Horizon 2020 programme and things were falling into place, though it wasn’t all easy says Mait: “You can use the analogy of sitting in a nice warm office at CERN, surrounded by beautiful mountains. In the office, you are safe and protected, but if you go outside and climb the mountains, you encounter rain and hail, it is an uphill struggle and very uncomfortable, but immensely satisfying when you reach the summit. Even if you work more than 100 hours per week.”
Lingvist currently has three million users, and Mait is convinced that the technology can be applied to all types of education. “What our data have demonstrated is that levels of learning in people are very different. Short-term memory capabilities can differ between five minutes and two seconds! Currently, based on our data, the older generation has much better memory characteristics. The benefit of our software is that it measures memory, and no matter one’s retention capabilities, the software will help improve retention rates.”
New talents
Faced with a future where artificial intelligence will make many jobs extinct, and many people will need to retrain, competitiveness will be derived from the speed at which people can learn, says Mait. He is now building Lingvist’s data-science research team to grow the company to its full potential, and is always on the lookout for new CERN talent. “Traditionally, physicists have excellent modelling, machine-learning and data-analysis skills, even though they might not be aware of it,” he says.
Besides the intrinsic worth of the knowledge that it generates, particle physics often acts as a trailblazer in developing cutting-edge technologies in the fields of accelerators, detectors and computing. These technologies, and the human expertise associated with them, find applications in a variety of areas, including the biomedical field, and can have a societal impact going way beyond their initial scope and expectations.
This webinar will introduce the knowledge-transfer goals of CERN, give an overview of the Laboratory’s medical-applications-related activities and give examples of the impact of CERN technologies on medtech: from hadrontherapy to medical imaging, flash radiotherapy, computing and simulation tools. It will also touch upon the challenges of transferring the technologies and know-how from CERN to the medtech industry and medical research.
Dr Manuela Cirilli is the deputy group leader of CERN’s Knowledge Transfer (KT) group, whose mission is to maximise the impact of CERN on society by creating opportunities for the transfer of the Laboratory’s technologies and know-how to fields outside particle physics. Manuela leads the Medical Applications section of the KT group and chairs the CERN Medical Applications Project Forum. She has an academic background in particle physics and science communication. In 1997, she started working on the NA48 experiment at CERN, designed to measure CP violation in the kaon system. In 2001, she began working on the construction, commissioning and calibration of the precision muon chambers of the ATLAS experiment at the LHC, until she joined CERN’s KT group in 2010.
In parallel to her career, Manuela has been actively engaging in science communication and popularisation since the early 2000s.
Evidence for the decay of the Higgs boson to a photon and a low-mass electron or muon pair, propagated predominantly by a virtual photon (γ*), H → γ*γ → ℓℓγ (where ℓ = e or μ), has been obtained at the LHC. At an LHC seminar today, the ATLAS collaboration reported a 3.2σ excess over background of H → ℓℓγ decay candidates with dilepton mass mℓℓ < 30 GeV.
The H → ℓℓγ decay is particularly interesting as it is a loop process
The measurement of rare decays of the Higgs boson is a crucial component of the Higgs-boson physics programme at the LHC, since they probe potential new interactions with the Higgs boson introduced by possible extensions of the Standard Model. The H → ℓℓγ decay is particularly interesting in this respect as it is a loop process and the three-body final state allows the CP structure of the Higgs boson to be probed. However, the small expected signal-to-background ratio and the typically low dilepton invariant mass make the search for H → ℓℓγ highly challenging.
The analysis performed by ATLAS searched for H → e+e–γ and H → μ+μ–γ decays. Special treatment was needed in particular for the electron channel: a dedicated electron trigger was developed as well as a specific identification algorithm. The predicted mℓℓ spectrum rises steeply towards lower values, with a kinematic cutoff at twice the final-state lepton mass. At such low electron–positron invariant masses, and given the large transverse momentum of their system, the electromagnetic showers induced by the electron and the positron in the ATLAS calorimeter can merge, requiring a specially developed reconstruction. Furthermore, a dedicated identification algorithm was developed for these topologies, and its efficiency was measured in data using photon detector-material conversions at low radius into an electron–positron pair from Z → ℓℓγ events.
The signal extraction is performed by searching in the ℓℓγ invariant mass (mℓℓγ) range between 110 and 160 GeV for a narrow signal peak over smooth background at the mass of the Higgs boson. The sensitivity to the H → ℓℓγ signal was increased by separating events in mutually exclusive categories based on lepton types and event topologies. ATLAS reports evidence in data for a H → ℓℓγ signal emerging over the background with a significance of 3.2σ (see figure). The Higgs boson production cross section times H → ℓℓγ branching fraction, measured for mℓℓ < 30 GeV, amounts to 8.7+2.8–2.7 fb. It corresponds to a signal strength – the ratio of the measured cross section times branching fraction to the Standard Model prediction – of 1.5 ± 0.5. With this, ATLAS has also extended the invariant-mass range of the lepton pair for the related Higgs-boson decay into a photon and a Z boson to lower masses, opening the door to future studies of three-body Higgs-boson decays and investigations of its underlying CP structure.
H Frederick Dylla is a “Sputnik kid”, whose curiosity and ingenuity led him on a successful 50-year career in physics, from plasma to accelerators and leading the American Institute of Physics. His debut book, Scientific Journeys: A Physicist Explores the Culture, History and Personalities of Science, is a collection of essays that puts a multidisciplinary historical perspective on the actors and events that shaped the world of science and scholarly publishing. Through geopolitical and economic context and a rich record of key events, he highlights innovations that have found their use in social and business applications. Those cited as having contributed to global technological progress range from the web and smartphones to medical imaging and renewable energy.
Dylla begins with the story of medieval German abbess, mystic, composer and medicinal botanist Hildegard of Bingen
The book is divided in five chapters: “signposts” (in the form of key people and events in scientific history); mentors and milestones in his life; science policy; communicating science; and finally a brief insight into the relationship between science and art. He begins with the story of medieval German abbess, mystic, composer and medicinal botanist Hildegard of Bingen: “a bright signpost of scholarship”. Dylla goes on to explore the idea that a single individual at the right time and place can change the course of history. Bounding through the centuries, he highlights the importance of science policy and science communication, the funding of big and small science alike, and the contemporary challenges linked to research, teaching science and scholarly publishing. Examples among these, says Dylla, are the protection of scientific integrity, new practices of distance learning and the weaknesses of the open-access model. The book ends bang up to date with a thought on the coronavirus pandemic and science’s key role in overcoming it.
Intended for teachers, science historians and students from high school to graduate school, Dylla’s book puts a face on scientific inventions. The weightiest chapter, mentors and milestones, focuses on personalities who have played an important role in his scientific voyage. Among the many named, however, Mildred Dresselhaus – the “queen of carbon” – is the only female scientist featured in the book besides Hildegard. Though by beginning the book with a brilliant but at best scientifically adjacent abbess who preceded Galileo by four centuries Dylla tacitly acknowledges the importance of representing diversity, the book unintentionally makes it discomfortingly clear how scarce role models for women can be in the white-male dominated world of science. The lack of a discussion on diversity is a missed opportunity in an otherwise excellent book.
The International Linear Collider (ILC) is a proposed electron–positron linear collider with a Higgs factory operating at a centre-of-mass energy of 250 GeV (ILC250) as a first stage. Its electron and positron beams can be longitudinally polarised, and the accelerator may be extended to operate at 500 GeV up to 1 TeV, and possibly beyond. In addition, the unique time structure of the ILC beams (which would collide at short bursts of 1312 bunches with 0.554 ms spacing at a frequency of 5 Hz) places much less stringent requirements on readout speed and radiation hardness than conditions at the LHC detectors. This allows the use of low-mass tracking and high-granularity sensors in the ILC detectors, giving unprecedented resolution in jet-energy measurements. It also results in an expected data rate of just a few GB/s, allowing collisions to be recorded without a trigger.
ILC250 primarily targets precision measurements of the Higgs boson (see Targeting a Higgs factory). However, fully exploiting these measurements demands substantial improvement in our knowledge about many other Standard Model (SM) observables. Here, ILC250 opens three avenues: the study of gauge-boson pair-production and fermion pair-production at 250 GeV; fermion-pair production at effective centre-of-mass energies lowered to about 91.2 GeV by prior emission of photons (radiative returns to the Z pole); and operation of the collider at both the Z pole and the WW threshold. In all of these cases, the polarisation of the electron and positron beams (at polarisations up to 80% and 30%–60%, respectively) boosts the statistical power of many measurements by factors between 2.5 (for Higgs measurements) and 10 (at the Z pole), thanks to the ability to exploit observables such as left–right asymmetries of production cross-sections. These additional polarisation-dependent observables are also essential to disentangle the unavoidable interference between Z and γ exchange in fermion pair-production at energies above the Z pole, enabling access to the chiral couplings of fermions to the Z and the photon. Broadly speaking, the polarised beams and the high luminosity of ILC250 will lead to at least one order of magnitude improvement over the current knowledge for many SM precision observables.
Other important inputs when interpreting Higgs measurements are charged triple-gauge couplings (TGCs), which are also probes of physics beyond the SM. ILC250 will measure these 100 times more precisely than LEP, with a further factor-of-two improvement possible at the higher-energy stage ILC500. These numbers refer to the case of extracting simultaneously all three TGCs relevant in SM effective field theory, which is currently the most favoured framework for the interpretation of precision Higgs-boson data, whereas TGC results from the LHC assume that only one of these couplings deviates from its SM value at a time. With both beams polarised and with full control over the orientation of the polarisation vectors, all 28 TGC parameters that exist in the most general case can potentially be determined simultaneously at the ILC.
Z-pole physics
Classic electroweak precision observables refer to the Z pole. ILC250 will produce about 90 million visible Z events via radiative return, which is about five times more than at LEP and 100 times more than SLC. Thanks to the polarised beams, these data will allow a direct measurement of the asymmetry Ae between the left- and right-handed electron’s coupling to the Z boson with 10 times better accuracy than today, and enable the asymmetries Af of the final-state fermions to the Z to be directly extracted. This is quite different from the case of unpolarised beams, where only the product Ae Af can be accessed. Compared to LEP/SLC results, the Z-pole asymmetries can be improved by typically a factor of 20 using only the radiative returns to the Z at ILC250. This would settle beyond doubt the long-standing question of whether the 3σ tension between the weak mixing-angle extractions from SLC and LEP originates from physics beyond the SM. With a few minor modifications, the ILC can also directly operate at the Z pole, improving fermion asymmetries by another factor 6 to 25 with respect to the radiative-return results.
The higher integrated luminosity of the ILC will provide new opportunities to search for physics beyond the SM
At energies above the Z pole, di-fermion production is sensitive to hypothetical, heavy siblings of the Z boson (so-called Z′ bosons) and to four-fermion operators, i.e. contact-interaction-like parametrisations of yet unknown interactions. ILC250 could indirectly discover Z′ particles with masses up to 6 TeV, while ILC1000 could extend the reach to 18 TeV. For contact interactions, depending on the details of the assumed model, compositeness scales of up to 160 TeV can be probed at ILC250, and up to nearly 400 TeV at ILC1000.
Direct searches for new physics
At first glance, it might seem that direct searches at ILC250 offer only a marginal improvement over LEP, which attained a collision energy of 209 GeV. Nevertheless, the higher integrated luminosity of the ILC (about 2000 times higher than LEP’s above the WW threshold), its polarised beams, much-improved detectors, and triggerless readout will provide new opportunities to search for physics beyond the SM. For example, ILC250 will improve on LEP searches for a new scalar particle produced in association with the Z boson by over an order of magnitude. Another example of a rate-limited search at LEP is the supersymmetric partner of the tau lepton, the tau slepton. In the most general case, tau-slepton masses above 26.3 GeV are not excluded, and in this case no improvement from HL-LHC is expected. The ILC, with its highly-granular detectors covering angles down to 6 mrad with respect to the collision axis, has the ability to cover masses up to nearly the kinematic limit of half the collision energy, also in the experimentally most difficult parts of the parameter space.
The absence of discoveries of new high-mass states at the LHC has led to increased interest in fermionic “Z-portal” models, with masses of dark-matter particles below the electroweak scale. A dark photon, for example, could be detected via its mixing with SM photons. In searching for such phenomena, ILC250 could cover the region between the reach of the B-factories, which is limited to below 10 GeV, and the LHC experiments, which start searching in a range above 150 GeV.
The ILC’s Higgs-factory stage will require only about 40% of the tunnel length available at the Kitakami Mountains in northern Japan, which is capable of housing a linear collider at least 50 km long. This is sufficient to reach a centre-of-mass energy of 1 TeV with current technology by extending the linacs and augmenting power and cryogenics. The upgrade to ILC500 is expected to cost approximately 60% of the ILC250 cost, while going to 1 TeV would require an estimated 100% of the ILC250 cost, assuming a modest increase of the accelerating gradient over what has been achieved (CERN Courier November/December 2020 p35). These upgrades offer the opportunity to optimise the exact energies of the post-Higgs-factory stages according to physics needs and technological advances.
ILC at higher energies
ILC500 targets the energy range 500–600 GeV, which would improve the precision on Higgs-boson couplings typically by a factor of two compared to ILC250 and on charged triple-gauge couplings by a factor of three to four. It would also offer optimal sensitivity in three important measurements. The first is the electroweak couplings of the top quark, for which a variety of new-physics models predict deviations for instance in its coupling to the Z (see “Model sensitivity” figure). The second is the Higgs self-coupling λ from double Higgs-strahlung (e+e–→ ZHH): while ILC500 could reach a precision of 27% on λ, at 1 TeV a measurement based on vector-boson fusion (VBF) reaches 10%. These numbers assume that λ takes the value predicted by the SM. However, the situation can be quite different if λ is larger, as is typically required by models of baryogenesis, and only the
combination of double Higgs-strahlung and VBF-based measurements can guarantee a precision of at least 10–20% for any value of λ (see “Higgs self-coupling” figure). A third physics target is the top-quark Yukawa coupling, for which a precision of 6.3% is projected at ILC500, 3.2% at 550 GeV and 1.6% at 1 TeV.
While ILC250 has interesting discovery potential in various rate-limited searches, ILC500 extends the kinematic reach significantly beyond LEP. For instance, in models of supersymmetry that adhere to naturalness, the supersymmetric partners of the Higgs boson (the higgsinos) must have masses that are not too far from the Z or Higgs bosons, typically around 100 to 300 GeV. While the lower range of these particles is already accessible at ILC250, the higher energy stages of the ILC will be able to cover the remainder of this search space. The ILC is also able to reconstruct decay chains when the mass differences among higgsinos are small, which is a challenging signature for the HL-LHC.
The ILC is the only future collider that is currently being discussed at the government level, by Japan, the US and various countries in Europe. It is also the most technologically established proposal, its cutting edge radio-frequency cavities already in operation at the European XFEL. The 2020 update of the European strategy for particle physics also noted that, should an ILC in Japan go ahead, the European particle-physics community would wish to collaborate. Recently, an ILC international development team was established to prepare for the creation of the ILC pre-laboratory, which will make all necessary technical preparations before construction can begin. If intergovernmental negotiations are successful, the ILC could undergo commissioning as early as the mid-2030s.
The ability to collide high-energy beams of hadrons under controlled conditions transformed the field of particle physics. Until the late 1960s, the high-energy frontier was dominated by the great proton synchrotrons. The Cosmotron at Brookhaven National Laboratory and the Bevatron at Lawrence Berkeley National Laboratory were soon followed by CERN’s Proton Synchrotron and Brookhaven’s Alternating Gradient Synchrotron, and later by the Proton Synchrotron at Serpukov near Moscow. In these machines protons were directed to internal or external targets in which secondary particles were produced.
The kinematical inefficiency of this process, whereby the centre-of-mass energy only increases as the square root of the beam energy, was recognised from the outset. In 1943, Norwegian engineer Rolf Widerøe proposed the idea of colliding beams, keeping the centre of mass at rest in order to exploit the full energy for the production of new particles. One of the main problems was to get colliding beam intensities high enough for a useful event rate to be achieved. In the 1950s the prolific group at the University of Wisconsin Midwestern Universities Research Association (MURA), led by Donald Kerst, worked on the problem of “stacking” particles, whereby successive pulses from an injector synchrotron are superposed to increase the beam intensity. They mainly concentrated on protons, where Liouville’s theorem (which states that for a continuous fluid under the action of conservative forces the density of phase space cannot be increased) was thought to apply. Only much later, ways to beat Liouville and to increase the beam density were found. At the 1956 International Accelerator Conference at CERN, Kerst made the first proposal to use stacking to produce colliding beams (not yet storage rings) of sufficient intensity.
At that same conference, Gerry O’Neill from Princeton presented a paper proposing that colliding electron beams could be achieved in storage rings by making use of the natural damping of particle amplitudes by synchrotron-radiation emission. A design for the 500 MeV Princeton–Stanford colliding beam experiment was published in 1958 and construction started that same year. At the same time, the Budker Institute for Nuclear Research in Novosibirsk started work on VEP-1, a pair of rings designed to collide electrons at 140 MeV. Then, in March 1960, Bruno Touschek gave a seminar at Laboratori Nazionali di Frascati in Italy where he first proposed a single-ring, 0.6 m-circumference 250 MeV electron–positron collider. “AdA” produced the first stored electron and positron beams less than one year later – a far cry from the time it takes today’s machines to go from conception to operation! From these trailblazers evolved the production machines, beginning with ADONE at Frascati and SPEAR at SLAC. However, it was always clear that the gift of synchrotron-radiation damping would become a hindrance to achieving very high energy collisions in a circular electron–positron collider because the power radiated increases as the fourth power of the beam energy and the inverse fourth power of mass, so is negligible for protons compared with electrons.
A step into the unknown
Meanwhile, in the early 1960s, discussion raged at CERN about the next best step for particle physics. Opinion was sharply divided between two camps, one pushing a very high-energy proton synchrotron for fixed-target physics and the other using the technique proposed at MURA to build an innovative colliding beam proton machine with about the same centre-of-mass energy as a conventional proton synchrotron of much larger dimensions. In order to resolve the conflict, in February 1964, 50 physicists from among Europe’s best met at CERN. From that meeting emerged a new committee, the European Committee for Future Accelerators, under the chairmanship of one of CERN’s founding fathers, Edoardo Amaldi. After about two years of deliberation, consensus was formed. The storage ring gained most support, although a high-energy proton synchrotron, the Super Proton Synchrotron (SPS), was built some years later and would go on to play an essential role in the development of hadron storage rings. On 15 December 1965, with the strong support of Amaldi, the CERN Council unanimously approved the construction of the Intersecting Storage Rings (ISR), launching the era of hadron colliders.
On 15 December 1965, the CERN Council unanimously approved the construction of the ISR, launching the era of hadron colliders
First collisions
Construction of the ISR began in 1966 and first collisions were observed on 27 January 1971. The machine, which needed to store beams for many hours without the help of synchrotron-radiation damping to combat inevitable magnetic field errors and instabilities, pushed the boundaries in accelerator science on all fronts. Several respected scientists doubted that it would ever work. In fact, the ISR worked beautifully, exceeding its design luminosity by an order of magnitude and providing an essential step in the development of the next generation of hadron colliders. A key element was the performance of its ultra-high-vacuum system, which was a source of continuous improvement throughout the 13 year-long lifetime of the machine.
For the experimentalists, the ISR’s collisions (which reached an energy of 63 GeV) opened an exciting adventure at the energy frontier. But they were also learning what kind of detectors to build to fully exploit the potential of the machine – a task made harder by the lack of clear physics benchmarks known at the time in the ISR energy regime. The concept of general-purpose instruments built by large collaborations, as we know them today, was not in the culture of the time. Instead, many small collaborations built experiments with relatively short lifecycles, which constituted a fruitful learning ground for what was to come at the next generation of hadron colliders.
There was initially a broad belief that physics action would be in the forward directions at a hadron collider. This led to the Split Field Magnet facility as one of the first detectors at the ISR, providing a high magnetic field in the forward directions but a negligible one at large angle with respect to the colliding beams (the nowadays so-important transverse direction). It was with subsequent detectors featuring transverse spectrometer arms over limited solid angles that physicists observed a large excess of high transverse momentum particles above low-energy extrapolations. With these first observations of point-like parton scattering, the ISR made a fundamental contribution to strong-interaction physics. Solid angles were too limited initially, and single-particle triggers too biased, to fully appreciate the hadronic jet structure. That feat required third-generation detectors, notably the Axial Field Spectrometer (AFS) at the end of the ISR era, offering full azimuthal central calorimeter coverage. The experiment provided evidence for the back-to-back two-jet structure of hard parton scattering.
For the detector builders, the original AFS concept was interesting as it provided an unobstructed phi-symmetric magnetic field in the centre of the detector, however, at the price of massive Helmholtz coil pole tips obscuring the forward directions. Indeed, the ISR enabled the development of many original experimental ideas. A very important one was the measurement of the total cross section using very forward detectors in close proximity to the beam. These “Roman Pots”, named for their inventors, made their appearance in all later hadron colliders, confirming the rising total pp cross section with energy.
It is easy to say after the fact, still with regrets, that with an earlier availability of more complete and selective (with electron-trigger capability) second- and third-generation experiments at the ISR, CERN would not have been left as a spectator during the famous November revolution of 1974 with the J/ψ discoveries at Brookhaven and SLAC. These, and the ϒ resonances discovered at Fermilab three years later, were clearly observed in the later-generation ISR experiments.
SPS opens new era
However, events were unfolding at CERN that would pave the way to the completion of the Standard Model. At the ISR in 1972, the phenomenon of Schottky noise (density fluctuations due to the granular nature of the beam in a storage ring) was first observed. It was this very same noise that Simon van der Meer speculated in a paper a few years earlier could be used for what he called “stochastic cooling” of a proton beam, beating Liouville’s theorem by the fact that a beam of particles is not a continuous fluid. Although it is unrealistic to detect the motion of individual particles and damp them to the nominal orbit, van der Meer showed that by correcting the mean transverse motion of a sample of particles continuously, and as long as the statistical nature of the Schottky signal was continuously regenerated, it would be theoretically possible to reduce the beam size and increase its density. With the bandwidth of electronics available at the time, van der Meer concluded that the cooling time would be too long to be of practical importance. But the challenge was taken up by Wolfgang Schnell, who built a state-of-the-art feedback system that demonstrated stochastic cooling of a proton beam for the first time. This would open the door to the idea of stacking and cooling of antiprotons, which later led to the SPS being converted into a proton–antiproton collider.
Another important step towards the next generation of hadron colliders occurred in 1973 when the collaboration working on the Gargamelle heavy-liquid bubble chamber published two papers revealing the first evidence for weak neutral currents. These were important observations in support of the unified theory of electromagnetic and weak interactions, for which Sheldon Glashow, Abdus Salam and Steven Weinberg were to receive the Nobel Prize in Physics in 1979. The electroweak theory predicted the existence and approximate masses of two vector bosons, the W and the Z, which were too high to be produced in any existing machine. However, Carlo Rubbia and collaborators proposed that, if the SPS could be converted into a collider with protons and antiprotons circulating in opposite directions, there would be enough energy to create them.
To achieve this the SPS would need to be converted into a storage ring like the ISR, but this time the beam would need to be kept “bunched” with the radio-frequency (RF) system working continuously to achieve a high enough luminosity (unlike the ISR where the beams were allowed to de-bunch all around the ring). The challenges here were two-fold. Noise in the RF system causes particles to diffuse rapidly from the bunch. This was solved by a dedicated feedback system. It was also predicted that the beam–beam interaction would limit the performance of a bunched-beam machine with no synchrotron-radiation damping due to the strongly nonlinear interactions between a particle in one beam with the global electromagnetic field in the other beam.
A much bigger challenge was to build an accumulator ring in which antiprotons could be stored and cooled by stochastic cooling until a sufficient intensity of antiprotons would be available to transfer into the SPS, accelerate to around 300 GeV and collide with protons. This was done in two stages. First a proof-of-principle was needed to show that the ideas developed at the ISR transferred to a dedicated accumulator ring specially designed for stochastic cooling. This ring was called the Initial Cooling Experiment (ICE), and operated at CERN in 1977–1978. In ICE transverse cooling was applied to reduce the beam size and a new technique for reducing the momentum spread in the beam was developed. The experiment proved to be a big success and the theory of stochastic cooling was refined to a point where a real accumulator ring (the Antiproton Accumulator) could be designed to accumulate and store antiprotons produced at 3.5 GeV by the proton beam from the 26 GeV Proton Synchrotron. First collisions of protons and antiprotons at 270 GeV were observed on the night of 10 July 1981, signalling the start of a new era in colliding beam physics.
A clear physics goal, namely the discovery of the W and Z intermediate vector bosons, drove the concepts for the two main SppS experiments UA1 and UA2 (in addition to a few smaller, specialised experiments). It was no coincidence that the leaders of both collaborations were pioneers of ISR experiments, and many lessons from the ISR were taken on board. UA1 pioneered the concept of a hermetic detector that covered as much as possible the full solid angle around the interaction region with calorimetry and tracking. This allows measurements of the missing transverse energy/momentum, signalling the escaping neutrino in the leptonic W decays. Both electrons and muons were measured, with tracking in a state-of-the-art drift chamber that provided bubble-chamber-like pictures of the interactions. The magnetic field was provided by a dipole-magnet configuration, an approach not favoured in later generation experiments because of its inherent lack of azimuthal symmetry. UA2 featured a (at the time) highly segmented electromagnetic and hadronic calorimeter in the central part (down to 40 degrees with respect to the beam axis), with 240 cells pointing to the interaction region. But it had no muon detection, and in its initial phase only limited electromagnetic coverage in the forward regions. There was no magnetic field except for the forward cones with toroids to probe the W polarisation.
In 1983 the SppS experiments made history with the direct discoveries of the W and Z. Many other results were obtained, including the first evidence of neutral B-meson particle–antiparticle mixing at UA1 thanks to its tracking and muon detection. The calorimetry of UA2 provided immediate unambiguous evidence for a two-jet structure in events with large transverse energy. Both UA1 and UA2 pushed QCD studies far ahead. The lack of hermeticity in UA2’s forward regions motivated a major upgrade (UA2′) for the second phase of the collider, complementing the central part with new fully hermetic calorimetry (both electromagnetic and hadronic), and also inserting a new tracking cylinder employing novel technologies (fibre tracking and silicon pad detectors). This enabled the experiment to improve searches for top quarks and supersymmetric particles, as well as making almost background-free first precision measurements of the W mass.
Meanwhile in America
At the time the SppS was driving new studies at CERN, the first large superconducting synchrotron (the Tevatron, with a design energy close to 1 TeV) was under construction at Fermilab. In view of the success of the stochastic cooling experiments, there was a strong lobby at the time to halt the construction of the Tevatron and to divert effort instead to emulate the SPS as a proton–antiproton collider using the Fermilab Main Ring. Wisely this proposal was rejected and construction of the Tevatron continued. It came into operation as a fixed-target synchrotron in 1984. Two years later it was also converted into a proton–antiproton collider and operated at the high-energy frontier until its closure in September 2011.
A huge step was made with the detector concepts for the Tevatron experiments, in terms of addressed physics signatures, sophistication and granularity of the detector components. This opened new and continuously evolving avenues in analysis methods at hadron colliders. Already the initial CDF and DØ detectors for Run I (which lasted until 1996) were designed with cylindrical concepts, characteristic of what we now call general-purpose collider experiments, albeit DØ still without a central magnetic field in contrast to CDF’s 1.4 T solenoid. In 1995 the experiments delivered the first Tevatron highlight: the discovery of the top quark. Both detectors underwent major upgrades for Run II (2001–2011) – a theme now seen for the LHC experiments – which had a great impact on the Tevatron’s physics results. CDF was equipped with a new tracker, a silicon vertex detector, new forward calorimeters and muon detectors, while DØ added a 1.9 T central solenoid, vertexing and fibre tracking, and new forward muon detectors. Alongside the instrumentation was a breath-taking evolution in real-time event selection (triggering) and data acquisition to keep up with the increasing luminosity of the collider.
The physics harvest of the Tevatron experiments during Run II was impressive, including a wealth of QCD measurements and major inroads in top-quark physics, heavy-flavour physics and searches for phenomena beyond the Standard Model. Still standing strong are its precision measurements of the W and top masses and of the electroweak mixing angle sin2θW. The story ended in around 2012 with a glimpse of the Higgs boson in associated production with a vector boson. The CDF and DØ experience influenced the LHC era in many ways: for example they were able to extract the very rare single-top production cross-section with sophisticated multivariate algorithms, and they demonstrated the power of combining mature single-experiment measurements in common analyses to achieve ultimate precision and sensitivity.
For the machine builders, the pioneering role of the Tevatron as the first large superconducting machine was also essential for further progress. Two other machines – the Relativistic Heavy Ion Collider at Brookhaven and the electron–proton collider HERA at DESY – derived directly from the experience of building the Tevatron. Lessons learned from that machine and from the SppS were also integrated into the design of the most powerful hadron collider yet built: the LHC.
The Large Hadron Collider
The LHC had a difficult birth. Although the idea of a large proton–proton collider at CERN had been around since at least 1977, the approval of the Superconducting Super Collider (SSC) in the US in 1987 put the whole project into doubt. The SSC, with a centre-of-mass energy of 40 TeV, was almost three times more powerful than what could ever be built using the existing infrastructure at CERN. It was only the resilience and conviction of Carlo Rubbia, who shared the 1984 Nobel Prize in Physics with van der Meer for the project leading to the discovery of the W and Z bosons, that kept the project alive. Rubbia, who became Director-General of CERN in 1989, argued that, in spite of its lower energy, the LHC could be competitive with the SSC by having a luminosity an order of magnitude higher, and at a fraction of the cost. He also argued that the LHC would be more versatile: as well as colliding protons, it would be able to accelerate heavy ions to record energies at little extra cost.
The SSC was eventually cancelled in 1993. This made the case for the LHC even stronger, but the financial climate in Europe at the time was not conducive to the approval of a large project. For example, CERN’s largest contributor, Germany, was struggling with the cost of reunification and many other countries were getting to grips with the introduction of the single European currency. In December 1993 a plan was presented to the CERN Council to build the machine over a 10-year period by reducing the other experimental programmes at CERN to the absolute minimum, with the exception of the full exploitation of the flagship Large Electron Positron (LEP) collider. Although the plan was generally well received, it became clear that Germany and the UK were unlikely to agree to the budget increase required. On the positive side, after the demise of the SSC, a US panel on the future of particle physics recommended that “the government should declare its intentions to join other nations in constructing the LHC”. Positive signals were also being received from India, Japan and Russia.
In June 1994 the proposal to build the LHC was made once more. However, approval was blocked by Germany and the UK, which demanded substantial additional contributions from the two host states, France and Switzerland. This forced CERN to propose a “missing magnet” machine where only two thirds of the dipole magnets would be installed in a first stage, allowing operation at reduced energy for a number of years. Although costing more in the long run, the plan would save some 300 million Swiss Francs in the first phase. This proposal was put to Council in December 1994 by the new Director-General Christopher Llewellyn Smith and, after a round of intense discussions, the project was finally approved for two-stage construction, to be reviewed in 1997 after non-Member States had made known their contributions. The first country to do so was Japan in 1995, followed by India, Russia and Canada the next year. A final sting in the tail came in June 1996 when Germany unilaterally announced that it intended to reduce its CERN subscription by between 8% and 9%, prompting the UK to demand a similar reduction and forcing CERN to take out loans. At the same time, the two-stage plan was dropped and, after a shaky start, the construction of the full LHC was given the green light.
The fact that the LHC was to be built at CERN, making full use of the existing infrastructure to reduce cost, imposed a number of strong constraints. The first was the 27 km-circumference of the LEP tunnel in which the machine was to be housed. For the LHC to achieve its design energy of 7 TeV per beam, its bending magnets would need to operate at a field of 8.3 T, about 60% higher than ever achieved in previous machines. This could only be done using affordable superconducting material by reducing the temperature of the liquid-helium coolant from its normal boiling point of 4.2 K to 1.9 K – where helium exists in a macroscopic quantum state with the loss of viscosity and a very large thermal conductivity. A second major constraint was the small (3.8 m) tunnel diameter, which made it impossible to house two independent rings like the ISR. Instead, a novel and elegant magnet design, first proposed by Bob Palmer at Brookhaven, with the two rings separated by only 19 cm in a common yoke and cryostat was developed. This also considerably reduced the cost.
This journey is now poised to continue, as we look ahead towards how a general-purpose detector at a future 100 TeV hadron collider might look like
At precisely 09:30 on 10 September 2008, almost 15 years after the project’s approval, the first beam was injected into the LHC, amid global media attention. In the days that followed good progress was made until disaster struck: during a ramp to full energy, one of the 10,000 superconducting joints between the magnets failed, causing extensive damage which took more than a year to recover from. Following repairs and consolidation, on 29 November 2009 beam was once more circulating and full commissioning and operation could start. Rapid progress in ramping up the luminosity followed, and the LHC physics programme, at an initial energy of 3.5 TeV per beam, began in earnest in March 2010.
LHC experiments
Yet a whole other level of sophistication was realised by the LHC detectors compared to those at previous colliders. The priority benchmark for the designs of the general-purpose detectors ATLAS and CMS was to unambiguously discover (or rule out) the Standard Model Higgs boson for all possible masses up to 1 TeV, which demanded the ability to measure a variety of final states. The challenges for the Higgs search also guaranteed the detectors’ potential for all kinds of searches for physics beyond the Standard Model, which was the other driving physics motivation at the energy frontier. These two very ambitious LHC detector designs integrated all the lessons learned from the experiments at the three predecessor machines, as well as further technology advances in other large experiments, most notably at HERA and LEP.
Just a few simple numbers illustrate the giant leap from the Tevatron to the LHC detectors. CDF and DØ, in their upgraded versions operating at a luminosity of up to 4 × 1032 cm–2s–1, typically had around a million channels and a triggered event rate of 100 Hz, with event sizes of 500 kB. The collaborations were each about 600 strong. By contrast, ATLAS and CMS operated during LHC Run 2 at a luminosity of 2 × 1034 cm–2s–1 with typically 100 million readout channels, and an event rate and size of 500 Hz and 1500 kB. Their publications have close to 3000 authors.
For many major LHC-detector components, complementary technologies were selected. This is most visible for the superconducting magnet systems, with an elegant and unique large 4 T solenoid in CMS serving both the muon and inner tracking measurements, and an air-core toroid system for the muon spectrometer in ATLAS together with a 2 T solenoid around the inner tracking cylinder. These choices drove the layout of the active detector components, for instance the electromagnetic calorimetry. Here again, different technologies were implemented: a novel-configuration liquid-argon sampling calorimeter for ATLAS and lead-tungstate crystals for CMS.
From the outset, the LHC was conceived as a highly versatile collider facility, not only for the exploration of high transverse-momentum physics. With its huge production of b and c quarks, it offered the possibility of a very fruitful programme in flavour physics, exploited with great success by the purposely designed LHCb experiment. Furthermore, in special runs the LHC provides heavy-ion collisions for studies of the quark–gluon plasma – the field of action for the ALICE experiment.
As the general-purpose experiments learned from the history of experiments in their field, the concepts of both LHCb and ALICE also evolved from a previous generation of experiments in their fields, which would be interesting to trace back. One remark is due: the designs of all four main detectors at the LHC have turned out to be so flexible that there are no strict boundaries between these three physics fields for them. All of them have learned to use features of their instruments to contribute at least in part to the full physics spectrum offered by the LHC, of which the highlight so far was the July 2012 announcement of the discovery of the Higgs boson by the ATLAS and CMS collaborations. The following year the collaborations were named in the citation for the 2013 Nobel Prize in Physics awarded to François Englert and Peter Higgs.
Since then, the LHC has exceeded its design luminosity by a factor of two and delivered an integrated luminosity of almost 200 fb–1 in proton–proton collisions, while its beam energy was increased to 6.5 TeV in 2015. The machine has also delivered heavy ion (lead–lead) and even lead–proton collisions. But the LHC still has a long way to go before its estimated end of operations in the mid-to-late 2030s. To this end, the machine was shut down in November 2018 for a major upgrade of the whole of the CERN injector complex as well as the detectors to prepare for operation at high luminosities, ultimately up to a “levelled” luminosity of 7 × 1034 cm–2s–1. The High Luminosity LHC (HL-LHC) upgrade is pushing the boundaries of superconducting magnet technology to the limit, particularly around the experiments where the present focusing elements will be replaced by new magnets built from high-performance Nb3Sn superconductor. The eventual objective is to accumulate 3000 fb–1 of integrated luminosity.
In parallel, the LHC-experiment collaborations are preparing and implementing major upgrades to their detectors using novel state-of-art technologies and revolutionary approaches to data collection to exploit the tenfold data volume promised by the HL-LHC. Hadron-collider detector concepts have come a long way in sophistication over the past 50 years. However, behind the scenes are other factors paramount to their success. These include an equally spectacular evolution in data-flow architectures, software and the computing approaches, and analysis methods – all of which have been driven into new territories by the extraordinary needs for dealing with rare events within the huge backgrounds of ordinary collisions at hadron colliders. Worthy of particular mention in the success of all LHC physics results is the Worldwide LHC Computing Grid. This journey is now poised to continue, as we look ahead towards how a general-purpose detector at a future 100 TeV hadron collider might look like.
Beyond the LHC
Although the LHC has at least 15 years of operations ahead of it, the question now arises, as it did in 1964: what is the next step for the field? The CERN Council has recently approved the recommendations of the 2020 update of the European strategy for particle physics, which includes, among other things, a thorough study of a very high-energy hadron collider to succeed the LHC. A technical and financial feasibility study for a 100 km circular collider at CERN with a collision energy of at least 100 TeV is now under way. While a decision to proceed with such a facility is to come later this decade, one thing is certain: lessons learned from 50 years of experience with hadron colliders and their detectors will be crucial to the success of our next step into the unknown.
Looking back on the great discoveries in particle physics, one can see two classes. The discovery of the Ω– in 1964 and of the top quark in 1995 were the final pieces of a puzzle – they completed an existing mathematical structure. In contrast, the discovery of CP violation in 1964 and of the J/ψ in 1974 opened up new vistas on the microscopic world. Paradoxically, although the Higgs boson was slated for discovery for almost half a century following the papers of Brout, Englert, Higgs, Weinberg and others, its discovery belongs in the second class. It constitutes a novel departure in the same way as the J/ψ and the discovery of CP violation, rather than the completion of a paradigm as represented by the discoveries of the Ω– and the top quark.
The novelty of the Higgs boson derives largely from its apparently scalar nature. It is the only fundamental particle without spin. Additionally, it is the only fundamental particle with a self-coupling (gluons also couple to other gluons, but only to those with different colour combinations). Measurements of the couplings of the Higgs boson to the W and Z bosons at the LHC have confirmed its role in the generation of their masses, likewise for the charged third-generation fermions. Despite this great success, the Higgs boson is connected to many of the most troublesome aspects of the Standard Model (see “Connecting the Higgs to Standard Model enigmas” panel). It is for this reason that the recently concluded update of the European strategy for particle physics advocated an electron–positron Higgs factory as the highest priority collider after the LHC, to allow detailed study of this novel and unique particle.
Circular vs linear
The discovery of the Higgs boson at the relatively light mass of 125 GeV, announced by the ATLAS and CMS collaborations in 2012, had two important consequences for experiment. The first was the large number of potentially observable branching fractions available. The second was that circular, as well as linear, e+e– machines could serve as Higgs factories. The two basic mechanisms for Higgs-boson production at such colliders are associated production, e+e–→ ZH, and vector-boson fusion. The former process is dominant at the low-energy first stage of the various Higgs factories under consideration, with vector-boson fusion becoming more important with increasing energy (see “Channeling the Higgs” figure). About a quarter of a million Higgs bosons would be produced per inverse attobarn of data, leading to substantial numbers of recorded events even after the branching ratios to observable modes are taken into account.
Four Higgs-factory designs are presently being considered. Two are linear accelerators, namely the International Linear Collider (ILC) under consideration in Japan and the Compact Linear Collider (CLIC) at CERN, while the other two are circular: the Future Circular Collider (FCC-ee) at CERN and the Circular Electron Positron Collider (CEPC) in China.
The beams in circular colliders continuously lose energy due to synchrotron radiation, causing the luminosity at circular colliders to decrease with beam energy roughly as Eb–3.5. The advantage of circular colliders is their high instantaneous luminosity, in particular at the centre-of-mass energy relevant for the Higgs-physics programme (250 GeV), but even more so at lower energies such as those corresponding to the Z-boson mass (91 GeV). Electron and positron beams in a circular machine naturally achieve transverse polarisation, which can be exploited to make precise measurements of the beam energy via the electron and positron spin-precession frequencies.
In contrast, for linear colliders the luminosity increases roughly linearly with the beam energy. The advantages of linear accelerators are that they can be extended to higher energies, and the beams can be polarised longitudinally. The ZH associated cross section can be increased by 40% with longitudinal polarisations of –80% and 30% for electrons and positrons, respectively. This increase, coupled with the ability to isolate certain components of Higgs-boson production by tuning the polarisation, enables a linear machine to achieve similar precisions on Higgs-boson measurements with half the integrated luminosity of a circular machine.
FCC-ee, CEPC and ILC are foreseen to run for several years at a centre-of-mass energy of around 250 GeV, where the ZH production cross section is largest. Instead, CLIC plans to run its first stage at 380 GeV where both WW fusion and ZH production contribute, and tt production is possible. The circular colliders FCC-ee and CEPC envisage running at the Z-pole and the WW production threshold for long enough to collect of the order 1012 Z bosons and 108 WW pairs, enabling powerful electroweak and flavour-physics programmes (see “Compare and contrast” table). To achieve design luminosity, all proposed e+e– colliders need beams focused to a very small size in one direction (30–70 nm for FCC-ee, 3–8 nm for ILC and 1–3 nm for CLIC), which are all below the values so far achieved at existing facilities.
Evolving designs
The proposed circular colliders are based on a combination of concepts that have been proven and used in previous and present colliders (LEP, SLC, PEP-II, KEKB, SuperKEKB, DAFNE). In Higgs-production mode the beam lifetime is limited by Bhabha scattering to about 30 minutes and therefore requires quasi-continuous injection or “top-up” as used by the B-factories. Each of the circular collider main concepts and parameters has been demonstrated in a previous machine, and thus the designs are considered mature. The total FCC-ee construction cost is estimated to be 10.5 billion CHF for energies up to 240 GeV, with an additional 1.1 billion CHF to go to the tt threshold. This includes 5.4 billion CHF for the tunnel, which could be reused later for a hadron collider. The CEPC cost has been estimated at $5 billion, including $1.3 billion for the tunnel. With the present design, the FCC-ee power consumption is 260–340 MW for the various energy stages (compared to 150 MW for the LHC).
The ILC was proposed in the late 1990s and a technical design report published in 2012. It uses superconducting RF cavities for the acceleration, as used in the currently operating European XFEL facility in Germany, to aim for gradients of 35 MV/m. The cost of the first energy stage (250 GeV) was estimated as $4.8–5.3 billion, with a power consumption of 130–200 MW, and an expression of interest to host the ILC as a global project is being considered in Japan. The CLIC accelerator uses a second beam, termed a drive-beam, to accelerate the primary beam, aiming for gradients in excess of 100 MV/m. This concept has been demonstrated with electron beams at the CLIC test facility, CTF3. The cost of the first energy stage of CLIC is estimated as 5.9 billion CHF with a power consumption of 170 MW, rising to 590 MW for final-stage operation at 3 TeV.
Another important difference between the proposed linear and circular colliders concerns the number of detectors they can host. Collisions at linear machines only occur at one interaction point, while in circular colliders at least two interaction points are proposed, doubling the luminosity available for analyses. Two detectors also offer the dual benefits of scientific competition and the cross-checking of results. At the ILC two detectors are proposed but they cannot run concurrently since they use the same interaction point.
FCC-ee and CLIC have both been proposed as CERN-hosted international projects, similar to the LHC or high-luminosity LHC (HL-LHC). At present, as recommended by the 2020 update of the European strategy for particle physics, a feasibility study for the FCC (including its post-FCC-ee hadron-collider stage, FCC-hh) is ongoing, with the goal of presenting an updated conceptual design report by the next strategy update in 2026. Among the e+e– colliders, CLIC has the greatest capacity to be extended to the multi-TeV energy range. In its low-energy incarnation it could be realised either with the drive-beam or conventional technology. CEPC is conceptually and technologically similar to FCC-ee and has also presented a conceptual design report. Nearly all statements about FCC-ee also hold for CEPC except that CEPC’s design luminosity is about a factor of two lower, and thus it takes longer to acquire the same integrated luminosity. At circular colliders, the multi-TeV regime (at least 100 TeV in the case of FCC-hh) would be reached by using proton beams, similar to what was done with LHC following LEP.
In addition to the vacuum expectation value of the Higgs field and the mass of the Higgs boson, the discovery of the Higgs boson introduces a large number of parameters into the Standard Model. Among them are the Yukawa couplings of the nine charged fermions (in contrast, the gauge sector of the SM has only three free parameters). The Yukawa forces, of which only three have been discovered corresponding to the couplings to the charged third-generation fermions, are completely new. They are of disparate strengths and, unlike the other forces, are not subject to the constraint of local gauge invariance. They provide a parameterisation of the theory of flavour, rather than an explanation. It is of primary importance to discover, bound and characterise the Yukawa forces. In particular, the discovery of CP violation in the Yukawa couplings would go beyond the confines of the Standard Model.
Famously, because of its scalar nature, the quantum corrections to the Higgs boson mass are only bounded by the cut-off on the theory, demanding large renormalisations to maintain the mass at 125 GeV as measured. This issue is not so much a problem for the Standard Model per se. However, in the context of a more complete theory that aims to supersede and encompass the Standard Model, it becomes much more troubling. In effect, the degree of cancellation necessary to maintain the Higgs mass at 125 GeV effectively sabotages the predictive power of any more complete theory. This sabotage becomes deadly as the scale of the new physics is pushed to higher and higher energies.
The electroweak potential is another area of importance in which our current knowledge is fragmentary. Within the confines of the Standard Model the potential is completely specified by the position of its minimum – the vacuum expectation value and the second derivative of the potential at the minimum, the mass of the Higgs boson (or equivalently its self-coupling). We have no direct knowledge of the behaviour of the potential at larger field values further from the minimum. In addition, extrapolation of the currently understood Higgs potential to higher energy reveals a world teetering between stability and instability. Further information about the behaviour of the potential could help us to interpret the meaning of this result. A modified electroweak potential might also give rise to a first-order phase transition at high temperature, rather than the smooth crossover expected for the Standard Model Higgs potential. This would fulfil one of the three Sakharov conditions necessary to generate an asymmetry between matter and antimatter in our universe.
To quantify the scientific reach of the proposed colliders compared to current knowledge or the expectations for the HL-LHC, it is necessary to define figures-of-merit for the observables that will be measured. For the Higgs boson the focus is on the coupling strengths to the Standard Model bosons and fermions, as well as the couplings to any new particles. The strength with which the Higgs boson couples to the various particles, i, is denoted by κi, defined such that κi = 1 corresponds to the Standard Model. Non-standard phenomena are included in this “kappa” framework by introducing two new quantities: the branching ratio into invisible particles (determined by measuring the missing energy in identified Higgs events), and the branching ratio to untagged particles (determined by measuring the contributions to the total width accounted for by the observed modes, or by directly searching for anomalous decays).
Higgs-boson observables
At hadron colliders, only ratios of κi parameters can be measured, since a precise measurement of the total width of the Higgs boson is lacking (the expected total width of the Higgs boson in the Standard Model is 4.2 MeV, which is far too small to be resolved experimentally). To determine the absolute κi values at a hadron collider a further assumption needs to be made, either on decay rates of the Higgs boson to new particles or on one of the κi values. An assumption that is often made, and valid in many beyond-the-Standard-Model theories, is that κZ≤ 1.
The kappa framework, however, by construction, does not parameterise possible effects coming from different Lorentz structures and/or the energy dependence of the Higgs couplings. Such effects could generically arise from the existence of new physics at higher scales and could lead not only to changes in the predicted rates, but also in distributions. Deviations of κi from 1 indicate a departure from the Standard Model, but do not provide a tool to diagnose its cause. This shortcoming is remedied in so-called effective-operator formalisms by including operators of mass dimension greater than four.
At e+e– colliders a Higgs boson produced via e+e–→ ZH can be identified without observing its decay products. This measurement, of primary importance, is unique to e+e– colliders. By measuring the Z decay products and with the precise knowledge of the momenta of the incoming e– and e+ beams, the presence of the Higgs boson in ZH events can be inferred based on energy and momentum conservation alone, without actually tagging the Higgs boson. In this way one directly measures the coupling between the Higgs and Z bosons. In combination with the Higgs branching ratio to Z pairs it can be interpreted as a measurement of the Higgs-boson width. The first-stage e+e– Higgs factories all constrain the total width at about the 2% level.
LHC and HL-LHC
To assess the potential impact of the e+e– Higgs factories it is important to examine the point of departure provided by the LHC and HL-LHC. Since its startup in 2010 the LHC has made a monumental impact on our understanding of the Higgs sector. After the Higgs discovery in 2012, a measurement programme started and now, with nearly 150 fb–1 of data analysed by ATLAS and CMS, much has been learned. The Higgs-boson mass has been measured with a precision of < 0.2%, its spin and parity confirmed as expected in the Standard Model, and its coupling to bosons and to third-generation charged fermions established with a precision of 5–10%.
With the HL-LHC and its experiments planned to operate from 2027, the precision on the coupling parameters and the branching ratios to new particles will be increased by a factor of 5–10 in all cases, typically resulting in a sensitivity of a few % (see “Kappa couplings” figure). The HL-LHC will also enable measurements of the very rare μ+μ– decay, the first evidence for which was recently reported by CMS and ATLAS, and thus show whether the Higgs boson also generates the mass of a second-generation fermion. With the full HL-LHC dataset, corresponding to 3000 fb–1 for each of ATLAS and CMS, it is expected that di-Higgs production will be established with a significance of four standard deviations. This will allow a determination of the Higgs-boson’s coupling to itself with a precision of 50%.
About a quarter of a million Higgs bosons could be produced per inverse attobarn of data
The LHC has also made enormous progress in the direct searches for new particles at high energies. With more than 1000 papers published on this topic, hunting down particles predicted by dozens of theoretical ideas, and no firm sign of a new particle anywhere, it is clear that the new physics is either heavier, or more weakly coupled or has other features that hides it in the LHC data. The LHC is also a precision machine for electroweak physics, having measured the W-boson mass and the top-quark mass with uncertainties of 0.02% and 0.3%, respectively. In addition, a large number of relevant cross-section measurements of multi-boson production have been made, probing the trilinear and quartic interactions of the gauge bosons with each other.
Higgs-factory impact
In terms of the measurement precision on the Higgs-boson couplings, the proposed Higgs factories are expected to bring a major improvement with respect to HL-LHC in most cases (see “Relative precision” figure). Only for the rare decays to muons, photons and Zγ, and for the very massive top quark, is this not the case. The highest precision (0.2% in the case of FCC-ee) is achieved on κZ since the main Higgs production mode, ZH, depends directly on it, regardless of the decay mode. For other Standard Model particles, improvement factors of two to four are typical. For the invisible and untagged decays, the constraints are improved to around 0.2% and 1%, respectively, for some of the Higgs factories. A new measurement, not possible at the LHC, is that of the charm–quark coupling, κc.
None of the initial stages of the proposed Higgs factories will be able to directly probe the self-coupling of the Higgs boson beyond the 50% expected from the HL-LHC, since the cross-sections for the relevant processes (e+e–→ ZHH and e+e–→ HHνν) are negligible at centre-of-mass energies below 400 GeV. The Higgs self-coupling, however, enters through loops also in single-Higgs production and indirect effects might therefore be observable, for instance as a small (< 1%) deviation in measurements of the inclusive ZH cross section. Measurements of the Higgs self-coupling exploiting the di-Higgs production process can only be performed at higher energy colliders. The ILC and CLIC project uncertainties of around 30% at their intermediate energies and around 10% at their ultimate energies, while FCC-hh projects a precision of around 5%. Similarly, for the Higgs coupling to the top quark, the HL-LHC precision of 3.2% will not be improved by the initial stages of any of the Higgs factories.
The proposed Higgs factories also have a rich physics programme at lower energies, particularly at the Z pole. FCC-ee, for instance, plans to run for four years at the Z pole to accumulate a total of more than 1012 Z bosons – 100,000 times more than at LEP. This will enable a rich and unprecedented electroweak physics programme, constraining so-called oblique parameters (which are sensitive to violations of weak isospin) at the per-mille level, 100 times better than today. It will also enable a B-physics programme, complementary to that at Belle II and LHCb. At CEPC, a similar programme is possible, while at ILC and CLIC the luminosity when running at the Z pole is much lower: the typical number of Z-bosons that can be accumulated here is 109, 100 times more than LEP but not at the same level as the circular colliders. FCC-ee’s electroweak programme also foresees a run at the WW threshold to enable a high-precision measurement of the W mass.
Concerning the large top-quark mass, measurements at the LHC suffer from uncertainties associated with renormalisation schemes and it is unlikely to improve the precision significantly at the HL-LHC beyond the currently achieved value of 400 MeV. At an e+e– collider operating at the tt threshold (~350 GeV), a measurement of the top mass with total uncertainty of around 50 MeV and with full control of the issues associated with the renormalisation scheme is possible. In addition to its importance as a fundamental parameter of the Standard Model, the top mass is the dominant term in the evolution of the Higgs potential with energy to determine vacuum stability (see “Connecting the Higgs to Standard Model enigmas” panel).
To assess the potential impact of the e+e– Higgs factories it is important to examine the point of departure provided by the LHC and HL-LHC
In short, a Higgs factory promises to expand our knowledge of nature at the smallest scales. The ZH cross-section measurement alone will probe fine tuning at a level of a few permille, about 30 times better than what we know today. This provides indirect sensitivity to new particles with masses up to 10–30 TeV, depending on their coupling strength, and could point to a new energy scale in nature.
But most of all the Higgs boson has not exhausted its ability to surprise. The rest of the Standard Model is a compact structure, exquisitely tested, and ruled by local gauge invariance and other symmetries. Compared to this, the Lagrangian of the Higgs sector is the wild west, where the final laws have yet to be written. Does the Higgs boson have a significant rate of invisible decays, which could be a key component in understanding the nature of dark matter in our universe? Does the Higgs boson act as a portal to other scalar degrees of freedom? Does the Higgs boson provide a source of CP violation? An electron–positron Higgs factory provides a tool to address these questions with unique clarity, when deviations between the measured and predicted values of observables are detected. Building on the data from the HL-LHC, it will be the perfect tool to elucidate the underlying laws of physics.
To explore all our coverage marking the 10th anniversary of the discovery of the Higgs boson ...
To provide the best experiences, we use technologies like cookies to store and/or access device information. Consenting to these technologies will allow us to process data such as browsing behavior or unique IDs on this site. Not consenting or withdrawing consent, may adversely affect certain features and functions.
Functional
Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.
