Interview with Oliver Brüning, CERN Director for Accelerators and Technology
June 29, 6.00 am: the LHC will be shut down for a four-year upgrade. What will this consist of? Which limitations of the current accelerator will it allow us to overcome?
It is a complex question, because over these four years we will be upgrading quite a number of elements, and there is a risk of oversimplifying things by merely saying that we are making LHC more performant. We are planning a profound renewal of CERN’s entire accelerator complex, and during Long Shutdown 2 (from 2018 to 2022) we already upgraded the injector complex in preparation for the major Long Shutdown 3 upgrade (2026-2030). The objective is to increase beam intensities in LHC and the luminosity, which is the measure of how many events we can produce per second per unit of space. At present, LHC has two fundamental intensity limitations: the overall intensity the machine can digest before beam instabilities develop, and the power of the radiofrequency system. As regards the first limitation, we are upgrading the collimation system, namely the sections dedicated to beam cleaning. Specifically, we will position absorbers as close as possible to the beam in order to remove stray particles, which generate unwanted electromagnetic fields and cause instabilities – an effect described by the so-called beam impedances. We will therefore replace the current collimators with “low-impedance collimators”, thereby paving the way for an increase in intensity. As for the power limitation of the radiofrequency system, this too will be upgraded and improved, so that it can sustain higher beam intensities once we emerge from Long Shutdown 3. The main hardware upgrades we will carry out, however, concern the sections to the left and right of the ATLAS and CMS detectors, the focusing elements that prepare the beams for collisions and for the detectors, amounting to a total of around 1.5 km of infrastructure. This intervention has two objectives: to produce a greater aperture for the beam, so that it can be focused into smaller points within the detector; and to increase the cryogenic power, which cools the magnets from the heat deposited by debris escaping the detectors. This debris ends up, for the most part, in the focusing elements at the sides of the detector, emitting radiation. There is an equipment lifetime limit due to radiation, because all insulating materials and the epoxy resins we use to keep the coils in position become brittle and lose mechanical stability. And since we are aiming for a tenfold increase in total data production, we are also expecting a tenfold increase in the radiation deposited in these focusing elements. We need to make them more resistant, and the only way to do this is to insert dedicated shielding elements, which in this case will be tungsten absorbers around the beam screens. Another area for improvements is the luminous region, namely the region in which the two beams overlap, which at present, operating with a crossing angle, do not overlap perfectly. With the LHC upgrade we will introduce transversely deflecting radiofrequency elements, the “crab cavities”, which deflect the beam so that the overlap within the detector is maximised, thereby increasing luminosity production. And the final major upgrade concerns the system through which we absorb the energy stored by the machine at the end of each operation, the “beam dump”, an absorber more than ten metres long. With HiLumi we will move from the 500 megajoules per beam of LHC to over 700 megajoules, and we must be able to discharge this safely. The current beam dump will be replaced by a new one that will be compatible with these higher intensities.
Which technological innovations will be introduced with HiLumi LHC? And what have been the most difficult challenges during these years of testing?
The “crown jewel” of the HiLumi project is the development of large-aperture quadruple magnets, that are nevertheless capable of maintaining the same focusing strength as the current machine. Quadrupoles are magnets that function as lenses for proton beams: they compress them and keep them focused before collision. But increasing the aperture – that is, the internal space crossed by the beam – while maintaining the same focusing capability requires much more intense magnetic fields in the superconducting coils. In LHC we reach peak fields of around 9 tesla, and with HiLumi we want to push this to around 12 tesla, with the result that the magnet technology used for LHC will no longer be sufficient. This technology uses a niobium-titanium alloy as a superconductor, a very interesting material because it is flexible, bendable, and therefore well suited for manufacturing. Whereas in order to reach the magnetic fields required by the upgrade we must move to a new superconductor, which we have identified in the niobium-tin (Nb3Sn) alloy. The challenge of this technology is that the material is fragile. It is strictly speaking not a ceramic, but it behaves almost like one. It is very rigid and, once the superconductor has been formed, it can no longer be bent or forced. If subjected to too much mechanical stress, it breaks, and the superconductor is compromised. From an engineering point of view, it is very difficult to produce coils for an accelerator using this material, and this has been one of our main occupations over the past twenty years. Their industrial production – now over 70% complete – represents a fundamental milestone for HiLumi.
Staying on the challenges, what technological challenges will HiLumi pose for the experiments?
With HiLumi we want to increase luminosity, and what really matters for the experiments is the integrated luminosity, namely the amount of data collected over a given period. This data is generally very complex. When a proton collides with a proton, in fact, these are not individual particles, since each proton is composed of quarks and gluons. There is significant uncertainty about what actually collided and at what energy: the detectors must be able to identify both the collision system and the produced particles. To do so unambiguously, the experiments would like to have the lowest possible particle multiplicity, ideally a single event for each beam crossing. At the same time, however, we need as much data as possible, and with LHC we have already increased the number of events considerably, to around 60 events per beam crossing, and with HiLumi we should reach 170-200 events per crossing. One truly obtains a multitude of events, and making sense of the data represents an enormous challenge for the detectors. The principal upgrades will therefore revolve around what we call the central tracker, which will be replaced with new silicon trackers: chips positioned as close as possible to the beam pipe, so as to achieve better granularity and see with greater precision what emerges from the collision point. Naturally, in the initial phases of HiLumi we will work with lower event rates, similar to those of LHC, and then gradually increase them. In order to do so, the machine must run like clockwork, without interruptions or downtime; it must be extremely reliable. To this end we built new underground areas where all active components (the power converters, quench heaters power supplies, the magnet protection systems) will be installed in galleries that are accessible during operation, where they will be well separated from the beam to minimise the failure of electronic components due to radiation emerging from the beam in the accelerator. If something does go wrong, technicians and engineers will be able to intervene while the machine is still running. This will help enormously to increase the availability and efficiency of HiLumi.
Returning to the new materials for the superconducting magnets: did the necessary expertise already exist within the European industry or did you have to develop it together with companies?
Superconducting magnet technology did not exist for the application we were trying to realise. We had to develop it together with industry, in an effort dating back to before 2000, when we began working on the conductor. From 2000 onwards we then launched a programme with our US colleagues in order to actually build magnets using this conductor made out of niobium-tin. And even today, despite the superconducting wire is coming from industry, production of the cables and coils is carried out in the laboratory. We worked together with industry on the production of niobium-tin coils for our 11-tesla dipole magnet programme. This transfer to industry is still ongoing, but now it is mainly related to fusion applications.
How important is CERN today as a platform for innovation for European industry?
It is very important, because we produce technologies that otherwise would not be implemented by industry. Clearly these are products that cannot be sold directly “off the shelf” to other customers, but they allow industries to develop techniques and manufacturing processes which enable them to realise the “unimaginable” in future. Let us take for example the aforementioned “crab cavities”, transversely deflecting radiofrequency systems. Unlike most radiofrequency structures produced industrially through consolidated processes, they are not simple cylindrical or elliptical cavities, but very compact and complex structures, lacking simple symmetry, which must be welded together from many pieces. We worked very closely with European industries in order to develop new manufacturing technologies for superconducting resonators; it was a complex learning process, which will subsequently allow them to produce structures of this type in an industrial context. The other example I would like to mention is the “superconducting links”. I have already referred to the new underground areas where active equipment will be separated from the machine in order to facilitate accessibility; this implies that we will have to cover distances of around 100 metres between the power converters and the magnets, and transport the current powering the magnets – tens of thousands of amperes for each one – across that distance. Altogether, this will amount to hundreds of thousands of amperes, a quantity of current that we could in the past only transport with extremely large water-cooled cables, with heat dissipation systems such as cooling towers. We therefore developed superconducting links: extremely thin links, themselves superconducting, capable of transporting hundreds of thousands of amperes without any resistance over a 100-metre distance. It is a truly extraordinary development, which did not previously exist. And the superconducting material itself, magnesium diboride, did not exist on the market. This conductor cable was developed in collaboration with the Italian company ASG Superconductors and cables were manufacture by ICAS. This technology will certainly find applications in industry, as, for example, in aluminium or steel plants, but also in all those sectors in which it is necessary to transport large electrical currents over long distances without the need for heavy cupper copper cables.
Speaking of applications, which sector do you think is likely to see the most growth over the coming years?
We have spoken about the niobium-tin superconductor, which I would still define as a “classical” superconductor, because it must be cooled to extremely low temperatures in order to become superconducting, between 2 and 4 Kelvin. It requires enormous cryogenic systems and high energy consumption (one need only think that a large part of LHC’s energy consumption is linked precisely to the cryogenic plants that keep the magnets cold). I therefore see the greatest growth potential in high-temperature superconductors: materials capable of becoming superconducting at much higher temperatures. The Holy Grail would be a superconductor operating at 70-80 Kelvin, which would allow the use of liquid nitrogen instead of helium as a refrigerant. But even a superconductor that required not 2 Kelvin but rather 10-15 Kelvin would allow an enormous amount of energy to be saved. I believe this is the great revolution awaiting us, and I expect a drastic development of it over the next 5 or 10 years.
On May 22 the European Strategy for Particle Physics, which defines the future of accelerators in Europe, was officially updated. Why is FCC considered the most promising option, and which scientific frontiers could it allow us to explore?
FCC is the most interesting project because it would allow us to investigate all particle production channels with the highest possible precision, opening a multitude of “windows” through which to search for new physics beyond the Standard Model. In particular, we could study the Higgs boson in great detail, discovered in 2012 at LHC as the particle that generates the mass of all the particles of the Standard Model, and which could help us solve other still-open questions. For example, what are dark energy and dark matter? We do not know, but I believe we should tackle these questions from multiple perspectives: astrophysics, gravitational waves, neutrinos, and even precision measurements of the Standard Model interactions, including the Higgs boson. We have said that it is the Higgs particle that generates mass, and dark matter must be composed of massive particles that therefore could interact with the Higgs particle. FCC could provide us with answers, indications as to where the Standard Model is incomplete and which extensions of it might include dark matter.
How difficult is it today to build a machine designed for the physics of 2040 (FCC-ee) and then 2070 (FCC-hh)?
If you were to ask an engineer or a physicist about the FCC-ee, they would answer that it is not so difficult, because all the technologies already exist. We know how to build FCC-ee. The real challenge lies in obtaining the support of governments and society to commit to such a project, which is very large in both cost and scale. Understanding what the Universe around us is made of, what dark matter is, is truly a fundamental question, and if this project is likely the primary tool for addressing it, I believe there is a very strong motivation, but it is important to succeed in conveying it. The challenge for the FCC-hh is more profound as the existing technology for the superconducting magnets implies very high costs. A breakthrough in Superconducting Magnet Technology to either drastically reduce the cost of the magnet or to significantly boost the reachable magnetic fields are still difficult R&D developments that are currently pursued.
Which area of accelerator R&D would you bet on?
I believe that high-temperature superconductors have the potential to truly revolutionise our sector. It is an area in which we can achieve extraordinary results, also for applications in society and medicine. Before the construction of LHC, niobium-titanium was the superconductor with the greatest potential, and when in the 1990s we began working with industry on developing the alloy, we were practically the only customers. Then, from the mid-2000s onwards, the market changed radically and the principal applications of niobium-titanium technologies became medical ones (magnetic resonance imaging, diagnostics in general…), eventually accounting for the bulk of the market. This is therefore a technology developed thanks to accelerator research, together with industry, and then brought to market by industry itself. Medical applications are truly the principal outlet nowadays. And in the medical field there are many other applications of accelerator technology: for example, hadron therapy, where we use hadrons for radiotherapy because they are more effective in protecting the rest of the body and targeting only the tumour. The major problem with hadron therapy is that the machines are very large, and the investment required from hospitals is enormous. High-temperature superconductors can also help here, having the potential to make everything much more compact, therefore far more widespread and economically accessible.
If you could access to one impossible technology, what breakthrough would you introduce immediately in the field of accelerators?
I repeat myself now, but I would truly like to have a high-temperature magnet capable of exceeding 20 tesla. It would allow us to more than double the magnetic field of LHC, with a technology that consumes far less energy during accelerator operation. It would even open up the possibility of having a higher-energy hadron collider in CERN’s current tunnel, the LHC tunnel, without the need for major civil engineering works. That would be fantastic, I would love that very much.
What discovery do you hope will be achieved with this next generation of accelerators?
With HiLumi I think we will manage to access Higgs self-coupling. Among the many unanswered questions, there is also the question if there exists only one Higgs or several Higgs bosons very close in energy, and how the Higgs interacts with itself. It is a very interesting question, because it is precisely this particle that generates mass, and HiLumi will be the first machine capable of attempting to obtain this type of data. Then, with FCC, we will have further information thanks to ultra-high-precision measurements, which will allow us not only to understand where the Standard Model requires extensions, but also to design the next great revolutionary machine.
BIO
Oliver Brüning took up the role of CERN Director for Accelerators and Technology in January 2026. Following his work at DESY on the stability of the HERA proton beam, he arrived at CERN in 1995, and since 1997 he has contributed to the LHC optics, commissioning and upgrades. From 2008, he served as LHeC Study Leader and held several leadership positions in beam physics. From 2015 to 2019, he led the LHC Full Energy Exploitation Study, and from 2020 to 2025, he headed the High-Luminosity LHC Upgrade project. He has served on several international committees, and was Chair of the DESY Machine Advisory Committee (2017–2020) and of the EPS Accelerator Group (2008–2011).
