Technology Archives – Page 18 of 116

Surface treatment: secrets of success in vacuum science

**Surface chemistry** Building 107 is a state-of-the-art 5000 m² facility that provides a diverse portfolio of chemical and electrochemical surface treatments for vacuum components and subsystems. Safety-critical thinking is hard-wired into the facility’s operations, underpinning the day-to-day storage, handling and large-scale use of chemical products. Credit: CERN

Materials exposed to the high-energy beams in a particle accelerator must fulfil a demanding checklist of mechanical, electrical and vacuum requirements. While the structural function comes from the bulk materials, many other properties are ascribed to a thin surface layer, sometimes just a few tens of nanometres thick. This is typically the case for the desorption caused by electron, photon and ion collisions; Joule-effect heating induced by the electromagnetic field associated with the particle beams; and electron multipacting phenomena (see “Collaboration yields vacuum innovation”). To deliver the required performance, dedicated chemical and electrochemical treatments are needed – and more often than not mandatory – to re-engineer the physical and chemical properties of vacuum component/subsystem surfaces.

The bigger drivers here are the construction and operation of the Large Hadron Collider (LHC) and the High-Luminosity LHC upgrade – projects that, in turn, have driven impressive developments in CERN’s capabilities and infrastructure for surface chemistry and surface modification. The most visible example of this synergy is the new Building 107, a state-of-the-art facility that combines a diverse portfolio of chemical and electrochemical surface treatments with a bulletproof approach to risk management for personnel and the environment. Operationally, that ability to characterise, re-engineer and fine-tune surface properties has scaled dramatically over the last decade, spurred by the recruitment of a world-class team of scientists and engineers, the purchase of advanced chemical processing systems, and the consolidation of our R&D collaborations with specialist research institutes across Europe.

Chemistry in action

Within CERN’s Building 107, an imposing structure located on the corner of Rue Salam and Rue Bloch, the simplest treatment to implement – as well as the most common – is chemical surface cleaning. After machining and handling, any accelerator component will be contaminated by a layer of dirt – mainly organic products, dust and salts. Successful cleaning requires the right choice of materials and production strategy. A typical error in the design of vacuum components, for example, is the presence of surfaces that are hidden (and so difficult to clean) or holes that cannot be rinsed or dried fully. Standard cleaning methods to tackle such issues are based on detergents that, in aqueous solution, will lower the surface tensions and so aid the rinsing of foreign materials like grease and dust.

Successful cleaning requires the right choice of materials and production strategy

The nature of the accelerator materials means there are also secondary effects of cleaning that must be considered at the design phase – e.g. removal of the oxide layer (pickling) for copper and etching for aluminium alloys. To improve the cleaning process, we apply mechanical agitation via circulation of cleaning fluids, oscillation of components and ultrasonic vibration. The last of these creates waves at a frequency higher than 25 kHz. In the expansion phase of the liquid waves, microbubbles of vapour are generated (cavitation), while in the compression phase the bubbles implode to generate pressures of around 1000 bar at the equipment surface – a pressure so high that the material can be eroded (though the higher the frequency, the smaller the gas bubbles and the less aggressive the surface interaction).

Chemical fine-tuning

An alternative cleaning method is based on non-aqueous solvents that act on contamination by dilution. Right now, modified alcohols are the most commonly used solvents at CERN – a result of their low selectivity and minimal toxicity – with the cleaning operation performed in a sealed machine to minimise the environmental impacts of the volatile chemicals. While the range of organic products on which solvents are effective is usually wider than that of detergents, they cannot efficiently remove polar contaminants like salt stains. Another drawback is the risk of contaminants recollecting on the component surface when the liquid does not flow adequately.

Ultimately, the choice of detergent versus solvent relies on the experience of the operator and on guidelines linked to the type of vacuum component and the nature of the contamination. In general, the coating of components destined for ultrahigh-vacuum (UHV) applications will require a preliminary cleaning phase with detergents. Meanwhile, solvents are the optimum choice when there are no stringent cleanliness requirements – e.g. degreasing of filters for cryoplants or during the component assembly phase – and for surfaces that are prone to react with or retain water – e.g. steel laminations for magnets, ceramics and welded bellows. (It is worth noting that trapped water is released in vacuum, compromising the achievement of the required pressure, while wet surfaces are seeds for corrosion in air.)

After rinsing and drying, the components are then ready for installation in the accelerator or for ongoing surface modification. In the case of the latter, the chemical treatments aim to generate a thinner, more compact oxide layer and/or a smoother surface – essential for subsequent plating processes. As such, the components can undergo etching, pickling and passivation (to reduce the chemical reactivity of the surface). Consider the copper components for the LHC’s current-lead support: before brazing (a joining process using a melted filler metal), these components are pickled in hydrochloric acid and passivated in chromic acid. Similarly, the aluminium contacts of busbars (for local high-current power distribution) must be pickled by caustic soda and/or a mixture of nitric and hydrofluoric acid before silver coating. Another instructive example is found in the LHCb’s Vertex Locator (VELO) detector, in which the aluminium RF-box window is thinned down to 150 microns by caustic soda.

Safety always comes first in Building 107

**Absolutely flawless** CERN’s new polishing facility during the treatment of the HL-LHC niobium RFD crab cavity (centre). The dedicated machine can provide rotation while chemically polishing with a mixture of nitric, hydrofluoric and phosphoric acids. Credit: Leonel Ferreira/CERN

Safety-critical thinking is hard-wired into the operational DNA of CERN’s Building 107, underpinning the day-to-day storage, handling and large-scale use of chemical products for surface treatments. That safety-first mantra means the 5000 m² facility is able to confine all hazards inside its walls, such that risks for the surrounding neighbourhood and environment are negligible. Among the key features of Building 107:

• There are retention basins that allow containment of the liquid from all surface-treatment tanks (plus, even in the unlikely case of a fire, there is enough retention capacity for the water pumped by the firefighting teams).

• The retention basins have leak detection sensors, pumping systems, buffer tanks and a special coating that’s able to withstand more than 100 types of chemical for several days in the event of a leak.

• Toxic and corrosive vapours are extracted continuously from the tanks and washed in dedicated scrubbers, while any escaped solvents are adsorbed on active carbon filters.

• A continuous spray of alkaline solution transfers toxic products (liquid phase) for decontamination at CERN’s wastewater treatment plant.

• In terms of fire prevention, all plastics used for the treatment tanks and extraction ducts are made of self-extinguishing polypropylene – removing the source of energy to sustain the flames.

• The safety of technicians is ensured by strict operating procedures (including regulated building access), enhanced air extraction and the storage of incompatible products in separate retention zones.

• State-of-the-art sensors provide permanent monitoring of critical airborne products and link to local and fire-brigade alarms.

Frequently, chemical or electrochemical polishing are required in addition to cleaning. Polishing removes the damaged subsurface layer generated by lamination and machining – essentially a tangle of voids, excess dislocations and impurities. In this context, it is worth highlighting the surface treatments for RF acceleration cavities. Best practice dictates that materials for such applications – essentially niobium and copper – must undergo chemical and/or electrochemical polishing to remove a surface layer of 150 micron thickness. As such, the final state of the material’s topmost layer is flawless and without residual stress. (Note that while mechanical polishing can achieve lower roughness, it leaves behind underlayer defects and abrasive contaminations that are incompatible with the high-voltage operation of RF cavities.) A related example is the niobium RFD crab cavity for the HL-LHC project. This complex-shaped object is treated by a dedicated machine that can provide rotation while chemically polishing with a mixture of nitric, hydrofluoric and phosphoric acids. In this chemical triple-whammy, the first acid oxidises niobium; the second fluorinates and “solubilises” the oxide; and the last acts as a buffer controlling the reaction rate.

Another intriguing opportunity is the switch from wet to dry chemistry for certain niche applications

The final set of treatments involves plating the component with a functional material. In outline, this process works by immersing the accelerator component (negatively biased) into an electrolytic solution containing the functional metal ions. The electrolytic solution is strongly acid or basic to ensure high electrical conductivity, with deposition occurring via reduction of the metallic ions on the component surface – all of which occurs in dedicated tanks where the solution is heated, agitated and monitored throughout.

At CERN, we have extensive experience in the electroplating of large components and can plate with copper, silver, nickel, gold and rhodium. Copper is by far the most common option and its thickness is frequently of the order of hundreds of microns (while gold and rhodium are rarely thicker than a few microns). Current capacity varies from 7 m-long pipes (around 10 cm diameter) to 3.5 m-long tanks (up to 0.8 m diameter). It is worth noting that these capabilities are also used to support other big-science facilities – including a recent implementation for the Drift Tube Linac tanks of the European Spallation Source (ESS) in Lund, Sweden.

Chemical innovation

Notwithstanding the day-to-day provision of a range of surface treatments, the Building 107 chemistry team is also tasked with driving process innovation. As safety is our priority, the main focus is on the replacement of toxic products with eco- and personnel-friendly chemicals. A key challenge in this regard is to substitute chromic acid and cyanate baths, and ideally limit the current extensive use of hydrofluoric acid – a development track inextricably linked to the commercialisation of new products and close cooperation with our partners in industry.

Custom surface-treatment services — **Bath time** The Building 107 team has extensive experience in the electroplating of large vacuum components and can plate with copper, silver, nickel, gold and rhodium. The laboratory also provides custom surface-treatment services for other big-science facilities, including the electroplating of the above component for the Drift Tube Linac tanks of the European Spallation Source (ESS). Credit: Pierre Maurin/CERN

Elsewhere, the chemistry team has registered impressive progress on several fronts. There’s the electroforming of tiny vacuum chambers for electron accelerators and RF cavities with seamless enclosure of flanges at the extremities. This R&D project is supported by CERN’s knowledge transfer funds and has already been proposed for the prototyping of the vacuum chamber of the Swiss Light Source II. A parallel line of enquiry includes production of self-supported graphite films for electron strippers that increase the positive charge of ions in beams – with the films fabricated either by etching the metallic support or by electrochemical delamination (a technique already proposed for the production of graphene foils).

Another intriguing opportunity is the switch from wet to dry chemistry for certain niche applications. A case in point is the use of oxygen plasmas for surface cleaning – a technique hitherto largely confined to industry but with one notable exception in accelerator science. The beryllium central beam pipes of the four main LHC experiments, for example, were cleaned by oxygen plasma before non-evaporable-getter coating, removing carbon contamination without dislodging atoms of the hazardous metal. Following on from this successful use case, we are presently studying oxygen plasmas for in situ decontamination and cleaning of radioactive components, a priority task for the chemistry team as the HL-LHC era approaches.

The future of surface chemistry at CERN looks bright – and noticeably greener. The Building 107 team, for its part, remains focused on developing chemical surface treatments that are, first and foremost, safer and, in some cases, drier.

Hyperloop: think big, win big

Tom Kammermeier is an industrial physicist in a hurry. Hardly surprising given that the commercial roadmap he’s following points to a multibillion-dollar opportunity for vacuum equipment makers – an opportunity that, in turn, promises to transform ground-based mass-transportation of people and goods over the coming decades using energy-efficient hyperloop technologies.

Put simply: if technology hype translates into commercial reality, today’s proof-of-principle hyperloop test facilities will, ultimately, scale up to enable the transit of passenger and freight capsules from A to B through steel tubes (roughly 4 m in diameter) maintained at partial vacuum (typically less than 1 mbar). The end-game: journeys of several hundred kilometres at speeds in excess of 1000 km/h – Los Angeles to San Francisco, Mumbai to Chennai, Montreal to Toronto are just some of the high-demand routes on the drawing board – with maglev technologies teed up to provide the required propulsion, acceleration and deceleration along the way.

The end-game for hyperloop is journeys of several hundred kilometres at speeds in excess of 1000 km/h

While the journey to commercial hyperloop deployment is only just beginning, a thriving and diverse innovation ecosystem is already hard at work, with heavily financed technology start-ups and dozens of academic groups and established manufacturers coalescing into a nascent hyperloop supply chain. As Leybold’s global application development manager (industrial vacuum), Kammermeier is front-and-centre in the German manufacturer’s efforts to establish itself as the “go-to” vacuum technology partner for the hyperloop development community. Here he talks to CERN Courier about the trade-offs, challenges and near-term benefits of playing the long game on technology strategy.

How does your application development team support technical and commercial outcomes within Leybold?

I coordinate a team of 20 application specialists worldwide who handle what we call third-level product support – essentially any unique or non-standard technical requests that get referred to us by our regional sales and field engineering colleagues. In each case, we’ll work closely with Leybold’s product engineering and R&D teams to come up with solutions, ensuring that any new learning and insights are shared across the organisation through a structured programme of knowledge dissemination – online webinars, tutorial videos and the like. Our remit also includes the investigation and development of new vacuum applications. This work is informed by emerging customer needs in markets where Leybold already has an established presence – for example, surface coatings, semiconductors, solar technology and food and drink – as well as evaluation of longer-range commercial applications like hyperloop transportation.

What’s the back-story to Leybold’s engagement with the hyperloop community?

The hyperloop opportunity was initially championed at Leybold back in 2015 by my colleague Carl Brockmeyer, who at the time was head of new business development (and is now president of Leybold’s scientific vacuum division). While Carl articulated the long-term commercial vision, my team focused on initial simulations and high-level requirements-gathering for the enabling vacuum technologies. At the outset, we worked closely with pioneering development companies such as Hyperloop Transportation Technologies (HTT) in the US and Virgin Hyperloop (US), while subsequent collaborations include TransPod (Canada) and the EuroTube Foundation (Switzerland).

I’m a physicist by training and, from the off, it was evident to me that there are no insurmountable technical barriers to hyperloop transportation. As such, it seems clear that the large-scale deployment of hyperloop systems will ultimately be driven by policy-makers and by commercial factors such as capital/operational costs versus return on investment.

Hyperloop represents a long-term commercial opportunity for Leybold. Are there any near-term upsides?

The calculus is simple: in the absence of volume orders, we invest time and resources in early-stage R&D collaborations with leading hyperloop companies in return for the publicity, benefits of association and the acquisition of technical and commercial domain knowledge. The work is bursty and comes in waves – essentially an R&D programme and reciprocal learning exercise at this stage. More widely, we’re seeing some payback in our established market sectors, where the hyperloop activity has opened doors with new customers who might not know Leybold so well. What we see is that hyperloop is a great topic for our sales teams to talk about – it’s very relatable.

What do these hyperloop collaborations typically involve?

Our approach is project-led, bringing together ad hoc teams of engineering, simulation and application specialists to address a range of customer requirements. Most of our collaborations to date have kicked off with simulation studies – a relatively cheap way to test the water and build a fundamental understanding of hyperloop vacuum systems and their core technologies.

**Going the distance** Leybold aims to establish itself as a key technology partner for the hyperloop development community and has supplied a large-scale vacuum pumping unit for Virgin Hyperloop’s DevLoop test facility (above) in the Nevada desert. Credit: Virgin Hyperloop

It wasn’t long, however, before our systems group began supplying one-off hardware orders, including a large-scale vacuum pumping unit for Virgin Hyperloop’s DevLoop test facility in the Nevada desert. While this is a custom installation, it’s
based on existing commercial pumping units that we sell into steel degassing applications, though with several modifications to the programmable controller.

There’s been lots of hype about hyperloop over the last five years. How do you see the market trajectory right now?

My take is that hyperloop R&D and commercialisation activities are gathering pace, as evidenced by the first successful demonstration of human travel in a hyperloop pod at Virgin Hyperloop’s DevLoop test site back in October. This represents a significant breakthrough after more than 400 previously unoccupied test-runs at DevLoop. Elsewhere, we recently sold another big pumping system into HTT for its work-in-progress test-track near Toulouse, France. We’re frequently in contact with them regarding simulation or engineering considerations, with safety-critical aspects very much to the fore as HTT also plans to transport human passengers in the near future.

What sort of technical challenges is Leybold being asked to address by hyperloop developers?

Pumping down a hyperloop vacuum tube over hundreds of kilometres is a non-trivial engineering challenge. From a vacuum perspective, you need to think carefully about the spacing of your pumping stations along the tube; optimisation of each pumping system; what happens in case of tube failures or accidents; and how the distributed pumping network can provide back-up pumping capacity and compensation (see “Hyperloop: rewriting the rules of large-scale vacuum”).

Hyperloop: rewriting the rules of large-scale vacuum

Custom pumping systems — **Big vacuum** Leybold is working to optimise its custom pumping systems (shown above) for the very long pump-downs (between 12–24 hours) that will be required in large-scale hyperloop tubes. Credit: Leybold GmbH/Virgin Hyperloop

“Pumping down a hyperloop vacuum tube over hundreds of kilometres is a non-trivial engineering challenge,” notes Leybold’s Tom Kammermeier in our accompanying interview. Here he outlines some of the key design and engineering considerations for hyperloop vacuum systems.

Location, location, location

The aspect ratio (diameter/length) of a hyperloop system is enormous – 1/1000,000 is easily within reach – and imposes inescapable design constraints in terms of vacuum pumping capability. A single-site pumping station, while minimising capital outlay, would result in some odd pressure distributions and gradients along the hyperloop track. During pump-down, for example, the operator might register the target base pressure at one end of the pipe while the other end is still at atmospheric pressure. What’s needed instead is an intelligent distribution of pumping capacity along the track – crucial for compensation of any leaks and pump failures, and doubly so in terms of reducing capital/operational expenditure (as every additional pumping site means more outlay in terms of enclosures, power supply, water supply and associated infrastructure).

Smart strategies for leak management

A vacuum system can be defined along a number of coordinates, not least in terms of its pump-down requirements and target operating pressure (where the total pumping speed equals the inleak flow rate). The higher the permissible operating pressure, the lower the pumping speed, and the greater the aggregate energy savings over time. A large-scale hyperloop system will therefore require a smart pumping network to optimise the distribution of pumping speed dynamically versus local inleak flow conditions – a capability that, in turn, will yield significant (and recurring) operational savings. It’s also worth noting that an understanding of the pumping-speed distribution (essentially a granular map of pressure along the tube) will enable efficient leak detection without recourse to a conventional and time-consuming leak search.

Gearing up, pumping down

Peak energy consumption for any hyperloop vacuum system will occur during end-to-end pump-down along the track. With this in mind, Leybold is working to optimise its multistage Roots pumping systems for the very long pump-downs (of the order of 12–24 hours) that will be required in large-scale hyperloop tubes. Roots pumps are an excellent option for high-volume flows at low pressures – i.e. the usual operating regime of hyperloop systems – but their efficient use for an extended pump-down from atmospheric pressure is problematic. Issues can include overheating due to gas compression; overload of the motor; or exceeding temperature limits due to low heat dissipation at low gas pressures. The answer is to employ variable-speed drives, which basically “know” the thermodynamics of each individual pump and enable optimised use. In this way, the programmable logic controller of the pumping system is able to orchestrate the individual pumps to yield the highest possible pumping speed during a pump-down – equating to some millions of m³/h for a 1000 km track.

What lessons have you learned from Leybold’s engagement with the hyperloop community?

A lot of the learning here has been around the simulation of large-scale distributed vacuum systems – because no-one has ever built a vacuum system on the scale necessary to support commercial hyperloop transportation. We’ve had plenty of discussions to date regarding our models and whether they’re still valid over distances of several hundred kilometres, while our technology roadmap focuses on what an optimised pumping system will look like for future “live” hyperloop deployments. To date, because the market is still not mature enough, we’ve created smart hyperloop pumping systems by adapting our existing product lines – specifically, units that we’ve developed for steel-industry applications.

Is cost a big driver of your hyperloop R&D priorities?

Always. Cost-of-ownership calculations feature prominently in discussions with all our hyperloop customers. We’ve given a lot of input, for example, on required pumping speed versus leak flow rate versus operating pressure. Fundamental studies like this help our partners to evaluate whether it’s worth focusing more of their investments on a leak-tight pipe or on the vacuum pumping systems. Another priority for developers is energy consumption, so our system-level simulations provide vital insights for the accurate calculation of pump-down time and vacuum performance versus energy budget. In this context, it’s worth noting that Leybold’s DRYVAC Energy Saver – which reduces the energy consumption of our dry compressing screw pumps and systems by as much as 50% – is emerging as a potential game-changer for the large-scale pumping systems that will underpin hyperloop installations.

Are vacuum equipment makers ready if hyperloop’s technology push translates into market pull?

If hyperloop transportation really takes off, it will represent a massive growth market for the vacuum industry. Even a mid-size hyperloop project will require significant focus and scale-up from suppliers like Leybold. The biggest challenge will be developing, then bringing to market, a new generation of application-specific pumping systems – at the required scale and the right price-points.

A joined-up vision for vacuum

**Big science** An aerial view over the ESS construction site in Lund, Sweden, taken in September 2020. Credit: Roger Eriksson/ESS

Neutron science 2.0 is evolving from concept to reality as construction progresses on the European Spallation Source (ESS), a €1.84 billion accelerator-driven neutron source in Lund, Sweden. ESS will deliver first science in 2023 and will, when in full operation, be the world’s most powerful neutron research facility – between 20 and 100 times brighter than the Institut Laue-Langevin (ILL) in Grenoble, France, and up to five times more powerful than the Spallation Neutron Source (SNS) in Oak Ridge, Tennessee, US.

This industrial-scale endeavour represents an amalgam of the most powerful linear proton accelerator ever built; a two-tonne, rotating tungsten target wheel (which produces neutrons via the spallation process); a reference set of 22 state-of-the-art neutron instruments for user experiments (of which 15 are under construction); and a high-performance data management and software development centre (located in Copenhagen). Here, Marcelo Juni Ferreira, vacuum group leader at ESS, tells CERN Courier how vacuum technologies are equally fundamental to the ESS’s scientific programme.

What does your role as ESS vacuum group leader involve?

I head up a 12-strong multidisciplinary team of engineers, scientists, designers and technicians who manage the international network of stakeholders developing the vacuum infrastructure for the ESS. Many of our partners, for example, make “in-kind” contributions of equipment and personnel rather than direct cash investments from the ESS member countries. As such, the ESS vacuum group is responsible for maintaining the facility’s integrated vacuum design approach across all of these contributions and all of our vacuum systems – the proton accelerator, target section, neutron beamlines and the full suite of neutron instruments that will ultimately support user experiments (see “ESS science, funding and partnership”).

In terms of specifics, what is meant by integrated vacuum design?

The integrated approach to vacuum design works on several levels. Cost reduction is a fundamental driver for ESS. The use of standard industry components where possible reduces maintenance and training requirements, minimises the need for expensive product inventory and, through a single framework agreement covering our in-kind partners and industry suppliers, we can work at scale to lower our overall procurement costs.

**Vacuum vision** Marcelo Juni Ferreira, ESS vacuum group leader, oversees an integrated approach to vacuum system design, with the use of standard industry components where possible to drive down training, maintenance and procurement costs. Credit: Monise Pavani Photo

Another motivation is to help the vacuum group support the diverse vacuum requirements across the neutron instruments. The goal in each case is to ensure sustainable, economical and long-term operation of each instrument’s vacuum plant to minimise downtime and maximise research output. To make this possible, each of the neutron instruments (and associated beamlines) has its own “vacuum interface” document summarising key technical specifications and performance requirements – all ultimately aligned with the ESS Vacuum Handbook, the main reference source promoting the use of common vacuum equipment and standards across all aspects of the project.

So, standardisation is a big part of your vacuum strategy?

Absolutely. It’s all about a unified approach to our vacuum equipment as well as the procurement policy for any major hardware/software purchases for the accelerator, the target and the neutron instruments. Another upside of standardisation is that it simplifies the interfaces between the ESS vacuum infrastructure and the ESS safety and control plant – for example, the personnel protection, machine protection and target safety systems.

ESS recently took delivery of the Target Monolith Vessel (TMV), one of the facility’s main vacuum sections. What is the TMV and who built it?

The TMV represents the core building block of the ESS target station and was assembled by our in-kind partners at ESS Bilbao, Spain, working in collaboration with local manufacturers such as Cadinox and AVS. When ESS goes online in 2023, the TMV will enclose all of the target subsystems – the target wheel, moderator, reflector plugs and cryogenic cooling – in a vacuum atmosphere and, with the help of 6000 tonnes of stainless-steel shielding, also confine any activated materials and ionising radiation in case of a highly unlikely event, such as an earthquake or accident (see “ESS operational highlights”).

The monolith is an impressive and complex piece of precision engineering in its own right. The vessel requires exacting and repeatable alignment tolerances (±25 μm) for the target wheel, the moderator and reflector assemblies relative to the incident proton beam as well as the neutron-beam extraction system. Ahead of shipping, ESS Bilbao successfully completed the leak and vacuum tests on the TMV with satisfactory measurements of dew-point temperature, pressure rise and leak detection. The final pressure obtained was 1 × 10^-6 mbar with a leakage < 1 × 10^–8 mbar.l/s.

In terms of the TMV, how does your team design and build for maximum uptime?

The focus on project risk is a collective effort across all support functions and is framed by the ESS Strategic Installation and Test Strategy. With the TMV, for example, our design choices seek to minimise service interruptions to the scientific experiments at ESS. Put another way: each vacuum component in the TMV must offer the longest “time before failure” available on the market. In the case of the rough vacuum pumps, for example, this comes from Kashiyama Industries of Japan through ESS’s supplier Low2High Vacuum in Sweden – offering a dry vacuum pump that’s capable of 24/7, maintenance-free operation for up to three years. We’ve actually tested six of these units running at the laboratory for more than five years and none of them have required any intervention.

ESS science, funding and partnership

• Large-scale neutron facilities are routinely used by academic and industrial researchers to understand material properties on the atomic scale, spurring advances across a spectrum of scientific discovery – from clean energy and environmental technology to pharma and healthcare, from structural biology and nanotech to food science and cultural heritage.

• ESS is a pan-European project with 13 European nations as members: the Czech Republic, Denmark, Estonia, France, Germany, Hungary, Italy, Norway, Poland, Spain, Sweden, Switzerland and the UK.

• Significant in-kind contributions of equipment and expertise – from more than 40 European partner laboratories – are expected to finance more than a third of the overall construction costs for ESS.

• ESS will deliver its first science in 2023, with up to 3000 visiting researchers expected every year once the lab is fully operational.

Smart choices like this add up and result in less maintenance, reduced manual handling of active materials (e.g. pump oil) and lower cost per unit life-cycle. Similar thinking informs our approach regarding the TMV’s vacuum “plumbing”. The use of aluminium gaskets and clamps, for example, streamlines installation (compared with CF flanges) and takes into account their low neutron activation in the case of maintenance removal and reassembly ahead of resumed operations (with hands-on manipulation being faster and simpler in each case).

What are the biggest operational challenges in terms of preparing the TMV for high-reliability vacuum performance?

The major effort on the vessel was – and still is – to qualify all in-vacuum parts and connections in terms of their leak rates, pressure-code requirements and surface finishing. This includes the water-cooled shielding blocks, hydrogen-cooled moderator/deflector, and the helium cooling unit for the rotating tungsten target wheel (which employs a ferrofluidic sealing system). It’s a huge collective effort in vacuum: there are more than 1000 flanges, around 20,000 bolts and 6000 tonnes of load in the fully configured TMV (which measures 6 m internal diameter and 11 m high).

There will be two possible modes of TMV operation, with the target residing in either high vacuum or helium at slightly below atmospheric pressure. What’s the rationale here?

One of the high-level design objectives for ESS states that the TMV should be built to last for 50 years of operation while satisfying all performance and safety criteria. Our initial simulations showed that “cleanliness” of the volume surrounding the collisions of the proton beam and the tungsten target wheel will be essential for slowing material degradation and therefore delivering against this objective. What’s more, the specification of a 5 MW proton beam means that secondary gamma and neutron radiation will be produced as a side-effect of the spallation process, further emphasising the need for a controlled environment as well as appropriate cooling of the shielding blocks to counter radiation-induced heating effects.

ESS operational highlights

Fundamental principles

ESS and Daresbury vacuum teams and components — **Test and measurement** ESS vacuum components and subsystems (left) are put through their paces in a dedicated vacuum test facility, one of the project’s first in-kind hardware shipments from STFC Daresbury Laboratory, UK, back in 2015. Right: members of the ESS and Daresbury vacuum teams. Credit: Julia Öberg/ESS

• At the heart of the ESS is a linear accelerator that produces up to a 5 MW beam of 2 GeV protons, with the bulk of the acceleration generated by more than 100 superconducting radio-frequency (RF) cavities.

• These accelerated protons strike a rotating tungsten target wheel (2.6 m diameter) to produce a beam of neutrons via nuclear spallation – i.e. the impact on the tungsten nuclei effectively “spalls” off free neutrons.

• The target wheel rotates at 23.3 rpm and is cooled by a flowing helium gas system interfaced with a secondary water system.

• The spalled neutrons pass through water premoderators, a supercritical hydrogen moderator (cooled to about 17 K) and a beryllium-lined reflector – all of which are housed in a replaceable plug – to slow the neutrons to useful energies before distribution to a suite of 15 neutron-science instruments.

• The TMV has an Active Cells Facility to perform remote handling, disassembly and storage of components that are taken out of the monolith after reaching the end of their lifetime; steel shielding blocks prevent the escape of neutron/gamma ionising radiation.

TMV vacuum considerations

• The TMV is designed to accommodate various leak-rate loads, including: outgassing of vacuum components; air leaks into the vacuum vessel; water leaks from internal piping plus humidity and condensation present during operations and pump down; and helium leaks from the target wheel.

• Total gas in-leakage is critical and, in conjunction with the capacity of the turbomolecular pumping system, will determine not only the TMV operating pressure but also the refrigeration capacity for the cryo-condensing coil for pumping of potential water leaks.

• In vacuum mode, TMV pressures < 10^–4 mbar will be required for interfacing with the UHV environment of the proton accelerator (i.e. to keep gas flows into the accelerator section to an acceptable level).

• TMV vacuum components (including polymer seals) must be compatible with operation up to 35 °C in harsh gamma/neutron radiation environments.

Operationally, the optimal mode of operation will be high vacuum (< 10^–4 mbar), which will negate the need for a proton beam window between the proton accelerator and the target. This, in turn, will lower the annual operating costs. Other advantages include up to 1% improved neutronic performance, reduced beam scattering on the TMV components (and therefore less heat load and radiation damage), as well as a cleaner image for the beam imaging diagnostics.

Nevertheless, we will design and build a proton beam window, so that it is ready to install for operation under helium should an unanticipated issue arise with the TMV vacuum. Worth noting that in this “helium mode” a pump-and-purge capability is provided to ensure high helium purity (> 99.9%).

What lessons can other big-science facilities learn from your experiences with the ESS vacuum project?

With ESS we are entering new territory and the reliability of all our components – vacuum and otherwise – requires close collaboration as well as consistent communication on all levels with our equipment vendors and in-kind partners. Operationally, there’s no doubt that the TMV and the other ESS vacuum systems have benefited from our dedicated vacuum laboratory – one of the first in-kind hardware shipments back in 2015 – and our efforts to recruit and build a skilled team of specialists in those early days of the project. The laboratory includes test facilities for vacuum integration, gauge calibration and materials outgassing studies – capabilities that allow us to iterate and optimise field solutions in good time ahead of deployment. All of which ultimately helps us to minimise project risk, with technical decisions informed by real-world testing and not just prior experience.

Collaboration yields vacuum innovation

**Gearing up** 2021 promises to be an intense period for the CERN vacuum team. An immediate concern is the restart of beam circulation in LHC vacuum systems that were open to the air for planned interventions and modification, including the ALICE beampipe (above). Credit: Maximilien Brice, Julien Marius Ordan/CERN

Vacuum represents a core enabling technology in particle accelerators. Without the required degree of vacuum, the rate of interaction between circulating particles and residual gas molecules would generate several adverse conditions. Particle beams would increase in size and so decrease in luminosity at the interaction points. Beam instability and the rate of particle loss would grow, endangering instrumentation and increasing the background noise in physics experiments. Induced radioactivity and bremsstrahlung radiation would increase risks for personnel and cause damage to the accelerator hardware. What’s more, vacuum is crucial for avoiding electrical breakdown in high-voltage equipment, as well as for thermal insulation of cryogenic fluids, reducing heat “inleaks” to acceptable levels.

Operationally, the level of vacuum required for particle accelerators spans a large range of residual gas densities – from high vacuum (HV, 10^–3 to 10^–9 mbar) through ultrahigh vacuum (UHV, 10^–9 to 10^–12 mbar) to extreme high vacuum (XHV, usually defined as 10^–12 mbar and lower). Applications in thermal insulation, for example, require a gas-molecule density 10 million times lower than sea-level atmospheric pressure – i.e. less than 10^–4 mbar. On the other hand, a modern synchrotron facility requires UHV residual gas densities of ≤ 10^–9 mbar, while some antimatter experiments impose a rarefaction requirement in the region of 10^–15 mbar. In the most challenging experiments, vacuum is an enclosed space where only several gas molecules per cm³ persist in their random motion, bouncing from one wall of the vacuum vessel to another and able to travel thousands of millions of km before striking another peer (roughly equivalent to the distance from the Sun to Jupiter).

Writ large, it is no surprise that, with more than 125 km of beampipes and liquid-helium transfer lines, CERN is home to one of the world’s largest vacuum systems – and certainly the longest and most sophisticated in terms of particle accelerators. From HV to the UHV/XHV regimes, the complexity of vacuum systems for the particle accelerators at CERN, and other big-science laboratories like it, stems largely from the interaction between particle beams and the surfaces that surround them.

Beam interactions

This “beam–surface dialogue” induces gas desorption from the vacuum system walls, an interaction that can be the dominant source of gas. Indeed, if atmospheric gas is evacuated rapidly from the vacuum system, with no in-leakage of air, it is possible to attain UHV conditions in just a few hours for chamber volumes of the order of a cubic metre. Although the vacuum-system walls release gases spontaneously – mainly water vapour and hydrogen – the choice of suitable materials and thermal treatments reduces the outgassing rates to an acceptable level before accelerator operation. As such, beam-induced gas desorption remains the biggest headache – and this effect, of course, arises only when the particle beams are in circulation.

Beam losses on the chamber walls can be a direct source of gas in the accelerator vacuum system. For the most part, however, beam-induced gas desorption occurs indirectly via the emission of synchrotron light and the beam-induced acceleration of electrons and ions created, for example, by residual gas ionisation. The synchrotron-light-induced desorption is mediated by surface–electron quantum transitions leading to the extraction of photoelectrons, which can desorb residual gas molecules in two ways – initially when leaving the chamber wall, also when striking the wall subsequently. This effect is by far the main source of gas in circular high-energy electron accelerators and plays a significant role in the Large Hadron Collider (LHC), where the critical energy of the emitted photons is around 40 eV (i.e. large enough to extract photoelectrons and induce desorption).

**Vacuum diversity** CERN is home to world-leading R&D capabilities in vacuum science and technology. Top left: prototyping a new set-up for thin-film coating of the HL-LHC vacuum chambers. Top right: the CMS beampipe traversing the inner tracker detector. The central part is made of 0.8 mm-thick pure beryllium (a metal highly transparent to elementary particles), while the rest of the vacuum vessel is made largely of aluminium alloy. Bottom left: a system for the study of new cathodes for sputter deposition of amorphous carbon thin films. Centre: thin-film coating of the RF shield of the LHC’s new target dump injection-segmented unit (a machine-protecting device). Right: an in-vacuum cable spool for electrical powering of a movable sputtering target designed for *in situ* amorphous carbon thin-film coating in the beam screen of the LHC’s standalone magnets. Credits: Pedro Costa Pinto/CERN; Micheal Hoch/CERN

It’s worth noting, though, that there’s no “instant fix” for excessive gas desorption. Even with appropriate chemical surface treatments, accelerator vacuum systems (particularly those for electrons) cannot cope with full beam performance on day one of commissioning. Instead, it is necessary to ramp up the performance of the vacuum system while the beam current is increased in a stepwise fashion. In this way, the dose of particles hitting the surfaces of the vacuum vessel increases (though without excessive beam losses), while desorption yields are reduced via surface cleaning and chemical modification. In the jargon, this optimisation of surface conditioning is known as a “scrubbing run”.

The time taken for surface conditioning can be cut dramatically with the help of nonevaporable getter (NEG) coatings, a concept developed at CERN in the late 1990s. Put simply: the beampipe walls are coated with a micrometre-thick film of Ti–Zr–V alloy that, once heated for a few hours in the accelerator at about 200 °C, provides a clean metal surface that also acts as a pump (i.e. gas molecules are adsorbed by chemical reaction at the surface). During heating, the main reservoir of gas is eliminated as the oxide passivation layer dissolves into the film; after which the cycle repeats whenever adsorption of gas molecules saturates the surface or air venting is necessary.

This NEG capability is deployed at scale by CERN. The 6 km-long beam lines of the LHC’s room-temperature straight sections, for example, are coated entirely with NEG materials, while uptake in several synchrotron research facilities is now envisaged after a pioneering implementation in MAX IV, the Swedish synchrotron. In summary: NEG coatings combine distributed, high-speed pumping with negligible space requirements – a win–win for small-diameter beampipes in the current generation of electron accelerators.

Another significant component of the beam–surface dialogue within particle accelerators is the heating of materials exposed to the circulating beams. One of two possible tracks for the transfer of thermal power is the interaction between the electromagnetic field generated by the beams with the surrounding materials, a process that induces electrical currents on the beam-facing surfaces.

Support for projects like the HL-LHC requires full cognisance of some pretty harsh operating environments

These currents may in turn give rise to Joule heating, typically mitigated by using a good electrical conductor (like copper) as the material of choice for the beampipes or as a layer deposited on stainless steel, usually via electrolytic techniques. Geometrical discontinuities of the vacuum chambers may also result in resonant interaction with the beam, creating enhanced local power dissipation in trapped modes – a problem that can be solved through optimised design of the vacuum chambers and their transitions.

Taken together, these mitigation measures have another highly beneficial side-effect. Beam-induced surface currents generate electromagnetic fields which, in turn, interact back with the beam, potentially disrupting its characteristics or its long-term stability in the accelerator. As such, the overall drive to reduce the impedance of the vacuum system (and of all in-vacuum components) results in longer beam lifetimes and preserved beam emittance, ultimately leading to higher collision rates in physics experiments.

The heat is still on

Ongoing innovation will be essential, however. In the next generation of high-energy proton accelerators operating with superfluid helium – the proposed Future Circular Collider (FCC-hh) is a case in point – the impedance of the beampipes could prove detrimental for the global heat-load balance of the cryogenic system. To counter this heat source, CERN has initiated an ambitious feasibility study in which the inner walls of vacuum chambers are coated with high-temperature superconductors (HTS). Owing to the much-reduced electrical losses of superconductors versus normal metals, successful use of HTS promises to yield a considerable impedance reduction. It’s early days, but initial results with HTS rare-earth barium copper oxide (ReBCO) test coatings are extremely encouraging.

At the same time, synchrotron radiation and electrons hitting the walls of the vacuum system also convey part of the beam power to the surrounding vessels. The multiplication of impinging electrons by the surface and their acceleration by the beam – a process known as electron multipacting – is of concern for cryogenic systems. In the LHC, for example, the heat load is intercepted by an intermediate wall that’s maintained at a temperature of 10–20 K rather than 1.9 K (which is the temperature of the cold bore – i.e. the chamber in tight contact with the magnet). Underpinning this arrangement is the insertion into the cold bore of an additional pipe – the so-called beam screen – which is made of copper-colaminated stainless-steel and cooled by a dedicated helium circuit. The beam screen and cold bore in turn communicate through pumping slots so that gas molecules are cryoadsorbed on the coldest surface.

CERN’s vacuum roadmap: collaboration is key

**Robotic access** The VAX vacuum module is installed in an area of limited access and relatively high radioactivity within the HL-LHC. The set-up (shown above in the prototyping phase) is fully compatible with robot intervention, enabling leak detection, gasket change and complete removal of parts to be carried out remotely and safely. Credit: Lukasz Piotr Krzempek/CERN

The evolution of vacuum technology and engineering at CERN is strictly aligned versus accelerator operation and priorities; the organisation’s fundamental science programme; and, at a high level, the 2020 update of the European Strategy for Particle Physics. As the restart of the LHC physics programme approaches (slated for early 2022), the reliability of the CERN vacuum system is our primary focus – especially after a shutdown that will have run to more than two years.

For sure, 2021 will be an intense period for the CERN vacuum team. An immediate concern is the restart of beam circulation in vacuum systems that were open to the air for planned interventions and modification – sometimes for several days or weeks. The heat load generated by the beams in the LHC’s arcs will be under the spotlight as well as the performance of the upgraded LHC’s injector chain. There is no doubt that our nights will be filled with worries – worries that will hopefully dissipate as new science breakthroughs are announced for the LHC’s beams and detectors.

Maintaining momentum

In parallel, we will maintain the pace of the HL-LHC programme, implementing vacuum innovations elaborated in the past five years. Chief among them are the new beam screens for the triplet magnets of the two high-luminosity experiments – CMS and ATLAS. This advanced concept integrates a carbon coating (as electron multipacting suppressor) and tungsten blocks (to absorb collision debris before it interacts with the magnets). Design optimisation required several iterations and the running of multiphysics programs. The vacuum team subsequently evaluated the mechanical stability of the HL-LHC beam screen during the electromagnetic and thermal transient generated by magnet quench (i.e. a sudden loss of superconducting properties). Experimental investigations of the vacuum performance – via measurement of adsorption isotherms – allowed us to choose 60 K as the operational temperature for the new beam screen.

Another notable HL-LHC achievement is the vacuum module installed between the last focusing magnet of the accelerator and the high-luminosity experiments. Referred to as VAX, this arrangement comprises a compact set of vacuum components, pumps, valves and gauges installed in an area of limited access and relatively high radioactivity. As such, the VAX design is fully compatible with robot intervention, enabling leak detection, gasket change and complete removal of parts to be carried out remotely and safely. The direction of travel is clear: robotic technologies will have a pivotal role to play in the vacuum systems of next-generation, high-intensity particle accelerators.

Joined-up thinking

Operationally, it is already time to prepare CERN and a new generation of vacuum experts for the post-LHC era. Our reference point is the aforementioned European Strategy for Particle Physics, with its initial prioritisation of an electron–positron Higgs factory to be followed, in the long run, by a 100 km-circumference proton–proton collider at “the highest achievable energy”.

These accelerators will push vacuum science and technology to the limit, amplifying the challenges that we have today with the LHC. Yet there’s plenty of encouraging progress to report. An optimised design for the vacuum chambers is already in the works, thanks to advanced simulations of synchrotron radiation and gas molecule distribution performed using CERN-maintained software. Furthermore, the Karlsruhe Research Accelerator (KARA) in Germany reports excellent results in its evaluations of the proton–proton prototype vacuum chamber. The biggest challenge remains cost: engineering solutions adopted at the km scale cannot be implemented for systems 10 to 100 times longer – the vacuum system would be prohibitively expensive.

Herein lies an opportunity – and more specifically a call to arms for vacuum specialists to work collaboratively across their respective disciplines to imagine, and subsequently deliver, the technology innovations that will address the economic challenges of big science in the 21st century. The potential synergies are already evident as the next generation of particle accelerators takes shape alongside new concepts for advanced gravitational-wave telescopes. Diverse physics initiatives with a common interest in driving down the cost of their enabling vacuum systems.

A granular understanding of the fundamental physics certainly helps here. While synchrotron radiation power depends only on the beam parameters, the contribution of electrons to the heat load depends on the surface parameters, above all the secondary electron yield – i.e. the ratio of emitted electrons versus incident electrons. This important characteristic of the surface walls decreases as the dose of impinging electrons accumulates – an additional outcome of beam conditioning. That said, such a decrease takes time and dedicated beam runs, while the mechanism of beam conditioning seems more complex than at first anticipated (as observed during Run 2 of the LHC from 2014–2018). In terms of specifics, the heat load transferred in the beam-screen cooling circuit was found to be higher than expected in four of the LHC’s eight arcs. CERN’s surface experts investigated several surface characteristics to understand this phenomenon and, finally, spotted anomalous behaviour in copper oxide that could lead to a less effective decrease of the secondary electron yield.

The sheer scale of CERN’s vacuum infrastructure represents an engineering challenge in its own right

To circumvent the need for additional beam conditioning, CERN’s vacuum group has developed amorphous carbon coatings with very low secondary-electron yields to effectively prevent electron multipacting. Such thin films are the baseline for the beampipes of the final focusing magnets for the High-Luminosity LHC (HL-LHC) upgrade, presently under way. The carbon coatings have also been implemented in selected areas of the Super Proton Synchrotron (which injects protons into the LHC) to reduce the direct effect of electron clouds on beam performance.

Another countermeasure to electron multipacting involves increasing the roughness of the walls of the vacuum vessel, such that secondary electrons are intercepted by the surfaces before they can be accelerated by the beam. In this instance, the CERN vacuum group is implementing laser treatments developed by two UK research centres – STFC Daresbury Laboratory and the University of Dundee. The laser, which is introduced into the beampipes using a dedicated robot from GE Inspection Robotics, engraves small grooves azimuthally, with a spacing of a few tens of micrometres. Furthermore, the redeposition of ablated material superposes nanometric particles that enhance the electron-capture effect.

Measurement and control

Zoom out from the esoteric complexity of beam–surface interactions and the sheer scale of CERN’s vacuum infrastructure represents an engineering challenge in its own right – not least in terms of vacuum metrology, diagnostics and control. In all, more than 12,000 vacuum instruments – gauges, pumps, valves and associated controllers with almost a million configuration settings – are managed via a flexible database running in the Cloud. Work is well advanced to mine the vast amounts of data generated by this network of vacuum systems – ultimately creating a “data-streaming pipeline” that will integrate the latest analytics software with a new generation of open-source diagnostic and reporting tools.

**Tunnel vision** Preparation of amorphous carbon coatings in the LHC during Long Shutdown 2. These thin films have very low secondary electron yields to prevent electron multipacting. Credit: Pedro Costa Pinto/CERN

Meanwhile, at the operational sharp-end, the measurement of extremely low pressures remains a core competency of the CERN vacuum team. This capability preserves, indeed builds on, the legacy of the Intersecting Storage Rings (ISR), the world’s first hadron collider and a pioneering environment for vacuum technology during the late 1960s and 1970s. The vacuum gauges operating at CERN today in the 10^–7–10^–12 mbar range are copies of the original models adopted for the ISR, while those in use in CERN’s R&D laboratories and in antimatter experiments (for measurement down to 10^–14 mbar) are the result of further developments in the late 1970s.

Studies of vacuum gauges to provide continuous measurement at even lower pressures are also under way at CERN, often in collaboration with Europe’s metrological community. In the framework of the EURAMET-EMPIR programme, for example, CERN vacuum experts have participated in the development and characterisation of a vacuum gauge with an ultrastable sensitivity for the transfer of vacuum standards amongst European research institutes (see “Vacuum metrology: made to measure”).

More broadly, support for projects like the HL-LHC requires full cognisance of some pretty harsh operating environments. Fundamentally, increasing beam currents means that vacuum systems and their electronic control circuits are more and more susceptible to radiation damage. A key determinant of the global cost/performance of a large-scale vacuum system is the deployment of electronics in the accelerator tunnels – with weaknesses in the devices gradually revealed through increasing radiation exposure. With this in mind, and by using radiation sources available on site as well as at other European research institutes, the CERN vacuum team has been busy evaluating the “radiation hardness” of hundreds of critical components and electronic devices.

Looking to the future, it’s evident that major accelerator initiatives such as the HL-LHC and the proposed FCC will maintain CERN’s role as one of the world’s leading R&D centres for vacuum science and technology – a specialist capability that will ultimately support fundamental scientific advances at CERN and beyond.

Vacuum metrology: made to measure

PTB scientists Karl Jousten and Claus Illgen — **Measurement science** ‘Design for manufacturability’ is a priority for the EMPIR 16NRM05 partners, such that any specialist manufacturer will be able to produce standardised, next-generation ionisation gauges at scale. Here, PTB scientists Karl Jousten (right) and Claus Illgen adjust the gauge assembly prior to bake-out and installation into a UHV system. Credit: PTB

Absence, it seems, can sometimes manifest as a ubiquitous presence. High and ultrahigh vacuum – broadly the “nothingness” defined by the pressure range spanning 0.1 Pa (0.001 mbar) through 10^–9 Pa – is a case in point. HV/UHV environments are, after all, indispensable features of all manner of scientific endeavours – from particle accelerators and fusion research to electron microscopy and surface analysis – as well as a fixture of diverse multibillion-dollar industries, including semiconductors, computing, solar cells and optical coatings.

For context, the ionisation vacuum gauge is the only instrument able to make pressure measurements in the HV/UHV regime, exploiting the electron-induced ionisation of gas molecules within the gauge volume to generate a current that’s proportional to pressure (see figure 1 in “Better traceability for big-science vacuum measurements”). Integrated within a residual gas analyser (RGA), for example, these workhorse instruments effectively “police” HV/UHV systems at a granular level – ensuring safe and reliable operation of large-scale research facilities by monitoring vacuum quality (detecting impurities at the sub-ppm level), providing in situ leak detection and checking the integrity of vacuum seals and feed-throughs.

Setting the standard

Notwithstanding the ubiquity of HV/UHV systems, it’s clear that many scientific and industrial users are sure to gain – and significantly so – from an enhanced approach to pressure measurement in this rarefied domain. For their part, HV/UHV end-users, metrology experts and the International Standards Organisation (ISO) all acknowledge the need for improved functionality and greater standardisation across commercial ionisation gauges – in short, enhanced accuracy and reproducibility plus more uniform sensitivity versus a broad spectrum of gas species.

That wish-list, it turns out, is the remit of an ambitious pan-European vacuum metrology initiative – the catchily titled 16NRM05 Ion Gauge – within the European Metrology Programme for Innovation Research (EMPIR), which in turn is overseen by the European Association of National Metrology Institutes (EURAMET). As completion of its three-year R&D effort approaches, it seems the EMPIR 16NRM05 consortium is well on its way to finalising the design parameters for a new ISO standard for ionisation vacuum gauges that will combine improved accuracy (total relative uncertainty of 1%), robustness and long-term stability with known relative gas sensitivity factors.

It’s a design that cannot be found on the market…The results have been very encouraging

Another priority for EMPIR 16NRM05 is “design for manufacturability”, such that any specialist manufacturer will be able to produce standardised, next-generation ionisation gauges at scale. “We work closely with the gauge manufacturers – VACOM of Germany and INFICON of Liechtenstein are consortium members – to make sure that any future standard will result in an instrument that is easy to use and economical to produce,” explains Karl Jousten, project lead and head of section for vacuum metrology at Physikalisch-Technische Bundesanstalt (PTB), Germany’s national measurement institute (NMI) in Berlin.

In fact, this engagement with industry underpins the project’s efforts to unify something of a fragmented supply chain. Put simply: manufacturers currently use a range of electrode materials, operating potentials and, most importantly, geometries to define their respective portfolios of ionisation gauges. “It’s no surprise,” Jousten adds, “that gauges from different vendors vary significantly in terms of their relative sensitivity factors. What’s more, all commercially available gauges lack long-term and transport stability – the instability being about 5% over one year.”

The EMPIR 16NRM05 project partners – five national measurement institutes (including PTB), VACOM and INFICON, along with vacuum experts from CERN and the University of Lisbon – have sought to bring order to this disorder by designing an ionisation gauge that is at once compatible with standardisation while exceeding current performance levels. When the project kicked off in summer 2017, for example, the partners set themselves the goal of improving the relative standard uncertainty due to long-term and transport instability from about 5% to below 1% for nitrogen gas. Another priority involves tightening the spread of sensitivity factors for different gas species (from about 10% to 2–3%) which, in turn, will help to streamline the calibration of relative gas sensitivity factors for individual gauges and multiple gas species.

It’s all about the detail

For starters, the consortium sought to identify and prioritise a set of high-level design parameters to underpin any future ISO-standardised gauge. A literature review of 260 relevant academic papers (from as far back as the 1950s) yielded some quick-wins and technical insights to inform subsequent simulations (using the commercial software packages OPERA and SIMION) of a v1.0 gauge design versus electrode positions, geometry and overall dimensions. Meanwhile, the partners carried out a statistical evaluation of the manufacturing tolerances for the electrode positions as well as a study of candidate electrode materials before settling on a “model gauge design” for further development.

“It’s a design that cannot be found on the market,” explains Jousten. “While somewhat risky, given that we can’t rely on prior experience with existing commercial products, the consortium took the view that the instabilities in current-generation gauges could not be overcome by modifying existing designs.” With a clear steer to rewrite the rulebook, VACOM and INFICON developed the technical drawings and produced 10 prototype gauges to be tested by NMI consortium members – a process that informed a further round of iteration and optimisation.

Better traceability for big-science vacuum measurements

**Fig. 1.** The EMPIR 16NRM05 ionisation vacuum gauge. (a) Electrons (red) leave the cathode (far left), are accelerated by the anode (green) to 200 eV, focused by the ion collector (blue) and collected by the Faraday cup (dark green) after passing a deflector (light blue). Some electrons will ionise gas molecules and those will be collected by the ion collector. The length of the electron path is well characterised and so the sensitivity of the gauge is predictable. (b) Ions (purple) generated by electrons within the anode grid are accelerated to the collector (dark blue) and measured as a current. Ions (light blue) generated outside the anode grid are caught by the deflector. Credit: EMPIR 16NRM05

The ionisation vacuum gauge is fundamental to the day-to-day work of the vacuum engineering teams at big-science laboratories like CERN. There’s commissioning of HV/UHV systems in the laboratory’s particle accelerators and detectors – monitoring of possible contamination or leaks between experimental runs of the LHC; pass/fail acceptance testing of vacuum components and subsystems prior to deployment; and a range of offline R&D activities, including low-temperature HV/UHV studies of advanced engineering materials.

“I see the primary use of the standardised gauge design in the testing of vacuum equipment and advanced materials prior to installation in the CERN accelerators,” explains Berthold Jenninger, a CERN vacuum specialist and the laboratory’s representative in the EMPIR 16NRM05 consortium. “The instrument will also provide an important reference to simplify the calibration of vacuum gauges and RGAs already deployed in our accelerator complex.”

The underlying issue is that commercial ionisation vacuum gauges are subject to significant drifts in their sensitivity during regular operation and handling – changes that are difficult to detect without access to an in-house calibration facility. Such facilities are the exception rather than the norm, however, given their significant overheads and the need for specialist metrology personnel to run them.

Owing to its stability, the EMPIR 16NRM05 gauge design promises to address this shortcoming by serving as a transfer reference for commercial ionisation vacuum gauges. “It will be possible to calibrate commercial vacuum gauges simply by comparing their readings with respect to that reference,” says Jenninger. “In this way, a research lab will get a clearer idea of the uncertainties of their gauges and, in turn, will be able to test and select the products best suited for their applications.”

The measurement of outgassing rate, pumping speed and vapour pressure at cryogenic temperatures will all benefit from the enhanced precision and traceability of the new-look gauge. Similarly, measurements of ionisation cross-section induced by electrons, ions or photons also rely on gas density measurement, so uncertainties in these properties will be reduced.

“Another bonus,” Jenninger notes, “will be enhanced traceability and comparability of vacuum measurements across different big-science facilities.”

“The results have been very encouraging,” explains Jousten. Specifically, the measured sensitivity of the latest model gauge design agrees with simulations, while the electron transmission through the ionisation region is close to 100%. As such, the electron path length is well-defined, and it can be expected that the relative sensitivities will relate exactly to the ionisation probabilities for different gases. For this reason, the fundamentals of the model gauge design are now largely fixed, with the only technical improvements in the works relating to robustness (for transport stability) and better electrical insulation between the gauge electrodes.

“Robustness appears fine, but is still under test at CMI [in the Czech Republic],” says Jousten. “Right now, the exchange of the emitting cathode – particularly its positioning – seems to depend a little too much on the skill of the technician, though this variability should be addressed by future industrial designs.”

Summarising progress as EMPIR 16NRM05 approaches the finishing line, Jousten points out that PTB and the consortium members originally set out to develop an ionisation vacuum gauge with good repeatability, reproducibility and transport robustness, so that relative sensitivity factors are consistent and can be accumulated over time for many gas species. “It seems that we have exceeded our target,” he explains, “since the sensitivity seems to be predictable for any gas for which the ionisation probability by electrons is known.” The variation of sensitivity for nitrogen between gauges appears to be < 5%, so that no calibration is necessary when the user is comfortable with that level of uncertainty. “At present,” Jousten concludes, “it looks like there is no need to calibrate the relative sensitivity factors, which represents enormous progress from the end-user perspective.”

Of course, much remains to be done. Jousten and his colleagues have already submitted a proposal to EURAMET for follow-on funding to develop the full ISO Technical Specification within the framework of ISO Technical Committee 112 (responsible for vacuum technology). In 2021, Covid permitting, the consortium members will then begin the hard graft of dissemination, presenting their new-look gauge design to manufacturers and end-users.

The High–Luminosity Large Hadron Collider Upgrade Project

By clicking the “Watch now” button you will be taken to our third-party webinar provider in order to register your details.

Want to learn more on this subject?

Watch now

This webinar will provide an overview of the High-Luminosity Large Hadron Collider (HL–LHC) upgrade project with highlights of the main challenges and technical innovations.

Presented by Oliver Brüning, the webinar will cover:

• An introduction to the HL–LHC project.
• An overview of the challenges of a high-energy, high-luminosity hadron collider.
• An outline of the performance reach in HE colliders over the next two decades.

Want to learn more on this subject?

Watch now

Dr Oliver Brüning is the project leader for the HL–LHC project, an upgrade project to the LHC that is scheduled to finish its implementation by 2026. Oliver has a background in accelerator design, beam dynamics and machine operation. He started his career in accelerator physics at DESY where he worked on non-linear beam dynamics studies for HERA and was part of the initial commissioning team of the HERA accelerator. He joined CERN in 1995 and became part of the LHC design team just before the formal LHC approval by the CERN council. Up to 2012, he was working on the design and commissioning of the LHC and from 2005 until 2015 he served as head of the CERN accelerator theory group. Since 2008 he has been co-ordinating the LHeC accelerator system studies and was the deputy project leader for the HL–LHC project between 2010–2020.

Russia’s particle-physics powerhouse

Timeline of INR RAS — **A timeline of INR RAS** Clockwise: (1) theoretical physicist Moisey Markov at the foundation of the Meson Factory building, Troitsk 1971; (2) construction of the first stage of the linear accelerator facility; (3) the top view of the linear accelerator facility; (4,8) deployment of the underwater optical module of the Baikal neutrino telescope; (5) the neutrino council at the Baksan Neutrino Observatory with the participation of Alexander Chudakov, Vadim Kuzmin, Bruno Pontecorvo, Petr Spivak and Albert Tavkhelidze; (6) the tunnel in the Andyrchi mountain for the Baksan Neutrino Observatory experiments; and (7) view of the Baksan Neutrino Observatory laboratories. Credit: INR RAS

Founded on 24 December 1970, the Institute for Nuclear Research of the Russian Academy of Sciences (INR RAS) is a large centre for particle physics in Moscow with wide participation in international projects. The INR RAS conducts work on cosmology, neutrino physics, astrophysics, high-energy physics, accelerator physics and technology, neutron research and nuclear medicine. It is most well-known for its unique research facilities that are spread all across Russia, and its large-scale collaborations in neutrino and high-energy physics. This includes experiments such as the Baksan Neutrino Observatory, and collaborations with a number of CERN experiments including CMS, ALICE, LHCb, NA61 and NA64.

The Institute was founded by the Decree of the Presidium of the USSR Academy of Sciences in accordance with the decision of the government. Theoretical physicist Moisey Markov had a crucial role in establishing the Institute and influenced the research that would later be undertaken. His ambition is seen in the decision to base INR RAS on three separate nuclear laboratories of the P.N. Lebedev Institute of Physics of the Academy of Sciences of the USSR. Each laboratory had a leading physicist in charge: the Atomic Nucleus Laboratory headed by Nobel laureate Ilya Frank; the Photonuclear Reactions Laboratory under the direction of Lyubov Lazareva; and a neutrino laboratory headed by Georgy Zatsepin and Alexander Chudakov. The man overseeing it all was the first director of INR RAS, Albert Tavkhelidze, a former researcher at the Joint Institute for Nuclear Research (JINR, Dubna). In 1987 he was replaced as director by Victor Matveev, then in 2014 by Leonid Kravchuk. Since 2020 the director of INR RAS is Maxim Libanov.

It (Troitsk) has the most powerful linear proton accelerator in the Euro-Asian region

From the very beginning, major efforts were focused on the construction and operation of large-scale research facilities. The hub of INR RAS was built 20 km outside of Moscow, in a town called Troitsk. In 1973 an accelerator division was created, with a long-term goal of creating a meson facility that would house a 600 MeV linear accelerator for protons and H- ions. The first beam was eventually accelerated to 20 MeV in 1988 and the facility was fully operational by 1993. Now known as the Moscow Meson Facility, it has the most powerful linear proton accelerator in the Euro-Asian region, providing fundamental and applied research in nuclear and neutron physics, condensed matter, development of technologies for the production of a wide range of radioisotopes, operation of a radiation therapy complex and many other applications.

A town called Neutrino
Over 1000 miles south from the Troitsk laboratory, an underground tunnel in the Caucasus mountains is the base of another INR RAS facility, the Baksan Neutrino Observatory (BNO). The facility was established in 1967 and the Baksan Underground Scintillation Telescope (BUST) started taking data in 1978. A town sensibly called “Neytrino” (Russian for neutrino) was constructed in parallel with the facility, and was where scientists and their families could live 1700 m above sea level next to the observatory. In 1987 BUST was one of the four neutrino detectors to first directly observe neutrinos from supernova SN1987A.

The observatory did not finish there, and the next step was the gallium-germanium neutrino telescope (GGNT), which was home to the Soviet–American Gallium Experiment (SAGE). The experiment contributed heavily towards solving the solar neutrino problem and simultaneously gave rise to a new problem known as the gallium anomaly, which is yet to be explained. SAGE is still well and truly alive, and with a recent upgrade of the GGNT completed in 2019, the team will now hunt for sterile neutrinos.

Modules being installed in Baikal-GVD — **Ice breaker** New clusters of optical modules being installed in the underwater Baikal-GVD Neutrino Telescope in April 2020. Credit: JINR

By 1990 another neutrino detector was under construction, following the original proposal of Markov and Chudakov. In collaboration with JINR, plans for an underwater neutrino telescope located at the world’s largest freshwater lake, Lake Baikal, took shape. Underwater telescopes use glass spheres that house photomultiplier tubes to detect Cherenkov light from the charged particles emerging from neutrino interactions in the lake water. The first detector developed for Lake Baikal was the NT200, which was constructed over five years from 1993–1998 and detected cosmic neutrinos for more than a decade. It has now been replaced with the Gigaton Volume Detector (Baikal-GVD), and plans were concluded in 2015 for the first phase of the telescope to be completed by 2021. Baikal-GVD has an effective volume of 1 km³ and is designed to register and study ultrahigh-energy neutrino fluxes from astrophysical sources.

Left a mark
There is no doubt that INR RAS has left its mark on high-energy physics. While the Institute’s most recognised work will be in neutrino physics, the Moscow Meson Facility has also contributed largely to other areas of the field. An experiment was created for direct measurement of the mass of the electron antineutrino via the beta decay of tritium. The “Troitsk nu-mass” experiment started in 1985 and its limit on the electron antineutrino mass was the world’s best for years. The improvement of this result became possible only in 2019 with the large-scale KATRIN experiment in Germany that was created in participation with INR RAS. In fact, the Troitsk nu-mass experiment was considered as a prototype for KATRIN.

Experimental data have been obtained on nuclear reactions with the participation of protons and neutrons of medium energies along with data on photonuclear reactions, including the study of the spin structure of a proton using an active polarised target. New effects in collisions of relativistic nuclei have been observed and a new scientific direction has been started, “nuclear photonics”. Two effects in astroparticle physics have been named after scientists from INR RAS: the “GZK cut-off”, which is high-energy cut-off in the spectrum of the ultrahigh-energy cosmic rays named after Kenneth Greisen (US), Georgy Zatsepin and Vadim Kuzmin (INR RAS); and the “Mikheyev–Smirnov–Wolfenstein effect” concerning neutrino oscillations in matter, named after Stanislav Mikheyev, Alexei Smirnov (INR RAS) and Lincoln Wolfenstein (US).

Theoretical studies at INR RAS are also widely known, including the development of perturbation theory methods, study of the ground state (vacuum) in gauge theories, methods for studying the dynamics of strong interactions of hadrons outside the framework of perturbation theory, the first ever brane-world models and the development of principles and the search for mechanisms for the formation of the baryon asymmetry of the universe.

There are plans to construct a large centre for nuclear medicine at the base of the linear accelerator centre

Scientists from INR RAS take an active part in the work of a number of large international experiments at CERN, JINR, Germany, Japan, Italy, USA, China, France, Spain and other countries. The Institute also conducts educational activities, having its own graduate school and teaching departments in nearby institutes such as the Moscow Institute for Physics and Technology.

The future of INR RAS is deeply rooted in its new large-scale infrastructures. Baikal-GVD will, along with the IceCube experiment at the South Pole, be able to register neutrinos of astrophysical origin in the hope of establishing their nature. A project has been prepared to modernise the linear proton accelerator in Troitsk using superconducting radio-frequency cavities, while there are also plans to construct a large centre for nuclear medicine at the base of the linear accelerator centre. There is a proposal to build the Baksan Large-Volume Scintillator Detector at BNO containing 10 ktons of ultra-pure liquid scintillator, which would be able to register neutrinos from the carbon–nitrogen–oxygen (CNO) fusion cycle in the Sun with a precision sufficient to discriminate between various solar models.

The past 50 years have seen consistent growth at INR RAS, and with world-leading future projects on the horizon, the Institute has no signs of slowing down.

Final stretch for LHC upgrades

The second long shutdown of the LHC and its injector complex began two years ago, at the start of 2019. Since then, sweeping upgrades in the accelerator complex and key maintenance work have resulted in a rejuvenated accelerator complex with injectors fit for a decade or more of high-brightness beam production. With major detector upgrades proceeding in parallel, physicists are eyeing the final stretch of the road to Run 3 – which promises to deliver to the experiments an integrated luminosity twice that of Run 1 and Run 2 combined in less than three years of operations.

A large number of physicists, engineers and technicians have strived day-in day-out
Rende Steerenberg

The ceremonial key to the Super Proton Synchrotron (SPS) was handed over to SPS operations on 4 December, signalling the successful completion of the LHC Injectors Upgrade (LIU) programme. “The amazing accomplishment of delivering the machine keys with only a small delay is thanks to the hard work, dedication and flexibility of many,” says head of the operations group Rende Steerenberg, who emphasised the thoroughness with which special measures to ensure the safety of personnel during the COVID-19 pandemic were observed. “A large number of physicists, engineers and technicians strived day-in day-out to complete the upgrade and consolidation of the accelerator complex safely and efficiently following the spring lockdown.”

Super synchrotrons

Major changes to the SPS include the dismantling and remounting of its radio-frequency cavities, the installation of new power amplifiers, and the installation of state-of-the-art beam-control and beam-dump systems. First beam from the new Linac 4 was injected into the upgraded Proton Synchrotron Booster (PSB) on 7 December. The PSB will undergo a commissioning period before injecting beam into the Proton Synchrotron (PS) on 1 March. It will then be the turn of the PS to be commissioned, before sending beam to the SPS on 12 April.

Among many changes to the LHC, all 1232 dipole-magnet interconnections were opened and their electrical insulation consolidated, removing the limitation that prevented the LHC from reaching 7 TeV per beam during Run 2. The cryogenics team cooled the first of the LHC’s eight sectors to its 1.9 K operational temperature on 15 November, with five other sectors being cooled in parallel and the full machine set to be cold by spring. After handing over to the electrical quality-assurance team for the final electrical tests, powering tests and a long campaign of quench training will take place to enable the LHC magnets to support fields in excess of those required during Run 2, when the beam energy was 6.5 TeV. Test beams are due to circulate at the end of September 2021, just four months later than planned before the COVID-19 pandemic.

All 1232 dipole-magnet interconnections were opened

Detector work

In parallel to work on CERN’s accelerator infrastructure, experimental physicists are working hard to complete major upgrades to the detectors which anticipate the stringent requirements of triggering and reconstructing events at the upgraded LHC. The refurbishment of trigger electronics for the ATLAS detector’s liquid-argon calorimeter is progressing quickly and the construction of the muon detector’s two new “small wheels” is set to be completed by October 2021. With a complex upgrade of the CMS detector’s muon system now complete, a newly built beam pipe will soon be fitted in the cavern, followed by the refurbished pixel detector with a new inner layer; magnet upgrades and shielding consolidation will then follow. With ALICE’s time-projection chamber now reinstalled, work is underway to install the detector’s new muon forward tracker, and a new 10 GPixel inner-tracking system will be installed in the first quarter of 2021. Meanwhile, the next steps for a significant revamp to the LHCb detector are the mounting of new vertex-locator modules and the first sensitive detector parts of the new ring-imaging Cherenkov detector during the first months of 2021. Following the completion of the upgrade programmes, Run 3 of the LHC will begin in March 2022.

Accelerator infrastructure relating to earlier stages in the lives of LHC protons is already beginning to be recommissioned. Hydrogen ions from a local source have been transferred to the ELENA ring to commission the newly installed transfer lines to CERN’s antimatter experiments. A newly developed source has fed lead ions into Linac 3, which provides ions to the LHC’s physics experiments, while pre-irradiated targets have provided stable isotopes to the ISOLDE nuclear-physics facility. Many experiments at ISOLDE and the PS-SPS complex will be able to start taking data in summer 2021.

No changes have been made to the LHC schedule beyond 2022. Following the completion of Run 3, the third long shutdown will begin at the start of 2025 for the LHC, and in early 2026 for the injector chain, and will end in mid-2027. During this time the installation of the High-Luminosity LHC (HL-LHC) will be completed, adding major high-technology upgrades to CERN’s flagship machine. In concert with the programme of injector upgrades completed in LS2, these will allow the HL-LHC to deliver an order-of-magnitude greater integrated luminosity to the experiments than its predecessor.

Very high-energy electrons for cancer therapy

Dosimetry experiment for VHEE studies — **High-tech** A dosimetry experiment for VHEE studies in collaboration with NPL in CERN’s CLEAR user facility. Credit: R Corsini/CERN

Radiotherapy (RT) is a fundamental component of effective cancer treatment and control. More than 10,000 electron linear accelerators are currently used worldwide to treat patients with RT, most operating in the low beam-energy range of 5–15 MeV. Usually the electrons are directed at high-density targets to generate bremsstrahlung, and it is the resulting photon beams that are used for therapy. While low-energy electrons have been used to treat cancer for more than five decades, their very low penetration depth tends to limit their application to superficial tumours. The use of high-energy electrons (up to 50 MeV) was studied in the 1980s, but not clinically implemented.

More recently, the idea of using very high-energy (50–250 MeV) electron beams for RT has gained interest. For higher energy electrons, the penetration becomes deeper and the transverse penumbra sharper, potentially enabling the treatment of deep-seated tumours. While the longitudinal dose deposition is also distributed over a larger area, this can be controlled by focusing the electron beam.

The production of very high-energy electrons (VHEE) for RT was the subject of the VHEE 2020 International Workshop, organised by CERN and held remotely from 5–7 October. More than 400 scientists, ranging from clinicians to biologists, and from accelerator physicists to dosimetry experts, gathered virtually to evaluate the perspectives of this novel technique.

FLASH effect

VHEE beams offer several benefits. First, small-diameter high-energy beams can be scanned and focused easily, enabling finer resolution for intensity-modulated treatments than is possible for photon beams. Second, electron accelerators are more compact and significantly cheaper than current installations required for proton therapy. Third, VHEE beams can operate at very high dose rates, possibly compatible with the generation of the “FLASH effect”.

FLASH-RT is a paradigm-shifting method for delivering ultra-high doses within an extremely short irradiation time (tenths of a second). The technique has recently been shown to preserve normal tissue in various species and organs while still maintaining anti-tumour efficacy equivalent to conventional RT at the same dose level, in part due to decreased production of toxic reactive oxygen species. The FLASH effect has been shown to take place with electron, photon and more recently proton beams. However, electron beams promise to deliver an intrinsically higher dose compared to protons and photons, especially over large areas as would be needed for large tumours. Most of the preclinical data demonstrating the increased therapeutic index of FLASH are based on a single fraction and hypo-fractionated regimen of RT and 4–6 MeV beams, which do not allow treatments of deep-seated tumours and trigger large lateral penumbra. This problem can be solved by increasing the electron energy to values higher than 50 MeV, where the penetration depth is larger.

Today, after three decades of research into linear colliders, it is possible to build compact high-gradient (~100 MV/m) linacs, making a compact and cost effective VHEE RT accelerator a reality. Furthermore, the use of novel accelerator techniques such as laser-plasma acceleration is also starting to be applied in the VHEE field. These are currently the subject of a wide international study, as was presented at the VHEE workshop.

At the same time pioneering preliminary work on FLASH was being carried out by researchers at Lausanne University Hospital (CHUV) in Switzerland and the Curie Institute in France, high-gradient linac technology advances for VHEE were being made at CERN for the proposed Compact Linear Collider (CLIC). An extensive R&D program on normal-conducting radio-frequency accelerating structures has been carried out to obtain the demanding performances of the CLIC linac: an accelerating gradient of 100 MV/m, low breakdown rate, micron-tolerance alignment and a high RF-to-beam efficiency (around 30%). All this is now being applied in the conceptual designs of new RT facilities, such as one jointly being developed by CHUV and CERN.

In deep Dose profile for various particle beams in water (for beam widths of 0.5 cm).

High-energy challenges

Many challenges, both technological and biological, have to be addressed and overcome for the ultimate goal of using VHEE and VHEE-FLASH as an innovative modality for effective cancer treatment with minimal damage to healthy tissues. All of these were extensively covered and discussed in the different sessions of VHEE 2020.

From the accelerator-technology point of view an important point is to assess the possibility of focusing and transversely scanning the beam, thereby overcoming the disadvantages associated in the past with low-energy-electron- and photon-beam irradiation. In particular, in the case of VHEE–FLASH it has to be ensured that the biological effect is maintained. Stability, reliability and repeatability are other mandatory ingredients for accelerators to be operated in a medical environment.

The major challenge for VHEE–FLASH is the delivery of a very high dose-rate, possibly over a large area, providing a uniform dose distribution throughout the target. Also the parameter window in which the FLASH effect takes place has still to be thoroughly defined, as does its effectiveness as a function of the physical parameters of the electron beam. This, together with a clear understanding of the underlying biological processes, will likely prove essential in order to fully optimise the FLASH RT technique. Of particular importance, as was repeatedly pointed out during the workshop, is the development of reliable online dosimetry for very high dose rates, a regime not adapted to the current standard dosimetry techniques for RT. Ionisation chambers, routinely used in medical linacs, suffer from nonlinear effects at very high dose rates. To obtain reliable measurements, R&D is needed to develop novel ion chambers or explore alternative possibilities such as solid-state detectors or the use of calibrated beam diagnostics.

All this demands a large test activity across different laboratories to experimentally characterise VHEE beams and their ability to produce the FLASH effect, and to provide a testbed for the associated technologies. It is also important to compare the properties of the electron beams depending on the way they are produced (radio-frequency or laser-plasma accelerator technologies).

A number of experimental test facilities are already available to perform these ambitious objectives: the CERN Linear Electron Accelerator for Research (CLEAR), so far rather unique in being able to provide both high-energy (50–250 MeV) and high-charge beams; VELA–CLARA at Daresbury Laboratory; PITZ at DESY and finally ELBE–HZDR using the superconducting radio-frequency technology at Dresden. Further radiobiology studies with laser-plasma accelerated electron beams are currently being performed at the DRACO PetaWatt laser facility at the ELBE Center at HZDR-Dresden and at the Laboratoire d’Optique Appliqué in the Institute Polytechnique de Paris. Future facilities, as exemplified by the previously mentioned CERN–CHUV facility or the PHASER proposal at SLAC, are also on the horizon.

Establishing innovative treatment modalities for cancer is a major 21st century health challenge. By 2040, cancer is predicted to be the leading cause of death, with approximatively 27.5 million newly diagnosed patients and 16.3 million related deaths per year. The October VHEE workshop demonstrated the continuing potential of accelerator physics to drive new RT treatments, and also included a lively session dedicated to industrial partners. The large increase in attendance since the first workshop in 2017 in Daresbury, UK, shows the vitality and increasing interest in this field.

CERN takes next step for hadron therapy

**Ions for therapy** How the SEEIIST Ion Therapy Research Infrastructure (left) in South East Europe might look, where the open roof shows the synchrotron and the treatment rooms with the gantry, and detail of the accelerator layout (right). Credits: D Kaprinis, SEEIIST/CERN (left) and D Kaprinis, SEEIIST/CERN (right)

Twenty years ago, pioneering work at CERN helped propel Europe to the forefront of cancer treatment with hadron beams. The Proton Ion Medical Machine Study (PIMMS), founded in 1996 by a CERN–TERA Foundation-MedAustron–Oncology2000 collaboration, paved the way to the construction of two hadron-therapy centres: CNAO in Pavia (Italy) and MedAustron in Wiener Neustadt (Austria). A parallel pioneering development at GSI produced two similar centres in Germany (HIT in Heidelberg and MIT in Marburg). Since the commissioning of the first facility in 2009, the four European hadron-therapy centres have treated more than 10,000 patients with protons or carbon ions. The improved health and life expectancy of these individuals is the best reward to the vision of all those at CERN and GSI who laid the foundations for this new type of cancer treatment.

Almost four million new cancer cases are diagnosed per year in Europe, around half of which can be effectively treated with X-rays at relatively low cost. Where hadrons are advantageous is in the treatment of deep tumours close to critical organs or of paediatric tumours. For these cancers, the “Bragg peak” energy-deposition characteristic of charged particles reduces the radiation dose to organs surrounding the tumour, increasing survival rates and reducing negative side effects and the risk of recurrency. With respect to protons, carbon ions have the additional advantages of hitting the target more precisely with higher biological effect, and of being effective against radioresistant hypoxic tumours, which constitute between 1 and 3% of all radiation-therapy cases. Present facilities treat only a small fraction of all patients who could take advantage of hadron therapy, however. The diffusion of this relatively novel cancer treatment is primarily limited by its cost, and by the need for more pre-clinical and clinical research to fully exploit its potential.

Given these limitations, how can the scientific community contribute to extending the benefits of hadron therapy to a larger number of cancer patients? To review this and similar questions, CERN has recently given a new boost to its medical accelerator activities, after a long interruption corresponding to the time when CERN resources where directed mainly towards LHC construction. The framework for this renewed effort was provided by the CERN Council in 2017 when it approved a strategy concerning knowledge-transfer for the benefit of medical applications. This strategy specifically encouraged new initiatives to leverage existing and upcoming CERN technologies and expertise in accelerator technologies towards the design of a new generation of light-ion accelerators for medicine.

**Combined windings** Solution for curved and straight “canted-cosine-theta” coils combining dipole and quadrupole in the same winding. Credits: G Kirby and J van Nugteren, CERN (left) and G Kirby and J van Nugteren, CERN (right)

The hadron-therapy landscape in 2020 is very different from what it was 20 years ago. The principal reason is that industry has entered the field and developed a new generation of compact cyclotrons for proton therapy. Beyond the four hadron (proton and ion) centres there are now 23 industry-built facilities in Europe providing only proton therapy to about 4000 patients per year. Thanks to this new set of facilities, proton therapy is now highly developed and is progressively extending its reach in competition with more conventional X-ray radiation therapy.

Despite its many advantages over X-rays and protons, therapy with ions (mainly carbon, but other ions like helium or oxygen are under study) is still administered in Europe only by the four large hadron-therapy facilities. In comparison, eight ion-therapy accelerators are in operation in Asia, most of them in Japan, and four others are under construction. The development of new specific instruments for cancer therapy with ions is an ideal application for CERN technologies, in line with CERN’s role of promoting the adoption of cutting-edge technologies that might result in innovative products and open new markets.

Next-generation accelerators

To propel the use of cancer therapy with ions we need a next-generation accelerator, capable of bringing beams of carbon ions to the 430 MeV/u energy required to cover the full body, with smaller dimensions and cost compared to the PIMMS-type machines. A new accelerator design with improved intensity and operational flexibility would also enable a wide research programme to optimise ion species and treatment modalities, in line with what was foreseen by the cancelled BioLEIR programme at CERN. This would allow the exploration of innovative paths to the treatment of cancer such as ultra-short FLASH therapy or the promising combination of ion therapy with immunotherapy, which is expected to trigger an immune response against diffused cancers and metastasis. Moreover, a more compact accelerator could be installed in, or very close to, existing hospitals to fully integrate ion therapy in cancer-treatment protocols while minimising the need to transport patients over long distances.

The development of new specific instruments for cancer therapy with ions is an ideal application for CERN technologies

These considerations are the foundation for the Next Ion Medical Machine Study (NIMMS), a new CERN initiative that aims to develop specific accelerator technologies for the next generation of ion-therapy facilities and help catalyse a new European collective action for therapy with ion beams. The NIMMS activities were launched in 2019, following a workshop at ESI Archamps in 2018 where the medical and accelerator communities agreed on basic specifications for a new-generation machine. In addition to smaller dimensions and cost, these include a higher beam current for faster treatment, operation with multiple ions, and irradiation from different angles using a gantry system.

In addressing the challenges of new designs with reduced dimensions, CERN is building on the development work promoted in the last decade by the TERA Foundation. Reducing the accelerator dimensions from the conventional synchrotrons used so far can take different directions, out of which two are particularly promising. The first is the classic approach of using superconductivity to increase the magnetic field and decrease the radius of the synchrotron, and the second consists of replacing the synchrotron with a high-gradient linear accelerator with a new design – in line with the proton therapy linac being developed by ADAM, a spin-off company of CERN and TERA now part of the AVO group. The goal in both designs is to reduce the surface occupied by the accelerator by more than a factor of two, from about 1200 to 500 m². With these considerations in mind, the NIMMS study has been structured in four work packages.

The main avenue to reduced dimensions is superconductivity, and the goal of the first work package is to develop new superconducting magnet designs for pulsed operation, with large apertures and curvatures – suitable for an ideal “square” synchrotron layout with only four 90 degree magnets. Different concepts are being explored, with some attention to the so-called canted cosine-theta design (see “Combined windings”) used for example in orbit correctors for the high-luminosity LHC, of which a team at Lawrence Berkeley National Laboratory has recently developed a curved prototype for medical applications. Other options under study are based on more traditional cosine-theta designs (see “Split yoke”), and on exploiting the potential of modern high-temperature superconductors.

**Split yoke** Model of a curved cosine-theta dipole and vertical split yoke with assembly clamps. Credit: C Kokkinos/FEAC Engineering (left) and C Kokkinos/FEAC Engineering (right).

The second work package covers the design of a compact linear accelerator optimised for installation in hospitals. Operating at 3 GHz with high field gradients, this linac design profits from the expertise gained with accelerating structures developed for the proposed Compact Linear Collider (CLIC), and uses as an injector a novel source for fully-stripped carbon based on the REX-ISOLDE design. The source is followed by a 750 MHz radio-frequency quadrupole using the design recently developed at CERN for medical and industrial applications.

The third NIMMS work package focuses on compact superconducting designs for the gantry, the large element required to precisely deliver ion beams to the patient that is critical for the cost and performance of an ion-therapy facility. The problem of integrating a large-acceptance beam optics with a compact superconducting magnetic system within a robust mechanical structure is an ideal challenge for the expertise of the CERN accelerator groups. Two designs are being considered: a lightweight rotational gantry covering only 180 degrees originally proposed by TERA, and the GaToroid toroidal gantry being developed at CERN.

NIMMS will consider new designs for the injector linac, with reduced cost and dimensions

The fourth work package is dedicated to the development of new high-current synchrotron designs, and to their integration in future cancer research and therapy facilities. To reduce treatment time, the goal is to accelerate more than an order of magnitude higher current than in the present European facilities. This requires careful multi-turn injection into the ring and strict control of beam optics, which add to other specific features of the new design, including a fast extraction that will make tests with the new ultra-fast FLASH treatment modality possible. Two synchrotron layouts are being considered, a more conventional one with room-temperature magnets (see “Ions for therapy”), and a very compact superconducting one of only 27 m circumference. The latter, equipped with a gantry of new design, would allow a single-room carbon-therapy facility to be realised in an area of about 1000 m². Additionally, NIMMS will consider new designs for the injector linac, with reduced cost and dimensions and including the option of being used for production of medical radioisotopes – for imaging and therapy – during the otherwise idle time between two synchrotron injections.

Ambitious work plan

This ambitious work plan exceeds the resources that CERN can allocate to this study, and its development requires collaborations at different levels. The first enthusiastic partner is the new SEEIIST (South East European International Institute for Sustainable Technologies) organisation, which aims at building a pan-European facility for cancer research and therapy with ions (see “Ions for therapy”). SEEIIST is already joining forces with NIMMS by supporting staff working at CERN on synchrotron and gantry design. The second partnership is with the ion therapy centres CNAO and MedAustron, which are evaluating the proposed superconducting gantry design in view of extending the treatment capabilities of their facilities. A third critical partner is CIEMAT, which will build the high-frequency linac pre-injector and validate it with beam. Other partners participating in the study at different levels are GSI, PSI, HIT, INFN, Melbourne University, Imperial College, and of course TERA which remains one of the driving forces behind medical-accelerator developments. This wide collaboration has been successful in attracting additional support from the European Commission via two recently approved projects beginning in 2021. The multidisciplinary HITRIplus project on ion therapy includes work packages dedicated to accelerator, gantry and superconducting magnet design, while the IFAST project for cutting-edge accelerator R&D contains an ambitious programme focusing on the optimisation and prototyping of superconducting magnets for ion therapy with industry.

Every technology starts from a dream, and particle accelerators are there to fulfil one of the oldest: looking inside the human body and curing it without bloodshed. It is up to us to further develop the tools to realise this dream.

Topics

Reset your password

Registration complete