Subject Grouping: Technology

Sweden’s MAX IV synchrotron radiation facility is among an elite cadre of advanced X-ray sources, shedding light on the structure and behaviour of matter at the atomic and molecular level across a range of fundamental and applied disciplines – from clean-energy technologies to pharma and healthcare, from structural biology and nanotech to food science and cultural heritage. 

In terms of core building blocks, this fourth-generation light source – which was inaugurated in 2016 – consists of a linear electron accelerator plus 1.5 and 3 GeV electron storage rings (with the two rings optimised for the production of soft and hard X rays, respectively). As well as delivering beam to a short-pulse facility, the linac serves as a full-energy injector to the two storage rings which, in turn, provide photons that are extracted for user experiments across 14 specialist beamlines.

Underpinning all of this is a ground-breaking implementation of ultrahigh-vacuum (UHV) technologies within MAX IV’s 3 GeV electron storage ring – the first synchrotron storage ring in which the inner surface of almost all the vacuum chambers along its circumference are coated with non-evaporable-getter (NEG) thin film for distributed pumping and low dynamic outgassing. Here, Marek Grabski, MAX IV vacuum section leader, gives CERN Courier the insider take on a unique vacuum installation and its subsequent operational validation. 

What are the main design challenges associated with the 3 GeV storage-ring vacuum system?

We were up against a number of technical constraints that necessitated an innovative approach to vacuum design. The vacuum chambers, for example, are encapsulated within the storage ring’s compact magnet blocks with bore apertures of 25 mm diameter (see “The MAX IV 3 GeV storage ring: unique technologies, unprecedented performance”). What’s more, there are requirements for long beam lifetime, space limitations imposed by the magnet design, the need for heat dissipation from incoming synchrotron radiation, as well as minimal beam-coupling impedance. 

The answer, it turned out, is a baseline design concept that exploits NEG thin-film coatings, a technology originally pioneered by CERN that combines distributed pumping of active residual gas species with low photon-stimulated desorption. The NEG coating was applied by magnetron sputtering to almost all the inner surfaces (98% lengthwise) of the vacuum chambers along the electron beam path. As a consequence, there are only three lumped ion pumps fitted on each standard “achromat” (20 achromats in all, with a single acromat measuring 26.4 m end-to-end). That’s far fewer than typically seen in other advanced synchrotron light sources. 

The MAX IV 3 GeV storage ring: unique technologies, unprecedented performance

Among the must-have user requirements for the 3 GeV storage ring was the specified design goal of reaching ultralow electron-beam emittance (and ultrahigh brightness) within a relatively small circumference (528 m). As such, the bare lattice natural emittance for the 3 GeV ring is 328 pm rad – more than an order of magnitude lower than typically achieved by previous third-generation storage rings in the same energy range.

Even though the fundamental concepts for realising ultralow emittance had been laid out in the early 1990s, many in the synchrotron community remained sceptical that the innovative technical solutions proposed for MAX IV would work. Despite the naysayers, on 25 August 2015 the first electron beam circulated in the 3 GeV storage ring and, over time, all design parameters were realised: the fourth generation of storage-ring-based light sources was born. 

Stringent beam parameters 

The MAX IV 3 GeV storage ring represents the first deployment of a so-called multibend achromat magnet lattice in an accelerator of this type, with the large number of bending magnets central to ensuring ultralow horizontal beam emittance. In all, there are seven bending magnets per achromat (and 20 achromats making up the complete storage ring). 

Not surprisingly, miniaturisation is a priority in order to accommodate the 140 magnet blocks – each consisting of a dipole magnet and other magnet types (quadrupoles, sextupoles, octupoles and correctors) – into the ring circumference. This was achieved by CNC machining the bending magnets from a single piece of solid steel (with high tolerances) and combining them with other magnet types into a single integrated block. All magnets within one block are mechanically referenced, with only the block as a whole aligned on a concrete girder.

Vacuum innovation

Meanwhile, the vacuum system design for the 3 GeV storage ring also required plenty of innovative thinking, key to which was the close collaboration between MAX IV and the vacuum team at the ALBA Synchrotron in Barcelona. For starters, the storage-ring vacuum vessels are made from extruded, oxygen-free, silver-bearing copper tubes (22 mm inner diameter, 1 mm wall thickness). 

Copper’s superior electrical and thermal conductivities are crucial when it comes to heat dissipation and electron beam impedance. The majority of the chamber walls act as heat absorbers, directly intercepting synchrotron radiation coming from the bending magnets. The resulting heat is dissipated by cooling water flowing in channels welded on the outer side of the vacuum chambers. Copper also absorbs unwanted radiation better than aluminium, offering enhanced protection for key hardware and instrumentation in the tunnel. 

The use of crotch absorbers for extraction of the photon beam is limited to one unit per achromat, while the section where synchrotron radiation is extracted to the beamlines is the only place where the vacuum vessels incorporate an antechamber. Herein the system design is particularly challenging, with the need for additional cooling blocks to be introduced on the vacuum chambers with the highest heat loads. 

Other important components of the vacuum system are the beam position monitors (BPMs), which are needed to keep the synchrotron beam on an optimised orbit. There are 10 BPMs in each of the 20 achromats, all of them decoupled thermally and mechanically from the vacuum chambers through RF-shielded bellows that also allow longitudinal expansion and small transversal movement of the chambers.

Ultimately, the space constraints imposed by the closed magnet block design – as well as the aggregate number of blocks along the ring circumference – was a big factor in the decision to implement a NEG-based pumping solution for MAX IV’s 3 GeV storage ring. It’s simply not possible to incorporate sufficient lumped ion pumps to keep the pressure inside the accelerator at the required level (below 1 × 10^–9 mbar) to achieve the desired beam lifetime while minimising residual gas–beam interactions. 

Operationally, it’s worth noting that a purified neon venting scheme (originally developed at CERN) has emerged as the best-practice solution for vacuum interventions and replacement or upgrade of vacuum chambers and components. As evidenced on two occasions so far (in 2018 and 2020), the benefits include significantly reduced downtime and risk management when splitting magnets and reactivating the NEG coating. 

How important was collaboration with CERN’s vacuum group on the NEG coatings?

Put simply, the large-scale deployment of NEG coatings as the core vacuum technology for the 3 GeV storage ring would not have been possible without the collaboration and support of CERN’s vacuum, surfaces and coatings (VSC) group. Working together, our main objective was to ensure that all the substrates used for chamber manufacturing, as well as the compact geometry of the 3 GeV storage-ring vacuum vessels, were compatible with the NEG coating process (in terms of coating adhesion, thickness, composition and activation behaviour). Key to success was the deep domain knowledge and proactive technical support of the VSC group, as well as access to CERN’s specialist facilities, including the mechanical workshop, vacuum laboratory and surface treatment plant. 

What did the manufacturing model look like for this vacuum system? 

Because of the technology and knowledge transfer from CERN to industry, it was possible for the majority of the vacuum chambers to be manufactured, cleaned, NEG-coated and tested by a single commercial supplier – in this case, FMB Feinwerk- und Messtechnik in Berlin, Germany. Lengthwise, 70% of the chambers were NEG-coated by the same vendor. Naturally, the manufacturing of all chambers had to be compatible with the NEG coating, which meant careful selection and verification of materials, joining methods (brazing) and handling. Equally important, the raw materials needed to undergo surface treatment compatible with the coating, with the final surface cleaning certified by CERN to ensure good film adhesion under all operating conditions – a potential bottleneck that was navigated thanks to excellent collaboration between the three parties involved. 

To spread the load, and to relieve the pressure on our commercial supplier ahead of system installation (which commenced in late 2014), it’s worth noting that most geometrically complicated chambers (including vacuum vessels with a 5 mm vertical aperture antechamber) were NEG-coated at CERN. Further NEG coating support was provided through a parallel collaboration with the European Synchrotron Radiation Facility (ESRF) in Grenoble. 

How did you handle the installation phase? 

This was a busy – and at times stressful – phase of the project, not least because all the vacuum chambers were being delivered “just-in-time” for final assembly in situ. This approach was possible thanks to exhaustive testing and qualification of all vacuum components prior to shipping from the commercial vendor, while extensive dialogue with the MAX IV team helped to resolve any issues arising before the vacuum components left the factory. 

Owing to the tight schedule for installation – just eight months – we initiated a collaboration with the Budker Institute of Nuclear Physics (BINP) in Russia to provide additional support. For the duration of the installation phase, we had two teams of specialists from BINP working alongside (and coordinated by) the MAX IV vacuum team. All vacuum-related processes – including assembly, testing, baking and NEG activation of each achromat (at 180 °C) – took place inside the accelerator tunnel directly above the opened lower magnet blocks of MAX IV’s multibend achromat (MBA) lattice. Our installation approach, though unconventional, yielded many advantages – not least, a reduction in the risks related to transportation of assembled vacuum sectors as well as reduced alignment issues. 

Presumably not everything went to plan through installation and acceptance?

One of the issues we encountered during the initial installation phase was a localised peeling of the NEG coating on the RF-shielded bellows assembly of several vacuum vessels. This was addressed as a matter of priority – NEG film fragments falling into the beam path is a show-stopper – and all the effected modules were replaced by the vendor in double-quick time. More broadly, the experience of the BINP staff meant difficulties with the geometry of a few chambers could also be resolved on the spot, while the just-in-time delivery of all the main vacuum components worked well, such that the installation was completed successfully and on time. After completion of several achromats, we installed straight sections in between while the RF cavities were integrated and conditioned in situ. 

How has the vacuum system performed from the commissioning phase and into regular operation? 

Bear in mind that MAX IV was the first synchrotron light source to apply NEG technology on such a scale. We were breaking new ground at the time, so there were credible concerns regarding the conditioning and long-term reliability of the NEG vacuum system – and, of course, possible effects on machine operation and performance. From commissioning into regular operations, however, it’s clear that the NEG pumping system is reliable, robust and efficient in delivering low dynamic pressure in the UHV regime.

Initial concerns around potential saturation of the NEG coating in the early stages of commissioning (when pressures are high) proved to be unfounded, while the same is true for the risk associated with peeling of the coating (and potential impacts on beam lifetime). We did address a few issues with hot-spots on the vacuum chambers during system conditioning, though again the overall impacts on machine performance were minimal. 

To sum up: the design current of 500 mA was successfully injected and stored in November 2018, proving that the vacuum system can handle the intense synchrotron radiation. After more than six years of operation, and 5000 Ah of accumulated beam dose, it is clear the vacuum system is reliable and provides sustained UHV conditions for the circulating beam – a performance, moreover, that matches or even exceeds that of conventional vacuum systems used in other storage rings.

What are the main lessons your team learned along the way through design, installation, commissioning and operation of the 3 GeV storage-ring vacuum system?

The unique parameters of the 3 GeV storage ring were delivered according to specification and per our anticipated timeline at the end of 2015. Successful project delivery was only possible by building on the collective experience and know-how of staff at MAX-lab (MAX IV’s predecessor) constructing and operating accelerators since the 1970s – and especially the lab’s “explorer mindset” for the early-adoption of new ideas and enabling technologies. Equally important, the commitment and team spirit of our technical staff, reinforced by our collaborations with colleagues at ALBA, CERN, ESRF and BINP, were fundamental to the realisation of a relatively simple, efficient and compact vacuum solution.

Operationally, it’s worth adding that there are many dependencies between the chosen enabling technologies in a project as complex as the MAX IV 3 GeV storage ring. As such, it was essential for us to take a holistic view of the vacuum system from the start, with the choice of a NEG pumping solution enforcing constraints across many aspects of the design – for example, chamber geometry, substrate type, surface treatment and the need for bellows. The earlier such knowledge is gathered within the laboratory, the more it pays off during construction and operation. Suffice to say, the design and technology solutions employed by MAX IV have opened the door for other advanced light sources to navigate and build on our experience.

By 2040, the annual global incidence of cancer is expected to rise by more than 42% from 19.3 million to 27.5 million cases, corresponding to approximately 16.3 million deaths. Shockingly, some 70% of these new cases will be in low- and middle-income countries (LMICs), which lack the healthcare programmes required to effectively manage their cancer burden. While it is estimated that about half of all cancer patients would benefit from radiotherapy (RT) for treatment, there is a significant shortage of RT machines outside high-income countries.

More than 10,000 electron linear accelerators (linacs) are currently used worldwide to treat patients with cancer. But only 10% of patients in low-income and 40% in middle-income countries who need RT have access to it. Patients face long waiting times, are forced to travel to neighbouring regions or face insurmountable expenditure to access treatment. In Africa alone, 27 out of 55 countries have no linac-based RT facilities. In those that do, the ratio of the number of machines to people ranges from one machine to 423,000 people in Mauritius, one machine to almost five million people in Kenya and one machine to more than 100 million people in Ethiopia (see “Out of balance” image). In high-income countries such as the US, Switzerland, Canada and the UK, by contrast, the ratio is one RT machine to 85,000, 102,000, 127,000 and 187,000 people, respectively. To draw another stark comparison, Africa has approximately 380 linacs for a population of 1.2 billion while the US has almost 4000 linacs for a population of 331 million.

Unique challenges

It is estimated that to meet the demand for RT in LMICs over the next two to three decades, the current projected need of 5000 RT machines is likely to become more than 12,000. To put these figures into perspective, Varian, the market leader in RT machines, has a current worldwide installation base of 8496 linacs. While many LMICs provide RT using cobalt-60 machines, linacs offer better dose-delivery parameters and better treatment without the environmental and potential terrorism risks associated with cobalt-60 sources. However, since linacs are more complex and labour-intensive to operate and maintain, their current costs are significantly higher than cobalt-60 machines, both in terms of initial capital costs and annual service contracts. These differences pose unique challenges in LMICs, where macro- and micro-economic conditions can influence the ability of these countries to provide linac-based RT. 

In November 2016 CERN hosted a first-of-its-kind workshop, sponsored by the International Cancer Expert Corps (ICEC), to discuss the design characteristics of RT linacs (see “Linac essentials” image) for the challenging environments of LMICs. Leading experts were invited from international organisations, government agencies, research institutes, universities and hospitals, and companies that produce equipment for conventional X-ray and particle therapy. The following October, CERN hosted a second workshop titled “Innovative, robust and affordable medical linear accelerators for challenging environments”, co-sponsored by the ICEC and the UK’s Science and Technology Facilities Council, STFC. Additional workshops have taken place in March 2018, hosted by STFC in collaboration with CERN and the ICEC, and in March 2019, hosted by STFC in Gaborone, Botswana (see “Healthy vision” image). These and other efforts have identified substantial opportunities for scientific and technical advancements in the design of the linac and the overall RT system for use in LMICs. In 2019, the ICEC, CERN, STFC and Lancaster University entered into a formal collaboration agreement to continue concerted efforts to develop this RT system. 

The idea of novel medical linacs is an excellent example of the impact of fundamental research on wider society

In June 2020, STFC funded a project called ITAR (Innovative Technologies towards building Affordable and equitable global Radiotherapy capacity) in partnership with the ICEC, CERN, Lancaster University, the University of Oxford and Swansea University. ITAR’s first phase was aimed at defining the persistent shortfalls in basic infrastructure, equipment and specialist workforce that remain barriers to effective RT delivery in LMICs. Clearly, a linac suitable for these conditions needs to be low-cost, robust and easy to maintain. Before specifying a detailed design, however, it was first essential to assess the challenges and difficulties RT facilities face in LMICs and in other demanding environments. Published in June 2021, an expansive study of RT facilities in 28 African countries was carried out and compared to western hospitals by the ITAR team to quantitatively and qualitatively assess and compare variables in several domains (see “Downtime” figure). The survey builds on a related 2018 study on the availability of RT services and barriers to providing such services in Botswana and Nigeria, which looked at the equipment maintenance logs of linacs in those countries and selected facilities in the UK.

Surveying the field

The absence of detailed data regarding linac downtime and failure modes makes it difficult to determine the exact impact of the LMIC environment on the performance of current technology. The ongoing ITAR design development and prototyping process identified a need for more information on equipment failures, maintenance and service shortcomings, personnel, training and country-specific healthcare challenges from a much larger representation of LMICs. A further-reaching ITAR survey obtained relevant information for defining design parameters and technological choices based on issues raised at the workshops. They include well-recognised factors such as ease and reliability of operation, machine self-diagnostics and a prominent display of impending or actual faults, ease of maintenance and repair, insensitivity to power interruptions, low power requirement and the consequent reduced heat production.

Based on the information from its surveys, ITAR produced a detailed specification and conceptual design for an RT linac that requires less maintenance, has fewer failures and offers fast repair. Over the next three years, under the umbrella of a larger project called STELLA (Smart Technologies to Extend Lives with Linear Accelerators) launched in June 2020, the project will progress to a prototype development phase at STFC’s Daresbury Laboratory. 

The design of the electron gun has been optimised to increase beam-capture. This has the dual advantage of reducing both the peak current required from the gun to deliver the requisite dose and “back bombardment”. It also allows for simpler replacement of the electron gun’s cathode by trained personnel (current designs require replacement of the full electron gun or even the full linac). Electron-beam capture is limited in medical linacs as the pulses from the electron gun are much longer in duration than the radiofrequency (RF) period, meaning electrons are injected at all RF phases. Some phases cause the bunch to be accelerated while others result in electrons being reflected back to the cathode. In typical linacs, less than 50% of electrons reach the target and many electrons reach the target with lower energies. In high-energy accelerators velocity bunching can be used to compress the bunch, however the space is limited in medical linacs and the energy gain per cell is often well in excess of the beam energy. To allow velocity bunching in a medical linac, the first cell needs to operate at a low gradient – such that less space is required to bunch as the average beam velocity is much lower and the deceleration is less than the beam energy. By adjusting the length of the first and second cells, the decelerated electrons can re-accelerate on the next RF cycle and synchronise with the accelerated electrons, capturing nearly all the electrons and transporting them to the target without a low-energy tail. This is achieved using techniques originally developed for the optimisation of klystrons as part of the Compact Linear Collider project at CERN. By adjusting cell-to-cell coupling, it is possible to make all the other cells at a higher gradient similar to a standard medical linac such that the total linac length remains the same (see “Strong coupling” figure).

The electrical power supply in LMICs can often be variable and protection equipment to isolate harmonics between pieces of equipment is not always installed, hence it is critical to consider this when designing the electrical system for RT machines. This in itself is relatively straightforward but is not normally considered as part of a RT machine design.

The failure of multi-leaf collimators (MLCs), which alter the intensity of the radiation so that it conforms to the tumour volume via several individually actuated leaves, is a major linac downtime issue. Designing MLCs that are less prone to failure will play a key role in RT in LMICs, with studies ongoing into ways to simplify the design without compromising on treatment quality.

Building a workforce

Making it simpler to diagnose and repair faults on linacs is another key area that needs improvement. Given the limited technical staff training in some LMICs, when a machine fails it can be challenging for local staff to make repairs. In addition, components that are degrading can be missed by staff, leading to loss of valuable time to order spares. An important component of the STELLA project, led by ICEC, is to enhance existing and establish new twinning programmes that provide mentoring and training to healthcare professionals in LMICs to build workforce capacity and capability in those regions.

The idea to address the need for a novel medical linac for challenging environments was first presented by Norman Coleman, senior scientific advisor to the ICEC, at the 2014 ICTR-PHE meeting in Geneva. This led to the creation of the STELLA project, led by Coleman and ICEC colleagues Nina Wendling and David Pistenmaa, which is now using technology originally developed for high-energy physics to bring this idea closer to reality – an excellent example of the impact of fundamental research on wider society. 

The next steps are to construct a full linac prototype to verify the higher capture, as well as to improve the ease of maintaining and repairing the machine. Then we need to have the RT machine manufactured for use in LMICs, which will require many practical and commercial challenges to be overcome. The aim of project STELLA to make RT truly accessible to all cancer patients brings to mind a quote from the famous Nigerian novelist Chinua Achebe: “While we do our good works let us not forget that the real solution lies in a world in which charity will have become unnecessary.”