The AWAKE experiment is adapting plasma-wakefield acceleration for applications in particle physics.
A laser ionises rubidium vapour, turning it into plasma. A proton bunch plunges inside, evolving into millimetre-long microbunches. The microbunches pull the plasma’s electrons, forming wakes in the plasma, like a speedboat displacing water. Crests and troughs of the plasma’s electric field trail the proton microbunches at almost the speed of light. If injected at just the right moment, relativistic electrons surf on the accelerating phase of the field over a distance of metres, gaining energy up to a factor of 1000 times faster than can be achieved in conventional accelerators.
Plasma wakefield acceleration is a cutting-edge technology that promises to revolutionise the field of particle acceleration by paving the way for smaller and more cost-effective linear accelerators. The technique traces back to a seminal paper published in 1979 by Toshiki Tajima and John Dawson which laid the foundations for subsequent breakthroughs. At its core, the principle involves using a driver to generate wakefields in a plasma, upon which a witness beam surfs to undergo acceleration. Since the publication of the first paper, the field has demonstrated remarkable success in achieving large accelerating gradients.
Traditionally, only laser pulses and electron bunches have been used as drive beams. However, since 2016 the Advanced Wakefield Experiment (AWAKE) at CERN has used proton bunches from the Super Proton Synchrotron (SPS) as drive beams – an innovative approach with profound implications. Thanks to their high stored energy, proton bunches enable AWAKE to accelerate an electron bunch to energies relevant for high-energy physics in a single plasma, circumventing the need for the multiple accelerating stages that are required when using lasers or electron bunches.
Bridging the divide
Relevant to any accelerator concept based on plasma wakefields, AWAKE technology promises to bridge the gap between global developments at small scales and possible future electron–positron colliders. The experiment is therefore an integral component of the European strategy for particle physics’ plasma roadmap, aiming to advance the concept to a level of technological maturity that would allow their application to particle-physics experiments. An international collaboration of approximately 100 people across 22 institutes worldwide, AWAKE has already published more than 90 papers, many in high-impact journals, alongside significant efforts to train the next generation, culminating in the completion of over 28 doctoral theses to date.
In the experiment, a 400 GeV proton bunch from the SPS is sent into a 10 m-long plasma source containing rubidium vapour at a temperature of around 200 °C (see “Rubidium source” figure). A laser pulse accompanies the proton bunch, ionising the vapour and transforming it into a plasma.
To induce the necessary wakefields, the drive bunch length must be of the order of the plasma wavelength, which corresponds to the natural oscillation period of the plasma. However, the length of the SPS proton bunch is around 6 cm, significantly longer than the 1 mm plasma wavelength in AWAKE, and short wavelengths are required to reach large accelerating gradients.
The solution is to take advantage of a beam-plasma instability, which transforms long particle bunches into microbunches with the period of the plasma through a process known as self-modulation. In other words, as the long proton bunch traverses the plasma, it can be coaxed into splitting into a train of shorter “microbunches”. The bunch train resonantly excites the plasma wave, like a pendulum or a child on a swing, being pushed with small kicks at its natural oscillation interval or resonant frequency. If applied at the right time, each kick increases the oscillation amplitude or height of the wave. When the amplitude is sufficiently high, a witness electron bunch from an external source is injected into the plasma wakefields, to ride the wakefields and gain energy.
The first phase of AWAKE (Run 1, from 2016 to 2018) served as a proof-of-concept demonstration of the acceleration scheme. First, it was shown that a plasma can be used as a compact device to self-modulate a highly relativistic and highly energetic proton bunch (see “Self-modulation” figure). Second, it was shown that the resulting bunch train resonantly excites strong wakefields. Third – the most direct demonstration – it was shown that externally injected electrons can be captured, focused and accelerated to GeV energies by the wakefields. The addition of a percent-level positive gradient in density along the plasma led to 20% boosts in the energy gained by the accelerated electrons.
Based on these proof-of-principle experimental results and expertise at CERN and in the collaboration, AWAKE developed a well-defined programme for Run 2, which launched in 2021 following Long Shutdown 2, and which will run for several more years from now. The goal is to achieve electron acceleration with GeV/m energy gain and beam quality similar to a normalised emittance of 10 mm-mrad and a relative energy spread of a few per cent. In parallel, scalable plasma sources are being developed that can be extended up to hundreds of metres in length (see “Helicon plasma source” and “Discharge source” figures). Once these goals are reached, the concepts of AWAKE could be used in particle-physics applications such as using electron beams with energy between 40 and 200 GeV impinging on a fixed target to search for new phenomena related to dark matter.
Controlled instability
The first Run 2 milestone, on track for completion by the end of the year, is to complete the self-modulator – the plasma that transforms the long proton bunch into a train of microbunches. The demonstration has been staged in two experimental phases.
The first phase was completed in 2022. The results prove that wakefields driven by a full proton bunch can have a reproducible and tunable timing. This is not at all a trivial demonstration given that the experiment is based on an instability!
Techniques to tune the instability are similar to those used with free-electron lasers: provide a controlled initial signal for the instability to grow from and operate in the saturated regime, for example. In AWAKE, the self-modulation instability is initiated by the wakefields driven by an electron bunch placed ahead of the proton bunch. The wakefields from the electron bunch imprint themselves on the proton bunch right from the start, leading to a well defined bunch train. This electron bunch is distinct from the witness bunches, which are later accelerated.
The second experimental phase for the completion of the self-modulator is to demonstrate that high-amplitude wakefields can be maintained over long distances. Numerical simulations predict that self-modulation can be optimised by tailoring the plasma’s density profile. For example, introducing a step in the plasma density should lead to higher accelerating fields that can be maintained over long distances. First measurements are very encouraging, with density steps already leading to increased energy gains for externally injected electrons. Work is ongoing to globally optimise the self-modulator.
AWAKE technology promises to bridge the gap between global developments at small scales and possible future electron–positron colliders
The second experimental milestone of Run 2 will be the acceleration of an electron bunch while demonstrating its sustained beam quality. The experimental setup designed to reach this milestone includes two plasmas: a self-modulator that prepares the proton bunch train, and a second “accelerator plasma” into which an external electron bunch is injected (see “Modulation and acceleration” figure). To make space for the installation of the additional equipment, CERN will in 2025 and 2026 dismantle the CNGS (CERN Neutrinos to Gran Sasso) target area that is installed in a 100m-long tunnel cavern downstream from the AWAKE experimental facility.
Accelerate ahead
Two enabling technologies are needed to achieve high-quality electron acceleration. The first is a source and transport line to inject the electron bunch on-axis into the accelerator plasma. A radio-frequency (RF) injector source was chosen because of the maturity of the technology, though the combination of S-band and X-band structures is novel, and forms a compact accelerator with possible medical applications. It is followed by a transport line that preserves the parameters of the 150 MeV 100 pC bunch, and allows for its tight focusing (5 to 10 µm) at the entrance of the accelerator plasma. External injection into plasma-based accelerators is challenging because of the high frequency (about 235 GHz in AWAKE) and thus small structure size (roughly 200 µm) at which they operate. The main goal is to demonstrate that the electron bunch can be accelerated to 4 to 10 GeV, with a relative energy spread of 5 to 8%, and emerge with approximately the same normalised emittance as at the entrance of the plasma (2–30 mm mrad).
For these experiments, rubidium vapour sources will be used for both the self-modulator and accelerator plasmas, as they provide the uniformity, tunability and reproducibility required for the acceleration process. However, the laser-ionisation process of the rubidium vapour does not scale to lengths beyond 20 m. The alternative enabling technology is therefore a plasma source whose length can be scaled to the 50 to 100 metres required for the bunch to reach 50–100 GeV energies. To achieve this, a laboratory to develop discharge and helicon-plasma sources has been set up at CERN (see “Discharge source” figure). Multiple units can in principle be stacked to reach the desired plasma length. The challenge with such sources is to demonstrate that they can produce required plasma parameters other than length.
The third and final experimental milestone for Run 2 will then be to replace the 10 m-long accelerator plasma with a longer source and achieve proportionally larger energy gains. The AWAKE acceleration concept will then essentially be mature to propose particle-physics experiments, for example with bunches of a billion or so 50 GeV electrons.
