To the terascale: catching the plasma wakefield

21 October 2013

A look at a design study of 1983 and subsequent progress.


Champs de sillage plasma à l’échelle du téra

Il y a 30 ans, John Lawson, du Laboratoire Rutherford Appleton, menait une étude de conception concernant un accélérateur linéaire à champs de sillage plasma excités par laser pour le XXIe siècle. Le modèle de conception produit alors reste une référence pour la définition d’une machine de plusieurs TeV, car la plupart des points importants soulevés à l’époque sont encore à l’étude. Certains problèmes majeurs ont été résolus, notamment la production de faisceaux d’électrons à haute énergie avec une bonne émittance, et la formation d’une colonne de plasma uniforme. Cependant, différentes expériences poursuivent les efforts visant à trouver des réponses à des questions essentielles de la physique sous-jacente.

Particle accelerators developed during the past century are approaching the energy frontier. Today at the terascale, the machines needed are extremely large and costly. However, for more than 30 years, plasma-based particle accelerators driven by either lasers or particle beams have shown promise for a route to high energies, primarily because of the extremely large accelerating electric fields they can support. About a thousand times greater than in conventional accelerators, the high fields would allow the possibility of compact accelerating structures. It is with this in mind that future facilities may incorporate aspects of plasma accelerators.

Plasma-based accelerators – the brainchild of John Dawson (who died in 2001) and his colleagues at the University of California, Los Angeles – are being investigated worldwide with a great deal of success. However, can they be a serious competitor and displace the conventional “dinosaur” variety? This is the question that the late John Lawson at the Rutherford Appleton Laboratory in the UK posed a few years after Dawson and his collaborator Toshi Tajima published their seminal paper on plasma-based accelerators (Tajima and Dawson 1979).

The accelerating fields in plasma are supported by the collective motion of plasma electrons forming a space-charge disturbance that moves close to the speed of light – commonly known as the plasma wakefield accelerator (CERN Courier June 2007 p28). The main advantage is that the plasma can support accelerating fields many orders of magnitude greater than conventional devices, which suffer from breakdown of the waveguide structure. In contrast, in a plasma-based system the plasma is already “broken down” and the collective electric field, E, supported by the plasma is determined by the electron density, n, such that E ∼ n1/2.

In 1982, Ronald Ruth and Alexander Chao in the US made a first qualitative design based on a simplified model for a linear laser-driven plasma accelerator that could yield electrons at 5 TeV (Ruth and Chao 1982). Spurred on by this, Lawson set up a study group – including Ruth – to investigate the idea further and produce a design based on a more realistic model for a particle collider at the terascale for the 21st century. Published in the summer of 1983, the reference design considered electron energies above 1 TeV and – because of synchrotron radiation losses – a linear collider (Lawson et al. 1983). Indeed, it is as components of linear colliders that plasma accelerators continue to be considered.

At the time it was already clear that because of the increased energy advantage of colliding beams, all future high-energy accelerators would work in the colliding-beam mode. Conventional machines were planned that would reach the 1 TeV energy scale for hadrons and the 100 GeV scale for electron–positron colliders. These became the Tevatron at Fermilab, the SLAC Linear Collider and the Large Electron–Positron collider and Large Hadron Collider at CERN, which have made remarkably precise measurements of the W and Z bosons, as well as discoveries of the top quark and – most recently – a Higgs boson.

In the RAL report, Lawson writes that “the maximum energies thought ever to be practically achievable are of order 0.4 TeV per beam for e+e and 20 TeV per beam for hadrons” and that these represent “the end of the dinosaurs”. However, radically new techniques such as plasma accelerators demand a long development time, so there is still ongoing development of conventional machines. These include the International Linear Collider for 350–500 GeV electrons and positrons with luminosity of the order of 1034 cm–2 s–1 and the more novel Compact Linear Collider for energies up to 3 TeV. In addition, there is the TLEP concept based on a synchrotron, which could also be used for protons (CERN Courier July/August 2013 p26).

The original plasma-accelerator schemes investigated were based on a long-pulse laser. Short-pulse lasers did not exist because chirp pulse amplification had not yet been demonstrated in the optical regime, only in the microwave regime. The Lawson design therefore incorporated the beat-wave mechanism of Tajima and Dawson, where two laser beams with a frequency difference equal to the plasma frequency drive a large-amplitude plasma wave with a phase velocity close to the speed of light. Most laser-driven and particle-driven particle accelerator experiments today are in the so-called bubble regime where the pulse length of the laser or particle beam is of the order of the plasma wavelength. In 2004, three independent groups in three different countries demonstrated mono-energetic electron beams with good emittance using short-pulse lasers and many groups worldwide now routinely produce electron beams at giga-electron-volt energies (CERN Courier November 2004 p5).

As a first attempt at finding a set of consistent parameters appropriate to a linear e+e collider at a few tera-electron-volts, Lawson’s design was optimistic but it laid the groundwork for later studies on the scalings in length of such a machine. For example, Carl Schroeder and colleagues at Lawrence Berkeley National Laboratory (LBNL) recently looked at physics considerations for a 1 TeV machine based on plasma acceleration driven by a short-pulse laser (Schroeder et al. 2010).

Despite the difference in the accelerating structure considered in 1983, the outline design of the RAL report is still a good blueprint for a multi-tera-electron-volt device. Its main points are still being investigated. These include the construction of uniform metre-long plasma columns, laser focusing and guiding schemes, the laser energy and efficiency requirements, staging, particle transport and focusing. There are also requirements on the particle bunch-density and on the beam quality in terms of energy spread, emittance and luminosity. Some of these issues are common to both plasma and conventional accelerators, although Lawson’s design brief was that plasma-based techniques should take particle accelerators to a new level, not easily achievable in a conventional machine.

In particular, in single-pass colliders, bunches are used only once and high densities are required. Self-fields add rather than cancel, causing a pinch that enhances the luminosity – provided the effect is not too strong. However, deflection of particles by the electric fields of the opposite bunch causes strong synchrotron-like radiation, known as beamstrahlung, which reduces the energy of the particles and increases the energy spread. This can be controlled by having fewer particles per bunch and instead having a train of “bunchlets” – a feature that Lawson’s report found to be advantageous.

Since then, researchers have solved many of the issues, an important one – highlighted by Lawson – being creation of a uniform plasma column. In an experiment at SLAC in 2007, Chan Joshi’s group at University of Califormia, Los Angeles, successfully demonstrated acceleration in metre-long plasma columns using lithium vapour that was fully ionized by the head of an electron beam. The resulting wakefield then accelerated particles near the back of the beam pulse. This idea will be incorporated in the group’s latest round of experiments to accelerate electrons and positrons using the electron-beam-driven plasma wakefield facility, FACET, at SLAC (CERN Courier March 2011 p23). It also underpins the proton-beam-driven plasma wakefield experiment, AWAKE, which will use a beam from CERN’s Super Proton Synchrotron and, initially, a 10-m-long plasma column, to produce giga-electron-volt electron beams (see AWAKE: to high energies in a single leap).

Today, most experiments such as the Berkeley Lab Laser Experiment (BELLA) at LBNL and FACET at SLAC, as well as other smaller-scale experiments, are aimed at addressing some of the key areas of the underlying physics that still have not been fully resolved. Joshi’s experiment not only showed that metre-length plasma columns can be built with the required density and homogeneity, it also demonstrated energy doubling of an electron beam from 42 GeV to 85 GeV in the lithium plasma – a remarkable result considering that it takes 3 km of the SLAC linac to accelerate electrons to 42 GeV (CERN Courier April 2007 p5). Laser wakefield experiments already demonstrate mono-energetic electron beams at the giga-electron-volt scale and BELLA, which is nearing completion of a petawatt-class laser, will demonstrate acceleration of electrons to 10 GeV (CERN Courier October 2012 p10).

Despite the successes of these experiments, it is still necessary to demonstrate beam quality – including low-energy spread and low emittance – and focusing of useful beams. In all cases, the experiments are guided by plasma simulations that require the largest computers. Such simulations have already demonstrated that in the range 10–50 GeV electron beams can be created in one stage of a plasma accelerator.

If plasma accelerators are to take over from conventional machines, a great deal of effort still needs to be put into efficient drivers. Suitable laser efficiency and pulse rate are looking likely with diode-pumped lasers or with fibre lasers (see Can fibre be the future of high-energy physics?), but effort has to be put into these schemes to meet the requirements necessary to drive a wakefield. For beam-driven systems, electron beams at 100 GeV and proton beams with tera-electron-volt energies are required. These exist at the LHC for protons and at the old SLAC linac for electrons. For an e+e system, a key challenge is positron acceleration and some groups are looking at positron acceleration in wakefields. Alternatively an ee collider or a photon (γ–γ) collider could be built, doing away with the need for positrons and so saving time and effort.

A number of other applications for plasma-based accelerators have been identified such as X-ray generators through betatron radiation, drivers for free-electron lasers, or low-energy proton machines. After more than 30 years it is time to develop the facilities that can answer some of the outstanding issues to demonstrate the full potential of high-energy plasma-based accelerators.

bright-rec iop pub iop-science physcis connect