S.M Hooker, D.A Jaroszynski, K Burnett


An overview is provided of the papers presented at the Discussion Meeting on ‘Laser-driven particle accelerators: new sources of energetic particles and radiation’ which was held at the Royal Society on 6 and 7 June 2005. A discussion of the future prospects of the field, held during the last session of the meeting, is summarized and the conclusions of this presented.


Relativistic plasma waves can be formed by intense laser pulses as they propagate through a plasma. The longitudinal electric fields within such waves can be three or more orders of magnitude greater than those used in conventional accelerators, and consequently laser-driven plasma accelerators offer the prospect of a new generation of extremely compact devices. Further, the output beams may be used to generate coherent radiation from terahertz frequencies to X-rays. The unique properties of these particle and radiation beams promise their widespread application in the biological and physical sciences.

This issue of Philosophical Transactions A contains papers presented at the Discussion Meeting on ‘Laser-driven particle accelerators: new sources of energetic particles and radiation’ held at the Royal Society on 6 and 7 June 2005. Scientists from Europe and the United States, with backgrounds in both the physical and biological sciences attended what proved to be a lively and fruitful meeting.

The basic concepts underlying plasma accelerators driven by either laser or particle beams are described in the paper by Bingham (2006). He discusses the operation of three types of plasma accelerators driven by intense laser pulses: beat-wave, laser wakefield and self-modulated plasma accelerators. He also describes plasma accelerators driven by energetic particle beams, and in particular the so-called ‘blow-out’ regime in which a driving electron beam completely expels plasma electrons from the vicinity of the driver pulse to form a cavity or bubble, within which exist intense radial and longitudinal electric fields.

The connection between plasma accelerators driven by charged particle beams and those driven by intense laser pulses is further explored in the paper by Reitsma & Jaroszynski (2006). They outline the so-called photon kinetic theory, in which the electromagnetic field is described in terms of classical quasi-particles, and compare the wakefields and collective effects—such as beam loading—created by electron and photon bunches.

Joshi & Mori (2006) describe experimental and theoretical work on electron-beam-driven plasma accelerators operating in the blow-out regime. They present the results of experiments performed with a 28.5 GeV electron beam at the Stanford Linear Accelerator Center (SLAC) in which acceleration of the trailing electrons in the driving bunch were accelerated by up to 4 GeV in a distance of only 10 cm. In those experiments intense X-rays at approximately 6.5 keV could also be generated by betatron emission arising from oscillation of the electron bunch in the ion channel formed in the bubble.

The blow-out or bubble regime may also be achieved with intense laser pulses if the ponderomotive force is able to expel electrons from the region of the laser pulse. In their paper, Pukhov & Gordienko (2006) develop a similarity theory for the bubble regime within which the trajectories of electrons following the laser–plasma interaction are formally similar for laser and plasma conditions which have a constant value of the similarity parameter, but which otherwise may be widely different. This approach enables scaling laws for the parameters of importance, such as the maximum electron energy, to be deduced for this acceleration regime.

The timing of the Discussion Meeting was very fortunate in that it provided a forum to discuss the results of three very recent experimental demonstrations of the generation of low-energy-spread electron beams by intense laser pulses. Previous experiments to demonstrate laser-driven plasma acceleration had only generated electron beams with a quasi-Boltzmann distribution of the electron energies, corresponding to a fractional energy spread of 100%. As a consequence, while electron beams with energies up to several hundred MeV had been produced, only a small fraction of the accelerated electrons were at the highest energies. The papers by Leemans et al. (2006), Mangles et al. (2006) and Malka et al. (2006) describe work undertaken at Lawrence Berkeley National Laboratory (LBNL), USA, Imperial College (IC), UK and Laboratoire d'Optique Appliquée (LOA), France, respectively. The pioneering work described therein demonstrates the generation of quasi-monoenergetic electron beams with energies in the 80–170 MeV range and an energy spread of only a few percent. Further, the normalized emittance of these beams is estimated to be a few π mm mrad, and the duration of the accelerated electron bunch is likely to be of order 10 fs—beam parameters which are beyond the reach of present day radio-frequency (RF) technology.

In the experiments performed by the groups at LBNL, IC and LOA the source of the accelerated electrons was the plasma itself, the electrons being injected from the background plasma by a process known as wave-breaking. For many applications it would be desirable further to decrease the energy spread and transverse emittance of the generated electron beam. One way of achieving this might be to inject a low-energy electron beam with good emittance and low energy spread into a plasma wave that is below the onset of wave-breaking. A variety of optical techniques for injecting electrons into plasma accelerators are presently being investigated by a large number of groups worldwide, including those at LBNL, IC and LOA. Alternative methods for injection based upon RF technology coupled with photo-injectors driven by femtosecond laser pulses are presented in the paper by van der Wiel et al. (2006). They describe the design of a state-of-the-art RF-based injector followed by a 50 mm long laser-driven plasma accelerator stage driven within a plasma channel. Modelling of the complete system suggests that the 300 ps duration bunches from the photo-injector are accelerated from an energy of a few MeV to approximately 85 MeV. Further, the bunches are compressed to less than 10 fs and have an energy spread below 10%. Additional improvements for decreasing the pulse duration, and increasing the brightness, of the electrons delivered by the photo-injector are also described.

A combination of RF techniques and plasma accelerators is also discussed by Kimura et al. (2006). They describe experiments in which electron micro-bunches were formed, and then accelerated in two inverse free-electron lasers driven by a CO2 laser. Energy gains of approximately 6 MeV, and a relative energy spread of order 1% were achieved. The authors also present calculations of two modifications of the laser wakefield accelerator (LWFA) concept—pseudo-resonant LWFA and seeded self-modified LWFA—which could be driven by a CO2 laser. For both methods, the acceleration gradients are predicted to be of order 1 GeV m−1.

Techniques for guiding the driving laser radiation can allow acceleration to be maintained over distances much longer than the limit set by refractive and diffractive defocusing of the pump radiation. Indeed, a unique feature of the work described by Leemans et al. (2006) is that the electron beams were generated within a so-called plasma waveguide. In a more detailed discussion of this topic, Milchberg et al. (2006) review current techniques for guiding laser pulses with intensities above 1017 W cm−2, before discussing a new approach for generating plasma waveguides. This method—hydrodynamic expansion of a laser-heated clustered medium—allows channels to be formed with a relatively low electron density thereby enabling acceleration over longer distances, and hence to higher energy, before the accelerated electron bunch enters the decelerating phase of the plasma wave.

Two novel laser-based techniques for particle acceleration are described in the paper by Kalmykov et al. (2006). The first operates by generating surface waves via the interaction of a solid structure with an incident laser field. Accelerating fields of order 1 GV m−1 are predicted for this structure, and the results of initial experiments are reported. The second approach is based on a variation of the well-known plasma beat-wave accelerator. In their revised approach, the authors allow the difference in the frequencies of the two lasers driving the plasma beat-wave to be slightly smaller than the plasma frequency. In this configuration a dynamic bi-stability can arise, leading to the formation of relativistic plasma waves with very large amplitudes.

Laser-driven accelerators are not limited to acceleration of electrons. McKenna et al. (2006) discuss the use of intense laser pulses to generate energetic beams of protons. After describing the mechanisms responsible for generating proton beams from the interaction of intense laser pulses with thin foils, the authors present the results of a new study of the influence on the proton generation of the level of the low-intensity pedestal which inevitably precedes an intense laser pulse.

An application of laser-driven plasma accelerators which is of potentially enormous importance is the generation of short pulse, tunable radiation from undulators and free-electron lasers driven by laser-accelerated particle beams. The investigation of this prospect forms the core of the programme of work described by Jaroszynski et al. (2006). In their paper, the authors describe in detail their theoretical and experimental programme including: the development of RF photoinjectors; a study of both geometric and plasma-based techniques for reducing the duration of the electron bunch derived from such injectors; the development and application of plasma channels for guiding the driving laser pulse; and investigation of the potential output of undulators and free-electron lasers driven by laser-accelerated beams.

Further discussion of the applications of electron beams with bunch durations of order 10 fs is provided by Dwyer et al. (2006), who describe the present state and future directions of femtosecond electron diffraction, and the applications of this exciting technique to the study of physical and biological systems.

The meeting was concluded by a discussion of the future prospects of the field. It was recognized that the very high acceleration gradients and intrinsically short bunch duration provided by laser-driven plasma accelerators gave them significant advantages over accelerators based on conventional techniques. It was acknowledged, however, that conventional accelerators and radiation sources were a highly successful and reliable technology with a well-established user base. As such, adoption of plasma-based acceleration techniques will only occur if they offer real advantages over conventional technology, and that these are demonstrated to potential users.

It was concluded that there would probably be several stages in the evolution of laser-driven accelerators. In the relatively near term it is likely that very compact, single-stage accelerators could be developed which would be capable of delivering electron beams with energies in the 100 MeV–1 GeV range, and pulse durations of the order of 10 fs. Devices of this type could be used directly in radiography experiments or in femtosecond electron diffraction with unprecedented temporal resolution. In order to use these compact accelerators for driving radiation sources such as X-ray free-electron lasers, it will be necessary to demonstrate reproducible acceleration of electron bunches which have low emittance, low energy spread, high peak current and are of femtosecond duration. In this regard, the experimental results reported on the generation of quasi-monoenergetic beams are extremely promising.

It was agreed that the development of ultra-high energy accelerators is likely to occur over a much longer timescale, and that many problems remain to be solved. For example, it will be necessary to maintain acceleration over many stages, operate at increased pulse repetition rates, and improve the total wall-plug energy efficiency. Developments in these areas are likely to rely on improvements in laser technology as much as in plasma accelerators.

A major conclusion of the Discussion Meeting was that it would be useful to draw a roadmap for the development of plasma-based accelerators. This would lay out the key goals and identify where advances in supporting technologies and theory are required. It was agreed that it would be imperative to involve potential user communities in identifying a useful route forward for establishing these potentially powerful new tools for researchers.


We would like to thank the speakers for an excellent series of papers, and the Royal Society for supporting the meeting and for providing delightful surroundings with excellent technical facilities. We would also like to acknowledge the invaluable support of Katherine Hardaker in organizing the meeting, and Cathy Brennan in co-ordinating the production of this issue.


  • One contribution of 15 to a Discussion Meeting Issue ‘Laser-driven particle accelerators: new sources of energetic particles and radiation’.


View Abstract