of 4

A laser–plasma accelerator producing monoenergetic electron beams

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
A laser–plasma accelerator producing monoenergetic electron beams
  iments 1,2,9 . No electrons were observed above 40MeV on thespectrometer phosphor screen (detection threshold 10 7 electrons).Plasma density was independently optimized for the unchannelledaccelerator, and highest charge and electron energies were obtainedat 4  £  10 19 cm 2 3 . The high density allowed acceleration before thelaser diffracted since self modulation and dephasing both occurmore quickly at high density, but this also reduced the peak energy compared to the channelled case. Operating the unchannelledaccelerator at the density used for the channelled accelerator(2  £  10 19 cm 2 3 ) produced low-charge low-energy beams, sincethe intensity of the drive beam was not maintained for sufficientdistance to allow acceleration at this density without channelling.Using a 600- m m-long plasma at 4  £  10 19 cm 2 3 , which is close to thedephasing length observed in PIC simulations at this density, theunchannelled accelerator produced the same peak energy as in the2mm plasma, but with more structure in the spectrum. Theseresults confirm that matching accelerator length to dephasinglength is critical for structuring the spectrum, and that extendingthe length (for instance, using a channel at lower density) results inhigher energies.Todifferentiatetheeffectsofchannellingfrompre-ionization,theigniter pulse was fired 80ps before the drive pulse. The plasma doesnot expand significantly over 80ps, so there was no shock wave andthe transverse density profile was flat and had no guiding properties.We observed no difference between the drive pulse only and pre-ionized cases, indicating that channelling and not pre-ionizationwasresponsible for differences in the electron beams described above.The beams from channel-guided accelerators such as thosedescribed here, with a few times 10 9 electrons in per cent levelenergy spread and mrad divergence, as well as intrinsic synchroni-zation to the laser beam, open up a new class of experiments withlaser accelerators. The channelling technology offers the ability tocontrol the laser beam propagation much as a copper structureprovides guiding and field shaping to RF accelerators. The invest-ment of power in channel formation is less than 5% of the drivepulse power (20% of the energy), yet the spectral density of thesebeams near 80MeV is at least a factor of 200 above previousunchannelled experiments using several times the laser power,andpeakenergyobservediscomparable 2 .Thenarrowenergyspreadof the channel-produced beams is consistent with simulations,which also indicate that the bunch length is near 10fs. Thechannel-guidedlaseraccelerator techniquewillhenceallowefficientgeneration of femtosecond X-rays 10 , coherent THz and infraredradiation 13,30 , and is an essential step towards the development of compactmultistageelectronacceleratorswithultrafastbunchesandwithfocusabilityandluminositycompetitivewithstateoftheartRFaccelerators.  A Received 4 June; accepted 29 July 2004; doi:10.1038/nature02900. 1. Modena, A.  et al.  Electron acceleration from the breaking of relativistic plasma waves.  Nature  377, 606–608 (1995).2. Malka,V.  et al.  Electronacceleration byawake fieldforcedbyan intenseultrashort laser pulse. Science 298,  1596–1600 (2002).3. Leemans, W. P.  et al.  Electron-yield enhancement in a laser-wakefield accelerator driven by asymmetric laser pulses.  Phys. Rev. Lett.  89,  174802 (2002).4. Tajima, T. & Dawson, J. M. Laser electron accelerator.  Phys. Rev. Lett.  43,  267–270 (1979).5. Esarey,E.,Sprangle,P.,Krall,J.&Ting,A.Overviewofplasma-basedacceleratorconcepts. IEEETrans.Plasma Sci.  24,  252–288 (1996).6. Esarey, E., Krall, J. & Sprangle, P. Envelope analysis of intense laser pulse self-modulation in plasmas. Phys. Rev. Lett.  72,  2887–2890 (1994).7. Esarey, E., Sprangle, P., Krall, J. & Ting, A. Self-focusing and guiding of short laser pulses in ionizinggases and plasmas.  IEEE J. Quant. Electron.  33,  1879–1914 (1997).8. Najmudin,Z. etal. Self-modulatedwakefieldandforcedlaserwakefieldaccelerationofelectrons. Phys.Plasmas  10,  2071–2077 (2003).9. Leemans, W. P.  et al.  Gamma-neutron activation experiments using laser wakefieldaccelerators.  Phys.Plasmas  8,  2510–2516 (2001).10. Leemans,W.P. etal. Observationofterahertzemissionfromalaser-plasmaacceleratedelectronbunchcrossing a plasma-vacuum boundary.  Phys. Rev. Lett.  91,  074802 (2003).11. Catravas, P., Esarey, E. & Leemans, W. P. Femtosecond x-rays from Thomson scattering using laserwakefield accelerators.  Meas. Sci. Technol.  12,  1828–1834 (2001).12. Wang, X. J., Qiu, X. & Ben-Zvi, I. Experimental observation of high-brightness microbunching in aphotocathode RF electron gun.  Phys. Rev. E   54,  R3121–R3124 (1996).13. Schoenlein, R. W.  et al.  Femtosecond X-ray pulses at 0.4A˚generated by 90 8  Thomson scattering — Atool for probing the structural dynamics of materials.  Science  274,  236–238 (1996).14. Sprangle,P.,Esarey,E.,Krall,J.&Joyce,G.Propagationandguidingofintenselaserpulsesinplasmas. Phys. Rev. Lett.  69,  2200–2203 (1992).15. Leemans, W. P.  et al.  Plasma guiding and wakefield generation for second-generation experiments. IEEE Trans. Plasma Sci.  24,  331–342 (1996).16. Umstadter, D., Kim, J. K. & Dodd, E. Laser injection of ultrashort electron pulses into wakefieldplasma waves.  Phys. Rev. Lett.  76,  2073–2076 (1996).17. Esarey, E.,Hubbard,R. F., Leemans, W. P., Ting,A. &Sprangle, P. Electroninjection into plasmawakefields by colliding laser pulses.  Phys. Rev. Lett.  79,  2682–2685 (1997).18. Durfee, C. G. & Milchberg, H. M. Light pipe for high intensity laser pulses.  Phys. Rev. Lett.  71, 2409–2412 (1993).19. Volfbeyn, P., Esarey, E. & Leemans, W. P. Guiding of laser pulses in plasma channels created by theignitor-heater technique.  Phys. Plasmas  6,  2269–2277 (1999).20. Kim, K. Y., Alexeev, I., Fan, J., Parra, E. & Milchberg, H. M. Plasma waveguides: Addition of endfunnels and generation in clustered gases.  AIP Conf. Proc.  647,  646–653 (2002).21. Gaul, E. W.  et al.  Production and characterization of a fully ionized He plasma channel.  Appl. Phys.Lett.  77,  4112–4114 (2000).22. Toth, C.  et al.  Powerful, pulsed, THz radiation from laser accelerated relativistic electron bunches. Proc. SPIE   5448,  491–504 (2004).23. Leemans,W. P.  etal.  Laser-drivenplasma-basedaccelerators — Wakefieldexcitation,channel guiding,and laser triggered particle injection.  Phys. Plasmas  5,  1615–1623 (1998).24. Strickland, D. & Mourou, G. Compression of amplified chirped optical pulses.  Opt. Commun.  56, 219–221 (1985).25. Leemans, W. P.  et al.  Terahertz radiation from laser accelerated electron bunches.  Phys. Plasmas  5, 2899–2906 (2004).26. Nieter, C. & Cary, J. R. VORPAL: A versatile plasma simulation code.  J. Comput. Phys.  196,  448–473(2004).27. Katsouleas, T., Wilks, S., Chen, S., Dawson, J. M. & Su, J. J. Beam loading in plasma accelerators.  Part. Accel.  22,  81–99 (1987).28. Reitsma, A. J. W.  et al.  Simulation of electron postacceleration in a two-stage laser wakefieldaccelerator.  Phys. Rev. ST Accel. Beams  5,  051301 (2002).29. Tsung, F. S.  et al.  Near GeVenergy laser wakefield acceleration of self-injected electrons in a cm scaleplasma channel.  Phys. Rev. Lett.  submitted.30. Saes, M.  et al.  A setup for ultrafast time-resolved x-ray absorption spectroscopy.  Rev. Sci. Instrum.  75, 24–30 (2004). Acknowledgements  This work was supported by the US Department of Energy and the NationalScience Foundation and used resources of the National Energy Research Scientific ComputingCenterat LBNL;C.G.wasalsosupported bytheHertzFoundation.C.G.acknowledgeshisfaculty advisor J. Wurtele. We appreciate contributions from G. Dugan, J. Faure, G. Fubiani, B. Nagler,K. Nakamura, N. Saleh, B. Shadwick, L. Archambault, M. Dickinson, S. Dimaggio, D. Syversrud, J. Wallig and N. Ybarrolaza. Competing interests statement  The authors declare that they have no competing financialinterests. Correspondence  and requests for materials should be addressed to W.P.L. (wpleemans@lbl.gov). ..............................................................  A laser–plasma acceleratorproducing monoenergeticelectron beams J.Faure 1 ,Y.Glinec 1 ,A.Pukhov  2 ,S.Kiselev  2 ,S.Gordienko 2 ,E.Lefebvre 3 ,J.-P. Rousseau 1 , F. Burgy  1 & V. Malka 1 1 Laboratoire d’Optique Applique´e, Ecole Polytechnique, ENSTA, CNRS,UMR 7639, 91761 Palaiseau, France 2 Institut fur Theoretische Physik, 1, Heinrich-Heine-Universitat Duesseldorf,40225 Duesseldorf, Germany  3 De´ partement de Physique The´orique et Applique´e, CEA/DAM Ile-de-France,91680 Bruye`res-le-Chaˆtel, France ............................................................................................................................................................................. Particle accelerators are used in a wide variety of fields, ranging from medicine and biology to high-energy physics. The accel-erating fields in conventional accelerators are limited to a few tensofMeVm 2 1 ,owingtomaterialbreakdownatthewallsofthestructure. Thus, the production of energetic particle beamscurrently requires large-scale accelerators and expensive infra-structures. Laser–plasma accelerators 1 have been proposed as anext generation of compact accelerators because of the huge letters to nature NATURE|VOL 431|30 SEPTEMBER 2004|www.nature.com/nature  541  ©    200 4 Nature PublishingGroup  electric fields they can sustain 2–5 ( > 100GeVm 2 1 ). However, ithas been difficult to use them efficiently for applicationsbecause they have produced poor-quality particle beams withlarge energy spreads 2–10 , owing to a randomization of electronsin phase space. Here we demonstrate that this randomizationcan be suppressed and that the quality of the electron beams canbe dramatically enhanced. Within a length of 3mm, the laserdrives a plasma bubble 11 that traps and accelerates plasmaelectrons. The resulting electron beam is extremely collimatedand quasi-monoenergetic, with a high charge of 0.5nC at170MeV. For most practical applications, high-quality particle beams withhigh spatial quality and monoenergetic energy distribution arerequired. A beam that does not satisfy these criteria would behard to use, because it would be difficult to transport it and/or tofocusit.Inordertoproducehigh-quality beamsfromplasma-basedaccelerators, two challenges have to be met: (1) the generation of anaccelerating structure in the plasma, and (2) the trapping andacceleration of injected beam loads into the accelerating structure.In order to generate accelerating structures in the plasma, a focusedultraintense laser pulse is used to drive large-amplitude plasmawaves.One possible method for achieving this is the laser wakefield,inwhich plasmawavesare excited by thelaser ponderomotiveforce.When the laser pulse length  c  t   (where  c   is the speed of light and  t   isthe pulse duration) is comparable to the plasma wavelength  l  p , theponderomotive force, which is proportional to the gradient of thelaser intensity, efficiently pushes plasma electrons out of the regionsof strong laser field. Thus, electrons are separated from the ions,which do not move because of their higher mass. This createsthe space charge field needed for particle acceleration, that is, theplasma wave.The generation of intense accelerating fields in plasmas has beendemonstrated in manyexperiments 8,12,13 . Proof-of-principle experi-ments have shown the feasibility of externally injecting electronsfrom a conventional accelerator into the laser-driven plasma accel-erating structure 8–10 . However, the output beam quality has beenpoor: the electron energy distribution has had a 100% energy spread. Until now, the most widespread method for producingelectron beams from plasmas has relied on the self-modulated laserwakefieldaccelerator 14–16 .Inthisaccelerator,thelaserpulseislongerthan the plasma wavelength. Under the influence of the self-modulation instability, its envelope modulates at the plasmafrequency and resonantly excites a plasma wave. When the plasmawave amplitude reaches the wave-breaking level, copious amountsof plasma background electrons are trapped in the plasmawave andaccelerated. Numerous experiments have produced electron beamswith nC charge and divergence varying from a few degreesto tens of degrees and maxwellian energy distributions 2–4 . More recently,several groups 5–7,17 have demonstrated that more compact laserscan be used to efficiently generate high-repetition-rate (10Hz)electron sources, which could be used for applications. However,these beams still have very large energy spreads and a low numberof electrons at high energy (typically   , 1pC at 200 ^ 10MeV).Previous experiments inherently produced poor-quality beams:wave-breaking occurred under the laser pulse envelope and theaccelerated electrons were also under the influence of the ultra-intense laser field. Direct laser acceleration 6,18 by transverse laserfield caused the spatial beam quality to deteriorate, causing emit-tance growth.Here we demonstrate the generation of high-quality electronbeams from ultraintense laser–plasma acceleration. Extremely col-limated beams with 10mrad divergence and 0.5 ^ 0.2nC of chargeat 170 ^ 20MeV have been produced. Contrary to all previousresults obtained from laser–plasma accelerators, the electron energy distribution is quasi-monoenergetic. The number of high-energy electrons (170MeV) is increased by at least three orders of magni-tude with respect to previous work.The experiment was performed by focusing a chirped pulseamplification laser 19,20 onto a helium gas jet (Fig. 1). Figure 2ashows a picture of the electron beam when no magnetic field isapplied. The electron beam is very well collimated, with a 10mraddivergence (full-width at half-maximum, FWHM); to our knowl- Figure 1  Experimental set-up. Top, picture of the experiment; bottom, diagram. Anultrashort and ultraintense laser pulse is focused onto a 3mm supersonic gas jet andproduces a highly collimated 170MeV electron beam. LANEX is a phosphor screen; CCD,charge-coupled device camera; ICT, integrating current transformer. Figure 2  Raw images obtained on the LANEX screen. The vertical axis represents thebeam angular divergence. When a magnetic field is applied, the horizontal axisrepresents electron energy. The white vertical dashed line is drawn at the intersectionof the laser axis with the LANEX screen.  a , Image of the electron beam spatialdistribution obtained from the LANEX screen when no magnetic field (  B   ) is applied. b , Image obtained when the magnetic field is applied, showing that the bulk of thebeam is deviated and its position corresponds to 170MeV electrons. The fact that thebeam trajectory is displaced when a magnetic field is applied confirms that the signalon the LANEX screen corresponds to electrons and not to photons.  c , Image obtainedwith a magnetic field and a higher plasma density (  n   e ¼ 2  £  10 19 cm 2 3  ). Thiselectron beam has a much larger divergence and a 100% energy spread with fewelectrons above 100MeV. letters to nature NATURE|VOL 431|30 SEPTEMBER 2004|www.nature.com/nature 542  ©    200 4 Nature PublishingGroup  edge, this is the smallest divergence ever measured for a beamemerging from a plasma accelerator. Figure 2b shows the deviationof the beam when a magnetic field is applied. The image shows anarrow peak around 170MeV, indicating efficient monoenergeticacceleration. For comparison, Fig. 2c shows an image obtained athigher electron density in the plasma ( n e ¼ 2  £  10 19 cm 2 3 ). Here,electronsarerandomly accelerated toallenergiesand thenumberof high-energy electrons is low. In addition, the beam divergenceis much larger than in Fig. 2b. Figure 3 shows an electronspectrum after deconvolution. The distribution is clearly quasi-monoenergetic and peaks at 170MeV, with a 24% energy spread(corresponding to the spectrometer resolution).Finally,thechargecontainedinthisbeamcanbeinferredusinganintegrating current transformer: the whole beam contains2 ^ 0.5nC, and the charge at 170 ^ 20MeV is 0.5 ^ 0.2nC.From the above, we can deduce that the electron beam energy was100mJ. Thus, the energy conversion from the laser to the electronbeam was 10%.Experimentally,thisregimecouldbereachedinanarrow rangeof parameters: stretching the pulse duration above 50fs was sufficientto lose the peaked energy distribution. Similarly, when the electrondensity was increased from 6  £  10 18 cm 2 3 to 7.5  £  10 18 cm 2 3 ,the energy distribution became a broad plateau, similar toprevious results 5 . Above 10 19 cm 2 3 , the electron distribution wasmaxwellian-like with very few electrons accelerated at high energy.Below 6  £  10 18 cm 2 3 , the number of accelerated electronsdecreased dramatically, although the distribution was still mono-energetic. The evolution of electron spectra with experimentalparameters indicates that using laser pulses shorter than the plasmaperiod is beneficial for high-quality and monoenergetic electronacceleration.To reach a deeper understanding of the experiment, we have runthree-dimensional (3D) particle-in-cell (PIC) simulations using thecode Virtual Laser Plasma Laboratory  21 . The simulation results areshown in Fig. 4a–c. The simulation suggests that our experimentalresults can be explained by the following scenario. (1) At thebeginning of the simulation, the laser pulse length (9 m m) is nearly resonant with the plasma wave ( l  p ¼ 13.6 m m); but its diameter(21 m m . l  p ) is larger than the matched diameter. (2) As the pulsepropagates in the plateau region of the gas jet, it self-focuses andundergoes longitudinal compression by plasma waves (Fig. 4a).Thisdecreasestheeffectiveradiusofthelaserpulseandincreasesthe Figure 3  Experimental and simulated electron spectra. Blue line with crosses, electronspectrum corresponding to Fig. 2b, after deconvolution. Dashed line, estimation of thebackground level. Red horizontal error bars, resolution of the spectrometer. Green line,electronspectrumobtainedfrom3DPICsimulations.d N   /d E   isthenumberofelectronsperMeV (  E   is the electron energy in MeV). Figure 4  3D PIC simulation results.  a ,  b , Distributions of laser intensity (  a  ) and electrondensity(  b  )inthe x  – z  plane,whichisperpendiculartothepolarizationdirectionandpassesthrough the laser axis. The laser pulse runs from left to right, and has propagated 2mm inthe plasma. The bubble structure is clearly visible. The laser pushes the electron fluidforward at the bubble head and creates a density compression there. Behind the laser wesee the cavitated region with nearly zero electron density. The radially expelled electronsflow along the cavity boundary and collide at the X-point at the bubble base. Someelectrons are trapped and accelerated in the bubble. The beam of accelerated electronsisseen as the black rod in  b . These electrons are propagating behind the laser pulse (  a  ) andare not disturbed by the laser field.  c , Electron phase space density  f   (  x  , g  ) in arbitraryunits.  g  is the relativistic factor of the electron:  g ¼ (1 2 v   2  /  c   2  ) 2 1/2 , and  v   is theelectron velocity. We see that the electrons have dephased and have self-bunched in thephase space around  g .. 350. This self-bunching results in the mono-energetic peak inthe energy spectrum (Fig. 3). The red horizontal dashed lines indicate the location of themono-energetic peak in the phase space. letters to nature NATURE|VOL 431|30 SEPTEMBER 2004|www.nature.com/nature  543  ©    200 4 Nature PublishingGroup  laser intensity by one order of magnitude. (3) This compressedlaser pulse is now resonant with the plasma wave and it drives ahighly nonlinear wakefield (Fig. 4b): the laser ponderomotivepotential expels the plasma electrons radially and leaves a cavitatedregion behind (this is referred to as the ‘cavitation’ or ‘blow-out’regime). In this regime, the 3D structure of the wakefieldresembles a plasma bubble 11 . (4) As the electron density at thewalls of the bubble becomes large, wave-breaking occurs andelectrons are injected and accelerated inside the bubble. (5) Asthe number of trapped electrons increases, the bubble elongates.Its effective group velocity decreases, and electrons start to dephasewith respect to the accelerating field. This dephasing causeselectron self-bunching in the phase space (Fig. 4c). This self-bunching results in the monoenergetic peak in the energy spectrum(Fig. 3).Simulations also show that the quality of the electron beam ishigher when trapped electrons do not interact with the laser field. If this wereto occur, the laser field would cause the electrons to scatterin phase space, degrading the low divergence as well as the mono-energetic distribution. This argument could explain why higher-quality beams are obtained experimentally for shorter pulses andlower electron densities.Figure 4a shows that the self-focused and compressed laser pulsestands in front of the trapped electrons (Fig. 4b), leaving themalmost undisturbed 5,11 . The electron energy spectrum obtainedfrom the simulations is shown in Fig. 3: it peaks at 175 ^ 25MeV,inagreementwiththeexperiment.Thedivergenceof10mradisalsoin agreement with experiments. Simulations also indicate that theelectron bunch duration is less than 30fs (here, the term ‘bunch’refers to the fact that electrons are created in short bursts). Becausethe electron distribution is quasi-monoenergetic, the bunch willstay short upon propagation: considering a 24% energy spread at170MeV, we can show that the bunch stretches by only 50fsm 2 1 asit propagates.Another important point is the apparent robustness of the ‘blow-out’regime.Theinitiallaserparameters — forexample,thefocalspotradius and intensity  — were far from the final values in the bubble(Fig.4).Yetself-focusingledtocompressionofthelaserpulseandtothe formation of an electron cavity. The energy of 1J for a 30fs laserpulse, as used in the experiment, seems to be close to the thresholdfor this regime. Simulations 11 suggest that with more laser energy and shorter pulses, the blow-out regime and the formation of thebubble will lead to the acceleration of monoenergetic beams athigher energies and higher charges.Our experimental results and 3D PIC simulations indicate that itis possible to generate a monoenergetic electron beam by carefully selecting laser and plasma parameters. The bunch duration( , 50fs), along with the present improvement in the charge (nC)and the quality of the electron beam (monoenergetic spectrum, low divergence),reinforcetherelevanceofplasma-basedacceleratorsformany applications (such as high-resolution radiography for non-destructive material inspection, radiotherapy, ultrafast chemistry,radiobiology and material science). With the rapid progress of laserscience, we expect that it will soon become possible to generatecompact, monoenergetic and high-quality electron beams with atunable energy range at a reasonable cost. Such a source wouldbe perfectly adapted as an injector for future GeV laser–plasmaaccelerator schemes. It would also be relevant for generatingultrashort X-ray sources, using undulators or lasers via Thomsonscattering.  A Methods Laser Thisnew regimewasreachedbyusingtheultrashortandultraintenselaserpulsegeneratedin a titanium-doped sapphire, chirped pulse amplification laser system. The laser pulsehad a 33 ^ 2fs duration (FWHM), and contained 1J of laser energy at centralwavelength820nm.Itwasfocusedontotheedgeofa3-mm-longsupersonicheliumgasjetusinga  f  /18off-axis parabola. The diffraction-limited focal spot had a diameter of   r  0 ¼ 21 m m atFWHM, producing avacuum-focused laser intensityof   I  ¼ 3.2  £  10 18 Wcm 2 2 , for whichthe corresponding normalized potential vector is  a 0 ¼ eA /( mc  2 ) ¼ 1.3 (  A  is the laservector potential,  e  and  m  are respectively the charge and mass of the electron). For thesehigh laser intensities, the helium gas was fully ionized by the foot of the laser pulse andionization did not play a role in the interaction. Electron diagnostics Electron detection was achieved using a LANEX phosphor screen, placed 25cm afterthe gas jet. As electrons passed through the screen, energy was deposited and re-emitted into visible photons which were then imaged onto a 16bit charge-coupleddevice (CCD) camera. For energy distribution measurements, a 0.45T, 5-cm-longpermanent magnet was inserted between the gas jet and the LANEX screen. TheLANEX screen was protected by a 100- m m-thick aluminium foil in order to avoiddirect exposure to the laser light. For deconvolution of the images obtained with theLANEX screen, electron deviation in the magnetic field has been considered as well asthe electron stopping power inside the LANEX screen. The resolution (red error bar inFig. 3) is limited by the electron beam spatial quality and by the dispersing power of the magnet. This gives a resolution of respectively 32MeV and 12MeV for 170MeVand 100MeV energies. Above 200MeV, the resolution quickly degrades. The charge of the electron beam was measured using an integrating current transformer placed 30cmbehind the LANEX screen. It allowed us to measure the total charge of the beam whenno magnetic field was applied, and the charge above 100MeV when the magnetic fieldwas applied. PIC simulations The simulation parameters corresponded to the optimal experimental case: the plasmaelectron density was  n e ¼ 6  £  10 18 cm 2 3 , the laser pulse duration was 30fs and the initiallaser spot size 21 m m FWHM. The laser pulse was assumed to be a perfect gaussiancontaining 1J of energy. The plasma profile was chosen to fit the experimental density profile of the gas jet. Received 5 July; accepted 25 August 2004; doi:10.1038/nature02963. 1. Tajima, T. & Dawson, J. M. Laser electron accelerator.  Phys. Rev. Lett.  43,  267–270 (1979).2. Modena, A.  et al.  Electron acceleration from the breaking of relativistic plasma waves.  Nature  337, 606–608 (1995).3. Umstadter, D.,Chen,S.-Y.,Maksimchuk,A.,Mourou,G.& Wagner,R.Nonlinearopticsin relativisticplasmas and laser wake field acceleration of electrons.  Science  273,  472–475 (1996).4. Moore, C. I.  et al.  Electron trapping in self-modulated laser wakefields by Raman backscatter.  Phys.Rev. Lett.  79,  3909–3912 (1997).5. Malka,V.  et al.  Electronacceleration byawake field forcedbyan intense ultrashort laser pulse.  Science 298,  1596–1600 (2002).6. Gahn, C.  et al.  Multi-MeVelectron beam generation by direct laser acceleration in high-density plasma channels.  Phys. Rev. Lett.  83,  4772–4775 (1999).7. Malka, V.  et al.  Characterization of electron beams produced by ultrashort (30fs) laser pulses.  Phys.Plasmas  8,  2605–2608 (2001).8. Kitagawa, Y.  et al.  Beat-wave excitation of plasmawaveand observation of accelerated electrons.  Phys.Rev. Lett.  68,  48–51 (1992).9. Everett, M.  et al.  Trapped electron acceleration by a laser-driven relativistic plasma wave.  Nature  368, 527–529 (1994).10. Amiranoff, F.  et al.  Observation of laser wakefield acceleration of electrons.  Phys. Rev. Lett.  81, 995–998 (1998).11. Pukhov, A. & Meyer-ter-Vehn, J. Laser wake field acceleration: the highly non-linear broken-waveregime.  Appl. Phys. B  74,  355–361 (2002).12. Clayton, C. E., Joshi, C., Darrow, C. & Umstadter, D. Relativistic plasma-wave excitation by collinearoptical mixing.  Phys. Rev. Lett.  54,  2343–2346 (1985).13. Amiranoff, F.  et al.  Observation of modulational instability in Nd-laser beat-wave experiments.  Phys.Rev. Lett.  68,  3710–3713 (1992).14. Andreev, N. E., Gorbunov, L. M., Kirsanov, V. I., Pogosova, A. A. & Ramazashvili, R. R. Resonantexcitation of wakefields by a laser pulse in a plasma.  JETP Lett.  55,  571–574 (1992).15. Sprangle, P., Esarey,E., Krall,J. &Joyce,G. Propagationandguiding of intenselaserpulsesinplasmas. Phys. Rev. Lett.  69,  2200–2203 (1992).16. Antonsen, T. M. & Mora, P. Self-focusing and Raman scattering of laser pulses in tenuous plasmas. Phys. Rev. Lett.  69,  2204–2207 (1992).17. Leemans, W. P.  et al.  Electron-yield enhancement in a laser-wakefield accelerator driven by asymmetric laser pulses.  Phys. Rev. Lett.  89,  174802 (2002).18. Pukhov,A., Sheng,Z.-M.& Meyer-ter-Vehn,J.Particle accelerationin relativisticlaserchannels. Phys.Plasmas  6,  2847–2854 (1999).19. Strickland, D. & Mourou, G. Compression of amplified chirped optical pulses.  Opt. Commun.  56, 219–221 (1985).20. Pittman, M.  et al.  Design and characterization of a near-diffraction-limited femtosecond 100-TW10-Hz high-intensity laser system.  Appl. Phys. B  74,  529–535 (2002).21. Pukhov,A.J.Three-dimensionalelectromagneticrelativisticparticle-in-cellcodeVLPL(VirtualLaserPlasma Lab).  J. Plasma Phys.  61,  425–433 (1999). Acknowledgements  We acknowledge support from the European Community ResearchInfrastructure Activity under the FP6 “Structuring the European Research Area” programme(CARE) and from the German Scientific Council (DFG). Competing interests statement  The authors declare that they have no competing financialinterests. Correspondence  and requests for materials should be addressed to V.M. (victor.malka@ensta.fr). letters to nature NATURE|VOL 431|30 SEPTEMBER 2004|www.nature.com/nature 544  ©    200 4 Nature PublishingGroup View publication statsView publication stats
Related Search
Related Docs
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks