Proton Driven Plasma Wakefield Acceleration

Transcription

1 Proton Driven Plasma Wakefield Acceleration Guoxing Xia Max-Planck Planck-Institute für f r Physik March 30, 2010

2 Large Hadron Collider-collision today! LHC: the world biggest accelerator, both in energy and size Grand start-up and perfect function at injection energy in September 2008 A electrical fault halted the machine running First collisions in late 2009 (2.36 TeV) 7 TeV collisions today at 13:06 CEST 4/1/2010 2

3 Complex machines 4/1/2010 3

4 International Linear Collider-ILC The next big thing. After LHC, a Linear Collider of over 30 km length, will probably be needed. 4/1/2010 4

5 Complex machines Two linear accelerators, with tiny intense beams of electrons and positrons colliding head-on on-head Total length ~ 30 km long (comparable scale to LHC) COM energy = 500 GeV, upgradeable to 1 TeV 4/1/2010 5

6 Livingston plot The progress in accelerator energy versus year since the first cyclotron in 1930s. The progress of plasma accelerators is illustrated by representative experiments (circles) T. Katsouleas, Phys. Plasmas 13, (2006) 4/1/2010 6

7 Outline Motivation Plasma wakefield acceleration-pwfa Proton driven PWFA Short proton bunch production Demonstration experiment at CERN Summary 4/1/2010 7

8 Motivation Particle physics enters terascale era LHC starts collisions with half energies just today (30/03/2010)! ILC is proposed as the next generation lepton collider However, the cost and size of these machines reach the limit New ideas to minimize the machine are highly demanding Plasma wakefield accelerator hold promise to reduce the machines scale Plasma, already in ionized state, can sustain very large electric c field than the conventional accelerating structures For a fixed energy machine, it can reduce the machine scale significantly ificantly! The plasma wakefield acceleration driven by laser pulse (LWFA) and a electron beam (PWFA) has already achieved remarkable progress; > 10~100 GeV/m acceleration gradients have been demonstrated in laboratory experiments. Since high energy proton beam is available now, investigation of the possibility of proton driven PWFA is possible 4/1/2010 8

9 Plasma Wakefield Acceleration (PWFA) I) Generate homogeneous plasma channel Gas Plasma II) Send dense electron beam towards plasma Ionization of gas via: Laser Beam RF = ion = electron Beam density n b Gas density n 0 4/1/2010 9

10 PWFA III) Excite plasma wakefields Electrons are expelled r z Ion channel ω = p λ 1mm p n p e ε m cm n 0 3 Space charge force of beam ejects all plasma electrons promptly along radial trajectories Pure ion channel is left: Ion focused regime, underdense plasma 4/1/

11 PWFA Electron motion solved with... driving force: Space charge of drive beam displaces plasma electrons. restoring force: Plasma ions exert restoring force Space charge oscillations (Harmonic oscillator) 1/ 2 2 m c = p p ε 0 e [ / ] 1/ 2 3 m n 100 n (cm ) E V Longitudinal fields can accelerate and decelerate! Approximately mm-wave length! 4/1/

12 PWFA Plasma structures are also super strong quadrupoles! (many thousand T/m) Plasma ions move relatively little W r r 2 T n0 = 2π n0 e = 960π 14 3 m 10 cm Constant focusing gradient need to handle acceleration and focusing! acceleration and focusing! 4/1/

13 PWFA Physical mechanism of the Plasma Wakefield Accelerator for previous single bunch experiments in the FFTB (top) and the two bunch case proposed for FACET (bottom) 4/1/

14 PWFA Electron beam (beam energy 42GeV, bunch length 50 fs, bunch charge 2.9 nc) Plasma (length 85 cm, density 2.7e17 cm-3) Max. energy gain 43 GeV (85 cm column) = 52 GeV/m! 29 GeV (113 cm column) Energy spectrum of the electrons in the 4/1/2010 Blumenfeld et al., Nature 445 (2007) GeV range as observed in plane 2 14

15 PWFA-based linear collider* *M. Hogan et al., Proceedings of PAC09 (TU1GRI01) 4/1/2010 *A. Seryi et al., Proceedings of PAC09 (WE6PFP081) 15

16 Energy gain E N 0.6 = 240(MeV/m) σ z (mm) 2 In the linear regime, the maximum achievable field scales as the bunch charge divided by bunch length squared, high intensity and short bunch length are required for high field production. σ z Transformer ratio limit * R E witness max = drive Emax 2 for longitudinal symmetric bunch *R.D. Ruth et al, SLAC PUB /1/

17 Existing proton machines HERA TEVATRON LHC Circumference [km] Maximum energy [TeV[ TeV] Energy spread [10-3 ] Bunch length [cm] Transverse emit. [10-9 π m rad] Particles per bunch [10 10 ] /1/

18 high energy proton beam as driver Hugh energy stored in current proton machines like Tevatron, HERA, SPS and LHC For example, the SPS/LHC beam carries significant stored energy for driving plasma waves SPS (450 GeV,, 1.3e11 p/bunch) ~ 10 kj LHC (1 TeV, 1.15e11 p/bunch) ~ 20 kj LHC (7 TeV, 1.15e11 p/bunch) ~ 140 kj SLAC (50 GeV,, 2e10 e-/bunch) e ~ 0.1 kj However, the current proton bunches are quite longer to use as driver directly. Need much effort to compress the beam How to couple the energy of driver to the plasma and the witness beam efficiently? 4/1/

19 Features of proton driven PWFA* Positive-charged driver makes the plasma electrons flow in the vicinity of the driver nonlinear response beam linear response Heavier mass means lower relativistic factor wave phase slippage 4/1/

20 Proton driven PWFA* In the linear regime of PWFA, the maximum achievable gradient scales as N/σ z2, therefore we need high current and short bunch Transformer ratio (the gained energy of witness beam / the energy of driver beam) is limited to 2 for longitudinal symmetric driven bunches Our idea is to use the existing high energy proton beam to drive plasma wakefield and then accelerate the electron beam to high energies* 2D and 3D Particle-In In-Cell simulations have given us very promising results One of the biggest hurdles is to produce the short proton beam,, in the scale of 100 micron. * A. Caldwell et al., Nature Physics 5, 363 (2009) 4/1/

21 Schematics of PDPWA A thin tube containing Li gas is surrounded by quadrupole magnets with alternating polarity. The magnification shows the plasma bubble created by the proton bunch (red). The electron bunch (yellow) undergoing acceleration is located at the back of the bubble. Note that the dimensions are not to scale. 4/1/

22 Parameter settings Drive Beam Symbol Value Protons in drive bunch [10 11 ] N p 1 Proton energy [TeV[ ] E p 1 Initial proton momentum spread σ p /p 0.1 Initial longitudinal spread [µm[ m ] σ z 100 Initial angular spread [mrad[ ] σ θ 0.03 Initial bunch transverse size [mm ] Witness Beam σ X,Y 0.4 Electrons in witness bunch [10 10 ] N e 1.5 Energy of electrons [GeV[ ] E e 10 Plasma Parameters Free electron density [cm -3 ] n p 6 10 Plasma wavelength [mm ] λ p 1.35 External Field Magnetic field gradient [T/m ] 1000 Magnetic length [m ] 0.7 4/1/

23 Simulations 2 D and 3 D Particle-In In-Cell (PIC) codes are employed to simulate the interactions between plasma and beams. a-d, Simulation results for the unloaded (no witness bunch) case (a,b) and in the presence of a witness bunch (c,d). The witness bunch is seen as the black spot in the first wave bucket in d). d) also shows the driving proton bunch at the wavefront (red). e) The on-axis accelerating field of the plasma wave for the unloaded (blue curve) and loaded (red curve) cases. 4/1/

24 Energy gain Simulation results 4/1/

25 Phase space of driver & witness a-h, Snapshots of the combined longitudinal phase space of the driver and the witness bunches (energy versus coordinate) (a d) and corresponding energy spectra (e h). The snapshots are taken at acceleration distances L=0, 150, 300, 450 m. The electrons are shown as blue points and the protons are depicted as red points. 4/1/

26 Energy gain & energy spread 620 GeV beam can be obtained a,b, The mean electron energy in TeV (a) and the r.m.s. variation of the energy in the bunch as a percentage (b) as a function of the distance travelled in the plasma. 4/1/

27 Simulation results Proton beam can indeed to be used as the drive beam for large amplitude wakefield production Proton driven PWFA can bring a bunch of electrons to the energy frontier in only one stage. An unsolved questions, short beam! 4/1/

28 Short proton bunch production Laser striking thin foil can produce short and low energy proton beam Emittance exchange technique, exchanges the longitudinal emittance to horizontal emittance Fast quads tunning for low momentum compaction factor before extraction in the ring Plasma wakefield beam slicing Fast nanochoppers to get microbeam Conventional magnetic bunch compression* * F.Zimmermann, G.Xia et al., Generation of short proton bunches in the CERN accelerator complex, Proceedings of PAC09 (FR5RFP004) 4/1/

29 Bunch compression σ σ zf zi ( dv ( dv RF RF / dz) α / dz) α i f f i Emittance exchange Schematics of fast bunch rotation in storage ring Nanochopper gets the microbunch F.Zimmermann, G. Xia et al., Proceedings of PAC09, (FR5RFP004) 4/1/

30 Magnetic bunch compression Beam compression can be achieved: (1) by introducing an energy-position correlation along the bunch with an RF section at zero- crossing of voltage (2) and passing beam through a region where path length is energy dependent: this is generated by bending magnets to create dispersive e regions. Tail ΔE/E (advance) -z Head (delay) lower energy trajectory center energy trajectory higher energy trajectory To compress a bunch longitudinally, trajectory in dispersive region must be shorter for tail of the bunch than it is for the head. 4/1/

31 Bunch compressor design* For a thousand-fold compression, one stage compression looks infeasible We expect that the ring could compress the bunch by a factor of 10 and the rest will be realized via magnetic chicane z( s δ ( s 2 2 ) 1 = ) 0 R R 65 0 z( s 1 δ ( s 0 0 ) 1+ R = ) R R56 R65 = R R 65 R ΔE = = = z z E R 56 1 ev E δ 0 2 = 2 L + ΔL B B 3 56 θ z( s δ ( s ω βc 0 0 ) ) Value Bunch charge Proton energy [TeV] 1 Initial energy spread [%] 0.01 Initial bunch length [cm] 1.0 Final bunch length [μm] 165 RF frequency [MHz] Average gradient of RF [MV/m] 25 Required RF voltage [MV] 65,000 RF phase [degree] -102 Compression ratio ~60 Momentum compaction (MC) [m] -1.0 Second order of MC [m] 1.5 Bending angle of dipole [rad.] 0.05 Length of dipole [m] 14.3 Drift space between dipoles [m] Total BC length [m] 4131 Final beam energy [GeV] Final energy spread [%] /1/2010 *T. Raubenheimer, SLAC-PUB

32 Phase space of beam 4,0x10-4 0,04 2,0x10-4 0,02 dp/p 0,0 dp/p 0,00-2,0x ,02 a) Initial phase space -4,0x ,04-0,02 0,00 0,02 0,04 z / m 0,04 c) Final phase space -0,04-5,00x ,50x10-4 0,00 2,50x10-4 5,00x10-4 z / m 0,02 dp/p 0,00-0,02 b) Phase space after RF section -0,04-0,04-0,02 0,00 0,02 0,04 z / m 4/1/2010 G. Xia, A. Caldwell et al., Proceedings of PAC09 (FR5RFP011) 32

33 Phase space of beam 0,04 b) c) 0,02 dp/p 0,00 a) -0,02 a) initial phase space b) phase space after RF c) phase space after bunch compressor -0,04-0,04-0,02 0,00 0,02 0,04 z / m G. Xia, A. Caldwell et al., Proceedings of PAC09 (FR5RFP011) 4/1/

34 Bunch length & energy spread bunch length 300 th1 Entries 1000 Mean e-05 RMS energy spread 50 th1y Entries 1000 Mean e-05 RMS Fitting data show that the bunch length of 165 micron and the relative energy spread of 9e-3 can be achieved. Futher optimization is onging. G. Xia et al., Plasma Phys. Control. Fusion (2010), to be published. 4/1/

35 Test experiment at CERN Near-term plan is to use the extracted proton beam from PS or SPS to demonstrate the energy gain of 1 GeV within 5 m of plasma Ultimate goal is to achieve 100 GeV energy gain in 100 m plasma The very high energy electron beam achieved by using the LHC-like high energy beam 4/1/

36 PS beam parameters 2 beamlines available from East hall of PS extraction area. However, the maximum space is limited to 60 m for experiment! Bunch compression, plasma cell, dignostic equiments and energy spectrometer can be accommodated within this space Proton Synchrotron-PS Momentum [GeV/c] 24 Maximum protons/bunch [10 11 ] 1.3 rms longitudinal emittance [evs] 0.03 rms bunch length [cm] 20 Relative rms energy spread [10-4 ] 5 rms transverse emittance [um] 3.0 Bunch spacing [ns] 25 # bunches / cycle 72 Cycle time [s] 3.6 4/1/

37 A trial for bunch compression 0,12 dp/p 0,10 0,08 0,06 0,04 0,02 0,00-0,02-0,04-0,06-0,08 pop experiment RF: -98 degree,400 cavities Quads: k1=0.05 total length of BC:605 m final energy GeV ,6-0,4-0,2 0,0 0,2 0,4 0,6 0,8 z / m A. Caldwell, G.Xia et al., Plasma wakefield excitation with a 24 GeV proton beam, Plasma Phys. Control. Fusion 52 (2010). 4/1/

38 Other RF systems 0,010 0,008 0,006 0,004 0,002 dp/p 0,000-0,002-0,004-0,006-0,008-0,010 comp5.mad8+comp5.xsif 8 RF cavities, 704MHz, 25MV/m total length of BC: 48m ct ,6-0,4-0,2 0,0 0,2 0,4 0,6 th1 z/m y Entries 1000 Mean RMS th1y Entries 1000 Mean RMS Bunchlets production /1/

39 Long bunch modulation by wakefield 4/1/2010 A. Pukhov 39

40 Beam modulation by wakefield half-cut beam full beam (smaller grid size) while the half cut bunch starts with a larger field, in the end the unseeded bunch has the larger gradient. The argument is that the larger number of protons makes the difference in the end. 4/1/2010 K. Lotov 40

41 Simulation of PS beam N. Kumar, A. Pukhov, K. Lotov, A. Caldwell, G. Xia, Selfmodulation instability of a long proton bunch in a plasma, to be submitted. 4/1/

42 SPS beam for test experiment Super Proton Synchrotron-SPS Momentum [GeV/c] 450 Maximum protons/bunch [10 11 ] 1.15 rms bunch length [cm] 12 rms energy spread [10-4 ] 2.8 rms transverse emittance [µm] 3.5 dipole Bunch spacing [ns] 25 beam dump steep tunnel 4/1/ I.Efthymiopoulos, EN/MEF - CERN

43 SPS beam for test experiment large space available for SPS tunnel The SPS/LHC proton bunch has excellent properties: Very stiff beam: can drive plasma without too much beam deterioration. Well controlled and maintained (for LHC, CNGS, HiRadMat, ). Variable in intensity (2e9-1.6e11) and emittance. Carries significant stored energy for driving plasma waves: If we can couple this energy into the plasma, a new plasma wakefield acceleration frontier can be opened. The issue is how to couple the proton energy to the plasma: CERN proton bunches are very long (120 mm), compared to the electron bunches used at SLAC (< 0.1 mm). Plasma wavelength is at the 1 mm scale. How do we couple the p bunch into the plasma? 4/1/

44 SPS beam compression 0,015 Two-stage bunch compression Firstly compress the beam from ring by a factor of 100 The rest compression realized by chicane 100 micron beam can be achieved within 300 m space σ σ zf zi bunch length ( dv ( dv RF RF / dz) α / dz) i f f α i th1 Entries 1000 Mean e-05 RMS dp/p 0,010 0,005 0,000-0,005-0, ,003-0,002-0,001 0,000 0,001 0,002 0,003 energy spread z/m sps100.mad8 sps100.xsif K1=0.05 Length of BC 262 m th1y Entries 1000 Mean RMS /1/

45 SPS beam compression 0,03 Intially compress by a factor 10 Magetic chicane will compress further 300 micron bunch can be achieved dp/p 0,02 0,01 0,00-0,01-0,02-0,03 SPS 1 cm initial beam (sps200.mad8) RF 11.4 GHz Chicane Lb=15m, 0.05 rad total length ,03-0,02-0,01 0,00 0,01 0,02 0,03 bunch length th1 Entries 1000 energy spread z/m th1y Entries Mean e-06 RMS Mean RMS /1/

46 SPS beam as driver K. Lotov 4/1/

47 SPS beam as driver A. Pukhov 4/1/

48 PPA09 Workshop-kick kick off meeting Participants from BINP,CERN,DESY, Duesseldorf Uni. IPP, KIT, MPP, RAL,SLAC, UCLA, USC University of College London, CI, JAI show interests recently 4/1/

49 SPS beamline Splitter Horizontal 42 mrad B6 B4 80 mrad HALL TT5 SCHEMATIC LAYOUT OF H3 Lau Gatignon, CERN SL EA TT61 42 mrad B3 T1 TCC6 4/1/

50 First experiment R. Assmann 4/1/

51 Experimental Layout Proton bunch Electro optical sampling system Plasma cell Foils for coherent transition radiation (CTR) Dipole magnet 10 T m Modulated proton bunch CTR Beam profile measurement Perhaps streak camera to also get time dependence Dump 4/1/

52 Possible Timeline 4/1/

53 Summary Simulation shows that proton driven plasma wakefield acceleration holds promise to accelerate the electron bunch to beyond 500 GeV in one stage Short high energy proton bunch could be produced via conventional magnetic bunch compression or combined other compression schemes. Demonstration experiment on PDPWA is planning as a future project Further investigations are ongoing on the key issues such as accelerating positron beam, high repetition rate for high luminosity, beam instabilities, mobile ions, etc. 4/1/

54 Thanks for your attention!

55 extra slides 4/1/

56 Tevatron at FNAL- another ideal place to test PDPWA 4/1/

57 PDPWA-based LC Concept for high repetition rate of proton driven plasma wakefield acceleration 3 ring + injectors + recovery V. Yakimenko, BNL, T. Katsouleas, Duke, LAPW09 4/1/

58 Luminosity Will need very small cross section beams for significant luminosity 4/1/

59 Possible problems with Proton Driven PWFA 1. Short bunch required! ee z linear N = 240(MeV/m) 4 10 σ = 100μm, N = 1* σ z (mm) yields 21GeV/m 2 Can such small beams be achieved with protons? Typical proton bunches in high energy accelerators have rms length >20 cm 4/1/

60 Possible problems with Proton Driven PWFA 2. Phase slippage because protons heavy (move more slowly than electrons) δ = πl 1 1 πl M 2 P c 4 λ p γ 1i γ 1 f γ 2i γ 2 f λ p E driver,i E driver, f L 1 E driver,i E driver, f 2 M 2 P c 4 λ p 300 m for E driver,i =1TeV,E driver, f = 0.5TeV,λ =1mm Few hundred meters possible but depends on plasma wavelength 4/1/

61 Possible problems with Proton Driven PWFA 3. Longitudinal growth of driving bunch due to energy spread d =Δv t Δβ L = ( γ γ 2 )L 2 ΔE E M P 2 c 4 E 2 L For d =100μm, L =100m, E =1.TeV, ΔE E = 0.5 Large momentum spread is allowed! 4/1/

62 Possible problems with Proton Driven PWFA 4. Proton interactions with plasma λ = 1 nσ < 1 n(10 23 cm 2 ) n = cm 3 λ =1000 km Only small fraction of protons will interact in plasma cell. What about the window for the plasma cell? Biggest issue identified so far is proton bunch length. 4/1/