The UCLA/NICADD Plasma Density Transition Traing Exeriment M.C. Thomson, J.B. Rosenzweig, G. Travish UCLA Deartment of Physics and Astronomy N. Barov Northern Illinois University Deartment of Physics H. Edwards, P. Piot, J. Santucci, R. Tikholav Fermi National Accelerator Laboratory PBPL DoE Review - May 2004
Why Build Plasma Electron Beam Sources? Better Emittance Higher Brightness Larger Gradients Thermionic Gun: Cathode Size ~ 10-2 m Emittance ~ 100 mm-mrad Current ~ 10 A/cm 2 Brightness ~ 10 9 A/(m-rad) 2 Gradient ~ 1 MV/m RF Photoinjector: Cathode Size ~ 10-3 m Emittance ~ 1 mm-mrad Current Density ~ 10 3 A/cm 2 Brightness ~ 10 12 A/(m-rad) 2 Gradient ~ 100 MV/m 10 17 cm -3 Plasma : Cathode Size ~ 10-5 m Emittance ~ 0.1 mm-mrad Current Density ~ 10 7 A/cm 2 Brightness ~ 10 14 A/(m-rad) 2 Gradient ~ 10 GV/m
The Problem of Injection All lasma accelerator schemes share two critical scalings: E max n Structure Size k 1 = c ω 1 This scaling makes the injection of charge into lasma waves difficult because of the technical roblems of ulse creation and timing at sub-s levels. n Externally Injected Witness Beam 88 sec 0.003 3G 0.002 2G 0.001 1G Tyical Frequency = 2856 MHz 0 ev/m -0.001-1G -0.002-2G -0.003 0.002 High Gradient RF Structures: 7 sec 0.000 0.004 0.006 0.008-3G 0.010 Plasma Wake Field Accelerator Oerated in the Non-Linear Regime at 5 x 1013 cm-3
Traing: The Injection of Plasma Electrons The alternative to injecting external charge, e.g. from a cathode, is creating a situation in which the lasma wave can tra and accelerate electrons directly out of the lasma. Automatic Traing: Gentle Convention Wave Breaking Disadvantage - High Energy Sread S. Bulanov, et al., Phys. Rev. E 58, R5257 (1998) Stimulated Traing: Multile Colliding Laser Pulses Disadvantage - Comlex Timing Requires E. Esarey, et al., Phys. Rev. Lett. 79, 2682 (1997) D. Umstadter, et al., Phys. Rev. Lett. 76, 2073 (1996) Ideally we would like to combine the simlicity of automatic traing with the beam quality of the stimulated methods...
Plasma Density Transition Traing Plasma Density Transition Traing is a self-traing scenario that uses the raid change in the wake field wavelength at a stee dro in the lasma density to dehase lasma elections into an accelerating hase of the wake. Transition Traing Fundamentals: Automatic Injection of Substantial Charge (~100 C) Into accelerating Phase Oerates in PWFA Blow Out regime where n beam > n lasma (underdense condition) Traing Condition: k L Transition < 1 where k -1 is the lasma skin deth k -1 = c/ω UCLA Work On Transition Traing: Concet Proosed - H. Suk, et al., Phys. Rev. Lett. 86, 1011 (2001) Traing Exeriment Proosed for the Netune Lab at UCLA - M.C.Thomson, et al., Proceedings PAC 2001, age 4014 (2001) Imroved analysis, Udated UCLA/NICADD Proosal - M.C. Thomson, et al., Proceedings PAC 2003, age 1870 (2003) Analysis of Scaling and Merits as a High Brightness Source - M.C. Thomson, et Al., Phys. Rev. STAB 7, 011301(2004)
Traing Regimes Strong Blowout Plasma: 5 x 10 13 cm -3 /3.5 x 10 13 cm -3 Beam: 1.2 x 10 14 cm -3 (63 nc) Weak Blowout Plasma: 2 x 10 13 cm -3 /3.5 x 10 12 cm -3 Beam: 4 x 10 13 cm -3 (5.9 nc)
Scaling of the Plasma Density Scaling of the transition traing system to higher Plasma Density ( n h ) requires that all charge densities be increased by the ratio n h / n and all lengths be decreased by the ratio: h h h h n n n n k k = = = 1 1 1 1 λ λ As the lasma density is increased, the catured beam arameters change according to the following scaling laws: n Q n n E λ λ λ = 3 max Constant = Constant Qc I λ λ λ ε n I B 2 2 1 λ ε
Transition Traing as a High Brightness Source Result of 2D PIC Simulations Examining the Scaling of a Weak Blowout Scenario Peak Density 2 x 10 13 cm -3 2 x 10 15 cm -3 2 x 10 17 cm -3 σ t, Diver 1.5 sec 150 fsec 15 fsec Q Driver 10 nc 1 nc 100 C σ t, Catured 2.7sec 270 fsec 28 fsec Q Catured 1.2 nc 120 C 12 C I Peak,Catured 163 Am 166 Am 166 Am ε x, normalized,catured 57 mm-mrad 5.9 mm-mrad 0.6 mm-mrad B normalized,catured 5 x 10 10 5 x 10 12 5 x 10 14 LCLS Injector Sec: ε x, normalized = 0.6 mm-mmrad I = 100 am B normalized = 2.8 x 10 14 Definitions: ε B 2 x, normalized x γ v = c x, normalized = = x ε 2 I 2 x 2 x, normalized x x 2
The Exerimental Plan Criteria for the first exeriment: Pre-Existing Drive Beam Parameters Low Plasma Density Easiest Possible Density Modification System
Argon Pulse Discharge Plasma Source The Plasma Source was redesigned, rebuilt, and tested During 2002-2003. Source shied to Fermilab Nov 2003 Originally develoed for a underdense lasma lens exeriment at Netune: H. Suk, C.E. Clayton, G. Hairaetian, et al., PAC Proceedings (1999). 3708
Source Testing Highly Reliable Plasma Production 2 ms, 40 kw eak ower ulses at 1 Hz Peak Density > 3 x 10 13 cm -3 Density (cm -3 ) 3.5x10 13 3.0x10 13 2.5x10 13 2.0x10 13 1.5x10 13 1.0x10 13 5.0x10 12 0.0-5 -4-3 -2-1 0 1 2 3 4 5 z (cm)
Tailoring Plasma Density Profiles The directionality of the lasma flow in the ulsed discharge source allows the density to be maniulated by lacing a erforated metal sheet between the source region and the interaction oint. If the density modifying obstruction is laced very close the drive beam ath the transition in density should be shar enough to show traing. Beam Path High Density Region 250 200 150 0 mm 0.5 mm 1 mm Plasma Flow Low Density Region C 100 50 0 1.5 mm 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Screen z x k P region 1 L Transition Deendence of the amount of charge catured on finite length of the lasma transition (MAGIC Simulation).
Density Screening Density Behind the Screen is Determined by the Oen Area of the screen. PIC Simulations and exeriments are consistent with the equation exected to govern the transition size: L Trans = 2x 150 L Transition < 1 mm 2 x 10 13 cm -3 Transitions short enough to exhibit traing have been roduced. Pre-transition density roll off not yet fully understood. Plasma Density (A.U.) 100 50 0 0.00 0.01 0.02 0.03 0.04 z distance (m) Simulation of a Screen with 500 µm holes and 50% oen area. Density is Integrated over a 400 µm band from 100 500 µm away from the foil.
Transition Measurements
Screens in 3-D In order to satisfy the traing condition the beam must ass within k -1 /2 of the screen. For a simle screen this means the beam must be within one k -1 /2 of the screen for the entire traing and acceleration rocess. Wakefield interactions with the screen can destroy the traing. By adding a solid metal baffle to the screen, transition sharness can be reserved while limiting the beam interaction with the screen. RACMST Artist s concetion of artial blocking of wake articles by the baffle. Preliminary simulations of baffle loses using MAGIC 3D.
Other Problems & Solutions Loss Mechanisms Other effects that may lead to traed charge loss: Beam ath bending during traing. Growth of the transition length across the wake. Longer than exected driver bunches. None of these effects should kill the traing entirely, but simulating their effects correctly is difficult. Charge Scaling A relatively small increase in driver beam charge can comensate for significant losses.
The First Exerimental Run Took lace at the Fermilab NICADD Photoinjector Laboratory (FNPL) from Jan 2004 to May 2004 FNPL Beamline: L-band Photoinjector Suerconducting 9- Cell Cavity Magnetic Comressor
Underdense Plasma Lens Plasma Density 1.3 x 10 12 cm -3 Plasma Thickness 2 cm Beam Charge 4 nc Beam Duration (FWHM) 30 sec Initial Beam Radius (σ r ) 400 µm Beam Density 2.6 x 10 12 cm -3 Netune Ex. H. Suk Proc. PAC 99,. 3708 Rudimentary Plasma Focusing Observed at FNPL Feb 2003: Beam Charge ~ 8 nc Beam Density ~ 2 x 10 13 cm -3 Beam Radius (σ r ) ~ 400 µm Beam Length (FWHM) ~ 15 s Screen ~ 30 cm from lasma Plasma Off Plasma On
Achieved Beam Parameters Problem Origins: UV Laser Comressor Energy Sread Isolation Foil Charge Transort FNPL Drive Laser Virtual Cathode
Drive Beam Deceleration Deceleration of the drive beam is direct evidence of resence of wake fields. Simulations indicate that assing through the traing exeriment lasma should roduce 2 3 MeV of drive beam deceleration. Max Deceleration of ~ 1 MeV observed. Kinetic Energy (ev) 14.0M 12.0M 2.0M 0.0 Drive Beam Traed Beam 0.070 0.075 0.080 0.085 0.090 z (m) Drive Beam Deceleration Exected from Simulations of The Design Case 11.58 MeV With Plasma 12.4 MeV 12.4 Just MeV Driver 13.3 MeV
The Search For Traed Electrons Possibility of small quantities of traed during good shots. Cylindrical lens added to the sectrometer light collection otics. Demonstrated sensitivity to < 100 C using the drive beam. Low energy signal observed, but it did not correlate to the resence of the lasma. Cylindrical lens added for enhanced light collection 0.87 MeV 1.4 MeV 2.2 MeV Weak Low Energy Signal
Future Plans Imroved Traing Exeriment: Requirements: Understanding and imrovement of the beam quality and charge transort. Construction of a new sectrometer exit ort. Sharer Transitions. Imroved lasma and beam diagnostics. Underdense Plasma Lens Exeriment: Requirements: Simulation studies to udate the exeriment to FNPL arameters Minor modifications to the lasma source. The 2 nd exerimental run is scheduled to begin late Setember 2004. The time allocation for these two activities has yet to be determined.