CW injector design studies at PITZ for European X-FEL Shankar Lal, Guan Shu, Hamed Shaker, Houjun Qian and Frank Stephan Photo Injector Test facility at DESY in Zeuthen (PITZ)
Outline FLASH overview European X-FEL overview Photo Injector Test facility at DESY in Zeuthen (PITZ) overview CW injector for European X-FEL upgrade CW injector beam dynamics simulations CW - NC- VHF gun RF design CW buncher RF design Summery Page 2
Free-Electron LASer in Hamburg (FLASH) overview DESY Hamburg Constructed in the early 2000s: UVV-FEL @ TTF2 Prototype for European X-FEL User operation of FLASH 1: Summer 2005 (world 1 st X-FEL) User operation of FLASH 2: April 2016 Ref: (1) https://flash.desy.de/, (2) http://accelconf.web.cern.ch/AccelConf/fel2017/papers/mod02.pdf Page 3
European X-FEL overview European XFEL TDR : 2002 DESY Hamburg Construction work started :2009 First lasing SASE1: May 2017 ( 0.9nm) First user run started : Sep 2017 SCRF accelerator :1.7 km Undulators : ~ 200 m Beam transport & instruments ~ 1km Soft to hard x-ray: 0.5-0.05 nm (0.25-25 keV) Ref : (1) http://xfel.desy.de, (2) Winfied Decking, IPAC2017, (3) Matthias Scholz, IPAC2018 Page 4
PITZ: Photocathode gun test facility for FLASH/European XFEL The Photo Injector Test Facility at DESY, Zeuthen (PITZ) focuses on development, testing and optimization of high brightness electron sources for SASE FELs (FLASH & European XFEL) Test-bed for injectors for FLASH and European XFEL Fundamental research in photo injector physics: cavities, cathode, photoemission etc. Application of high brightness beam: plasma acceleration, THz and UED etc. Ref : (1) http://pitz.desy.de/, (2) M. Krasilnikov et al. PRSTAB15, 100701 (2012) Page 5
Electron source for FLASH/European X-FEL L-band (1.3 GHz) Photocathode RF gun Main Solenoid 1.6 cell, normal conducting RF Gun Laser Electric field at cathode : ~60 MV/m RF Feed Beam energy : ~6.5 MeV/c Bucking Pulse length : 650 us (micro bunch spacing 220 ns) Solenoid Electron Bunch length : up to 20ps (variable using laser) Cathode : Cs 2 Te (QE~5-10%) Peak charge: up to 5nC/bunch Average power :6.5 MW x 650us x 10Hz :~42 kW Ref : (1) http://pitz.desy.de/, (2) M. Krasilnikov et al. PRSTAB15, 100701 (2012) Page 6
European X-FEL : Present and Future operation mode Present operation Electron source Up to 27 µA NC pulse gun developed at PITZ Time [sec] Electron source Possible future upgrade: Need CW gun for operation Up to 10 µA Up to 20 µA SCRF gun R&D is ongoing at NC CW RF gun is under design DESY in Hamburg study at PITZ as a backup solution Page 7
European X-FEL : CW gun for future upgrade CW SRF Gun Pulsed NC Gun APEX-1 CW Gun Design parameters for EXFEL for EXFEL for LCLS II (NC) Duty cycle [%] 100 ~0.65 100 Operation frequency [MHz] 1300 1300 186 RF input power [kW] 0.75 ~ 42 ~ 100 Cathode gradient [MV/m] 40 60 19.5 Beam energy at gun exit [MeV] 3 6.1 0.75 NC CW RF GUN SRF GUN Advantages: Advantages: • Mature RF and mechanical technology • Advanced technology, SRF→ intrinsic CW operation • Easy cathode exchange • Potential for high gradients & high beam energy, • APEX experiments indicate high beam performance for better beam performance XFEL, adopted by LCLS-II. Challenges: Challenges: • Integration of cathode in a SC cavity, i.e. cathode • Cathode gradient and beam energy limitation exchange, cathode lifetime, multipacting, cavity • Operation stability with significant power loss contamination etc. Ref: (1) Elmar Vogel, and (2) Guan Shu, Meeting of Hamburg Alliance New Beams and Accelerators , DESY Hamburg, September 2018 Page 8
DESY VHF gun : Goals and Constraints 1. Gun resonant frequency 187 MHz (APEX) (1300/7) is not compatible with XFEL timing system. 217 MHz (1300/6) and 162 MHz (1300/8) are candidates. 2. Cathode gradient: up to 30 MV/m Higher gradient improves beam brightness Breakdown limits: APEX gun tested ~1.9 Kilp w/o breakdown in CW mode 2*Kilp @217 MHz is ~30.4 MV/m 3. Gun power :<100 kW (demonstrated in APEX gun test) 4. Higher beam voltage: >750 kV Page 9
VHF gun concept: LBNL Inspired by ERL injector design: DC gun VHF RF gun Quasi DC gun beam dynamics: • Beam bunch length: ~60ps, <4 deg • Photoemission phase: ~90 degree Low power dissipation: • Lower thermal power density • 90 kW, <30 W/cm 2 , <70 deg C Ultra high vacuum • Vacuum slots around cavity wall • Vacuum ~10 -10 -10 -11 torr Size compared to Pillbox cavity • Compact cavity size • Enhanced cathode field Ref : F. Sannibale et al . Phys. Rev. ST Accel. Beams 15, 103501 (2012) Page 10
29 cm DESY VHF gun: RF simulation results EXFEL Peak H Parameter APEX Unit Peak E Mode 1 Mode 2 Operation mode CW CW CW 68 cm Frequency 186 217 217 MHz Voltage 750 864 690 kV Cathode gradient 19.47 30.0 24.0 MV/m E field H field Intrinsic quality 30900 32160 32160 factor, Q0 Shunt impedance 6.5 7.5 7.5 M Ω No MP near operating point Nominal RF power 87.5 100 64 kW for Q0 Stored energy 2.3 2.4 1.5 J Maximum surface 24.1 38.5 30.0 MV/m EXFEL gun field (1.7 Kilp.) (2.5 kIilp) (2.0 Kilp.) operation 690 APEX gun kV~860 kV Maximum wall power operation~750 kV 25.0 35.2 22.5 W/cm 2 density Page 11
DESY VHF gun based injector: Beam dynamics simulations Genetic optimizer setup Validating gun performance using LCLS-II injector 100 pC case Gun: Resonance frequency = 217 MHz, E cathode = 30 MV/m, phase variable Thermal emittance : 1 ~ 0.5 mm.mrad/mm Laser temporal profile: flattop 60 ps with 2 ps edges, radially Gaussian truncation at 1-sigma, both variable Injector layout 1 st solenoid position changed according to new gun geometry Other elements position stay the same as LCLS-II Solenoids: focusing variable Buncher: voltage increase from 240 kV to ~400 kV Accelerating structure: TESLA cavity Cav 1 amplitude and phase variable Injector optimization by ASTRA simulations, driven by genetic optimizer Optimize emittance and high order energy spread of 10 A solutions Page 12
Beam dynamics simulations: Preliminary results DESY VHF gun (mode 1) vs APEX gun, PITZ gun LBNL gun DESY VHF gun • Emittance - 30% reduction vs APEX gun, close to PITZ gun. - ~0.1 um emittance achieved with low thermal emittance. • H.O. energy spread - ~3 keV (1/3 of APEX gun) and close to PITZ gun. 100 pC APEX DESY VHF PITZ gun Unit 1 0.5 0.85 Thermal 1 μ m.rad/mm PITZ gun DESY VHF gun Ecath 20 30 30 60 MV/m I peak 15 10 11 4 A 100% ε (projected) 0.29 0.20 0.12 0.17 μ m.rad 95% ε (projected) 0.21 0.15 0.09 0.11 μ m.rad H.O. energy spread 9.6 2.4 2.7 3.7 keV DESY VHF gun (mode 2) vs APEX gun Longitudinal phase space with 1 st and 2 nd order chirp correction • Both transverse and longitudinal beam quality is similar Page 13
Buncher design: Goals and Constraints 1300 MHz 1300 MHz 217/162 MHz Frequency : 1.3 GHz (SCRF accelerating structure frequency) Accelerating voltage : 400 kV ( beam dynamics optimization underway) No of cells 2 or 3: Availability of space Power dissipation < 5 kW/cell (demonstrated in APEX ) Configuration : one - 2 cell / two- 2 cell / one -3 cell (beam dynamics optimization underway) Page 14
1300 MHz buncher : Literature survey Parameters\ Laboratory Cornell/ Jlab :ERL KEK:ERL LBNL:APEX Geometry No. of cells 1 1 2 𝑆 𝑡ℎ = 𝑊 2 4.2 5.33 7.8 𝑄 𝑑 Nominal Acc. Voltage (kV) 120 130 240 Power dissipation (kW) 3.42 3.17 7.4 Proposed PITZ/DESY design: KEK design (highest shunt impedance/cell ) with multiple cells Ref: (1). V. Veshcherevich and S. Belomestnykh , “ Buncher cavity for ERL”, PAC 2003; (2) T. Takahashi et al., “Development of a 1.3 GHz buncher cavity for the compact ERL”, IPAC 2014; (3). H. Qian et al., “Design of a 1.3 GHz two -cell buncher for APEX”, IPAC 2014 Page 15
1300 MHz pre-buncher: RF Design Alternative designs Two-cell re-entrant type Option 1: TESLA shape with re-entrant at end Electric field array plot Magnetic energy distribution Electric field array plot Magnetic energy distribution Higher mode separation increased to ~3 MHz Shunt impedance: ~9 M (15% higher compare to LBNL) Peak magnetic field shifted near beam pipes: easy for cooling Option2 :Geometry similar to TESLA (SCRF) cavities On- axis Electric field profile Multipacting simulations results Shunt impedance: 9.9 M (25% higher compare to LBNL) Power dissipation: 16 kW (8 kW/cell, higher compare to LBNL) Practical issues Small Mode separation (~ 1 MHz) Electric field array plot Magnetic energy distribution Higher mode separation increased to ~3 MHz Maximum heat near inter-cell coupling iris: difficult to remove Shunt impedance: 7.7 M (similar to LBNL design) Peak magnetic field shifted near beam pipes: easy for cooling Page 16
Recommend
More recommend