1. CLIC meeting, October 16-18, 2007 High Polarization DC High Voltage GaAs Photoguns P. Adderley, J. Brittian, J.Clark, J. Grames, J. Hansknecht, M.Poelker, M. Stutzman, R. Suleiman Students: A. Jayaprakash, J. McCarter, K. Surles-Law

2. CLIC meeting, October 16-18, 2007 Basic Requirements: 100 nm Superlattice GaAs: Layers of GaAs on GaAsP No strain relaxation QE ~ 0.8% Pol ~ 85% @ 780 nm chekc 14 pairs Fiber-based Laser Bake/Vent Guns Good gunGood PhotocathodeGood Laser Examples at CEBAF/Jefferson Lab – high average current CW

3. CLIC meeting, October 16-18, 2007 DC High Voltage GaAs Photoguns FacilityBeam StructureHV [kV]Avg. CurrentBunch ChargePolarization Bonn – ELSA50 Hz505 μA100 nC / 1 μs+ CEBAFcw: 1497 MHz100200 μA0.13 pC+ Mainz Microtroncw: 2450 MHz10050 μA0.02 pC+ MIT Bates600 Hz, DF=1.5%60120 μA250 nC (25 μs)+ (bulk) Nagoya/JLC200+ NIKHEV2μs pulses @ 1 Hz1000.04 μA SLAC–fixed Target0.3μs pulses @ 120 Hz120+ SLAC – SLC2ns pulses @ 120 Hz1202 μA16 nC+ JLAB FELcw: 75 MHz35010 mA135 pC- Daresbury ERLPcw: 75 MHz350- Cornellcw: 1300 MHz750100 mA77 pC- JAEA25050 mA- JLab 100kW FELcw: 750 MHz500100 mA135 pC- ILC1ms w/ 1ns pulses @ 5Hz140 - 200~100 μA5 nC+ CLIC207ns w/ 100ps pulses @ 50Hz10 μA0.6 nC+ EIC – ELIC100100 μA+ EIC - eRHICcw>10025 mA+ Historical or diminished scopeUnder constructionProposedoperating

4. CLIC meeting, October 16-18, 2007 CLIC e-Beam Time Structure 50 Hz Repetition Rate 207 ns, 311 micro-bunches ~ 100 ps 667 ps, 1497 MHz

5. CLIC meeting, October 16-18, 2007 CLIC e-Beam Source Parameters ParameterSymbolValue Number Electrons per microbunchNeNe 4 x 10 9 Number of microbunchesnbnb 311 Width of microbunchtbtb ~ 100 ps Time between microbunches tbtb 667 ps Microbunch rep ratefbfb 1497 MHz Width of macropulseTBTB 207 ns Macropulse repetition rateFBFB 50 Hz Charge per micropulseCbCb 0.64 nC Charge per macropulseCBCB 199 nC Average current from gun (C B x F B )I ave 10uA Average current in macropulse (C B / T B )IBIB 1 A Duty Factor w/in macropulse (100ps/667ps)DF0.15 Peak current of micropulse (I B / DF)I peak 6.4 A

6. CLIC meeting, October 16-18, 2007 CLIC Polarized e-Source Shared Challenges (compared to Jefferson Lab CW experience) Photocathode material – polarization > 80% Photocathode preparation techniques – high QE Machine-friendly gun design: minimize downtime Ultrahigh vacuum requirement High voltage and high field gradient: no high voltage breakdown, no field emission Cathode/anode design: manage ALL of the extracted beam Unique Challenges High bunch charge and high peak current: space charge and surface charge limit Injector design Drive laser

7. CLIC meeting, October 16-18, 2007 Shared Challenges: Recent Developments Commercial strained-superlattice photocathode –Consistent 85% polarization, ~ 1% QE CEBAF load-locked gun –Improved vacuum and accelerator-friendly ops High Voltage R&D (just beginning) –Reduce field emission –Push value of “routine” operation beyond 100kV –Reduce complexity and cost of HV insulator Cathode/Anode Design (just beginning) –Optimize geometry to support loss-free beam delivery across entire photocathode surface

8. CLIC meeting, October 16-18, 2007 Photocathode Material High QE ~ 20% Pol ~ 35% Bulk GaAs “conventional” material QE ~ 0.15% Pol ~ 75% @ 850 nm Strained GaAs: GaAs on GaAsP 100 nm Superlattice GaAs: Layers of GaAs on GaAsP No strain relaxation QE ~ 1% Pol ~ 85% @ 780 nm 100 nm 14 pairs Commercial Products

9. CLIC meeting, October 16-18, 2007 Significant FOM Improvement P I 2 2 sup. str. = 1.38 But we could not operate with long lifetime….

10. CLIC meeting, October 16-18, 2007 Commercial Superlattice Photocathodes Success required ~ 1 year of effort Cannot be hydrogen cleaned Arsenic capped (worked with vendor SVT) No solvents during preparation! Anodized edge: a critical step. Eliminates electrons that hit beampipe walls M. Baylac et al., “Effects of atomic hydrogen and deuterium exposure on high polarization GaAs photocathodes” PRST-AB 8, 123501 (2005)

11. CLIC meeting, October 16-18, 2007 CEBAF 100kV polarized electron source Two-Gun Photoinjector - One gun providing beam, one “hot” spare vent/bake guns – 4 days to replace photocathode (can’t run beam from one gun while other is baking) Activate photocathode inside gun – no HV breakdown after 7 full activations (re-bake gun after 7 th full activation) 13 mm photocathode, but use only center portion, 5 mm dia. Extract ~ 2000 Coulombs per year Beam current ~ 100uA, laser 0.5mm dia., lifetime: ~ 100C, 1x10 5 C/cm 2

12. CLIC meeting, October 16-18, 2007 What limits photogun lifetime? QE scan of photocathode QE (%) Imperfect vacuum and Ion Backbombardment photocathode anode cathode Laser IN e beam OUT Note, other factors can limit lifetime: Field emission, photocathode material, laser wavelength, laser radial position at photocathode, beam optics, gun voltage, gap size, …..

13. CLIC meeting, October 16-18, 2007 New CEBAF load-locked gun “suitcase” Loading chamber Preparation/activation chamber HV chamber

14. CLIC meeting, October 16-18, 2007 Key Features: Smaller surface area Electropolished and vacuum fired to limit outgassing NEG-coated Never vented Multiple pucks (8 hours to heat/activate new sample) Suitcase for installing new photocathodes (one day to replace all pucks) Mask to limit active area, no more anodizing

15. CLIC meeting, October 16-18, 2007 LL Gun and Test Beamline Y-scale: multiple variables 10 mA, 47C7.5 mA, 54C5 mA, 95C Time (hours) QE scan

16. CLIC meeting, October 16-18, 2007 Compare NEW and OLD load locked guns Photogun Lifetime - the best vacuum gauge OLD NEW Bulk GaAs, Green light and DC beam “Further Measurements of Photocathode Operational Lifetime at Beam Current > 1mA using an Improved 100 kV DC High Voltage GaAs Photogun,” J. Grames, et al., Proceedings Polarized Electron Source Workshop, SPIN06, Tokyo, Japan Vacuum gauges indicated same pressure in both guns, suggesting our gauges don’t work below 1.5x10 -11 Torr

17. CLIC meeting, October 16-18, 2007 1mA at High Polarization* ParameterValue Laser Rep Rate499 MHz Laser Pulselength30 ps Wavelength780 nm Laser Spot Size450 mm Current1 mA Duration8.25 hr Charge30.3 C Lifetime210 C Charge Lifetime160 kC/cm 2 Note High Initial QE Vacuum signals Laser Power Beam Current * Note: did not actually measure polarization “Lifetime Measurements of High Polarization Strained Superlattice Gallium Arsenide at Beam Current > 1 mA Using a New 100 kV Load Lock Photogun”, J. Grames et al., Particle Accelerator Conference, Albuquerque, NM, June 25-29, 2007

18. CLIC meeting, October 16-18, 2007 Increase Gun Voltage: Why? Address current density limitation due to Child’s Law Reduce space-charge-induced emittance growth, maintain smaller transverse beam profile and short bunchlength Reduce problems associated with surface charge limit (i.e., QE reduction at high laser power) Prolong Operating Lifetime? Historically, Labs have had difficulty operating DC high voltage guns above field gradient ~ 5 MV/m and bias voltage ~ 100kV. That said, it would be beneficial to build a CLIC gun with higher voltage to...

19. CLIC meeting, October 16-18, 2007 Space Charge Limit Peak current at source ~ 6 A Assume laser spot size 1cm Current density j = 7.6 A/cm 2 Space Charge Limit (Child’s Law) V (kV) j 0 (A/cm 2 ) 14014 20023 35053 At lower gun voltages, large laser spot is required. Must also consider charge limit at anode… Slide info courtesy Jym Clendenin, SLAC for 3 cm cathode/anode gap

20. CLIC meeting, October 16-18, 2007 Surface Charge Limit Peak to peak spacing 2.8ns, bunchwidth 0.7ns, Charge: 1nC/bunch Nagoya 5.5 A/cm2 measured @ SLAC for 780 nm, 75 ns pulse 9.7 A/cm2 @ Nagoya for 780 nm, 30 ps Heavily doped surface: viable solution? CLIC current density comparable to these values…something to worry about QE reduction at high laser power Slide info courtesy Takashi Maruyama, SLAC

21. CLIC meeting, October 16-18, 2007 Improve Lifetime with Larger Laser Spot? (Best Solution – Improve Vacuum, but not easy) Bigger laser spot, same # electrons, same # ions Ionized residual gas strikes photocathode Ion damage distributed over larger area

22. CLIC meeting, October 16-18, 2007 Lifetime with Large/Small Laser Spots 1500 350 2 ≈ 18 Expectation: “Further Measurements of Photocathode Operational Lifetime at Beam Current > 1mA using an Improved 100 kV DC High Voltage GaAs Photogun,” J. Grames, et al., Proceedings Polarized Electron Source Workshop, SPIN06, Tokyo, Japan Tough to measure large Coulomb lifetimes with only 100-200 C runs! Factor of 5 to 10 improvement with larger laser spot size

23. CLIC meeting, October 16-18, 2007 Improve Lifetime? Hypothesis: Double the gun voltage, halve the # of “bad” ions, improve lifetime by 2 Ionization cross section for H 2 100kV 250kV Most ions created at low energy, < 10kV Low energy ion column for 100kV gun Low energy ion column for 200kV gun Ion energy

24. CLIC meeting, October 16-18, 2007 Must Eliminate Field Emission Cornell Technique, high pressure rinsing (B. Dunham ERL07) Recent tests at JLab with shaped electrodes support Cornell results Work of M. Chetsova, K. Surles-Law CEBAF gun

25. CLIC meeting, October 16-18, 2007 Cathode/Anode Design We learned at CEBAF that it is extremely important to manage ALL of the extracted beam –Anodized edge: beam from outside 5 mm active area can hit beampipe walls, degrade vacuum, reduce operating lifetime CLIC requires large laser beam to reduce current density and overcome space and surface charge limit Suggest detailed modeling of cathode/anode optic and first few meters of beamline –Perhaps using multivariate optimization?

26. CLIC meeting, October 16-18, 2007 Goals of Cathode/Anode Design Create cathode/anode optic with small aberration across large photocathode active area, with very little beam loss. What to optimize? –Size of cathode electrode diameter, size of photocathode active area –Size of laser beam: lowest possible current density but with adequate emittance –Cathode/anode shape for adequate focusing –Cathode voltage/gradient: higher voltage to reduce space charge and provide possibility of extracting higher peak current with more narrow laser pulsewidth, to reduce SHB requirements

27. CLIC meeting, October 16-18, 2007 Conclusions: CLIC Polarized Source CLIC gun to incorporate state-of-the art vacuum features (small volume, low outgassing rate, NEG pumps and coating) plus reliable load-lock design for quick photocathode replacement Push gun voltage > 120kV to reduce space charge and surface charge limitations Cathode/anode design: managing ALL of the beam

28. CLIC meeting, October 16-18, 2007 Improving Gun Vacuum Ultimate Pressure = Outgassing Rate x Surface Area Pump Speed test chambers CEBAF guns New LL gun How to explain this discrepancy?  Outgassing rate higher than assumed “standard” value; 1x10 -12 Torr·L/s·cm 2 ?  NEG pump speed smaller than SAES says? Measured pressure always much greater than predicted

29. CLIC meeting, October 16-18, 2007 Outgassing Rate “Characterization of the CEBAF 100 kV DC GaAs Photoelectron Gun Vacuum System,” M.L. Stutzman, et al., Nucl. Instrum. Meth. A, 574 (2007) p. 213-220 Preprocessing In situ bake parameters Outgassing Rate (Torr·L/s·cm 2 ) Chambert(h)T(°C)EP Surface roughness t(h)T(°C) Orifice Method Rate of Rise Method Old 304no3.7 μm4002509.7x10 -13 1x10 -12 New 304no3.7 μm1802501.9x10 -12 2.5x10 -12 EP 3044900yes2.1 μm 30 then 90 150 250 8.9x10 -13 Orifice and Rate of Rise Methods Studied 304, 316L and 6061 Al Degreasing and solvent cleaning vs EP and vacuum firing

30. CLIC meeting, October 16-18, 2007 Benefit of EP and Vacuum Firing Electropolishing and vacuum firing provides low rate with fewer bakes Extremely low values (e.g., 10 -14 to 10 -15 ) reported in literature elude us Conclusion: We have the “industry-standard” outgassing rate ~ 1x10 -12 Torr·L/s·cm 2

31. CLIC meeting, October 16-18, 2007 Recent High Temperature Bake of JLab FEL Gun 316 LN Stainless Steel, Baked at 400°C for 10 days Vacuum inside, hot air outside, Strip heaters instead of hot air guns Outgassing rate: 1.49x10 -13 Torr L/s cm 2 Lessons for CEBAF/ILC? We should have vacuum fired our end flanges Welding introduces hydrogen 250C for 30 hours not adequate

32. CLIC meeting, October 16-18, 2007 NEG Pump Speed Full NEG activation better than passive activation via bake NEG pump speed very good, at least at high pressure Conclusion: Can’t explain reduced pump speed at low pressure – a real effect? More likely an indication of gauge limitations

33. CLIC meeting, October 16-18, 2007 NEG Coating NEG coating turns a gas source into pump ~0.02 L/s·cm 2 : Modest pump speed can be improved

34. CLIC meeting, October 16-18, 2007 ILC e- Beam Time Structure 5 Hz Repetition Rate 1 ms, 2820 micro-bunches 2 ns 337 ns 3 MHz

35. CLIC meeting, October 16-18, 2007 ILC e-Beam Source Parameters ParameterSymbolValue Number Electrons per microbunchNeNe 3 x 10 10 Number of microbunchesnbnb 3000 Width of microbunchtbtb ~ 1 ns Time between microbunches tbtb 337 ns Microbunch rep ratefbfb 3 MHz Width of macropulseTBTB 1 ms Macropulse repetition rateFBFB 5 Hz Charge per micropulseCbCb 4.8 nC Charge per macropulseCBCB 14420 nC Average current from gun (C B x F B )I ave 72 uA Average current macropulse (C B / T B )IBIB 14.4 mA Duty Factor w/in macropulse (1ns/337ns)DF3x10 -3 Peak current of micropulse (I B / DF)I peak 4.8 A

36. CLIC meeting, October 16-18, 2007 Fiber-Based Drive Laser Gain-switched seed ErYb-doped fiber amplifier Frequency-doubler 1560nm 780nm

37. CLIC meeting, October 16-18, 2007 Fiber-Based Drive Laser e-bunch ~ 30 ps autocorrelator trace  CEBAF’s last laser!  Gain-switching better than modelocking; no phase lock problems, no feedback  Very high power  Telecom industry spurs growth, ensures availability  Useful because of superlattice photocathode (requires 780nm) Ti-sapp