High Intensity Polarized Electron Sources 07182006 Evgeni Tsentalovich

High Intensity Polarized Electron Sources 07/18/2006 Evgeni Tsentalovich MIT

Progress over past two decades 15 years ago • Unreliable guns at development stage • Dreams to exceed 40% polarization 07/18/2006 Now • Routinely operated productive quality guns (SLAC, JLAB, Mainz, Bates…) • Strained, superlattice crystals with polarization approaching 90% • New photocathode materials • New gun concepts

New requirements New generation of accelerators (e. RHIC, ILC) demand polarized injectors with extreme parameters • Very high current • Very high polarization • Low emittance Another application: Energy Recovery Linac (ERL) 07/18/2006 • Very high current • No polarization • Very low emittance

Ga. As photocathodes Requirements: high QE and polarization • Remains the only material for polarized electron guns • Very high QE • Very high polarization • But ! Very demanding technology ( Ultra-high vacuum requirements) 07/18/2006

Semiconductor band structure E Band gap Conducting band Valence band - low Doping (Z, Be) is used to control the concentration of carriers: - medium - high 07/18/2006

Band structure of Ga. As E Conducting band -1/2 1 1/2 3 1. 6 e. V -3/2 0. 3 e. V k 07/18/2006 -1/2 1/2 3/2

Strained crystal -1/2 E 1 1/2 3 1. 6 e. V -3/2 -1/2 0. 3 e. V k 07/18/2006 -1/2 1/2 3/2

Ga. As-based photocathodes Strained Ga. As: Ga. As on Ga. As. P High QE ~ 1 -10% Pol ~ 35 -45% 07/18/2006 100 nm 14 pairs 100 nm Bulk Ga. As Superlattice Ga. As: Layers of Ga. As on Ga. As. P QE ~ 0. 15% Pol ~ 75% QE ~ 0. 8% Pol ~ 85%

Negative electron affinity Most (but not all!) electrons reaching the surface are thermolized E Conductive band Vacuum level Band gap (forbidden zone) Cs, O(F) deposition Valence band surface x 07/18/2006

Photocathodes degradation Poisoning by residual gases • Oxygen- and carbon-containing species are more harmful • Hydrogen and noble gases are more tolerable • This degradation can be healed by heat-cleaning at moderate temperatures (<550 C) 07/18/2006 Ion bombardment • Most harmful • Only high-temperature (~600 C) heat cleaning restores QE, and only partially • Effect is proportional to pressure in the chamber and to average current

Charge saturation E Vacuum level surface x 07/18/2006

Charge saturation (SLAC data) High doping →low polarization ! 07/18/2006

High gradient doping High ( )doped layer ~ 5 nm • Works very well Superlattice • The high-doped layer is thin enough to preserve high polarization Buffer • Charge saturation is highly suppressed (at least for fresh crystals) Substrate • The top layer can survive only few high-temperature (~600 C) activations • Might be problematic for high-current guns 07/18/2006

DC gun design Cylindrical symmetry Cathode Anode r' r Normalized emittance 07/18/2006 Emittance: doesn’t change with acceleration

DC gun design Infinitely small beam spot, no space charge, no nonlinear transverse forces r' r Cathode 07/18/2006 Emittance:

DC gun design Finite beam spot, no space charge, no nonlinear transverse forces r' r Emittance: Cathode With perfectly linear transverse forces only thermal emittance remains 07/18/2006

Neglecting thermal emittance r'r' r Emittance: Cathode Nonlinearity in the gun optics may introduce the emittance growth. 07/18/2006

Space charge Cathode Anode J Space charge may change the beam profile and increase the beam emittance J r J r r Emittance growth might be suppressed by shaping the laser profile 07/18/2006

Space charge • Space charge effects are strongest when electrons have low energy (no space charge effects for relativistic beam) • Accelerate as fast as possible – high gradient in the gun • Accelerate as high as possible – high gun voltage, to reduce space charge effects between the gun and the accelerator 07/18/2006

Space charge Worst case scenario: large emitting spot AND Very strong Strong Weak high current density Child’s law: - microperveance; d – distance between cathode and anode Space charge influence: Space charge effects could be reduced by • Increasing gun voltage • Reducing cathode – anode gap 07/18/2006 • Increasing the emitting spot Limited (breakdowns) Non-linear transverse forces

Emittance: • • • Thermal Ga. As cathode (room temperature) ~0. 2 mm·mrad ·R(mm) Thermal Cu, Cs 2 Te cathodes ~1. 2 mm·mrad ·R(mm) Real gun with small emitting spot (JLAB) ~ 5 mm·mrad Real gun with large emitting spot (Bates) ~15 mm·mrad Beam after RF chopping/bunching ~ 20 -100 mm·mrad Estimations for RF (SRF) gun ~ 1 -5 mm·mrad • ILC requirements 07/18/2006 ~. 05 mm·mrad

Polarized electron guns: DC Approved technology (at least for ~ 100 k. V) Require RF chopping/bunching RF bunching could be avoided with appropriate laser system RF No working Ga. As-based RF gun yet Beam from the gun is bunched High acceleration rate, high electron energy from the gun Low energy beam (space charge! ) Better suited for large emitting spot BEST FOR CONVENTIONAL APPLICATIONS OR WHEN VERY HIGH CURRENT IS NEEDED 07/18/2006 BEST FOR APPLICATIONS WITH VERY HIGH BRIGHTNESS AND LOW EMITTANCE

DC Guns: Mainz V = 100 k. V Active spot. 25 mm 07/18/2006

DC Guns: JLAB V = 100 k. V Active spot 0. 2 mm 07/18/2006

DC Guns: Bates V = 60 k. V Active spot 12 mm 07/18/2006

DC Guns: SLAC V = 120 k. V Active spot 15 mm 07/18/2006

DC Guns: Nagoya V = 200 k. V Active spot 18 mm 07/18/2006

DC Guns: Cornell V = 500 k. V (800 ? ) 07/18/2006

RF guns • The only practical experience: BINP (Novosibirsk) • Good vacuum conditions with RF on and unactivated Ga. As crystal installed • Activated Ga. As crystal survived just several RF cycles • Severe back-bombardment resulted in a very short life time 07/18/2006

RF guns (SLAC) 1. 6 cell pill box Higher Order Mode (HOM) single cell • More open structure • No internal irises • More effective vacuum pumping 07/18/2006

RF guns (BNL & AES) 07/18/2006

RF guns: Warm Significant practical experience Unclear if Ga. As-based cathode will survive RF gun conditions New, more robust cathode materials may appear (Ga. N) Much easier to do 07/18/2006 SRF Very expensive and untested technology Best vacuum possible Wide open apertures (eliminates back bombardment) Better chances of success

Laser development Fiber lasers: • Very short pulses • Mode – locked, but rep. rate limited to MHz • Wavelength 1030 – 1500 nm, but could be frequencydoubled • Reliable • Relatively expensive 07/18/2006

Laser development Elliptical beams (SLAC) • Suppression of non-linear space charge effects • Maximizing brightness • Might be very useful for RF guns • Very challenging task 07/18/2006

ILC gun • DC or RF gun could be used • ILC emittance requirements are so high that even RF gun is unlikely to meet them without dumping ring • Although dumping ring is still required for RF gun, it might be of much simpler design, saving millions • Conclusion: RF gun would be a better option, but it requires significant R&D and the success is not guaranteed 07/18/2006

e. RHIC gun (ring-ring) • Modest intensity and emittance requirements • Regular DC gun is well suited for the task • Two options: mode-locked laser or RF chopper/buncher Polarized electron gun for ring e. RHIC version is based on proven and Mode-locked laser: technology RF chopper/buncher: doesn’t require any significant • Simplifies injector • Complicates injector R&D • No emittance growth in chopper • Emittance growth in chopper • Beam compression reduces peak current demand from the gun 07/18/2006

e. RHIC gun (linac-ring) Extremely high current demand !!! I(average) ~ 500 m. A I(peak) ~ 200 A High polarization → strained Ga. As → QE ~ 0. 1% Average laser power ~ 800 W Such lasers do not exist. Possible solutions: a) array of diode lasers b) dedicated FEL – almost unlimited laser power, tunable 07/18/2006

Problems without known solution Heat load (800 W on the cathode) HEAT t=1 mm Ga. As ACTIVE COOLING With a conventional cathode stalk system, the cathode would heat up to stellar temperatures, but, fortunately, melt first. New problem: dynamic cooling (gun off !) 07/18/2006

Problems without known solution Peak current (~200 A) For DC gun : Larger cathodes? Ring-like cathodes ? Emitting spot : What about emittance ? ? ? 07/18/2006

Can we relax the requirements? • With I(average) ~ 40 -50 m. A the luminosity is the same as in ring-ring version • 40 -50 m. A gun is still a very difficult task, but it is a LOT easier than 500 m. A • Heat load and perveance problems go away • Life time of the cathode is still a major problem 07/18/2006