Precision Electron Beam Polarimetry Wolfgang Lorenzon (Michigan) SPIN 2008 Symposium 6-October 2008

How to measure polarization of e  /e  beams? Three different targets used currently: 1. e  - nucleus: Mott scattering 30 – 300 keV (5 MeV: JLab) spin-orbit coupling of electron spin with (large Z) target nucleus 2. e  - electrons: Møller (Bhabha) scat. MeV – GeV atomic electron in Fe (or Fe-alloy) polarized by external magnetic field 3. e  - photons: Compton scattering > 1 GeV laser photons scatter off lepton beam Goal: measure  P/P ≈ 1% (Hall C, EIC)  P/P ≈ 0.25% (0.1%) (ILC) realistic? 2

Electron Polarimetry Many polarimeters are, have been in use, or a planned: Compton Polarimeters: ILC 45.6 – 500 GeV EIC 3 – 20 GeV LEP mainly used as machine tool for resonant depolarization SLAC SLD 46 GeV DESY HERA, storage ring 27.5 GeV (three polarimeters) JLab Hall A < 8 GeV / Hall C < 12 GeV Bates South Hall Ring < 1 GeV Nikhef AmPS, storage ring < 1 GeV Møller / Bhabha Polarimeters: Bates linear accelerator < 1 GeV Mainz Mainz Microtron MAMI < 1 GeV Jlab Hall A, B, C 3

LaboratoryPolarimeterRelative precisionDominant systematic uncertainty JLab5 MeV Mott~1%Sherman function Hall A Møller~2-3%target polarization Hall B Møller1.6% (?) 2-3% (realistic ?) target polarization, Levchuk effect Hall C Møller1.3% (best quoted) 0.5% (possible ?) target polarization, Levchuk effect, high current extrapolation Hall A Compton1% > 3 GeV)detector acceptance + response HERALPol Compton1.6%analyzing power TPol Compton3.1%focus correction + analyzing power Cavity LPol Compton?still unknown MIT-BatesMott~3%Sherman function + detector response Transmission>4%analyzing power Compton~4%analyzing power SLACCompton0.5%analyzing power Polarimeter Roundup

The “Spin Dance” Experiment (2000) Source Strained GaAs photocathode (  = 850 nm, P b >75 %) Accelerator 5.7 GeV, 5 pass recirculation Wien filter in injector was varied from -110 o to 110 o to vary degree of longitudinal polarization in each hall → precise cross-comparison of JLab polarimeters 5 Phys. Rev. ST Accel. Beams 7, (2004)

“Spin Dance” 2000 Data P meas cos(  Wien +  )

Polarization Results Results shown include statistical errors only → some amplification to account for non-sinusoidal behavior Statistically significant disagreement Systematics shown: Mott Møller C 1% Compton Møller B 1.6% Møller A 3% Even including systematic errors, discrepancy still significant

Polarization Results- Reduced Data Set Hall A, B Møllers sensitive to transverse components of beam polarization Normally – these components eliminated via measurements with foil tilt reversed, but some systematic effects may remain closed circles = full data set open circles = reduced data set Agreement improves, but still statistically significant deviations  when systematics included, discrepancy less significant

Lessons Learned Providing/proving precision at 1% level challenging Including polarization diagnostics and monitoring in beam lattice design is crucial Measure polarization at (or close to) IP Measure beam polarization continuously – protects against drifts or systematic current-dependence to polarization Flip electron and laser polarization – fast enough to protect against drifts Multiple devices/techniques to measure polarization – cross-comparisons of individual polarimeters are crucial for testing systematics of each device – at least one polarimeter needs to measure absolute polarization, others might do relative measurements (fast and precise) – absolute measurement does not have to be fast New ideas? 9

532 nm HERA (27.5 GeV) EIC (10 GeV) Jlab HERA EIC -7/9 Compton edge: Compton vs Møller Polarimetry Detect  at 0 °, e - < E e Strong  need <<1 at E e < 20 GeV P laser ~ 100% non-invasive measurement syst. Error: 3 → 50 GeV ( ~ 1 → 0.5%) hard at < 1 GeV: (Jlab project: ~ 0.8%) rad. corr. to Born < 0.1% Detect e - at  CM ~ 90 °  good systematics beam energy independent ferromagnetic target P T ~ 8% beam heating (I e < 2-4  A), Levchuck eff. invasive measurement syst. error 2-3% typically 0.5% (1%?) at high magn. field rad. corr. to Born < 0.3% ILC

Polarized atomic hydrogen in a cold magnetic trap use ultra-cold traps (at 300 mK: P e ~ , density ~ 3 ∙ cm -3, stat. 1% in 10 min at 100  A) expected depolarization for 100  A CEBAF < limitations: beam heating → “continuous” beam & complexity of target advantages: expected accuracy < 0.5% & non-invasive, continuous, the same beam Problem: very unlikely to work for high beam currents for EIC (due to gas and cell heating) Jet Target: avoids these problems – VEPP mA, transverse – stat 20% in 8 minutes (5 ∙ e - /cm 2, 100% polarization) – What is electron polarization in a jet? E. Chudakov et al., IEEE Trans. Nucl. Sci. 51, 1533 (2004) Jet Target: need to address these problems –Breit-Rabi measurement analyzes only part of jet → uniformity of jet has to be understood –large background from ions in the beam: most of them associated with jet (hard to measure) –origin of background observed in Novosibirsk still unclear? –clarification of depolarization by beam RF needed → might be considerable

New Fiber Laser Technology (Hall C) ps pulses at 499 MHz - external to beamline vacuum → easy access - huge lumi boost when phase locked - excellent stability, low maintenance Electron BeamLaserBeam Jeff Martin (Winnipeg) Gain switched

Compact, Off-The Shelf, Rack Mountable… 13 RF locked low-power 1560 nm fiber diode ErYb-doped fiber amplifier Frequency doubler

Fiber Laser for Hall C Compton 14 Seed laser at 1064 nm Fiber amplifier (50 W output at 1064 nm) Frequency doubling cavity Result: 25 W, 532 nm, 30 ps pulses at 499 MHz JLab Polarized source group is building laser (J. Grames) problems with amplification

Dominant Challenge: determine A z 15 Best tool to measure e - polarization → Compton e - (integrating mode) Challenge accurate knowledge of ∫Bdl must calibrate the electron detector fit the asymmetry shape or use Compton Edge

Electron Polarimetry 9/14/200716W. Lorenzon PSTP 2007 Kent Paschke

Future Efforts 17 EIC – e-p and e-ion collisions at c.m. energies: GeV – 10 GeV (~ GeV) electrons/positrons – Longitudinal polarization at IP: ~ 70% or better – Needed accuracy:  P/P = 1% – Bunch separation: ns – Luminosity: L(ep) ~ cm -2 s -1 per IP ILC – e - - e + collisions at c.m. energies: GeV – Longitudinal polarization at (IP) – P(e - ) > 80% – P(e + ) > 50% – Needed accuracy:  P/P = 0.25% (0.1%) → new territory

EIC Compton Polarimeter 18 No serious obstacles are foreseen to achieve 1% precision for electron beam polarimetry at the EIC (3-20 GeV) JLAB at 12 GeV will be a natural testbed for future EIC e - /e + Polarimeter tests – evaluate new ideas/technologies for the EIC There are issues that need attention (crossing frequency 3-35 ns; beam-beam effects at high currents; crab crossing effect on polarization)

ILC Polarimeters 19 Three ways to measure polarization at the ILC upstream Compton polarimeter downstream Compton polarimeter (e + e - →) W + W - production (has large cross section & very sensitive to polarization) Complication polarization at IP = lumi-weighted polarization ≠ polarization at polarimeter depolarization and spin transport effects estimated at 0.1%-0.4% levels! same level as required accuracy keep errors in these effects small Need upstream, downstream & e + e - physics measurements determine best values for each polarimeter separately (hide from each other) compare and see whether they agree final calibration with e + e - → W + W - to e + e - IP: 1.8 km upstream polarimeter downstream polarimeter ~ 150 m behind IP Goal:  P/P ≈ 0.25% (0.1%)

Summary 20 Electron beam polarization can be measured with high precision – no serious obstacles are foreseen to achieve 1% – imperative to include polarimetry in beam lattice design – use multiple devices/techniques to control systematics Big challenge to reach  P/P = 0.25% – no fundamental show stoppers (but high beam energy needed)  P/P = 0.1% enters new territory – goes beyond traditional polarimetry – use physics processes directly (W + W - production) Build on experience, but be open to new ideas – maybe use e-p elastic scattering at low energies (measures P e P T ) – … Extensive modeling necessary – A Z, depolarization, BMT, etc. – much work done, much work still ahead